diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,165508 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 5910,
+  "global_step": 23638,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.2304763516371946e-05,
+      "grad_norm": 0.8915818929672241,
+      "learning_rate": 0.001,
+      "loss": 2.2708,
+      "step": 1
+    },
+    {
+      "epoch": 8.460952703274389e-05,
+      "grad_norm": 0.4329769015312195,
+      "learning_rate": 0.001,
+      "loss": 2.483,
+      "step": 2
+    },
+    {
+      "epoch": 0.00012691429054911582,
+      "grad_norm": 0.45692744851112366,
+      "learning_rate": 0.001,
+      "loss": 3.5579,
+      "step": 3
+    },
+    {
+      "epoch": 0.00016921905406548778,
+      "grad_norm": 0.40001946687698364,
+      "learning_rate": 0.001,
+      "loss": 2.4916,
+      "step": 4
+    },
+    {
+      "epoch": 0.00021152381758185972,
+      "grad_norm": 0.530651867389679,
+      "learning_rate": 0.001,
+      "loss": 1.7408,
+      "step": 5
+    },
+    {
+      "epoch": 0.00025382858109823165,
+      "grad_norm": 1.0012280941009521,
+      "learning_rate": 0.001,
+      "loss": 2.3514,
+      "step": 6
+    },
+    {
+      "epoch": 0.0002961333446146036,
+      "grad_norm": 0.3786524534225464,
+      "learning_rate": 0.001,
+      "loss": 2.6244,
+      "step": 7
+    },
+    {
+      "epoch": 0.00033843810813097557,
+      "grad_norm": 0.7540691494941711,
+      "learning_rate": 0.001,
+      "loss": 2.4326,
+      "step": 8
+    },
+    {
+      "epoch": 0.0003807428716473475,
+      "grad_norm": 0.44959110021591187,
+      "learning_rate": 0.001,
+      "loss": 2.1347,
+      "step": 9
+    },
+    {
+      "epoch": 0.00042304763516371943,
+      "grad_norm": 0.3813895881175995,
+      "learning_rate": 0.001,
+      "loss": 2.6791,
+      "step": 10
+    },
+    {
+      "epoch": 0.00046535239868009137,
+      "grad_norm": 0.42967715859413147,
+      "learning_rate": 0.001,
+      "loss": 1.9963,
+      "step": 11
+    },
+    {
+      "epoch": 0.0005076571621964633,
+      "grad_norm": 0.8154281377792358,
+      "learning_rate": 0.001,
+      "loss": 3.3512,
+      "step": 12
+    },
+    {
+      "epoch": 0.0005499619257128352,
+      "grad_norm": 0.4104110896587372,
+      "learning_rate": 0.001,
+      "loss": 1.9401,
+      "step": 13
+    },
+    {
+      "epoch": 0.0005922666892292072,
+      "grad_norm": 0.536148190498352,
+      "learning_rate": 0.001,
+      "loss": 2.0377,
+      "step": 14
+    },
+    {
+      "epoch": 0.0006345714527455792,
+      "grad_norm": 0.3982076644897461,
+      "learning_rate": 0.001,
+      "loss": 2.6555,
+      "step": 15
+    },
+    {
+      "epoch": 0.0006768762162619511,
+      "grad_norm": 0.41156458854675293,
+      "learning_rate": 0.001,
+      "loss": 2.0036,
+      "step": 16
+    },
+    {
+      "epoch": 0.0007191809797783231,
+      "grad_norm": 0.5118016004562378,
+      "learning_rate": 0.001,
+      "loss": 2.4408,
+      "step": 17
+    },
+    {
+      "epoch": 0.000761485743294695,
+      "grad_norm": 0.44030845165252686,
+      "learning_rate": 0.001,
+      "loss": 3.1642,
+      "step": 18
+    },
+    {
+      "epoch": 0.0008037905068110669,
+      "grad_norm": 0.3349159061908722,
+      "learning_rate": 0.001,
+      "loss": 1.947,
+      "step": 19
+    },
+    {
+      "epoch": 0.0008460952703274389,
+      "grad_norm": 0.44204267859458923,
+      "learning_rate": 0.001,
+      "loss": 2.3666,
+      "step": 20
+    },
+    {
+      "epoch": 0.0008884000338438108,
+      "grad_norm": 0.4936717748641968,
+      "learning_rate": 0.001,
+      "loss": 3.2801,
+      "step": 21
+    },
+    {
+      "epoch": 0.0009307047973601827,
+      "grad_norm": 0.48639440536499023,
+      "learning_rate": 0.001,
+      "loss": 2.9606,
+      "step": 22
+    },
+    {
+      "epoch": 0.0009730095608765547,
+      "grad_norm": 0.3897702991962433,
+      "learning_rate": 0.001,
+      "loss": 2.0694,
+      "step": 23
+    },
+    {
+      "epoch": 0.0010153143243929266,
+      "grad_norm": 0.47153130173683167,
+      "learning_rate": 0.001,
+      "loss": 1.8757,
+      "step": 24
+    },
+    {
+      "epoch": 0.0010576190879092986,
+      "grad_norm": 0.7353273034095764,
+      "learning_rate": 0.001,
+      "loss": 3.3964,
+      "step": 25
+    },
+    {
+      "epoch": 0.0010999238514256705,
+      "grad_norm": 0.4378684461116791,
+      "learning_rate": 0.001,
+      "loss": 3.3866,
+      "step": 26
+    },
+    {
+      "epoch": 0.0011422286149420425,
+      "grad_norm": 0.5399907231330872,
+      "learning_rate": 0.001,
+      "loss": 3.5229,
+      "step": 27
+    },
+    {
+      "epoch": 0.0011845333784584143,
+      "grad_norm": 0.4380228519439697,
+      "learning_rate": 0.001,
+      "loss": 3.3026,
+      "step": 28
+    },
+    {
+      "epoch": 0.0012268381419747864,
+      "grad_norm": 0.5237672924995422,
+      "learning_rate": 0.001,
+      "loss": 2.7964,
+      "step": 29
+    },
+    {
+      "epoch": 0.0012691429054911584,
+      "grad_norm": 0.5567352771759033,
+      "learning_rate": 0.001,
+      "loss": 2.2648,
+      "step": 30
+    },
+    {
+      "epoch": 0.0013114476690075302,
+      "grad_norm": 0.4474937915802002,
+      "learning_rate": 0.001,
+      "loss": 2.6151,
+      "step": 31
+    },
+    {
+      "epoch": 0.0013537524325239023,
+      "grad_norm": 0.5182416439056396,
+      "learning_rate": 0.001,
+      "loss": 3.4561,
+      "step": 32
+    },
+    {
+      "epoch": 0.001396057196040274,
+      "grad_norm": 0.4204736649990082,
+      "learning_rate": 0.001,
+      "loss": 3.0416,
+      "step": 33
+    },
+    {
+      "epoch": 0.0014383619595566461,
+      "grad_norm": 0.45871949195861816,
+      "learning_rate": 0.001,
+      "loss": 2.2159,
+      "step": 34
+    },
+    {
+      "epoch": 0.001480666723073018,
+      "grad_norm": 0.4581148028373718,
+      "learning_rate": 0.001,
+      "loss": 2.0835,
+      "step": 35
+    },
+    {
+      "epoch": 0.00152297148658939,
+      "grad_norm": 0.41536086797714233,
+      "learning_rate": 0.001,
+      "loss": 2.5392,
+      "step": 36
+    },
+    {
+      "epoch": 0.0015652762501057618,
+      "grad_norm": 0.48674535751342773,
+      "learning_rate": 0.001,
+      "loss": 2.6758,
+      "step": 37
+    },
+    {
+      "epoch": 0.0016075810136221339,
+      "grad_norm": 0.8809964656829834,
+      "learning_rate": 0.001,
+      "loss": 2.8721,
+      "step": 38
+    },
+    {
+      "epoch": 0.0016498857771385057,
+      "grad_norm": 0.6416422724723816,
+      "learning_rate": 0.001,
+      "loss": 2.2551,
+      "step": 39
+    },
+    {
+      "epoch": 0.0016921905406548777,
+      "grad_norm": 0.8650534152984619,
+      "learning_rate": 0.001,
+      "loss": 2.4497,
+      "step": 40
+    },
+    {
+      "epoch": 0.0017344953041712498,
+      "grad_norm": 0.42286792397499084,
+      "learning_rate": 0.001,
+      "loss": 2.813,
+      "step": 41
+    },
+    {
+      "epoch": 0.0017768000676876216,
+      "grad_norm": 0.7012924551963806,
+      "learning_rate": 0.001,
+      "loss": 3.2002,
+      "step": 42
+    },
+    {
+      "epoch": 0.0018191048312039936,
+      "grad_norm": 0.6307465434074402,
+      "learning_rate": 0.001,
+      "loss": 2.1075,
+      "step": 43
+    },
+    {
+      "epoch": 0.0018614095947203655,
+      "grad_norm": 0.3289887309074402,
+      "learning_rate": 0.001,
+      "loss": 1.9497,
+      "step": 44
+    },
+    {
+      "epoch": 0.0019037143582367375,
+      "grad_norm": 0.5121589303016663,
+      "learning_rate": 0.001,
+      "loss": 2.9249,
+      "step": 45
+    },
+    {
+      "epoch": 0.0019460191217531093,
+      "grad_norm": 0.545623242855072,
+      "learning_rate": 0.001,
+      "loss": 2.364,
+      "step": 46
+    },
+    {
+      "epoch": 0.001988323885269481,
+      "grad_norm": 0.427653431892395,
+      "learning_rate": 0.001,
+      "loss": 2.1296,
+      "step": 47
+    },
+    {
+      "epoch": 0.002030628648785853,
+      "grad_norm": 0.4887496531009674,
+      "learning_rate": 0.001,
+      "loss": 2.9931,
+      "step": 48
+    },
+    {
+      "epoch": 0.0020729334123022252,
+      "grad_norm": 0.5210931897163391,
+      "learning_rate": 0.001,
+      "loss": 2.232,
+      "step": 49
+    },
+    {
+      "epoch": 0.0021152381758185973,
+      "grad_norm": 0.4477267861366272,
+      "learning_rate": 0.001,
+      "loss": 2.4165,
+      "step": 50
+    },
+    {
+      "epoch": 0.0021575429393349693,
+      "grad_norm": 1.6815810203552246,
+      "learning_rate": 0.001,
+      "loss": 2.0357,
+      "step": 51
+    },
+    {
+      "epoch": 0.002199847702851341,
+      "grad_norm": 0.3510444164276123,
+      "learning_rate": 0.001,
+      "loss": 2.9298,
+      "step": 52
+    },
+    {
+      "epoch": 0.002242152466367713,
+      "grad_norm": 0.42857974767684937,
+      "learning_rate": 0.001,
+      "loss": 2.8284,
+      "step": 53
+    },
+    {
+      "epoch": 0.002284457229884085,
+      "grad_norm": 0.6068593263626099,
+      "learning_rate": 0.001,
+      "loss": 3.6971,
+      "step": 54
+    },
+    {
+      "epoch": 0.002326761993400457,
+      "grad_norm": 0.46816545724868774,
+      "learning_rate": 0.001,
+      "loss": 2.0175,
+      "step": 55
+    },
+    {
+      "epoch": 0.0023690667569168287,
+      "grad_norm": 0.44248470664024353,
+      "learning_rate": 0.001,
+      "loss": 2.1154,
+      "step": 56
+    },
+    {
+      "epoch": 0.0024113715204332007,
+      "grad_norm": 0.4355832636356354,
+      "learning_rate": 0.001,
+      "loss": 2.5438,
+      "step": 57
+    },
+    {
+      "epoch": 0.0024536762839495727,
+      "grad_norm": 0.39491739869117737,
+      "learning_rate": 0.001,
+      "loss": 2.5997,
+      "step": 58
+    },
+    {
+      "epoch": 0.002495981047465945,
+      "grad_norm": 23.964534759521484,
+      "learning_rate": 0.001,
+      "loss": 2.3666,
+      "step": 59
+    },
+    {
+      "epoch": 0.002538285810982317,
+      "grad_norm": 0.5575146079063416,
+      "learning_rate": 0.001,
+      "loss": 2.6327,
+      "step": 60
+    },
+    {
+      "epoch": 0.0025805905744986884,
+      "grad_norm": 0.44521743059158325,
+      "learning_rate": 0.001,
+      "loss": 3.0179,
+      "step": 61
+    },
+    {
+      "epoch": 0.0026228953380150605,
+      "grad_norm": 0.891205370426178,
+      "learning_rate": 0.001,
+      "loss": 1.837,
+      "step": 62
+    },
+    {
+      "epoch": 0.0026652001015314325,
+      "grad_norm": 0.33089524507522583,
+      "learning_rate": 0.001,
+      "loss": 2.9327,
+      "step": 63
+    },
+    {
+      "epoch": 0.0027075048650478046,
+      "grad_norm": 0.3559476435184479,
+      "learning_rate": 0.001,
+      "loss": 2.6628,
+      "step": 64
+    },
+    {
+      "epoch": 0.002749809628564176,
+      "grad_norm": 0.4208162724971771,
+      "learning_rate": 0.001,
+      "loss": 2.1463,
+      "step": 65
+    },
+    {
+      "epoch": 0.002792114392080548,
+      "grad_norm": 0.48232489824295044,
+      "learning_rate": 0.001,
+      "loss": 2.1562,
+      "step": 66
+    },
+    {
+      "epoch": 0.0028344191555969202,
+      "grad_norm": 0.44821107387542725,
+      "learning_rate": 0.001,
+      "loss": 2.3234,
+      "step": 67
+    },
+    {
+      "epoch": 0.0028767239191132923,
+      "grad_norm": 0.42845988273620605,
+      "learning_rate": 0.001,
+      "loss": 2.1317,
+      "step": 68
+    },
+    {
+      "epoch": 0.002919028682629664,
+      "grad_norm": 0.6322043538093567,
+      "learning_rate": 0.001,
+      "loss": 1.8463,
+      "step": 69
+    },
+    {
+      "epoch": 0.002961333446146036,
+      "grad_norm": 0.4234910011291504,
+      "learning_rate": 0.001,
+      "loss": 2.8342,
+      "step": 70
+    },
+    {
+      "epoch": 0.003003638209662408,
+      "grad_norm": 0.4582638740539551,
+      "learning_rate": 0.001,
+      "loss": 2.1037,
+      "step": 71
+    },
+    {
+      "epoch": 0.00304594297317878,
+      "grad_norm": 0.41287997364997864,
+      "learning_rate": 0.001,
+      "loss": 2.4761,
+      "step": 72
+    },
+    {
+      "epoch": 0.003088247736695152,
+      "grad_norm": 0.6007514595985413,
+      "learning_rate": 0.001,
+      "loss": 2.221,
+      "step": 73
+    },
+    {
+      "epoch": 0.0031305525002115237,
+      "grad_norm": 0.6128891110420227,
+      "learning_rate": 0.001,
+      "loss": 2.847,
+      "step": 74
+    },
+    {
+      "epoch": 0.0031728572637278957,
+      "grad_norm": 0.713722288608551,
+      "learning_rate": 0.001,
+      "loss": 2.3765,
+      "step": 75
+    },
+    {
+      "epoch": 0.0032151620272442677,
+      "grad_norm": 0.5172343254089355,
+      "learning_rate": 0.001,
+      "loss": 2.3498,
+      "step": 76
+    },
+    {
+      "epoch": 0.00325746679076064,
+      "grad_norm": 0.44072654843330383,
+      "learning_rate": 0.001,
+      "loss": 2.6489,
+      "step": 77
+    },
+    {
+      "epoch": 0.0032997715542770114,
+      "grad_norm": 0.4492231011390686,
+      "learning_rate": 0.001,
+      "loss": 2.786,
+      "step": 78
+    },
+    {
+      "epoch": 0.0033420763177933834,
+      "grad_norm": 0.38566169142723083,
+      "learning_rate": 0.001,
+      "loss": 2.5199,
+      "step": 79
+    },
+    {
+      "epoch": 0.0033843810813097555,
+      "grad_norm": 0.44315439462661743,
+      "learning_rate": 0.001,
+      "loss": 3.2877,
+      "step": 80
+    },
+    {
+      "epoch": 0.0034266858448261275,
+      "grad_norm": 0.4293360710144043,
+      "learning_rate": 0.001,
+      "loss": 3.0169,
+      "step": 81
+    },
+    {
+      "epoch": 0.0034689906083424996,
+      "grad_norm": 0.4374128580093384,
+      "learning_rate": 0.001,
+      "loss": 1.8099,
+      "step": 82
+    },
+    {
+      "epoch": 0.003511295371858871,
+      "grad_norm": 0.3986782133579254,
+      "learning_rate": 0.001,
+      "loss": 1.953,
+      "step": 83
+    },
+    {
+      "epoch": 0.003553600135375243,
+      "grad_norm": 0.4710613787174225,
+      "learning_rate": 0.001,
+      "loss": 2.3669,
+      "step": 84
+    },
+    {
+      "epoch": 0.0035959048988916152,
+      "grad_norm": 0.8579991459846497,
+      "learning_rate": 0.001,
+      "loss": 1.8707,
+      "step": 85
+    },
+    {
+      "epoch": 0.0036382096624079873,
+      "grad_norm": 0.5034980177879333,
+      "learning_rate": 0.001,
+      "loss": 3.2651,
+      "step": 86
+    },
+    {
+      "epoch": 0.003680514425924359,
+      "grad_norm": 0.8276948928833008,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 87
+    },
+    {
+      "epoch": 0.003722819189440731,
+      "grad_norm": 0.4035736322402954,
+      "learning_rate": 0.001,
+      "loss": 2.4765,
+      "step": 88
+    },
+    {
+      "epoch": 0.003765123952957103,
+      "grad_norm": 0.6690497994422913,
+      "learning_rate": 0.001,
+      "loss": 3.0138,
+      "step": 89
+    },
+    {
+      "epoch": 0.003807428716473475,
+      "grad_norm": 0.3947993814945221,
+      "learning_rate": 0.001,
+      "loss": 2.2625,
+      "step": 90
+    },
+    {
+      "epoch": 0.003849733479989847,
+      "grad_norm": 0.7430382966995239,
+      "learning_rate": 0.001,
+      "loss": 1.8479,
+      "step": 91
+    },
+    {
+      "epoch": 0.0038920382435062187,
+      "grad_norm": 0.4126277267932892,
+      "learning_rate": 0.001,
+      "loss": 1.9788,
+      "step": 92
+    },
+    {
+      "epoch": 0.003934343007022591,
+      "grad_norm": 0.39753755927085876,
+      "learning_rate": 0.001,
+      "loss": 2.1964,
+      "step": 93
+    },
+    {
+      "epoch": 0.003976647770538962,
+      "grad_norm": 0.4967736601829529,
+      "learning_rate": 0.001,
+      "loss": 3.2626,
+      "step": 94
+    },
+    {
+      "epoch": 0.004018952534055335,
+      "grad_norm": 0.3497101664543152,
+      "learning_rate": 0.001,
+      "loss": 2.0893,
+      "step": 95
+    },
+    {
+      "epoch": 0.004061257297571706,
+      "grad_norm": 0.3811435401439667,
+      "learning_rate": 0.001,
+      "loss": 2.0541,
+      "step": 96
+    },
+    {
+      "epoch": 0.004103562061088079,
+      "grad_norm": 0.43603935837745667,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 97
+    },
+    {
+      "epoch": 0.0041458668246044505,
+      "grad_norm": 0.7152244448661804,
+      "learning_rate": 0.001,
+      "loss": 2.3025,
+      "step": 98
+    },
+    {
+      "epoch": 0.004188171588120822,
+      "grad_norm": 0.43975409865379333,
+      "learning_rate": 0.001,
+      "loss": 2.2561,
+      "step": 99
+    },
+    {
+      "epoch": 0.0042304763516371946,
+      "grad_norm": 0.41654467582702637,
+      "learning_rate": 0.001,
+      "loss": 2.5949,
+      "step": 100
+    },
+    {
+      "epoch": 0.004272781115153566,
+      "grad_norm": 0.5210958123207092,
+      "learning_rate": 0.001,
+      "loss": 2.3369,
+      "step": 101
+    },
+    {
+      "epoch": 0.004315085878669939,
+      "grad_norm": 0.35408666729927063,
+      "learning_rate": 0.001,
+      "loss": 2.4174,
+      "step": 102
+    },
+    {
+      "epoch": 0.00435739064218631,
+      "grad_norm": 0.38895490765571594,
+      "learning_rate": 0.001,
+      "loss": 2.6246,
+      "step": 103
+    },
+    {
+      "epoch": 0.004399695405702682,
+      "grad_norm": 0.34460920095443726,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 104
+    },
+    {
+      "epoch": 0.004442000169219054,
+      "grad_norm": 0.3262885510921478,
+      "learning_rate": 0.001,
+      "loss": 2.4733,
+      "step": 105
+    },
+    {
+      "epoch": 0.004484304932735426,
+      "grad_norm": 1.2628955841064453,
+      "learning_rate": 0.001,
+      "loss": 3.0509,
+      "step": 106
+    },
+    {
+      "epoch": 0.0045266096962517976,
+      "grad_norm": 0.42988988757133484,
+      "learning_rate": 0.001,
+      "loss": 2.5068,
+      "step": 107
+    },
+    {
+      "epoch": 0.00456891445976817,
+      "grad_norm": 0.42915594577789307,
+      "learning_rate": 0.001,
+      "loss": 2.6055,
+      "step": 108
+    },
+    {
+      "epoch": 0.004611219223284542,
+      "grad_norm": 0.5522227883338928,
+      "learning_rate": 0.001,
+      "loss": 2.6447,
+      "step": 109
+    },
+    {
+      "epoch": 0.004653523986800914,
+      "grad_norm": 0.3777415156364441,
+      "learning_rate": 0.001,
+      "loss": 2.4052,
+      "step": 110
+    },
+    {
+      "epoch": 0.004695828750317286,
+      "grad_norm": 0.5317783951759338,
+      "learning_rate": 0.001,
+      "loss": 2.5688,
+      "step": 111
+    },
+    {
+      "epoch": 0.004738133513833657,
+      "grad_norm": 0.4777368903160095,
+      "learning_rate": 0.001,
+      "loss": 2.3497,
+      "step": 112
+    },
+    {
+      "epoch": 0.00478043827735003,
+      "grad_norm": 1.6108906269073486,
+      "learning_rate": 0.001,
+      "loss": 2.4328,
+      "step": 113
+    },
+    {
+      "epoch": 0.004822743040866401,
+      "grad_norm": 0.3473729193210602,
+      "learning_rate": 0.001,
+      "loss": 2.3263,
+      "step": 114
+    },
+    {
+      "epoch": 0.004865047804382774,
+      "grad_norm": 0.43898317217826843,
+      "learning_rate": 0.001,
+      "loss": 3.2201,
+      "step": 115
+    },
+    {
+      "epoch": 0.0049073525678991455,
+      "grad_norm": 0.6918124556541443,
+      "learning_rate": 0.001,
+      "loss": 3.625,
+      "step": 116
+    },
+    {
+      "epoch": 0.004949657331415517,
+      "grad_norm": 9.253972053527832,
+      "learning_rate": 0.001,
+      "loss": 2.8214,
+      "step": 117
+    },
+    {
+      "epoch": 0.00499196209493189,
+      "grad_norm": 0.8446986079216003,
+      "learning_rate": 0.001,
+      "loss": 2.1835,
+      "step": 118
+    },
+    {
+      "epoch": 0.005034266858448261,
+      "grad_norm": 2.1206488609313965,
+      "learning_rate": 0.001,
+      "loss": 2.2383,
+      "step": 119
+    },
+    {
+      "epoch": 0.005076571621964634,
+      "grad_norm": 0.4840485453605652,
+      "learning_rate": 0.001,
+      "loss": 2.4211,
+      "step": 120
+    },
+    {
+      "epoch": 0.005118876385481005,
+      "grad_norm": 0.39015278220176697,
+      "learning_rate": 0.001,
+      "loss": 2.548,
+      "step": 121
+    },
+    {
+      "epoch": 0.005161181148997377,
+      "grad_norm": 0.3702382743358612,
+      "learning_rate": 0.001,
+      "loss": 2.992,
+      "step": 122
+    },
+    {
+      "epoch": 0.005203485912513749,
+      "grad_norm": 0.4635108709335327,
+      "learning_rate": 0.001,
+      "loss": 2.5117,
+      "step": 123
+    },
+    {
+      "epoch": 0.005245790676030121,
+      "grad_norm": 0.34755992889404297,
+      "learning_rate": 0.001,
+      "loss": 2.3573,
+      "step": 124
+    },
+    {
+      "epoch": 0.0052880954395464926,
+      "grad_norm": 0.5302094221115112,
+      "learning_rate": 0.001,
+      "loss": 2.4309,
+      "step": 125
+    },
+    {
+      "epoch": 0.005330400203062865,
+      "grad_norm": 0.562074601650238,
+      "learning_rate": 0.001,
+      "loss": 2.0844,
+      "step": 126
+    },
+    {
+      "epoch": 0.005372704966579237,
+      "grad_norm": 0.42684710025787354,
+      "learning_rate": 0.001,
+      "loss": 2.7916,
+      "step": 127
+    },
+    {
+      "epoch": 0.005415009730095609,
+      "grad_norm": 0.5416858792304993,
+      "learning_rate": 0.001,
+      "loss": 2.5613,
+      "step": 128
+    },
+    {
+      "epoch": 0.005457314493611981,
+      "grad_norm": 0.4300224483013153,
+      "learning_rate": 0.001,
+      "loss": 3.3811,
+      "step": 129
+    },
+    {
+      "epoch": 0.005499619257128352,
+      "grad_norm": 0.46849384903907776,
+      "learning_rate": 0.001,
+      "loss": 2.5776,
+      "step": 130
+    },
+    {
+      "epoch": 0.005541924020644725,
+      "grad_norm": 0.46816426515579224,
+      "learning_rate": 0.001,
+      "loss": 2.3562,
+      "step": 131
+    },
+    {
+      "epoch": 0.005584228784161096,
+      "grad_norm": 0.4470241069793701,
+      "learning_rate": 0.001,
+      "loss": 2.3042,
+      "step": 132
+    },
+    {
+      "epoch": 0.005626533547677469,
+      "grad_norm": 0.7884194254875183,
+      "learning_rate": 0.001,
+      "loss": 2.8377,
+      "step": 133
+    },
+    {
+      "epoch": 0.0056688383111938405,
+      "grad_norm": 0.6185025572776794,
+      "learning_rate": 0.001,
+      "loss": 2.3049,
+      "step": 134
+    },
+    {
+      "epoch": 0.005711143074710212,
+      "grad_norm": 1.048901915550232,
+      "learning_rate": 0.001,
+      "loss": 2.1582,
+      "step": 135
+    },
+    {
+      "epoch": 0.005753447838226585,
+      "grad_norm": 0.47046688199043274,
+      "learning_rate": 0.001,
+      "loss": 3.4005,
+      "step": 136
+    },
+    {
+      "epoch": 0.005795752601742956,
+      "grad_norm": 0.435097873210907,
+      "learning_rate": 0.001,
+      "loss": 2.4624,
+      "step": 137
+    },
+    {
+      "epoch": 0.005838057365259328,
+      "grad_norm": 0.41452470421791077,
+      "learning_rate": 0.001,
+      "loss": 2.7871,
+      "step": 138
+    },
+    {
+      "epoch": 0.0058803621287757,
+      "grad_norm": 0.8315361142158508,
+      "learning_rate": 0.001,
+      "loss": 2.007,
+      "step": 139
+    },
+    {
+      "epoch": 0.005922666892292072,
+      "grad_norm": 1.842594027519226,
+      "learning_rate": 0.001,
+      "loss": 2.3946,
+      "step": 140
+    },
+    {
+      "epoch": 0.005964971655808444,
+      "grad_norm": 0.6004571318626404,
+      "learning_rate": 0.001,
+      "loss": 2.3669,
+      "step": 141
+    },
+    {
+      "epoch": 0.006007276419324816,
+      "grad_norm": 0.534443199634552,
+      "learning_rate": 0.001,
+      "loss": 2.5264,
+      "step": 142
+    },
+    {
+      "epoch": 0.0060495811828411876,
+      "grad_norm": 0.4516150951385498,
+      "learning_rate": 0.001,
+      "loss": 2.3313,
+      "step": 143
+    },
+    {
+      "epoch": 0.00609188594635756,
+      "grad_norm": 0.4618544280529022,
+      "learning_rate": 0.001,
+      "loss": 2.4899,
+      "step": 144
+    },
+    {
+      "epoch": 0.006134190709873932,
+      "grad_norm": 0.7234168648719788,
+      "learning_rate": 0.001,
+      "loss": 2.6252,
+      "step": 145
+    },
+    {
+      "epoch": 0.006176495473390304,
+      "grad_norm": 0.5123864412307739,
+      "learning_rate": 0.001,
+      "loss": 2.9488,
+      "step": 146
+    },
+    {
+      "epoch": 0.006218800236906676,
+      "grad_norm": 0.5958865284919739,
+      "learning_rate": 0.001,
+      "loss": 3.4051,
+      "step": 147
+    },
+    {
+      "epoch": 0.006261105000423047,
+      "grad_norm": 0.5423961877822876,
+      "learning_rate": 0.001,
+      "loss": 4.0919,
+      "step": 148
+    },
+    {
+      "epoch": 0.00630340976393942,
+      "grad_norm": 1.3664882183074951,
+      "learning_rate": 0.001,
+      "loss": 2.0117,
+      "step": 149
+    },
+    {
+      "epoch": 0.006345714527455791,
+      "grad_norm": 0.35952356457710266,
+      "learning_rate": 0.001,
+      "loss": 2.8003,
+      "step": 150
+    },
+    {
+      "epoch": 0.006388019290972164,
+      "grad_norm": 0.43103012442588806,
+      "learning_rate": 0.001,
+      "loss": 2.4838,
+      "step": 151
+    },
+    {
+      "epoch": 0.0064303240544885355,
+      "grad_norm": 0.6071809530258179,
+      "learning_rate": 0.001,
+      "loss": 2.6764,
+      "step": 152
+    },
+    {
+      "epoch": 0.006472628818004907,
+      "grad_norm": 0.38204053044319153,
+      "learning_rate": 0.001,
+      "loss": 2.2442,
+      "step": 153
+    },
+    {
+      "epoch": 0.00651493358152128,
+      "grad_norm": 0.301871120929718,
+      "learning_rate": 0.001,
+      "loss": 1.855,
+      "step": 154
+    },
+    {
+      "epoch": 0.006557238345037651,
+      "grad_norm": 0.4744417667388916,
+      "learning_rate": 0.001,
+      "loss": 2.5978,
+      "step": 155
+    },
+    {
+      "epoch": 0.006599543108554023,
+      "grad_norm": 0.5253042578697205,
+      "learning_rate": 0.001,
+      "loss": 1.9776,
+      "step": 156
+    },
+    {
+      "epoch": 0.006641847872070395,
+      "grad_norm": 0.611546516418457,
+      "learning_rate": 0.001,
+      "loss": 2.3661,
+      "step": 157
+    },
+    {
+      "epoch": 0.006684152635586767,
+      "grad_norm": 0.40985289216041565,
+      "learning_rate": 0.001,
+      "loss": 2.6935,
+      "step": 158
+    },
+    {
+      "epoch": 0.006726457399103139,
+      "grad_norm": 0.561894953250885,
+      "learning_rate": 0.001,
+      "loss": 2.7189,
+      "step": 159
+    },
+    {
+      "epoch": 0.006768762162619511,
+      "grad_norm": 0.7344470620155334,
+      "learning_rate": 0.001,
+      "loss": 2.0996,
+      "step": 160
+    },
+    {
+      "epoch": 0.0068110669261358826,
+      "grad_norm": 0.6005984544754028,
+      "learning_rate": 0.001,
+      "loss": 2.9467,
+      "step": 161
+    },
+    {
+      "epoch": 0.006853371689652255,
+      "grad_norm": 1.274715542793274,
+      "learning_rate": 0.001,
+      "loss": 3.6441,
+      "step": 162
+    },
+    {
+      "epoch": 0.006895676453168627,
+      "grad_norm": 0.44184061884880066,
+      "learning_rate": 0.001,
+      "loss": 2.4605,
+      "step": 163
+    },
+    {
+      "epoch": 0.006937981216684999,
+      "grad_norm": 0.4935331642627716,
+      "learning_rate": 0.001,
+      "loss": 3.2329,
+      "step": 164
+    },
+    {
+      "epoch": 0.006980285980201371,
+      "grad_norm": 1.1876336336135864,
+      "learning_rate": 0.001,
+      "loss": 1.9255,
+      "step": 165
+    },
+    {
+      "epoch": 0.007022590743717742,
+      "grad_norm": 0.3959547281265259,
+      "learning_rate": 0.001,
+      "loss": 2.0602,
+      "step": 166
+    },
+    {
+      "epoch": 0.007064895507234115,
+      "grad_norm": 0.6868635416030884,
+      "learning_rate": 0.001,
+      "loss": 3.2132,
+      "step": 167
+    },
+    {
+      "epoch": 0.007107200270750486,
+      "grad_norm": 1.7616485357284546,
+      "learning_rate": 0.001,
+      "loss": 2.568,
+      "step": 168
+    },
+    {
+      "epoch": 0.007149505034266858,
+      "grad_norm": 0.7364703416824341,
+      "learning_rate": 0.001,
+      "loss": 3.4709,
+      "step": 169
+    },
+    {
+      "epoch": 0.0071918097977832305,
+      "grad_norm": 0.3650343716144562,
+      "learning_rate": 0.001,
+      "loss": 2.5879,
+      "step": 170
+    },
+    {
+      "epoch": 0.007234114561299602,
+      "grad_norm": 0.5033729076385498,
+      "learning_rate": 0.001,
+      "loss": 2.3781,
+      "step": 171
+    },
+    {
+      "epoch": 0.007276419324815975,
+      "grad_norm": 7.086312294006348,
+      "learning_rate": 0.001,
+      "loss": 3.8916,
+      "step": 172
+    },
+    {
+      "epoch": 0.007318724088332346,
+      "grad_norm": 0.6279006600379944,
+      "learning_rate": 0.001,
+      "loss": 2.9944,
+      "step": 173
+    },
+    {
+      "epoch": 0.007361028851848718,
+      "grad_norm": 0.4906074106693268,
+      "learning_rate": 0.001,
+      "loss": 3.5258,
+      "step": 174
+    },
+    {
+      "epoch": 0.00740333361536509,
+      "grad_norm": 0.3576771318912506,
+      "learning_rate": 0.001,
+      "loss": 1.7132,
+      "step": 175
+    },
+    {
+      "epoch": 0.007445638378881462,
+      "grad_norm": 0.39911970496177673,
+      "learning_rate": 0.001,
+      "loss": 3.7563,
+      "step": 176
+    },
+    {
+      "epoch": 0.007487943142397834,
+      "grad_norm": 2.3106889724731445,
+      "learning_rate": 0.001,
+      "loss": 3.3329,
+      "step": 177
+    },
+    {
+      "epoch": 0.007530247905914206,
+      "grad_norm": 1.1456013917922974,
+      "learning_rate": 0.001,
+      "loss": 1.9892,
+      "step": 178
+    },
+    {
+      "epoch": 0.007572552669430578,
+      "grad_norm": 0.457940012216568,
+      "learning_rate": 0.001,
+      "loss": 2.1332,
+      "step": 179
+    },
+    {
+      "epoch": 0.00761485743294695,
+      "grad_norm": 0.44287726283073425,
+      "learning_rate": 0.001,
+      "loss": 2.6956,
+      "step": 180
+    },
+    {
+      "epoch": 0.007657162196463322,
+      "grad_norm": 0.4767186641693115,
+      "learning_rate": 0.001,
+      "loss": 2.0452,
+      "step": 181
+    },
+    {
+      "epoch": 0.007699466959979694,
+      "grad_norm": 26.49453353881836,
+      "learning_rate": 0.001,
+      "loss": 2.0203,
+      "step": 182
+    },
+    {
+      "epoch": 0.007741771723496066,
+      "grad_norm": 0.6047090888023376,
+      "learning_rate": 0.001,
+      "loss": 2.731,
+      "step": 183
+    },
+    {
+      "epoch": 0.007784076487012437,
+      "grad_norm": 0.8945805430412292,
+      "learning_rate": 0.001,
+      "loss": 2.3029,
+      "step": 184
+    },
+    {
+      "epoch": 0.00782638125052881,
+      "grad_norm": 0.8593363761901855,
+      "learning_rate": 0.001,
+      "loss": 2.5228,
+      "step": 185
+    },
+    {
+      "epoch": 0.007868686014045181,
+      "grad_norm": 2.482999801635742,
+      "learning_rate": 0.001,
+      "loss": 3.0185,
+      "step": 186
+    },
+    {
+      "epoch": 0.007910990777561553,
+      "grad_norm": 0.7577206492424011,
+      "learning_rate": 0.001,
+      "loss": 2.948,
+      "step": 187
+    },
+    {
+      "epoch": 0.007953295541077925,
+      "grad_norm": 2.415463924407959,
+      "learning_rate": 0.001,
+      "loss": 3.2312,
+      "step": 188
+    },
+    {
+      "epoch": 0.007995600304594298,
+      "grad_norm": 0.9507613182067871,
+      "learning_rate": 0.001,
+      "loss": 2.5924,
+      "step": 189
+    },
+    {
+      "epoch": 0.00803790506811067,
+      "grad_norm": 0.7477337121963501,
+      "learning_rate": 0.001,
+      "loss": 3.3504,
+      "step": 190
+    },
+    {
+      "epoch": 0.008080209831627041,
+      "grad_norm": 1.864358901977539,
+      "learning_rate": 0.001,
+      "loss": 2.9727,
+      "step": 191
+    },
+    {
+      "epoch": 0.008122514595143413,
+      "grad_norm": 1.1556525230407715,
+      "learning_rate": 0.001,
+      "loss": 2.8587,
+      "step": 192
+    },
+    {
+      "epoch": 0.008164819358659784,
+      "grad_norm": 1.382063388824463,
+      "learning_rate": 0.001,
+      "loss": 2.4826,
+      "step": 193
+    },
+    {
+      "epoch": 0.008207124122176158,
+      "grad_norm": 0.43554991483688354,
+      "learning_rate": 0.001,
+      "loss": 2.5445,
+      "step": 194
+    },
+    {
+      "epoch": 0.00824942888569253,
+      "grad_norm": 0.4753492772579193,
+      "learning_rate": 0.001,
+      "loss": 3.262,
+      "step": 195
+    },
+    {
+      "epoch": 0.008291733649208901,
+      "grad_norm": 1.1016758680343628,
+      "learning_rate": 0.001,
+      "loss": 2.1801,
+      "step": 196
+    },
+    {
+      "epoch": 0.008334038412725273,
+      "grad_norm": 0.6898307800292969,
+      "learning_rate": 0.001,
+      "loss": 2.3774,
+      "step": 197
+    },
+    {
+      "epoch": 0.008376343176241644,
+      "grad_norm": 4.887700080871582,
+      "learning_rate": 0.001,
+      "loss": 3.0435,
+      "step": 198
+    },
+    {
+      "epoch": 0.008418647939758018,
+      "grad_norm": 1.1673232316970825,
+      "learning_rate": 0.001,
+      "loss": 2.4091,
+      "step": 199
+    },
+    {
+      "epoch": 0.008460952703274389,
+      "grad_norm": 0.5722813010215759,
+      "learning_rate": 0.001,
+      "loss": 2.1268,
+      "step": 200
+    },
+    {
+      "epoch": 0.00850325746679076,
+      "grad_norm": 0.45928850769996643,
+      "learning_rate": 0.001,
+      "loss": 2.4795,
+      "step": 201
+    },
+    {
+      "epoch": 0.008545562230307132,
+      "grad_norm": 0.429862380027771,
+      "learning_rate": 0.001,
+      "loss": 2.8722,
+      "step": 202
+    },
+    {
+      "epoch": 0.008587866993823504,
+      "grad_norm": 0.45352330803871155,
+      "learning_rate": 0.001,
+      "loss": 2.2689,
+      "step": 203
+    },
+    {
+      "epoch": 0.008630171757339877,
+      "grad_norm": 0.4726592004299164,
+      "learning_rate": 0.001,
+      "loss": 3.3355,
+      "step": 204
+    },
+    {
+      "epoch": 0.008672476520856249,
+      "grad_norm": 0.44824856519699097,
+      "learning_rate": 0.001,
+      "loss": 2.5926,
+      "step": 205
+    },
+    {
+      "epoch": 0.00871478128437262,
+      "grad_norm": 0.39711594581604004,
+      "learning_rate": 0.001,
+      "loss": 2.2368,
+      "step": 206
+    },
+    {
+      "epoch": 0.008757086047888992,
+      "grad_norm": 0.47843319177627563,
+      "learning_rate": 0.001,
+      "loss": 2.983,
+      "step": 207
+    },
+    {
+      "epoch": 0.008799390811405364,
+      "grad_norm": 0.33170491456985474,
+      "learning_rate": 0.001,
+      "loss": 2.1638,
+      "step": 208
+    },
+    {
+      "epoch": 0.008841695574921735,
+      "grad_norm": 0.4759407937526703,
+      "learning_rate": 0.001,
+      "loss": 2.4175,
+      "step": 209
+    },
+    {
+      "epoch": 0.008884000338438109,
+      "grad_norm": 0.5208505392074585,
+      "learning_rate": 0.001,
+      "loss": 3.6394,
+      "step": 210
+    },
+    {
+      "epoch": 0.00892630510195448,
+      "grad_norm": 0.5411098599433899,
+      "learning_rate": 0.001,
+      "loss": 3.212,
+      "step": 211
+    },
+    {
+      "epoch": 0.008968609865470852,
+      "grad_norm": 0.4449928104877472,
+      "learning_rate": 0.001,
+      "loss": 2.8441,
+      "step": 212
+    },
+    {
+      "epoch": 0.009010914628987223,
+      "grad_norm": 0.7975568175315857,
+      "learning_rate": 0.001,
+      "loss": 2.4782,
+      "step": 213
+    },
+    {
+      "epoch": 0.009053219392503595,
+      "grad_norm": 1.3413524627685547,
+      "learning_rate": 0.001,
+      "loss": 1.8429,
+      "step": 214
+    },
+    {
+      "epoch": 0.009095524156019968,
+      "grad_norm": 1.463435411453247,
+      "learning_rate": 0.001,
+      "loss": 2.138,
+      "step": 215
+    },
+    {
+      "epoch": 0.00913782891953634,
+      "grad_norm": 0.5830798745155334,
+      "learning_rate": 0.001,
+      "loss": 2.7381,
+      "step": 216
+    },
+    {
+      "epoch": 0.009180133683052712,
+      "grad_norm": 1.1858150959014893,
+      "learning_rate": 0.001,
+      "loss": 2.9576,
+      "step": 217
+    },
+    {
+      "epoch": 0.009222438446569083,
+      "grad_norm": 0.5822926163673401,
+      "learning_rate": 0.001,
+      "loss": 2.9898,
+      "step": 218
+    },
+    {
+      "epoch": 0.009264743210085455,
+      "grad_norm": 0.38207611441612244,
+      "learning_rate": 0.001,
+      "loss": 2.5692,
+      "step": 219
+    },
+    {
+      "epoch": 0.009307047973601828,
+      "grad_norm": 0.3924141228199005,
+      "learning_rate": 0.001,
+      "loss": 2.1596,
+      "step": 220
+    },
+    {
+      "epoch": 0.0093493527371182,
+      "grad_norm": 0.3632948398590088,
+      "learning_rate": 0.001,
+      "loss": 1.9336,
+      "step": 221
+    },
+    {
+      "epoch": 0.009391657500634571,
+      "grad_norm": 0.6469441652297974,
+      "learning_rate": 0.001,
+      "loss": 3.3174,
+      "step": 222
+    },
+    {
+      "epoch": 0.009433962264150943,
+      "grad_norm": 0.4201594293117523,
+      "learning_rate": 0.001,
+      "loss": 2.3425,
+      "step": 223
+    },
+    {
+      "epoch": 0.009476267027667315,
+      "grad_norm": 0.4190143942832947,
+      "learning_rate": 0.001,
+      "loss": 2.306,
+      "step": 224
+    },
+    {
+      "epoch": 0.009518571791183688,
+      "grad_norm": 0.7149160504341125,
+      "learning_rate": 0.001,
+      "loss": 2.82,
+      "step": 225
+    },
+    {
+      "epoch": 0.00956087655470006,
+      "grad_norm": 0.6239504814147949,
+      "learning_rate": 0.001,
+      "loss": 2.1841,
+      "step": 226
+    },
+    {
+      "epoch": 0.009603181318216431,
+      "grad_norm": 0.5916045904159546,
+      "learning_rate": 0.001,
+      "loss": 2.1635,
+      "step": 227
+    },
+    {
+      "epoch": 0.009645486081732803,
+      "grad_norm": 0.4300955832004547,
+      "learning_rate": 0.001,
+      "loss": 2.3918,
+      "step": 228
+    },
+    {
+      "epoch": 0.009687790845249174,
+      "grad_norm": 0.7572110891342163,
+      "learning_rate": 0.001,
+      "loss": 2.1596,
+      "step": 229
+    },
+    {
+      "epoch": 0.009730095608765548,
+      "grad_norm": 0.4874956011772156,
+      "learning_rate": 0.001,
+      "loss": 3.3631,
+      "step": 230
+    },
+    {
+      "epoch": 0.00977240037228192,
+      "grad_norm": 0.43173378705978394,
+      "learning_rate": 0.001,
+      "loss": 2.446,
+      "step": 231
+    },
+    {
+      "epoch": 0.009814705135798291,
+      "grad_norm": 0.6573549509048462,
+      "learning_rate": 0.001,
+      "loss": 3.0706,
+      "step": 232
+    },
+    {
+      "epoch": 0.009857009899314663,
+      "grad_norm": 0.42250698804855347,
+      "learning_rate": 0.001,
+      "loss": 2.9873,
+      "step": 233
+    },
+    {
+      "epoch": 0.009899314662831034,
+      "grad_norm": 3.47033953666687,
+      "learning_rate": 0.001,
+      "loss": 2.4825,
+      "step": 234
+    },
+    {
+      "epoch": 0.009941619426347408,
+      "grad_norm": 0.39495372772216797,
+      "learning_rate": 0.001,
+      "loss": 2.5977,
+      "step": 235
+    },
+    {
+      "epoch": 0.00998392418986378,
+      "grad_norm": 0.5418169498443604,
+      "learning_rate": 0.001,
+      "loss": 2.0364,
+      "step": 236
+    },
+    {
+      "epoch": 0.01002622895338015,
+      "grad_norm": 0.4567147195339203,
+      "learning_rate": 0.001,
+      "loss": 2.9259,
+      "step": 237
+    },
+    {
+      "epoch": 0.010068533716896522,
+      "grad_norm": 0.8719536066055298,
+      "learning_rate": 0.001,
+      "loss": 2.8369,
+      "step": 238
+    },
+    {
+      "epoch": 0.010110838480412894,
+      "grad_norm": 0.44655323028564453,
+      "learning_rate": 0.001,
+      "loss": 2.4415,
+      "step": 239
+    },
+    {
+      "epoch": 0.010153143243929267,
+      "grad_norm": 0.5208900570869446,
+      "learning_rate": 0.001,
+      "loss": 2.4879,
+      "step": 240
+    },
+    {
+      "epoch": 0.010195448007445639,
+      "grad_norm": 0.3061939775943756,
+      "learning_rate": 0.001,
+      "loss": 1.9186,
+      "step": 241
+    },
+    {
+      "epoch": 0.01023775277096201,
+      "grad_norm": 0.48679229617118835,
+      "learning_rate": 0.001,
+      "loss": 2.2632,
+      "step": 242
+    },
+    {
+      "epoch": 0.010280057534478382,
+      "grad_norm": 0.5101296305656433,
+      "learning_rate": 0.001,
+      "loss": 3.1684,
+      "step": 243
+    },
+    {
+      "epoch": 0.010322362297994754,
+      "grad_norm": 1.0889201164245605,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 244
+    },
+    {
+      "epoch": 0.010364667061511125,
+      "grad_norm": 0.4110303819179535,
+      "learning_rate": 0.001,
+      "loss": 2.9342,
+      "step": 245
+    },
+    {
+      "epoch": 0.010406971825027499,
+      "grad_norm": 0.528121292591095,
+      "learning_rate": 0.001,
+      "loss": 3.7985,
+      "step": 246
+    },
+    {
+      "epoch": 0.01044927658854387,
+      "grad_norm": 2.397808074951172,
+      "learning_rate": 0.001,
+      "loss": 3.0319,
+      "step": 247
+    },
+    {
+      "epoch": 0.010491581352060242,
+      "grad_norm": 1.5704766511917114,
+      "learning_rate": 0.001,
+      "loss": 2.6339,
+      "step": 248
+    },
+    {
+      "epoch": 0.010533886115576614,
+      "grad_norm": 2.5879034996032715,
+      "learning_rate": 0.001,
+      "loss": 2.4711,
+      "step": 249
+    },
+    {
+      "epoch": 0.010576190879092985,
+      "grad_norm": 0.5719565153121948,
+      "learning_rate": 0.001,
+      "loss": 2.0952,
+      "step": 250
+    },
+    {
+      "epoch": 0.010618495642609358,
+      "grad_norm": 0.48833173513412476,
+      "learning_rate": 0.001,
+      "loss": 1.9318,
+      "step": 251
+    },
+    {
+      "epoch": 0.01066080040612573,
+      "grad_norm": 0.4771743416786194,
+      "learning_rate": 0.001,
+      "loss": 2.4739,
+      "step": 252
+    },
+    {
+      "epoch": 0.010703105169642102,
+      "grad_norm": 0.6914018392562866,
+      "learning_rate": 0.001,
+      "loss": 3.0228,
+      "step": 253
+    },
+    {
+      "epoch": 0.010745409933158473,
+      "grad_norm": 0.655381441116333,
+      "learning_rate": 0.001,
+      "loss": 2.2491,
+      "step": 254
+    },
+    {
+      "epoch": 0.010787714696674845,
+      "grad_norm": 0.47917628288269043,
+      "learning_rate": 0.001,
+      "loss": 3.1589,
+      "step": 255
+    },
+    {
+      "epoch": 0.010830019460191218,
+      "grad_norm": 0.5288345217704773,
+      "learning_rate": 0.001,
+      "loss": 2.5891,
+      "step": 256
+    },
+    {
+      "epoch": 0.01087232422370759,
+      "grad_norm": 0.7679047584533691,
+      "learning_rate": 0.001,
+      "loss": 2.1992,
+      "step": 257
+    },
+    {
+      "epoch": 0.010914628987223961,
+      "grad_norm": 0.4910162091255188,
+      "learning_rate": 0.001,
+      "loss": 2.5994,
+      "step": 258
+    },
+    {
+      "epoch": 0.010956933750740333,
+      "grad_norm": 0.40495947003364563,
+      "learning_rate": 0.001,
+      "loss": 2.7402,
+      "step": 259
+    },
+    {
+      "epoch": 0.010999238514256705,
+      "grad_norm": 0.32625123858451843,
+      "learning_rate": 0.001,
+      "loss": 2.6868,
+      "step": 260
+    },
+    {
+      "epoch": 0.011041543277773078,
+      "grad_norm": 0.4822738468647003,
+      "learning_rate": 0.001,
+      "loss": 2.6367,
+      "step": 261
+    },
+    {
+      "epoch": 0.01108384804128945,
+      "grad_norm": 0.5915836095809937,
+      "learning_rate": 0.001,
+      "loss": 2.3544,
+      "step": 262
+    },
+    {
+      "epoch": 0.011126152804805821,
+      "grad_norm": 0.32710862159729004,
+      "learning_rate": 0.001,
+      "loss": 2.3255,
+      "step": 263
+    },
+    {
+      "epoch": 0.011168457568322193,
+      "grad_norm": 0.4875565767288208,
+      "learning_rate": 0.001,
+      "loss": 2.5211,
+      "step": 264
+    },
+    {
+      "epoch": 0.011210762331838564,
+      "grad_norm": 1.2803752422332764,
+      "learning_rate": 0.001,
+      "loss": 2.407,
+      "step": 265
+    },
+    {
+      "epoch": 0.011253067095354938,
+      "grad_norm": 0.7037876844406128,
+      "learning_rate": 0.001,
+      "loss": 3.2321,
+      "step": 266
+    },
+    {
+      "epoch": 0.01129537185887131,
+      "grad_norm": 0.36865493655204773,
+      "learning_rate": 0.001,
+      "loss": 2.4842,
+      "step": 267
+    },
+    {
+      "epoch": 0.011337676622387681,
+      "grad_norm": 0.4892154037952423,
+      "learning_rate": 0.001,
+      "loss": 3.5454,
+      "step": 268
+    },
+    {
+      "epoch": 0.011379981385904053,
+      "grad_norm": 0.5366520285606384,
+      "learning_rate": 0.001,
+      "loss": 2.1276,
+      "step": 269
+    },
+    {
+      "epoch": 0.011422286149420424,
+      "grad_norm": 0.47337809205055237,
+      "learning_rate": 0.001,
+      "loss": 3.3193,
+      "step": 270
+    },
+    {
+      "epoch": 0.011464590912936798,
+      "grad_norm": 0.4499555230140686,
+      "learning_rate": 0.001,
+      "loss": 3.5789,
+      "step": 271
+    },
+    {
+      "epoch": 0.01150689567645317,
+      "grad_norm": 0.3636587858200073,
+      "learning_rate": 0.001,
+      "loss": 2.3532,
+      "step": 272
+    },
+    {
+      "epoch": 0.01154920043996954,
+      "grad_norm": 0.5930542349815369,
+      "learning_rate": 0.001,
+      "loss": 2.3694,
+      "step": 273
+    },
+    {
+      "epoch": 0.011591505203485912,
+      "grad_norm": 0.7394574284553528,
+      "learning_rate": 0.001,
+      "loss": 3.1853,
+      "step": 274
+    },
+    {
+      "epoch": 0.011633809967002284,
+      "grad_norm": 0.6907839775085449,
+      "learning_rate": 0.001,
+      "loss": 2.5174,
+      "step": 275
+    },
+    {
+      "epoch": 0.011676114730518656,
+      "grad_norm": 1.15705406665802,
+      "learning_rate": 0.001,
+      "loss": 2.315,
+      "step": 276
+    },
+    {
+      "epoch": 0.011718419494035029,
+      "grad_norm": 0.5256633758544922,
+      "learning_rate": 0.001,
+      "loss": 2.429,
+      "step": 277
+    },
+    {
+      "epoch": 0.0117607242575514,
+      "grad_norm": 0.9328014850616455,
+      "learning_rate": 0.001,
+      "loss": 2.4643,
+      "step": 278
+    },
+    {
+      "epoch": 0.011803029021067772,
+      "grad_norm": 0.5655497312545776,
+      "learning_rate": 0.001,
+      "loss": 3.344,
+      "step": 279
+    },
+    {
+      "epoch": 0.011845333784584144,
+      "grad_norm": 0.6000591516494751,
+      "learning_rate": 0.001,
+      "loss": 3.2783,
+      "step": 280
+    },
+    {
+      "epoch": 0.011887638548100515,
+      "grad_norm": 1.3328900337219238,
+      "learning_rate": 0.001,
+      "loss": 4.1927,
+      "step": 281
+    },
+    {
+      "epoch": 0.011929943311616889,
+      "grad_norm": 2.8590199947357178,
+      "learning_rate": 0.001,
+      "loss": 2.3905,
+      "step": 282
+    },
+    {
+      "epoch": 0.01197224807513326,
+      "grad_norm": 0.7419810891151428,
+      "learning_rate": 0.001,
+      "loss": 2.6781,
+      "step": 283
+    },
+    {
+      "epoch": 0.012014552838649632,
+      "grad_norm": 0.6407875418663025,
+      "learning_rate": 0.001,
+      "loss": 2.5239,
+      "step": 284
+    },
+    {
+      "epoch": 0.012056857602166004,
+      "grad_norm": 0.7808129191398621,
+      "learning_rate": 0.001,
+      "loss": 2.41,
+      "step": 285
+    },
+    {
+      "epoch": 0.012099162365682375,
+      "grad_norm": 0.5127401947975159,
+      "learning_rate": 0.001,
+      "loss": 2.5634,
+      "step": 286
+    },
+    {
+      "epoch": 0.012141467129198748,
+      "grad_norm": 0.7659389972686768,
+      "learning_rate": 0.001,
+      "loss": 2.6071,
+      "step": 287
+    },
+    {
+      "epoch": 0.01218377189271512,
+      "grad_norm": 0.4947212040424347,
+      "learning_rate": 0.001,
+      "loss": 2.8672,
+      "step": 288
+    },
+    {
+      "epoch": 0.012226076656231492,
+      "grad_norm": 0.7338204979896545,
+      "learning_rate": 0.001,
+      "loss": 2.9333,
+      "step": 289
+    },
+    {
+      "epoch": 0.012268381419747863,
+      "grad_norm": 0.5832387804985046,
+      "learning_rate": 0.001,
+      "loss": 3.106,
+      "step": 290
+    },
+    {
+      "epoch": 0.012310686183264235,
+      "grad_norm": 0.3840453028678894,
+      "learning_rate": 0.001,
+      "loss": 1.9948,
+      "step": 291
+    },
+    {
+      "epoch": 0.012352990946780608,
+      "grad_norm": 0.6502733826637268,
+      "learning_rate": 0.001,
+      "loss": 2.5404,
+      "step": 292
+    },
+    {
+      "epoch": 0.01239529571029698,
+      "grad_norm": 0.4887555241584778,
+      "learning_rate": 0.001,
+      "loss": 2.4991,
+      "step": 293
+    },
+    {
+      "epoch": 0.012437600473813351,
+      "grad_norm": 0.9977436661720276,
+      "learning_rate": 0.001,
+      "loss": 2.0015,
+      "step": 294
+    },
+    {
+      "epoch": 0.012479905237329723,
+      "grad_norm": 0.6219228506088257,
+      "learning_rate": 0.001,
+      "loss": 2.7717,
+      "step": 295
+    },
+    {
+      "epoch": 0.012522210000846095,
+      "grad_norm": 0.4836757481098175,
+      "learning_rate": 0.001,
+      "loss": 2.207,
+      "step": 296
+    },
+    {
+      "epoch": 0.012564514764362468,
+      "grad_norm": 0.43116042017936707,
+      "learning_rate": 0.001,
+      "loss": 2.7609,
+      "step": 297
+    },
+    {
+      "epoch": 0.01260681952787884,
+      "grad_norm": 0.990545928478241,
+      "learning_rate": 0.001,
+      "loss": 2.7015,
+      "step": 298
+    },
+    {
+      "epoch": 0.012649124291395211,
+      "grad_norm": 0.39893725514411926,
+      "learning_rate": 0.001,
+      "loss": 1.6939,
+      "step": 299
+    },
+    {
+      "epoch": 0.012691429054911583,
+      "grad_norm": 0.6461228132247925,
+      "learning_rate": 0.001,
+      "loss": 2.8863,
+      "step": 300
+    },
+    {
+      "epoch": 0.012733733818427954,
+      "grad_norm": 1.7560275793075562,
+      "learning_rate": 0.001,
+      "loss": 2.7424,
+      "step": 301
+    },
+    {
+      "epoch": 0.012776038581944328,
+      "grad_norm": 0.40187498927116394,
+      "learning_rate": 0.001,
+      "loss": 2.6793,
+      "step": 302
+    },
+    {
+      "epoch": 0.0128183433454607,
+      "grad_norm": 1.5104907751083374,
+      "learning_rate": 0.001,
+      "loss": 3.1125,
+      "step": 303
+    },
+    {
+      "epoch": 0.012860648108977071,
+      "grad_norm": 0.4994620382785797,
+      "learning_rate": 0.001,
+      "loss": 2.9088,
+      "step": 304
+    },
+    {
+      "epoch": 0.012902952872493443,
+      "grad_norm": 0.4981285035610199,
+      "learning_rate": 0.001,
+      "loss": 2.1621,
+      "step": 305
+    },
+    {
+      "epoch": 0.012945257636009814,
+      "grad_norm": 0.5565193295478821,
+      "learning_rate": 0.001,
+      "loss": 2.0941,
+      "step": 306
+    },
+    {
+      "epoch": 0.012987562399526186,
+      "grad_norm": 0.5400714874267578,
+      "learning_rate": 0.001,
+      "loss": 2.6684,
+      "step": 307
+    },
+    {
+      "epoch": 0.01302986716304256,
+      "grad_norm": 0.6254021525382996,
+      "learning_rate": 0.001,
+      "loss": 2.0742,
+      "step": 308
+    },
+    {
+      "epoch": 0.01307217192655893,
+      "grad_norm": 0.6768165230751038,
+      "learning_rate": 0.001,
+      "loss": 2.6983,
+      "step": 309
+    },
+    {
+      "epoch": 0.013114476690075302,
+      "grad_norm": 0.4339453876018524,
+      "learning_rate": 0.001,
+      "loss": 2.8612,
+      "step": 310
+    },
+    {
+      "epoch": 0.013156781453591674,
+      "grad_norm": 0.5363678336143494,
+      "learning_rate": 0.001,
+      "loss": 2.3609,
+      "step": 311
+    },
+    {
+      "epoch": 0.013199086217108046,
+      "grad_norm": 0.5207034945487976,
+      "learning_rate": 0.001,
+      "loss": 2.3486,
+      "step": 312
+    },
+    {
+      "epoch": 0.013241390980624419,
+      "grad_norm": 0.6771058440208435,
+      "learning_rate": 0.001,
+      "loss": 1.5712,
+      "step": 313
+    },
+    {
+      "epoch": 0.01328369574414079,
+      "grad_norm": 0.5648268461227417,
+      "learning_rate": 0.001,
+      "loss": 3.9228,
+      "step": 314
+    },
+    {
+      "epoch": 0.013326000507657162,
+      "grad_norm": 0.464399516582489,
+      "learning_rate": 0.001,
+      "loss": 2.0737,
+      "step": 315
+    },
+    {
+      "epoch": 0.013368305271173534,
+      "grad_norm": 0.47292885184288025,
+      "learning_rate": 0.001,
+      "loss": 2.1631,
+      "step": 316
+    },
+    {
+      "epoch": 0.013410610034689905,
+      "grad_norm": 0.5086976885795593,
+      "learning_rate": 0.001,
+      "loss": 2.0718,
+      "step": 317
+    },
+    {
+      "epoch": 0.013452914798206279,
+      "grad_norm": 0.9637677669525146,
+      "learning_rate": 0.001,
+      "loss": 2.6751,
+      "step": 318
+    },
+    {
+      "epoch": 0.01349521956172265,
+      "grad_norm": 0.9571125507354736,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 319
+    },
+    {
+      "epoch": 0.013537524325239022,
+      "grad_norm": 0.37801995873451233,
+      "learning_rate": 0.001,
+      "loss": 1.9653,
+      "step": 320
+    },
+    {
+      "epoch": 0.013579829088755394,
+      "grad_norm": 0.5243141651153564,
+      "learning_rate": 0.001,
+      "loss": 2.2679,
+      "step": 321
+    },
+    {
+      "epoch": 0.013622133852271765,
+      "grad_norm": 0.6378395557403564,
+      "learning_rate": 0.001,
+      "loss": 1.8723,
+      "step": 322
+    },
+    {
+      "epoch": 0.013664438615788138,
+      "grad_norm": 0.45354440808296204,
+      "learning_rate": 0.001,
+      "loss": 2.2213,
+      "step": 323
+    },
+    {
+      "epoch": 0.01370674337930451,
+      "grad_norm": 0.976237952709198,
+      "learning_rate": 0.001,
+      "loss": 2.6927,
+      "step": 324
+    },
+    {
+      "epoch": 0.013749048142820882,
+      "grad_norm": 0.4840424358844757,
+      "learning_rate": 0.001,
+      "loss": 2.3498,
+      "step": 325
+    },
+    {
+      "epoch": 0.013791352906337253,
+      "grad_norm": 0.5081096291542053,
+      "learning_rate": 0.001,
+      "loss": 2.9997,
+      "step": 326
+    },
+    {
+      "epoch": 0.013833657669853625,
+      "grad_norm": 0.5560334324836731,
+      "learning_rate": 0.001,
+      "loss": 3.1096,
+      "step": 327
+    },
+    {
+      "epoch": 0.013875962433369998,
+      "grad_norm": 0.41798263788223267,
+      "learning_rate": 0.001,
+      "loss": 2.0079,
+      "step": 328
+    },
+    {
+      "epoch": 0.01391826719688637,
+      "grad_norm": 0.40997782349586487,
+      "learning_rate": 0.001,
+      "loss": 1.4621,
+      "step": 329
+    },
+    {
+      "epoch": 0.013960571960402741,
+      "grad_norm": 0.4821615517139435,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 330
+    },
+    {
+      "epoch": 0.014002876723919113,
+      "grad_norm": 0.332916796207428,
+      "learning_rate": 0.001,
+      "loss": 1.9644,
+      "step": 331
+    },
+    {
+      "epoch": 0.014045181487435485,
+      "grad_norm": 0.49843454360961914,
+      "learning_rate": 0.001,
+      "loss": 3.7125,
+      "step": 332
+    },
+    {
+      "epoch": 0.014087486250951858,
+      "grad_norm": 0.6166870594024658,
+      "learning_rate": 0.001,
+      "loss": 1.9159,
+      "step": 333
+    },
+    {
+      "epoch": 0.01412979101446823,
+      "grad_norm": 0.6194549202919006,
+      "learning_rate": 0.001,
+      "loss": 2.4323,
+      "step": 334
+    },
+    {
+      "epoch": 0.014172095777984601,
+      "grad_norm": 0.5313667058944702,
+      "learning_rate": 0.001,
+      "loss": 2.9331,
+      "step": 335
+    },
+    {
+      "epoch": 0.014214400541500973,
+      "grad_norm": 0.3673984408378601,
+      "learning_rate": 0.001,
+      "loss": 2.2797,
+      "step": 336
+    },
+    {
+      "epoch": 0.014256705305017344,
+      "grad_norm": 0.4663550853729248,
+      "learning_rate": 0.001,
+      "loss": 3.1726,
+      "step": 337
+    },
+    {
+      "epoch": 0.014299010068533716,
+      "grad_norm": 1.455588459968567,
+      "learning_rate": 0.001,
+      "loss": 1.7567,
+      "step": 338
+    },
+    {
+      "epoch": 0.01434131483205009,
+      "grad_norm": 1.234546184539795,
+      "learning_rate": 0.001,
+      "loss": 2.1583,
+      "step": 339
+    },
+    {
+      "epoch": 0.014383619595566461,
+      "grad_norm": 2.307666301727295,
+      "learning_rate": 0.001,
+      "loss": 2.2571,
+      "step": 340
+    },
+    {
+      "epoch": 0.014425924359082833,
+      "grad_norm": 2.126053810119629,
+      "learning_rate": 0.001,
+      "loss": 2.0934,
+      "step": 341
+    },
+    {
+      "epoch": 0.014468229122599204,
+      "grad_norm": 0.5785415768623352,
+      "learning_rate": 0.001,
+      "loss": 2.4343,
+      "step": 342
+    },
+    {
+      "epoch": 0.014510533886115576,
+      "grad_norm": 0.5483373999595642,
+      "learning_rate": 0.001,
+      "loss": 1.878,
+      "step": 343
+    },
+    {
+      "epoch": 0.01455283864963195,
+      "grad_norm": 0.7590854167938232,
+      "learning_rate": 0.001,
+      "loss": 3.4764,
+      "step": 344
+    },
+    {
+      "epoch": 0.01459514341314832,
+      "grad_norm": 1.6678118705749512,
+      "learning_rate": 0.001,
+      "loss": 1.8089,
+      "step": 345
+    },
+    {
+      "epoch": 0.014637448176664692,
+      "grad_norm": 0.5652883052825928,
+      "learning_rate": 0.001,
+      "loss": 4.2131,
+      "step": 346
+    },
+    {
+      "epoch": 0.014679752940181064,
+      "grad_norm": 0.5262782573699951,
+      "learning_rate": 0.001,
+      "loss": 3.3063,
+      "step": 347
+    },
+    {
+      "epoch": 0.014722057703697436,
+      "grad_norm": 0.5519827604293823,
+      "learning_rate": 0.001,
+      "loss": 2.9829,
+      "step": 348
+    },
+    {
+      "epoch": 0.014764362467213809,
+      "grad_norm": 0.4387628436088562,
+      "learning_rate": 0.001,
+      "loss": 2.4457,
+      "step": 349
+    },
+    {
+      "epoch": 0.01480666723073018,
+      "grad_norm": 0.5566309690475464,
+      "learning_rate": 0.001,
+      "loss": 2.4278,
+      "step": 350
+    },
+    {
+      "epoch": 0.014848971994246552,
+      "grad_norm": 0.3900257647037506,
+      "learning_rate": 0.001,
+      "loss": 2.6352,
+      "step": 351
+    },
+    {
+      "epoch": 0.014891276757762924,
+      "grad_norm": 0.4080370366573334,
+      "learning_rate": 0.001,
+      "loss": 2.1755,
+      "step": 352
+    },
+    {
+      "epoch": 0.014933581521279295,
+      "grad_norm": 0.6504301428794861,
+      "learning_rate": 0.001,
+      "loss": 2.5511,
+      "step": 353
+    },
+    {
+      "epoch": 0.014975886284795669,
+      "grad_norm": 0.9981370568275452,
+      "learning_rate": 0.001,
+      "loss": 2.4473,
+      "step": 354
+    },
+    {
+      "epoch": 0.01501819104831204,
+      "grad_norm": 1.0626752376556396,
+      "learning_rate": 0.001,
+      "loss": 2.398,
+      "step": 355
+    },
+    {
+      "epoch": 0.015060495811828412,
+      "grad_norm": 0.487619549036026,
+      "learning_rate": 0.001,
+      "loss": 2.8647,
+      "step": 356
+    },
+    {
+      "epoch": 0.015102800575344784,
+      "grad_norm": 0.47419092059135437,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 357
+    },
+    {
+      "epoch": 0.015145105338861155,
+      "grad_norm": 0.6504938006401062,
+      "learning_rate": 0.001,
+      "loss": 2.5824,
+      "step": 358
+    },
+    {
+      "epoch": 0.015187410102377528,
+      "grad_norm": 0.5113835334777832,
+      "learning_rate": 0.001,
+      "loss": 3.355,
+      "step": 359
+    },
+    {
+      "epoch": 0.0152297148658939,
+      "grad_norm": 0.38182953000068665,
+      "learning_rate": 0.001,
+      "loss": 2.3596,
+      "step": 360
+    },
+    {
+      "epoch": 0.015272019629410272,
+      "grad_norm": 0.5668423771858215,
+      "learning_rate": 0.001,
+      "loss": 2.5567,
+      "step": 361
+    },
+    {
+      "epoch": 0.015314324392926643,
+      "grad_norm": 0.6734775304794312,
+      "learning_rate": 0.001,
+      "loss": 1.5854,
+      "step": 362
+    },
+    {
+      "epoch": 0.015356629156443015,
+      "grad_norm": 3.227863073348999,
+      "learning_rate": 0.001,
+      "loss": 2.8558,
+      "step": 363
+    },
+    {
+      "epoch": 0.015398933919959388,
+      "grad_norm": 1.0969175100326538,
+      "learning_rate": 0.001,
+      "loss": 2.2552,
+      "step": 364
+    },
+    {
+      "epoch": 0.01544123868347576,
+      "grad_norm": 0.8865289092063904,
+      "learning_rate": 0.001,
+      "loss": 2.6505,
+      "step": 365
+    },
+    {
+      "epoch": 0.015483543446992131,
+      "grad_norm": 0.506199836730957,
+      "learning_rate": 0.001,
+      "loss": 2.7035,
+      "step": 366
+    },
+    {
+      "epoch": 0.015525848210508503,
+      "grad_norm": 0.4529375731945038,
+      "learning_rate": 0.001,
+      "loss": 3.187,
+      "step": 367
+    },
+    {
+      "epoch": 0.015568152974024875,
+      "grad_norm": 0.49568095803260803,
+      "learning_rate": 0.001,
+      "loss": 2.9369,
+      "step": 368
+    },
+    {
+      "epoch": 0.015610457737541246,
+      "grad_norm": 3.762782096862793,
+      "learning_rate": 0.001,
+      "loss": 3.8394,
+      "step": 369
+    },
+    {
+      "epoch": 0.01565276250105762,
+      "grad_norm": 1.1418230533599854,
+      "learning_rate": 0.001,
+      "loss": 2.9679,
+      "step": 370
+    },
+    {
+      "epoch": 0.01569506726457399,
+      "grad_norm": 1.2121808528900146,
+      "learning_rate": 0.001,
+      "loss": 3.15,
+      "step": 371
+    },
+    {
+      "epoch": 0.015737372028090363,
+      "grad_norm": 0.39264196157455444,
+      "learning_rate": 0.001,
+      "loss": 2.4917,
+      "step": 372
+    },
+    {
+      "epoch": 0.015779676791606734,
+      "grad_norm": 0.44038763642311096,
+      "learning_rate": 0.001,
+      "loss": 3.1301,
+      "step": 373
+    },
+    {
+      "epoch": 0.015821981555123106,
+      "grad_norm": 0.5781055688858032,
+      "learning_rate": 0.001,
+      "loss": 2.0814,
+      "step": 374
+    },
+    {
+      "epoch": 0.015864286318639478,
+      "grad_norm": 0.4739765226840973,
+      "learning_rate": 0.001,
+      "loss": 3.3641,
+      "step": 375
+    },
+    {
+      "epoch": 0.01590659108215585,
+      "grad_norm": 0.45817050337791443,
+      "learning_rate": 0.001,
+      "loss": 2.6101,
+      "step": 376
+    },
+    {
+      "epoch": 0.015948895845672224,
+      "grad_norm": 0.45967790484428406,
+      "learning_rate": 0.001,
+      "loss": 2.9429,
+      "step": 377
+    },
+    {
+      "epoch": 0.015991200609188596,
+      "grad_norm": 0.512673556804657,
+      "learning_rate": 0.001,
+      "loss": 2.7608,
+      "step": 378
+    },
+    {
+      "epoch": 0.016033505372704968,
+      "grad_norm": 0.32649508118629456,
+      "learning_rate": 0.001,
+      "loss": 2.9877,
+      "step": 379
+    },
+    {
+      "epoch": 0.01607581013622134,
+      "grad_norm": 0.45146769285202026,
+      "learning_rate": 0.001,
+      "loss": 3.3331,
+      "step": 380
+    },
+    {
+      "epoch": 0.01611811489973771,
+      "grad_norm": 0.7151511907577515,
+      "learning_rate": 0.001,
+      "loss": 2.1693,
+      "step": 381
+    },
+    {
+      "epoch": 0.016160419663254082,
+      "grad_norm": 0.4157434403896332,
+      "learning_rate": 0.001,
+      "loss": 2.2174,
+      "step": 382
+    },
+    {
+      "epoch": 0.016202724426770454,
+      "grad_norm": 0.4276852607727051,
+      "learning_rate": 0.001,
+      "loss": 2.6254,
+      "step": 383
+    },
+    {
+      "epoch": 0.016245029190286826,
+      "grad_norm": 1.40143620967865,
+      "learning_rate": 0.001,
+      "loss": 3.6675,
+      "step": 384
+    },
+    {
+      "epoch": 0.016287333953803197,
+      "grad_norm": 2.487680673599243,
+      "learning_rate": 0.001,
+      "loss": 1.7132,
+      "step": 385
+    },
+    {
+      "epoch": 0.01632963871731957,
+      "grad_norm": 0.7442037463188171,
+      "learning_rate": 0.001,
+      "loss": 2.906,
+      "step": 386
+    },
+    {
+      "epoch": 0.01637194348083594,
+      "grad_norm": 0.5867713093757629,
+      "learning_rate": 0.001,
+      "loss": 2.4791,
+      "step": 387
+    },
+    {
+      "epoch": 0.016414248244352315,
+      "grad_norm": 3.043729066848755,
+      "learning_rate": 0.001,
+      "loss": 3.2289,
+      "step": 388
+    },
+    {
+      "epoch": 0.016456553007868687,
+      "grad_norm": 1.29078209400177,
+      "learning_rate": 0.001,
+      "loss": 3.2969,
+      "step": 389
+    },
+    {
+      "epoch": 0.01649885777138506,
+      "grad_norm": 0.46137046813964844,
+      "learning_rate": 0.001,
+      "loss": 2.3551,
+      "step": 390
+    },
+    {
+      "epoch": 0.01654116253490143,
+      "grad_norm": 6.1766510009765625,
+      "learning_rate": 0.001,
+      "loss": 2.9738,
+      "step": 391
+    },
+    {
+      "epoch": 0.016583467298417802,
+      "grad_norm": 0.5667681694030762,
+      "learning_rate": 0.001,
+      "loss": 2.3875,
+      "step": 392
+    },
+    {
+      "epoch": 0.016625772061934174,
+      "grad_norm": 0.6917624473571777,
+      "learning_rate": 0.001,
+      "loss": 2.6381,
+      "step": 393
+    },
+    {
+      "epoch": 0.016668076825450545,
+      "grad_norm": 0.7266809344291687,
+      "learning_rate": 0.001,
+      "loss": 1.8559,
+      "step": 394
+    },
+    {
+      "epoch": 0.016710381588966917,
+      "grad_norm": 2.0730323791503906,
+      "learning_rate": 0.001,
+      "loss": 2.3098,
+      "step": 395
+    },
+    {
+      "epoch": 0.01675268635248329,
+      "grad_norm": 0.48305729031562805,
+      "learning_rate": 0.001,
+      "loss": 2.1544,
+      "step": 396
+    },
+    {
+      "epoch": 0.01679499111599966,
+      "grad_norm": 0.7128873467445374,
+      "learning_rate": 0.001,
+      "loss": 2.3416,
+      "step": 397
+    },
+    {
+      "epoch": 0.016837295879516035,
+      "grad_norm": 2.950924873352051,
+      "learning_rate": 0.001,
+      "loss": 3.7996,
+      "step": 398
+    },
+    {
+      "epoch": 0.016879600643032407,
+      "grad_norm": 0.8031635284423828,
+      "learning_rate": 0.001,
+      "loss": 2.3264,
+      "step": 399
+    },
+    {
+      "epoch": 0.016921905406548778,
+      "grad_norm": 0.6791091561317444,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 400
+    },
+    {
+      "epoch": 0.01696421017006515,
+      "grad_norm": 0.445100873708725,
+      "learning_rate": 0.001,
+      "loss": 2.1986,
+      "step": 401
+    },
+    {
+      "epoch": 0.01700651493358152,
+      "grad_norm": 0.43192756175994873,
+      "learning_rate": 0.001,
+      "loss": 2.4621,
+      "step": 402
+    },
+    {
+      "epoch": 0.017048819697097893,
+      "grad_norm": 0.6217234134674072,
+      "learning_rate": 0.001,
+      "loss": 2.6163,
+      "step": 403
+    },
+    {
+      "epoch": 0.017091124460614265,
+      "grad_norm": 7.2630791664123535,
+      "learning_rate": 0.001,
+      "loss": 2.2638,
+      "step": 404
+    },
+    {
+      "epoch": 0.017133429224130636,
+      "grad_norm": 0.48048147559165955,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 405
+    },
+    {
+      "epoch": 0.017175733987647008,
+      "grad_norm": 1.1334668397903442,
+      "learning_rate": 0.001,
+      "loss": 3.1583,
+      "step": 406
+    },
+    {
+      "epoch": 0.01721803875116338,
+      "grad_norm": 0.3903160095214844,
+      "learning_rate": 0.001,
+      "loss": 3.0678,
+      "step": 407
+    },
+    {
+      "epoch": 0.017260343514679755,
+      "grad_norm": 0.8138815760612488,
+      "learning_rate": 0.001,
+      "loss": 2.5374,
+      "step": 408
+    },
+    {
+      "epoch": 0.017302648278196126,
+      "grad_norm": 0.8442886471748352,
+      "learning_rate": 0.001,
+      "loss": 2.5721,
+      "step": 409
+    },
+    {
+      "epoch": 0.017344953041712498,
+      "grad_norm": 0.6825190186500549,
+      "learning_rate": 0.001,
+      "loss": 2.7989,
+      "step": 410
+    },
+    {
+      "epoch": 0.01738725780522887,
+      "grad_norm": 0.625694215297699,
+      "learning_rate": 0.001,
+      "loss": 3.3783,
+      "step": 411
+    },
+    {
+      "epoch": 0.01742956256874524,
+      "grad_norm": 0.5650992393493652,
+      "learning_rate": 0.001,
+      "loss": 3.2683,
+      "step": 412
+    },
+    {
+      "epoch": 0.017471867332261613,
+      "grad_norm": 0.5415240526199341,
+      "learning_rate": 0.001,
+      "loss": 3.2396,
+      "step": 413
+    },
+    {
+      "epoch": 0.017514172095777984,
+      "grad_norm": 1.5903420448303223,
+      "learning_rate": 0.001,
+      "loss": 3.1021,
+      "step": 414
+    },
+    {
+      "epoch": 0.017556476859294356,
+      "grad_norm": 0.4088042676448822,
+      "learning_rate": 0.001,
+      "loss": 2.5221,
+      "step": 415
+    },
+    {
+      "epoch": 0.017598781622810727,
+      "grad_norm": 0.367471307516098,
+      "learning_rate": 0.001,
+      "loss": 2.0654,
+      "step": 416
+    },
+    {
+      "epoch": 0.0176410863863271,
+      "grad_norm": 0.6564489603042603,
+      "learning_rate": 0.001,
+      "loss": 2.2583,
+      "step": 417
+    },
+    {
+      "epoch": 0.01768339114984347,
+      "grad_norm": 0.681338369846344,
+      "learning_rate": 0.001,
+      "loss": 2.0275,
+      "step": 418
+    },
+    {
+      "epoch": 0.017725695913359846,
+      "grad_norm": 0.36002209782600403,
+      "learning_rate": 0.001,
+      "loss": 2.3529,
+      "step": 419
+    },
+    {
+      "epoch": 0.017768000676876217,
+      "grad_norm": 0.8276366591453552,
+      "learning_rate": 0.001,
+      "loss": 1.9392,
+      "step": 420
+    },
+    {
+      "epoch": 0.01781030544039259,
+      "grad_norm": 0.6230726838111877,
+      "learning_rate": 0.001,
+      "loss": 2.0964,
+      "step": 421
+    },
+    {
+      "epoch": 0.01785261020390896,
+      "grad_norm": 0.5634045004844666,
+      "learning_rate": 0.001,
+      "loss": 2.8694,
+      "step": 422
+    },
+    {
+      "epoch": 0.017894914967425332,
+      "grad_norm": 0.6966080665588379,
+      "learning_rate": 0.001,
+      "loss": 3.0085,
+      "step": 423
+    },
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 1.007291316986084,
+      "learning_rate": 0.001,
+      "loss": 2.2582,
+      "step": 424
+    },
+    {
+      "epoch": 0.017979524494458075,
+      "grad_norm": 1.6382334232330322,
+      "learning_rate": 0.001,
+      "loss": 3.448,
+      "step": 425
+    },
+    {
+      "epoch": 0.018021829257974447,
+      "grad_norm": 0.46437203884124756,
+      "learning_rate": 0.001,
+      "loss": 2.4852,
+      "step": 426
+    },
+    {
+      "epoch": 0.01806413402149082,
+      "grad_norm": 0.6570785641670227,
+      "learning_rate": 0.001,
+      "loss": 2.6239,
+      "step": 427
+    },
+    {
+      "epoch": 0.01810643878500719,
+      "grad_norm": 0.501592755317688,
+      "learning_rate": 0.001,
+      "loss": 2.7384,
+      "step": 428
+    },
+    {
+      "epoch": 0.018148743548523565,
+      "grad_norm": 0.44482702016830444,
+      "learning_rate": 0.001,
+      "loss": 2.3664,
+      "step": 429
+    },
+    {
+      "epoch": 0.018191048312039937,
+      "grad_norm": 0.4519723951816559,
+      "learning_rate": 0.001,
+      "loss": 3.4909,
+      "step": 430
+    },
+    {
+      "epoch": 0.01823335307555631,
+      "grad_norm": 0.37969911098480225,
+      "learning_rate": 0.001,
+      "loss": 2.9331,
+      "step": 431
+    },
+    {
+      "epoch": 0.01827565783907268,
+      "grad_norm": 0.35430458188056946,
+      "learning_rate": 0.001,
+      "loss": 1.9049,
+      "step": 432
+    },
+    {
+      "epoch": 0.01831796260258905,
+      "grad_norm": 0.3929635286331177,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 433
+    },
+    {
+      "epoch": 0.018360267366105423,
+      "grad_norm": 0.6036363244056702,
+      "learning_rate": 0.001,
+      "loss": 2.2359,
+      "step": 434
+    },
+    {
+      "epoch": 0.018402572129621795,
+      "grad_norm": 0.9288585186004639,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 435
+    },
+    {
+      "epoch": 0.018444876893138167,
+      "grad_norm": 0.5024868249893188,
+      "learning_rate": 0.001,
+      "loss": 2.3324,
+      "step": 436
+    },
+    {
+      "epoch": 0.018487181656654538,
+      "grad_norm": 0.3713514804840088,
+      "learning_rate": 0.001,
+      "loss": 1.9664,
+      "step": 437
+    },
+    {
+      "epoch": 0.01852948642017091,
+      "grad_norm": 0.3861111104488373,
+      "learning_rate": 0.001,
+      "loss": 1.807,
+      "step": 438
+    },
+    {
+      "epoch": 0.018571791183687285,
+      "grad_norm": 0.42646774649620056,
+      "learning_rate": 0.001,
+      "loss": 2.1702,
+      "step": 439
+    },
+    {
+      "epoch": 0.018614095947203656,
+      "grad_norm": 1.0648521184921265,
+      "learning_rate": 0.001,
+      "loss": 2.8879,
+      "step": 440
+    },
+    {
+      "epoch": 0.018656400710720028,
+      "grad_norm": 0.5311859846115112,
+      "learning_rate": 0.001,
+      "loss": 2.2553,
+      "step": 441
+    },
+    {
+      "epoch": 0.0186987054742364,
+      "grad_norm": 0.4035789966583252,
+      "learning_rate": 0.001,
+      "loss": 2.5759,
+      "step": 442
+    },
+    {
+      "epoch": 0.01874101023775277,
+      "grad_norm": 0.4712333083152771,
+      "learning_rate": 0.001,
+      "loss": 3.1349,
+      "step": 443
+    },
+    {
+      "epoch": 0.018783315001269143,
+      "grad_norm": 0.3229919373989105,
+      "learning_rate": 0.001,
+      "loss": 2.1652,
+      "step": 444
+    },
+    {
+      "epoch": 0.018825619764785514,
+      "grad_norm": 0.3765084147453308,
+      "learning_rate": 0.001,
+      "loss": 2.9515,
+      "step": 445
+    },
+    {
+      "epoch": 0.018867924528301886,
+      "grad_norm": 0.33263400197029114,
+      "learning_rate": 0.001,
+      "loss": 2.7202,
+      "step": 446
+    },
+    {
+      "epoch": 0.018910229291818258,
+      "grad_norm": 1.0643253326416016,
+      "learning_rate": 0.001,
+      "loss": 2.129,
+      "step": 447
+    },
+    {
+      "epoch": 0.01895253405533463,
+      "grad_norm": 0.5070205926895142,
+      "learning_rate": 0.001,
+      "loss": 2.2968,
+      "step": 448
+    },
+    {
+      "epoch": 0.018994838818851,
+      "grad_norm": 0.8920692801475525,
+      "learning_rate": 0.001,
+      "loss": 2.2572,
+      "step": 449
+    },
+    {
+      "epoch": 0.019037143582367376,
+      "grad_norm": 0.5254008173942566,
+      "learning_rate": 0.001,
+      "loss": 2.6386,
+      "step": 450
+    },
+    {
+      "epoch": 0.019079448345883748,
+      "grad_norm": 0.7540894746780396,
+      "learning_rate": 0.001,
+      "loss": 3.8766,
+      "step": 451
+    },
+    {
+      "epoch": 0.01912175310940012,
+      "grad_norm": 3.530369281768799,
+      "learning_rate": 0.001,
+      "loss": 2.6652,
+      "step": 452
+    },
+    {
+      "epoch": 0.01916405787291649,
+      "grad_norm": 0.385513573884964,
+      "learning_rate": 0.001,
+      "loss": 2.0282,
+      "step": 453
+    },
+    {
+      "epoch": 0.019206362636432862,
+      "grad_norm": 0.3878650367259979,
+      "learning_rate": 0.001,
+      "loss": 2.3184,
+      "step": 454
+    },
+    {
+      "epoch": 0.019248667399949234,
+      "grad_norm": 0.47612032294273376,
+      "learning_rate": 0.001,
+      "loss": 2.2334,
+      "step": 455
+    },
+    {
+      "epoch": 0.019290972163465606,
+      "grad_norm": 0.4401743412017822,
+      "learning_rate": 0.001,
+      "loss": 1.9957,
+      "step": 456
+    },
+    {
+      "epoch": 0.019333276926981977,
+      "grad_norm": 0.4817490875720978,
+      "learning_rate": 0.001,
+      "loss": 2.1504,
+      "step": 457
+    },
+    {
+      "epoch": 0.01937558169049835,
+      "grad_norm": 0.461186021566391,
+      "learning_rate": 0.001,
+      "loss": 2.0759,
+      "step": 458
+    },
+    {
+      "epoch": 0.01941788645401472,
+      "grad_norm": 0.572381317615509,
+      "learning_rate": 0.001,
+      "loss": 2.7651,
+      "step": 459
+    },
+    {
+      "epoch": 0.019460191217531096,
+      "grad_norm": 0.4275853931903839,
+      "learning_rate": 0.001,
+      "loss": 2.5997,
+      "step": 460
+    },
+    {
+      "epoch": 0.019502495981047467,
+      "grad_norm": 0.4098559021949768,
+      "learning_rate": 0.001,
+      "loss": 2.641,
+      "step": 461
+    },
+    {
+      "epoch": 0.01954480074456384,
+      "grad_norm": 0.4118167757987976,
+      "learning_rate": 0.001,
+      "loss": 1.7914,
+      "step": 462
+    },
+    {
+      "epoch": 0.01958710550808021,
+      "grad_norm": 0.46635180711746216,
+      "learning_rate": 0.001,
+      "loss": 3.0353,
+      "step": 463
+    },
+    {
+      "epoch": 0.019629410271596582,
+      "grad_norm": 1.0905091762542725,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 464
+    },
+    {
+      "epoch": 0.019671715035112954,
+      "grad_norm": 0.42049458622932434,
+      "learning_rate": 0.001,
+      "loss": 2.4426,
+      "step": 465
+    },
+    {
+      "epoch": 0.019714019798629325,
+      "grad_norm": 0.5391716361045837,
+      "learning_rate": 0.001,
+      "loss": 2.2722,
+      "step": 466
+    },
+    {
+      "epoch": 0.019756324562145697,
+      "grad_norm": 1.7512651681900024,
+      "learning_rate": 0.001,
+      "loss": 1.9385,
+      "step": 467
+    },
+    {
+      "epoch": 0.01979862932566207,
+      "grad_norm": 0.33571651577949524,
+      "learning_rate": 0.001,
+      "loss": 2.7864,
+      "step": 468
+    },
+    {
+      "epoch": 0.01984093408917844,
+      "grad_norm": 0.4144052267074585,
+      "learning_rate": 0.001,
+      "loss": 2.1114,
+      "step": 469
+    },
+    {
+      "epoch": 0.019883238852694815,
+      "grad_norm": 0.39613720774650574,
+      "learning_rate": 0.001,
+      "loss": 2.5951,
+      "step": 470
+    },
+    {
+      "epoch": 0.019925543616211187,
+      "grad_norm": 0.5358018279075623,
+      "learning_rate": 0.001,
+      "loss": 3.1943,
+      "step": 471
+    },
+    {
+      "epoch": 0.01996784837972756,
+      "grad_norm": 0.5452474355697632,
+      "learning_rate": 0.001,
+      "loss": 2.9164,
+      "step": 472
+    },
+    {
+      "epoch": 0.02001015314324393,
+      "grad_norm": 0.3648831248283386,
+      "learning_rate": 0.001,
+      "loss": 2.3703,
+      "step": 473
+    },
+    {
+      "epoch": 0.0200524579067603,
+      "grad_norm": 0.3464433252811432,
+      "learning_rate": 0.001,
+      "loss": 2.7931,
+      "step": 474
+    },
+    {
+      "epoch": 0.020094762670276673,
+      "grad_norm": 1.0535080432891846,
+      "learning_rate": 0.001,
+      "loss": 2.5576,
+      "step": 475
+    },
+    {
+      "epoch": 0.020137067433793045,
+      "grad_norm": 0.2924720346927643,
+      "learning_rate": 0.001,
+      "loss": 2.6184,
+      "step": 476
+    },
+    {
+      "epoch": 0.020179372197309416,
+      "grad_norm": 0.388220876455307,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 477
+    },
+    {
+      "epoch": 0.020221676960825788,
+      "grad_norm": 1.577642560005188,
+      "learning_rate": 0.001,
+      "loss": 2.0881,
+      "step": 478
+    },
+    {
+      "epoch": 0.02026398172434216,
+      "grad_norm": 0.795059323310852,
+      "learning_rate": 0.001,
+      "loss": 2.5844,
+      "step": 479
+    },
+    {
+      "epoch": 0.020306286487858535,
+      "grad_norm": 0.7602189183235168,
+      "learning_rate": 0.001,
+      "loss": 2.3127,
+      "step": 480
+    },
+    {
+      "epoch": 0.020348591251374906,
+      "grad_norm": 0.44378194212913513,
+      "learning_rate": 0.001,
+      "loss": 2.1581,
+      "step": 481
+    },
+    {
+      "epoch": 0.020390896014891278,
+      "grad_norm": 0.5162745714187622,
+      "learning_rate": 0.001,
+      "loss": 2.8813,
+      "step": 482
+    },
+    {
+      "epoch": 0.02043320077840765,
+      "grad_norm": 0.39076024293899536,
+      "learning_rate": 0.001,
+      "loss": 2.1318,
+      "step": 483
+    },
+    {
+      "epoch": 0.02047550554192402,
+      "grad_norm": 0.3666277825832367,
+      "learning_rate": 0.001,
+      "loss": 2.2807,
+      "step": 484
+    },
+    {
+      "epoch": 0.020517810305440393,
+      "grad_norm": 0.5351219177246094,
+      "learning_rate": 0.001,
+      "loss": 2.2821,
+      "step": 485
+    },
+    {
+      "epoch": 0.020560115068956764,
+      "grad_norm": 0.3654989004135132,
+      "learning_rate": 0.001,
+      "loss": 2.062,
+      "step": 486
+    },
+    {
+      "epoch": 0.020602419832473136,
+      "grad_norm": 0.3425626754760742,
+      "learning_rate": 0.001,
+      "loss": 1.9772,
+      "step": 487
+    },
+    {
+      "epoch": 0.020644724595989507,
+      "grad_norm": 6.2479166984558105,
+      "learning_rate": 0.001,
+      "loss": 1.9888,
+      "step": 488
+    },
+    {
+      "epoch": 0.02068702935950588,
+      "grad_norm": 0.381218820810318,
+      "learning_rate": 0.001,
+      "loss": 2.1181,
+      "step": 489
+    },
+    {
+      "epoch": 0.02072933412302225,
+      "grad_norm": 0.37633004784584045,
+      "learning_rate": 0.001,
+      "loss": 1.9825,
+      "step": 490
+    },
+    {
+      "epoch": 0.020771638886538626,
+      "grad_norm": 0.5049352049827576,
+      "learning_rate": 0.001,
+      "loss": 2.3675,
+      "step": 491
+    },
+    {
+      "epoch": 0.020813943650054997,
+      "grad_norm": 0.3255545496940613,
+      "learning_rate": 0.001,
+      "loss": 1.4964,
+      "step": 492
+    },
+    {
+      "epoch": 0.02085624841357137,
+      "grad_norm": 0.43696948885917664,
+      "learning_rate": 0.001,
+      "loss": 2.4114,
+      "step": 493
+    },
+    {
+      "epoch": 0.02089855317708774,
+      "grad_norm": 0.37952399253845215,
+      "learning_rate": 0.001,
+      "loss": 1.6125,
+      "step": 494
+    },
+    {
+      "epoch": 0.020940857940604112,
+      "grad_norm": 0.41570955514907837,
+      "learning_rate": 0.001,
+      "loss": 3.0489,
+      "step": 495
+    },
+    {
+      "epoch": 0.020983162704120484,
+      "grad_norm": 0.46523749828338623,
+      "learning_rate": 0.001,
+      "loss": 2.5591,
+      "step": 496
+    },
+    {
+      "epoch": 0.021025467467636855,
+      "grad_norm": 0.8702875971794128,
+      "learning_rate": 0.001,
+      "loss": 2.686,
+      "step": 497
+    },
+    {
+      "epoch": 0.021067772231153227,
+      "grad_norm": 2.071568012237549,
+      "learning_rate": 0.001,
+      "loss": 2.4348,
+      "step": 498
+    },
+    {
+      "epoch": 0.0211100769946696,
+      "grad_norm": 0.8501281142234802,
+      "learning_rate": 0.001,
+      "loss": 1.8937,
+      "step": 499
+    },
+    {
+      "epoch": 0.02115238175818597,
+      "grad_norm": 22.378358840942383,
+      "learning_rate": 0.001,
+      "loss": 2.139,
+      "step": 500
+    },
+    {
+      "epoch": 0.021194686521702345,
+      "grad_norm": 56.75684356689453,
+      "learning_rate": 0.001,
+      "loss": 3.3663,
+      "step": 501
+    },
+    {
+      "epoch": 0.021236991285218717,
+      "grad_norm": 0.6316158771514893,
+      "learning_rate": 0.001,
+      "loss": 2.0616,
+      "step": 502
+    },
+    {
+      "epoch": 0.02127929604873509,
+      "grad_norm": 0.6567540764808655,
+      "learning_rate": 0.001,
+      "loss": 3.1767,
+      "step": 503
+    },
+    {
+      "epoch": 0.02132160081225146,
+      "grad_norm": 3.0959877967834473,
+      "learning_rate": 0.001,
+      "loss": 3.7518,
+      "step": 504
+    },
+    {
+      "epoch": 0.02136390557576783,
+      "grad_norm": 0.4583059549331665,
+      "learning_rate": 0.001,
+      "loss": 2.7923,
+      "step": 505
+    },
+    {
+      "epoch": 0.021406210339284203,
+      "grad_norm": 0.7765421271324158,
+      "learning_rate": 0.001,
+      "loss": 3.0666,
+      "step": 506
+    },
+    {
+      "epoch": 0.021448515102800575,
+      "grad_norm": 0.5920457243919373,
+      "learning_rate": 0.001,
+      "loss": 2.7377,
+      "step": 507
+    },
+    {
+      "epoch": 0.021490819866316947,
+      "grad_norm": 0.5121840834617615,
+      "learning_rate": 0.001,
+      "loss": 2.056,
+      "step": 508
+    },
+    {
+      "epoch": 0.021533124629833318,
+      "grad_norm": 0.3588508367538452,
+      "learning_rate": 0.001,
+      "loss": 2.4136,
+      "step": 509
+    },
+    {
+      "epoch": 0.02157542939334969,
+      "grad_norm": 0.5358933806419373,
+      "learning_rate": 0.001,
+      "loss": 2.65,
+      "step": 510
+    },
+    {
+      "epoch": 0.021617734156866065,
+      "grad_norm": 0.45659172534942627,
+      "learning_rate": 0.001,
+      "loss": 3.4225,
+      "step": 511
+    },
+    {
+      "epoch": 0.021660038920382436,
+      "grad_norm": 0.4603763520717621,
+      "learning_rate": 0.001,
+      "loss": 2.3112,
+      "step": 512
+    },
+    {
+      "epoch": 0.021702343683898808,
+      "grad_norm": 0.4111917316913605,
+      "learning_rate": 0.001,
+      "loss": 2.4298,
+      "step": 513
+    },
+    {
+      "epoch": 0.02174464844741518,
+      "grad_norm": 0.3395131826400757,
+      "learning_rate": 0.001,
+      "loss": 2.235,
+      "step": 514
+    },
+    {
+      "epoch": 0.02178695321093155,
+      "grad_norm": 0.4213927388191223,
+      "learning_rate": 0.001,
+      "loss": 2.3075,
+      "step": 515
+    },
+    {
+      "epoch": 0.021829257974447923,
+      "grad_norm": 0.6425489187240601,
+      "learning_rate": 0.001,
+      "loss": 3.7484,
+      "step": 516
+    },
+    {
+      "epoch": 0.021871562737964294,
+      "grad_norm": 0.7505455613136292,
+      "learning_rate": 0.001,
+      "loss": 1.9084,
+      "step": 517
+    },
+    {
+      "epoch": 0.021913867501480666,
+      "grad_norm": 0.35921424627304077,
+      "learning_rate": 0.001,
+      "loss": 2.0054,
+      "step": 518
+    },
+    {
+      "epoch": 0.021956172264997038,
+      "grad_norm": 0.31863313913345337,
+      "learning_rate": 0.001,
+      "loss": 2.5181,
+      "step": 519
+    },
+    {
+      "epoch": 0.02199847702851341,
+      "grad_norm": 0.34935393929481506,
+      "learning_rate": 0.001,
+      "loss": 2.407,
+      "step": 520
+    },
+    {
+      "epoch": 0.02204078179202978,
+      "grad_norm": 0.34255489706993103,
+      "learning_rate": 0.001,
+      "loss": 2.0617,
+      "step": 521
+    },
+    {
+      "epoch": 0.022083086555546156,
+      "grad_norm": 0.3920708894729614,
+      "learning_rate": 0.001,
+      "loss": 2.9539,
+      "step": 522
+    },
+    {
+      "epoch": 0.022125391319062528,
+      "grad_norm": 0.3370257318019867,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 523
+    },
+    {
+      "epoch": 0.0221676960825789,
+      "grad_norm": 0.4249412417411804,
+      "learning_rate": 0.001,
+      "loss": 1.8416,
+      "step": 524
+    },
+    {
+      "epoch": 0.02221000084609527,
+      "grad_norm": 0.35336267948150635,
+      "learning_rate": 0.001,
+      "loss": 2.1076,
+      "step": 525
+    },
+    {
+      "epoch": 0.022252305609611642,
+      "grad_norm": 0.3903971016407013,
+      "learning_rate": 0.001,
+      "loss": 2.2353,
+      "step": 526
+    },
+    {
+      "epoch": 0.022294610373128014,
+      "grad_norm": 0.3893096446990967,
+      "learning_rate": 0.001,
+      "loss": 2.6471,
+      "step": 527
+    },
+    {
+      "epoch": 0.022336915136644386,
+      "grad_norm": 0.3965369462966919,
+      "learning_rate": 0.001,
+      "loss": 2.8242,
+      "step": 528
+    },
+    {
+      "epoch": 0.022379219900160757,
+      "grad_norm": 0.5321797728538513,
+      "learning_rate": 0.001,
+      "loss": 2.8859,
+      "step": 529
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 0.4475747346878052,
+      "learning_rate": 0.001,
+      "loss": 3.02,
+      "step": 530
+    },
+    {
+      "epoch": 0.0224638294271935,
+      "grad_norm": 1.6769299507141113,
+      "learning_rate": 0.001,
+      "loss": 3.5111,
+      "step": 531
+    },
+    {
+      "epoch": 0.022506134190709876,
+      "grad_norm": 0.40092411637306213,
+      "learning_rate": 0.001,
+      "loss": 1.5887,
+      "step": 532
+    },
+    {
+      "epoch": 0.022548438954226247,
+      "grad_norm": 1.476195216178894,
+      "learning_rate": 0.001,
+      "loss": 2.5106,
+      "step": 533
+    },
+    {
+      "epoch": 0.02259074371774262,
+      "grad_norm": 1.381837248802185,
+      "learning_rate": 0.001,
+      "loss": 1.9169,
+      "step": 534
+    },
+    {
+      "epoch": 0.02263304848125899,
+      "grad_norm": 0.8170017004013062,
+      "learning_rate": 0.001,
+      "loss": 1.8514,
+      "step": 535
+    },
+    {
+      "epoch": 0.022675353244775362,
+      "grad_norm": 0.6372575163841248,
+      "learning_rate": 0.001,
+      "loss": 2.8408,
+      "step": 536
+    },
+    {
+      "epoch": 0.022717658008291734,
+      "grad_norm": 0.5463494658470154,
+      "learning_rate": 0.001,
+      "loss": 2.3254,
+      "step": 537
+    },
+    {
+      "epoch": 0.022759962771808105,
+      "grad_norm": 0.5595137476921082,
+      "learning_rate": 0.001,
+      "loss": 3.0805,
+      "step": 538
+    },
+    {
+      "epoch": 0.022802267535324477,
+      "grad_norm": 0.7195737361907959,
+      "learning_rate": 0.001,
+      "loss": 2.4751,
+      "step": 539
+    },
+    {
+      "epoch": 0.02284457229884085,
+      "grad_norm": 0.3829958438873291,
+      "learning_rate": 0.001,
+      "loss": 2.0523,
+      "step": 540
+    },
+    {
+      "epoch": 0.02288687706235722,
+      "grad_norm": 1.4294495582580566,
+      "learning_rate": 0.001,
+      "loss": 2.5157,
+      "step": 541
+    },
+    {
+      "epoch": 0.022929181825873595,
+      "grad_norm": 0.5000084638595581,
+      "learning_rate": 0.001,
+      "loss": 2.1641,
+      "step": 542
+    },
+    {
+      "epoch": 0.022971486589389967,
+      "grad_norm": 0.8625162243843079,
+      "learning_rate": 0.001,
+      "loss": 3.6346,
+      "step": 543
+    },
+    {
+      "epoch": 0.02301379135290634,
+      "grad_norm": 0.5158907771110535,
+      "learning_rate": 0.001,
+      "loss": 2.6786,
+      "step": 544
+    },
+    {
+      "epoch": 0.02305609611642271,
+      "grad_norm": 1.403883457183838,
+      "learning_rate": 0.001,
+      "loss": 2.3591,
+      "step": 545
+    },
+    {
+      "epoch": 0.02309840087993908,
+      "grad_norm": 1.1025398969650269,
+      "learning_rate": 0.001,
+      "loss": 4.0568,
+      "step": 546
+    },
+    {
+      "epoch": 0.023140705643455453,
+      "grad_norm": 0.7397220730781555,
+      "learning_rate": 0.001,
+      "loss": 2.6037,
+      "step": 547
+    },
+    {
+      "epoch": 0.023183010406971825,
+      "grad_norm": 1.0023000240325928,
+      "learning_rate": 0.001,
+      "loss": 2.5716,
+      "step": 548
+    },
+    {
+      "epoch": 0.023225315170488196,
+      "grad_norm": 0.9285343885421753,
+      "learning_rate": 0.001,
+      "loss": 2.373,
+      "step": 549
+    },
+    {
+      "epoch": 0.023267619934004568,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.001,
+      "loss": 3.0031,
+      "step": 550
+    },
+    {
+      "epoch": 0.02330992469752094,
+      "grad_norm": 0.6443062424659729,
+      "learning_rate": 0.001,
+      "loss": 2.2552,
+      "step": 551
+    },
+    {
+      "epoch": 0.02335222946103731,
+      "grad_norm": 0.4823873043060303,
+      "learning_rate": 0.001,
+      "loss": 3.3305,
+      "step": 552
+    },
+    {
+      "epoch": 0.023394534224553686,
+      "grad_norm": 0.37984201312065125,
+      "learning_rate": 0.001,
+      "loss": 2.3388,
+      "step": 553
+    },
+    {
+      "epoch": 0.023436838988070058,
+      "grad_norm": 2.326221466064453,
+      "learning_rate": 0.001,
+      "loss": 2.7738,
+      "step": 554
+    },
+    {
+      "epoch": 0.02347914375158643,
+      "grad_norm": 0.5138685703277588,
+      "learning_rate": 0.001,
+      "loss": 2.2619,
+      "step": 555
+    },
+    {
+      "epoch": 0.0235214485151028,
+      "grad_norm": 1.178184151649475,
+      "learning_rate": 0.001,
+      "loss": 2.28,
+      "step": 556
+    },
+    {
+      "epoch": 0.023563753278619173,
+      "grad_norm": 0.4544781446456909,
+      "learning_rate": 0.001,
+      "loss": 2.7337,
+      "step": 557
+    },
+    {
+      "epoch": 0.023606058042135544,
+      "grad_norm": 0.589658260345459,
+      "learning_rate": 0.001,
+      "loss": 3.867,
+      "step": 558
+    },
+    {
+      "epoch": 0.023648362805651916,
+      "grad_norm": 0.7427922487258911,
+      "learning_rate": 0.001,
+      "loss": 2.7503,
+      "step": 559
+    },
+    {
+      "epoch": 0.023690667569168287,
+      "grad_norm": 0.3987663686275482,
+      "learning_rate": 0.001,
+      "loss": 1.9502,
+      "step": 560
+    },
+    {
+      "epoch": 0.02373297233268466,
+      "grad_norm": 0.45180320739746094,
+      "learning_rate": 0.001,
+      "loss": 2.6439,
+      "step": 561
+    },
+    {
+      "epoch": 0.02377527709620103,
+      "grad_norm": 0.677946150302887,
+      "learning_rate": 0.001,
+      "loss": 2.3775,
+      "step": 562
+    },
+    {
+      "epoch": 0.023817581859717406,
+      "grad_norm": 0.6633649468421936,
+      "learning_rate": 0.001,
+      "loss": 3.2441,
+      "step": 563
+    },
+    {
+      "epoch": 0.023859886623233777,
+      "grad_norm": 0.44053682684898376,
+      "learning_rate": 0.001,
+      "loss": 2.1561,
+      "step": 564
+    },
+    {
+      "epoch": 0.02390219138675015,
+      "grad_norm": 0.47400352358818054,
+      "learning_rate": 0.001,
+      "loss": 2.6964,
+      "step": 565
+    },
+    {
+      "epoch": 0.02394449615026652,
+      "grad_norm": 0.394731730222702,
+      "learning_rate": 0.001,
+      "loss": 2.7092,
+      "step": 566
+    },
+    {
+      "epoch": 0.023986800913782892,
+      "grad_norm": 0.34284764528274536,
+      "learning_rate": 0.001,
+      "loss": 2.6047,
+      "step": 567
+    },
+    {
+      "epoch": 0.024029105677299264,
+      "grad_norm": 0.8231601119041443,
+      "learning_rate": 0.001,
+      "loss": 2.5821,
+      "step": 568
+    },
+    {
+      "epoch": 0.024071410440815635,
+      "grad_norm": 0.47505271434783936,
+      "learning_rate": 0.001,
+      "loss": 2.4732,
+      "step": 569
+    },
+    {
+      "epoch": 0.024113715204332007,
+      "grad_norm": 0.40913763642311096,
+      "learning_rate": 0.001,
+      "loss": 2.8477,
+      "step": 570
+    },
+    {
+      "epoch": 0.02415601996784838,
+      "grad_norm": 0.3416093587875366,
+      "learning_rate": 0.001,
+      "loss": 2.3009,
+      "step": 571
+    },
+    {
+      "epoch": 0.02419832473136475,
+      "grad_norm": 0.3416256010532379,
+      "learning_rate": 0.001,
+      "loss": 2.4531,
+      "step": 572
+    },
+    {
+      "epoch": 0.024240629494881125,
+      "grad_norm": 0.355182409286499,
+      "learning_rate": 0.001,
+      "loss": 2.9286,
+      "step": 573
+    },
+    {
+      "epoch": 0.024282934258397497,
+      "grad_norm": 0.5212019085884094,
+      "learning_rate": 0.001,
+      "loss": 2.0053,
+      "step": 574
+    },
+    {
+      "epoch": 0.02432523902191387,
+      "grad_norm": 0.4017347991466522,
+      "learning_rate": 0.001,
+      "loss": 3.4724,
+      "step": 575
+    },
+    {
+      "epoch": 0.02436754378543024,
+      "grad_norm": 0.34895554184913635,
+      "learning_rate": 0.001,
+      "loss": 2.7992,
+      "step": 576
+    },
+    {
+      "epoch": 0.024409848548946612,
+      "grad_norm": 0.3911406993865967,
+      "learning_rate": 0.001,
+      "loss": 2.4038,
+      "step": 577
+    },
+    {
+      "epoch": 0.024452153312462983,
+      "grad_norm": 0.38320183753967285,
+      "learning_rate": 0.001,
+      "loss": 3.2096,
+      "step": 578
+    },
+    {
+      "epoch": 0.024494458075979355,
+      "grad_norm": 0.4434884786605835,
+      "learning_rate": 0.001,
+      "loss": 2.6099,
+      "step": 579
+    },
+    {
+      "epoch": 0.024536762839495727,
+      "grad_norm": 0.34895849227905273,
+      "learning_rate": 0.001,
+      "loss": 2.2875,
+      "step": 580
+    },
+    {
+      "epoch": 0.024579067603012098,
+      "grad_norm": 0.38560599088668823,
+      "learning_rate": 0.001,
+      "loss": 2.8941,
+      "step": 581
+    },
+    {
+      "epoch": 0.02462137236652847,
+      "grad_norm": 1.215502381324768,
+      "learning_rate": 0.001,
+      "loss": 3.2749,
+      "step": 582
+    },
+    {
+      "epoch": 0.02466367713004484,
+      "grad_norm": 0.38922634720802307,
+      "learning_rate": 0.001,
+      "loss": 3.0296,
+      "step": 583
+    },
+    {
+      "epoch": 0.024705981893561216,
+      "grad_norm": 0.39207783341407776,
+      "learning_rate": 0.001,
+      "loss": 2.1744,
+      "step": 584
+    },
+    {
+      "epoch": 0.024748286657077588,
+      "grad_norm": 0.4072285294532776,
+      "learning_rate": 0.001,
+      "loss": 3.4887,
+      "step": 585
+    },
+    {
+      "epoch": 0.02479059142059396,
+      "grad_norm": 0.4496638774871826,
+      "learning_rate": 0.001,
+      "loss": 2.5288,
+      "step": 586
+    },
+    {
+      "epoch": 0.02483289618411033,
+      "grad_norm": 0.6576820015907288,
+      "learning_rate": 0.001,
+      "loss": 2.7665,
+      "step": 587
+    },
+    {
+      "epoch": 0.024875200947626703,
+      "grad_norm": 0.32960543036460876,
+      "learning_rate": 0.001,
+      "loss": 2.256,
+      "step": 588
+    },
+    {
+      "epoch": 0.024917505711143074,
+      "grad_norm": 0.44357311725616455,
+      "learning_rate": 0.001,
+      "loss": 3.0581,
+      "step": 589
+    },
+    {
+      "epoch": 0.024959810474659446,
+      "grad_norm": 0.5914167165756226,
+      "learning_rate": 0.001,
+      "loss": 2.8782,
+      "step": 590
+    },
+    {
+      "epoch": 0.025002115238175818,
+      "grad_norm": 0.6511535048484802,
+      "learning_rate": 0.001,
+      "loss": 3.7855,
+      "step": 591
+    },
+    {
+      "epoch": 0.02504442000169219,
+      "grad_norm": 0.42436838150024414,
+      "learning_rate": 0.001,
+      "loss": 1.9582,
+      "step": 592
+    },
+    {
+      "epoch": 0.02508672476520856,
+      "grad_norm": 0.38704079389572144,
+      "learning_rate": 0.001,
+      "loss": 3.0496,
+      "step": 593
+    },
+    {
+      "epoch": 0.025129029528724936,
+      "grad_norm": 13.88198184967041,
+      "learning_rate": 0.001,
+      "loss": 2.714,
+      "step": 594
+    },
+    {
+      "epoch": 0.025171334292241308,
+      "grad_norm": 0.7542757987976074,
+      "learning_rate": 0.001,
+      "loss": 2.4161,
+      "step": 595
+    },
+    {
+      "epoch": 0.02521363905575768,
+      "grad_norm": 0.475299209356308,
+      "learning_rate": 0.001,
+      "loss": 2.1213,
+      "step": 596
+    },
+    {
+      "epoch": 0.02525594381927405,
+      "grad_norm": 0.599808931350708,
+      "learning_rate": 0.001,
+      "loss": 2.5554,
+      "step": 597
+    },
+    {
+      "epoch": 0.025298248582790422,
+      "grad_norm": 0.45108288526535034,
+      "learning_rate": 0.001,
+      "loss": 2.0113,
+      "step": 598
+    },
+    {
+      "epoch": 0.025340553346306794,
+      "grad_norm": 0.471336305141449,
+      "learning_rate": 0.001,
+      "loss": 2.5406,
+      "step": 599
+    },
+    {
+      "epoch": 0.025382858109823166,
+      "grad_norm": 0.5292516946792603,
+      "learning_rate": 0.001,
+      "loss": 2.4529,
+      "step": 600
+    },
+    {
+      "epoch": 0.025425162873339537,
+      "grad_norm": 0.7199414968490601,
+      "learning_rate": 0.001,
+      "loss": 2.9323,
+      "step": 601
+    },
+    {
+      "epoch": 0.02546746763685591,
+      "grad_norm": 0.48330041766166687,
+      "learning_rate": 0.001,
+      "loss": 2.0881,
+      "step": 602
+    },
+    {
+      "epoch": 0.02550977240037228,
+      "grad_norm": 0.3391055464744568,
+      "learning_rate": 0.001,
+      "loss": 1.897,
+      "step": 603
+    },
+    {
+      "epoch": 0.025552077163888656,
+      "grad_norm": 0.4490862190723419,
+      "learning_rate": 0.001,
+      "loss": 2.6384,
+      "step": 604
+    },
+    {
+      "epoch": 0.025594381927405027,
+      "grad_norm": 0.7531641125679016,
+      "learning_rate": 0.001,
+      "loss": 3.3676,
+      "step": 605
+    },
+    {
+      "epoch": 0.0256366866909214,
+      "grad_norm": 0.3735019266605377,
+      "learning_rate": 0.001,
+      "loss": 2.1751,
+      "step": 606
+    },
+    {
+      "epoch": 0.02567899145443777,
+      "grad_norm": 0.5445840954780579,
+      "learning_rate": 0.001,
+      "loss": 2.6547,
+      "step": 607
+    },
+    {
+      "epoch": 0.025721296217954142,
+      "grad_norm": 0.4098416566848755,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 608
+    },
+    {
+      "epoch": 0.025763600981470514,
+      "grad_norm": 1.7501581907272339,
+      "learning_rate": 0.001,
+      "loss": 3.0225,
+      "step": 609
+    },
+    {
+      "epoch": 0.025805905744986885,
+      "grad_norm": 0.3755500912666321,
+      "learning_rate": 0.001,
+      "loss": 3.3205,
+      "step": 610
+    },
+    {
+      "epoch": 0.025848210508503257,
+      "grad_norm": 0.37451812624931335,
+      "learning_rate": 0.001,
+      "loss": 2.1556,
+      "step": 611
+    },
+    {
+      "epoch": 0.02589051527201963,
+      "grad_norm": 0.3815903663635254,
+      "learning_rate": 0.001,
+      "loss": 2.7517,
+      "step": 612
+    },
+    {
+      "epoch": 0.025932820035536,
+      "grad_norm": 0.327603280544281,
+      "learning_rate": 0.001,
+      "loss": 2.0281,
+      "step": 613
+    },
+    {
+      "epoch": 0.02597512479905237,
+      "grad_norm": 0.40311387181282043,
+      "learning_rate": 0.001,
+      "loss": 2.5982,
+      "step": 614
+    },
+    {
+      "epoch": 0.026017429562568747,
+      "grad_norm": 0.412945032119751,
+      "learning_rate": 0.001,
+      "loss": 2.4766,
+      "step": 615
+    },
+    {
+      "epoch": 0.02605973432608512,
+      "grad_norm": 0.7048538327217102,
+      "learning_rate": 0.001,
+      "loss": 2.8419,
+      "step": 616
+    },
+    {
+      "epoch": 0.02610203908960149,
+      "grad_norm": 0.3429849147796631,
+      "learning_rate": 0.001,
+      "loss": 2.3492,
+      "step": 617
+    },
+    {
+      "epoch": 0.02614434385311786,
+      "grad_norm": 0.39611124992370605,
+      "learning_rate": 0.001,
+      "loss": 2.307,
+      "step": 618
+    },
+    {
+      "epoch": 0.026186648616634233,
+      "grad_norm": 0.8335956931114197,
+      "learning_rate": 0.001,
+      "loss": 1.8694,
+      "step": 619
+    },
+    {
+      "epoch": 0.026228953380150605,
+      "grad_norm": 0.4806004762649536,
+      "learning_rate": 0.001,
+      "loss": 4.0374,
+      "step": 620
+    },
+    {
+      "epoch": 0.026271258143666976,
+      "grad_norm": 0.5001326203346252,
+      "learning_rate": 0.001,
+      "loss": 2.9919,
+      "step": 621
+    },
+    {
+      "epoch": 0.026313562907183348,
+      "grad_norm": 0.32230761647224426,
+      "learning_rate": 0.001,
+      "loss": 1.7651,
+      "step": 622
+    },
+    {
+      "epoch": 0.02635586767069972,
+      "grad_norm": 0.351005494594574,
+      "learning_rate": 0.001,
+      "loss": 2.1334,
+      "step": 623
+    },
+    {
+      "epoch": 0.02639817243421609,
+      "grad_norm": 0.3018239438533783,
+      "learning_rate": 0.001,
+      "loss": 2.0244,
+      "step": 624
+    },
+    {
+      "epoch": 0.026440477197732466,
+      "grad_norm": 0.4218401610851288,
+      "learning_rate": 0.001,
+      "loss": 2.0751,
+      "step": 625
+    },
+    {
+      "epoch": 0.026482781961248838,
+      "grad_norm": 0.4317582845687866,
+      "learning_rate": 0.001,
+      "loss": 3.2093,
+      "step": 626
+    },
+    {
+      "epoch": 0.02652508672476521,
+      "grad_norm": 0.604067325592041,
+      "learning_rate": 0.001,
+      "loss": 1.9582,
+      "step": 627
+    },
+    {
+      "epoch": 0.02656739148828158,
+      "grad_norm": 0.7366513013839722,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 628
+    },
+    {
+      "epoch": 0.026609696251797953,
+      "grad_norm": 0.4882150888442993,
+      "learning_rate": 0.001,
+      "loss": 2.1566,
+      "step": 629
+    },
+    {
+      "epoch": 0.026652001015314324,
+      "grad_norm": 0.4116131663322449,
+      "learning_rate": 0.001,
+      "loss": 2.4123,
+      "step": 630
+    },
+    {
+      "epoch": 0.026694305778830696,
+      "grad_norm": 0.3892541527748108,
+      "learning_rate": 0.001,
+      "loss": 2.4224,
+      "step": 631
+    },
+    {
+      "epoch": 0.026736610542347067,
+      "grad_norm": 1.0133605003356934,
+      "learning_rate": 0.001,
+      "loss": 2.1201,
+      "step": 632
+    },
+    {
+      "epoch": 0.02677891530586344,
+      "grad_norm": 0.4348303973674774,
+      "learning_rate": 0.001,
+      "loss": 2.1447,
+      "step": 633
+    },
+    {
+      "epoch": 0.02682122006937981,
+      "grad_norm": 0.5337515473365784,
+      "learning_rate": 0.001,
+      "loss": 2.2251,
+      "step": 634
+    },
+    {
+      "epoch": 0.026863524832896186,
+      "grad_norm": 0.696890652179718,
+      "learning_rate": 0.001,
+      "loss": 2.4125,
+      "step": 635
+    },
+    {
+      "epoch": 0.026905829596412557,
+      "grad_norm": 0.4275985360145569,
+      "learning_rate": 0.001,
+      "loss": 2.661,
+      "step": 636
+    },
+    {
+      "epoch": 0.02694813435992893,
+      "grad_norm": 0.6423033475875854,
+      "learning_rate": 0.001,
+      "loss": 2.4138,
+      "step": 637
+    },
+    {
+      "epoch": 0.0269904391234453,
+      "grad_norm": 0.39524486660957336,
+      "learning_rate": 0.001,
+      "loss": 2.1282,
+      "step": 638
+    },
+    {
+      "epoch": 0.027032743886961672,
+      "grad_norm": 0.37805479764938354,
+      "learning_rate": 0.001,
+      "loss": 2.4532,
+      "step": 639
+    },
+    {
+      "epoch": 0.027075048650478044,
+      "grad_norm": 0.38140320777893066,
+      "learning_rate": 0.001,
+      "loss": 2.911,
+      "step": 640
+    },
+    {
+      "epoch": 0.027117353413994415,
+      "grad_norm": 0.33297136425971985,
+      "learning_rate": 0.001,
+      "loss": 1.9244,
+      "step": 641
+    },
+    {
+      "epoch": 0.027159658177510787,
+      "grad_norm": 0.5293309688568115,
+      "learning_rate": 0.001,
+      "loss": 2.1008,
+      "step": 642
+    },
+    {
+      "epoch": 0.02720196294102716,
+      "grad_norm": 0.7821404933929443,
+      "learning_rate": 0.001,
+      "loss": 2.258,
+      "step": 643
+    },
+    {
+      "epoch": 0.02724426770454353,
+      "grad_norm": 0.3198223412036896,
+      "learning_rate": 0.001,
+      "loss": 2.4415,
+      "step": 644
+    },
+    {
+      "epoch": 0.027286572468059902,
+      "grad_norm": 0.38337278366088867,
+      "learning_rate": 0.001,
+      "loss": 3.3935,
+      "step": 645
+    },
+    {
+      "epoch": 0.027328877231576277,
+      "grad_norm": 0.34149906039237976,
+      "learning_rate": 0.001,
+      "loss": 2.4166,
+      "step": 646
+    },
+    {
+      "epoch": 0.02737118199509265,
+      "grad_norm": 0.3562192916870117,
+      "learning_rate": 0.001,
+      "loss": 1.8025,
+      "step": 647
+    },
+    {
+      "epoch": 0.02741348675860902,
+      "grad_norm": 0.42415452003479004,
+      "learning_rate": 0.001,
+      "loss": 4.2058,
+      "step": 648
+    },
+    {
+      "epoch": 0.027455791522125392,
+      "grad_norm": 0.5879824757575989,
+      "learning_rate": 0.001,
+      "loss": 1.9211,
+      "step": 649
+    },
+    {
+      "epoch": 0.027498096285641763,
+      "grad_norm": 0.27752256393432617,
+      "learning_rate": 0.001,
+      "loss": 2.2914,
+      "step": 650
+    },
+    {
+      "epoch": 0.027540401049158135,
+      "grad_norm": 0.399371474981308,
+      "learning_rate": 0.001,
+      "loss": 2.4115,
+      "step": 651
+    },
+    {
+      "epoch": 0.027582705812674507,
+      "grad_norm": 0.3457944691181183,
+      "learning_rate": 0.001,
+      "loss": 3.1942,
+      "step": 652
+    },
+    {
+      "epoch": 0.027625010576190878,
+      "grad_norm": 0.30785703659057617,
+      "learning_rate": 0.001,
+      "loss": 2.3755,
+      "step": 653
+    },
+    {
+      "epoch": 0.02766731533970725,
+      "grad_norm": 2.2121496200561523,
+      "learning_rate": 0.001,
+      "loss": 3.7683,
+      "step": 654
+    },
+    {
+      "epoch": 0.02770962010322362,
+      "grad_norm": 0.5387346148490906,
+      "learning_rate": 0.001,
+      "loss": 3.003,
+      "step": 655
+    },
+    {
+      "epoch": 0.027751924866739996,
+      "grad_norm": 0.3397753834724426,
+      "learning_rate": 0.001,
+      "loss": 2.1727,
+      "step": 656
+    },
+    {
+      "epoch": 0.027794229630256368,
+      "grad_norm": 0.5060691237449646,
+      "learning_rate": 0.001,
+      "loss": 1.9395,
+      "step": 657
+    },
+    {
+      "epoch": 0.02783653439377274,
+      "grad_norm": 0.8157616853713989,
+      "learning_rate": 0.001,
+      "loss": 2.0495,
+      "step": 658
+    },
+    {
+      "epoch": 0.02787883915728911,
+      "grad_norm": 0.2972865402698517,
+      "learning_rate": 0.001,
+      "loss": 2.1643,
+      "step": 659
+    },
+    {
+      "epoch": 0.027921143920805483,
+      "grad_norm": 0.3544447124004364,
+      "learning_rate": 0.001,
+      "loss": 1.8471,
+      "step": 660
+    },
+    {
+      "epoch": 0.027963448684321855,
+      "grad_norm": 0.4713522791862488,
+      "learning_rate": 0.001,
+      "loss": 2.3587,
+      "step": 661
+    },
+    {
+      "epoch": 0.028005753447838226,
+      "grad_norm": 0.6973052024841309,
+      "learning_rate": 0.001,
+      "loss": 2.1898,
+      "step": 662
+    },
+    {
+      "epoch": 0.028048058211354598,
+      "grad_norm": 0.30767449736595154,
+      "learning_rate": 0.001,
+      "loss": 1.8533,
+      "step": 663
+    },
+    {
+      "epoch": 0.02809036297487097,
+      "grad_norm": 0.4573628902435303,
+      "learning_rate": 0.001,
+      "loss": 2.3476,
+      "step": 664
+    },
+    {
+      "epoch": 0.02813266773838734,
+      "grad_norm": 0.3506143391132355,
+      "learning_rate": 0.001,
+      "loss": 2.35,
+      "step": 665
+    },
+    {
+      "epoch": 0.028174972501903716,
+      "grad_norm": 0.3929474353790283,
+      "learning_rate": 0.001,
+      "loss": 2.7286,
+      "step": 666
+    },
+    {
+      "epoch": 0.028217277265420088,
+      "grad_norm": 0.5329930782318115,
+      "learning_rate": 0.001,
+      "loss": 2.3529,
+      "step": 667
+    },
+    {
+      "epoch": 0.02825958202893646,
+      "grad_norm": 0.41064202785491943,
+      "learning_rate": 0.001,
+      "loss": 2.9625,
+      "step": 668
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 13.079408645629883,
+      "learning_rate": 0.001,
+      "loss": 1.974,
+      "step": 669
+    },
+    {
+      "epoch": 0.028344191555969202,
+      "grad_norm": 0.3811262845993042,
+      "learning_rate": 0.001,
+      "loss": 3.0237,
+      "step": 670
+    },
+    {
+      "epoch": 0.028386496319485574,
+      "grad_norm": 0.48839476704597473,
+      "learning_rate": 0.001,
+      "loss": 2.8337,
+      "step": 671
+    },
+    {
+      "epoch": 0.028428801083001946,
+      "grad_norm": 0.6662212014198303,
+      "learning_rate": 0.001,
+      "loss": 2.888,
+      "step": 672
+    },
+    {
+      "epoch": 0.028471105846518317,
+      "grad_norm": 0.3799011707305908,
+      "learning_rate": 0.001,
+      "loss": 3.3166,
+      "step": 673
+    },
+    {
+      "epoch": 0.02851341061003469,
+      "grad_norm": 0.4989493191242218,
+      "learning_rate": 0.001,
+      "loss": 2.6335,
+      "step": 674
+    },
+    {
+      "epoch": 0.02855571537355106,
+      "grad_norm": 0.3663240075111389,
+      "learning_rate": 0.001,
+      "loss": 2.1542,
+      "step": 675
+    },
+    {
+      "epoch": 0.028598020137067432,
+      "grad_norm": 0.3251192271709442,
+      "learning_rate": 0.001,
+      "loss": 2.1317,
+      "step": 676
+    },
+    {
+      "epoch": 0.028640324900583807,
+      "grad_norm": 0.36757683753967285,
+      "learning_rate": 0.001,
+      "loss": 2.2327,
+      "step": 677
+    },
+    {
+      "epoch": 0.02868262966410018,
+      "grad_norm": 0.34744536876678467,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 678
+    },
+    {
+      "epoch": 0.02872493442761655,
+      "grad_norm": 0.6469220519065857,
+      "learning_rate": 0.001,
+      "loss": 2.9052,
+      "step": 679
+    },
+    {
+      "epoch": 0.028767239191132922,
+      "grad_norm": 0.3472583591938019,
+      "learning_rate": 0.001,
+      "loss": 2.1404,
+      "step": 680
+    },
+    {
+      "epoch": 0.028809543954649294,
+      "grad_norm": 0.3611651062965393,
+      "learning_rate": 0.001,
+      "loss": 2.7106,
+      "step": 681
+    },
+    {
+      "epoch": 0.028851848718165665,
+      "grad_norm": 1.0853683948516846,
+      "learning_rate": 0.001,
+      "loss": 2.0889,
+      "step": 682
+    },
+    {
+      "epoch": 0.028894153481682037,
+      "grad_norm": 0.5070203542709351,
+      "learning_rate": 0.001,
+      "loss": 2.1454,
+      "step": 683
+    },
+    {
+      "epoch": 0.02893645824519841,
+      "grad_norm": 1.9084913730621338,
+      "learning_rate": 0.001,
+      "loss": 2.6574,
+      "step": 684
+    },
+    {
+      "epoch": 0.02897876300871478,
+      "grad_norm": 0.4699901044368744,
+      "learning_rate": 0.001,
+      "loss": 2.2962,
+      "step": 685
+    },
+    {
+      "epoch": 0.02902106777223115,
+      "grad_norm": 0.5868147611618042,
+      "learning_rate": 0.001,
+      "loss": 1.9462,
+      "step": 686
+    },
+    {
+      "epoch": 0.029063372535747527,
+      "grad_norm": 0.4815481901168823,
+      "learning_rate": 0.001,
+      "loss": 2.3967,
+      "step": 687
+    },
+    {
+      "epoch": 0.0291056772992639,
+      "grad_norm": 1.597350001335144,
+      "learning_rate": 0.001,
+      "loss": 2.6351,
+      "step": 688
+    },
+    {
+      "epoch": 0.02914798206278027,
+      "grad_norm": 0.3565700352191925,
+      "learning_rate": 0.001,
+      "loss": 2.237,
+      "step": 689
+    },
+    {
+      "epoch": 0.02919028682629664,
+      "grad_norm": 0.362229585647583,
+      "learning_rate": 0.001,
+      "loss": 2.5553,
+      "step": 690
+    },
+    {
+      "epoch": 0.029232591589813013,
+      "grad_norm": 0.3516606390476227,
+      "learning_rate": 0.001,
+      "loss": 2.0371,
+      "step": 691
+    },
+    {
+      "epoch": 0.029274896353329385,
+      "grad_norm": 0.3809257447719574,
+      "learning_rate": 0.001,
+      "loss": 2.0524,
+      "step": 692
+    },
+    {
+      "epoch": 0.029317201116845756,
+      "grad_norm": 0.3532337248325348,
+      "learning_rate": 0.001,
+      "loss": 2.1929,
+      "step": 693
+    },
+    {
+      "epoch": 0.029359505880362128,
+      "grad_norm": 0.5409374833106995,
+      "learning_rate": 0.001,
+      "loss": 2.7353,
+      "step": 694
+    },
+    {
+      "epoch": 0.0294018106438785,
+      "grad_norm": 0.3616213798522949,
+      "learning_rate": 0.001,
+      "loss": 1.8789,
+      "step": 695
+    },
+    {
+      "epoch": 0.02944411540739487,
+      "grad_norm": 0.31754282116889954,
+      "learning_rate": 0.001,
+      "loss": 2.617,
+      "step": 696
+    },
+    {
+      "epoch": 0.029486420170911246,
+      "grad_norm": 0.4599725604057312,
+      "learning_rate": 0.001,
+      "loss": 2.535,
+      "step": 697
+    },
+    {
+      "epoch": 0.029528724934427618,
+      "grad_norm": 1.1604878902435303,
+      "learning_rate": 0.001,
+      "loss": 2.6166,
+      "step": 698
+    },
+    {
+      "epoch": 0.02957102969794399,
+      "grad_norm": 0.3580074906349182,
+      "learning_rate": 0.001,
+      "loss": 2.5489,
+      "step": 699
+    },
+    {
+      "epoch": 0.02961333446146036,
+      "grad_norm": 0.3962548077106476,
+      "learning_rate": 0.001,
+      "loss": 1.9984,
+      "step": 700
+    },
+    {
+      "epoch": 0.029655639224976733,
+      "grad_norm": 0.342292845249176,
+      "learning_rate": 0.001,
+      "loss": 2.9691,
+      "step": 701
+    },
+    {
+      "epoch": 0.029697943988493104,
+      "grad_norm": 0.4738942086696625,
+      "learning_rate": 0.001,
+      "loss": 2.2661,
+      "step": 702
+    },
+    {
+      "epoch": 0.029740248752009476,
+      "grad_norm": 0.4526740610599518,
+      "learning_rate": 0.001,
+      "loss": 2.084,
+      "step": 703
+    },
+    {
+      "epoch": 0.029782553515525848,
+      "grad_norm": 0.37444889545440674,
+      "learning_rate": 0.001,
+      "loss": 2.6041,
+      "step": 704
+    },
+    {
+      "epoch": 0.02982485827904222,
+      "grad_norm": 0.3665851950645447,
+      "learning_rate": 0.001,
+      "loss": 2.6698,
+      "step": 705
+    },
+    {
+      "epoch": 0.02986716304255859,
+      "grad_norm": 0.3661838471889496,
+      "learning_rate": 0.001,
+      "loss": 2.8261,
+      "step": 706
+    },
+    {
+      "epoch": 0.029909467806074962,
+      "grad_norm": 9.27115249633789,
+      "learning_rate": 0.001,
+      "loss": 2.6095,
+      "step": 707
+    },
+    {
+      "epoch": 0.029951772569591337,
+      "grad_norm": 0.7346689701080322,
+      "learning_rate": 0.001,
+      "loss": 2.503,
+      "step": 708
+    },
+    {
+      "epoch": 0.02999407733310771,
+      "grad_norm": 45.74531555175781,
+      "learning_rate": 0.001,
+      "loss": 2.5024,
+      "step": 709
+    },
+    {
+      "epoch": 0.03003638209662408,
+      "grad_norm": 0.5024700164794922,
+      "learning_rate": 0.001,
+      "loss": 2.0991,
+      "step": 710
+    },
+    {
+      "epoch": 0.030078686860140452,
+      "grad_norm": 0.44415217638015747,
+      "learning_rate": 0.001,
+      "loss": 3.5278,
+      "step": 711
+    },
+    {
+      "epoch": 0.030120991623656824,
+      "grad_norm": 0.3363064229488373,
+      "learning_rate": 0.001,
+      "loss": 2.9116,
+      "step": 712
+    },
+    {
+      "epoch": 0.030163296387173195,
+      "grad_norm": 0.4409061670303345,
+      "learning_rate": 0.001,
+      "loss": 2.5084,
+      "step": 713
+    },
+    {
+      "epoch": 0.030205601150689567,
+      "grad_norm": 0.9301056265830994,
+      "learning_rate": 0.001,
+      "loss": 3.2272,
+      "step": 714
+    },
+    {
+      "epoch": 0.03024790591420594,
+      "grad_norm": 0.3383950889110565,
+      "learning_rate": 0.001,
+      "loss": 2.6067,
+      "step": 715
+    },
+    {
+      "epoch": 0.03029021067772231,
+      "grad_norm": 2.455920934677124,
+      "learning_rate": 0.001,
+      "loss": 3.4956,
+      "step": 716
+    },
+    {
+      "epoch": 0.030332515441238682,
+      "grad_norm": 0.31969428062438965,
+      "learning_rate": 0.001,
+      "loss": 2.7223,
+      "step": 717
+    },
+    {
+      "epoch": 0.030374820204755057,
+      "grad_norm": 0.4326886534690857,
+      "learning_rate": 0.001,
+      "loss": 2.7745,
+      "step": 718
+    },
+    {
+      "epoch": 0.03041712496827143,
+      "grad_norm": 0.2943534553050995,
+      "learning_rate": 0.001,
+      "loss": 1.881,
+      "step": 719
+    },
+    {
+      "epoch": 0.0304594297317878,
+      "grad_norm": 0.31899815797805786,
+      "learning_rate": 0.001,
+      "loss": 1.6605,
+      "step": 720
+    },
+    {
+      "epoch": 0.030501734495304172,
+      "grad_norm": 1.0850050449371338,
+      "learning_rate": 0.001,
+      "loss": 2.5878,
+      "step": 721
+    },
+    {
+      "epoch": 0.030544039258820543,
+      "grad_norm": 0.4966890215873718,
+      "learning_rate": 0.001,
+      "loss": 3.5874,
+      "step": 722
+    },
+    {
+      "epoch": 0.030586344022336915,
+      "grad_norm": 0.44982099533081055,
+      "learning_rate": 0.001,
+      "loss": 2.1516,
+      "step": 723
+    },
+    {
+      "epoch": 0.030628648785853287,
+      "grad_norm": 0.5142163038253784,
+      "learning_rate": 0.001,
+      "loss": 2.5352,
+      "step": 724
+    },
+    {
+      "epoch": 0.030670953549369658,
+      "grad_norm": 0.4751744270324707,
+      "learning_rate": 0.001,
+      "loss": 2.1874,
+      "step": 725
+    },
+    {
+      "epoch": 0.03071325831288603,
+      "grad_norm": 0.5114376544952393,
+      "learning_rate": 0.001,
+      "loss": 2.2583,
+      "step": 726
+    },
+    {
+      "epoch": 0.0307555630764024,
+      "grad_norm": 0.4771519601345062,
+      "learning_rate": 0.001,
+      "loss": 2.4049,
+      "step": 727
+    },
+    {
+      "epoch": 0.030797867839918776,
+      "grad_norm": 0.3221670389175415,
+      "learning_rate": 0.001,
+      "loss": 1.871,
+      "step": 728
+    },
+    {
+      "epoch": 0.030840172603435148,
+      "grad_norm": 0.6503878235816956,
+      "learning_rate": 0.001,
+      "loss": 2.5809,
+      "step": 729
+    },
+    {
+      "epoch": 0.03088247736695152,
+      "grad_norm": 0.6677407622337341,
+      "learning_rate": 0.001,
+      "loss": 3.2025,
+      "step": 730
+    },
+    {
+      "epoch": 0.03092478213046789,
+      "grad_norm": 0.46367889642715454,
+      "learning_rate": 0.001,
+      "loss": 2.6524,
+      "step": 731
+    },
+    {
+      "epoch": 0.030967086893984263,
+      "grad_norm": 0.5044416189193726,
+      "learning_rate": 0.001,
+      "loss": 2.579,
+      "step": 732
+    },
+    {
+      "epoch": 0.031009391657500635,
+      "grad_norm": 1.4938032627105713,
+      "learning_rate": 0.001,
+      "loss": 2.0139,
+      "step": 733
+    },
+    {
+      "epoch": 0.031051696421017006,
+      "grad_norm": 0.6672400236129761,
+      "learning_rate": 0.001,
+      "loss": 2.7173,
+      "step": 734
+    },
+    {
+      "epoch": 0.031094001184533378,
+      "grad_norm": 1.8434799909591675,
+      "learning_rate": 0.001,
+      "loss": 2.4786,
+      "step": 735
+    },
+    {
+      "epoch": 0.03113630594804975,
+      "grad_norm": 0.4245956540107727,
+      "learning_rate": 0.001,
+      "loss": 2.2883,
+      "step": 736
+    },
+    {
+      "epoch": 0.03117861071156612,
+      "grad_norm": 0.6225207448005676,
+      "learning_rate": 0.001,
+      "loss": 2.9485,
+      "step": 737
+    },
+    {
+      "epoch": 0.031220915475082493,
+      "grad_norm": 1.0561782121658325,
+      "learning_rate": 0.001,
+      "loss": 2.7129,
+      "step": 738
+    },
+    {
+      "epoch": 0.031263220238598864,
+      "grad_norm": 0.46243247389793396,
+      "learning_rate": 0.001,
+      "loss": 2.3125,
+      "step": 739
+    },
+    {
+      "epoch": 0.03130552500211524,
+      "grad_norm": 0.8395326137542725,
+      "learning_rate": 0.001,
+      "loss": 2.0963,
+      "step": 740
+    },
+    {
+      "epoch": 0.03134782976563161,
+      "grad_norm": 0.49101972579956055,
+      "learning_rate": 0.001,
+      "loss": 2.9097,
+      "step": 741
+    },
+    {
+      "epoch": 0.03139013452914798,
+      "grad_norm": 1.3258086442947388,
+      "learning_rate": 0.001,
+      "loss": 2.8042,
+      "step": 742
+    },
+    {
+      "epoch": 0.03143243929266435,
+      "grad_norm": 3.282763719558716,
+      "learning_rate": 0.001,
+      "loss": 2.2387,
+      "step": 743
+    },
+    {
+      "epoch": 0.031474744056180726,
+      "grad_norm": 1.5577962398529053,
+      "learning_rate": 0.001,
+      "loss": 3.241,
+      "step": 744
+    },
+    {
+      "epoch": 0.0315170488196971,
+      "grad_norm": 0.40975457429885864,
+      "learning_rate": 0.001,
+      "loss": 2.8257,
+      "step": 745
+    },
+    {
+      "epoch": 0.03155935358321347,
+      "grad_norm": 0.9148398041725159,
+      "learning_rate": 0.001,
+      "loss": 3.0109,
+      "step": 746
+    },
+    {
+      "epoch": 0.031601658346729844,
+      "grad_norm": 0.45257365703582764,
+      "learning_rate": 0.001,
+      "loss": 2.0356,
+      "step": 747
+    },
+    {
+      "epoch": 0.03164396311024621,
+      "grad_norm": 1.465282917022705,
+      "learning_rate": 0.001,
+      "loss": 2.6872,
+      "step": 748
+    },
+    {
+      "epoch": 0.03168626787376259,
+      "grad_norm": 0.41279497742652893,
+      "learning_rate": 0.001,
+      "loss": 2.452,
+      "step": 749
+    },
+    {
+      "epoch": 0.031728572637278955,
+      "grad_norm": 0.48798638582229614,
+      "learning_rate": 0.001,
+      "loss": 2.6751,
+      "step": 750
+    },
+    {
+      "epoch": 0.03177087740079533,
+      "grad_norm": 7.742496967315674,
+      "learning_rate": 0.001,
+      "loss": 1.8392,
+      "step": 751
+    },
+    {
+      "epoch": 0.0318131821643117,
+      "grad_norm": 0.8577462434768677,
+      "learning_rate": 0.001,
+      "loss": 2.1078,
+      "step": 752
+    },
+    {
+      "epoch": 0.031855486927828074,
+      "grad_norm": 0.454903781414032,
+      "learning_rate": 0.001,
+      "loss": 2.7662,
+      "step": 753
+    },
+    {
+      "epoch": 0.03189779169134445,
+      "grad_norm": 0.7064248919487,
+      "learning_rate": 0.001,
+      "loss": 2.8522,
+      "step": 754
+    },
+    {
+      "epoch": 0.03194009645486082,
+      "grad_norm": 2.0568008422851562,
+      "learning_rate": 0.001,
+      "loss": 2.7169,
+      "step": 755
+    },
+    {
+      "epoch": 0.03198240121837719,
+      "grad_norm": 0.5723927617073059,
+      "learning_rate": 0.001,
+      "loss": 2.5269,
+      "step": 756
+    },
+    {
+      "epoch": 0.03202470598189356,
+      "grad_norm": 0.37855952978134155,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 757
+    },
+    {
+      "epoch": 0.032067010745409935,
+      "grad_norm": 0.6945350170135498,
+      "learning_rate": 0.001,
+      "loss": 2.7209,
+      "step": 758
+    },
+    {
+      "epoch": 0.0321093155089263,
+      "grad_norm": 0.45397865772247314,
+      "learning_rate": 0.001,
+      "loss": 2.3489,
+      "step": 759
+    },
+    {
+      "epoch": 0.03215162027244268,
+      "grad_norm": 0.36017417907714844,
+      "learning_rate": 0.001,
+      "loss": 1.7156,
+      "step": 760
+    },
+    {
+      "epoch": 0.032193925035959046,
+      "grad_norm": 0.5965908765792847,
+      "learning_rate": 0.001,
+      "loss": 2.2643,
+      "step": 761
+    },
+    {
+      "epoch": 0.03223622979947542,
+      "grad_norm": 0.4590226709842682,
+      "learning_rate": 0.001,
+      "loss": 2.5763,
+      "step": 762
+    },
+    {
+      "epoch": 0.03227853456299179,
+      "grad_norm": 1.0792022943496704,
+      "learning_rate": 0.001,
+      "loss": 2.0132,
+      "step": 763
+    },
+    {
+      "epoch": 0.032320839326508165,
+      "grad_norm": 0.4761914908885956,
+      "learning_rate": 0.001,
+      "loss": 1.9762,
+      "step": 764
+    },
+    {
+      "epoch": 0.03236314409002454,
+      "grad_norm": 0.5108850598335266,
+      "learning_rate": 0.001,
+      "loss": 2.6186,
+      "step": 765
+    },
+    {
+      "epoch": 0.03240544885354091,
+      "grad_norm": 0.47809892892837524,
+      "learning_rate": 0.001,
+      "loss": 2.1778,
+      "step": 766
+    },
+    {
+      "epoch": 0.03244775361705728,
+      "grad_norm": 1.9831953048706055,
+      "learning_rate": 0.001,
+      "loss": 3.3741,
+      "step": 767
+    },
+    {
+      "epoch": 0.03249005838057365,
+      "grad_norm": 1.5907195806503296,
+      "learning_rate": 0.001,
+      "loss": 2.9347,
+      "step": 768
+    },
+    {
+      "epoch": 0.032532363144090026,
+      "grad_norm": 0.43444883823394775,
+      "learning_rate": 0.001,
+      "loss": 1.9747,
+      "step": 769
+    },
+    {
+      "epoch": 0.032574667907606394,
+      "grad_norm": 0.4542893171310425,
+      "learning_rate": 0.001,
+      "loss": 2.4121,
+      "step": 770
+    },
+    {
+      "epoch": 0.03261697267112277,
+      "grad_norm": 0.41902869939804077,
+      "learning_rate": 0.001,
+      "loss": 2.1587,
+      "step": 771
+    },
+    {
+      "epoch": 0.03265927743463914,
+      "grad_norm": 5.790309906005859,
+      "learning_rate": 0.001,
+      "loss": 2.8927,
+      "step": 772
+    },
+    {
+      "epoch": 0.03270158219815551,
+      "grad_norm": 0.6397790908813477,
+      "learning_rate": 0.001,
+      "loss": 2.8784,
+      "step": 773
+    },
+    {
+      "epoch": 0.03274388696167188,
+      "grad_norm": 1.084983468055725,
+      "learning_rate": 0.001,
+      "loss": 3.2401,
+      "step": 774
+    },
+    {
+      "epoch": 0.032786191725188256,
+      "grad_norm": 0.4556136429309845,
+      "learning_rate": 0.001,
+      "loss": 2.6491,
+      "step": 775
+    },
+    {
+      "epoch": 0.03282849648870463,
+      "grad_norm": 1.9562443494796753,
+      "learning_rate": 0.001,
+      "loss": 3.3097,
+      "step": 776
+    },
+    {
+      "epoch": 0.032870801252221,
+      "grad_norm": 0.3327184319496155,
+      "learning_rate": 0.001,
+      "loss": 2.3438,
+      "step": 777
+    },
+    {
+      "epoch": 0.032913106015737374,
+      "grad_norm": 0.29970988631248474,
+      "learning_rate": 0.001,
+      "loss": 1.7802,
+      "step": 778
+    },
+    {
+      "epoch": 0.03295541077925374,
+      "grad_norm": 3.424224615097046,
+      "learning_rate": 0.001,
+      "loss": 3.0777,
+      "step": 779
+    },
+    {
+      "epoch": 0.03299771554277012,
+      "grad_norm": 0.7678899168968201,
+      "learning_rate": 0.001,
+      "loss": 3.1746,
+      "step": 780
+    },
+    {
+      "epoch": 0.033040020306286486,
+      "grad_norm": 11.198646545410156,
+      "learning_rate": 0.001,
+      "loss": 4.0651,
+      "step": 781
+    },
+    {
+      "epoch": 0.03308232506980286,
+      "grad_norm": 0.5121546387672424,
+      "learning_rate": 0.001,
+      "loss": 2.7573,
+      "step": 782
+    },
+    {
+      "epoch": 0.03312462983331923,
+      "grad_norm": 0.449379563331604,
+      "learning_rate": 0.001,
+      "loss": 2.4772,
+      "step": 783
+    },
+    {
+      "epoch": 0.033166934596835604,
+      "grad_norm": 1.0964363813400269,
+      "learning_rate": 0.001,
+      "loss": 2.0905,
+      "step": 784
+    },
+    {
+      "epoch": 0.03320923936035198,
+      "grad_norm": 60.26259994506836,
+      "learning_rate": 0.001,
+      "loss": 2.596,
+      "step": 785
+    },
+    {
+      "epoch": 0.03325154412386835,
+      "grad_norm": 0.7746313810348511,
+      "learning_rate": 0.001,
+      "loss": 3.0838,
+      "step": 786
+    },
+    {
+      "epoch": 0.03329384888738472,
+      "grad_norm": 0.37214627861976624,
+      "learning_rate": 0.001,
+      "loss": 2.0005,
+      "step": 787
+    },
+    {
+      "epoch": 0.03333615365090109,
+      "grad_norm": 0.42488759756088257,
+      "learning_rate": 0.001,
+      "loss": 2.6669,
+      "step": 788
+    },
+    {
+      "epoch": 0.033378458414417465,
+      "grad_norm": 0.41743898391723633,
+      "learning_rate": 0.001,
+      "loss": 2.7365,
+      "step": 789
+    },
+    {
+      "epoch": 0.033420763177933833,
+      "grad_norm": 0.3555799424648285,
+      "learning_rate": 0.001,
+      "loss": 2.8075,
+      "step": 790
+    },
+    {
+      "epoch": 0.03346306794145021,
+      "grad_norm": 0.4383140504360199,
+      "learning_rate": 0.001,
+      "loss": 2.7863,
+      "step": 791
+    },
+    {
+      "epoch": 0.03350537270496658,
+      "grad_norm": 0.38277480006217957,
+      "learning_rate": 0.001,
+      "loss": 1.9109,
+      "step": 792
+    },
+    {
+      "epoch": 0.03354767746848295,
+      "grad_norm": 0.3921913802623749,
+      "learning_rate": 0.001,
+      "loss": 2.5242,
+      "step": 793
+    },
+    {
+      "epoch": 0.03358998223199932,
+      "grad_norm": 0.6960230469703674,
+      "learning_rate": 0.001,
+      "loss": 2.7836,
+      "step": 794
+    },
+    {
+      "epoch": 0.033632286995515695,
+      "grad_norm": 0.38849663734436035,
+      "learning_rate": 0.001,
+      "loss": 2.4461,
+      "step": 795
+    },
+    {
+      "epoch": 0.03367459175903207,
+      "grad_norm": 0.37054547667503357,
+      "learning_rate": 0.001,
+      "loss": 2.7659,
+      "step": 796
+    },
+    {
+      "epoch": 0.03371689652254844,
+      "grad_norm": 0.3550194501876831,
+      "learning_rate": 0.001,
+      "loss": 2.3983,
+      "step": 797
+    },
+    {
+      "epoch": 0.03375920128606481,
+      "grad_norm": 0.48828840255737305,
+      "learning_rate": 0.001,
+      "loss": 2.2206,
+      "step": 798
+    },
+    {
+      "epoch": 0.03380150604958118,
+      "grad_norm": 1.516994833946228,
+      "learning_rate": 0.001,
+      "loss": 1.6824,
+      "step": 799
+    },
+    {
+      "epoch": 0.033843810813097557,
+      "grad_norm": 0.5133375525474548,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 800
+    },
+    {
+      "epoch": 0.033886115576613925,
+      "grad_norm": 1.752416729927063,
+      "learning_rate": 0.001,
+      "loss": 2.2985,
+      "step": 801
+    },
+    {
+      "epoch": 0.0339284203401303,
+      "grad_norm": 0.3678469657897949,
+      "learning_rate": 0.001,
+      "loss": 2.3368,
+      "step": 802
+    },
+    {
+      "epoch": 0.03397072510364667,
+      "grad_norm": 0.5483198165893555,
+      "learning_rate": 0.001,
+      "loss": 2.8708,
+      "step": 803
+    },
+    {
+      "epoch": 0.03401302986716304,
+      "grad_norm": 0.7874423265457153,
+      "learning_rate": 0.001,
+      "loss": 2.8457,
+      "step": 804
+    },
+    {
+      "epoch": 0.03405533463067941,
+      "grad_norm": 1.40812087059021,
+      "learning_rate": 0.001,
+      "loss": 2.5861,
+      "step": 805
+    },
+    {
+      "epoch": 0.034097639394195786,
+      "grad_norm": 1.082838535308838,
+      "learning_rate": 0.001,
+      "loss": 1.8972,
+      "step": 806
+    },
+    {
+      "epoch": 0.03413994415771216,
+      "grad_norm": 1.259365200996399,
+      "learning_rate": 0.001,
+      "loss": 3.0101,
+      "step": 807
+    },
+    {
+      "epoch": 0.03418224892122853,
+      "grad_norm": 2.4589128494262695,
+      "learning_rate": 0.001,
+      "loss": 2.2503,
+      "step": 808
+    },
+    {
+      "epoch": 0.034224553684744904,
+      "grad_norm": 0.39903524518013,
+      "learning_rate": 0.001,
+      "loss": 3.306,
+      "step": 809
+    },
+    {
+      "epoch": 0.03426685844826127,
+      "grad_norm": 1.4496986865997314,
+      "learning_rate": 0.001,
+      "loss": 2.5615,
+      "step": 810
+    },
+    {
+      "epoch": 0.03430916321177765,
+      "grad_norm": 1.3750876188278198,
+      "learning_rate": 0.001,
+      "loss": 3.1478,
+      "step": 811
+    },
+    {
+      "epoch": 0.034351467975294016,
+      "grad_norm": 0.6277172565460205,
+      "learning_rate": 0.001,
+      "loss": 3.0823,
+      "step": 812
+    },
+    {
+      "epoch": 0.03439377273881039,
+      "grad_norm": 1.379515290260315,
+      "learning_rate": 0.001,
+      "loss": 2.5736,
+      "step": 813
+    },
+    {
+      "epoch": 0.03443607750232676,
+      "grad_norm": 0.5735925436019897,
+      "learning_rate": 0.001,
+      "loss": 2.4369,
+      "step": 814
+    },
+    {
+      "epoch": 0.034478382265843134,
+      "grad_norm": 0.3423369228839874,
+      "learning_rate": 0.001,
+      "loss": 2.274,
+      "step": 815
+    },
+    {
+      "epoch": 0.03452068702935951,
+      "grad_norm": 1.0366361141204834,
+      "learning_rate": 0.001,
+      "loss": 2.4531,
+      "step": 816
+    },
+    {
+      "epoch": 0.03456299179287588,
+      "grad_norm": 0.40095871686935425,
+      "learning_rate": 0.001,
+      "loss": 1.8635,
+      "step": 817
+    },
+    {
+      "epoch": 0.03460529655639225,
+      "grad_norm": 0.44527915120124817,
+      "learning_rate": 0.001,
+      "loss": 2.0943,
+      "step": 818
+    },
+    {
+      "epoch": 0.03464760131990862,
+      "grad_norm": 0.6049659848213196,
+      "learning_rate": 0.001,
+      "loss": 1.9373,
+      "step": 819
+    },
+    {
+      "epoch": 0.034689906083424996,
+      "grad_norm": 0.3163706064224243,
+      "learning_rate": 0.001,
+      "loss": 2.163,
+      "step": 820
+    },
+    {
+      "epoch": 0.034732210846941364,
+      "grad_norm": 0.7788311243057251,
+      "learning_rate": 0.001,
+      "loss": 2.3464,
+      "step": 821
+    },
+    {
+      "epoch": 0.03477451561045774,
+      "grad_norm": 0.46176034212112427,
+      "learning_rate": 0.001,
+      "loss": 3.2508,
+      "step": 822
+    },
+    {
+      "epoch": 0.03481682037397411,
+      "grad_norm": 0.3383532464504242,
+      "learning_rate": 0.001,
+      "loss": 1.9256,
+      "step": 823
+    },
+    {
+      "epoch": 0.03485912513749048,
+      "grad_norm": 0.4552175998687744,
+      "learning_rate": 0.001,
+      "loss": 2.0759,
+      "step": 824
+    },
+    {
+      "epoch": 0.03490142990100685,
+      "grad_norm": 1.0574527978897095,
+      "learning_rate": 0.001,
+      "loss": 2.2718,
+      "step": 825
+    },
+    {
+      "epoch": 0.034943734664523225,
+      "grad_norm": 0.3664693236351013,
+      "learning_rate": 0.001,
+      "loss": 2.5472,
+      "step": 826
+    },
+    {
+      "epoch": 0.0349860394280396,
+      "grad_norm": 1.357131838798523,
+      "learning_rate": 0.001,
+      "loss": 2.9158,
+      "step": 827
+    },
+    {
+      "epoch": 0.03502834419155597,
+      "grad_norm": 0.5285866260528564,
+      "learning_rate": 0.001,
+      "loss": 2.1292,
+      "step": 828
+    },
+    {
+      "epoch": 0.035070648955072344,
+      "grad_norm": 0.3371189534664154,
+      "learning_rate": 0.001,
+      "loss": 2.5495,
+      "step": 829
+    },
+    {
+      "epoch": 0.03511295371858871,
+      "grad_norm": 0.33508920669555664,
+      "learning_rate": 0.001,
+      "loss": 1.6964,
+      "step": 830
+    },
+    {
+      "epoch": 0.03515525848210509,
+      "grad_norm": 4.2542643547058105,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 831
+    },
+    {
+      "epoch": 0.035197563245621455,
+      "grad_norm": 0.29248476028442383,
+      "learning_rate": 0.001,
+      "loss": 2.0451,
+      "step": 832
+    },
+    {
+      "epoch": 0.03523986800913783,
+      "grad_norm": 0.2993159592151642,
+      "learning_rate": 0.001,
+      "loss": 2.3614,
+      "step": 833
+    },
+    {
+      "epoch": 0.0352821727726542,
+      "grad_norm": 0.34885692596435547,
+      "learning_rate": 0.001,
+      "loss": 2.5053,
+      "step": 834
+    },
+    {
+      "epoch": 0.03532447753617057,
+      "grad_norm": 0.3291875422000885,
+      "learning_rate": 0.001,
+      "loss": 2.3983,
+      "step": 835
+    },
+    {
+      "epoch": 0.03536678229968694,
+      "grad_norm": 0.34914126992225647,
+      "learning_rate": 0.001,
+      "loss": 2.0887,
+      "step": 836
+    },
+    {
+      "epoch": 0.035409087063203316,
+      "grad_norm": 0.38519594073295593,
+      "learning_rate": 0.001,
+      "loss": 2.7274,
+      "step": 837
+    },
+    {
+      "epoch": 0.03545139182671969,
+      "grad_norm": 5.37352991104126,
+      "learning_rate": 0.001,
+      "loss": 2.5118,
+      "step": 838
+    },
+    {
+      "epoch": 0.03549369659023606,
+      "grad_norm": 0.6151306629180908,
+      "learning_rate": 0.001,
+      "loss": 2.654,
+      "step": 839
+    },
+    {
+      "epoch": 0.035536001353752435,
+      "grad_norm": 0.504226803779602,
+      "learning_rate": 0.001,
+      "loss": 2.674,
+      "step": 840
+    },
+    {
+      "epoch": 0.0355783061172688,
+      "grad_norm": 0.3421787619590759,
+      "learning_rate": 0.001,
+      "loss": 2.526,
+      "step": 841
+    },
+    {
+      "epoch": 0.03562061088078518,
+      "grad_norm": 0.3042963445186615,
+      "learning_rate": 0.001,
+      "loss": 3.2472,
+      "step": 842
+    },
+    {
+      "epoch": 0.035662915644301546,
+      "grad_norm": 0.3438011705875397,
+      "learning_rate": 0.001,
+      "loss": 2.2514,
+      "step": 843
+    },
+    {
+      "epoch": 0.03570522040781792,
+      "grad_norm": 0.3311542570590973,
+      "learning_rate": 0.001,
+      "loss": 2.3859,
+      "step": 844
+    },
+    {
+      "epoch": 0.03574752517133429,
+      "grad_norm": 0.5320329666137695,
+      "learning_rate": 0.001,
+      "loss": 1.6051,
+      "step": 845
+    },
+    {
+      "epoch": 0.035789829934850664,
+      "grad_norm": 0.3768901526927948,
+      "learning_rate": 0.001,
+      "loss": 2.8369,
+      "step": 846
+    },
+    {
+      "epoch": 0.03583213469836704,
+      "grad_norm": 1.0556464195251465,
+      "learning_rate": 0.001,
+      "loss": 2.1928,
+      "step": 847
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 0.40488889813423157,
+      "learning_rate": 0.001,
+      "loss": 1.6987,
+      "step": 848
+    },
+    {
+      "epoch": 0.03591674422539978,
+      "grad_norm": 2.706275224685669,
+      "learning_rate": 0.001,
+      "loss": 2.6072,
+      "step": 849
+    },
+    {
+      "epoch": 0.03595904898891615,
+      "grad_norm": 0.8498132824897766,
+      "learning_rate": 0.001,
+      "loss": 3.2416,
+      "step": 850
+    },
+    {
+      "epoch": 0.036001353752432526,
+      "grad_norm": 0.38959065079689026,
+      "learning_rate": 0.001,
+      "loss": 2.3458,
+      "step": 851
+    },
+    {
+      "epoch": 0.036043658515948894,
+      "grad_norm": 0.41819003224372864,
+      "learning_rate": 0.001,
+      "loss": 2.1264,
+      "step": 852
+    },
+    {
+      "epoch": 0.03608596327946527,
+      "grad_norm": 0.48070141673088074,
+      "learning_rate": 0.001,
+      "loss": 2.6504,
+      "step": 853
+    },
+    {
+      "epoch": 0.03612826804298164,
+      "grad_norm": 0.5976713299751282,
+      "learning_rate": 0.001,
+      "loss": 3.0266,
+      "step": 854
+    },
+    {
+      "epoch": 0.03617057280649801,
+      "grad_norm": 0.3724801242351532,
+      "learning_rate": 0.001,
+      "loss": 3.0966,
+      "step": 855
+    },
+    {
+      "epoch": 0.03621287757001438,
+      "grad_norm": 0.35969069600105286,
+      "learning_rate": 0.001,
+      "loss": 3.1912,
+      "step": 856
+    },
+    {
+      "epoch": 0.036255182333530755,
+      "grad_norm": 4.334183692932129,
+      "learning_rate": 0.001,
+      "loss": 3.7079,
+      "step": 857
+    },
+    {
+      "epoch": 0.03629748709704713,
+      "grad_norm": 1.282253384590149,
+      "learning_rate": 0.001,
+      "loss": 2.2486,
+      "step": 858
+    },
+    {
+      "epoch": 0.0363397918605635,
+      "grad_norm": 0.44701820611953735,
+      "learning_rate": 0.001,
+      "loss": 2.8982,
+      "step": 859
+    },
+    {
+      "epoch": 0.036382096624079874,
+      "grad_norm": 0.37352272868156433,
+      "learning_rate": 0.001,
+      "loss": 3.1632,
+      "step": 860
+    },
+    {
+      "epoch": 0.03642440138759624,
+      "grad_norm": 0.33013972640037537,
+      "learning_rate": 0.001,
+      "loss": 2.0044,
+      "step": 861
+    },
+    {
+      "epoch": 0.03646670615111262,
+      "grad_norm": 0.37367743253707886,
+      "learning_rate": 0.001,
+      "loss": 3.3361,
+      "step": 862
+    },
+    {
+      "epoch": 0.036509010914628985,
+      "grad_norm": 0.5262033343315125,
+      "learning_rate": 0.001,
+      "loss": 4.0203,
+      "step": 863
+    },
+    {
+      "epoch": 0.03655131567814536,
+      "grad_norm": 0.8196743726730347,
+      "learning_rate": 0.001,
+      "loss": 1.9911,
+      "step": 864
+    },
+    {
+      "epoch": 0.03659362044166173,
+      "grad_norm": 1.1237256526947021,
+      "learning_rate": 0.001,
+      "loss": 3.1662,
+      "step": 865
+    },
+    {
+      "epoch": 0.0366359252051781,
+      "grad_norm": 2.1621487140655518,
+      "learning_rate": 0.001,
+      "loss": 1.6604,
+      "step": 866
+    },
+    {
+      "epoch": 0.03667822996869447,
+      "grad_norm": 0.5566956400871277,
+      "learning_rate": 0.001,
+      "loss": 2.1015,
+      "step": 867
+    },
+    {
+      "epoch": 0.03672053473221085,
+      "grad_norm": 0.35224634408950806,
+      "learning_rate": 0.001,
+      "loss": 2.6242,
+      "step": 868
+    },
+    {
+      "epoch": 0.03676283949572722,
+      "grad_norm": 0.33982589840888977,
+      "learning_rate": 0.001,
+      "loss": 2.6346,
+      "step": 869
+    },
+    {
+      "epoch": 0.03680514425924359,
+      "grad_norm": 0.9707451462745667,
+      "learning_rate": 0.001,
+      "loss": 2.7817,
+      "step": 870
+    },
+    {
+      "epoch": 0.036847449022759965,
+      "grad_norm": 0.32144203782081604,
+      "learning_rate": 0.001,
+      "loss": 1.9405,
+      "step": 871
+    },
+    {
+      "epoch": 0.03688975378627633,
+      "grad_norm": 0.31683671474456787,
+      "learning_rate": 0.001,
+      "loss": 2.9229,
+      "step": 872
+    },
+    {
+      "epoch": 0.03693205854979271,
+      "grad_norm": 0.36165180802345276,
+      "learning_rate": 0.001,
+      "loss": 3.1064,
+      "step": 873
+    },
+    {
+      "epoch": 0.036974363313309076,
+      "grad_norm": 0.311084508895874,
+      "learning_rate": 0.001,
+      "loss": 2.1846,
+      "step": 874
+    },
+    {
+      "epoch": 0.03701666807682545,
+      "grad_norm": 0.3572933077812195,
+      "learning_rate": 0.001,
+      "loss": 2.8808,
+      "step": 875
+    },
+    {
+      "epoch": 0.03705897284034182,
+      "grad_norm": 1.0521661043167114,
+      "learning_rate": 0.001,
+      "loss": 2.187,
+      "step": 876
+    },
+    {
+      "epoch": 0.037101277603858195,
+      "grad_norm": 0.41594183444976807,
+      "learning_rate": 0.001,
+      "loss": 2.4503,
+      "step": 877
+    },
+    {
+      "epoch": 0.03714358236737457,
+      "grad_norm": 1.126168131828308,
+      "learning_rate": 0.001,
+      "loss": 1.9875,
+      "step": 878
+    },
+    {
+      "epoch": 0.03718588713089094,
+      "grad_norm": 0.3451598882675171,
+      "learning_rate": 0.001,
+      "loss": 2.6444,
+      "step": 879
+    },
+    {
+      "epoch": 0.03722819189440731,
+      "grad_norm": 0.7697860598564148,
+      "learning_rate": 0.001,
+      "loss": 2.1816,
+      "step": 880
+    },
+    {
+      "epoch": 0.03727049665792368,
+      "grad_norm": 0.7067566514015198,
+      "learning_rate": 0.001,
+      "loss": 2.5282,
+      "step": 881
+    },
+    {
+      "epoch": 0.037312801421440056,
+      "grad_norm": 0.4576680064201355,
+      "learning_rate": 0.001,
+      "loss": 2.2394,
+      "step": 882
+    },
+    {
+      "epoch": 0.037355106184956424,
+      "grad_norm": 0.47805696725845337,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 883
+    },
+    {
+      "epoch": 0.0373974109484728,
+      "grad_norm": 18.085206985473633,
+      "learning_rate": 0.001,
+      "loss": 3.4886,
+      "step": 884
+    },
+    {
+      "epoch": 0.03743971571198917,
+      "grad_norm": 0.35168933868408203,
+      "learning_rate": 0.001,
+      "loss": 2.0473,
+      "step": 885
+    },
+    {
+      "epoch": 0.03748202047550554,
+      "grad_norm": 0.4354495108127594,
+      "learning_rate": 0.001,
+      "loss": 2.42,
+      "step": 886
+    },
+    {
+      "epoch": 0.03752432523902191,
+      "grad_norm": 0.4347662329673767,
+      "learning_rate": 0.001,
+      "loss": 2.7022,
+      "step": 887
+    },
+    {
+      "epoch": 0.037566630002538286,
+      "grad_norm": 2.830768346786499,
+      "learning_rate": 0.001,
+      "loss": 2.2454,
+      "step": 888
+    },
+    {
+      "epoch": 0.03760893476605466,
+      "grad_norm": 0.7625808119773865,
+      "learning_rate": 0.001,
+      "loss": 2.8905,
+      "step": 889
+    },
+    {
+      "epoch": 0.03765123952957103,
+      "grad_norm": 0.4779190719127655,
+      "learning_rate": 0.001,
+      "loss": 2.4054,
+      "step": 890
+    },
+    {
+      "epoch": 0.037693544293087404,
+      "grad_norm": 0.3955821692943573,
+      "learning_rate": 0.001,
+      "loss": 2.4408,
+      "step": 891
+    },
+    {
+      "epoch": 0.03773584905660377,
+      "grad_norm": 1.995564579963684,
+      "learning_rate": 0.001,
+      "loss": 1.9623,
+      "step": 892
+    },
+    {
+      "epoch": 0.03777815382012015,
+      "grad_norm": 1.702072024345398,
+      "learning_rate": 0.001,
+      "loss": 2.2569,
+      "step": 893
+    },
+    {
+      "epoch": 0.037820458583636515,
+      "grad_norm": 0.8360947370529175,
+      "learning_rate": 0.001,
+      "loss": 2.1785,
+      "step": 894
+    },
+    {
+      "epoch": 0.03786276334715289,
+      "grad_norm": 0.4631440043449402,
+      "learning_rate": 0.001,
+      "loss": 3.6477,
+      "step": 895
+    },
+    {
+      "epoch": 0.03790506811066926,
+      "grad_norm": 0.4883536100387573,
+      "learning_rate": 0.001,
+      "loss": 2.1303,
+      "step": 896
+    },
+    {
+      "epoch": 0.037947372874185634,
+      "grad_norm": 1.4146642684936523,
+      "learning_rate": 0.001,
+      "loss": 2.8125,
+      "step": 897
+    },
+    {
+      "epoch": 0.037989677637702,
+      "grad_norm": 33.48441696166992,
+      "learning_rate": 0.001,
+      "loss": 2.3088,
+      "step": 898
+    },
+    {
+      "epoch": 0.03803198240121838,
+      "grad_norm": 1.0009665489196777,
+      "learning_rate": 0.001,
+      "loss": 3.2976,
+      "step": 899
+    },
+    {
+      "epoch": 0.03807428716473475,
+      "grad_norm": 0.6075432896614075,
+      "learning_rate": 0.001,
+      "loss": 3.0491,
+      "step": 900
+    },
+    {
+      "epoch": 0.03811659192825112,
+      "grad_norm": 0.6014549136161804,
+      "learning_rate": 0.001,
+      "loss": 2.6115,
+      "step": 901
+    },
+    {
+      "epoch": 0.038158896691767495,
+      "grad_norm": 4.742966175079346,
+      "learning_rate": 0.001,
+      "loss": 3.1309,
+      "step": 902
+    },
+    {
+      "epoch": 0.03820120145528386,
+      "grad_norm": 1.2808791399002075,
+      "learning_rate": 0.001,
+      "loss": 2.7616,
+      "step": 903
+    },
+    {
+      "epoch": 0.03824350621880024,
+      "grad_norm": 0.6155859231948853,
+      "learning_rate": 0.001,
+      "loss": 3.0479,
+      "step": 904
+    },
+    {
+      "epoch": 0.038285810982316607,
+      "grad_norm": 0.4951868951320648,
+      "learning_rate": 0.001,
+      "loss": 2.4222,
+      "step": 905
+    },
+    {
+      "epoch": 0.03832811574583298,
+      "grad_norm": 3.6468491554260254,
+      "learning_rate": 0.001,
+      "loss": 2.7734,
+      "step": 906
+    },
+    {
+      "epoch": 0.03837042050934935,
+      "grad_norm": 0.6409814953804016,
+      "learning_rate": 0.001,
+      "loss": 2.7391,
+      "step": 907
+    },
+    {
+      "epoch": 0.038412725272865725,
+      "grad_norm": 1.037394642829895,
+      "learning_rate": 0.001,
+      "loss": 3.3225,
+      "step": 908
+    },
+    {
+      "epoch": 0.0384550300363821,
+      "grad_norm": 0.7472317218780518,
+      "learning_rate": 0.001,
+      "loss": 4.0019,
+      "step": 909
+    },
+    {
+      "epoch": 0.03849733479989847,
+      "grad_norm": 0.5227680802345276,
+      "learning_rate": 0.001,
+      "loss": 2.0158,
+      "step": 910
+    },
+    {
+      "epoch": 0.03853963956341484,
+      "grad_norm": 0.5070827603340149,
+      "learning_rate": 0.001,
+      "loss": 2.1145,
+      "step": 911
+    },
+    {
+      "epoch": 0.03858194432693121,
+      "grad_norm": 1.9357705116271973,
+      "learning_rate": 0.001,
+      "loss": 2.4728,
+      "step": 912
+    },
+    {
+      "epoch": 0.038624249090447586,
+      "grad_norm": 3.635646104812622,
+      "learning_rate": 0.001,
+      "loss": 2.7857,
+      "step": 913
+    },
+    {
+      "epoch": 0.038666553853963954,
+      "grad_norm": 0.9932013154029846,
+      "learning_rate": 0.001,
+      "loss": 2.1222,
+      "step": 914
+    },
+    {
+      "epoch": 0.03870885861748033,
+      "grad_norm": 1.2177910804748535,
+      "learning_rate": 0.001,
+      "loss": 2.5434,
+      "step": 915
+    },
+    {
+      "epoch": 0.0387511633809967,
+      "grad_norm": 0.6072919964790344,
+      "learning_rate": 0.001,
+      "loss": 2.2818,
+      "step": 916
+    },
+    {
+      "epoch": 0.03879346814451307,
+      "grad_norm": 1.0168566703796387,
+      "learning_rate": 0.001,
+      "loss": 3.5772,
+      "step": 917
+    },
+    {
+      "epoch": 0.03883577290802944,
+      "grad_norm": 12.94654655456543,
+      "learning_rate": 0.001,
+      "loss": 3.1531,
+      "step": 918
+    },
+    {
+      "epoch": 0.038878077671545816,
+      "grad_norm": 0.9608505964279175,
+      "learning_rate": 0.001,
+      "loss": 4.1271,
+      "step": 919
+    },
+    {
+      "epoch": 0.03892038243506219,
+      "grad_norm": 0.9271876215934753,
+      "learning_rate": 0.001,
+      "loss": 3.0042,
+      "step": 920
+    },
+    {
+      "epoch": 0.03896268719857856,
+      "grad_norm": 0.7582124471664429,
+      "learning_rate": 0.001,
+      "loss": 2.4802,
+      "step": 921
+    },
+    {
+      "epoch": 0.039004991962094934,
+      "grad_norm": 3.8157336711883545,
+      "learning_rate": 0.001,
+      "loss": 4.0794,
+      "step": 922
+    },
+    {
+      "epoch": 0.0390472967256113,
+      "grad_norm": 0.4885561764240265,
+      "learning_rate": 0.001,
+      "loss": 3.57,
+      "step": 923
+    },
+    {
+      "epoch": 0.03908960148912768,
+      "grad_norm": 0.5436772704124451,
+      "learning_rate": 0.001,
+      "loss": 2.4098,
+      "step": 924
+    },
+    {
+      "epoch": 0.039131906252644046,
+      "grad_norm": 0.5220960378646851,
+      "learning_rate": 0.001,
+      "loss": 2.5934,
+      "step": 925
+    },
+    {
+      "epoch": 0.03917421101616042,
+      "grad_norm": 0.4250609278678894,
+      "learning_rate": 0.001,
+      "loss": 2.1337,
+      "step": 926
+    },
+    {
+      "epoch": 0.03921651577967679,
+      "grad_norm": 0.45072197914123535,
+      "learning_rate": 0.001,
+      "loss": 2.2017,
+      "step": 927
+    },
+    {
+      "epoch": 0.039258820543193164,
+      "grad_norm": 0.5734637379646301,
+      "learning_rate": 0.001,
+      "loss": 3.231,
+      "step": 928
+    },
+    {
+      "epoch": 0.03930112530670953,
+      "grad_norm": 2.7498257160186768,
+      "learning_rate": 0.001,
+      "loss": 2.0577,
+      "step": 929
+    },
+    {
+      "epoch": 0.03934343007022591,
+      "grad_norm": 1.0663455724716187,
+      "learning_rate": 0.001,
+      "loss": 2.2796,
+      "step": 930
+    },
+    {
+      "epoch": 0.03938573483374228,
+      "grad_norm": 0.4455733895301819,
+      "learning_rate": 0.001,
+      "loss": 2.7131,
+      "step": 931
+    },
+    {
+      "epoch": 0.03942803959725865,
+      "grad_norm": 10.824166297912598,
+      "learning_rate": 0.001,
+      "loss": 3.0632,
+      "step": 932
+    },
+    {
+      "epoch": 0.039470344360775025,
+      "grad_norm": 0.9566563367843628,
+      "learning_rate": 0.001,
+      "loss": 2.4747,
+      "step": 933
+    },
+    {
+      "epoch": 0.039512649124291394,
+      "grad_norm": 0.694932222366333,
+      "learning_rate": 0.001,
+      "loss": 3.1625,
+      "step": 934
+    },
+    {
+      "epoch": 0.03955495388780777,
+      "grad_norm": 2.0650901794433594,
+      "learning_rate": 0.001,
+      "loss": 2.1686,
+      "step": 935
+    },
+    {
+      "epoch": 0.03959725865132414,
+      "grad_norm": 0.47802698612213135,
+      "learning_rate": 0.001,
+      "loss": 3.7587,
+      "step": 936
+    },
+    {
+      "epoch": 0.03963956341484051,
+      "grad_norm": 1.5605882406234741,
+      "learning_rate": 0.001,
+      "loss": 2.1477,
+      "step": 937
+    },
+    {
+      "epoch": 0.03968186817835688,
+      "grad_norm": 3.1594226360321045,
+      "learning_rate": 0.001,
+      "loss": 2.8316,
+      "step": 938
+    },
+    {
+      "epoch": 0.039724172941873255,
+      "grad_norm": 1.2251648902893066,
+      "learning_rate": 0.001,
+      "loss": 3.0162,
+      "step": 939
+    },
+    {
+      "epoch": 0.03976647770538963,
+      "grad_norm": 0.4009862542152405,
+      "learning_rate": 0.001,
+      "loss": 2.5269,
+      "step": 940
+    },
+    {
+      "epoch": 0.039808782468906,
+      "grad_norm": 0.426503449678421,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 941
+    },
+    {
+      "epoch": 0.03985108723242237,
+      "grad_norm": 0.5413312315940857,
+      "learning_rate": 0.001,
+      "loss": 2.9823,
+      "step": 942
+    },
+    {
+      "epoch": 0.03989339199593874,
+      "grad_norm": 0.4892127811908722,
+      "learning_rate": 0.001,
+      "loss": 1.9651,
+      "step": 943
+    },
+    {
+      "epoch": 0.03993569675945512,
+      "grad_norm": 0.47731220722198486,
+      "learning_rate": 0.001,
+      "loss": 1.7261,
+      "step": 944
+    },
+    {
+      "epoch": 0.039978001522971485,
+      "grad_norm": 0.44389474391937256,
+      "learning_rate": 0.001,
+      "loss": 2.0669,
+      "step": 945
+    },
+    {
+      "epoch": 0.04002030628648786,
+      "grad_norm": 0.6225860714912415,
+      "learning_rate": 0.001,
+      "loss": 2.501,
+      "step": 946
+    },
+    {
+      "epoch": 0.04006261105000423,
+      "grad_norm": 0.5539319515228271,
+      "learning_rate": 0.001,
+      "loss": 2.2267,
+      "step": 947
+    },
+    {
+      "epoch": 0.0401049158135206,
+      "grad_norm": 0.34385305643081665,
+      "learning_rate": 0.001,
+      "loss": 2.3557,
+      "step": 948
+    },
+    {
+      "epoch": 0.04014722057703697,
+      "grad_norm": 0.41491398215293884,
+      "learning_rate": 0.001,
+      "loss": 2.2824,
+      "step": 949
+    },
+    {
+      "epoch": 0.040189525340553346,
+      "grad_norm": 0.8437483906745911,
+      "learning_rate": 0.001,
+      "loss": 3.2014,
+      "step": 950
+    },
+    {
+      "epoch": 0.04023183010406972,
+      "grad_norm": 0.3619152307510376,
+      "learning_rate": 0.001,
+      "loss": 2.7257,
+      "step": 951
+    },
+    {
+      "epoch": 0.04027413486758609,
+      "grad_norm": 0.8494651317596436,
+      "learning_rate": 0.001,
+      "loss": 2.5558,
+      "step": 952
+    },
+    {
+      "epoch": 0.040316439631102464,
+      "grad_norm": 1.8171448707580566,
+      "learning_rate": 0.001,
+      "loss": 3.2215,
+      "step": 953
+    },
+    {
+      "epoch": 0.04035874439461883,
+      "grad_norm": 1.0728228092193604,
+      "learning_rate": 0.001,
+      "loss": 3.5815,
+      "step": 954
+    },
+    {
+      "epoch": 0.04040104915813521,
+      "grad_norm": 3.6907944679260254,
+      "learning_rate": 0.001,
+      "loss": 3.3357,
+      "step": 955
+    },
+    {
+      "epoch": 0.040443353921651576,
+      "grad_norm": 0.7766746282577515,
+      "learning_rate": 0.001,
+      "loss": 3.0304,
+      "step": 956
+    },
+    {
+      "epoch": 0.04048565868516795,
+      "grad_norm": 1.010837197303772,
+      "learning_rate": 0.001,
+      "loss": 2.4535,
+      "step": 957
+    },
+    {
+      "epoch": 0.04052796344868432,
+      "grad_norm": 0.6098260879516602,
+      "learning_rate": 0.001,
+      "loss": 3.0176,
+      "step": 958
+    },
+    {
+      "epoch": 0.040570268212200694,
+      "grad_norm": 1.3032490015029907,
+      "learning_rate": 0.001,
+      "loss": 2.7713,
+      "step": 959
+    },
+    {
+      "epoch": 0.04061257297571707,
+      "grad_norm": 6.024738788604736,
+      "learning_rate": 0.001,
+      "loss": 2.4266,
+      "step": 960
+    },
+    {
+      "epoch": 0.04065487773923344,
+      "grad_norm": 0.8786823749542236,
+      "learning_rate": 0.001,
+      "loss": 2.61,
+      "step": 961
+    },
+    {
+      "epoch": 0.04069718250274981,
+      "grad_norm": 0.5354873538017273,
+      "learning_rate": 0.001,
+      "loss": 2.3966,
+      "step": 962
+    },
+    {
+      "epoch": 0.04073948726626618,
+      "grad_norm": 0.8180838823318481,
+      "learning_rate": 0.001,
+      "loss": 3.0533,
+      "step": 963
+    },
+    {
+      "epoch": 0.040781792029782556,
+      "grad_norm": 0.6282289028167725,
+      "learning_rate": 0.001,
+      "loss": 3.1841,
+      "step": 964
+    },
+    {
+      "epoch": 0.040824096793298924,
+      "grad_norm": 0.783220648765564,
+      "learning_rate": 0.001,
+      "loss": 2.6245,
+      "step": 965
+    },
+    {
+      "epoch": 0.0408664015568153,
+      "grad_norm": 0.8866094946861267,
+      "learning_rate": 0.001,
+      "loss": 3.3324,
+      "step": 966
+    },
+    {
+      "epoch": 0.04090870632033167,
+      "grad_norm": 0.9526342153549194,
+      "learning_rate": 0.001,
+      "loss": 2.7561,
+      "step": 967
+    },
+    {
+      "epoch": 0.04095101108384804,
+      "grad_norm": 0.7338128685951233,
+      "learning_rate": 0.001,
+      "loss": 3.8083,
+      "step": 968
+    },
+    {
+      "epoch": 0.04099331584736441,
+      "grad_norm": 0.7096972465515137,
+      "learning_rate": 0.001,
+      "loss": 2.2787,
+      "step": 969
+    },
+    {
+      "epoch": 0.041035620610880785,
+      "grad_norm": 0.459326833486557,
+      "learning_rate": 0.001,
+      "loss": 2.6205,
+      "step": 970
+    },
+    {
+      "epoch": 0.04107792537439716,
+      "grad_norm": 0.5549113154411316,
+      "learning_rate": 0.001,
+      "loss": 3.9624,
+      "step": 971
+    },
+    {
+      "epoch": 0.04112023013791353,
+      "grad_norm": 0.5863713026046753,
+      "learning_rate": 0.001,
+      "loss": 2.6585,
+      "step": 972
+    },
+    {
+      "epoch": 0.041162534901429904,
+      "grad_norm": 0.435687780380249,
+      "learning_rate": 0.001,
+      "loss": 3.875,
+      "step": 973
+    },
+    {
+      "epoch": 0.04120483966494627,
+      "grad_norm": 0.43353310227394104,
+      "learning_rate": 0.001,
+      "loss": 3.5305,
+      "step": 974
+    },
+    {
+      "epoch": 0.04124714442846265,
+      "grad_norm": 0.3369213342666626,
+      "learning_rate": 0.001,
+      "loss": 3.0293,
+      "step": 975
+    },
+    {
+      "epoch": 0.041289449191979015,
+      "grad_norm": 0.37502485513687134,
+      "learning_rate": 0.001,
+      "loss": 2.6536,
+      "step": 976
+    },
+    {
+      "epoch": 0.04133175395549539,
+      "grad_norm": 3.277416229248047,
+      "learning_rate": 0.001,
+      "loss": 2.3418,
+      "step": 977
+    },
+    {
+      "epoch": 0.04137405871901176,
+      "grad_norm": 0.30327439308166504,
+      "learning_rate": 0.001,
+      "loss": 2.6483,
+      "step": 978
+    },
+    {
+      "epoch": 0.04141636348252813,
+      "grad_norm": 0.3673608601093292,
+      "learning_rate": 0.001,
+      "loss": 2.884,
+      "step": 979
+    },
+    {
+      "epoch": 0.0414586682460445,
+      "grad_norm": 0.49414071440696716,
+      "learning_rate": 0.001,
+      "loss": 2.3366,
+      "step": 980
+    },
+    {
+      "epoch": 0.041500973009560876,
+      "grad_norm": 2.8874402046203613,
+      "learning_rate": 0.001,
+      "loss": 2.0754,
+      "step": 981
+    },
+    {
+      "epoch": 0.04154327777307725,
+      "grad_norm": 0.5763828754425049,
+      "learning_rate": 0.001,
+      "loss": 3.3845,
+      "step": 982
+    },
+    {
+      "epoch": 0.04158558253659362,
+      "grad_norm": 0.5612279772758484,
+      "learning_rate": 0.001,
+      "loss": 2.315,
+      "step": 983
+    },
+    {
+      "epoch": 0.041627887300109995,
+      "grad_norm": 1.2401258945465088,
+      "learning_rate": 0.001,
+      "loss": 3.435,
+      "step": 984
+    },
+    {
+      "epoch": 0.04167019206362636,
+      "grad_norm": 0.9611271023750305,
+      "learning_rate": 0.001,
+      "loss": 2.7392,
+      "step": 985
+    },
+    {
+      "epoch": 0.04171249682714274,
+      "grad_norm": 0.4988934397697449,
+      "learning_rate": 0.001,
+      "loss": 2.2294,
+      "step": 986
+    },
+    {
+      "epoch": 0.041754801590659106,
+      "grad_norm": 1.0643762350082397,
+      "learning_rate": 0.001,
+      "loss": 2.4758,
+      "step": 987
+    },
+    {
+      "epoch": 0.04179710635417548,
+      "grad_norm": 0.3355838656425476,
+      "learning_rate": 0.001,
+      "loss": 2.1455,
+      "step": 988
+    },
+    {
+      "epoch": 0.04183941111769185,
+      "grad_norm": 17.996070861816406,
+      "learning_rate": 0.001,
+      "loss": 3.2964,
+      "step": 989
+    },
+    {
+      "epoch": 0.041881715881208224,
+      "grad_norm": 4.5193610191345215,
+      "learning_rate": 0.001,
+      "loss": 2.0008,
+      "step": 990
+    },
+    {
+      "epoch": 0.0419240206447246,
+      "grad_norm": 0.4277860224246979,
+      "learning_rate": 0.001,
+      "loss": 3.1608,
+      "step": 991
+    },
+    {
+      "epoch": 0.04196632540824097,
+      "grad_norm": 0.47893789410591125,
+      "learning_rate": 0.001,
+      "loss": 2.6076,
+      "step": 992
+    },
+    {
+      "epoch": 0.04200863017175734,
+      "grad_norm": 0.5489811301231384,
+      "learning_rate": 0.001,
+      "loss": 2.1761,
+      "step": 993
+    },
+    {
+      "epoch": 0.04205093493527371,
+      "grad_norm": 0.5169845819473267,
+      "learning_rate": 0.001,
+      "loss": 2.3756,
+      "step": 994
+    },
+    {
+      "epoch": 0.042093239698790086,
+      "grad_norm": 1.397851586341858,
+      "learning_rate": 0.001,
+      "loss": 3.6065,
+      "step": 995
+    },
+    {
+      "epoch": 0.042135544462306454,
+      "grad_norm": 0.43019843101501465,
+      "learning_rate": 0.001,
+      "loss": 2.4526,
+      "step": 996
+    },
+    {
+      "epoch": 0.04217784922582283,
+      "grad_norm": 0.45522528886795044,
+      "learning_rate": 0.001,
+      "loss": 1.9483,
+      "step": 997
+    },
+    {
+      "epoch": 0.0422201539893392,
+      "grad_norm": 0.405224472284317,
+      "learning_rate": 0.001,
+      "loss": 2.6085,
+      "step": 998
+    },
+    {
+      "epoch": 0.04226245875285557,
+      "grad_norm": 0.6403412818908691,
+      "learning_rate": 0.001,
+      "loss": 3.2959,
+      "step": 999
+    },
+    {
+      "epoch": 0.04230476351637194,
+      "grad_norm": 0.4327405095100403,
+      "learning_rate": 0.001,
+      "loss": 2.4858,
+      "step": 1000
+    },
+    {
+      "epoch": 0.042347068279888316,
+      "grad_norm": 1.4853562116622925,
+      "learning_rate": 0.001,
+      "loss": 2.0621,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04238937304340469,
+      "grad_norm": 0.42528292536735535,
+      "learning_rate": 0.001,
+      "loss": 2.4629,
+      "step": 1002
+    },
+    {
+      "epoch": 0.04243167780692106,
+      "grad_norm": 0.30595675110816956,
+      "learning_rate": 0.001,
+      "loss": 2.1818,
+      "step": 1003
+    },
+    {
+      "epoch": 0.042473982570437434,
+      "grad_norm": 0.4422406554222107,
+      "learning_rate": 0.001,
+      "loss": 1.949,
+      "step": 1004
+    },
+    {
+      "epoch": 0.0425162873339538,
+      "grad_norm": 0.3010110557079315,
+      "learning_rate": 0.001,
+      "loss": 2.3222,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04255859209747018,
+      "grad_norm": 12.802899360656738,
+      "learning_rate": 0.001,
+      "loss": 2.388,
+      "step": 1006
+    },
+    {
+      "epoch": 0.042600896860986545,
+      "grad_norm": 0.438760906457901,
+      "learning_rate": 0.001,
+      "loss": 2.7597,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04264320162450292,
+      "grad_norm": 0.9350125789642334,
+      "learning_rate": 0.001,
+      "loss": 2.3453,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04268550638801929,
+      "grad_norm": 0.4918479919433594,
+      "learning_rate": 0.001,
+      "loss": 2.1911,
+      "step": 1009
+    },
+    {
+      "epoch": 0.04272781115153566,
+      "grad_norm": 0.47994211316108704,
+      "learning_rate": 0.001,
+      "loss": 2.7534,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04277011591505203,
+      "grad_norm": 0.34078550338745117,
+      "learning_rate": 0.001,
+      "loss": 2.1767,
+      "step": 1011
+    },
+    {
+      "epoch": 0.04281242067856841,
+      "grad_norm": 0.6441370844841003,
+      "learning_rate": 0.001,
+      "loss": 2.1742,
+      "step": 1012
+    },
+    {
+      "epoch": 0.04285472544208478,
+      "grad_norm": 1.00253427028656,
+      "learning_rate": 0.001,
+      "loss": 2.6111,
+      "step": 1013
+    },
+    {
+      "epoch": 0.04289703020560115,
+      "grad_norm": 0.42675742506980896,
+      "learning_rate": 0.001,
+      "loss": 3.7831,
+      "step": 1014
+    },
+    {
+      "epoch": 0.042939334969117525,
+      "grad_norm": 0.5508003830909729,
+      "learning_rate": 0.001,
+      "loss": 2.6718,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04298163973263389,
+      "grad_norm": 0.5830764174461365,
+      "learning_rate": 0.001,
+      "loss": 3.2418,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04302394449615027,
+      "grad_norm": 0.5782317519187927,
+      "learning_rate": 0.001,
+      "loss": 2.5268,
+      "step": 1017
+    },
+    {
+      "epoch": 0.043066249259666636,
+      "grad_norm": 0.39523276686668396,
+      "learning_rate": 0.001,
+      "loss": 2.3105,
+      "step": 1018
+    },
+    {
+      "epoch": 0.04310855402318301,
+      "grad_norm": 0.5414249300956726,
+      "learning_rate": 0.001,
+      "loss": 2.5811,
+      "step": 1019
+    },
+    {
+      "epoch": 0.04315085878669938,
+      "grad_norm": 0.5326202511787415,
+      "learning_rate": 0.001,
+      "loss": 2.5664,
+      "step": 1020
+    },
+    {
+      "epoch": 0.043193163550215755,
+      "grad_norm": 0.3412122428417206,
+      "learning_rate": 0.001,
+      "loss": 2.6221,
+      "step": 1021
+    },
+    {
+      "epoch": 0.04323546831373213,
+      "grad_norm": 0.3936321437358856,
+      "learning_rate": 0.001,
+      "loss": 2.1314,
+      "step": 1022
+    },
+    {
+      "epoch": 0.0432777730772485,
+      "grad_norm": 1.0190694332122803,
+      "learning_rate": 0.001,
+      "loss": 2.6642,
+      "step": 1023
+    },
+    {
+      "epoch": 0.04332007784076487,
+      "grad_norm": 2.318169116973877,
+      "learning_rate": 0.001,
+      "loss": 3.3849,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04336238260428124,
+      "grad_norm": 0.29711923003196716,
+      "learning_rate": 0.001,
+      "loss": 2.2662,
+      "step": 1025
+    },
+    {
+      "epoch": 0.043404687367797616,
+      "grad_norm": 0.4354766011238098,
+      "learning_rate": 0.001,
+      "loss": 2.5792,
+      "step": 1026
+    },
+    {
+      "epoch": 0.043446992131313984,
+      "grad_norm": 0.3557169735431671,
+      "learning_rate": 0.001,
+      "loss": 3.1672,
+      "step": 1027
+    },
+    {
+      "epoch": 0.04348929689483036,
+      "grad_norm": 0.48224884271621704,
+      "learning_rate": 0.001,
+      "loss": 2.0798,
+      "step": 1028
+    },
+    {
+      "epoch": 0.04353160165834673,
+      "grad_norm": 0.3722802698612213,
+      "learning_rate": 0.001,
+      "loss": 2.6505,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0435739064218631,
+      "grad_norm": 0.40177568793296814,
+      "learning_rate": 0.001,
+      "loss": 2.6731,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04361621118537947,
+      "grad_norm": 0.34859174489974976,
+      "learning_rate": 0.001,
+      "loss": 2.0711,
+      "step": 1031
+    },
+    {
+      "epoch": 0.043658515948895846,
+      "grad_norm": 0.35582149028778076,
+      "learning_rate": 0.001,
+      "loss": 2.5389,
+      "step": 1032
+    },
+    {
+      "epoch": 0.04370082071241222,
+      "grad_norm": 0.35008880496025085,
+      "learning_rate": 0.001,
+      "loss": 2.0354,
+      "step": 1033
+    },
+    {
+      "epoch": 0.04374312547592859,
+      "grad_norm": 0.7123366594314575,
+      "learning_rate": 0.001,
+      "loss": 2.6122,
+      "step": 1034
+    },
+    {
+      "epoch": 0.043785430239444964,
+      "grad_norm": 0.6655288338661194,
+      "learning_rate": 0.001,
+      "loss": 2.3297,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04382773500296133,
+      "grad_norm": 2.140918493270874,
+      "learning_rate": 0.001,
+      "loss": 2.2363,
+      "step": 1036
+    },
+    {
+      "epoch": 0.04387003976647771,
+      "grad_norm": 1.8651494979858398,
+      "learning_rate": 0.001,
+      "loss": 3.1583,
+      "step": 1037
+    },
+    {
+      "epoch": 0.043912344529994075,
+      "grad_norm": 0.5514461994171143,
+      "learning_rate": 0.001,
+      "loss": 3.5841,
+      "step": 1038
+    },
+    {
+      "epoch": 0.04395464929351045,
+      "grad_norm": 0.48088452219963074,
+      "learning_rate": 0.001,
+      "loss": 2.3557,
+      "step": 1039
+    },
+    {
+      "epoch": 0.04399695405702682,
+      "grad_norm": 1.1052600145339966,
+      "learning_rate": 0.001,
+      "loss": 2.5092,
+      "step": 1040
+    },
+    {
+      "epoch": 0.044039258820543194,
+      "grad_norm": 0.4182952344417572,
+      "learning_rate": 0.001,
+      "loss": 3.016,
+      "step": 1041
+    },
+    {
+      "epoch": 0.04408156358405956,
+      "grad_norm": 0.30450159311294556,
+      "learning_rate": 0.001,
+      "loss": 2.5603,
+      "step": 1042
+    },
+    {
+      "epoch": 0.04412386834757594,
+      "grad_norm": 0.2754423916339874,
+      "learning_rate": 0.001,
+      "loss": 1.9932,
+      "step": 1043
+    },
+    {
+      "epoch": 0.04416617311109231,
+      "grad_norm": 2.453713893890381,
+      "learning_rate": 0.001,
+      "loss": 3.1293,
+      "step": 1044
+    },
+    {
+      "epoch": 0.04420847787460868,
+      "grad_norm": 0.606838583946228,
+      "learning_rate": 0.001,
+      "loss": 3.1959,
+      "step": 1045
+    },
+    {
+      "epoch": 0.044250782638125055,
+      "grad_norm": 0.580777645111084,
+      "learning_rate": 0.001,
+      "loss": 2.0586,
+      "step": 1046
+    },
+    {
+      "epoch": 0.04429308740164142,
+      "grad_norm": 5.384679317474365,
+      "learning_rate": 0.001,
+      "loss": 2.4947,
+      "step": 1047
+    },
+    {
+      "epoch": 0.0443353921651578,
+      "grad_norm": 0.3704220652580261,
+      "learning_rate": 0.001,
+      "loss": 2.4293,
+      "step": 1048
+    },
+    {
+      "epoch": 0.04437769692867417,
+      "grad_norm": 0.3065285086631775,
+      "learning_rate": 0.001,
+      "loss": 2.7632,
+      "step": 1049
+    },
+    {
+      "epoch": 0.04442000169219054,
+      "grad_norm": 0.3605315089225769,
+      "learning_rate": 0.001,
+      "loss": 2.3277,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04446230645570691,
+      "grad_norm": 0.46234673261642456,
+      "learning_rate": 0.001,
+      "loss": 2.1531,
+      "step": 1051
+    },
+    {
+      "epoch": 0.044504611219223285,
+      "grad_norm": 0.43919745087623596,
+      "learning_rate": 0.001,
+      "loss": 2.3706,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04454691598273966,
+      "grad_norm": 0.37155017256736755,
+      "learning_rate": 0.001,
+      "loss": 3.1104,
+      "step": 1053
+    },
+    {
+      "epoch": 0.04458922074625603,
+      "grad_norm": 0.31615880131721497,
+      "learning_rate": 0.001,
+      "loss": 3.1019,
+      "step": 1054
+    },
+    {
+      "epoch": 0.0446315255097724,
+      "grad_norm": 1.206450343132019,
+      "learning_rate": 0.001,
+      "loss": 3.4134,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04467383027328877,
+      "grad_norm": 0.6831628084182739,
+      "learning_rate": 0.001,
+      "loss": 2.5709,
+      "step": 1056
+    },
+    {
+      "epoch": 0.044716135036805146,
+      "grad_norm": 0.4207130968570709,
+      "learning_rate": 0.001,
+      "loss": 2.2903,
+      "step": 1057
+    },
+    {
+      "epoch": 0.044758439800321514,
+      "grad_norm": 0.3084408938884735,
+      "learning_rate": 0.001,
+      "loss": 2.8131,
+      "step": 1058
+    },
+    {
+      "epoch": 0.04480074456383789,
+      "grad_norm": 0.6155023574829102,
+      "learning_rate": 0.001,
+      "loss": 2.3744,
+      "step": 1059
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 0.6356950998306274,
+      "learning_rate": 0.001,
+      "loss": 2.5759,
+      "step": 1060
+    },
+    {
+      "epoch": 0.04488535409087063,
+      "grad_norm": 0.41177189350128174,
+      "learning_rate": 0.001,
+      "loss": 3.0073,
+      "step": 1061
+    },
+    {
+      "epoch": 0.044927658854387,
+      "grad_norm": 0.30051976442337036,
+      "learning_rate": 0.001,
+      "loss": 2.8997,
+      "step": 1062
+    },
+    {
+      "epoch": 0.044969963617903376,
+      "grad_norm": 0.28739258646965027,
+      "learning_rate": 0.001,
+      "loss": 1.8987,
+      "step": 1063
+    },
+    {
+      "epoch": 0.04501226838141975,
+      "grad_norm": 0.950710117816925,
+      "learning_rate": 0.001,
+      "loss": 2.3666,
+      "step": 1064
+    },
+    {
+      "epoch": 0.04505457314493612,
+      "grad_norm": 0.5854257941246033,
+      "learning_rate": 0.001,
+      "loss": 2.323,
+      "step": 1065
+    },
+    {
+      "epoch": 0.045096877908452494,
+      "grad_norm": 0.3358246088027954,
+      "learning_rate": 0.001,
+      "loss": 2.5187,
+      "step": 1066
+    },
+    {
+      "epoch": 0.04513918267196886,
+      "grad_norm": 1.6332861185073853,
+      "learning_rate": 0.001,
+      "loss": 2.3758,
+      "step": 1067
+    },
+    {
+      "epoch": 0.04518148743548524,
+      "grad_norm": 1.7713162899017334,
+      "learning_rate": 0.001,
+      "loss": 2.3417,
+      "step": 1068
+    },
+    {
+      "epoch": 0.045223792199001606,
+      "grad_norm": 0.3794574439525604,
+      "learning_rate": 0.001,
+      "loss": 2.0666,
+      "step": 1069
+    },
+    {
+      "epoch": 0.04526609696251798,
+      "grad_norm": 0.2956782281398773,
+      "learning_rate": 0.001,
+      "loss": 2.8441,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04530840172603435,
+      "grad_norm": 0.2897772192955017,
+      "learning_rate": 0.001,
+      "loss": 2.3111,
+      "step": 1071
+    },
+    {
+      "epoch": 0.045350706489550724,
+      "grad_norm": 4.40648078918457,
+      "learning_rate": 0.001,
+      "loss": 2.0011,
+      "step": 1072
+    },
+    {
+      "epoch": 0.04539301125306709,
+      "grad_norm": 7.493185520172119,
+      "learning_rate": 0.001,
+      "loss": 3.1528,
+      "step": 1073
+    },
+    {
+      "epoch": 0.04543531601658347,
+      "grad_norm": 0.34881994128227234,
+      "learning_rate": 0.001,
+      "loss": 2.0333,
+      "step": 1074
+    },
+    {
+      "epoch": 0.04547762078009984,
+      "grad_norm": 0.9052633047103882,
+      "learning_rate": 0.001,
+      "loss": 2.4499,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04551992554361621,
+      "grad_norm": 0.7866429090499878,
+      "learning_rate": 0.001,
+      "loss": 2.5423,
+      "step": 1076
+    },
+    {
+      "epoch": 0.045562230307132585,
+      "grad_norm": 0.8085727691650391,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 1077
+    },
+    {
+      "epoch": 0.045604535070648954,
+      "grad_norm": 0.517099142074585,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 1078
+    },
+    {
+      "epoch": 0.04564683983416533,
+      "grad_norm": 0.5327157378196716,
+      "learning_rate": 0.001,
+      "loss": 3.1626,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0456891445976817,
+      "grad_norm": 0.28011730313301086,
+      "learning_rate": 0.001,
+      "loss": 1.6698,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04573144936119807,
+      "grad_norm": 2.291832208633423,
+      "learning_rate": 0.001,
+      "loss": 2.2235,
+      "step": 1081
+    },
+    {
+      "epoch": 0.04577375412471444,
+      "grad_norm": 0.5037325024604797,
+      "learning_rate": 0.001,
+      "loss": 2.2568,
+      "step": 1082
+    },
+    {
+      "epoch": 0.045816058888230815,
+      "grad_norm": 0.7182853817939758,
+      "learning_rate": 0.001,
+      "loss": 2.6144,
+      "step": 1083
+    },
+    {
+      "epoch": 0.04585836365174719,
+      "grad_norm": 0.3494111895561218,
+      "learning_rate": 0.001,
+      "loss": 3.3147,
+      "step": 1084
+    },
+    {
+      "epoch": 0.04590066841526356,
+      "grad_norm": 0.35580259561538696,
+      "learning_rate": 0.001,
+      "loss": 3.1962,
+      "step": 1085
+    },
+    {
+      "epoch": 0.04594297317877993,
+      "grad_norm": 0.5036289691925049,
+      "learning_rate": 0.001,
+      "loss": 2.1262,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0459852779422963,
+      "grad_norm": 0.5229755640029907,
+      "learning_rate": 0.001,
+      "loss": 1.7954,
+      "step": 1087
+    },
+    {
+      "epoch": 0.04602758270581268,
+      "grad_norm": 0.5959165096282959,
+      "learning_rate": 0.001,
+      "loss": 3.3961,
+      "step": 1088
+    },
+    {
+      "epoch": 0.046069887469329045,
+      "grad_norm": 0.43716666102409363,
+      "learning_rate": 0.001,
+      "loss": 2.0928,
+      "step": 1089
+    },
+    {
+      "epoch": 0.04611219223284542,
+      "grad_norm": 0.38793283700942993,
+      "learning_rate": 0.001,
+      "loss": 2.1611,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04615449699636179,
+      "grad_norm": 0.3964957892894745,
+      "learning_rate": 0.001,
+      "loss": 2.4315,
+      "step": 1091
+    },
+    {
+      "epoch": 0.04619680175987816,
+      "grad_norm": 0.7259196043014526,
+      "learning_rate": 0.001,
+      "loss": 2.5655,
+      "step": 1092
+    },
+    {
+      "epoch": 0.04623910652339453,
+      "grad_norm": 20.704673767089844,
+      "learning_rate": 0.001,
+      "loss": 2.1808,
+      "step": 1093
+    },
+    {
+      "epoch": 0.046281411286910906,
+      "grad_norm": 1.8599133491516113,
+      "learning_rate": 0.001,
+      "loss": 3.0483,
+      "step": 1094
+    },
+    {
+      "epoch": 0.04632371605042728,
+      "grad_norm": 0.3958137035369873,
+      "learning_rate": 0.001,
+      "loss": 2.8665,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04636602081394365,
+      "grad_norm": 0.46354708075523376,
+      "learning_rate": 0.001,
+      "loss": 2.1897,
+      "step": 1096
+    },
+    {
+      "epoch": 0.046408325577460025,
+      "grad_norm": 0.5653846263885498,
+      "learning_rate": 0.001,
+      "loss": 2.2344,
+      "step": 1097
+    },
+    {
+      "epoch": 0.04645063034097639,
+      "grad_norm": 0.3511918783187866,
+      "learning_rate": 0.001,
+      "loss": 1.8777,
+      "step": 1098
+    },
+    {
+      "epoch": 0.04649293510449277,
+      "grad_norm": 0.4230622351169586,
+      "learning_rate": 0.001,
+      "loss": 2.2688,
+      "step": 1099
+    },
+    {
+      "epoch": 0.046535239868009136,
+      "grad_norm": 0.5649506449699402,
+      "learning_rate": 0.001,
+      "loss": 1.8348,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04657754463152551,
+      "grad_norm": 0.3206387162208557,
+      "learning_rate": 0.001,
+      "loss": 2.2039,
+      "step": 1101
+    },
+    {
+      "epoch": 0.04661984939504188,
+      "grad_norm": 0.3505328297615051,
+      "learning_rate": 0.001,
+      "loss": 2.8421,
+      "step": 1102
+    },
+    {
+      "epoch": 0.046662154158558254,
+      "grad_norm": 0.8784352540969849,
+      "learning_rate": 0.001,
+      "loss": 3.6394,
+      "step": 1103
+    },
+    {
+      "epoch": 0.04670445892207462,
+      "grad_norm": 0.3867740035057068,
+      "learning_rate": 0.001,
+      "loss": 2.2488,
+      "step": 1104
+    },
+    {
+      "epoch": 0.046746763685591,
+      "grad_norm": 0.3754059374332428,
+      "learning_rate": 0.001,
+      "loss": 2.4647,
+      "step": 1105
+    },
+    {
+      "epoch": 0.04678906844910737,
+      "grad_norm": 0.3429076671600342,
+      "learning_rate": 0.001,
+      "loss": 2.658,
+      "step": 1106
+    },
+    {
+      "epoch": 0.04683137321262374,
+      "grad_norm": 0.35917940735816956,
+      "learning_rate": 0.001,
+      "loss": 2.8441,
+      "step": 1107
+    },
+    {
+      "epoch": 0.046873677976140116,
+      "grad_norm": 0.4225330948829651,
+      "learning_rate": 0.001,
+      "loss": 3.4419,
+      "step": 1108
+    },
+    {
+      "epoch": 0.046915982739656484,
+      "grad_norm": 0.2866974174976349,
+      "learning_rate": 0.001,
+      "loss": 1.7171,
+      "step": 1109
+    },
+    {
+      "epoch": 0.04695828750317286,
+      "grad_norm": 0.655272364616394,
+      "learning_rate": 0.001,
+      "loss": 2.1575,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04700059226668923,
+      "grad_norm": 0.32389822602272034,
+      "learning_rate": 0.001,
+      "loss": 2.9878,
+      "step": 1111
+    },
+    {
+      "epoch": 0.0470428970302056,
+      "grad_norm": 0.33824819326400757,
+      "learning_rate": 0.001,
+      "loss": 2.6459,
+      "step": 1112
+    },
+    {
+      "epoch": 0.04708520179372197,
+      "grad_norm": 0.2966809570789337,
+      "learning_rate": 0.001,
+      "loss": 2.6308,
+      "step": 1113
+    },
+    {
+      "epoch": 0.047127506557238345,
+      "grad_norm": 0.3179897665977478,
+      "learning_rate": 0.001,
+      "loss": 2.525,
+      "step": 1114
+    },
+    {
+      "epoch": 0.04716981132075472,
+      "grad_norm": 1.4565528631210327,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 1115
+    },
+    {
+      "epoch": 0.04721211608427109,
+      "grad_norm": 0.4614794850349426,
+      "learning_rate": 0.001,
+      "loss": 3.2028,
+      "step": 1116
+    },
+    {
+      "epoch": 0.047254420847787464,
+      "grad_norm": 0.3581281900405884,
+      "learning_rate": 0.001,
+      "loss": 2.3384,
+      "step": 1117
+    },
+    {
+      "epoch": 0.04729672561130383,
+      "grad_norm": 0.3326326310634613,
+      "learning_rate": 0.001,
+      "loss": 4.0398,
+      "step": 1118
+    },
+    {
+      "epoch": 0.04733903037482021,
+      "grad_norm": 0.3846263289451599,
+      "learning_rate": 0.001,
+      "loss": 2.437,
+      "step": 1119
+    },
+    {
+      "epoch": 0.047381335138336575,
+      "grad_norm": 0.3471648395061493,
+      "learning_rate": 0.001,
+      "loss": 2.8839,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04742363990185295,
+      "grad_norm": 0.8793920874595642,
+      "learning_rate": 0.001,
+      "loss": 2.2851,
+      "step": 1121
+    },
+    {
+      "epoch": 0.04746594466536932,
+      "grad_norm": 0.4244942367076874,
+      "learning_rate": 0.001,
+      "loss": 2.1305,
+      "step": 1122
+    },
+    {
+      "epoch": 0.04750824942888569,
+      "grad_norm": 0.272441565990448,
+      "learning_rate": 0.001,
+      "loss": 2.3609,
+      "step": 1123
+    },
+    {
+      "epoch": 0.04755055419240206,
+      "grad_norm": 0.5901132225990295,
+      "learning_rate": 0.001,
+      "loss": 2.5462,
+      "step": 1124
+    },
+    {
+      "epoch": 0.047592858955918436,
+      "grad_norm": 1.15731680393219,
+      "learning_rate": 0.001,
+      "loss": 2.0628,
+      "step": 1125
+    },
+    {
+      "epoch": 0.04763516371943481,
+      "grad_norm": 0.35768887400627136,
+      "learning_rate": 0.001,
+      "loss": 2.415,
+      "step": 1126
+    },
+    {
+      "epoch": 0.04767746848295118,
+      "grad_norm": 0.5913015604019165,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 1127
+    },
+    {
+      "epoch": 0.047719773246467555,
+      "grad_norm": 0.7240321040153503,
+      "learning_rate": 0.001,
+      "loss": 2.7932,
+      "step": 1128
+    },
+    {
+      "epoch": 0.04776207800998392,
+      "grad_norm": 0.29713061451911926,
+      "learning_rate": 0.001,
+      "loss": 1.7194,
+      "step": 1129
+    },
+    {
+      "epoch": 0.0478043827735003,
+      "grad_norm": 0.30949392914772034,
+      "learning_rate": 0.001,
+      "loss": 2.3119,
+      "step": 1130
+    },
+    {
+      "epoch": 0.047846687537016666,
+      "grad_norm": 0.7915238738059998,
+      "learning_rate": 0.001,
+      "loss": 2.2115,
+      "step": 1131
+    },
+    {
+      "epoch": 0.04788899230053304,
+      "grad_norm": 0.33516794443130493,
+      "learning_rate": 0.001,
+      "loss": 2.4067,
+      "step": 1132
+    },
+    {
+      "epoch": 0.04793129706404941,
+      "grad_norm": 0.4222777783870697,
+      "learning_rate": 0.001,
+      "loss": 2.3102,
+      "step": 1133
+    },
+    {
+      "epoch": 0.047973601827565784,
+      "grad_norm": 0.3512939512729645,
+      "learning_rate": 0.001,
+      "loss": 4.4249,
+      "step": 1134
+    },
+    {
+      "epoch": 0.04801590659108215,
+      "grad_norm": 0.7207156419754028,
+      "learning_rate": 0.001,
+      "loss": 2.2423,
+      "step": 1135
+    },
+    {
+      "epoch": 0.04805821135459853,
+      "grad_norm": 0.3659185767173767,
+      "learning_rate": 0.001,
+      "loss": 2.3806,
+      "step": 1136
+    },
+    {
+      "epoch": 0.0481005161181149,
+      "grad_norm": 0.30134308338165283,
+      "learning_rate": 0.001,
+      "loss": 2.3715,
+      "step": 1137
+    },
+    {
+      "epoch": 0.04814282088163127,
+      "grad_norm": 0.5804815888404846,
+      "learning_rate": 0.001,
+      "loss": 4.0196,
+      "step": 1138
+    },
+    {
+      "epoch": 0.048185125645147646,
+      "grad_norm": 0.2848808169364929,
+      "learning_rate": 0.001,
+      "loss": 2.0329,
+      "step": 1139
+    },
+    {
+      "epoch": 0.048227430408664014,
+      "grad_norm": 0.5376827716827393,
+      "learning_rate": 0.001,
+      "loss": 2.5122,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04826973517218039,
+      "grad_norm": 0.393577516078949,
+      "learning_rate": 0.001,
+      "loss": 3.0434,
+      "step": 1141
+    },
+    {
+      "epoch": 0.04831203993569676,
+      "grad_norm": 1.210608959197998,
+      "learning_rate": 0.001,
+      "loss": 2.1217,
+      "step": 1142
+    },
+    {
+      "epoch": 0.04835434469921313,
+      "grad_norm": 0.9917371273040771,
+      "learning_rate": 0.001,
+      "loss": 4.2548,
+      "step": 1143
+    },
+    {
+      "epoch": 0.0483966494627295,
+      "grad_norm": 4.173092365264893,
+      "learning_rate": 0.001,
+      "loss": 2.1519,
+      "step": 1144
+    },
+    {
+      "epoch": 0.048438954226245876,
+      "grad_norm": 0.4106460213661194,
+      "learning_rate": 0.001,
+      "loss": 3.6214,
+      "step": 1145
+    },
+    {
+      "epoch": 0.04848125898976225,
+      "grad_norm": 1.5029926300048828,
+      "learning_rate": 0.001,
+      "loss": 3.2724,
+      "step": 1146
+    },
+    {
+      "epoch": 0.04852356375327862,
+      "grad_norm": 0.5879022479057312,
+      "learning_rate": 0.001,
+      "loss": 1.8398,
+      "step": 1147
+    },
+    {
+      "epoch": 0.048565868516794994,
+      "grad_norm": 0.36776503920555115,
+      "learning_rate": 0.001,
+      "loss": 2.2025,
+      "step": 1148
+    },
+    {
+      "epoch": 0.04860817328031136,
+      "grad_norm": 0.9697319269180298,
+      "learning_rate": 0.001,
+      "loss": 2.4534,
+      "step": 1149
+    },
+    {
+      "epoch": 0.04865047804382774,
+      "grad_norm": 0.640369176864624,
+      "learning_rate": 0.001,
+      "loss": 4.1821,
+      "step": 1150
+    },
+    {
+      "epoch": 0.048692782807344105,
+      "grad_norm": 0.5271949768066406,
+      "learning_rate": 0.001,
+      "loss": 2.5097,
+      "step": 1151
+    },
+    {
+      "epoch": 0.04873508757086048,
+      "grad_norm": 0.498152494430542,
+      "learning_rate": 0.001,
+      "loss": 2.097,
+      "step": 1152
+    },
+    {
+      "epoch": 0.04877739233437685,
+      "grad_norm": 0.45384323596954346,
+      "learning_rate": 0.001,
+      "loss": 2.8774,
+      "step": 1153
+    },
+    {
+      "epoch": 0.048819697097893223,
+      "grad_norm": 0.6459994912147522,
+      "learning_rate": 0.001,
+      "loss": 2.267,
+      "step": 1154
+    },
+    {
+      "epoch": 0.04886200186140959,
+      "grad_norm": 0.8564043045043945,
+      "learning_rate": 0.001,
+      "loss": 2.7848,
+      "step": 1155
+    },
+    {
+      "epoch": 0.04890430662492597,
+      "grad_norm": 0.33357423543930054,
+      "learning_rate": 0.001,
+      "loss": 2.5382,
+      "step": 1156
+    },
+    {
+      "epoch": 0.04894661138844234,
+      "grad_norm": 1.739807367324829,
+      "learning_rate": 0.001,
+      "loss": 2.3483,
+      "step": 1157
+    },
+    {
+      "epoch": 0.04898891615195871,
+      "grad_norm": 0.6629720330238342,
+      "learning_rate": 0.001,
+      "loss": 2.345,
+      "step": 1158
+    },
+    {
+      "epoch": 0.049031220915475085,
+      "grad_norm": 0.42855751514434814,
+      "learning_rate": 0.001,
+      "loss": 2.7333,
+      "step": 1159
+    },
+    {
+      "epoch": 0.04907352567899145,
+      "grad_norm": 0.5585654973983765,
+      "learning_rate": 0.001,
+      "loss": 2.7187,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04911583044250783,
+      "grad_norm": 1.3134695291519165,
+      "learning_rate": 0.001,
+      "loss": 2.2319,
+      "step": 1161
+    },
+    {
+      "epoch": 0.049158135206024196,
+      "grad_norm": 16.80616569519043,
+      "learning_rate": 0.001,
+      "loss": 3.13,
+      "step": 1162
+    },
+    {
+      "epoch": 0.04920043996954057,
+      "grad_norm": 0.4731372594833374,
+      "learning_rate": 0.001,
+      "loss": 2.9317,
+      "step": 1163
+    },
+    {
+      "epoch": 0.04924274473305694,
+      "grad_norm": 0.7382914423942566,
+      "learning_rate": 0.001,
+      "loss": 3.251,
+      "step": 1164
+    },
+    {
+      "epoch": 0.049285049496573315,
+      "grad_norm": 1.2850393056869507,
+      "learning_rate": 0.001,
+      "loss": 2.4139,
+      "step": 1165
+    },
+    {
+      "epoch": 0.04932735426008968,
+      "grad_norm": 0.45804908871650696,
+      "learning_rate": 0.001,
+      "loss": 2.1216,
+      "step": 1166
+    },
+    {
+      "epoch": 0.04936965902360606,
+      "grad_norm": 0.41841673851013184,
+      "learning_rate": 0.001,
+      "loss": 2.3593,
+      "step": 1167
+    },
+    {
+      "epoch": 0.04941196378712243,
+      "grad_norm": 0.3579320013523102,
+      "learning_rate": 0.001,
+      "loss": 2.733,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0494542685506388,
+      "grad_norm": 0.460112988948822,
+      "learning_rate": 0.001,
+      "loss": 1.9814,
+      "step": 1169
+    },
+    {
+      "epoch": 0.049496573314155176,
+      "grad_norm": 0.6798352599143982,
+      "learning_rate": 0.001,
+      "loss": 2.1811,
+      "step": 1170
+    },
+    {
+      "epoch": 0.049538878077671544,
+      "grad_norm": 13.641300201416016,
+      "learning_rate": 0.001,
+      "loss": 2.4548,
+      "step": 1171
+    },
+    {
+      "epoch": 0.04958118284118792,
+      "grad_norm": 1.915924310684204,
+      "learning_rate": 0.001,
+      "loss": 2.3664,
+      "step": 1172
+    },
+    {
+      "epoch": 0.04962348760470429,
+      "grad_norm": 0.34269511699676514,
+      "learning_rate": 0.001,
+      "loss": 2.571,
+      "step": 1173
+    },
+    {
+      "epoch": 0.04966579236822066,
+      "grad_norm": 1.606256127357483,
+      "learning_rate": 0.001,
+      "loss": 1.7933,
+      "step": 1174
+    },
+    {
+      "epoch": 0.04970809713173703,
+      "grad_norm": 0.6439388394355774,
+      "learning_rate": 0.001,
+      "loss": 2.7446,
+      "step": 1175
+    },
+    {
+      "epoch": 0.049750401895253406,
+      "grad_norm": 0.554309606552124,
+      "learning_rate": 0.001,
+      "loss": 2.9953,
+      "step": 1176
+    },
+    {
+      "epoch": 0.04979270665876978,
+      "grad_norm": 0.36114075779914856,
+      "learning_rate": 0.001,
+      "loss": 2.1207,
+      "step": 1177
+    },
+    {
+      "epoch": 0.04983501142228615,
+      "grad_norm": 0.36802929639816284,
+      "learning_rate": 0.001,
+      "loss": 2.7589,
+      "step": 1178
+    },
+    {
+      "epoch": 0.049877316185802524,
+      "grad_norm": 0.38982877135276794,
+      "learning_rate": 0.001,
+      "loss": 2.7034,
+      "step": 1179
+    },
+    {
+      "epoch": 0.04991962094931889,
+      "grad_norm": 0.46610337495803833,
+      "learning_rate": 0.001,
+      "loss": 2.2733,
+      "step": 1180
+    },
+    {
+      "epoch": 0.04996192571283527,
+      "grad_norm": 1.6383532285690308,
+      "learning_rate": 0.001,
+      "loss": 2.6442,
+      "step": 1181
+    },
+    {
+      "epoch": 0.050004230476351635,
+      "grad_norm": 0.3796882927417755,
+      "learning_rate": 0.001,
+      "loss": 2.8773,
+      "step": 1182
+    },
+    {
+      "epoch": 0.05004653523986801,
+      "grad_norm": 0.3811698257923126,
+      "learning_rate": 0.001,
+      "loss": 2.2252,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05008884000338438,
+      "grad_norm": 0.6414541602134705,
+      "learning_rate": 0.001,
+      "loss": 2.8175,
+      "step": 1184
+    },
+    {
+      "epoch": 0.050131144766900754,
+      "grad_norm": 0.8355805277824402,
+      "learning_rate": 0.001,
+      "loss": 2.3944,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05017344953041712,
+      "grad_norm": 1.5542443990707397,
+      "learning_rate": 0.001,
+      "loss": 2.544,
+      "step": 1186
+    },
+    {
+      "epoch": 0.0502157542939335,
+      "grad_norm": 1.7193444967269897,
+      "learning_rate": 0.001,
+      "loss": 2.0804,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05025805905744987,
+      "grad_norm": 1.0862737894058228,
+      "learning_rate": 0.001,
+      "loss": 2.5055,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05030036382096624,
+      "grad_norm": 0.4850480258464813,
+      "learning_rate": 0.001,
+      "loss": 3.1904,
+      "step": 1189
+    },
+    {
+      "epoch": 0.050342668584482615,
+      "grad_norm": 0.5566414594650269,
+      "learning_rate": 0.001,
+      "loss": 1.9222,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05038497334799898,
+      "grad_norm": 7.770880699157715,
+      "learning_rate": 0.001,
+      "loss": 3.1213,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05042727811151536,
+      "grad_norm": 0.9301654100418091,
+      "learning_rate": 0.001,
+      "loss": 2.1625,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05046958287503173,
+      "grad_norm": 0.28750085830688477,
+      "learning_rate": 0.001,
+      "loss": 2.0768,
+      "step": 1193
+    },
+    {
+      "epoch": 0.0505118876385481,
+      "grad_norm": 0.33357667922973633,
+      "learning_rate": 0.001,
+      "loss": 1.2852,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05055419240206447,
+      "grad_norm": 0.4574757218360901,
+      "learning_rate": 0.001,
+      "loss": 1.9372,
+      "step": 1195
+    },
+    {
+      "epoch": 0.050596497165580845,
+      "grad_norm": 2.3129217624664307,
+      "learning_rate": 0.001,
+      "loss": 3.4428,
+      "step": 1196
+    },
+    {
+      "epoch": 0.05063880192909721,
+      "grad_norm": 11.059386253356934,
+      "learning_rate": 0.001,
+      "loss": 2.063,
+      "step": 1197
+    },
+    {
+      "epoch": 0.05068110669261359,
+      "grad_norm": 0.5615996718406677,
+      "learning_rate": 0.001,
+      "loss": 2.9518,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05072341145612996,
+      "grad_norm": 1.5756652355194092,
+      "learning_rate": 0.001,
+      "loss": 2.7033,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05076571621964633,
+      "grad_norm": 0.5411480069160461,
+      "learning_rate": 0.001,
+      "loss": 2.3017,
+      "step": 1200
+    },
+    {
+      "epoch": 0.050808020983162706,
+      "grad_norm": 0.41354984045028687,
+      "learning_rate": 0.001,
+      "loss": 3.1848,
+      "step": 1201
+    },
+    {
+      "epoch": 0.050850325746679075,
+      "grad_norm": 1.6077523231506348,
+      "learning_rate": 0.001,
+      "loss": 3.4731,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05089263051019545,
+      "grad_norm": 0.8017259240150452,
+      "learning_rate": 0.001,
+      "loss": 1.9944,
+      "step": 1203
+    },
+    {
+      "epoch": 0.05093493527371182,
+      "grad_norm": 1.3641940355300903,
+      "learning_rate": 0.001,
+      "loss": 3.2556,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05097724003722819,
+      "grad_norm": 0.9652554392814636,
+      "learning_rate": 0.001,
+      "loss": 3.2552,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05101954480074456,
+      "grad_norm": 0.44206491112709045,
+      "learning_rate": 0.001,
+      "loss": 2.1191,
+      "step": 1206
+    },
+    {
+      "epoch": 0.051061849564260936,
+      "grad_norm": 0.7324578762054443,
+      "learning_rate": 0.001,
+      "loss": 3.5191,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05110415432777731,
+      "grad_norm": 1.9216930866241455,
+      "learning_rate": 0.001,
+      "loss": 3.5364,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05114645909129368,
+      "grad_norm": 0.8697239756584167,
+      "learning_rate": 0.001,
+      "loss": 2.362,
+      "step": 1209
+    },
+    {
+      "epoch": 0.051188763854810054,
+      "grad_norm": 0.6998451948165894,
+      "learning_rate": 0.001,
+      "loss": 2.7623,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05123106861832642,
+      "grad_norm": 0.9253522753715515,
+      "learning_rate": 0.001,
+      "loss": 3.4004,
+      "step": 1211
+    },
+    {
+      "epoch": 0.0512733733818428,
+      "grad_norm": 0.9361282587051392,
+      "learning_rate": 0.001,
+      "loss": 3.1784,
+      "step": 1212
+    },
+    {
+      "epoch": 0.051315678145359166,
+      "grad_norm": 0.38276004791259766,
+      "learning_rate": 0.001,
+      "loss": 2.448,
+      "step": 1213
+    },
+    {
+      "epoch": 0.05135798290887554,
+      "grad_norm": 0.5177932977676392,
+      "learning_rate": 0.001,
+      "loss": 3.3615,
+      "step": 1214
+    },
+    {
+      "epoch": 0.05140028767239191,
+      "grad_norm": 1.2303301095962524,
+      "learning_rate": 0.001,
+      "loss": 2.7344,
+      "step": 1215
+    },
+    {
+      "epoch": 0.051442592435908284,
+      "grad_norm": 25.516271591186523,
+      "learning_rate": 0.001,
+      "loss": 2.5343,
+      "step": 1216
+    },
+    {
+      "epoch": 0.05148489719942465,
+      "grad_norm": 0.7592980861663818,
+      "learning_rate": 0.001,
+      "loss": 3.7762,
+      "step": 1217
+    },
+    {
+      "epoch": 0.05152720196294103,
+      "grad_norm": 0.9871262311935425,
+      "learning_rate": 0.001,
+      "loss": 2.7732,
+      "step": 1218
+    },
+    {
+      "epoch": 0.0515695067264574,
+      "grad_norm": 0.6601408123970032,
+      "learning_rate": 0.001,
+      "loss": 2.7274,
+      "step": 1219
+    },
+    {
+      "epoch": 0.05161181148997377,
+      "grad_norm": 0.8926149606704712,
+      "learning_rate": 0.001,
+      "loss": 3.5338,
+      "step": 1220
+    },
+    {
+      "epoch": 0.051654116253490145,
+      "grad_norm": 0.6787442564964294,
+      "learning_rate": 0.001,
+      "loss": 2.4065,
+      "step": 1221
+    },
+    {
+      "epoch": 0.051696421017006514,
+      "grad_norm": 0.8640452027320862,
+      "learning_rate": 0.001,
+      "loss": 2.5657,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05173872578052289,
+      "grad_norm": 1.1135860681533813,
+      "learning_rate": 0.001,
+      "loss": 2.0519,
+      "step": 1223
+    },
+    {
+      "epoch": 0.05178103054403926,
+      "grad_norm": 0.481350302696228,
+      "learning_rate": 0.001,
+      "loss": 2.7644,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05182333530755563,
+      "grad_norm": 0.4704929292201996,
+      "learning_rate": 0.001,
+      "loss": 2.403,
+      "step": 1225
+    },
+    {
+      "epoch": 0.051865640071072,
+      "grad_norm": 6.178280353546143,
+      "learning_rate": 0.001,
+      "loss": 2.2048,
+      "step": 1226
+    },
+    {
+      "epoch": 0.051907944834588375,
+      "grad_norm": 0.5791310667991638,
+      "learning_rate": 0.001,
+      "loss": 3.0232,
+      "step": 1227
+    },
+    {
+      "epoch": 0.05195024959810474,
+      "grad_norm": 2.9677908420562744,
+      "learning_rate": 0.001,
+      "loss": 3.8193,
+      "step": 1228
+    },
+    {
+      "epoch": 0.05199255436162112,
+      "grad_norm": 1.5064809322357178,
+      "learning_rate": 0.001,
+      "loss": 3.0974,
+      "step": 1229
+    },
+    {
+      "epoch": 0.05203485912513749,
+      "grad_norm": 0.3698960542678833,
+      "learning_rate": 0.001,
+      "loss": 3.0401,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05207716388865386,
+      "grad_norm": 0.5180457830429077,
+      "learning_rate": 0.001,
+      "loss": 3.2249,
+      "step": 1231
+    },
+    {
+      "epoch": 0.05211946865217024,
+      "grad_norm": 0.3810598850250244,
+      "learning_rate": 0.001,
+      "loss": 2.5986,
+      "step": 1232
+    },
+    {
+      "epoch": 0.052161773415686605,
+      "grad_norm": 0.33539965748786926,
+      "learning_rate": 0.001,
+      "loss": 3.7299,
+      "step": 1233
+    },
+    {
+      "epoch": 0.05220407817920298,
+      "grad_norm": 0.5705146789550781,
+      "learning_rate": 0.001,
+      "loss": 3.6431,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05224638294271935,
+      "grad_norm": 0.4359552562236786,
+      "learning_rate": 0.001,
+      "loss": 3.0,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05228868770623572,
+      "grad_norm": 0.9745590686798096,
+      "learning_rate": 0.001,
+      "loss": 2.4393,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05233099246975209,
+      "grad_norm": 0.30867111682891846,
+      "learning_rate": 0.001,
+      "loss": 2.227,
+      "step": 1237
+    },
+    {
+      "epoch": 0.052373297233268466,
+      "grad_norm": 0.335039883852005,
+      "learning_rate": 0.001,
+      "loss": 2.2415,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05241560199678484,
+      "grad_norm": 0.6804436445236206,
+      "learning_rate": 0.001,
+      "loss": 2.7192,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05245790676030121,
+      "grad_norm": 0.41609200835227966,
+      "learning_rate": 0.001,
+      "loss": 3.3984,
+      "step": 1240
+    },
+    {
+      "epoch": 0.052500211523817585,
+      "grad_norm": 0.3207823932170868,
+      "learning_rate": 0.001,
+      "loss": 2.5688,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05254251628733395,
+      "grad_norm": 0.33397310972213745,
+      "learning_rate": 0.001,
+      "loss": 2.5408,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05258482105085033,
+      "grad_norm": 0.38078972697257996,
+      "learning_rate": 0.001,
+      "loss": 2.1729,
+      "step": 1243
+    },
+    {
+      "epoch": 0.052627125814366696,
+      "grad_norm": 0.34104421734809875,
+      "learning_rate": 0.001,
+      "loss": 3.0408,
+      "step": 1244
+    },
+    {
+      "epoch": 0.05266943057788307,
+      "grad_norm": 0.32771986722946167,
+      "learning_rate": 0.001,
+      "loss": 2.7228,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05271173534139944,
+      "grad_norm": 0.8869002461433411,
+      "learning_rate": 0.001,
+      "loss": 3.8551,
+      "step": 1246
+    },
+    {
+      "epoch": 0.052754040104915814,
+      "grad_norm": 0.2952006459236145,
+      "learning_rate": 0.001,
+      "loss": 2.3158,
+      "step": 1247
+    },
+    {
+      "epoch": 0.05279634486843218,
+      "grad_norm": 0.8608886003494263,
+      "learning_rate": 0.001,
+      "loss": 2.2941,
+      "step": 1248
+    },
+    {
+      "epoch": 0.05283864963194856,
+      "grad_norm": 0.5748282074928284,
+      "learning_rate": 0.001,
+      "loss": 2.9898,
+      "step": 1249
+    },
+    {
+      "epoch": 0.05288095439546493,
+      "grad_norm": 0.6410783529281616,
+      "learning_rate": 0.001,
+      "loss": 1.971,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0529232591589813,
+      "grad_norm": 1.4829449653625488,
+      "learning_rate": 0.001,
+      "loss": 3.4999,
+      "step": 1251
+    },
+    {
+      "epoch": 0.052965563922497676,
+      "grad_norm": 1.1416040658950806,
+      "learning_rate": 0.001,
+      "loss": 2.4466,
+      "step": 1252
+    },
+    {
+      "epoch": 0.053007868686014044,
+      "grad_norm": 1.1560173034667969,
+      "learning_rate": 0.001,
+      "loss": 2.3162,
+      "step": 1253
+    },
+    {
+      "epoch": 0.05305017344953042,
+      "grad_norm": 1.0749090909957886,
+      "learning_rate": 0.001,
+      "loss": 3.186,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05309247821304679,
+      "grad_norm": 0.3261003792285919,
+      "learning_rate": 0.001,
+      "loss": 2.3392,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05313478297656316,
+      "grad_norm": 0.4178333282470703,
+      "learning_rate": 0.001,
+      "loss": 1.8991,
+      "step": 1256
+    },
+    {
+      "epoch": 0.05317708774007953,
+      "grad_norm": 0.3390306234359741,
+      "learning_rate": 0.001,
+      "loss": 2.6713,
+      "step": 1257
+    },
+    {
+      "epoch": 0.053219392503595905,
+      "grad_norm": 0.46924909949302673,
+      "learning_rate": 0.001,
+      "loss": 3.4529,
+      "step": 1258
+    },
+    {
+      "epoch": 0.053261697267112273,
+      "grad_norm": 0.45819172263145447,
+      "learning_rate": 0.001,
+      "loss": 2.3935,
+      "step": 1259
+    },
+    {
+      "epoch": 0.05330400203062865,
+      "grad_norm": 0.36325275897979736,
+      "learning_rate": 0.001,
+      "loss": 1.8038,
+      "step": 1260
+    },
+    {
+      "epoch": 0.053346306794145024,
+      "grad_norm": 0.4765758216381073,
+      "learning_rate": 0.001,
+      "loss": 2.941,
+      "step": 1261
+    },
+    {
+      "epoch": 0.05338861155766139,
+      "grad_norm": 0.5178735256195068,
+      "learning_rate": 0.001,
+      "loss": 2.8948,
+      "step": 1262
+    },
+    {
+      "epoch": 0.05343091632117777,
+      "grad_norm": 0.3561374545097351,
+      "learning_rate": 0.001,
+      "loss": 3.3945,
+      "step": 1263
+    },
+    {
+      "epoch": 0.053473221084694135,
+      "grad_norm": 0.29234689474105835,
+      "learning_rate": 0.001,
+      "loss": 1.8964,
+      "step": 1264
+    },
+    {
+      "epoch": 0.05351552584821051,
+      "grad_norm": 1.8673244714736938,
+      "learning_rate": 0.001,
+      "loss": 2.2186,
+      "step": 1265
+    },
+    {
+      "epoch": 0.05355783061172688,
+      "grad_norm": 1.1225836277008057,
+      "learning_rate": 0.001,
+      "loss": 1.8668,
+      "step": 1266
+    },
+    {
+      "epoch": 0.05360013537524325,
+      "grad_norm": 0.8716728091239929,
+      "learning_rate": 0.001,
+      "loss": 2.8729,
+      "step": 1267
+    },
+    {
+      "epoch": 0.05364244013875962,
+      "grad_norm": 2.1635050773620605,
+      "learning_rate": 0.001,
+      "loss": 2.645,
+      "step": 1268
+    },
+    {
+      "epoch": 0.053684744902275996,
+      "grad_norm": 0.38754305243492126,
+      "learning_rate": 0.001,
+      "loss": 3.3657,
+      "step": 1269
+    },
+    {
+      "epoch": 0.05372704966579237,
+      "grad_norm": 0.35894379019737244,
+      "learning_rate": 0.001,
+      "loss": 1.9849,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05376935442930874,
+      "grad_norm": 0.45211154222488403,
+      "learning_rate": 0.001,
+      "loss": 3.4621,
+      "step": 1271
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 1.4661130905151367,
+      "learning_rate": 0.001,
+      "loss": 2.2278,
+      "step": 1272
+    },
+    {
+      "epoch": 0.05385396395634148,
+      "grad_norm": 0.3701130151748657,
+      "learning_rate": 0.001,
+      "loss": 1.8248,
+      "step": 1273
+    },
+    {
+      "epoch": 0.05389626871985786,
+      "grad_norm": 0.3744787871837616,
+      "learning_rate": 0.001,
+      "loss": 2.5166,
+      "step": 1274
+    },
+    {
+      "epoch": 0.053938573483374226,
+      "grad_norm": 0.6254411935806274,
+      "learning_rate": 0.001,
+      "loss": 2.1034,
+      "step": 1275
+    },
+    {
+      "epoch": 0.0539808782468906,
+      "grad_norm": 0.5083110928535461,
+      "learning_rate": 0.001,
+      "loss": 2.5448,
+      "step": 1276
+    },
+    {
+      "epoch": 0.05402318301040697,
+      "grad_norm": 0.4066379964351654,
+      "learning_rate": 0.001,
+      "loss": 2.726,
+      "step": 1277
+    },
+    {
+      "epoch": 0.054065487773923344,
+      "grad_norm": 0.326869398355484,
+      "learning_rate": 0.001,
+      "loss": 2.1421,
+      "step": 1278
+    },
+    {
+      "epoch": 0.05410779253743971,
+      "grad_norm": 0.51470547914505,
+      "learning_rate": 0.001,
+      "loss": 3.0139,
+      "step": 1279
+    },
+    {
+      "epoch": 0.05415009730095609,
+      "grad_norm": 1.5484551191329956,
+      "learning_rate": 0.001,
+      "loss": 2.947,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05419240206447246,
+      "grad_norm": 0.4610329270362854,
+      "learning_rate": 0.001,
+      "loss": 2.0814,
+      "step": 1281
+    },
+    {
+      "epoch": 0.05423470682798883,
+      "grad_norm": 1.545684576034546,
+      "learning_rate": 0.001,
+      "loss": 4.4303,
+      "step": 1282
+    },
+    {
+      "epoch": 0.054277011591505206,
+      "grad_norm": 1.627662181854248,
+      "learning_rate": 0.001,
+      "loss": 2.4313,
+      "step": 1283
+    },
+    {
+      "epoch": 0.054319316355021574,
+      "grad_norm": 0.37047141790390015,
+      "learning_rate": 0.001,
+      "loss": 2.1884,
+      "step": 1284
+    },
+    {
+      "epoch": 0.05436162111853795,
+      "grad_norm": 0.46785968542099,
+      "learning_rate": 0.001,
+      "loss": 3.1745,
+      "step": 1285
+    },
+    {
+      "epoch": 0.05440392588205432,
+      "grad_norm": 0.3685908615589142,
+      "learning_rate": 0.001,
+      "loss": 2.1659,
+      "step": 1286
+    },
+    {
+      "epoch": 0.05444623064557069,
+      "grad_norm": 0.8433362245559692,
+      "learning_rate": 0.001,
+      "loss": 2.7851,
+      "step": 1287
+    },
+    {
+      "epoch": 0.05448853540908706,
+      "grad_norm": 1.0135656595230103,
+      "learning_rate": 0.001,
+      "loss": 2.2148,
+      "step": 1288
+    },
+    {
+      "epoch": 0.054530840172603436,
+      "grad_norm": 0.9696979522705078,
+      "learning_rate": 0.001,
+      "loss": 2.5305,
+      "step": 1289
+    },
+    {
+      "epoch": 0.054573144936119804,
+      "grad_norm": 0.38324859738349915,
+      "learning_rate": 0.001,
+      "loss": 2.2144,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05461544969963618,
+      "grad_norm": 0.4518929421901703,
+      "learning_rate": 0.001,
+      "loss": 2.0715,
+      "step": 1291
+    },
+    {
+      "epoch": 0.054657754463152554,
+      "grad_norm": 0.29709556698799133,
+      "learning_rate": 0.001,
+      "loss": 1.878,
+      "step": 1292
+    },
+    {
+      "epoch": 0.05470005922666892,
+      "grad_norm": 0.41647666692733765,
+      "learning_rate": 0.001,
+      "loss": 3.2047,
+      "step": 1293
+    },
+    {
+      "epoch": 0.0547423639901853,
+      "grad_norm": 0.40319982171058655,
+      "learning_rate": 0.001,
+      "loss": 2.8011,
+      "step": 1294
+    },
+    {
+      "epoch": 0.054784668753701665,
+      "grad_norm": 1.4222816228866577,
+      "learning_rate": 0.001,
+      "loss": 2.4923,
+      "step": 1295
+    },
+    {
+      "epoch": 0.05482697351721804,
+      "grad_norm": 0.6435485482215881,
+      "learning_rate": 0.001,
+      "loss": 2.7776,
+      "step": 1296
+    },
+    {
+      "epoch": 0.05486927828073441,
+      "grad_norm": 0.3867262899875641,
+      "learning_rate": 0.001,
+      "loss": 2.9806,
+      "step": 1297
+    },
+    {
+      "epoch": 0.054911583044250784,
+      "grad_norm": 0.32293111085891724,
+      "learning_rate": 0.001,
+      "loss": 2.2782,
+      "step": 1298
+    },
+    {
+      "epoch": 0.05495388780776715,
+      "grad_norm": 1.7177414894104004,
+      "learning_rate": 0.001,
+      "loss": 2.5676,
+      "step": 1299
+    },
+    {
+      "epoch": 0.05499619257128353,
+      "grad_norm": 4.7346272468566895,
+      "learning_rate": 0.001,
+      "loss": 2.7508,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0550384973347999,
+      "grad_norm": 3.0460281372070312,
+      "learning_rate": 0.001,
+      "loss": 2.946,
+      "step": 1301
+    },
+    {
+      "epoch": 0.05508080209831627,
+      "grad_norm": 3.609449863433838,
+      "learning_rate": 0.001,
+      "loss": 2.9066,
+      "step": 1302
+    },
+    {
+      "epoch": 0.055123106861832645,
+      "grad_norm": 0.6072207689285278,
+      "learning_rate": 0.001,
+      "loss": 2.9887,
+      "step": 1303
+    },
+    {
+      "epoch": 0.05516541162534901,
+      "grad_norm": 0.3145444691181183,
+      "learning_rate": 0.001,
+      "loss": 2.0579,
+      "step": 1304
+    },
+    {
+      "epoch": 0.05520771638886539,
+      "grad_norm": 0.4488667845726013,
+      "learning_rate": 0.001,
+      "loss": 3.0136,
+      "step": 1305
+    },
+    {
+      "epoch": 0.055250021152381756,
+      "grad_norm": 2.8112523555755615,
+      "learning_rate": 0.001,
+      "loss": 2.0162,
+      "step": 1306
+    },
+    {
+      "epoch": 0.05529232591589813,
+      "grad_norm": 0.3468235731124878,
+      "learning_rate": 0.001,
+      "loss": 2.5913,
+      "step": 1307
+    },
+    {
+      "epoch": 0.0553346306794145,
+      "grad_norm": 0.32054612040519714,
+      "learning_rate": 0.001,
+      "loss": 2.9339,
+      "step": 1308
+    },
+    {
+      "epoch": 0.055376935442930875,
+      "grad_norm": 1.1245250701904297,
+      "learning_rate": 0.001,
+      "loss": 2.1141,
+      "step": 1309
+    },
+    {
+      "epoch": 0.05541924020644724,
+      "grad_norm": 0.8368834257125854,
+      "learning_rate": 0.001,
+      "loss": 2.7699,
+      "step": 1310
+    },
+    {
+      "epoch": 0.05546154496996362,
+      "grad_norm": 0.5362788438796997,
+      "learning_rate": 0.001,
+      "loss": 3.1062,
+      "step": 1311
+    },
+    {
+      "epoch": 0.05550384973347999,
+      "grad_norm": 0.3507515490055084,
+      "learning_rate": 0.001,
+      "loss": 2.5279,
+      "step": 1312
+    },
+    {
+      "epoch": 0.05554615449699636,
+      "grad_norm": 0.4399697184562683,
+      "learning_rate": 0.001,
+      "loss": 2.6847,
+      "step": 1313
+    },
+    {
+      "epoch": 0.055588459260512736,
+      "grad_norm": 0.693962812423706,
+      "learning_rate": 0.001,
+      "loss": 1.86,
+      "step": 1314
+    },
+    {
+      "epoch": 0.055630764024029104,
+      "grad_norm": 0.5763394236564636,
+      "learning_rate": 0.001,
+      "loss": 2.8907,
+      "step": 1315
+    },
+    {
+      "epoch": 0.05567306878754548,
+      "grad_norm": 0.3348402678966522,
+      "learning_rate": 0.001,
+      "loss": 2.6682,
+      "step": 1316
+    },
+    {
+      "epoch": 0.05571537355106185,
+      "grad_norm": 2.0695788860321045,
+      "learning_rate": 0.001,
+      "loss": 2.2931,
+      "step": 1317
+    },
+    {
+      "epoch": 0.05575767831457822,
+      "grad_norm": 1.0537198781967163,
+      "learning_rate": 0.001,
+      "loss": 2.7975,
+      "step": 1318
+    },
+    {
+      "epoch": 0.05579998307809459,
+      "grad_norm": 0.38374683260917664,
+      "learning_rate": 0.001,
+      "loss": 3.1988,
+      "step": 1319
+    },
+    {
+      "epoch": 0.055842287841610966,
+      "grad_norm": 16.446990966796875,
+      "learning_rate": 0.001,
+      "loss": 2.2754,
+      "step": 1320
+    },
+    {
+      "epoch": 0.055884592605127334,
+      "grad_norm": 0.378416508436203,
+      "learning_rate": 0.001,
+      "loss": 3.6752,
+      "step": 1321
+    },
+    {
+      "epoch": 0.05592689736864371,
+      "grad_norm": 3.8668839931488037,
+      "learning_rate": 0.001,
+      "loss": 3.5075,
+      "step": 1322
+    },
+    {
+      "epoch": 0.055969202132160084,
+      "grad_norm": 0.40051063895225525,
+      "learning_rate": 0.001,
+      "loss": 1.935,
+      "step": 1323
+    },
+    {
+      "epoch": 0.05601150689567645,
+      "grad_norm": 0.3199704587459564,
+      "learning_rate": 0.001,
+      "loss": 2.8692,
+      "step": 1324
+    },
+    {
+      "epoch": 0.05605381165919283,
+      "grad_norm": 0.39655637741088867,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 1325
+    },
+    {
+      "epoch": 0.056096116422709195,
+      "grad_norm": 0.32414257526397705,
+      "learning_rate": 0.001,
+      "loss": 2.3473,
+      "step": 1326
+    },
+    {
+      "epoch": 0.05613842118622557,
+      "grad_norm": 0.30689844489097595,
+      "learning_rate": 0.001,
+      "loss": 1.9908,
+      "step": 1327
+    },
+    {
+      "epoch": 0.05618072594974194,
+      "grad_norm": 0.3849036693572998,
+      "learning_rate": 0.001,
+      "loss": 2.2868,
+      "step": 1328
+    },
+    {
+      "epoch": 0.056223030713258314,
+      "grad_norm": 0.7097468376159668,
+      "learning_rate": 0.001,
+      "loss": 2.9331,
+      "step": 1329
+    },
+    {
+      "epoch": 0.05626533547677468,
+      "grad_norm": 0.3276157081127167,
+      "learning_rate": 0.001,
+      "loss": 2.3295,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05630764024029106,
+      "grad_norm": 0.30598339438438416,
+      "learning_rate": 0.001,
+      "loss": 2.8663,
+      "step": 1331
+    },
+    {
+      "epoch": 0.05634994500380743,
+      "grad_norm": 5.709721565246582,
+      "learning_rate": 0.001,
+      "loss": 2.2997,
+      "step": 1332
+    },
+    {
+      "epoch": 0.0563922497673238,
+      "grad_norm": 1.1615700721740723,
+      "learning_rate": 0.001,
+      "loss": 2.2019,
+      "step": 1333
+    },
+    {
+      "epoch": 0.056434554530840175,
+      "grad_norm": 0.8189941644668579,
+      "learning_rate": 0.001,
+      "loss": 1.9522,
+      "step": 1334
+    },
+    {
+      "epoch": 0.05647685929435654,
+      "grad_norm": 0.4313448965549469,
+      "learning_rate": 0.001,
+      "loss": 2.1067,
+      "step": 1335
+    },
+    {
+      "epoch": 0.05651916405787292,
+      "grad_norm": 0.38705217838287354,
+      "learning_rate": 0.001,
+      "loss": 2.9772,
+      "step": 1336
+    },
+    {
+      "epoch": 0.05656146882138929,
+      "grad_norm": 0.4954543709754944,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 1337
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.5692545771598816,
+      "learning_rate": 0.001,
+      "loss": 2.3607,
+      "step": 1338
+    },
+    {
+      "epoch": 0.05664607834842203,
+      "grad_norm": 1.1173979043960571,
+      "learning_rate": 0.001,
+      "loss": 2.9379,
+      "step": 1339
+    },
+    {
+      "epoch": 0.056688383111938405,
+      "grad_norm": 2.1702587604522705,
+      "learning_rate": 0.001,
+      "loss": 2.4834,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05673068787545477,
+      "grad_norm": 3.090153694152832,
+      "learning_rate": 0.001,
+      "loss": 1.9018,
+      "step": 1341
+    },
+    {
+      "epoch": 0.05677299263897115,
+      "grad_norm": 0.7734284996986389,
+      "learning_rate": 0.001,
+      "loss": 2.3851,
+      "step": 1342
+    },
+    {
+      "epoch": 0.05681529740248752,
+      "grad_norm": 0.2983352243900299,
+      "learning_rate": 0.001,
+      "loss": 2.5298,
+      "step": 1343
+    },
+    {
+      "epoch": 0.05685760216600389,
+      "grad_norm": 0.8630567193031311,
+      "learning_rate": 0.001,
+      "loss": 3.0804,
+      "step": 1344
+    },
+    {
+      "epoch": 0.056899906929520266,
+      "grad_norm": 0.3336477279663086,
+      "learning_rate": 0.001,
+      "loss": 2.2696,
+      "step": 1345
+    },
+    {
+      "epoch": 0.056942211693036635,
+      "grad_norm": 0.30996400117874146,
+      "learning_rate": 0.001,
+      "loss": 2.3269,
+      "step": 1346
+    },
+    {
+      "epoch": 0.05698451645655301,
+      "grad_norm": 0.38334864377975464,
+      "learning_rate": 0.001,
+      "loss": 2.4821,
+      "step": 1347
+    },
+    {
+      "epoch": 0.05702682122006938,
+      "grad_norm": 0.29219967126846313,
+      "learning_rate": 0.001,
+      "loss": 2.192,
+      "step": 1348
+    },
+    {
+      "epoch": 0.05706912598358575,
+      "grad_norm": 1.8894317150115967,
+      "learning_rate": 0.001,
+      "loss": 2.2862,
+      "step": 1349
+    },
+    {
+      "epoch": 0.05711143074710212,
+      "grad_norm": 3.3716726303100586,
+      "learning_rate": 0.001,
+      "loss": 2.0416,
+      "step": 1350
+    },
+    {
+      "epoch": 0.057153735510618496,
+      "grad_norm": 0.3227848708629608,
+      "learning_rate": 0.001,
+      "loss": 2.7895,
+      "step": 1351
+    },
+    {
+      "epoch": 0.057196040274134864,
+      "grad_norm": 0.31334635615348816,
+      "learning_rate": 0.001,
+      "loss": 2.7555,
+      "step": 1352
+    },
+    {
+      "epoch": 0.05723834503765124,
+      "grad_norm": 0.32888343930244446,
+      "learning_rate": 0.001,
+      "loss": 2.6685,
+      "step": 1353
+    },
+    {
+      "epoch": 0.057280649801167614,
+      "grad_norm": 1.3726975917816162,
+      "learning_rate": 0.001,
+      "loss": 2.4858,
+      "step": 1354
+    },
+    {
+      "epoch": 0.05732295456468398,
+      "grad_norm": 0.296441912651062,
+      "learning_rate": 0.001,
+      "loss": 2.4981,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05736525932820036,
+      "grad_norm": 0.40023407340049744,
+      "learning_rate": 0.001,
+      "loss": 2.3818,
+      "step": 1356
+    },
+    {
+      "epoch": 0.057407564091716726,
+      "grad_norm": 0.23572492599487305,
+      "learning_rate": 0.001,
+      "loss": 1.7868,
+      "step": 1357
+    },
+    {
+      "epoch": 0.0574498688552331,
+      "grad_norm": 0.3053801357746124,
+      "learning_rate": 0.001,
+      "loss": 2.115,
+      "step": 1358
+    },
+    {
+      "epoch": 0.05749217361874947,
+      "grad_norm": 0.4453347623348236,
+      "learning_rate": 0.001,
+      "loss": 2.3336,
+      "step": 1359
+    },
+    {
+      "epoch": 0.057534478382265844,
+      "grad_norm": 0.29795461893081665,
+      "learning_rate": 0.001,
+      "loss": 2.3624,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05757678314578221,
+      "grad_norm": 0.3462240993976593,
+      "learning_rate": 0.001,
+      "loss": 2.4219,
+      "step": 1361
+    },
+    {
+      "epoch": 0.05761908790929859,
+      "grad_norm": 0.7424085140228271,
+      "learning_rate": 0.001,
+      "loss": 2.0359,
+      "step": 1362
+    },
+    {
+      "epoch": 0.05766139267281496,
+      "grad_norm": 0.33206912875175476,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 1363
+    },
+    {
+      "epoch": 0.05770369743633133,
+      "grad_norm": 0.5105282068252563,
+      "learning_rate": 0.001,
+      "loss": 2.5832,
+      "step": 1364
+    },
+    {
+      "epoch": 0.057746002199847705,
+      "grad_norm": 0.29361414909362793,
+      "learning_rate": 0.001,
+      "loss": 1.5788,
+      "step": 1365
+    },
+    {
+      "epoch": 0.057788306963364074,
+      "grad_norm": 0.7224523425102234,
+      "learning_rate": 0.001,
+      "loss": 1.6817,
+      "step": 1366
+    },
+    {
+      "epoch": 0.05783061172688045,
+      "grad_norm": 0.33219602704048157,
+      "learning_rate": 0.001,
+      "loss": 2.8996,
+      "step": 1367
+    },
+    {
+      "epoch": 0.05787291649039682,
+      "grad_norm": 0.6146803498268127,
+      "learning_rate": 0.001,
+      "loss": 2.8432,
+      "step": 1368
+    },
+    {
+      "epoch": 0.05791522125391319,
+      "grad_norm": 0.4107002019882202,
+      "learning_rate": 0.001,
+      "loss": 2.1509,
+      "step": 1369
+    },
+    {
+      "epoch": 0.05795752601742956,
+      "grad_norm": 0.7845795750617981,
+      "learning_rate": 0.001,
+      "loss": 2.5862,
+      "step": 1370
+    },
+    {
+      "epoch": 0.057999830780945935,
+      "grad_norm": 0.2706119120121002,
+      "learning_rate": 0.001,
+      "loss": 2.4746,
+      "step": 1371
+    },
+    {
+      "epoch": 0.0580421355444623,
+      "grad_norm": 0.26624995470046997,
+      "learning_rate": 0.001,
+      "loss": 2.1325,
+      "step": 1372
+    },
+    {
+      "epoch": 0.05808444030797868,
+      "grad_norm": 1.2387701272964478,
+      "learning_rate": 0.001,
+      "loss": 3.0806,
+      "step": 1373
+    },
+    {
+      "epoch": 0.05812674507149505,
+      "grad_norm": 0.30188238620758057,
+      "learning_rate": 0.001,
+      "loss": 2.6137,
+      "step": 1374
+    },
+    {
+      "epoch": 0.05816904983501142,
+      "grad_norm": 0.47407734394073486,
+      "learning_rate": 0.001,
+      "loss": 2.949,
+      "step": 1375
+    },
+    {
+      "epoch": 0.0582113545985278,
+      "grad_norm": 13.492250442504883,
+      "learning_rate": 0.001,
+      "loss": 2.3764,
+      "step": 1376
+    },
+    {
+      "epoch": 0.058253659362044165,
+      "grad_norm": 1.4784618616104126,
+      "learning_rate": 0.001,
+      "loss": 2.6514,
+      "step": 1377
+    },
+    {
+      "epoch": 0.05829596412556054,
+      "grad_norm": 0.5186235904693604,
+      "learning_rate": 0.001,
+      "loss": 2.5127,
+      "step": 1378
+    },
+    {
+      "epoch": 0.05833826888907691,
+      "grad_norm": 0.33871689438819885,
+      "learning_rate": 0.001,
+      "loss": 1.9724,
+      "step": 1379
+    },
+    {
+      "epoch": 0.05838057365259328,
+      "grad_norm": 0.6185868382453918,
+      "learning_rate": 0.001,
+      "loss": 2.2309,
+      "step": 1380
+    },
+    {
+      "epoch": 0.05842287841610965,
+      "grad_norm": 0.3206985890865326,
+      "learning_rate": 0.001,
+      "loss": 2.2691,
+      "step": 1381
+    },
+    {
+      "epoch": 0.058465183179626026,
+      "grad_norm": 0.34144043922424316,
+      "learning_rate": 0.001,
+      "loss": 2.1018,
+      "step": 1382
+    },
+    {
+      "epoch": 0.058507487943142394,
+      "grad_norm": 2.2303543090820312,
+      "learning_rate": 0.001,
+      "loss": 2.0798,
+      "step": 1383
+    },
+    {
+      "epoch": 0.05854979270665877,
+      "grad_norm": 0.36561813950538635,
+      "learning_rate": 0.001,
+      "loss": 2.8894,
+      "step": 1384
+    },
+    {
+      "epoch": 0.058592097470175145,
+      "grad_norm": 0.5910583138465881,
+      "learning_rate": 0.001,
+      "loss": 2.6603,
+      "step": 1385
+    },
+    {
+      "epoch": 0.05863440223369151,
+      "grad_norm": 0.3511381149291992,
+      "learning_rate": 0.001,
+      "loss": 3.2612,
+      "step": 1386
+    },
+    {
+      "epoch": 0.05867670699720789,
+      "grad_norm": 1.0417660474777222,
+      "learning_rate": 0.001,
+      "loss": 2.0464,
+      "step": 1387
+    },
+    {
+      "epoch": 0.058719011760724256,
+      "grad_norm": 0.3064709007740021,
+      "learning_rate": 0.001,
+      "loss": 1.7901,
+      "step": 1388
+    },
+    {
+      "epoch": 0.05876131652424063,
+      "grad_norm": 0.6421626210212708,
+      "learning_rate": 0.001,
+      "loss": 2.4181,
+      "step": 1389
+    },
+    {
+      "epoch": 0.058803621287757,
+      "grad_norm": 0.5801759958267212,
+      "learning_rate": 0.001,
+      "loss": 1.9541,
+      "step": 1390
+    },
+    {
+      "epoch": 0.058845926051273374,
+      "grad_norm": 0.32130327820777893,
+      "learning_rate": 0.001,
+      "loss": 2.5807,
+      "step": 1391
+    },
+    {
+      "epoch": 0.05888823081478974,
+      "grad_norm": 0.31312790513038635,
+      "learning_rate": 0.001,
+      "loss": 3.0668,
+      "step": 1392
+    },
+    {
+      "epoch": 0.05893053557830612,
+      "grad_norm": 0.5199546813964844,
+      "learning_rate": 0.001,
+      "loss": 2.9197,
+      "step": 1393
+    },
+    {
+      "epoch": 0.05897284034182249,
+      "grad_norm": 0.45629438757896423,
+      "learning_rate": 0.001,
+      "loss": 1.9592,
+      "step": 1394
+    },
+    {
+      "epoch": 0.05901514510533886,
+      "grad_norm": 0.2890334129333496,
+      "learning_rate": 0.001,
+      "loss": 2.716,
+      "step": 1395
+    },
+    {
+      "epoch": 0.059057449868855236,
+      "grad_norm": 0.47239354252815247,
+      "learning_rate": 0.001,
+      "loss": 2.2804,
+      "step": 1396
+    },
+    {
+      "epoch": 0.059099754632371604,
+      "grad_norm": 0.989652693271637,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 1397
+    },
+    {
+      "epoch": 0.05914205939588798,
+      "grad_norm": 0.40251368284225464,
+      "learning_rate": 0.001,
+      "loss": 1.9777,
+      "step": 1398
+    },
+    {
+      "epoch": 0.05918436415940435,
+      "grad_norm": 0.2854582965373993,
+      "learning_rate": 0.001,
+      "loss": 1.8564,
+      "step": 1399
+    },
+    {
+      "epoch": 0.05922666892292072,
+      "grad_norm": 0.4429548382759094,
+      "learning_rate": 0.001,
+      "loss": 3.2216,
+      "step": 1400
+    },
+    {
+      "epoch": 0.05926897368643709,
+      "grad_norm": 0.3628709316253662,
+      "learning_rate": 0.001,
+      "loss": 1.898,
+      "step": 1401
+    },
+    {
+      "epoch": 0.059311278449953465,
+      "grad_norm": 0.37840116024017334,
+      "learning_rate": 0.001,
+      "loss": 3.3639,
+      "step": 1402
+    },
+    {
+      "epoch": 0.059353583213469834,
+      "grad_norm": 1.245588779449463,
+      "learning_rate": 0.001,
+      "loss": 3.1546,
+      "step": 1403
+    },
+    {
+      "epoch": 0.05939588797698621,
+      "grad_norm": 0.304123193025589,
+      "learning_rate": 0.001,
+      "loss": 3.7022,
+      "step": 1404
+    },
+    {
+      "epoch": 0.059438192740502584,
+      "grad_norm": 0.4658697843551636,
+      "learning_rate": 0.001,
+      "loss": 2.8618,
+      "step": 1405
+    },
+    {
+      "epoch": 0.05948049750401895,
+      "grad_norm": 0.3281252682209015,
+      "learning_rate": 0.001,
+      "loss": 2.5643,
+      "step": 1406
+    },
+    {
+      "epoch": 0.05952280226753533,
+      "grad_norm": 0.28439781069755554,
+      "learning_rate": 0.001,
+      "loss": 3.376,
+      "step": 1407
+    },
+    {
+      "epoch": 0.059565107031051695,
+      "grad_norm": 0.6219123601913452,
+      "learning_rate": 0.001,
+      "loss": 4.5267,
+      "step": 1408
+    },
+    {
+      "epoch": 0.05960741179456807,
+      "grad_norm": 0.30366063117980957,
+      "learning_rate": 0.001,
+      "loss": 2.8077,
+      "step": 1409
+    },
+    {
+      "epoch": 0.05964971655808444,
+      "grad_norm": 0.2643471658229828,
+      "learning_rate": 0.001,
+      "loss": 2.5821,
+      "step": 1410
+    },
+    {
+      "epoch": 0.05969202132160081,
+      "grad_norm": 0.5798410177230835,
+      "learning_rate": 0.001,
+      "loss": 2.406,
+      "step": 1411
+    },
+    {
+      "epoch": 0.05973432608511718,
+      "grad_norm": 0.2849283218383789,
+      "learning_rate": 0.001,
+      "loss": 1.9369,
+      "step": 1412
+    },
+    {
+      "epoch": 0.059776630848633557,
+      "grad_norm": 0.35845881700515747,
+      "learning_rate": 0.001,
+      "loss": 2.6967,
+      "step": 1413
+    },
+    {
+      "epoch": 0.059818935612149925,
+      "grad_norm": 1.3240238428115845,
+      "learning_rate": 0.001,
+      "loss": 3.4985,
+      "step": 1414
+    },
+    {
+      "epoch": 0.0598612403756663,
+      "grad_norm": 0.2925819754600525,
+      "learning_rate": 0.001,
+      "loss": 2.6216,
+      "step": 1415
+    },
+    {
+      "epoch": 0.059903545139182675,
+      "grad_norm": 0.8406017422676086,
+      "learning_rate": 0.001,
+      "loss": 2.1828,
+      "step": 1416
+    },
+    {
+      "epoch": 0.05994584990269904,
+      "grad_norm": 0.344137966632843,
+      "learning_rate": 0.001,
+      "loss": 2.3818,
+      "step": 1417
+    },
+    {
+      "epoch": 0.05998815466621542,
+      "grad_norm": 0.3151387870311737,
+      "learning_rate": 0.001,
+      "loss": 2.53,
+      "step": 1418
+    },
+    {
+      "epoch": 0.060030459429731786,
+      "grad_norm": 0.3660660982131958,
+      "learning_rate": 0.001,
+      "loss": 2.4767,
+      "step": 1419
+    },
+    {
+      "epoch": 0.06007276419324816,
+      "grad_norm": 3.185765266418457,
+      "learning_rate": 0.001,
+      "loss": 1.7922,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06011506895676453,
+      "grad_norm": 0.4721246659755707,
+      "learning_rate": 0.001,
+      "loss": 2.1482,
+      "step": 1421
+    },
+    {
+      "epoch": 0.060157373720280904,
+      "grad_norm": 0.281838983297348,
+      "learning_rate": 0.001,
+      "loss": 2.1035,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06019967848379727,
+      "grad_norm": 0.5046917796134949,
+      "learning_rate": 0.001,
+      "loss": 2.3257,
+      "step": 1423
+    },
+    {
+      "epoch": 0.06024198324731365,
+      "grad_norm": 0.36857858300209045,
+      "learning_rate": 0.001,
+      "loss": 2.5842,
+      "step": 1424
+    },
+    {
+      "epoch": 0.06028428801083002,
+      "grad_norm": 0.6424447894096375,
+      "learning_rate": 0.001,
+      "loss": 2.5619,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06032659277434639,
+      "grad_norm": 0.3190594017505646,
+      "learning_rate": 0.001,
+      "loss": 2.6482,
+      "step": 1426
+    },
+    {
+      "epoch": 0.060368897537862766,
+      "grad_norm": 0.3126913905143738,
+      "learning_rate": 0.001,
+      "loss": 3.0516,
+      "step": 1427
+    },
+    {
+      "epoch": 0.060411202301379134,
+      "grad_norm": 0.41238340735435486,
+      "learning_rate": 0.001,
+      "loss": 3.3109,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06045350706489551,
+      "grad_norm": 0.44219493865966797,
+      "learning_rate": 0.001,
+      "loss": 3.6026,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06049581182841188,
+      "grad_norm": 0.32209640741348267,
+      "learning_rate": 0.001,
+      "loss": 2.6428,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06053811659192825,
+      "grad_norm": 0.39168989658355713,
+      "learning_rate": 0.001,
+      "loss": 3.429,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06058042135544462,
+      "grad_norm": 0.27055805921554565,
+      "learning_rate": 0.001,
+      "loss": 3.8608,
+      "step": 1432
+    },
+    {
+      "epoch": 0.060622726118960996,
+      "grad_norm": 0.49823957681655884,
+      "learning_rate": 0.001,
+      "loss": 3.4317,
+      "step": 1433
+    },
+    {
+      "epoch": 0.060665030882477364,
+      "grad_norm": 0.32492002844810486,
+      "learning_rate": 0.001,
+      "loss": 2.4563,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06070733564599374,
+      "grad_norm": 0.34951508045196533,
+      "learning_rate": 0.001,
+      "loss": 3.2943,
+      "step": 1435
+    },
+    {
+      "epoch": 0.060749640409510114,
+      "grad_norm": 0.4263649880886078,
+      "learning_rate": 0.001,
+      "loss": 2.6012,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06079194517302648,
+      "grad_norm": 1.0415583848953247,
+      "learning_rate": 0.001,
+      "loss": 2.7861,
+      "step": 1437
+    },
+    {
+      "epoch": 0.06083424993654286,
+      "grad_norm": 0.331173837184906,
+      "learning_rate": 0.001,
+      "loss": 2.4296,
+      "step": 1438
+    },
+    {
+      "epoch": 0.060876554700059225,
+      "grad_norm": 0.88392573595047,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 1439
+    },
+    {
+      "epoch": 0.0609188594635756,
+      "grad_norm": 1.8237016201019287,
+      "learning_rate": 0.001,
+      "loss": 2.9384,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06096116422709197,
+      "grad_norm": 0.40476563572883606,
+      "learning_rate": 0.001,
+      "loss": 3.0706,
+      "step": 1441
+    },
+    {
+      "epoch": 0.061003468990608344,
+      "grad_norm": 2.381662607192993,
+      "learning_rate": 0.001,
+      "loss": 2.1478,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06104577375412471,
+      "grad_norm": 0.3319908082485199,
+      "learning_rate": 0.001,
+      "loss": 3.0221,
+      "step": 1443
+    },
+    {
+      "epoch": 0.06108807851764109,
+      "grad_norm": 1.0969241857528687,
+      "learning_rate": 0.001,
+      "loss": 2.5479,
+      "step": 1444
+    },
+    {
+      "epoch": 0.061130383281157455,
+      "grad_norm": 0.510847806930542,
+      "learning_rate": 0.001,
+      "loss": 1.8381,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06117268804467383,
+      "grad_norm": 0.5893344879150391,
+      "learning_rate": 0.001,
+      "loss": 2.661,
+      "step": 1446
+    },
+    {
+      "epoch": 0.061214992808190205,
+      "grad_norm": 0.36038708686828613,
+      "learning_rate": 0.001,
+      "loss": 3.1335,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06125729757170657,
+      "grad_norm": 0.3727112412452698,
+      "learning_rate": 0.001,
+      "loss": 3.2217,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06129960233522295,
+      "grad_norm": 0.31264758110046387,
+      "learning_rate": 0.001,
+      "loss": 2.0928,
+      "step": 1449
+    },
+    {
+      "epoch": 0.061341907098739316,
+      "grad_norm": 0.3144207298755646,
+      "learning_rate": 0.001,
+      "loss": 2.5247,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06138421186225569,
+      "grad_norm": 0.5916879773139954,
+      "learning_rate": 0.001,
+      "loss": 3.4771,
+      "step": 1451
+    },
+    {
+      "epoch": 0.06142651662577206,
+      "grad_norm": 0.34228265285491943,
+      "learning_rate": 0.001,
+      "loss": 1.9271,
+      "step": 1452
+    },
+    {
+      "epoch": 0.061468821389288435,
+      "grad_norm": 0.31869953870773315,
+      "learning_rate": 0.001,
+      "loss": 1.9089,
+      "step": 1453
+    },
+    {
+      "epoch": 0.0615111261528048,
+      "grad_norm": 0.2902565598487854,
+      "learning_rate": 0.001,
+      "loss": 1.9665,
+      "step": 1454
+    },
+    {
+      "epoch": 0.06155343091632118,
+      "grad_norm": 0.8417028188705444,
+      "learning_rate": 0.001,
+      "loss": 2.1853,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06159573567983755,
+      "grad_norm": 0.29936543107032776,
+      "learning_rate": 0.001,
+      "loss": 2.2901,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06163804044335392,
+      "grad_norm": 1.0548635721206665,
+      "learning_rate": 0.001,
+      "loss": 3.0753,
+      "step": 1457
+    },
+    {
+      "epoch": 0.061680345206870296,
+      "grad_norm": 0.5437618494033813,
+      "learning_rate": 0.001,
+      "loss": 2.3487,
+      "step": 1458
+    },
+    {
+      "epoch": 0.061722649970386664,
+      "grad_norm": 0.35127514600753784,
+      "learning_rate": 0.001,
+      "loss": 2.761,
+      "step": 1459
+    },
+    {
+      "epoch": 0.06176495473390304,
+      "grad_norm": 0.3495043218135834,
+      "learning_rate": 0.001,
+      "loss": 2.1056,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06180725949741941,
+      "grad_norm": 0.3617175817489624,
+      "learning_rate": 0.001,
+      "loss": 2.226,
+      "step": 1461
+    },
+    {
+      "epoch": 0.06184956426093578,
+      "grad_norm": 2.682561159133911,
+      "learning_rate": 0.001,
+      "loss": 2.2108,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06189186902445215,
+      "grad_norm": 0.6340761184692383,
+      "learning_rate": 0.001,
+      "loss": 3.0965,
+      "step": 1463
+    },
+    {
+      "epoch": 0.061934173787968526,
+      "grad_norm": 0.43284621834754944,
+      "learning_rate": 0.001,
+      "loss": 3.3314,
+      "step": 1464
+    },
+    {
+      "epoch": 0.061976478551484894,
+      "grad_norm": 0.9071413278579712,
+      "learning_rate": 0.001,
+      "loss": 1.9998,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06201878331500127,
+      "grad_norm": 0.32309386134147644,
+      "learning_rate": 0.001,
+      "loss": 1.6943,
+      "step": 1466
+    },
+    {
+      "epoch": 0.062061088078517644,
+      "grad_norm": 1.5545170307159424,
+      "learning_rate": 0.001,
+      "loss": 2.5362,
+      "step": 1467
+    },
+    {
+      "epoch": 0.06210339284203401,
+      "grad_norm": 0.4889759123325348,
+      "learning_rate": 0.001,
+      "loss": 2.4579,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06214569760555039,
+      "grad_norm": 0.3919735252857208,
+      "learning_rate": 0.001,
+      "loss": 2.5643,
+      "step": 1469
+    },
+    {
+      "epoch": 0.062188002369066755,
+      "grad_norm": 0.8115111589431763,
+      "learning_rate": 0.001,
+      "loss": 2.6917,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06223030713258313,
+      "grad_norm": 1.6906379461288452,
+      "learning_rate": 0.001,
+      "loss": 1.7673,
+      "step": 1471
+    },
+    {
+      "epoch": 0.0622726118960995,
+      "grad_norm": 0.33348721265792847,
+      "learning_rate": 0.001,
+      "loss": 1.9239,
+      "step": 1472
+    },
+    {
+      "epoch": 0.062314916659615874,
+      "grad_norm": 0.5403802990913391,
+      "learning_rate": 0.001,
+      "loss": 2.6563,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06235722142313224,
+      "grad_norm": 0.6966021060943604,
+      "learning_rate": 0.001,
+      "loss": 1.9165,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06239952618664862,
+      "grad_norm": 0.720427930355072,
+      "learning_rate": 0.001,
+      "loss": 2.4165,
+      "step": 1475
+    },
+    {
+      "epoch": 0.062441830950164985,
+      "grad_norm": 0.617408275604248,
+      "learning_rate": 0.001,
+      "loss": 3.6807,
+      "step": 1476
+    },
+    {
+      "epoch": 0.06248413571368136,
+      "grad_norm": 0.5452759265899658,
+      "learning_rate": 0.001,
+      "loss": 3.4443,
+      "step": 1477
+    },
+    {
+      "epoch": 0.06252644047719773,
+      "grad_norm": 0.4260646402835846,
+      "learning_rate": 0.001,
+      "loss": 3.0329,
+      "step": 1478
+    },
+    {
+      "epoch": 0.06256874524071411,
+      "grad_norm": 0.3713223338127136,
+      "learning_rate": 0.001,
+      "loss": 3.1337,
+      "step": 1479
+    },
+    {
+      "epoch": 0.06261105000423048,
+      "grad_norm": 0.4795995056629181,
+      "learning_rate": 0.001,
+      "loss": 2.6778,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06265335476774685,
+      "grad_norm": 1.03213369846344,
+      "learning_rate": 0.001,
+      "loss": 2.7868,
+      "step": 1481
+    },
+    {
+      "epoch": 0.06269565953126321,
+      "grad_norm": 0.7824976444244385,
+      "learning_rate": 0.001,
+      "loss": 2.982,
+      "step": 1482
+    },
+    {
+      "epoch": 0.0627379642947796,
+      "grad_norm": 1.2638541460037231,
+      "learning_rate": 0.001,
+      "loss": 3.0771,
+      "step": 1483
+    },
+    {
+      "epoch": 0.06278026905829596,
+      "grad_norm": 1.2298089265823364,
+      "learning_rate": 0.001,
+      "loss": 2.0408,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06282257382181233,
+      "grad_norm": 0.8441680073738098,
+      "learning_rate": 0.001,
+      "loss": 2.4082,
+      "step": 1485
+    },
+    {
+      "epoch": 0.0628648785853287,
+      "grad_norm": 0.34452885389328003,
+      "learning_rate": 0.001,
+      "loss": 2.2197,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06290718334884508,
+      "grad_norm": 0.39541852474212646,
+      "learning_rate": 0.001,
+      "loss": 2.9002,
+      "step": 1487
+    },
+    {
+      "epoch": 0.06294948811236145,
+      "grad_norm": 0.49740156531333923,
+      "learning_rate": 0.001,
+      "loss": 2.4369,
+      "step": 1488
+    },
+    {
+      "epoch": 0.06299179287587782,
+      "grad_norm": 1.0628348588943481,
+      "learning_rate": 0.001,
+      "loss": 2.9808,
+      "step": 1489
+    },
+    {
+      "epoch": 0.0630340976393942,
+      "grad_norm": 0.3743760883808136,
+      "learning_rate": 0.001,
+      "loss": 2.8068,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06307640240291057,
+      "grad_norm": 0.6638221740722656,
+      "learning_rate": 0.001,
+      "loss": 2.0364,
+      "step": 1491
+    },
+    {
+      "epoch": 0.06311870716642694,
+      "grad_norm": 0.34382185339927673,
+      "learning_rate": 0.001,
+      "loss": 2.6025,
+      "step": 1492
+    },
+    {
+      "epoch": 0.0631610119299433,
+      "grad_norm": 0.378578782081604,
+      "learning_rate": 0.001,
+      "loss": 2.762,
+      "step": 1493
+    },
+    {
+      "epoch": 0.06320331669345969,
+      "grad_norm": 0.38604044914245605,
+      "learning_rate": 0.001,
+      "loss": 2.0935,
+      "step": 1494
+    },
+    {
+      "epoch": 0.06324562145697606,
+      "grad_norm": 1.129136562347412,
+      "learning_rate": 0.001,
+      "loss": 2.8483,
+      "step": 1495
+    },
+    {
+      "epoch": 0.06328792622049242,
+      "grad_norm": 0.30852529406547546,
+      "learning_rate": 0.001,
+      "loss": 2.3716,
+      "step": 1496
+    },
+    {
+      "epoch": 0.0633302309840088,
+      "grad_norm": 2.0638482570648193,
+      "learning_rate": 0.001,
+      "loss": 2.6199,
+      "step": 1497
+    },
+    {
+      "epoch": 0.06337253574752517,
+      "grad_norm": 0.25943711400032043,
+      "learning_rate": 0.001,
+      "loss": 2.2946,
+      "step": 1498
+    },
+    {
+      "epoch": 0.06341484051104154,
+      "grad_norm": 0.31075453758239746,
+      "learning_rate": 0.001,
+      "loss": 2.1379,
+      "step": 1499
+    },
+    {
+      "epoch": 0.06345714527455791,
+      "grad_norm": 0.34476733207702637,
+      "learning_rate": 0.001,
+      "loss": 3.0886,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06349945003807429,
+      "grad_norm": 0.3541569709777832,
+      "learning_rate": 0.001,
+      "loss": 2.0473,
+      "step": 1501
+    },
+    {
+      "epoch": 0.06354175480159066,
+      "grad_norm": 0.6460628509521484,
+      "learning_rate": 0.001,
+      "loss": 2.4141,
+      "step": 1502
+    },
+    {
+      "epoch": 0.06358405956510703,
+      "grad_norm": 0.30732765793800354,
+      "learning_rate": 0.001,
+      "loss": 2.4192,
+      "step": 1503
+    },
+    {
+      "epoch": 0.0636263643286234,
+      "grad_norm": 0.3573780655860901,
+      "learning_rate": 0.001,
+      "loss": 3.1885,
+      "step": 1504
+    },
+    {
+      "epoch": 0.06366866909213978,
+      "grad_norm": 0.26040375232696533,
+      "learning_rate": 0.001,
+      "loss": 1.8493,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06371097385565615,
+      "grad_norm": 0.2959687411785126,
+      "learning_rate": 0.001,
+      "loss": 2.4594,
+      "step": 1506
+    },
+    {
+      "epoch": 0.06375327861917252,
+      "grad_norm": 0.3453410267829895,
+      "learning_rate": 0.001,
+      "loss": 3.036,
+      "step": 1507
+    },
+    {
+      "epoch": 0.0637955833826889,
+      "grad_norm": 0.35461366176605225,
+      "learning_rate": 0.001,
+      "loss": 3.1924,
+      "step": 1508
+    },
+    {
+      "epoch": 0.06383788814620527,
+      "grad_norm": 0.34828194975852966,
+      "learning_rate": 0.001,
+      "loss": 2.1851,
+      "step": 1509
+    },
+    {
+      "epoch": 0.06388019290972163,
+      "grad_norm": 2.283674955368042,
+      "learning_rate": 0.001,
+      "loss": 2.278,
+      "step": 1510
+    },
+    {
+      "epoch": 0.063922497673238,
+      "grad_norm": 0.3838748335838318,
+      "learning_rate": 0.001,
+      "loss": 2.2081,
+      "step": 1511
+    },
+    {
+      "epoch": 0.06396480243675438,
+      "grad_norm": 0.2536729574203491,
+      "learning_rate": 0.001,
+      "loss": 1.9966,
+      "step": 1512
+    },
+    {
+      "epoch": 0.06400710720027075,
+      "grad_norm": 3.2230656147003174,
+      "learning_rate": 0.001,
+      "loss": 2.9543,
+      "step": 1513
+    },
+    {
+      "epoch": 0.06404941196378712,
+      "grad_norm": 0.2882685661315918,
+      "learning_rate": 0.001,
+      "loss": 2.0072,
+      "step": 1514
+    },
+    {
+      "epoch": 0.06409171672730349,
+      "grad_norm": 0.34128835797309875,
+      "learning_rate": 0.001,
+      "loss": 2.4368,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06413402149081987,
+      "grad_norm": 0.326494425535202,
+      "learning_rate": 0.001,
+      "loss": 2.5767,
+      "step": 1516
+    },
+    {
+      "epoch": 0.06417632625433624,
+      "grad_norm": 0.4839684069156647,
+      "learning_rate": 0.001,
+      "loss": 2.0286,
+      "step": 1517
+    },
+    {
+      "epoch": 0.0642186310178526,
+      "grad_norm": 0.43224430084228516,
+      "learning_rate": 0.001,
+      "loss": 3.4441,
+      "step": 1518
+    },
+    {
+      "epoch": 0.06426093578136899,
+      "grad_norm": 0.41439104080200195,
+      "learning_rate": 0.001,
+      "loss": 2.9004,
+      "step": 1519
+    },
+    {
+      "epoch": 0.06430324054488536,
+      "grad_norm": 0.35360193252563477,
+      "learning_rate": 0.001,
+      "loss": 2.659,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06434554530840172,
+      "grad_norm": 0.25587019324302673,
+      "learning_rate": 0.001,
+      "loss": 2.0744,
+      "step": 1521
+    },
+    {
+      "epoch": 0.06438785007191809,
+      "grad_norm": 0.2605937719345093,
+      "learning_rate": 0.001,
+      "loss": 2.4958,
+      "step": 1522
+    },
+    {
+      "epoch": 0.06443015483543447,
+      "grad_norm": 0.26113244891166687,
+      "learning_rate": 0.001,
+      "loss": 2.1607,
+      "step": 1523
+    },
+    {
+      "epoch": 0.06447245959895084,
+      "grad_norm": 0.3792099356651306,
+      "learning_rate": 0.001,
+      "loss": 2.3428,
+      "step": 1524
+    },
+    {
+      "epoch": 0.06451476436246721,
+      "grad_norm": 0.44862857460975647,
+      "learning_rate": 0.001,
+      "loss": 2.1672,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06455706912598358,
+      "grad_norm": 0.2584918737411499,
+      "learning_rate": 0.001,
+      "loss": 2.2928,
+      "step": 1526
+    },
+    {
+      "epoch": 0.06459937388949996,
+      "grad_norm": 0.2768254280090332,
+      "learning_rate": 0.001,
+      "loss": 1.6838,
+      "step": 1527
+    },
+    {
+      "epoch": 0.06464167865301633,
+      "grad_norm": 0.30833518505096436,
+      "learning_rate": 0.001,
+      "loss": 1.8115,
+      "step": 1528
+    },
+    {
+      "epoch": 0.0646839834165327,
+      "grad_norm": 0.8284555077552795,
+      "learning_rate": 0.001,
+      "loss": 2.5659,
+      "step": 1529
+    },
+    {
+      "epoch": 0.06472628818004908,
+      "grad_norm": 0.5380930304527283,
+      "learning_rate": 0.001,
+      "loss": 2.6105,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06476859294356545,
+      "grad_norm": 0.5618930459022522,
+      "learning_rate": 0.001,
+      "loss": 2.2628,
+      "step": 1531
+    },
+    {
+      "epoch": 0.06481089770708182,
+      "grad_norm": 0.27346742153167725,
+      "learning_rate": 0.001,
+      "loss": 2.1463,
+      "step": 1532
+    },
+    {
+      "epoch": 0.06485320247059818,
+      "grad_norm": 0.30352386832237244,
+      "learning_rate": 0.001,
+      "loss": 3.6644,
+      "step": 1533
+    },
+    {
+      "epoch": 0.06489550723411457,
+      "grad_norm": 0.31220725178718567,
+      "learning_rate": 0.001,
+      "loss": 2.6265,
+      "step": 1534
+    },
+    {
+      "epoch": 0.06493781199763093,
+      "grad_norm": 0.27084285020828247,
+      "learning_rate": 0.001,
+      "loss": 2.3142,
+      "step": 1535
+    },
+    {
+      "epoch": 0.0649801167611473,
+      "grad_norm": 0.5739223957061768,
+      "learning_rate": 0.001,
+      "loss": 2.2806,
+      "step": 1536
+    },
+    {
+      "epoch": 0.06502242152466367,
+      "grad_norm": 0.7535676956176758,
+      "learning_rate": 0.001,
+      "loss": 2.3798,
+      "step": 1537
+    },
+    {
+      "epoch": 0.06506472628818005,
+      "grad_norm": 0.27977368235588074,
+      "learning_rate": 0.001,
+      "loss": 2.7091,
+      "step": 1538
+    },
+    {
+      "epoch": 0.06510703105169642,
+      "grad_norm": 0.6584524512290955,
+      "learning_rate": 0.001,
+      "loss": 2.9503,
+      "step": 1539
+    },
+    {
+      "epoch": 0.06514933581521279,
+      "grad_norm": 0.34777507185935974,
+      "learning_rate": 0.001,
+      "loss": 2.8689,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06519164057872917,
+      "grad_norm": 0.5156388282775879,
+      "learning_rate": 0.001,
+      "loss": 2.4906,
+      "step": 1541
+    },
+    {
+      "epoch": 0.06523394534224554,
+      "grad_norm": 0.28216788172721863,
+      "learning_rate": 0.001,
+      "loss": 1.8898,
+      "step": 1542
+    },
+    {
+      "epoch": 0.06527625010576191,
+      "grad_norm": 1.3255605697631836,
+      "learning_rate": 0.001,
+      "loss": 1.7489,
+      "step": 1543
+    },
+    {
+      "epoch": 0.06531855486927828,
+      "grad_norm": 0.28111153841018677,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 1544
+    },
+    {
+      "epoch": 0.06536085963279466,
+      "grad_norm": 0.2950698435306549,
+      "learning_rate": 0.001,
+      "loss": 1.6588,
+      "step": 1545
+    },
+    {
+      "epoch": 0.06540316439631103,
+      "grad_norm": 0.2627398371696472,
+      "learning_rate": 0.001,
+      "loss": 2.2555,
+      "step": 1546
+    },
+    {
+      "epoch": 0.0654454691598274,
+      "grad_norm": 0.545285701751709,
+      "learning_rate": 0.001,
+      "loss": 2.238,
+      "step": 1547
+    },
+    {
+      "epoch": 0.06548777392334376,
+      "grad_norm": 0.2821153402328491,
+      "learning_rate": 0.001,
+      "loss": 2.5748,
+      "step": 1548
+    },
+    {
+      "epoch": 0.06553007868686014,
+      "grad_norm": 0.32898885011672974,
+      "learning_rate": 0.001,
+      "loss": 2.6032,
+      "step": 1549
+    },
+    {
+      "epoch": 0.06557238345037651,
+      "grad_norm": 0.6069597601890564,
+      "learning_rate": 0.001,
+      "loss": 2.3282,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06561468821389288,
+      "grad_norm": 0.33791178464889526,
+      "learning_rate": 0.001,
+      "loss": 2.3444,
+      "step": 1551
+    },
+    {
+      "epoch": 0.06565699297740926,
+      "grad_norm": 0.630935788154602,
+      "learning_rate": 0.001,
+      "loss": 2.574,
+      "step": 1552
+    },
+    {
+      "epoch": 0.06569929774092563,
+      "grad_norm": 0.35226690769195557,
+      "learning_rate": 0.001,
+      "loss": 2.3944,
+      "step": 1553
+    },
+    {
+      "epoch": 0.065741602504442,
+      "grad_norm": 0.254507452249527,
+      "learning_rate": 0.001,
+      "loss": 2.4492,
+      "step": 1554
+    },
+    {
+      "epoch": 0.06578390726795837,
+      "grad_norm": 1.1583757400512695,
+      "learning_rate": 0.001,
+      "loss": 2.0208,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06582621203147475,
+      "grad_norm": 0.27121299505233765,
+      "learning_rate": 0.001,
+      "loss": 2.2889,
+      "step": 1556
+    },
+    {
+      "epoch": 0.06586851679499112,
+      "grad_norm": 0.6531930565834045,
+      "learning_rate": 0.001,
+      "loss": 3.3624,
+      "step": 1557
+    },
+    {
+      "epoch": 0.06591082155850748,
+      "grad_norm": 0.42395731806755066,
+      "learning_rate": 0.001,
+      "loss": 2.3959,
+      "step": 1558
+    },
+    {
+      "epoch": 0.06595312632202387,
+      "grad_norm": 0.3243398368358612,
+      "learning_rate": 0.001,
+      "loss": 2.0876,
+      "step": 1559
+    },
+    {
+      "epoch": 0.06599543108554023,
+      "grad_norm": 0.3197888135910034,
+      "learning_rate": 0.001,
+      "loss": 2.2143,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0660377358490566,
+      "grad_norm": 0.33776190876960754,
+      "learning_rate": 0.001,
+      "loss": 2.6401,
+      "step": 1561
+    },
+    {
+      "epoch": 0.06608004061257297,
+      "grad_norm": 0.34774449467658997,
+      "learning_rate": 0.001,
+      "loss": 2.5507,
+      "step": 1562
+    },
+    {
+      "epoch": 0.06612234537608935,
+      "grad_norm": 0.5661643147468567,
+      "learning_rate": 0.001,
+      "loss": 2.9336,
+      "step": 1563
+    },
+    {
+      "epoch": 0.06616465013960572,
+      "grad_norm": 0.3465413749217987,
+      "learning_rate": 0.001,
+      "loss": 1.8256,
+      "step": 1564
+    },
+    {
+      "epoch": 0.06620695490312209,
+      "grad_norm": 0.3140278458595276,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 1565
+    },
+    {
+      "epoch": 0.06624925966663846,
+      "grad_norm": 0.6140979528427124,
+      "learning_rate": 0.001,
+      "loss": 2.4981,
+      "step": 1566
+    },
+    {
+      "epoch": 0.06629156443015484,
+      "grad_norm": 0.36587056517601013,
+      "learning_rate": 0.001,
+      "loss": 2.1976,
+      "step": 1567
+    },
+    {
+      "epoch": 0.06633386919367121,
+      "grad_norm": 0.2791414260864258,
+      "learning_rate": 0.001,
+      "loss": 2.1657,
+      "step": 1568
+    },
+    {
+      "epoch": 0.06637617395718758,
+      "grad_norm": 0.974829375743866,
+      "learning_rate": 0.001,
+      "loss": 2.604,
+      "step": 1569
+    },
+    {
+      "epoch": 0.06641847872070396,
+      "grad_norm": 0.3969157934188843,
+      "learning_rate": 0.001,
+      "loss": 3.1704,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06646078348422033,
+      "grad_norm": 0.31570157408714294,
+      "learning_rate": 0.001,
+      "loss": 2.6029,
+      "step": 1571
+    },
+    {
+      "epoch": 0.0665030882477367,
+      "grad_norm": 0.5296083688735962,
+      "learning_rate": 0.001,
+      "loss": 2.375,
+      "step": 1572
+    },
+    {
+      "epoch": 0.06654539301125306,
+      "grad_norm": 0.8200310468673706,
+      "learning_rate": 0.001,
+      "loss": 1.9998,
+      "step": 1573
+    },
+    {
+      "epoch": 0.06658769777476944,
+      "grad_norm": 0.3001185953617096,
+      "learning_rate": 0.001,
+      "loss": 2.5948,
+      "step": 1574
+    },
+    {
+      "epoch": 0.06663000253828581,
+      "grad_norm": 0.7405486106872559,
+      "learning_rate": 0.001,
+      "loss": 3.2701,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06667230730180218,
+      "grad_norm": 1.4987202882766724,
+      "learning_rate": 0.001,
+      "loss": 3.5723,
+      "step": 1576
+    },
+    {
+      "epoch": 0.06671461206531855,
+      "grad_norm": 0.7644462585449219,
+      "learning_rate": 0.001,
+      "loss": 2.1558,
+      "step": 1577
+    },
+    {
+      "epoch": 0.06675691682883493,
+      "grad_norm": 0.3527809977531433,
+      "learning_rate": 0.001,
+      "loss": 2.5948,
+      "step": 1578
+    },
+    {
+      "epoch": 0.0667992215923513,
+      "grad_norm": 1.6607075929641724,
+      "learning_rate": 0.001,
+      "loss": 2.2302,
+      "step": 1579
+    },
+    {
+      "epoch": 0.06684152635586767,
+      "grad_norm": 0.42727428674697876,
+      "learning_rate": 0.001,
+      "loss": 2.4521,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06688383111938405,
+      "grad_norm": 0.5329534411430359,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 1581
+    },
+    {
+      "epoch": 0.06692613588290042,
+      "grad_norm": 0.32826146483421326,
+      "learning_rate": 0.001,
+      "loss": 3.578,
+      "step": 1582
+    },
+    {
+      "epoch": 0.06696844064641679,
+      "grad_norm": 0.5298377275466919,
+      "learning_rate": 0.001,
+      "loss": 2.7983,
+      "step": 1583
+    },
+    {
+      "epoch": 0.06701074540993315,
+      "grad_norm": 0.3430579900741577,
+      "learning_rate": 0.001,
+      "loss": 1.9132,
+      "step": 1584
+    },
+    {
+      "epoch": 0.06705305017344954,
+      "grad_norm": 0.31380248069763184,
+      "learning_rate": 0.001,
+      "loss": 2.7339,
+      "step": 1585
+    },
+    {
+      "epoch": 0.0670953549369659,
+      "grad_norm": 0.4660479724407196,
+      "learning_rate": 0.001,
+      "loss": 2.183,
+      "step": 1586
+    },
+    {
+      "epoch": 0.06713765970048227,
+      "grad_norm": 1.5428110361099243,
+      "learning_rate": 0.001,
+      "loss": 2.3808,
+      "step": 1587
+    },
+    {
+      "epoch": 0.06717996446399864,
+      "grad_norm": 1.2535374164581299,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 1588
+    },
+    {
+      "epoch": 0.06722226922751502,
+      "grad_norm": 0.438996285200119,
+      "learning_rate": 0.001,
+      "loss": 2.3571,
+      "step": 1589
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 0.3860979378223419,
+      "learning_rate": 0.001,
+      "loss": 3.1402,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06730687875454776,
+      "grad_norm": 0.32492074370384216,
+      "learning_rate": 0.001,
+      "loss": 2.5053,
+      "step": 1591
+    },
+    {
+      "epoch": 0.06734918351806414,
+      "grad_norm": 0.4713243246078491,
+      "learning_rate": 0.001,
+      "loss": 3.3592,
+      "step": 1592
+    },
+    {
+      "epoch": 0.06739148828158051,
+      "grad_norm": 0.30050158500671387,
+      "learning_rate": 0.001,
+      "loss": 2.1694,
+      "step": 1593
+    },
+    {
+      "epoch": 0.06743379304509688,
+      "grad_norm": 0.34924885630607605,
+      "learning_rate": 0.001,
+      "loss": 1.9499,
+      "step": 1594
+    },
+    {
+      "epoch": 0.06747609780861324,
+      "grad_norm": 0.4567183256149292,
+      "learning_rate": 0.001,
+      "loss": 2.2653,
+      "step": 1595
+    },
+    {
+      "epoch": 0.06751840257212963,
+      "grad_norm": 4.099420070648193,
+      "learning_rate": 0.001,
+      "loss": 2.9475,
+      "step": 1596
+    },
+    {
+      "epoch": 0.067560707335646,
+      "grad_norm": 0.3680780827999115,
+      "learning_rate": 0.001,
+      "loss": 2.3772,
+      "step": 1597
+    },
+    {
+      "epoch": 0.06760301209916236,
+      "grad_norm": 3.2965519428253174,
+      "learning_rate": 0.001,
+      "loss": 1.8148,
+      "step": 1598
+    },
+    {
+      "epoch": 0.06764531686267873,
+      "grad_norm": 0.4864022433757782,
+      "learning_rate": 0.001,
+      "loss": 2.3886,
+      "step": 1599
+    },
+    {
+      "epoch": 0.06768762162619511,
+      "grad_norm": 0.5703839063644409,
+      "learning_rate": 0.001,
+      "loss": 2.1813,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06772992638971148,
+      "grad_norm": 0.28068217635154724,
+      "learning_rate": 0.001,
+      "loss": 2.4237,
+      "step": 1601
+    },
+    {
+      "epoch": 0.06777223115322785,
+      "grad_norm": 0.27317923307418823,
+      "learning_rate": 0.001,
+      "loss": 2.2093,
+      "step": 1602
+    },
+    {
+      "epoch": 0.06781453591674423,
+      "grad_norm": 0.3868054151535034,
+      "learning_rate": 0.001,
+      "loss": 2.4154,
+      "step": 1603
+    },
+    {
+      "epoch": 0.0678568406802606,
+      "grad_norm": 0.36695969104766846,
+      "learning_rate": 0.001,
+      "loss": 2.3344,
+      "step": 1604
+    },
+    {
+      "epoch": 0.06789914544377697,
+      "grad_norm": 0.37811872363090515,
+      "learning_rate": 0.001,
+      "loss": 1.7114,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06794145020729334,
+      "grad_norm": 0.3852631449699402,
+      "learning_rate": 0.001,
+      "loss": 3.0529,
+      "step": 1606
+    },
+    {
+      "epoch": 0.06798375497080972,
+      "grad_norm": 0.9019683003425598,
+      "learning_rate": 0.001,
+      "loss": 2.0123,
+      "step": 1607
+    },
+    {
+      "epoch": 0.06802605973432609,
+      "grad_norm": 1.809870958328247,
+      "learning_rate": 0.001,
+      "loss": 2.5859,
+      "step": 1608
+    },
+    {
+      "epoch": 0.06806836449784245,
+      "grad_norm": 0.31933197379112244,
+      "learning_rate": 0.001,
+      "loss": 2.0505,
+      "step": 1609
+    },
+    {
+      "epoch": 0.06811066926135882,
+      "grad_norm": 0.37392449378967285,
+      "learning_rate": 0.001,
+      "loss": 3.3611,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0681529740248752,
+      "grad_norm": 0.3455560803413391,
+      "learning_rate": 0.001,
+      "loss": 2.3174,
+      "step": 1611
+    },
+    {
+      "epoch": 0.06819527878839157,
+      "grad_norm": 0.5688546299934387,
+      "learning_rate": 0.001,
+      "loss": 2.2655,
+      "step": 1612
+    },
+    {
+      "epoch": 0.06823758355190794,
+      "grad_norm": 0.6909453868865967,
+      "learning_rate": 0.001,
+      "loss": 1.8641,
+      "step": 1613
+    },
+    {
+      "epoch": 0.06827988831542432,
+      "grad_norm": 0.31672239303588867,
+      "learning_rate": 0.001,
+      "loss": 1.9912,
+      "step": 1614
+    },
+    {
+      "epoch": 0.06832219307894069,
+      "grad_norm": 0.8048245310783386,
+      "learning_rate": 0.001,
+      "loss": 2.0888,
+      "step": 1615
+    },
+    {
+      "epoch": 0.06836449784245706,
+      "grad_norm": 0.28497985005378723,
+      "learning_rate": 0.001,
+      "loss": 2.5377,
+      "step": 1616
+    },
+    {
+      "epoch": 0.06840680260597343,
+      "grad_norm": 0.8506115674972534,
+      "learning_rate": 0.001,
+      "loss": 2.3879,
+      "step": 1617
+    },
+    {
+      "epoch": 0.06844910736948981,
+      "grad_norm": 1.2071772813796997,
+      "learning_rate": 0.001,
+      "loss": 2.7813,
+      "step": 1618
+    },
+    {
+      "epoch": 0.06849141213300618,
+      "grad_norm": 0.37529999017715454,
+      "learning_rate": 0.001,
+      "loss": 2.6804,
+      "step": 1619
+    },
+    {
+      "epoch": 0.06853371689652255,
+      "grad_norm": 0.5811559557914734,
+      "learning_rate": 0.001,
+      "loss": 2.3751,
+      "step": 1620
+    },
+    {
+      "epoch": 0.06857602166003893,
+      "grad_norm": 1.8456511497497559,
+      "learning_rate": 0.001,
+      "loss": 2.3837,
+      "step": 1621
+    },
+    {
+      "epoch": 0.0686183264235553,
+      "grad_norm": 0.7621955871582031,
+      "learning_rate": 0.001,
+      "loss": 2.696,
+      "step": 1622
+    },
+    {
+      "epoch": 0.06866063118707166,
+      "grad_norm": 0.8707615733146667,
+      "learning_rate": 0.001,
+      "loss": 2.1565,
+      "step": 1623
+    },
+    {
+      "epoch": 0.06870293595058803,
+      "grad_norm": 0.6202014684677124,
+      "learning_rate": 0.001,
+      "loss": 1.9679,
+      "step": 1624
+    },
+    {
+      "epoch": 0.06874524071410441,
+      "grad_norm": 0.38461774587631226,
+      "learning_rate": 0.001,
+      "loss": 2.1663,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06878754547762078,
+      "grad_norm": 0.3248027265071869,
+      "learning_rate": 0.001,
+      "loss": 2.6196,
+      "step": 1626
+    },
+    {
+      "epoch": 0.06882985024113715,
+      "grad_norm": 0.5045840740203857,
+      "learning_rate": 0.001,
+      "loss": 4.0164,
+      "step": 1627
+    },
+    {
+      "epoch": 0.06887215500465352,
+      "grad_norm": 0.31521186232566833,
+      "learning_rate": 0.001,
+      "loss": 1.885,
+      "step": 1628
+    },
+    {
+      "epoch": 0.0689144597681699,
+      "grad_norm": 0.3150392770767212,
+      "learning_rate": 0.001,
+      "loss": 2.2383,
+      "step": 1629
+    },
+    {
+      "epoch": 0.06895676453168627,
+      "grad_norm": 0.45550379157066345,
+      "learning_rate": 0.001,
+      "loss": 2.2817,
+      "step": 1630
+    },
+    {
+      "epoch": 0.06899906929520264,
+      "grad_norm": 0.30408960580825806,
+      "learning_rate": 0.001,
+      "loss": 2.213,
+      "step": 1631
+    },
+    {
+      "epoch": 0.06904137405871902,
+      "grad_norm": 0.36207836866378784,
+      "learning_rate": 0.001,
+      "loss": 2.593,
+      "step": 1632
+    },
+    {
+      "epoch": 0.06908367882223539,
+      "grad_norm": 0.37449923157691956,
+      "learning_rate": 0.001,
+      "loss": 2.0836,
+      "step": 1633
+    },
+    {
+      "epoch": 0.06912598358575175,
+      "grad_norm": 0.29214954376220703,
+      "learning_rate": 0.001,
+      "loss": 2.2552,
+      "step": 1634
+    },
+    {
+      "epoch": 0.06916828834926812,
+      "grad_norm": 0.8513506650924683,
+      "learning_rate": 0.001,
+      "loss": 1.975,
+      "step": 1635
+    },
+    {
+      "epoch": 0.0692105931127845,
+      "grad_norm": 0.5883151888847351,
+      "learning_rate": 0.001,
+      "loss": 3.0927,
+      "step": 1636
+    },
+    {
+      "epoch": 0.06925289787630087,
+      "grad_norm": 0.350691020488739,
+      "learning_rate": 0.001,
+      "loss": 2.0645,
+      "step": 1637
+    },
+    {
+      "epoch": 0.06929520263981724,
+      "grad_norm": 0.8605084419250488,
+      "learning_rate": 0.001,
+      "loss": 1.8086,
+      "step": 1638
+    },
+    {
+      "epoch": 0.06933750740333361,
+      "grad_norm": 1.2016950845718384,
+      "learning_rate": 0.001,
+      "loss": 1.9914,
+      "step": 1639
+    },
+    {
+      "epoch": 0.06937981216684999,
+      "grad_norm": 0.2884148955345154,
+      "learning_rate": 0.001,
+      "loss": 2.1365,
+      "step": 1640
+    },
+    {
+      "epoch": 0.06942211693036636,
+      "grad_norm": 0.337169349193573,
+      "learning_rate": 0.001,
+      "loss": 2.0075,
+      "step": 1641
+    },
+    {
+      "epoch": 0.06946442169388273,
+      "grad_norm": 0.6522495150566101,
+      "learning_rate": 0.001,
+      "loss": 3.0513,
+      "step": 1642
+    },
+    {
+      "epoch": 0.06950672645739911,
+      "grad_norm": 0.44939300417900085,
+      "learning_rate": 0.001,
+      "loss": 2.8989,
+      "step": 1643
+    },
+    {
+      "epoch": 0.06954903122091548,
+      "grad_norm": 0.30578479170799255,
+      "learning_rate": 0.001,
+      "loss": 3.2959,
+      "step": 1644
+    },
+    {
+      "epoch": 0.06959133598443185,
+      "grad_norm": 45.22916793823242,
+      "learning_rate": 0.001,
+      "loss": 2.4688,
+      "step": 1645
+    },
+    {
+      "epoch": 0.06963364074794821,
+      "grad_norm": 0.2697926461696625,
+      "learning_rate": 0.001,
+      "loss": 3.0138,
+      "step": 1646
+    },
+    {
+      "epoch": 0.0696759455114646,
+      "grad_norm": 0.32543253898620605,
+      "learning_rate": 0.001,
+      "loss": 2.884,
+      "step": 1647
+    },
+    {
+      "epoch": 0.06971825027498096,
+      "grad_norm": 1.5244373083114624,
+      "learning_rate": 0.001,
+      "loss": 2.605,
+      "step": 1648
+    },
+    {
+      "epoch": 0.06976055503849733,
+      "grad_norm": 0.2659820318222046,
+      "learning_rate": 0.001,
+      "loss": 2.1016,
+      "step": 1649
+    },
+    {
+      "epoch": 0.0698028598020137,
+      "grad_norm": 0.30054983496665955,
+      "learning_rate": 0.001,
+      "loss": 2.4681,
+      "step": 1650
+    },
+    {
+      "epoch": 0.06984516456553008,
+      "grad_norm": 0.3234836161136627,
+      "learning_rate": 0.001,
+      "loss": 1.9912,
+      "step": 1651
+    },
+    {
+      "epoch": 0.06988746932904645,
+      "grad_norm": 0.27745023369789124,
+      "learning_rate": 0.001,
+      "loss": 2.2214,
+      "step": 1652
+    },
+    {
+      "epoch": 0.06992977409256282,
+      "grad_norm": 0.6211910247802734,
+      "learning_rate": 0.001,
+      "loss": 2.1174,
+      "step": 1653
+    },
+    {
+      "epoch": 0.0699720788560792,
+      "grad_norm": 6.255242824554443,
+      "learning_rate": 0.001,
+      "loss": 2.5594,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07001438361959557,
+      "grad_norm": 0.3245622217655182,
+      "learning_rate": 0.001,
+      "loss": 1.9634,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07005668838311194,
+      "grad_norm": 0.4616556167602539,
+      "learning_rate": 0.001,
+      "loss": 2.2543,
+      "step": 1656
+    },
+    {
+      "epoch": 0.0700989931466283,
+      "grad_norm": 0.2637854516506195,
+      "learning_rate": 0.001,
+      "loss": 2.29,
+      "step": 1657
+    },
+    {
+      "epoch": 0.07014129791014469,
+      "grad_norm": 0.9866844415664673,
+      "learning_rate": 0.001,
+      "loss": 2.8994,
+      "step": 1658
+    },
+    {
+      "epoch": 0.07018360267366106,
+      "grad_norm": 0.7608810663223267,
+      "learning_rate": 0.001,
+      "loss": 2.2228,
+      "step": 1659
+    },
+    {
+      "epoch": 0.07022590743717742,
+      "grad_norm": 0.31685671210289,
+      "learning_rate": 0.001,
+      "loss": 2.2006,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07026821220069379,
+      "grad_norm": 1.3476111888885498,
+      "learning_rate": 0.001,
+      "loss": 3.0012,
+      "step": 1661
+    },
+    {
+      "epoch": 0.07031051696421017,
+      "grad_norm": 5.591330051422119,
+      "learning_rate": 0.001,
+      "loss": 2.1398,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07035282172772654,
+      "grad_norm": 0.6403492093086243,
+      "learning_rate": 0.001,
+      "loss": 3.7632,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07039512649124291,
+      "grad_norm": 0.4311082661151886,
+      "learning_rate": 0.001,
+      "loss": 2.0528,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07043743125475929,
+      "grad_norm": 0.9901355504989624,
+      "learning_rate": 0.001,
+      "loss": 2.5533,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07047973601827566,
+      "grad_norm": 2.0582873821258545,
+      "learning_rate": 0.001,
+      "loss": 3.0359,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07052204078179203,
+      "grad_norm": 0.428562194108963,
+      "learning_rate": 0.001,
+      "loss": 3.0728,
+      "step": 1667
+    },
+    {
+      "epoch": 0.0705643455453084,
+      "grad_norm": 0.2768668532371521,
+      "learning_rate": 0.001,
+      "loss": 1.7863,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07060665030882478,
+      "grad_norm": 0.593871533870697,
+      "learning_rate": 0.001,
+      "loss": 2.8978,
+      "step": 1669
+    },
+    {
+      "epoch": 0.07064895507234115,
+      "grad_norm": 0.41889962553977966,
+      "learning_rate": 0.001,
+      "loss": 2.3901,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07069125983585751,
+      "grad_norm": 0.8739703297615051,
+      "learning_rate": 0.001,
+      "loss": 2.9089,
+      "step": 1671
+    },
+    {
+      "epoch": 0.07073356459937388,
+      "grad_norm": 0.3771756589412689,
+      "learning_rate": 0.001,
+      "loss": 2.338,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07077586936289026,
+      "grad_norm": 0.29926592111587524,
+      "learning_rate": 0.001,
+      "loss": 3.1151,
+      "step": 1673
+    },
+    {
+      "epoch": 0.07081817412640663,
+      "grad_norm": 0.43917933106422424,
+      "learning_rate": 0.001,
+      "loss": 2.3074,
+      "step": 1674
+    },
+    {
+      "epoch": 0.070860478889923,
+      "grad_norm": 4.316728591918945,
+      "learning_rate": 0.001,
+      "loss": 2.0206,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07090278365343938,
+      "grad_norm": 0.5438099503517151,
+      "learning_rate": 0.001,
+      "loss": 2.57,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07094508841695575,
+      "grad_norm": 0.4939629137516022,
+      "learning_rate": 0.001,
+      "loss": 2.4069,
+      "step": 1677
+    },
+    {
+      "epoch": 0.07098739318047212,
+      "grad_norm": 0.4457460343837738,
+      "learning_rate": 0.001,
+      "loss": 2.7809,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07102969794398849,
+      "grad_norm": 0.45648688077926636,
+      "learning_rate": 0.001,
+      "loss": 2.7326,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07107200270750487,
+      "grad_norm": 0.3010053038597107,
+      "learning_rate": 0.001,
+      "loss": 2.2902,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07111430747102124,
+      "grad_norm": 0.42142006754875183,
+      "learning_rate": 0.001,
+      "loss": 1.9107,
+      "step": 1681
+    },
+    {
+      "epoch": 0.0711566122345376,
+      "grad_norm": 0.3928402364253998,
+      "learning_rate": 0.001,
+      "loss": 3.0004,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07119891699805399,
+      "grad_norm": 0.28071898221969604,
+      "learning_rate": 0.001,
+      "loss": 2.289,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07124122176157036,
+      "grad_norm": 0.3247803747653961,
+      "learning_rate": 0.001,
+      "loss": 2.2083,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07128352652508672,
+      "grad_norm": 0.5170036554336548,
+      "learning_rate": 0.001,
+      "loss": 2.1761,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07132583128860309,
+      "grad_norm": 1.5233675241470337,
+      "learning_rate": 0.001,
+      "loss": 2.3741,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07136813605211947,
+      "grad_norm": 0.33674490451812744,
+      "learning_rate": 0.001,
+      "loss": 2.2987,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07141044081563584,
+      "grad_norm": 3.1619656085968018,
+      "learning_rate": 0.001,
+      "loss": 1.9981,
+      "step": 1688
+    },
+    {
+      "epoch": 0.07145274557915221,
+      "grad_norm": 0.3328041732311249,
+      "learning_rate": 0.001,
+      "loss": 3.4872,
+      "step": 1689
+    },
+    {
+      "epoch": 0.07149505034266858,
+      "grad_norm": 0.3385929465293884,
+      "learning_rate": 0.001,
+      "loss": 2.4917,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07153735510618496,
+      "grad_norm": 0.3292979300022125,
+      "learning_rate": 0.001,
+      "loss": 1.7646,
+      "step": 1691
+    },
+    {
+      "epoch": 0.07157965986970133,
+      "grad_norm": 9.468561172485352,
+      "learning_rate": 0.001,
+      "loss": 3.2305,
+      "step": 1692
+    },
+    {
+      "epoch": 0.0716219646332177,
+      "grad_norm": 0.339837908744812,
+      "learning_rate": 0.001,
+      "loss": 4.1912,
+      "step": 1693
+    },
+    {
+      "epoch": 0.07166426939673408,
+      "grad_norm": 0.7992456555366516,
+      "learning_rate": 0.001,
+      "loss": 1.7365,
+      "step": 1694
+    },
+    {
+      "epoch": 0.07170657416025045,
+      "grad_norm": 0.3533322513103485,
+      "learning_rate": 0.001,
+      "loss": 1.6018,
+      "step": 1695
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 0.4883526563644409,
+      "learning_rate": 0.001,
+      "loss": 2.5816,
+      "step": 1696
+    },
+    {
+      "epoch": 0.07179118368728318,
+      "grad_norm": 0.3907789885997772,
+      "learning_rate": 0.001,
+      "loss": 1.906,
+      "step": 1697
+    },
+    {
+      "epoch": 0.07183348845079957,
+      "grad_norm": 0.5203762054443359,
+      "learning_rate": 0.001,
+      "loss": 2.2438,
+      "step": 1698
+    },
+    {
+      "epoch": 0.07187579321431593,
+      "grad_norm": 1.2718244791030884,
+      "learning_rate": 0.001,
+      "loss": 2.1247,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0719180979778323,
+      "grad_norm": 0.3142852485179901,
+      "learning_rate": 0.001,
+      "loss": 2.0239,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07196040274134867,
+      "grad_norm": 0.4279821515083313,
+      "learning_rate": 0.001,
+      "loss": 2.2428,
+      "step": 1701
+    },
+    {
+      "epoch": 0.07200270750486505,
+      "grad_norm": 0.6683199405670166,
+      "learning_rate": 0.001,
+      "loss": 2.4512,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07204501226838142,
+      "grad_norm": 2.1548545360565186,
+      "learning_rate": 0.001,
+      "loss": 2.6076,
+      "step": 1703
+    },
+    {
+      "epoch": 0.07208731703189779,
+      "grad_norm": 0.5022167563438416,
+      "learning_rate": 0.001,
+      "loss": 2.6147,
+      "step": 1704
+    },
+    {
+      "epoch": 0.07212962179541417,
+      "grad_norm": 4.610467910766602,
+      "learning_rate": 0.001,
+      "loss": 2.8257,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07217192655893054,
+      "grad_norm": 0.35856351256370544,
+      "learning_rate": 0.001,
+      "loss": 2.0853,
+      "step": 1706
+    },
+    {
+      "epoch": 0.0722142313224469,
+      "grad_norm": 0.45485618710517883,
+      "learning_rate": 0.001,
+      "loss": 2.5217,
+      "step": 1707
+    },
+    {
+      "epoch": 0.07225653608596327,
+      "grad_norm": 0.3743135333061218,
+      "learning_rate": 0.001,
+      "loss": 2.2795,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07229884084947966,
+      "grad_norm": 0.286580890417099,
+      "learning_rate": 0.001,
+      "loss": 2.0651,
+      "step": 1709
+    },
+    {
+      "epoch": 0.07234114561299602,
+      "grad_norm": 0.574466347694397,
+      "learning_rate": 0.001,
+      "loss": 2.1261,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07238345037651239,
+      "grad_norm": 0.35179415345191956,
+      "learning_rate": 0.001,
+      "loss": 2.6743,
+      "step": 1711
+    },
+    {
+      "epoch": 0.07242575514002876,
+      "grad_norm": 0.3282983601093292,
+      "learning_rate": 0.001,
+      "loss": 2.7438,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07246805990354514,
+      "grad_norm": 0.41283783316612244,
+      "learning_rate": 0.001,
+      "loss": 2.4672,
+      "step": 1713
+    },
+    {
+      "epoch": 0.07251036466706151,
+      "grad_norm": 0.4331008791923523,
+      "learning_rate": 0.001,
+      "loss": 2.4509,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07255266943057788,
+      "grad_norm": 0.4134253263473511,
+      "learning_rate": 0.001,
+      "loss": 2.7329,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07259497419409426,
+      "grad_norm": 0.3073793649673462,
+      "learning_rate": 0.001,
+      "loss": 2.2086,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07263727895761063,
+      "grad_norm": 0.3595843017101288,
+      "learning_rate": 0.001,
+      "loss": 2.2972,
+      "step": 1717
+    },
+    {
+      "epoch": 0.072679583721127,
+      "grad_norm": 1.8754022121429443,
+      "learning_rate": 0.001,
+      "loss": 2.6782,
+      "step": 1718
+    },
+    {
+      "epoch": 0.07272188848464337,
+      "grad_norm": 0.2694728374481201,
+      "learning_rate": 0.001,
+      "loss": 2.5075,
+      "step": 1719
+    },
+    {
+      "epoch": 0.07276419324815975,
+      "grad_norm": 0.266966849565506,
+      "learning_rate": 0.001,
+      "loss": 2.7838,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07280649801167612,
+      "grad_norm": 0.6708036661148071,
+      "learning_rate": 0.001,
+      "loss": 2.5704,
+      "step": 1721
+    },
+    {
+      "epoch": 0.07284880277519248,
+      "grad_norm": 0.32971417903900146,
+      "learning_rate": 0.001,
+      "loss": 2.1911,
+      "step": 1722
+    },
+    {
+      "epoch": 0.07289110753870885,
+      "grad_norm": 0.5026300549507141,
+      "learning_rate": 0.001,
+      "loss": 2.1365,
+      "step": 1723
+    },
+    {
+      "epoch": 0.07293341230222523,
+      "grad_norm": 0.6025309562683105,
+      "learning_rate": 0.001,
+      "loss": 2.7638,
+      "step": 1724
+    },
+    {
+      "epoch": 0.0729757170657416,
+      "grad_norm": 0.35378944873809814,
+      "learning_rate": 0.001,
+      "loss": 2.5012,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07301802182925797,
+      "grad_norm": 0.2955608665943146,
+      "learning_rate": 0.001,
+      "loss": 3.3105,
+      "step": 1726
+    },
+    {
+      "epoch": 0.07306032659277435,
+      "grad_norm": 1.2950598001480103,
+      "learning_rate": 0.001,
+      "loss": 1.4412,
+      "step": 1727
+    },
+    {
+      "epoch": 0.07310263135629072,
+      "grad_norm": 0.34854191541671753,
+      "learning_rate": 0.001,
+      "loss": 2.1458,
+      "step": 1728
+    },
+    {
+      "epoch": 0.07314493611980709,
+      "grad_norm": 0.2604203224182129,
+      "learning_rate": 0.001,
+      "loss": 2.5409,
+      "step": 1729
+    },
+    {
+      "epoch": 0.07318724088332346,
+      "grad_norm": 0.28911593556404114,
+      "learning_rate": 0.001,
+      "loss": 2.3813,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07322954564683984,
+      "grad_norm": 0.4041121304035187,
+      "learning_rate": 0.001,
+      "loss": 2.1355,
+      "step": 1731
+    },
+    {
+      "epoch": 0.0732718504103562,
+      "grad_norm": 3.5361766815185547,
+      "learning_rate": 0.001,
+      "loss": 2.6676,
+      "step": 1732
+    },
+    {
+      "epoch": 0.07331415517387257,
+      "grad_norm": 0.38924169540405273,
+      "learning_rate": 0.001,
+      "loss": 2.1344,
+      "step": 1733
+    },
+    {
+      "epoch": 0.07335645993738894,
+      "grad_norm": 0.36835405230522156,
+      "learning_rate": 0.001,
+      "loss": 2.3473,
+      "step": 1734
+    },
+    {
+      "epoch": 0.07339876470090533,
+      "grad_norm": 2.3047337532043457,
+      "learning_rate": 0.001,
+      "loss": 2.1468,
+      "step": 1735
+    },
+    {
+      "epoch": 0.0734410694644217,
+      "grad_norm": 0.47445395588874817,
+      "learning_rate": 0.001,
+      "loss": 2.6338,
+      "step": 1736
+    },
+    {
+      "epoch": 0.07348337422793806,
+      "grad_norm": 0.5361562967300415,
+      "learning_rate": 0.001,
+      "loss": 2.8039,
+      "step": 1737
+    },
+    {
+      "epoch": 0.07352567899145444,
+      "grad_norm": 0.362362802028656,
+      "learning_rate": 0.001,
+      "loss": 1.9508,
+      "step": 1738
+    },
+    {
+      "epoch": 0.07356798375497081,
+      "grad_norm": 1.4122487306594849,
+      "learning_rate": 0.001,
+      "loss": 2.8017,
+      "step": 1739
+    },
+    {
+      "epoch": 0.07361028851848718,
+      "grad_norm": 0.5565959215164185,
+      "learning_rate": 0.001,
+      "loss": 2.2053,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07365259328200355,
+      "grad_norm": 0.3117307722568512,
+      "learning_rate": 0.001,
+      "loss": 2.1428,
+      "step": 1741
+    },
+    {
+      "epoch": 0.07369489804551993,
+      "grad_norm": 0.4470238983631134,
+      "learning_rate": 0.001,
+      "loss": 3.3104,
+      "step": 1742
+    },
+    {
+      "epoch": 0.0737372028090363,
+      "grad_norm": 0.31840646266937256,
+      "learning_rate": 0.001,
+      "loss": 3.3071,
+      "step": 1743
+    },
+    {
+      "epoch": 0.07377950757255267,
+      "grad_norm": 0.29003921151161194,
+      "learning_rate": 0.001,
+      "loss": 1.8832,
+      "step": 1744
+    },
+    {
+      "epoch": 0.07382181233606905,
+      "grad_norm": 0.2651999890804291,
+      "learning_rate": 0.001,
+      "loss": 2.1535,
+      "step": 1745
+    },
+    {
+      "epoch": 0.07386411709958542,
+      "grad_norm": 0.34940871596336365,
+      "learning_rate": 0.001,
+      "loss": 2.5327,
+      "step": 1746
+    },
+    {
+      "epoch": 0.07390642186310178,
+      "grad_norm": 0.6055050492286682,
+      "learning_rate": 0.001,
+      "loss": 2.8376,
+      "step": 1747
+    },
+    {
+      "epoch": 0.07394872662661815,
+      "grad_norm": 0.372227281332016,
+      "learning_rate": 0.001,
+      "loss": 2.1704,
+      "step": 1748
+    },
+    {
+      "epoch": 0.07399103139013453,
+      "grad_norm": 0.31508174538612366,
+      "learning_rate": 0.001,
+      "loss": 2.3462,
+      "step": 1749
+    },
+    {
+      "epoch": 0.0740333361536509,
+      "grad_norm": 0.40334755182266235,
+      "learning_rate": 0.001,
+      "loss": 2.2007,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07407564091716727,
+      "grad_norm": 0.23412242531776428,
+      "learning_rate": 0.001,
+      "loss": 1.9882,
+      "step": 1751
+    },
+    {
+      "epoch": 0.07411794568068364,
+      "grad_norm": 0.47561314702033997,
+      "learning_rate": 0.001,
+      "loss": 2.3195,
+      "step": 1752
+    },
+    {
+      "epoch": 0.07416025044420002,
+      "grad_norm": 0.3126150965690613,
+      "learning_rate": 0.001,
+      "loss": 3.2531,
+      "step": 1753
+    },
+    {
+      "epoch": 0.07420255520771639,
+      "grad_norm": 0.6425087451934814,
+      "learning_rate": 0.001,
+      "loss": 1.549,
+      "step": 1754
+    },
+    {
+      "epoch": 0.07424485997123276,
+      "grad_norm": 0.3294941484928131,
+      "learning_rate": 0.001,
+      "loss": 3.2919,
+      "step": 1755
+    },
+    {
+      "epoch": 0.07428716473474914,
+      "grad_norm": 0.32722482085227966,
+      "learning_rate": 0.001,
+      "loss": 2.3305,
+      "step": 1756
+    },
+    {
+      "epoch": 0.07432946949826551,
+      "grad_norm": 0.30937889218330383,
+      "learning_rate": 0.001,
+      "loss": 1.9746,
+      "step": 1757
+    },
+    {
+      "epoch": 0.07437177426178188,
+      "grad_norm": 1.4851760864257812,
+      "learning_rate": 0.001,
+      "loss": 2.4941,
+      "step": 1758
+    },
+    {
+      "epoch": 0.07441407902529824,
+      "grad_norm": 0.2669662833213806,
+      "learning_rate": 0.001,
+      "loss": 2.0385,
+      "step": 1759
+    },
+    {
+      "epoch": 0.07445638378881463,
+      "grad_norm": 0.44635874032974243,
+      "learning_rate": 0.001,
+      "loss": 2.5292,
+      "step": 1760
+    },
+    {
+      "epoch": 0.074498688552331,
+      "grad_norm": 0.28977322578430176,
+      "learning_rate": 0.001,
+      "loss": 2.7165,
+      "step": 1761
+    },
+    {
+      "epoch": 0.07454099331584736,
+      "grad_norm": 0.25588658452033997,
+      "learning_rate": 0.001,
+      "loss": 2.1741,
+      "step": 1762
+    },
+    {
+      "epoch": 0.07458329807936373,
+      "grad_norm": 0.28512755036354065,
+      "learning_rate": 0.001,
+      "loss": 2.8152,
+      "step": 1763
+    },
+    {
+      "epoch": 0.07462560284288011,
+      "grad_norm": 0.8499886393547058,
+      "learning_rate": 0.001,
+      "loss": 2.7018,
+      "step": 1764
+    },
+    {
+      "epoch": 0.07466790760639648,
+      "grad_norm": 0.5321717262268066,
+      "learning_rate": 0.001,
+      "loss": 2.3566,
+      "step": 1765
+    },
+    {
+      "epoch": 0.07471021236991285,
+      "grad_norm": 0.3150945007801056,
+      "learning_rate": 0.001,
+      "loss": 2.2778,
+      "step": 1766
+    },
+    {
+      "epoch": 0.07475251713342923,
+      "grad_norm": 1.8098750114440918,
+      "learning_rate": 0.001,
+      "loss": 2.2745,
+      "step": 1767
+    },
+    {
+      "epoch": 0.0747948218969456,
+      "grad_norm": 0.2686639726161957,
+      "learning_rate": 0.001,
+      "loss": 3.0812,
+      "step": 1768
+    },
+    {
+      "epoch": 0.07483712666046197,
+      "grad_norm": 0.3387921154499054,
+      "learning_rate": 0.001,
+      "loss": 2.8251,
+      "step": 1769
+    },
+    {
+      "epoch": 0.07487943142397833,
+      "grad_norm": 0.5623756647109985,
+      "learning_rate": 0.001,
+      "loss": 2.115,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07492173618749472,
+      "grad_norm": 0.2617107927799225,
+      "learning_rate": 0.001,
+      "loss": 2.468,
+      "step": 1771
+    },
+    {
+      "epoch": 0.07496404095101109,
+      "grad_norm": 0.6408830285072327,
+      "learning_rate": 0.001,
+      "loss": 2.4926,
+      "step": 1772
+    },
+    {
+      "epoch": 0.07500634571452745,
+      "grad_norm": 0.38034066557884216,
+      "learning_rate": 0.001,
+      "loss": 2.867,
+      "step": 1773
+    },
+    {
+      "epoch": 0.07504865047804382,
+      "grad_norm": 0.39489755034446716,
+      "learning_rate": 0.001,
+      "loss": 3.4283,
+      "step": 1774
+    },
+    {
+      "epoch": 0.0750909552415602,
+      "grad_norm": 2.249427318572998,
+      "learning_rate": 0.001,
+      "loss": 2.2082,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07513326000507657,
+      "grad_norm": 4.038248062133789,
+      "learning_rate": 0.001,
+      "loss": 3.7159,
+      "step": 1776
+    },
+    {
+      "epoch": 0.07517556476859294,
+      "grad_norm": 0.28461596369743347,
+      "learning_rate": 0.001,
+      "loss": 2.4496,
+      "step": 1777
+    },
+    {
+      "epoch": 0.07521786953210932,
+      "grad_norm": 0.3818615674972534,
+      "learning_rate": 0.001,
+      "loss": 3.797,
+      "step": 1778
+    },
+    {
+      "epoch": 0.07526017429562569,
+      "grad_norm": 0.3321216404438019,
+      "learning_rate": 0.001,
+      "loss": 2.7859,
+      "step": 1779
+    },
+    {
+      "epoch": 0.07530247905914206,
+      "grad_norm": 0.394345223903656,
+      "learning_rate": 0.001,
+      "loss": 2.7472,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07534478382265843,
+      "grad_norm": 0.32292380928993225,
+      "learning_rate": 0.001,
+      "loss": 2.2914,
+      "step": 1781
+    },
+    {
+      "epoch": 0.07538708858617481,
+      "grad_norm": 0.24881044030189514,
+      "learning_rate": 0.001,
+      "loss": 1.5237,
+      "step": 1782
+    },
+    {
+      "epoch": 0.07542939334969118,
+      "grad_norm": 0.2507427930831909,
+      "learning_rate": 0.001,
+      "loss": 2.339,
+      "step": 1783
+    },
+    {
+      "epoch": 0.07547169811320754,
+      "grad_norm": 0.340909868478775,
+      "learning_rate": 0.001,
+      "loss": 3.373,
+      "step": 1784
+    },
+    {
+      "epoch": 0.07551400287672391,
+      "grad_norm": 0.32472652196884155,
+      "learning_rate": 0.001,
+      "loss": 2.0487,
+      "step": 1785
+    },
+    {
+      "epoch": 0.0755563076402403,
+      "grad_norm": 0.3497520983219147,
+      "learning_rate": 0.001,
+      "loss": 3.3913,
+      "step": 1786
+    },
+    {
+      "epoch": 0.07559861240375666,
+      "grad_norm": 1.4078129529953003,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 1787
+    },
+    {
+      "epoch": 0.07564091716727303,
+      "grad_norm": 0.5234339237213135,
+      "learning_rate": 0.001,
+      "loss": 3.5763,
+      "step": 1788
+    },
+    {
+      "epoch": 0.07568322193078941,
+      "grad_norm": 0.32184237241744995,
+      "learning_rate": 0.001,
+      "loss": 2.3567,
+      "step": 1789
+    },
+    {
+      "epoch": 0.07572552669430578,
+      "grad_norm": 0.32312852144241333,
+      "learning_rate": 0.001,
+      "loss": 1.9989,
+      "step": 1790
+    },
+    {
+      "epoch": 0.07576783145782215,
+      "grad_norm": 0.47937512397766113,
+      "learning_rate": 0.001,
+      "loss": 3.2448,
+      "step": 1791
+    },
+    {
+      "epoch": 0.07581013622133852,
+      "grad_norm": 0.6073363423347473,
+      "learning_rate": 0.001,
+      "loss": 2.1826,
+      "step": 1792
+    },
+    {
+      "epoch": 0.0758524409848549,
+      "grad_norm": 0.24768002331256866,
+      "learning_rate": 0.001,
+      "loss": 2.2067,
+      "step": 1793
+    },
+    {
+      "epoch": 0.07589474574837127,
+      "grad_norm": 0.7116523385047913,
+      "learning_rate": 0.001,
+      "loss": 2.6321,
+      "step": 1794
+    },
+    {
+      "epoch": 0.07593705051188764,
+      "grad_norm": 0.3589756488800049,
+      "learning_rate": 0.001,
+      "loss": 3.1839,
+      "step": 1795
+    },
+    {
+      "epoch": 0.075979355275404,
+      "grad_norm": 0.3133421540260315,
+      "learning_rate": 0.001,
+      "loss": 2.097,
+      "step": 1796
+    },
+    {
+      "epoch": 0.07602166003892039,
+      "grad_norm": 0.28984275460243225,
+      "learning_rate": 0.001,
+      "loss": 3.0274,
+      "step": 1797
+    },
+    {
+      "epoch": 0.07606396480243675,
+      "grad_norm": 0.3593312203884125,
+      "learning_rate": 0.001,
+      "loss": 2.0219,
+      "step": 1798
+    },
+    {
+      "epoch": 0.07610626956595312,
+      "grad_norm": 0.4126046299934387,
+      "learning_rate": 0.001,
+      "loss": 1.9161,
+      "step": 1799
+    },
+    {
+      "epoch": 0.0761485743294695,
+      "grad_norm": 0.7341791987419128,
+      "learning_rate": 0.001,
+      "loss": 2.4186,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07619087909298587,
+      "grad_norm": 0.27145180106163025,
+      "learning_rate": 0.001,
+      "loss": 1.9538,
+      "step": 1801
+    },
+    {
+      "epoch": 0.07623318385650224,
+      "grad_norm": 0.5113809108734131,
+      "learning_rate": 0.001,
+      "loss": 2.2766,
+      "step": 1802
+    },
+    {
+      "epoch": 0.07627548862001861,
+      "grad_norm": 0.27664855122566223,
+      "learning_rate": 0.001,
+      "loss": 2.3126,
+      "step": 1803
+    },
+    {
+      "epoch": 0.07631779338353499,
+      "grad_norm": 0.23161768913269043,
+      "learning_rate": 0.001,
+      "loss": 1.7487,
+      "step": 1804
+    },
+    {
+      "epoch": 0.07636009814705136,
+      "grad_norm": 1.3314400911331177,
+      "learning_rate": 0.001,
+      "loss": 2.0632,
+      "step": 1805
+    },
+    {
+      "epoch": 0.07640240291056773,
+      "grad_norm": 0.3629261553287506,
+      "learning_rate": 0.001,
+      "loss": 2.7344,
+      "step": 1806
+    },
+    {
+      "epoch": 0.07644470767408411,
+      "grad_norm": 0.2677980363368988,
+      "learning_rate": 0.001,
+      "loss": 1.9242,
+      "step": 1807
+    },
+    {
+      "epoch": 0.07648701243760048,
+      "grad_norm": 0.26410120725631714,
+      "learning_rate": 0.001,
+      "loss": 2.2882,
+      "step": 1808
+    },
+    {
+      "epoch": 0.07652931720111684,
+      "grad_norm": 1.109229326248169,
+      "learning_rate": 0.001,
+      "loss": 2.2232,
+      "step": 1809
+    },
+    {
+      "epoch": 0.07657162196463321,
+      "grad_norm": 0.533427357673645,
+      "learning_rate": 0.001,
+      "loss": 2.5592,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0766139267281496,
+      "grad_norm": 0.32029929757118225,
+      "learning_rate": 0.001,
+      "loss": 2.2293,
+      "step": 1811
+    },
+    {
+      "epoch": 0.07665623149166596,
+      "grad_norm": 0.3942636251449585,
+      "learning_rate": 0.001,
+      "loss": 2.5888,
+      "step": 1812
+    },
+    {
+      "epoch": 0.07669853625518233,
+      "grad_norm": 0.28876644372940063,
+      "learning_rate": 0.001,
+      "loss": 2.1704,
+      "step": 1813
+    },
+    {
+      "epoch": 0.0767408410186987,
+      "grad_norm": 0.30077075958251953,
+      "learning_rate": 0.001,
+      "loss": 3.1774,
+      "step": 1814
+    },
+    {
+      "epoch": 0.07678314578221508,
+      "grad_norm": 0.6673346161842346,
+      "learning_rate": 0.001,
+      "loss": 2.055,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07682545054573145,
+      "grad_norm": 0.2500164806842804,
+      "learning_rate": 0.001,
+      "loss": 2.6555,
+      "step": 1816
+    },
+    {
+      "epoch": 0.07686775530924782,
+      "grad_norm": 0.2818383574485779,
+      "learning_rate": 0.001,
+      "loss": 2.6336,
+      "step": 1817
+    },
+    {
+      "epoch": 0.0769100600727642,
+      "grad_norm": 0.4379955232143402,
+      "learning_rate": 0.001,
+      "loss": 1.9602,
+      "step": 1818
+    },
+    {
+      "epoch": 0.07695236483628057,
+      "grad_norm": 0.27242425084114075,
+      "learning_rate": 0.001,
+      "loss": 2.0259,
+      "step": 1819
+    },
+    {
+      "epoch": 0.07699466959979694,
+      "grad_norm": 0.2450658231973648,
+      "learning_rate": 0.001,
+      "loss": 2.2079,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0770369743633133,
+      "grad_norm": 0.3834247291088104,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 1821
+    },
+    {
+      "epoch": 0.07707927912682969,
+      "grad_norm": 0.3078935742378235,
+      "learning_rate": 0.001,
+      "loss": 2.2373,
+      "step": 1822
+    },
+    {
+      "epoch": 0.07712158389034605,
+      "grad_norm": 0.31686580181121826,
+      "learning_rate": 0.001,
+      "loss": 2.4674,
+      "step": 1823
+    },
+    {
+      "epoch": 0.07716388865386242,
+      "grad_norm": 0.28138643503189087,
+      "learning_rate": 0.001,
+      "loss": 1.9822,
+      "step": 1824
+    },
+    {
+      "epoch": 0.07720619341737879,
+      "grad_norm": 0.6008328795433044,
+      "learning_rate": 0.001,
+      "loss": 2.6718,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07724849818089517,
+      "grad_norm": 0.3610805869102478,
+      "learning_rate": 0.001,
+      "loss": 3.121,
+      "step": 1826
+    },
+    {
+      "epoch": 0.07729080294441154,
+      "grad_norm": 0.44791126251220703,
+      "learning_rate": 0.001,
+      "loss": 2.4597,
+      "step": 1827
+    },
+    {
+      "epoch": 0.07733310770792791,
+      "grad_norm": 0.6357278823852539,
+      "learning_rate": 0.001,
+      "loss": 2.6391,
+      "step": 1828
+    },
+    {
+      "epoch": 0.07737541247144429,
+      "grad_norm": 0.47252053022384644,
+      "learning_rate": 0.001,
+      "loss": 2.4176,
+      "step": 1829
+    },
+    {
+      "epoch": 0.07741771723496066,
+      "grad_norm": 0.3132396340370178,
+      "learning_rate": 0.001,
+      "loss": 2.105,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07746002199847703,
+      "grad_norm": 0.2682265341281891,
+      "learning_rate": 0.001,
+      "loss": 2.9813,
+      "step": 1831
+    },
+    {
+      "epoch": 0.0775023267619934,
+      "grad_norm": 0.44940343499183655,
+      "learning_rate": 0.001,
+      "loss": 1.8856,
+      "step": 1832
+    },
+    {
+      "epoch": 0.07754463152550978,
+      "grad_norm": 0.2868255376815796,
+      "learning_rate": 0.001,
+      "loss": 2.9346,
+      "step": 1833
+    },
+    {
+      "epoch": 0.07758693628902615,
+      "grad_norm": 0.3130701780319214,
+      "learning_rate": 0.001,
+      "loss": 2.0693,
+      "step": 1834
+    },
+    {
+      "epoch": 0.07762924105254251,
+      "grad_norm": 0.3155618906021118,
+      "learning_rate": 0.001,
+      "loss": 2.8602,
+      "step": 1835
+    },
+    {
+      "epoch": 0.07767154581605888,
+      "grad_norm": 0.3293514847755432,
+      "learning_rate": 0.001,
+      "loss": 2.2591,
+      "step": 1836
+    },
+    {
+      "epoch": 0.07771385057957526,
+      "grad_norm": 0.2850898504257202,
+      "learning_rate": 0.001,
+      "loss": 2.0444,
+      "step": 1837
+    },
+    {
+      "epoch": 0.07775615534309163,
+      "grad_norm": 0.8841787576675415,
+      "learning_rate": 0.001,
+      "loss": 2.7438,
+      "step": 1838
+    },
+    {
+      "epoch": 0.077798460106608,
+      "grad_norm": 0.4035090506076813,
+      "learning_rate": 0.001,
+      "loss": 2.5178,
+      "step": 1839
+    },
+    {
+      "epoch": 0.07784076487012438,
+      "grad_norm": 0.2477097362279892,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 1840
+    },
+    {
+      "epoch": 0.07788306963364075,
+      "grad_norm": 0.32553189992904663,
+      "learning_rate": 0.001,
+      "loss": 2.4148,
+      "step": 1841
+    },
+    {
+      "epoch": 0.07792537439715712,
+      "grad_norm": 0.3204468786716461,
+      "learning_rate": 0.001,
+      "loss": 3.0719,
+      "step": 1842
+    },
+    {
+      "epoch": 0.07796767916067349,
+      "grad_norm": 0.30373045802116394,
+      "learning_rate": 0.001,
+      "loss": 2.8824,
+      "step": 1843
+    },
+    {
+      "epoch": 0.07800998392418987,
+      "grad_norm": 2.558303117752075,
+      "learning_rate": 0.001,
+      "loss": 3.2411,
+      "step": 1844
+    },
+    {
+      "epoch": 0.07805228868770624,
+      "grad_norm": 0.3434048295021057,
+      "learning_rate": 0.001,
+      "loss": 2.8942,
+      "step": 1845
+    },
+    {
+      "epoch": 0.0780945934512226,
+      "grad_norm": 0.30965694785118103,
+      "learning_rate": 0.001,
+      "loss": 3.0761,
+      "step": 1846
+    },
+    {
+      "epoch": 0.07813689821473897,
+      "grad_norm": 0.3516073226928711,
+      "learning_rate": 0.001,
+      "loss": 2.8424,
+      "step": 1847
+    },
+    {
+      "epoch": 0.07817920297825535,
+      "grad_norm": 0.3779698312282562,
+      "learning_rate": 0.001,
+      "loss": 2.8813,
+      "step": 1848
+    },
+    {
+      "epoch": 0.07822150774177172,
+      "grad_norm": 0.30136117339134216,
+      "learning_rate": 0.001,
+      "loss": 1.8542,
+      "step": 1849
+    },
+    {
+      "epoch": 0.07826381250528809,
+      "grad_norm": 0.25570622086524963,
+      "learning_rate": 0.001,
+      "loss": 2.0118,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07830611726880447,
+      "grad_norm": 0.45675748586654663,
+      "learning_rate": 0.001,
+      "loss": 2.1994,
+      "step": 1851
+    },
+    {
+      "epoch": 0.07834842203232084,
+      "grad_norm": 0.5405208468437195,
+      "learning_rate": 0.001,
+      "loss": 2.8306,
+      "step": 1852
+    },
+    {
+      "epoch": 0.07839072679583721,
+      "grad_norm": 0.28319215774536133,
+      "learning_rate": 0.001,
+      "loss": 2.2087,
+      "step": 1853
+    },
+    {
+      "epoch": 0.07843303155935358,
+      "grad_norm": 0.5920930504798889,
+      "learning_rate": 0.001,
+      "loss": 2.1219,
+      "step": 1854
+    },
+    {
+      "epoch": 0.07847533632286996,
+      "grad_norm": 0.6683090329170227,
+      "learning_rate": 0.001,
+      "loss": 2.048,
+      "step": 1855
+    },
+    {
+      "epoch": 0.07851764108638633,
+      "grad_norm": 0.28934597969055176,
+      "learning_rate": 0.001,
+      "loss": 2.4109,
+      "step": 1856
+    },
+    {
+      "epoch": 0.0785599458499027,
+      "grad_norm": 1.1490424871444702,
+      "learning_rate": 0.001,
+      "loss": 2.8251,
+      "step": 1857
+    },
+    {
+      "epoch": 0.07860225061341906,
+      "grad_norm": 0.5859717130661011,
+      "learning_rate": 0.001,
+      "loss": 2.2749,
+      "step": 1858
+    },
+    {
+      "epoch": 0.07864455537693545,
+      "grad_norm": 0.614921510219574,
+      "learning_rate": 0.001,
+      "loss": 2.7844,
+      "step": 1859
+    },
+    {
+      "epoch": 0.07868686014045181,
+      "grad_norm": 0.3503010869026184,
+      "learning_rate": 0.001,
+      "loss": 2.4979,
+      "step": 1860
+    },
+    {
+      "epoch": 0.07872916490396818,
+      "grad_norm": 0.42067447304725647,
+      "learning_rate": 0.001,
+      "loss": 2.9724,
+      "step": 1861
+    },
+    {
+      "epoch": 0.07877146966748456,
+      "grad_norm": 0.45878085494041443,
+      "learning_rate": 0.001,
+      "loss": 1.98,
+      "step": 1862
+    },
+    {
+      "epoch": 0.07881377443100093,
+      "grad_norm": 1.4618866443634033,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 1863
+    },
+    {
+      "epoch": 0.0788560791945173,
+      "grad_norm": 0.33975115418434143,
+      "learning_rate": 0.001,
+      "loss": 2.7163,
+      "step": 1864
+    },
+    {
+      "epoch": 0.07889838395803367,
+      "grad_norm": 27.44309425354004,
+      "learning_rate": 0.001,
+      "loss": 3.4932,
+      "step": 1865
+    },
+    {
+      "epoch": 0.07894068872155005,
+      "grad_norm": 0.4054664373397827,
+      "learning_rate": 0.001,
+      "loss": 2.5312,
+      "step": 1866
+    },
+    {
+      "epoch": 0.07898299348506642,
+      "grad_norm": 0.3030809462070465,
+      "learning_rate": 0.001,
+      "loss": 2.5558,
+      "step": 1867
+    },
+    {
+      "epoch": 0.07902529824858279,
+      "grad_norm": 0.40705251693725586,
+      "learning_rate": 0.001,
+      "loss": 3.12,
+      "step": 1868
+    },
+    {
+      "epoch": 0.07906760301209917,
+      "grad_norm": 16.77686882019043,
+      "learning_rate": 0.001,
+      "loss": 3.2163,
+      "step": 1869
+    },
+    {
+      "epoch": 0.07910990777561554,
+      "grad_norm": 1.4358597993850708,
+      "learning_rate": 0.001,
+      "loss": 3.0532,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0791522125391319,
+      "grad_norm": 0.3366057872772217,
+      "learning_rate": 0.001,
+      "loss": 2.595,
+      "step": 1871
+    },
+    {
+      "epoch": 0.07919451730264827,
+      "grad_norm": 0.3896106779575348,
+      "learning_rate": 0.001,
+      "loss": 2.9058,
+      "step": 1872
+    },
+    {
+      "epoch": 0.07923682206616466,
+      "grad_norm": 0.34666091203689575,
+      "learning_rate": 0.001,
+      "loss": 1.9996,
+      "step": 1873
+    },
+    {
+      "epoch": 0.07927912682968102,
+      "grad_norm": 0.34630462527275085,
+      "learning_rate": 0.001,
+      "loss": 2.6526,
+      "step": 1874
+    },
+    {
+      "epoch": 0.07932143159319739,
+      "grad_norm": 0.43369296193122864,
+      "learning_rate": 0.001,
+      "loss": 3.2299,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07936373635671376,
+      "grad_norm": 0.323746919631958,
+      "learning_rate": 0.001,
+      "loss": 2.6891,
+      "step": 1876
+    },
+    {
+      "epoch": 0.07940604112023014,
+      "grad_norm": 3.8398473262786865,
+      "learning_rate": 0.001,
+      "loss": 4.0965,
+      "step": 1877
+    },
+    {
+      "epoch": 0.07944834588374651,
+      "grad_norm": 0.5422288775444031,
+      "learning_rate": 0.001,
+      "loss": 2.9482,
+      "step": 1878
+    },
+    {
+      "epoch": 0.07949065064726288,
+      "grad_norm": 0.701977014541626,
+      "learning_rate": 0.001,
+      "loss": 2.8954,
+      "step": 1879
+    },
+    {
+      "epoch": 0.07953295541077926,
+      "grad_norm": 0.7025458216667175,
+      "learning_rate": 0.001,
+      "loss": 2.7868,
+      "step": 1880
+    },
+    {
+      "epoch": 0.07957526017429563,
+      "grad_norm": 0.8017560243606567,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 1881
+    },
+    {
+      "epoch": 0.079617564937812,
+      "grad_norm": 0.45160236954689026,
+      "learning_rate": 0.001,
+      "loss": 2.2321,
+      "step": 1882
+    },
+    {
+      "epoch": 0.07965986970132836,
+      "grad_norm": 0.4654178023338318,
+      "learning_rate": 0.001,
+      "loss": 3.4246,
+      "step": 1883
+    },
+    {
+      "epoch": 0.07970217446484475,
+      "grad_norm": 0.3444591164588928,
+      "learning_rate": 0.001,
+      "loss": 2.8039,
+      "step": 1884
+    },
+    {
+      "epoch": 0.07974447922836111,
+      "grad_norm": 1.3339020013809204,
+      "learning_rate": 0.001,
+      "loss": 3.6179,
+      "step": 1885
+    },
+    {
+      "epoch": 0.07978678399187748,
+      "grad_norm": 0.23645636439323425,
+      "learning_rate": 0.001,
+      "loss": 1.7768,
+      "step": 1886
+    },
+    {
+      "epoch": 0.07982908875539385,
+      "grad_norm": 0.3817785680294037,
+      "learning_rate": 0.001,
+      "loss": 2.7453,
+      "step": 1887
+    },
+    {
+      "epoch": 0.07987139351891023,
+      "grad_norm": 0.41465362906455994,
+      "learning_rate": 0.001,
+      "loss": 3.0675,
+      "step": 1888
+    },
+    {
+      "epoch": 0.0799136982824266,
+      "grad_norm": 0.37037500739097595,
+      "learning_rate": 0.001,
+      "loss": 2.3173,
+      "step": 1889
+    },
+    {
+      "epoch": 0.07995600304594297,
+      "grad_norm": 0.31565260887145996,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07999830780945935,
+      "grad_norm": 6.897204399108887,
+      "learning_rate": 0.001,
+      "loss": 2.5709,
+      "step": 1891
+    },
+    {
+      "epoch": 0.08004061257297572,
+      "grad_norm": 1.544342279434204,
+      "learning_rate": 0.001,
+      "loss": 2.1807,
+      "step": 1892
+    },
+    {
+      "epoch": 0.08008291733649209,
+      "grad_norm": 1.1234506368637085,
+      "learning_rate": 0.001,
+      "loss": 2.368,
+      "step": 1893
+    },
+    {
+      "epoch": 0.08012522210000846,
+      "grad_norm": 0.2884044051170349,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 1894
+    },
+    {
+      "epoch": 0.08016752686352484,
+      "grad_norm": 0.38709041476249695,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0802098316270412,
+      "grad_norm": 41.802032470703125,
+      "learning_rate": 0.001,
+      "loss": 1.8604,
+      "step": 1896
+    },
+    {
+      "epoch": 0.08025213639055757,
+      "grad_norm": 0.44042840600013733,
+      "learning_rate": 0.001,
+      "loss": 1.6504,
+      "step": 1897
+    },
+    {
+      "epoch": 0.08029444115407394,
+      "grad_norm": 1.7440584897994995,
+      "learning_rate": 0.001,
+      "loss": 2.5165,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08033674591759032,
+      "grad_norm": 1.517181396484375,
+      "learning_rate": 0.001,
+      "loss": 2.1762,
+      "step": 1899
+    },
+    {
+      "epoch": 0.08037905068110669,
+      "grad_norm": 0.5725038051605225,
+      "learning_rate": 0.001,
+      "loss": 2.5455,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08042135544462306,
+      "grad_norm": 0.45639267563819885,
+      "learning_rate": 0.001,
+      "loss": 2.379,
+      "step": 1901
+    },
+    {
+      "epoch": 0.08046366020813944,
+      "grad_norm": 1.4812275171279907,
+      "learning_rate": 0.001,
+      "loss": 2.4805,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08050596497165581,
+      "grad_norm": 0.3487553000450134,
+      "learning_rate": 0.001,
+      "loss": 1.9803,
+      "step": 1903
+    },
+    {
+      "epoch": 0.08054826973517218,
+      "grad_norm": 0.36132708191871643,
+      "learning_rate": 0.001,
+      "loss": 2.0808,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08059057449868855,
+      "grad_norm": 0.3573780357837677,
+      "learning_rate": 0.001,
+      "loss": 1.9821,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08063287926220493,
+      "grad_norm": 2.551344156265259,
+      "learning_rate": 0.001,
+      "loss": 2.4851,
+      "step": 1906
+    },
+    {
+      "epoch": 0.0806751840257213,
+      "grad_norm": 1.2544894218444824,
+      "learning_rate": 0.001,
+      "loss": 2.0348,
+      "step": 1907
+    },
+    {
+      "epoch": 0.08071748878923767,
+      "grad_norm": 0.3534647524356842,
+      "learning_rate": 0.001,
+      "loss": 4.1648,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08075979355275403,
+      "grad_norm": 0.5673227310180664,
+      "learning_rate": 0.001,
+      "loss": 2.953,
+      "step": 1909
+    },
+    {
+      "epoch": 0.08080209831627042,
+      "grad_norm": 5.431023120880127,
+      "learning_rate": 0.001,
+      "loss": 2.5128,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08084440307978678,
+      "grad_norm": 15.899325370788574,
+      "learning_rate": 0.001,
+      "loss": 2.9833,
+      "step": 1911
+    },
+    {
+      "epoch": 0.08088670784330315,
+      "grad_norm": 0.5933341383934021,
+      "learning_rate": 0.001,
+      "loss": 2.0336,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08092901260681953,
+      "grad_norm": 0.572313129901886,
+      "learning_rate": 0.001,
+      "loss": 2.8636,
+      "step": 1913
+    },
+    {
+      "epoch": 0.0809713173703359,
+      "grad_norm": 2.6062185764312744,
+      "learning_rate": 0.001,
+      "loss": 2.3718,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08101362213385227,
+      "grad_norm": 0.769607424736023,
+      "learning_rate": 0.001,
+      "loss": 2.6415,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08105592689736864,
+      "grad_norm": 0.4920905828475952,
+      "learning_rate": 0.001,
+      "loss": 2.1552,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08109823166088502,
+      "grad_norm": 0.5874459147453308,
+      "learning_rate": 0.001,
+      "loss": 2.376,
+      "step": 1917
+    },
+    {
+      "epoch": 0.08114053642440139,
+      "grad_norm": 0.48925819993019104,
+      "learning_rate": 0.001,
+      "loss": 1.7734,
+      "step": 1918
+    },
+    {
+      "epoch": 0.08118284118791776,
+      "grad_norm": 0.7119834423065186,
+      "learning_rate": 0.001,
+      "loss": 2.6034,
+      "step": 1919
+    },
+    {
+      "epoch": 0.08122514595143414,
+      "grad_norm": 3.461428165435791,
+      "learning_rate": 0.001,
+      "loss": 2.6797,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0812674507149505,
+      "grad_norm": 0.7845181822776794,
+      "learning_rate": 0.001,
+      "loss": 4.2376,
+      "step": 1921
+    },
+    {
+      "epoch": 0.08130975547846687,
+      "grad_norm": 0.5165002942085266,
+      "learning_rate": 0.001,
+      "loss": 3.3205,
+      "step": 1922
+    },
+    {
+      "epoch": 0.08135206024198324,
+      "grad_norm": 1.0320379734039307,
+      "learning_rate": 0.001,
+      "loss": 2.7671,
+      "step": 1923
+    },
+    {
+      "epoch": 0.08139436500549962,
+      "grad_norm": 7.306519031524658,
+      "learning_rate": 0.001,
+      "loss": 3.0783,
+      "step": 1924
+    },
+    {
+      "epoch": 0.08143666976901599,
+      "grad_norm": 0.51955646276474,
+      "learning_rate": 0.001,
+      "loss": 2.2582,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08147897453253236,
+      "grad_norm": 0.405947744846344,
+      "learning_rate": 0.001,
+      "loss": 3.2719,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08152127929604873,
+      "grad_norm": 2.2772488594055176,
+      "learning_rate": 0.001,
+      "loss": 2.8051,
+      "step": 1927
+    },
+    {
+      "epoch": 0.08156358405956511,
+      "grad_norm": 0.4286925494670868,
+      "learning_rate": 0.001,
+      "loss": 2.4886,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08160588882308148,
+      "grad_norm": 0.3987908661365509,
+      "learning_rate": 0.001,
+      "loss": 2.1242,
+      "step": 1929
+    },
+    {
+      "epoch": 0.08164819358659785,
+      "grad_norm": 0.42939138412475586,
+      "learning_rate": 0.001,
+      "loss": 3.0869,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08169049835011423,
+      "grad_norm": 0.7904039025306702,
+      "learning_rate": 0.001,
+      "loss": 2.6297,
+      "step": 1931
+    },
+    {
+      "epoch": 0.0817328031136306,
+      "grad_norm": 0.2795187532901764,
+      "learning_rate": 0.001,
+      "loss": 1.8094,
+      "step": 1932
+    },
+    {
+      "epoch": 0.08177510787714697,
+      "grad_norm": 0.7740998268127441,
+      "learning_rate": 0.001,
+      "loss": 2.6684,
+      "step": 1933
+    },
+    {
+      "epoch": 0.08181741264066333,
+      "grad_norm": 0.5167595148086548,
+      "learning_rate": 0.001,
+      "loss": 2.8719,
+      "step": 1934
+    },
+    {
+      "epoch": 0.08185971740417972,
+      "grad_norm": 0.354657381772995,
+      "learning_rate": 0.001,
+      "loss": 2.6624,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08190202216769608,
+      "grad_norm": 0.35921430587768555,
+      "learning_rate": 0.001,
+      "loss": 2.4581,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08194432693121245,
+      "grad_norm": 0.6091867089271545,
+      "learning_rate": 0.001,
+      "loss": 2.978,
+      "step": 1937
+    },
+    {
+      "epoch": 0.08198663169472882,
+      "grad_norm": 0.3409866988658905,
+      "learning_rate": 0.001,
+      "loss": 2.5926,
+      "step": 1938
+    },
+    {
+      "epoch": 0.0820289364582452,
+      "grad_norm": 0.5260420441627502,
+      "learning_rate": 0.001,
+      "loss": 3.2647,
+      "step": 1939
+    },
+    {
+      "epoch": 0.08207124122176157,
+      "grad_norm": 0.2958446741104126,
+      "learning_rate": 0.001,
+      "loss": 2.509,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08211354598527794,
+      "grad_norm": 0.2943398058414459,
+      "learning_rate": 0.001,
+      "loss": 2.0493,
+      "step": 1941
+    },
+    {
+      "epoch": 0.08215585074879432,
+      "grad_norm": 0.3319268226623535,
+      "learning_rate": 0.001,
+      "loss": 2.3304,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08219815551231069,
+      "grad_norm": 0.5134836435317993,
+      "learning_rate": 0.001,
+      "loss": 2.693,
+      "step": 1943
+    },
+    {
+      "epoch": 0.08224046027582706,
+      "grad_norm": 0.3512554168701172,
+      "learning_rate": 0.001,
+      "loss": 3.3824,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08228276503934343,
+      "grad_norm": 1.8263537883758545,
+      "learning_rate": 0.001,
+      "loss": 2.3778,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08232506980285981,
+      "grad_norm": 0.3825913667678833,
+      "learning_rate": 0.001,
+      "loss": 2.2277,
+      "step": 1946
+    },
+    {
+      "epoch": 0.08236737456637618,
+      "grad_norm": 0.3037860691547394,
+      "learning_rate": 0.001,
+      "loss": 2.3199,
+      "step": 1947
+    },
+    {
+      "epoch": 0.08240967932989254,
+      "grad_norm": 0.4019972085952759,
+      "learning_rate": 0.001,
+      "loss": 2.3972,
+      "step": 1948
+    },
+    {
+      "epoch": 0.08245198409340891,
+      "grad_norm": 1.4788627624511719,
+      "learning_rate": 0.001,
+      "loss": 2.4204,
+      "step": 1949
+    },
+    {
+      "epoch": 0.0824942888569253,
+      "grad_norm": 0.5321111083030701,
+      "learning_rate": 0.001,
+      "loss": 2.8231,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08253659362044166,
+      "grad_norm": 0.5626177191734314,
+      "learning_rate": 0.001,
+      "loss": 3.8786,
+      "step": 1951
+    },
+    {
+      "epoch": 0.08257889838395803,
+      "grad_norm": 0.5114405751228333,
+      "learning_rate": 0.001,
+      "loss": 2.941,
+      "step": 1952
+    },
+    {
+      "epoch": 0.08262120314747441,
+      "grad_norm": 0.48595815896987915,
+      "learning_rate": 0.001,
+      "loss": 2.6246,
+      "step": 1953
+    },
+    {
+      "epoch": 0.08266350791099078,
+      "grad_norm": 0.28523850440979004,
+      "learning_rate": 0.001,
+      "loss": 2.7918,
+      "step": 1954
+    },
+    {
+      "epoch": 0.08270581267450715,
+      "grad_norm": 0.7027163505554199,
+      "learning_rate": 0.001,
+      "loss": 2.7862,
+      "step": 1955
+    },
+    {
+      "epoch": 0.08274811743802352,
+      "grad_norm": 0.2511918544769287,
+      "learning_rate": 0.001,
+      "loss": 1.811,
+      "step": 1956
+    },
+    {
+      "epoch": 0.0827904222015399,
+      "grad_norm": 0.2575523555278778,
+      "learning_rate": 0.001,
+      "loss": 3.1751,
+      "step": 1957
+    },
+    {
+      "epoch": 0.08283272696505627,
+      "grad_norm": 0.3014431297779083,
+      "learning_rate": 0.001,
+      "loss": 2.4499,
+      "step": 1958
+    },
+    {
+      "epoch": 0.08287503172857263,
+      "grad_norm": 0.24527455866336823,
+      "learning_rate": 0.001,
+      "loss": 2.0457,
+      "step": 1959
+    },
+    {
+      "epoch": 0.082917336492089,
+      "grad_norm": 0.312063992023468,
+      "learning_rate": 0.001,
+      "loss": 2.063,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08295964125560538,
+      "grad_norm": 1.2686574459075928,
+      "learning_rate": 0.001,
+      "loss": 3.0543,
+      "step": 1961
+    },
+    {
+      "epoch": 0.08300194601912175,
+      "grad_norm": 0.3444327712059021,
+      "learning_rate": 0.001,
+      "loss": 2.5528,
+      "step": 1962
+    },
+    {
+      "epoch": 0.08304425078263812,
+      "grad_norm": 0.5791609287261963,
+      "learning_rate": 0.001,
+      "loss": 3.4888,
+      "step": 1963
+    },
+    {
+      "epoch": 0.0830865555461545,
+      "grad_norm": 1.6436971426010132,
+      "learning_rate": 0.001,
+      "loss": 2.6672,
+      "step": 1964
+    },
+    {
+      "epoch": 0.08312886030967087,
+      "grad_norm": 0.25417083501815796,
+      "learning_rate": 0.001,
+      "loss": 2.1457,
+      "step": 1965
+    },
+    {
+      "epoch": 0.08317116507318724,
+      "grad_norm": 0.3741660714149475,
+      "learning_rate": 0.001,
+      "loss": 1.8138,
+      "step": 1966
+    },
+    {
+      "epoch": 0.08321346983670361,
+      "grad_norm": 0.3511454164981842,
+      "learning_rate": 0.001,
+      "loss": 1.9674,
+      "step": 1967
+    },
+    {
+      "epoch": 0.08325577460021999,
+      "grad_norm": 0.3857306241989136,
+      "learning_rate": 0.001,
+      "loss": 3.0311,
+      "step": 1968
+    },
+    {
+      "epoch": 0.08329807936373636,
+      "grad_norm": 0.6367977261543274,
+      "learning_rate": 0.001,
+      "loss": 2.3158,
+      "step": 1969
+    },
+    {
+      "epoch": 0.08334038412725273,
+      "grad_norm": 0.30183205008506775,
+      "learning_rate": 0.001,
+      "loss": 2.4227,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0833826888907691,
+      "grad_norm": 0.3112057149410248,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 1971
+    },
+    {
+      "epoch": 0.08342499365428548,
+      "grad_norm": 0.3398885726928711,
+      "learning_rate": 0.001,
+      "loss": 2.5514,
+      "step": 1972
+    },
+    {
+      "epoch": 0.08346729841780184,
+      "grad_norm": 0.5639985799789429,
+      "learning_rate": 0.001,
+      "loss": 3.3706,
+      "step": 1973
+    },
+    {
+      "epoch": 0.08350960318131821,
+      "grad_norm": 0.2720838785171509,
+      "learning_rate": 0.001,
+      "loss": 2.0734,
+      "step": 1974
+    },
+    {
+      "epoch": 0.0835519079448346,
+      "grad_norm": 0.3382203280925751,
+      "learning_rate": 0.001,
+      "loss": 2.1143,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08359421270835096,
+      "grad_norm": 0.9592907428741455,
+      "learning_rate": 0.001,
+      "loss": 1.722,
+      "step": 1976
+    },
+    {
+      "epoch": 0.08363651747186733,
+      "grad_norm": 0.4691549241542816,
+      "learning_rate": 0.001,
+      "loss": 3.0425,
+      "step": 1977
+    },
+    {
+      "epoch": 0.0836788222353837,
+      "grad_norm": 0.4217078983783722,
+      "learning_rate": 0.001,
+      "loss": 2.8026,
+      "step": 1978
+    },
+    {
+      "epoch": 0.08372112699890008,
+      "grad_norm": 0.9768161773681641,
+      "learning_rate": 0.001,
+      "loss": 3.152,
+      "step": 1979
+    },
+    {
+      "epoch": 0.08376343176241645,
+      "grad_norm": 0.4018516540527344,
+      "learning_rate": 0.001,
+      "loss": 1.9038,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08380573652593282,
+      "grad_norm": 0.3423294126987457,
+      "learning_rate": 0.001,
+      "loss": 3.554,
+      "step": 1981
+    },
+    {
+      "epoch": 0.0838480412894492,
+      "grad_norm": 4.468573093414307,
+      "learning_rate": 0.001,
+      "loss": 1.7724,
+      "step": 1982
+    },
+    {
+      "epoch": 0.08389034605296557,
+      "grad_norm": 2.0579144954681396,
+      "learning_rate": 0.001,
+      "loss": 2.9118,
+      "step": 1983
+    },
+    {
+      "epoch": 0.08393265081648194,
+      "grad_norm": 0.3974030911922455,
+      "learning_rate": 0.001,
+      "loss": 1.8146,
+      "step": 1984
+    },
+    {
+      "epoch": 0.0839749555799983,
+      "grad_norm": 2.071157693862915,
+      "learning_rate": 0.001,
+      "loss": 3.3286,
+      "step": 1985
+    },
+    {
+      "epoch": 0.08401726034351469,
+      "grad_norm": 0.8951895236968994,
+      "learning_rate": 0.001,
+      "loss": 2.1807,
+      "step": 1986
+    },
+    {
+      "epoch": 0.08405956510703105,
+      "grad_norm": 0.3309991955757141,
+      "learning_rate": 0.001,
+      "loss": 2.9089,
+      "step": 1987
+    },
+    {
+      "epoch": 0.08410186987054742,
+      "grad_norm": 0.3316001892089844,
+      "learning_rate": 0.001,
+      "loss": 2.4278,
+      "step": 1988
+    },
+    {
+      "epoch": 0.08414417463406379,
+      "grad_norm": 0.4614253342151642,
+      "learning_rate": 0.001,
+      "loss": 1.8007,
+      "step": 1989
+    },
+    {
+      "epoch": 0.08418647939758017,
+      "grad_norm": 0.4860163629055023,
+      "learning_rate": 0.001,
+      "loss": 2.394,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08422878416109654,
+      "grad_norm": 0.30347874760627747,
+      "learning_rate": 0.001,
+      "loss": 2.4783,
+      "step": 1991
+    },
+    {
+      "epoch": 0.08427108892461291,
+      "grad_norm": 0.3198263943195343,
+      "learning_rate": 0.001,
+      "loss": 2.4848,
+      "step": 1992
+    },
+    {
+      "epoch": 0.08431339368812929,
+      "grad_norm": 6.526590824127197,
+      "learning_rate": 0.001,
+      "loss": 2.6791,
+      "step": 1993
+    },
+    {
+      "epoch": 0.08435569845164566,
+      "grad_norm": 0.4787849485874176,
+      "learning_rate": 0.001,
+      "loss": 3.7781,
+      "step": 1994
+    },
+    {
+      "epoch": 0.08439800321516203,
+      "grad_norm": 0.34780532121658325,
+      "learning_rate": 0.001,
+      "loss": 2.3788,
+      "step": 1995
+    },
+    {
+      "epoch": 0.0844403079786784,
+      "grad_norm": 0.30479609966278076,
+      "learning_rate": 0.001,
+      "loss": 2.1992,
+      "step": 1996
+    },
+    {
+      "epoch": 0.08448261274219478,
+      "grad_norm": 1.0337961912155151,
+      "learning_rate": 0.001,
+      "loss": 2.4048,
+      "step": 1997
+    },
+    {
+      "epoch": 0.08452491750571114,
+      "grad_norm": 0.5477790832519531,
+      "learning_rate": 0.001,
+      "loss": 2.2986,
+      "step": 1998
+    },
+    {
+      "epoch": 0.08456722226922751,
+      "grad_norm": 0.7585234045982361,
+      "learning_rate": 0.001,
+      "loss": 2.7247,
+      "step": 1999
+    },
+    {
+      "epoch": 0.08460952703274388,
+      "grad_norm": 12.078346252441406,
+      "learning_rate": 0.001,
+      "loss": 1.9748,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08465183179626026,
+      "grad_norm": 0.26165372133255005,
+      "learning_rate": 0.001,
+      "loss": 2.2865,
+      "step": 2001
+    },
+    {
+      "epoch": 0.08469413655977663,
+      "grad_norm": 0.21518190205097198,
+      "learning_rate": 0.001,
+      "loss": 1.9516,
+      "step": 2002
+    },
+    {
+      "epoch": 0.084736441323293,
+      "grad_norm": 1.5820046663284302,
+      "learning_rate": 0.001,
+      "loss": 2.2892,
+      "step": 2003
+    },
+    {
+      "epoch": 0.08477874608680938,
+      "grad_norm": 0.3572734296321869,
+      "learning_rate": 0.001,
+      "loss": 3.4287,
+      "step": 2004
+    },
+    {
+      "epoch": 0.08482105085032575,
+      "grad_norm": 3.6497509479522705,
+      "learning_rate": 0.001,
+      "loss": 3.2188,
+      "step": 2005
+    },
+    {
+      "epoch": 0.08486335561384212,
+      "grad_norm": 0.38023465871810913,
+      "learning_rate": 0.001,
+      "loss": 2.0518,
+      "step": 2006
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.24291670322418213,
+      "learning_rate": 0.001,
+      "loss": 2.0684,
+      "step": 2007
+    },
+    {
+      "epoch": 0.08494796514087487,
+      "grad_norm": 0.3124212622642517,
+      "learning_rate": 0.001,
+      "loss": 2.3747,
+      "step": 2008
+    },
+    {
+      "epoch": 0.08499026990439124,
+      "grad_norm": 1.313674807548523,
+      "learning_rate": 0.001,
+      "loss": 1.7892,
+      "step": 2009
+    },
+    {
+      "epoch": 0.0850325746679076,
+      "grad_norm": 0.5036904215812683,
+      "learning_rate": 0.001,
+      "loss": 2.5027,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08507487943142397,
+      "grad_norm": 0.4840260148048401,
+      "learning_rate": 0.001,
+      "loss": 2.3056,
+      "step": 2011
+    },
+    {
+      "epoch": 0.08511718419494035,
+      "grad_norm": 0.48675820231437683,
+      "learning_rate": 0.001,
+      "loss": 2.6466,
+      "step": 2012
+    },
+    {
+      "epoch": 0.08515948895845672,
+      "grad_norm": 6.131993770599365,
+      "learning_rate": 0.001,
+      "loss": 2.6512,
+      "step": 2013
+    },
+    {
+      "epoch": 0.08520179372197309,
+      "grad_norm": 0.37686774134635925,
+      "learning_rate": 0.001,
+      "loss": 3.5639,
+      "step": 2014
+    },
+    {
+      "epoch": 0.08524409848548947,
+      "grad_norm": 0.6002863645553589,
+      "learning_rate": 0.001,
+      "loss": 3.1445,
+      "step": 2015
+    },
+    {
+      "epoch": 0.08528640324900584,
+      "grad_norm": 0.46876654028892517,
+      "learning_rate": 0.001,
+      "loss": 2.5458,
+      "step": 2016
+    },
+    {
+      "epoch": 0.08532870801252221,
+      "grad_norm": 0.38201484084129333,
+      "learning_rate": 0.001,
+      "loss": 2.5112,
+      "step": 2017
+    },
+    {
+      "epoch": 0.08537101277603858,
+      "grad_norm": 0.31709811091423035,
+      "learning_rate": 0.001,
+      "loss": 1.8442,
+      "step": 2018
+    },
+    {
+      "epoch": 0.08541331753955496,
+      "grad_norm": 3.5930492877960205,
+      "learning_rate": 0.001,
+      "loss": 2.654,
+      "step": 2019
+    },
+    {
+      "epoch": 0.08545562230307133,
+      "grad_norm": 0.48756277561187744,
+      "learning_rate": 0.001,
+      "loss": 2.5555,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0854979270665877,
+      "grad_norm": 0.3280019164085388,
+      "learning_rate": 0.001,
+      "loss": 3.0868,
+      "step": 2021
+    },
+    {
+      "epoch": 0.08554023183010406,
+      "grad_norm": 3.57021164894104,
+      "learning_rate": 0.001,
+      "loss": 2.7273,
+      "step": 2022
+    },
+    {
+      "epoch": 0.08558253659362045,
+      "grad_norm": 2.2349965572357178,
+      "learning_rate": 0.001,
+      "loss": 3.3427,
+      "step": 2023
+    },
+    {
+      "epoch": 0.08562484135713681,
+      "grad_norm": 0.333631694316864,
+      "learning_rate": 0.001,
+      "loss": 2.8953,
+      "step": 2024
+    },
+    {
+      "epoch": 0.08566714612065318,
+      "grad_norm": 0.31778064370155334,
+      "learning_rate": 0.001,
+      "loss": 2.4624,
+      "step": 2025
+    },
+    {
+      "epoch": 0.08570945088416956,
+      "grad_norm": 0.3415403664112091,
+      "learning_rate": 0.001,
+      "loss": 2.1366,
+      "step": 2026
+    },
+    {
+      "epoch": 0.08575175564768593,
+      "grad_norm": 3.173630952835083,
+      "learning_rate": 0.001,
+      "loss": 2.8478,
+      "step": 2027
+    },
+    {
+      "epoch": 0.0857940604112023,
+      "grad_norm": 0.27932387590408325,
+      "learning_rate": 0.001,
+      "loss": 2.1341,
+      "step": 2028
+    },
+    {
+      "epoch": 0.08583636517471867,
+      "grad_norm": 0.4122964143753052,
+      "learning_rate": 0.001,
+      "loss": 2.9483,
+      "step": 2029
+    },
+    {
+      "epoch": 0.08587866993823505,
+      "grad_norm": 2.471703052520752,
+      "learning_rate": 0.001,
+      "loss": 3.5857,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08592097470175142,
+      "grad_norm": 0.3559750020503998,
+      "learning_rate": 0.001,
+      "loss": 3.2138,
+      "step": 2031
+    },
+    {
+      "epoch": 0.08596327946526779,
+      "grad_norm": 0.4111155867576599,
+      "learning_rate": 0.001,
+      "loss": 2.3594,
+      "step": 2032
+    },
+    {
+      "epoch": 0.08600558422878415,
+      "grad_norm": 0.2714095115661621,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 2033
+    },
+    {
+      "epoch": 0.08604788899230054,
+      "grad_norm": 0.313900887966156,
+      "learning_rate": 0.001,
+      "loss": 2.274,
+      "step": 2034
+    },
+    {
+      "epoch": 0.0860901937558169,
+      "grad_norm": 0.2859230637550354,
+      "learning_rate": 0.001,
+      "loss": 2.1988,
+      "step": 2035
+    },
+    {
+      "epoch": 0.08613249851933327,
+      "grad_norm": 0.323250949382782,
+      "learning_rate": 0.001,
+      "loss": 2.0196,
+      "step": 2036
+    },
+    {
+      "epoch": 0.08617480328284965,
+      "grad_norm": 0.2705003321170807,
+      "learning_rate": 0.001,
+      "loss": 1.9217,
+      "step": 2037
+    },
+    {
+      "epoch": 0.08621710804636602,
+      "grad_norm": 0.25685760378837585,
+      "learning_rate": 0.001,
+      "loss": 1.9932,
+      "step": 2038
+    },
+    {
+      "epoch": 0.08625941280988239,
+      "grad_norm": 0.2940180003643036,
+      "learning_rate": 0.001,
+      "loss": 3.1947,
+      "step": 2039
+    },
+    {
+      "epoch": 0.08630171757339876,
+      "grad_norm": 0.23758481442928314,
+      "learning_rate": 0.001,
+      "loss": 2.1624,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08634402233691514,
+      "grad_norm": 0.2667909264564514,
+      "learning_rate": 0.001,
+      "loss": 2.5975,
+      "step": 2041
+    },
+    {
+      "epoch": 0.08638632710043151,
+      "grad_norm": 0.23966266214847565,
+      "learning_rate": 0.001,
+      "loss": 2.7902,
+      "step": 2042
+    },
+    {
+      "epoch": 0.08642863186394788,
+      "grad_norm": 0.3106350898742676,
+      "learning_rate": 0.001,
+      "loss": 1.6657,
+      "step": 2043
+    },
+    {
+      "epoch": 0.08647093662746426,
+      "grad_norm": 0.25842878222465515,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 2044
+    },
+    {
+      "epoch": 0.08651324139098063,
+      "grad_norm": 0.6965135931968689,
+      "learning_rate": 0.001,
+      "loss": 2.3328,
+      "step": 2045
+    },
+    {
+      "epoch": 0.086555546154497,
+      "grad_norm": 0.30815958976745605,
+      "learning_rate": 0.001,
+      "loss": 1.7124,
+      "step": 2046
+    },
+    {
+      "epoch": 0.08659785091801336,
+      "grad_norm": 0.32658496499061584,
+      "learning_rate": 0.001,
+      "loss": 3.9796,
+      "step": 2047
+    },
+    {
+      "epoch": 0.08664015568152975,
+      "grad_norm": 0.25080788135528564,
+      "learning_rate": 0.001,
+      "loss": 2.4049,
+      "step": 2048
+    },
+    {
+      "epoch": 0.08668246044504611,
+      "grad_norm": 0.31165841221809387,
+      "learning_rate": 0.001,
+      "loss": 2.0898,
+      "step": 2049
+    },
+    {
+      "epoch": 0.08672476520856248,
+      "grad_norm": 1.0614737272262573,
+      "learning_rate": 0.001,
+      "loss": 1.7963,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08676706997207885,
+      "grad_norm": 0.44003593921661377,
+      "learning_rate": 0.001,
+      "loss": 3.2018,
+      "step": 2051
+    },
+    {
+      "epoch": 0.08680937473559523,
+      "grad_norm": 0.44519808888435364,
+      "learning_rate": 0.001,
+      "loss": 2.6672,
+      "step": 2052
+    },
+    {
+      "epoch": 0.0868516794991116,
+      "grad_norm": 1.5413126945495605,
+      "learning_rate": 0.001,
+      "loss": 1.6514,
+      "step": 2053
+    },
+    {
+      "epoch": 0.08689398426262797,
+      "grad_norm": 1.6876548528671265,
+      "learning_rate": 0.001,
+      "loss": 2.9417,
+      "step": 2054
+    },
+    {
+      "epoch": 0.08693628902614435,
+      "grad_norm": 0.26920443773269653,
+      "learning_rate": 0.001,
+      "loss": 2.1858,
+      "step": 2055
+    },
+    {
+      "epoch": 0.08697859378966072,
+      "grad_norm": 0.4876200556755066,
+      "learning_rate": 0.001,
+      "loss": 2.9616,
+      "step": 2056
+    },
+    {
+      "epoch": 0.08702089855317709,
+      "grad_norm": 0.32026639580726624,
+      "learning_rate": 0.001,
+      "loss": 2.398,
+      "step": 2057
+    },
+    {
+      "epoch": 0.08706320331669345,
+      "grad_norm": 0.34720298647880554,
+      "learning_rate": 0.001,
+      "loss": 2.5501,
+      "step": 2058
+    },
+    {
+      "epoch": 0.08710550808020984,
+      "grad_norm": 0.3236474394798279,
+      "learning_rate": 0.001,
+      "loss": 1.9898,
+      "step": 2059
+    },
+    {
+      "epoch": 0.0871478128437262,
+      "grad_norm": 0.4026321470737457,
+      "learning_rate": 0.001,
+      "loss": 2.4051,
+      "step": 2060
+    },
+    {
+      "epoch": 0.08719011760724257,
+      "grad_norm": 0.9391548037528992,
+      "learning_rate": 0.001,
+      "loss": 2.2811,
+      "step": 2061
+    },
+    {
+      "epoch": 0.08723242237075894,
+      "grad_norm": 0.29237544536590576,
+      "learning_rate": 0.001,
+      "loss": 2.1954,
+      "step": 2062
+    },
+    {
+      "epoch": 0.08727472713427532,
+      "grad_norm": 0.26706722378730774,
+      "learning_rate": 0.001,
+      "loss": 2.0118,
+      "step": 2063
+    },
+    {
+      "epoch": 0.08731703189779169,
+      "grad_norm": 0.6832917928695679,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 2064
+    },
+    {
+      "epoch": 0.08735933666130806,
+      "grad_norm": 3.363959312438965,
+      "learning_rate": 0.001,
+      "loss": 2.8937,
+      "step": 2065
+    },
+    {
+      "epoch": 0.08740164142482444,
+      "grad_norm": 0.6167427897453308,
+      "learning_rate": 0.001,
+      "loss": 2.4164,
+      "step": 2066
+    },
+    {
+      "epoch": 0.08744394618834081,
+      "grad_norm": 0.40449997782707214,
+      "learning_rate": 0.001,
+      "loss": 2.503,
+      "step": 2067
+    },
+    {
+      "epoch": 0.08748625095185718,
+      "grad_norm": 0.77544105052948,
+      "learning_rate": 0.001,
+      "loss": 3.1617,
+      "step": 2068
+    },
+    {
+      "epoch": 0.08752855571537355,
+      "grad_norm": 0.6678179502487183,
+      "learning_rate": 0.001,
+      "loss": 2.4034,
+      "step": 2069
+    },
+    {
+      "epoch": 0.08757086047888993,
+      "grad_norm": 0.46908748149871826,
+      "learning_rate": 0.001,
+      "loss": 2.4388,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0876131652424063,
+      "grad_norm": 0.3415198028087616,
+      "learning_rate": 0.001,
+      "loss": 2.1139,
+      "step": 2071
+    },
+    {
+      "epoch": 0.08765547000592266,
+      "grad_norm": 0.3351193070411682,
+      "learning_rate": 0.001,
+      "loss": 2.3951,
+      "step": 2072
+    },
+    {
+      "epoch": 0.08769777476943903,
+      "grad_norm": 0.27987435460090637,
+      "learning_rate": 0.001,
+      "loss": 2.0981,
+      "step": 2073
+    },
+    {
+      "epoch": 0.08774007953295541,
+      "grad_norm": 1.4388349056243896,
+      "learning_rate": 0.001,
+      "loss": 1.9596,
+      "step": 2074
+    },
+    {
+      "epoch": 0.08778238429647178,
+      "grad_norm": 2.4498531818389893,
+      "learning_rate": 0.001,
+      "loss": 2.667,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08782468905998815,
+      "grad_norm": 0.259954035282135,
+      "learning_rate": 0.001,
+      "loss": 2.7094,
+      "step": 2076
+    },
+    {
+      "epoch": 0.08786699382350453,
+      "grad_norm": 0.24030877649784088,
+      "learning_rate": 0.001,
+      "loss": 2.7154,
+      "step": 2077
+    },
+    {
+      "epoch": 0.0879092985870209,
+      "grad_norm": 4.155751705169678,
+      "learning_rate": 0.001,
+      "loss": 1.8017,
+      "step": 2078
+    },
+    {
+      "epoch": 0.08795160335053727,
+      "grad_norm": 0.30022862553596497,
+      "learning_rate": 0.001,
+      "loss": 2.6318,
+      "step": 2079
+    },
+    {
+      "epoch": 0.08799390811405364,
+      "grad_norm": 4.097969055175781,
+      "learning_rate": 0.001,
+      "loss": 2.7386,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08803621287757002,
+      "grad_norm": 0.8870471715927124,
+      "learning_rate": 0.001,
+      "loss": 2.6592,
+      "step": 2081
+    },
+    {
+      "epoch": 0.08807851764108639,
+      "grad_norm": 0.27012670040130615,
+      "learning_rate": 0.001,
+      "loss": 2.5062,
+      "step": 2082
+    },
+    {
+      "epoch": 0.08812082240460276,
+      "grad_norm": 0.28547555208206177,
+      "learning_rate": 0.001,
+      "loss": 2.3858,
+      "step": 2083
+    },
+    {
+      "epoch": 0.08816312716811912,
+      "grad_norm": 0.34155458211898804,
+      "learning_rate": 0.001,
+      "loss": 3.2342,
+      "step": 2084
+    },
+    {
+      "epoch": 0.0882054319316355,
+      "grad_norm": 0.26611489057540894,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 2085
+    },
+    {
+      "epoch": 0.08824773669515187,
+      "grad_norm": 0.35170432925224304,
+      "learning_rate": 0.001,
+      "loss": 2.7927,
+      "step": 2086
+    },
+    {
+      "epoch": 0.08829004145866824,
+      "grad_norm": 0.6946236491203308,
+      "learning_rate": 0.001,
+      "loss": 1.982,
+      "step": 2087
+    },
+    {
+      "epoch": 0.08833234622218462,
+      "grad_norm": 0.8609383702278137,
+      "learning_rate": 0.001,
+      "loss": 1.992,
+      "step": 2088
+    },
+    {
+      "epoch": 0.08837465098570099,
+      "grad_norm": 0.2939896583557129,
+      "learning_rate": 0.001,
+      "loss": 2.2666,
+      "step": 2089
+    },
+    {
+      "epoch": 0.08841695574921736,
+      "grad_norm": 0.26682353019714355,
+      "learning_rate": 0.001,
+      "loss": 2.5193,
+      "step": 2090
+    },
+    {
+      "epoch": 0.08845926051273373,
+      "grad_norm": 0.9872138500213623,
+      "learning_rate": 0.001,
+      "loss": 2.6831,
+      "step": 2091
+    },
+    {
+      "epoch": 0.08850156527625011,
+      "grad_norm": 0.9422023296356201,
+      "learning_rate": 0.001,
+      "loss": 3.6744,
+      "step": 2092
+    },
+    {
+      "epoch": 0.08854387003976648,
+      "grad_norm": 0.4777209758758545,
+      "learning_rate": 0.001,
+      "loss": 3.7379,
+      "step": 2093
+    },
+    {
+      "epoch": 0.08858617480328285,
+      "grad_norm": 104.47978210449219,
+      "learning_rate": 0.001,
+      "loss": 1.8196,
+      "step": 2094
+    },
+    {
+      "epoch": 0.08862847956679921,
+      "grad_norm": 17.303163528442383,
+      "learning_rate": 0.001,
+      "loss": 3.1286,
+      "step": 2095
+    },
+    {
+      "epoch": 0.0886707843303156,
+      "grad_norm": 0.3178851008415222,
+      "learning_rate": 0.001,
+      "loss": 3.5087,
+      "step": 2096
+    },
+    {
+      "epoch": 0.08871308909383196,
+      "grad_norm": 0.3228803873062134,
+      "learning_rate": 0.001,
+      "loss": 2.0294,
+      "step": 2097
+    },
+    {
+      "epoch": 0.08875539385734833,
+      "grad_norm": 0.6258364319801331,
+      "learning_rate": 0.001,
+      "loss": 2.5332,
+      "step": 2098
+    },
+    {
+      "epoch": 0.08879769862086472,
+      "grad_norm": 0.7965203523635864,
+      "learning_rate": 0.001,
+      "loss": 2.7295,
+      "step": 2099
+    },
+    {
+      "epoch": 0.08884000338438108,
+      "grad_norm": 0.4302770793437958,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08888230814789745,
+      "grad_norm": 0.5029553771018982,
+      "learning_rate": 0.001,
+      "loss": 2.9496,
+      "step": 2101
+    },
+    {
+      "epoch": 0.08892461291141382,
+      "grad_norm": 0.3942144215106964,
+      "learning_rate": 0.001,
+      "loss": 2.313,
+      "step": 2102
+    },
+    {
+      "epoch": 0.0889669176749302,
+      "grad_norm": 0.43013280630111694,
+      "learning_rate": 0.001,
+      "loss": 1.9479,
+      "step": 2103
+    },
+    {
+      "epoch": 0.08900922243844657,
+      "grad_norm": 1.0882439613342285,
+      "learning_rate": 0.001,
+      "loss": 2.8823,
+      "step": 2104
+    },
+    {
+      "epoch": 0.08905152720196294,
+      "grad_norm": 0.3308344781398773,
+      "learning_rate": 0.001,
+      "loss": 2.302,
+      "step": 2105
+    },
+    {
+      "epoch": 0.08909383196547932,
+      "grad_norm": 0.29911816120147705,
+      "learning_rate": 0.001,
+      "loss": 3.0091,
+      "step": 2106
+    },
+    {
+      "epoch": 0.08913613672899569,
+      "grad_norm": 1.754325270652771,
+      "learning_rate": 0.001,
+      "loss": 3.21,
+      "step": 2107
+    },
+    {
+      "epoch": 0.08917844149251206,
+      "grad_norm": 2.3082916736602783,
+      "learning_rate": 0.001,
+      "loss": 2.3856,
+      "step": 2108
+    },
+    {
+      "epoch": 0.08922074625602842,
+      "grad_norm": 0.6127236485481262,
+      "learning_rate": 0.001,
+      "loss": 2.2969,
+      "step": 2109
+    },
+    {
+      "epoch": 0.0892630510195448,
+      "grad_norm": 0.41946130990982056,
+      "learning_rate": 0.001,
+      "loss": 2.8269,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08930535578306117,
+      "grad_norm": 6.0178327560424805,
+      "learning_rate": 0.001,
+      "loss": 2.7265,
+      "step": 2111
+    },
+    {
+      "epoch": 0.08934766054657754,
+      "grad_norm": 0.38182544708251953,
+      "learning_rate": 0.001,
+      "loss": 3.1223,
+      "step": 2112
+    },
+    {
+      "epoch": 0.08938996531009391,
+      "grad_norm": 1.5331692695617676,
+      "learning_rate": 0.001,
+      "loss": 2.4643,
+      "step": 2113
+    },
+    {
+      "epoch": 0.08943227007361029,
+      "grad_norm": 0.29090458154678345,
+      "learning_rate": 0.001,
+      "loss": 1.958,
+      "step": 2114
+    },
+    {
+      "epoch": 0.08947457483712666,
+      "grad_norm": 0.3298831284046173,
+      "learning_rate": 0.001,
+      "loss": 2.4861,
+      "step": 2115
+    },
+    {
+      "epoch": 0.08951687960064303,
+      "grad_norm": 0.9520841240882874,
+      "learning_rate": 0.001,
+      "loss": 2.2431,
+      "step": 2116
+    },
+    {
+      "epoch": 0.08955918436415941,
+      "grad_norm": 0.3201279938220978,
+      "learning_rate": 0.001,
+      "loss": 1.9121,
+      "step": 2117
+    },
+    {
+      "epoch": 0.08960148912767578,
+      "grad_norm": 0.3153676390647888,
+      "learning_rate": 0.001,
+      "loss": 2.2245,
+      "step": 2118
+    },
+    {
+      "epoch": 0.08964379389119215,
+      "grad_norm": 0.425874263048172,
+      "learning_rate": 0.001,
+      "loss": 3.1083,
+      "step": 2119
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 0.6781908869743347,
+      "learning_rate": 0.001,
+      "loss": 3.0335,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0897284034182249,
+      "grad_norm": 0.38351890444755554,
+      "learning_rate": 0.001,
+      "loss": 3.6367,
+      "step": 2121
+    },
+    {
+      "epoch": 0.08977070818174127,
+      "grad_norm": 0.26305243372917175,
+      "learning_rate": 0.001,
+      "loss": 1.9731,
+      "step": 2122
+    },
+    {
+      "epoch": 0.08981301294525763,
+      "grad_norm": 6.707549571990967,
+      "learning_rate": 0.001,
+      "loss": 2.2013,
+      "step": 2123
+    },
+    {
+      "epoch": 0.089855317708774,
+      "grad_norm": 0.31812310218811035,
+      "learning_rate": 0.001,
+      "loss": 1.9723,
+      "step": 2124
+    },
+    {
+      "epoch": 0.08989762247229038,
+      "grad_norm": 0.30152562260627747,
+      "learning_rate": 0.001,
+      "loss": 2.6349,
+      "step": 2125
+    },
+    {
+      "epoch": 0.08993992723580675,
+      "grad_norm": 0.33307674527168274,
+      "learning_rate": 0.001,
+      "loss": 2.295,
+      "step": 2126
+    },
+    {
+      "epoch": 0.08998223199932312,
+      "grad_norm": 0.26768437027931213,
+      "learning_rate": 0.001,
+      "loss": 3.2849,
+      "step": 2127
+    },
+    {
+      "epoch": 0.0900245367628395,
+      "grad_norm": 0.39628762006759644,
+      "learning_rate": 0.001,
+      "loss": 3.1855,
+      "step": 2128
+    },
+    {
+      "epoch": 0.09006684152635587,
+      "grad_norm": 0.3016011416912079,
+      "learning_rate": 0.001,
+      "loss": 3.3723,
+      "step": 2129
+    },
+    {
+      "epoch": 0.09010914628987224,
+      "grad_norm": 0.6254957318305969,
+      "learning_rate": 0.001,
+      "loss": 4.1078,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0901514510533886,
+      "grad_norm": 0.23875102400779724,
+      "learning_rate": 0.001,
+      "loss": 1.7937,
+      "step": 2131
+    },
+    {
+      "epoch": 0.09019375581690499,
+      "grad_norm": 0.2940853536128998,
+      "learning_rate": 0.001,
+      "loss": 2.0315,
+      "step": 2132
+    },
+    {
+      "epoch": 0.09023606058042136,
+      "grad_norm": 2.5391368865966797,
+      "learning_rate": 0.001,
+      "loss": 2.5625,
+      "step": 2133
+    },
+    {
+      "epoch": 0.09027836534393772,
+      "grad_norm": 5.3973388671875,
+      "learning_rate": 0.001,
+      "loss": 1.8511,
+      "step": 2134
+    },
+    {
+      "epoch": 0.09032067010745409,
+      "grad_norm": 0.25178593397140503,
+      "learning_rate": 0.001,
+      "loss": 2.3371,
+      "step": 2135
+    },
+    {
+      "epoch": 0.09036297487097047,
+      "grad_norm": 0.31066349148750305,
+      "learning_rate": 0.001,
+      "loss": 2.7473,
+      "step": 2136
+    },
+    {
+      "epoch": 0.09040527963448684,
+      "grad_norm": 0.42618098855018616,
+      "learning_rate": 0.001,
+      "loss": 3.0467,
+      "step": 2137
+    },
+    {
+      "epoch": 0.09044758439800321,
+      "grad_norm": 0.29961666464805603,
+      "learning_rate": 0.001,
+      "loss": 2.192,
+      "step": 2138
+    },
+    {
+      "epoch": 0.0904898891615196,
+      "grad_norm": 0.31987568736076355,
+      "learning_rate": 0.001,
+      "loss": 2.7925,
+      "step": 2139
+    },
+    {
+      "epoch": 0.09053219392503596,
+      "grad_norm": 0.309462308883667,
+      "learning_rate": 0.001,
+      "loss": 2.0587,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09057449868855233,
+      "grad_norm": 0.3607901930809021,
+      "learning_rate": 0.001,
+      "loss": 2.1792,
+      "step": 2141
+    },
+    {
+      "epoch": 0.0906168034520687,
+      "grad_norm": 0.281773179769516,
+      "learning_rate": 0.001,
+      "loss": 1.8093,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09065910821558508,
+      "grad_norm": 0.5456468462944031,
+      "learning_rate": 0.001,
+      "loss": 3.7337,
+      "step": 2143
+    },
+    {
+      "epoch": 0.09070141297910145,
+      "grad_norm": 0.25825467705726624,
+      "learning_rate": 0.001,
+      "loss": 1.9948,
+      "step": 2144
+    },
+    {
+      "epoch": 0.09074371774261782,
+      "grad_norm": 0.9076284766197205,
+      "learning_rate": 0.001,
+      "loss": 2.3712,
+      "step": 2145
+    },
+    {
+      "epoch": 0.09078602250613418,
+      "grad_norm": 0.41249004006385803,
+      "learning_rate": 0.001,
+      "loss": 2.3084,
+      "step": 2146
+    },
+    {
+      "epoch": 0.09082832726965057,
+      "grad_norm": 1.1865334510803223,
+      "learning_rate": 0.001,
+      "loss": 3.0776,
+      "step": 2147
+    },
+    {
+      "epoch": 0.09087063203316693,
+      "grad_norm": 0.3349679410457611,
+      "learning_rate": 0.001,
+      "loss": 2.3653,
+      "step": 2148
+    },
+    {
+      "epoch": 0.0909129367966833,
+      "grad_norm": 0.4010538160800934,
+      "learning_rate": 0.001,
+      "loss": 2.3073,
+      "step": 2149
+    },
+    {
+      "epoch": 0.09095524156019968,
+      "grad_norm": 0.38532817363739014,
+      "learning_rate": 0.001,
+      "loss": 2.9012,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09099754632371605,
+      "grad_norm": 0.33620357513427734,
+      "learning_rate": 0.001,
+      "loss": 2.9892,
+      "step": 2151
+    },
+    {
+      "epoch": 0.09103985108723242,
+      "grad_norm": 5.379061222076416,
+      "learning_rate": 0.001,
+      "loss": 2.203,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09108215585074879,
+      "grad_norm": 8.182388305664062,
+      "learning_rate": 0.001,
+      "loss": 3.635,
+      "step": 2153
+    },
+    {
+      "epoch": 0.09112446061426517,
+      "grad_norm": 0.23541434109210968,
+      "learning_rate": 0.001,
+      "loss": 2.343,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09116676537778154,
+      "grad_norm": 0.27592089772224426,
+      "learning_rate": 0.001,
+      "loss": 2.7213,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09120907014129791,
+      "grad_norm": 0.30198514461517334,
+      "learning_rate": 0.001,
+      "loss": 3.0081,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09125137490481428,
+      "grad_norm": 0.26055535674095154,
+      "learning_rate": 0.001,
+      "loss": 2.4518,
+      "step": 2157
+    },
+    {
+      "epoch": 0.09129367966833066,
+      "grad_norm": 0.2565707564353943,
+      "learning_rate": 0.001,
+      "loss": 2.2259,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09133598443184703,
+      "grad_norm": 0.2991856634616852,
+      "learning_rate": 0.001,
+      "loss": 2.2095,
+      "step": 2159
+    },
+    {
+      "epoch": 0.0913782891953634,
+      "grad_norm": 0.6168603897094727,
+      "learning_rate": 0.001,
+      "loss": 2.6562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09142059395887978,
+      "grad_norm": 1.3440738916397095,
+      "learning_rate": 0.001,
+      "loss": 2.2589,
+      "step": 2161
+    },
+    {
+      "epoch": 0.09146289872239614,
+      "grad_norm": 0.2948399782180786,
+      "learning_rate": 0.001,
+      "loss": 2.4666,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09150520348591251,
+      "grad_norm": 0.3604365885257721,
+      "learning_rate": 0.001,
+      "loss": 2.1146,
+      "step": 2163
+    },
+    {
+      "epoch": 0.09154750824942888,
+      "grad_norm": 0.33632999658584595,
+      "learning_rate": 0.001,
+      "loss": 2.7935,
+      "step": 2164
+    },
+    {
+      "epoch": 0.09158981301294526,
+      "grad_norm": 0.49976199865341187,
+      "learning_rate": 0.001,
+      "loss": 2.0044,
+      "step": 2165
+    },
+    {
+      "epoch": 0.09163211777646163,
+      "grad_norm": 3.4067957401275635,
+      "learning_rate": 0.001,
+      "loss": 2.8011,
+      "step": 2166
+    },
+    {
+      "epoch": 0.091674422539978,
+      "grad_norm": 0.3377350866794586,
+      "learning_rate": 0.001,
+      "loss": 3.0546,
+      "step": 2167
+    },
+    {
+      "epoch": 0.09171672730349438,
+      "grad_norm": 0.3675040304660797,
+      "learning_rate": 0.001,
+      "loss": 2.6498,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09175903206701075,
+      "grad_norm": 0.30007094144821167,
+      "learning_rate": 0.001,
+      "loss": 3.8178,
+      "step": 2169
+    },
+    {
+      "epoch": 0.09180133683052712,
+      "grad_norm": 0.27063828706741333,
+      "learning_rate": 0.001,
+      "loss": 3.6035,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09184364159404348,
+      "grad_norm": 0.23841188848018646,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 2171
+    },
+    {
+      "epoch": 0.09188594635755987,
+      "grad_norm": 1.4631080627441406,
+      "learning_rate": 0.001,
+      "loss": 2.8047,
+      "step": 2172
+    },
+    {
+      "epoch": 0.09192825112107623,
+      "grad_norm": 0.26148828864097595,
+      "learning_rate": 0.001,
+      "loss": 1.7367,
+      "step": 2173
+    },
+    {
+      "epoch": 0.0919705558845926,
+      "grad_norm": 3.1010782718658447,
+      "learning_rate": 0.001,
+      "loss": 2.1477,
+      "step": 2174
+    },
+    {
+      "epoch": 0.09201286064810897,
+      "grad_norm": 1.398153305053711,
+      "learning_rate": 0.001,
+      "loss": 2.5751,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09205516541162535,
+      "grad_norm": 1.0668129920959473,
+      "learning_rate": 0.001,
+      "loss": 2.3153,
+      "step": 2176
+    },
+    {
+      "epoch": 0.09209747017514172,
+      "grad_norm": 0.2531263828277588,
+      "learning_rate": 0.001,
+      "loss": 3.1233,
+      "step": 2177
+    },
+    {
+      "epoch": 0.09213977493865809,
+      "grad_norm": 0.26637500524520874,
+      "learning_rate": 0.001,
+      "loss": 3.5828,
+      "step": 2178
+    },
+    {
+      "epoch": 0.09218207970217447,
+      "grad_norm": 0.2811260223388672,
+      "learning_rate": 0.001,
+      "loss": 2.9139,
+      "step": 2179
+    },
+    {
+      "epoch": 0.09222438446569084,
+      "grad_norm": 0.39359262585639954,
+      "learning_rate": 0.001,
+      "loss": 3.4906,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09226668922920721,
+      "grad_norm": 0.3012666702270508,
+      "learning_rate": 0.001,
+      "loss": 1.9615,
+      "step": 2181
+    },
+    {
+      "epoch": 0.09230899399272358,
+      "grad_norm": 0.5200200080871582,
+      "learning_rate": 0.001,
+      "loss": 2.8205,
+      "step": 2182
+    },
+    {
+      "epoch": 0.09235129875623996,
+      "grad_norm": 0.3044484257698059,
+      "learning_rate": 0.001,
+      "loss": 2.9967,
+      "step": 2183
+    },
+    {
+      "epoch": 0.09239360351975633,
+      "grad_norm": 0.3814965784549713,
+      "learning_rate": 0.001,
+      "loss": 2.1881,
+      "step": 2184
+    },
+    {
+      "epoch": 0.0924359082832727,
+      "grad_norm": 0.3739076554775238,
+      "learning_rate": 0.001,
+      "loss": 2.8411,
+      "step": 2185
+    },
+    {
+      "epoch": 0.09247821304678906,
+      "grad_norm": 0.3986753225326538,
+      "learning_rate": 0.001,
+      "loss": 2.6639,
+      "step": 2186
+    },
+    {
+      "epoch": 0.09252051781030544,
+      "grad_norm": 0.35847851634025574,
+      "learning_rate": 0.001,
+      "loss": 3.6791,
+      "step": 2187
+    },
+    {
+      "epoch": 0.09256282257382181,
+      "grad_norm": 0.3173019289970398,
+      "learning_rate": 0.001,
+      "loss": 2.2659,
+      "step": 2188
+    },
+    {
+      "epoch": 0.09260512733733818,
+      "grad_norm": 0.25434383749961853,
+      "learning_rate": 0.001,
+      "loss": 1.8084,
+      "step": 2189
+    },
+    {
+      "epoch": 0.09264743210085456,
+      "grad_norm": 0.2687484323978424,
+      "learning_rate": 0.001,
+      "loss": 2.9027,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09268973686437093,
+      "grad_norm": 0.29788804054260254,
+      "learning_rate": 0.001,
+      "loss": 3.081,
+      "step": 2191
+    },
+    {
+      "epoch": 0.0927320416278873,
+      "grad_norm": 2.9321951866149902,
+      "learning_rate": 0.001,
+      "loss": 4.4009,
+      "step": 2192
+    },
+    {
+      "epoch": 0.09277434639140367,
+      "grad_norm": 0.29507976770401,
+      "learning_rate": 0.001,
+      "loss": 2.3275,
+      "step": 2193
+    },
+    {
+      "epoch": 0.09281665115492005,
+      "grad_norm": 0.29394271969795227,
+      "learning_rate": 0.001,
+      "loss": 2.6158,
+      "step": 2194
+    },
+    {
+      "epoch": 0.09285895591843642,
+      "grad_norm": 1.7109583616256714,
+      "learning_rate": 0.001,
+      "loss": 2.0686,
+      "step": 2195
+    },
+    {
+      "epoch": 0.09290126068195279,
+      "grad_norm": 0.28951358795166016,
+      "learning_rate": 0.001,
+      "loss": 2.8997,
+      "step": 2196
+    },
+    {
+      "epoch": 0.09294356544546915,
+      "grad_norm": 0.49977269768714905,
+      "learning_rate": 0.001,
+      "loss": 2.8217,
+      "step": 2197
+    },
+    {
+      "epoch": 0.09298587020898554,
+      "grad_norm": 0.35717108845710754,
+      "learning_rate": 0.001,
+      "loss": 2.0279,
+      "step": 2198
+    },
+    {
+      "epoch": 0.0930281749725019,
+      "grad_norm": 0.22373569011688232,
+      "learning_rate": 0.001,
+      "loss": 1.7682,
+      "step": 2199
+    },
+    {
+      "epoch": 0.09307047973601827,
+      "grad_norm": 0.25105729699134827,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09311278449953465,
+      "grad_norm": 0.24033679068088531,
+      "learning_rate": 0.001,
+      "loss": 2.2573,
+      "step": 2201
+    },
+    {
+      "epoch": 0.09315508926305102,
+      "grad_norm": 0.3561880588531494,
+      "learning_rate": 0.001,
+      "loss": 3.0528,
+      "step": 2202
+    },
+    {
+      "epoch": 0.09319739402656739,
+      "grad_norm": 0.26465436816215515,
+      "learning_rate": 0.001,
+      "loss": 3.0083,
+      "step": 2203
+    },
+    {
+      "epoch": 0.09323969879008376,
+      "grad_norm": 1.4494379758834839,
+      "learning_rate": 0.001,
+      "loss": 2.5106,
+      "step": 2204
+    },
+    {
+      "epoch": 0.09328200355360014,
+      "grad_norm": 0.28080251812934875,
+      "learning_rate": 0.001,
+      "loss": 2.9088,
+      "step": 2205
+    },
+    {
+      "epoch": 0.09332430831711651,
+      "grad_norm": 0.6943084597587585,
+      "learning_rate": 0.001,
+      "loss": 2.1316,
+      "step": 2206
+    },
+    {
+      "epoch": 0.09336661308063288,
+      "grad_norm": 0.2942868173122406,
+      "learning_rate": 0.001,
+      "loss": 2.6873,
+      "step": 2207
+    },
+    {
+      "epoch": 0.09340891784414924,
+      "grad_norm": 0.4104631245136261,
+      "learning_rate": 0.001,
+      "loss": 2.3427,
+      "step": 2208
+    },
+    {
+      "epoch": 0.09345122260766563,
+      "grad_norm": 0.22013962268829346,
+      "learning_rate": 0.001,
+      "loss": 1.803,
+      "step": 2209
+    },
+    {
+      "epoch": 0.093493527371182,
+      "grad_norm": 0.24783039093017578,
+      "learning_rate": 0.001,
+      "loss": 2.1961,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09353583213469836,
+      "grad_norm": 0.34850063920021057,
+      "learning_rate": 0.001,
+      "loss": 2.5243,
+      "step": 2211
+    },
+    {
+      "epoch": 0.09357813689821474,
+      "grad_norm": 0.35803404450416565,
+      "learning_rate": 0.001,
+      "loss": 2.6485,
+      "step": 2212
+    },
+    {
+      "epoch": 0.09362044166173111,
+      "grad_norm": 4.5896124839782715,
+      "learning_rate": 0.001,
+      "loss": 2.4434,
+      "step": 2213
+    },
+    {
+      "epoch": 0.09366274642524748,
+      "grad_norm": 0.34777218103408813,
+      "learning_rate": 0.001,
+      "loss": 2.0321,
+      "step": 2214
+    },
+    {
+      "epoch": 0.09370505118876385,
+      "grad_norm": 0.3306042551994324,
+      "learning_rate": 0.001,
+      "loss": 2.4836,
+      "step": 2215
+    },
+    {
+      "epoch": 0.09374735595228023,
+      "grad_norm": 0.7642030715942383,
+      "learning_rate": 0.001,
+      "loss": 2.1946,
+      "step": 2216
+    },
+    {
+      "epoch": 0.0937896607157966,
+      "grad_norm": 0.37406593561172485,
+      "learning_rate": 0.001,
+      "loss": 3.6855,
+      "step": 2217
+    },
+    {
+      "epoch": 0.09383196547931297,
+      "grad_norm": 0.6254957318305969,
+      "learning_rate": 0.001,
+      "loss": 3.1232,
+      "step": 2218
+    },
+    {
+      "epoch": 0.09387427024282934,
+      "grad_norm": 0.31491386890411377,
+      "learning_rate": 0.001,
+      "loss": 2.757,
+      "step": 2219
+    },
+    {
+      "epoch": 0.09391657500634572,
+      "grad_norm": 0.37096840143203735,
+      "learning_rate": 0.001,
+      "loss": 2.4591,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09395887976986209,
+      "grad_norm": 0.30345040559768677,
+      "learning_rate": 0.001,
+      "loss": 2.2993,
+      "step": 2221
+    },
+    {
+      "epoch": 0.09400118453337845,
+      "grad_norm": 0.8886005282402039,
+      "learning_rate": 0.001,
+      "loss": 2.9874,
+      "step": 2222
+    },
+    {
+      "epoch": 0.09404348929689484,
+      "grad_norm": 0.5755462646484375,
+      "learning_rate": 0.001,
+      "loss": 2.5191,
+      "step": 2223
+    },
+    {
+      "epoch": 0.0940857940604112,
+      "grad_norm": 0.28378450870513916,
+      "learning_rate": 0.001,
+      "loss": 2.1818,
+      "step": 2224
+    },
+    {
+      "epoch": 0.09412809882392757,
+      "grad_norm": 0.4646637737751007,
+      "learning_rate": 0.001,
+      "loss": 2.155,
+      "step": 2225
+    },
+    {
+      "epoch": 0.09417040358744394,
+      "grad_norm": 0.28239914774894714,
+      "learning_rate": 0.001,
+      "loss": 3.0008,
+      "step": 2226
+    },
+    {
+      "epoch": 0.09421270835096032,
+      "grad_norm": 0.3244176506996155,
+      "learning_rate": 0.001,
+      "loss": 2.2789,
+      "step": 2227
+    },
+    {
+      "epoch": 0.09425501311447669,
+      "grad_norm": 0.25214463472366333,
+      "learning_rate": 0.001,
+      "loss": 1.8638,
+      "step": 2228
+    },
+    {
+      "epoch": 0.09429731787799306,
+      "grad_norm": 0.29388561844825745,
+      "learning_rate": 0.001,
+      "loss": 2.6425,
+      "step": 2229
+    },
+    {
+      "epoch": 0.09433962264150944,
+      "grad_norm": 0.7651423811912537,
+      "learning_rate": 0.001,
+      "loss": 2.5595,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09438192740502581,
+      "grad_norm": 0.7599115967750549,
+      "learning_rate": 0.001,
+      "loss": 2.6142,
+      "step": 2231
+    },
+    {
+      "epoch": 0.09442423216854218,
+      "grad_norm": 0.28916943073272705,
+      "learning_rate": 0.001,
+      "loss": 1.9283,
+      "step": 2232
+    },
+    {
+      "epoch": 0.09446653693205855,
+      "grad_norm": 0.2711436152458191,
+      "learning_rate": 0.001,
+      "loss": 2.0485,
+      "step": 2233
+    },
+    {
+      "epoch": 0.09450884169557493,
+      "grad_norm": 0.3689520061016083,
+      "learning_rate": 0.001,
+      "loss": 2.4108,
+      "step": 2234
+    },
+    {
+      "epoch": 0.0945511464590913,
+      "grad_norm": 0.3966144621372223,
+      "learning_rate": 0.001,
+      "loss": 2.4453,
+      "step": 2235
+    },
+    {
+      "epoch": 0.09459345122260766,
+      "grad_norm": 0.8278962969779968,
+      "learning_rate": 0.001,
+      "loss": 2.281,
+      "step": 2236
+    },
+    {
+      "epoch": 0.09463575598612403,
+      "grad_norm": 0.29549941420555115,
+      "learning_rate": 0.001,
+      "loss": 1.8836,
+      "step": 2237
+    },
+    {
+      "epoch": 0.09467806074964041,
+      "grad_norm": 0.563795268535614,
+      "learning_rate": 0.001,
+      "loss": 3.243,
+      "step": 2238
+    },
+    {
+      "epoch": 0.09472036551315678,
+      "grad_norm": 0.2910638749599457,
+      "learning_rate": 0.001,
+      "loss": 2.7178,
+      "step": 2239
+    },
+    {
+      "epoch": 0.09476267027667315,
+      "grad_norm": 1.0017096996307373,
+      "learning_rate": 0.001,
+      "loss": 2.5647,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09480497504018953,
+      "grad_norm": 0.2674773037433624,
+      "learning_rate": 0.001,
+      "loss": 2.4521,
+      "step": 2241
+    },
+    {
+      "epoch": 0.0948472798037059,
+      "grad_norm": 0.33499273657798767,
+      "learning_rate": 0.001,
+      "loss": 1.9591,
+      "step": 2242
+    },
+    {
+      "epoch": 0.09488958456722227,
+      "grad_norm": 0.28287482261657715,
+      "learning_rate": 0.001,
+      "loss": 1.8664,
+      "step": 2243
+    },
+    {
+      "epoch": 0.09493188933073864,
+      "grad_norm": 0.3287140130996704,
+      "learning_rate": 0.001,
+      "loss": 2.2835,
+      "step": 2244
+    },
+    {
+      "epoch": 0.09497419409425502,
+      "grad_norm": 0.28710776567459106,
+      "learning_rate": 0.001,
+      "loss": 2.1672,
+      "step": 2245
+    },
+    {
+      "epoch": 0.09501649885777139,
+      "grad_norm": 0.5307938456535339,
+      "learning_rate": 0.001,
+      "loss": 3.1055,
+      "step": 2246
+    },
+    {
+      "epoch": 0.09505880362128775,
+      "grad_norm": 0.2794020473957062,
+      "learning_rate": 0.001,
+      "loss": 1.4943,
+      "step": 2247
+    },
+    {
+      "epoch": 0.09510110838480412,
+      "grad_norm": 0.483715295791626,
+      "learning_rate": 0.001,
+      "loss": 1.7752,
+      "step": 2248
+    },
+    {
+      "epoch": 0.0951434131483205,
+      "grad_norm": 0.8411474227905273,
+      "learning_rate": 0.001,
+      "loss": 2.9864,
+      "step": 2249
+    },
+    {
+      "epoch": 0.09518571791183687,
+      "grad_norm": 0.30003300309181213,
+      "learning_rate": 0.001,
+      "loss": 4.0293,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09522802267535324,
+      "grad_norm": 0.2558903396129608,
+      "learning_rate": 0.001,
+      "loss": 2.1845,
+      "step": 2251
+    },
+    {
+      "epoch": 0.09527032743886962,
+      "grad_norm": 0.30151990056037903,
+      "learning_rate": 0.001,
+      "loss": 2.3792,
+      "step": 2252
+    },
+    {
+      "epoch": 0.09531263220238599,
+      "grad_norm": 0.29656192660331726,
+      "learning_rate": 0.001,
+      "loss": 3.2587,
+      "step": 2253
+    },
+    {
+      "epoch": 0.09535493696590236,
+      "grad_norm": 0.8995860815048218,
+      "learning_rate": 0.001,
+      "loss": 2.6727,
+      "step": 2254
+    },
+    {
+      "epoch": 0.09539724172941873,
+      "grad_norm": 0.3521322011947632,
+      "learning_rate": 0.001,
+      "loss": 2.4839,
+      "step": 2255
+    },
+    {
+      "epoch": 0.09543954649293511,
+      "grad_norm": 0.2579616904258728,
+      "learning_rate": 0.001,
+      "loss": 2.0373,
+      "step": 2256
+    },
+    {
+      "epoch": 0.09548185125645148,
+      "grad_norm": 2.846085548400879,
+      "learning_rate": 0.001,
+      "loss": 2.3873,
+      "step": 2257
+    },
+    {
+      "epoch": 0.09552415601996785,
+      "grad_norm": 0.28332656621932983,
+      "learning_rate": 0.001,
+      "loss": 2.426,
+      "step": 2258
+    },
+    {
+      "epoch": 0.09556646078348421,
+      "grad_norm": 1.6657432317733765,
+      "learning_rate": 0.001,
+      "loss": 2.0307,
+      "step": 2259
+    },
+    {
+      "epoch": 0.0956087655470006,
+      "grad_norm": 0.3576366901397705,
+      "learning_rate": 0.001,
+      "loss": 2.6188,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09565107031051696,
+      "grad_norm": 0.8456769585609436,
+      "learning_rate": 0.001,
+      "loss": 2.3674,
+      "step": 2261
+    },
+    {
+      "epoch": 0.09569337507403333,
+      "grad_norm": 0.3482474088668823,
+      "learning_rate": 0.001,
+      "loss": 3.4504,
+      "step": 2262
+    },
+    {
+      "epoch": 0.09573567983754971,
+      "grad_norm": 0.42225900292396545,
+      "learning_rate": 0.001,
+      "loss": 2.8715,
+      "step": 2263
+    },
+    {
+      "epoch": 0.09577798460106608,
+      "grad_norm": 1.451746940612793,
+      "learning_rate": 0.001,
+      "loss": 2.8526,
+      "step": 2264
+    },
+    {
+      "epoch": 0.09582028936458245,
+      "grad_norm": 0.3721246123313904,
+      "learning_rate": 0.001,
+      "loss": 2.4929,
+      "step": 2265
+    },
+    {
+      "epoch": 0.09586259412809882,
+      "grad_norm": 0.9242255091667175,
+      "learning_rate": 0.001,
+      "loss": 2.1786,
+      "step": 2266
+    },
+    {
+      "epoch": 0.0959048988916152,
+      "grad_norm": 0.4310086965560913,
+      "learning_rate": 0.001,
+      "loss": 3.3944,
+      "step": 2267
+    },
+    {
+      "epoch": 0.09594720365513157,
+      "grad_norm": 14.419943809509277,
+      "learning_rate": 0.001,
+      "loss": 1.8957,
+      "step": 2268
+    },
+    {
+      "epoch": 0.09598950841864794,
+      "grad_norm": 0.373689740896225,
+      "learning_rate": 0.001,
+      "loss": 2.3684,
+      "step": 2269
+    },
+    {
+      "epoch": 0.0960318131821643,
+      "grad_norm": 0.3977186381816864,
+      "learning_rate": 0.001,
+      "loss": 1.9556,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09607411794568069,
+      "grad_norm": 0.6978753209114075,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 2271
+    },
+    {
+      "epoch": 0.09611642270919706,
+      "grad_norm": 0.3750711679458618,
+      "learning_rate": 0.001,
+      "loss": 2.1204,
+      "step": 2272
+    },
+    {
+      "epoch": 0.09615872747271342,
+      "grad_norm": 0.5839630365371704,
+      "learning_rate": 0.001,
+      "loss": 2.1901,
+      "step": 2273
+    },
+    {
+      "epoch": 0.0962010322362298,
+      "grad_norm": 0.27384617924690247,
+      "learning_rate": 0.001,
+      "loss": 3.4993,
+      "step": 2274
+    },
+    {
+      "epoch": 0.09624333699974617,
+      "grad_norm": 0.3717498183250427,
+      "learning_rate": 0.001,
+      "loss": 1.9125,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09628564176326254,
+      "grad_norm": 0.34626662731170654,
+      "learning_rate": 0.001,
+      "loss": 2.2712,
+      "step": 2276
+    },
+    {
+      "epoch": 0.09632794652677891,
+      "grad_norm": 1.6960554122924805,
+      "learning_rate": 0.001,
+      "loss": 2.3714,
+      "step": 2277
+    },
+    {
+      "epoch": 0.09637025129029529,
+      "grad_norm": 0.8217235803604126,
+      "learning_rate": 0.001,
+      "loss": 2.2524,
+      "step": 2278
+    },
+    {
+      "epoch": 0.09641255605381166,
+      "grad_norm": 0.2900844216346741,
+      "learning_rate": 0.001,
+      "loss": 2.7666,
+      "step": 2279
+    },
+    {
+      "epoch": 0.09645486081732803,
+      "grad_norm": 0.47249212861061096,
+      "learning_rate": 0.001,
+      "loss": 2.5579,
+      "step": 2280
+    },
+    {
+      "epoch": 0.0964971655808444,
+      "grad_norm": 0.2934323847293854,
+      "learning_rate": 0.001,
+      "loss": 2.2327,
+      "step": 2281
+    },
+    {
+      "epoch": 0.09653947034436078,
+      "grad_norm": 0.7934165596961975,
+      "learning_rate": 0.001,
+      "loss": 2.5456,
+      "step": 2282
+    },
+    {
+      "epoch": 0.09658177510787715,
+      "grad_norm": 0.442386269569397,
+      "learning_rate": 0.001,
+      "loss": 2.7316,
+      "step": 2283
+    },
+    {
+      "epoch": 0.09662407987139351,
+      "grad_norm": 0.3879510760307312,
+      "learning_rate": 0.001,
+      "loss": 1.8956,
+      "step": 2284
+    },
+    {
+      "epoch": 0.0966663846349099,
+      "grad_norm": 0.42023125290870667,
+      "learning_rate": 0.001,
+      "loss": 3.0717,
+      "step": 2285
+    },
+    {
+      "epoch": 0.09670868939842626,
+      "grad_norm": 0.21355478465557098,
+      "learning_rate": 0.001,
+      "loss": 2.0076,
+      "step": 2286
+    },
+    {
+      "epoch": 0.09675099416194263,
+      "grad_norm": 10.379256248474121,
+      "learning_rate": 0.001,
+      "loss": 2.8718,
+      "step": 2287
+    },
+    {
+      "epoch": 0.096793298925459,
+      "grad_norm": 0.3485308289527893,
+      "learning_rate": 0.001,
+      "loss": 1.9105,
+      "step": 2288
+    },
+    {
+      "epoch": 0.09683560368897538,
+      "grad_norm": 1.0859586000442505,
+      "learning_rate": 0.001,
+      "loss": 1.7195,
+      "step": 2289
+    },
+    {
+      "epoch": 0.09687790845249175,
+      "grad_norm": 0.43755942583084106,
+      "learning_rate": 0.001,
+      "loss": 2.1547,
+      "step": 2290
+    },
+    {
+      "epoch": 0.09692021321600812,
+      "grad_norm": 0.26755252480506897,
+      "learning_rate": 0.001,
+      "loss": 1.9872,
+      "step": 2291
+    },
+    {
+      "epoch": 0.0969625179795245,
+      "grad_norm": 0.3288552463054657,
+      "learning_rate": 0.001,
+      "loss": 3.2151,
+      "step": 2292
+    },
+    {
+      "epoch": 0.09700482274304087,
+      "grad_norm": 0.3707546293735504,
+      "learning_rate": 0.001,
+      "loss": 2.7954,
+      "step": 2293
+    },
+    {
+      "epoch": 0.09704712750655724,
+      "grad_norm": 0.33741122484207153,
+      "learning_rate": 0.001,
+      "loss": 2.5806,
+      "step": 2294
+    },
+    {
+      "epoch": 0.0970894322700736,
+      "grad_norm": 0.29366111755371094,
+      "learning_rate": 0.001,
+      "loss": 2.161,
+      "step": 2295
+    },
+    {
+      "epoch": 0.09713173703358999,
+      "grad_norm": 0.29789918661117554,
+      "learning_rate": 0.001,
+      "loss": 2.6224,
+      "step": 2296
+    },
+    {
+      "epoch": 0.09717404179710636,
+      "grad_norm": 2.258840322494507,
+      "learning_rate": 0.001,
+      "loss": 1.8646,
+      "step": 2297
+    },
+    {
+      "epoch": 0.09721634656062272,
+      "grad_norm": 0.2928341329097748,
+      "learning_rate": 0.001,
+      "loss": 2.6383,
+      "step": 2298
+    },
+    {
+      "epoch": 0.09725865132413909,
+      "grad_norm": 0.42597708106040955,
+      "learning_rate": 0.001,
+      "loss": 2.1843,
+      "step": 2299
+    },
+    {
+      "epoch": 0.09730095608765547,
+      "grad_norm": 2.620967149734497,
+      "learning_rate": 0.001,
+      "loss": 2.0323,
+      "step": 2300
+    },
+    {
+      "epoch": 0.09734326085117184,
+      "grad_norm": 0.5456410646438599,
+      "learning_rate": 0.001,
+      "loss": 1.7493,
+      "step": 2301
+    },
+    {
+      "epoch": 0.09738556561468821,
+      "grad_norm": 0.28042513132095337,
+      "learning_rate": 0.001,
+      "loss": 2.5266,
+      "step": 2302
+    },
+    {
+      "epoch": 0.09742787037820459,
+      "grad_norm": 0.23970133066177368,
+      "learning_rate": 0.001,
+      "loss": 1.9479,
+      "step": 2303
+    },
+    {
+      "epoch": 0.09747017514172096,
+      "grad_norm": 0.40894538164138794,
+      "learning_rate": 0.001,
+      "loss": 1.8509,
+      "step": 2304
+    },
+    {
+      "epoch": 0.09751247990523733,
+      "grad_norm": 0.23067353665828705,
+      "learning_rate": 0.001,
+      "loss": 2.7972,
+      "step": 2305
+    },
+    {
+      "epoch": 0.0975547846687537,
+      "grad_norm": 0.36500483751296997,
+      "learning_rate": 0.001,
+      "loss": 2.0178,
+      "step": 2306
+    },
+    {
+      "epoch": 0.09759708943227008,
+      "grad_norm": 0.4092254936695099,
+      "learning_rate": 0.001,
+      "loss": 2.9206,
+      "step": 2307
+    },
+    {
+      "epoch": 0.09763939419578645,
+      "grad_norm": 0.49489879608154297,
+      "learning_rate": 0.001,
+      "loss": 2.1519,
+      "step": 2308
+    },
+    {
+      "epoch": 0.09768169895930282,
+      "grad_norm": 0.39644548296928406,
+      "learning_rate": 0.001,
+      "loss": 2.3865,
+      "step": 2309
+    },
+    {
+      "epoch": 0.09772400372281918,
+      "grad_norm": 0.4900401830673218,
+      "learning_rate": 0.001,
+      "loss": 1.6001,
+      "step": 2310
+    },
+    {
+      "epoch": 0.09776630848633557,
+      "grad_norm": 0.2798866629600525,
+      "learning_rate": 0.001,
+      "loss": 2.7082,
+      "step": 2311
+    },
+    {
+      "epoch": 0.09780861324985193,
+      "grad_norm": 0.23535476624965668,
+      "learning_rate": 0.001,
+      "loss": 1.938,
+      "step": 2312
+    },
+    {
+      "epoch": 0.0978509180133683,
+      "grad_norm": 0.27713868021965027,
+      "learning_rate": 0.001,
+      "loss": 2.7315,
+      "step": 2313
+    },
+    {
+      "epoch": 0.09789322277688468,
+      "grad_norm": 0.3740001618862152,
+      "learning_rate": 0.001,
+      "loss": 2.2113,
+      "step": 2314
+    },
+    {
+      "epoch": 0.09793552754040105,
+      "grad_norm": 0.23348340392112732,
+      "learning_rate": 0.001,
+      "loss": 2.2677,
+      "step": 2315
+    },
+    {
+      "epoch": 0.09797783230391742,
+      "grad_norm": 2.2800018787384033,
+      "learning_rate": 0.001,
+      "loss": 2.1095,
+      "step": 2316
+    },
+    {
+      "epoch": 0.09802013706743379,
+      "grad_norm": 0.2242593616247177,
+      "learning_rate": 0.001,
+      "loss": 2.0112,
+      "step": 2317
+    },
+    {
+      "epoch": 0.09806244183095017,
+      "grad_norm": 0.40007051825523376,
+      "learning_rate": 0.001,
+      "loss": 2.9323,
+      "step": 2318
+    },
+    {
+      "epoch": 0.09810474659446654,
+      "grad_norm": 0.32281118631362915,
+      "learning_rate": 0.001,
+      "loss": 2.8329,
+      "step": 2319
+    },
+    {
+      "epoch": 0.0981470513579829,
+      "grad_norm": 0.32091641426086426,
+      "learning_rate": 0.001,
+      "loss": 2.1386,
+      "step": 2320
+    },
+    {
+      "epoch": 0.09818935612149927,
+      "grad_norm": 0.372270792722702,
+      "learning_rate": 0.001,
+      "loss": 3.7813,
+      "step": 2321
+    },
+    {
+      "epoch": 0.09823166088501566,
+      "grad_norm": 0.23315389454364777,
+      "learning_rate": 0.001,
+      "loss": 1.7683,
+      "step": 2322
+    },
+    {
+      "epoch": 0.09827396564853202,
+      "grad_norm": 0.287507027387619,
+      "learning_rate": 0.001,
+      "loss": 2.4763,
+      "step": 2323
+    },
+    {
+      "epoch": 0.09831627041204839,
+      "grad_norm": 0.33321303129196167,
+      "learning_rate": 0.001,
+      "loss": 2.5295,
+      "step": 2324
+    },
+    {
+      "epoch": 0.09835857517556477,
+      "grad_norm": 0.40260010957717896,
+      "learning_rate": 0.001,
+      "loss": 2.4879,
+      "step": 2325
+    },
+    {
+      "epoch": 0.09840087993908114,
+      "grad_norm": 0.5056272745132446,
+      "learning_rate": 0.001,
+      "loss": 3.2216,
+      "step": 2326
+    },
+    {
+      "epoch": 0.09844318470259751,
+      "grad_norm": 6.566803455352783,
+      "learning_rate": 0.001,
+      "loss": 2.4137,
+      "step": 2327
+    },
+    {
+      "epoch": 0.09848548946611388,
+      "grad_norm": 0.37313172221183777,
+      "learning_rate": 0.001,
+      "loss": 2.8742,
+      "step": 2328
+    },
+    {
+      "epoch": 0.09852779422963026,
+      "grad_norm": 0.7859897613525391,
+      "learning_rate": 0.001,
+      "loss": 1.96,
+      "step": 2329
+    },
+    {
+      "epoch": 0.09857009899314663,
+      "grad_norm": 0.2522431015968323,
+      "learning_rate": 0.001,
+      "loss": 2.2632,
+      "step": 2330
+    },
+    {
+      "epoch": 0.098612403756663,
+      "grad_norm": 8.351020812988281,
+      "learning_rate": 0.001,
+      "loss": 3.1826,
+      "step": 2331
+    },
+    {
+      "epoch": 0.09865470852017937,
+      "grad_norm": 0.4760158360004425,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 2332
+    },
+    {
+      "epoch": 0.09869701328369575,
+      "grad_norm": 0.27335453033447266,
+      "learning_rate": 0.001,
+      "loss": 1.8718,
+      "step": 2333
+    },
+    {
+      "epoch": 0.09873931804721212,
+      "grad_norm": 0.3000699281692505,
+      "learning_rate": 0.001,
+      "loss": 2.1485,
+      "step": 2334
+    },
+    {
+      "epoch": 0.09878162281072848,
+      "grad_norm": 0.2664180099964142,
+      "learning_rate": 0.001,
+      "loss": 2.0335,
+      "step": 2335
+    },
+    {
+      "epoch": 0.09882392757424487,
+      "grad_norm": 1.341559886932373,
+      "learning_rate": 0.001,
+      "loss": 1.9941,
+      "step": 2336
+    },
+    {
+      "epoch": 0.09886623233776123,
+      "grad_norm": 0.2919972836971283,
+      "learning_rate": 0.001,
+      "loss": 2.7002,
+      "step": 2337
+    },
+    {
+      "epoch": 0.0989085371012776,
+      "grad_norm": 0.3330998718738556,
+      "learning_rate": 0.001,
+      "loss": 2.831,
+      "step": 2338
+    },
+    {
+      "epoch": 0.09895084186479397,
+      "grad_norm": 2.2501633167266846,
+      "learning_rate": 0.001,
+      "loss": 2.2368,
+      "step": 2339
+    },
+    {
+      "epoch": 0.09899314662831035,
+      "grad_norm": 0.7335495352745056,
+      "learning_rate": 0.001,
+      "loss": 3.2086,
+      "step": 2340
+    },
+    {
+      "epoch": 0.09903545139182672,
+      "grad_norm": 0.36334070563316345,
+      "learning_rate": 0.001,
+      "loss": 3.0316,
+      "step": 2341
+    },
+    {
+      "epoch": 0.09907775615534309,
+      "grad_norm": 0.35820385813713074,
+      "learning_rate": 0.001,
+      "loss": 2.6684,
+      "step": 2342
+    },
+    {
+      "epoch": 0.09912006091885946,
+      "grad_norm": 0.5417790412902832,
+      "learning_rate": 0.001,
+      "loss": 3.4955,
+      "step": 2343
+    },
+    {
+      "epoch": 0.09916236568237584,
+      "grad_norm": 0.6243907809257507,
+      "learning_rate": 0.001,
+      "loss": 2.7038,
+      "step": 2344
+    },
+    {
+      "epoch": 0.0992046704458922,
+      "grad_norm": 0.4269777834415436,
+      "learning_rate": 0.001,
+      "loss": 3.0679,
+      "step": 2345
+    },
+    {
+      "epoch": 0.09924697520940857,
+      "grad_norm": 0.45058220624923706,
+      "learning_rate": 0.001,
+      "loss": 2.0214,
+      "step": 2346
+    },
+    {
+      "epoch": 0.09928927997292496,
+      "grad_norm": 0.35213160514831543,
+      "learning_rate": 0.001,
+      "loss": 1.6954,
+      "step": 2347
+    },
+    {
+      "epoch": 0.09933158473644133,
+      "grad_norm": 0.36422258615493774,
+      "learning_rate": 0.001,
+      "loss": 2.605,
+      "step": 2348
+    },
+    {
+      "epoch": 0.0993738894999577,
+      "grad_norm": 0.4111798405647278,
+      "learning_rate": 0.001,
+      "loss": 2.9701,
+      "step": 2349
+    },
+    {
+      "epoch": 0.09941619426347406,
+      "grad_norm": 0.2559814453125,
+      "learning_rate": 0.001,
+      "loss": 2.0841,
+      "step": 2350
+    },
+    {
+      "epoch": 0.09945849902699044,
+      "grad_norm": 0.4313149154186249,
+      "learning_rate": 0.001,
+      "loss": 2.1517,
+      "step": 2351
+    },
+    {
+      "epoch": 0.09950080379050681,
+      "grad_norm": 0.7786276340484619,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 2352
+    },
+    {
+      "epoch": 0.09954310855402318,
+      "grad_norm": 0.30976659059524536,
+      "learning_rate": 0.001,
+      "loss": 2.1551,
+      "step": 2353
+    },
+    {
+      "epoch": 0.09958541331753956,
+      "grad_norm": 0.2895033061504364,
+      "learning_rate": 0.001,
+      "loss": 2.9111,
+      "step": 2354
+    },
+    {
+      "epoch": 0.09962771808105593,
+      "grad_norm": 0.28255394101142883,
+      "learning_rate": 0.001,
+      "loss": 2.9254,
+      "step": 2355
+    },
+    {
+      "epoch": 0.0996700228445723,
+      "grad_norm": 1.1467900276184082,
+      "learning_rate": 0.001,
+      "loss": 2.8531,
+      "step": 2356
+    },
+    {
+      "epoch": 0.09971232760808867,
+      "grad_norm": 0.2727132737636566,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 2357
+    },
+    {
+      "epoch": 0.09975463237160505,
+      "grad_norm": 0.2686462104320526,
+      "learning_rate": 0.001,
+      "loss": 3.5868,
+      "step": 2358
+    },
+    {
+      "epoch": 0.09979693713512142,
+      "grad_norm": 0.888546884059906,
+      "learning_rate": 0.001,
+      "loss": 2.3209,
+      "step": 2359
+    },
+    {
+      "epoch": 0.09983924189863778,
+      "grad_norm": 0.96397465467453,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09988154666215415,
+      "grad_norm": 0.3050881028175354,
+      "learning_rate": 0.001,
+      "loss": 2.7307,
+      "step": 2361
+    },
+    {
+      "epoch": 0.09992385142567053,
+      "grad_norm": 0.2523465156555176,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 2362
+    },
+    {
+      "epoch": 0.0999661561891869,
+      "grad_norm": 0.28606629371643066,
+      "learning_rate": 0.001,
+      "loss": 3.3748,
+      "step": 2363
+    },
+    {
+      "epoch": 0.10000846095270327,
+      "grad_norm": 0.25041037797927856,
+      "learning_rate": 0.001,
+      "loss": 1.5891,
+      "step": 2364
+    },
+    {
+      "epoch": 0.10005076571621965,
+      "grad_norm": 0.36506375670433044,
+      "learning_rate": 0.001,
+      "loss": 2.7191,
+      "step": 2365
+    },
+    {
+      "epoch": 0.10009307047973602,
+      "grad_norm": 0.30195456743240356,
+      "learning_rate": 0.001,
+      "loss": 2.418,
+      "step": 2366
+    },
+    {
+      "epoch": 0.10013537524325239,
+      "grad_norm": 0.2500908374786377,
+      "learning_rate": 0.001,
+      "loss": 1.9814,
+      "step": 2367
+    },
+    {
+      "epoch": 0.10017768000676876,
+      "grad_norm": 0.29026201367378235,
+      "learning_rate": 0.001,
+      "loss": 2.4507,
+      "step": 2368
+    },
+    {
+      "epoch": 0.10021998477028514,
+      "grad_norm": 0.28172945976257324,
+      "learning_rate": 0.001,
+      "loss": 2.6413,
+      "step": 2369
+    },
+    {
+      "epoch": 0.10026228953380151,
+      "grad_norm": 0.2788645625114441,
+      "learning_rate": 0.001,
+      "loss": 2.1827,
+      "step": 2370
+    },
+    {
+      "epoch": 0.10030459429731788,
+      "grad_norm": 0.24209265410900116,
+      "learning_rate": 0.001,
+      "loss": 2.1193,
+      "step": 2371
+    },
+    {
+      "epoch": 0.10034689906083424,
+      "grad_norm": 0.25518038868904114,
+      "learning_rate": 0.001,
+      "loss": 2.3543,
+      "step": 2372
+    },
+    {
+      "epoch": 0.10038920382435063,
+      "grad_norm": 0.28497177362442017,
+      "learning_rate": 0.001,
+      "loss": 2.3745,
+      "step": 2373
+    },
+    {
+      "epoch": 0.100431508587867,
+      "grad_norm": 0.49267375469207764,
+      "learning_rate": 0.001,
+      "loss": 1.9744,
+      "step": 2374
+    },
+    {
+      "epoch": 0.10047381335138336,
+      "grad_norm": 0.26035183668136597,
+      "learning_rate": 0.001,
+      "loss": 1.834,
+      "step": 2375
+    },
+    {
+      "epoch": 0.10051611811489974,
+      "grad_norm": 0.28036877512931824,
+      "learning_rate": 0.001,
+      "loss": 3.1908,
+      "step": 2376
+    },
+    {
+      "epoch": 0.10055842287841611,
+      "grad_norm": 1.9085135459899902,
+      "learning_rate": 0.001,
+      "loss": 3.0537,
+      "step": 2377
+    },
+    {
+      "epoch": 0.10060072764193248,
+      "grad_norm": 0.26107195019721985,
+      "learning_rate": 0.001,
+      "loss": 2.0188,
+      "step": 2378
+    },
+    {
+      "epoch": 0.10064303240544885,
+      "grad_norm": 0.2460990846157074,
+      "learning_rate": 0.001,
+      "loss": 3.2141,
+      "step": 2379
+    },
+    {
+      "epoch": 0.10068533716896523,
+      "grad_norm": 0.3403857946395874,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1007276419324816,
+      "grad_norm": 0.36075910925865173,
+      "learning_rate": 0.001,
+      "loss": 2.6347,
+      "step": 2381
+    },
+    {
+      "epoch": 0.10076994669599797,
+      "grad_norm": 2.8513433933258057,
+      "learning_rate": 0.001,
+      "loss": 2.6082,
+      "step": 2382
+    },
+    {
+      "epoch": 0.10081225145951433,
+      "grad_norm": 0.30714812874794006,
+      "learning_rate": 0.001,
+      "loss": 2.1303,
+      "step": 2383
+    },
+    {
+      "epoch": 0.10085455622303072,
+      "grad_norm": 0.26115673780441284,
+      "learning_rate": 0.001,
+      "loss": 2.4109,
+      "step": 2384
+    },
+    {
+      "epoch": 0.10089686098654709,
+      "grad_norm": 3.011416435241699,
+      "learning_rate": 0.001,
+      "loss": 2.6895,
+      "step": 2385
+    },
+    {
+      "epoch": 0.10093916575006345,
+      "grad_norm": 0.2893314063549042,
+      "learning_rate": 0.001,
+      "loss": 1.6992,
+      "step": 2386
+    },
+    {
+      "epoch": 0.10098147051357984,
+      "grad_norm": 0.411714106798172,
+      "learning_rate": 0.001,
+      "loss": 3.7138,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1010237752770962,
+      "grad_norm": 0.37683776021003723,
+      "learning_rate": 0.001,
+      "loss": 2.6581,
+      "step": 2388
+    },
+    {
+      "epoch": 0.10106608004061257,
+      "grad_norm": 0.8233429789543152,
+      "learning_rate": 0.001,
+      "loss": 3.1353,
+      "step": 2389
+    },
+    {
+      "epoch": 0.10110838480412894,
+      "grad_norm": 0.7344207167625427,
+      "learning_rate": 0.001,
+      "loss": 2.6045,
+      "step": 2390
+    },
+    {
+      "epoch": 0.10115068956764532,
+      "grad_norm": 0.28514331579208374,
+      "learning_rate": 0.001,
+      "loss": 1.9666,
+      "step": 2391
+    },
+    {
+      "epoch": 0.10119299433116169,
+      "grad_norm": 0.2906087040901184,
+      "learning_rate": 0.001,
+      "loss": 1.729,
+      "step": 2392
+    },
+    {
+      "epoch": 0.10123529909467806,
+      "grad_norm": 0.4669470191001892,
+      "learning_rate": 0.001,
+      "loss": 1.5726,
+      "step": 2393
+    },
+    {
+      "epoch": 0.10127760385819443,
+      "grad_norm": 0.44565969705581665,
+      "learning_rate": 0.001,
+      "loss": 2.1407,
+      "step": 2394
+    },
+    {
+      "epoch": 0.10131990862171081,
+      "grad_norm": 0.2725061774253845,
+      "learning_rate": 0.001,
+      "loss": 2.1667,
+      "step": 2395
+    },
+    {
+      "epoch": 0.10136221338522718,
+      "grad_norm": 0.49515077471733093,
+      "learning_rate": 0.001,
+      "loss": 2.4108,
+      "step": 2396
+    },
+    {
+      "epoch": 0.10140451814874354,
+      "grad_norm": 0.2847639322280884,
+      "learning_rate": 0.001,
+      "loss": 2.9909,
+      "step": 2397
+    },
+    {
+      "epoch": 0.10144682291225993,
+      "grad_norm": 0.8457419872283936,
+      "learning_rate": 0.001,
+      "loss": 3.0259,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1014891276757763,
+      "grad_norm": 9.33295726776123,
+      "learning_rate": 0.001,
+      "loss": 3.2422,
+      "step": 2399
+    },
+    {
+      "epoch": 0.10153143243929266,
+      "grad_norm": 1.8597520589828491,
+      "learning_rate": 0.001,
+      "loss": 2.6769,
+      "step": 2400
+    },
+    {
+      "epoch": 0.10157373720280903,
+      "grad_norm": 0.26994621753692627,
+      "learning_rate": 0.001,
+      "loss": 2.2029,
+      "step": 2401
+    },
+    {
+      "epoch": 0.10161604196632541,
+      "grad_norm": 1.4038335084915161,
+      "learning_rate": 0.001,
+      "loss": 3.4995,
+      "step": 2402
+    },
+    {
+      "epoch": 0.10165834672984178,
+      "grad_norm": 0.6501076221466064,
+      "learning_rate": 0.001,
+      "loss": 2.7349,
+      "step": 2403
+    },
+    {
+      "epoch": 0.10170065149335815,
+      "grad_norm": 0.21357211470603943,
+      "learning_rate": 0.001,
+      "loss": 1.9009,
+      "step": 2404
+    },
+    {
+      "epoch": 0.10174295625687452,
+      "grad_norm": 0.272206574678421,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1017852610203909,
+      "grad_norm": 0.23111356794834137,
+      "learning_rate": 0.001,
+      "loss": 2.2551,
+      "step": 2406
+    },
+    {
+      "epoch": 0.10182756578390727,
+      "grad_norm": 0.21537300944328308,
+      "learning_rate": 0.001,
+      "loss": 3.186,
+      "step": 2407
+    },
+    {
+      "epoch": 0.10186987054742364,
+      "grad_norm": 0.26508206129074097,
+      "learning_rate": 0.001,
+      "loss": 2.1786,
+      "step": 2408
+    },
+    {
+      "epoch": 0.10191217531094002,
+      "grad_norm": 3.251347303390503,
+      "learning_rate": 0.001,
+      "loss": 2.1077,
+      "step": 2409
+    },
+    {
+      "epoch": 0.10195448007445639,
+      "grad_norm": 0.2736663818359375,
+      "learning_rate": 0.001,
+      "loss": 2.0577,
+      "step": 2410
+    },
+    {
+      "epoch": 0.10199678483797275,
+      "grad_norm": 0.3979964852333069,
+      "learning_rate": 0.001,
+      "loss": 3.457,
+      "step": 2411
+    },
+    {
+      "epoch": 0.10203908960148912,
+      "grad_norm": 0.2330218404531479,
+      "learning_rate": 0.001,
+      "loss": 2.2031,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1020813943650055,
+      "grad_norm": 0.24427588284015656,
+      "learning_rate": 0.001,
+      "loss": 1.7257,
+      "step": 2413
+    },
+    {
+      "epoch": 0.10212369912852187,
+      "grad_norm": 0.7020580172538757,
+      "learning_rate": 0.001,
+      "loss": 3.6614,
+      "step": 2414
+    },
+    {
+      "epoch": 0.10216600389203824,
+      "grad_norm": 0.35481932759284973,
+      "learning_rate": 0.001,
+      "loss": 2.5094,
+      "step": 2415
+    },
+    {
+      "epoch": 0.10220830865555462,
+      "grad_norm": 10.207537651062012,
+      "learning_rate": 0.001,
+      "loss": 2.1178,
+      "step": 2416
+    },
+    {
+      "epoch": 0.10225061341907099,
+      "grad_norm": 0.3748272657394409,
+      "learning_rate": 0.001,
+      "loss": 2.5037,
+      "step": 2417
+    },
+    {
+      "epoch": 0.10229291818258736,
+      "grad_norm": 1.0302212238311768,
+      "learning_rate": 0.001,
+      "loss": 1.8157,
+      "step": 2418
+    },
+    {
+      "epoch": 0.10233522294610373,
+      "grad_norm": 0.35969147086143494,
+      "learning_rate": 0.001,
+      "loss": 2.2622,
+      "step": 2419
+    },
+    {
+      "epoch": 0.10237752770962011,
+      "grad_norm": 0.3533392548561096,
+      "learning_rate": 0.001,
+      "loss": 2.4072,
+      "step": 2420
+    },
+    {
+      "epoch": 0.10241983247313648,
+      "grad_norm": 0.3274601101875305,
+      "learning_rate": 0.001,
+      "loss": 2.4427,
+      "step": 2421
+    },
+    {
+      "epoch": 0.10246213723665284,
+      "grad_norm": 0.35946276783943176,
+      "learning_rate": 0.001,
+      "loss": 1.8442,
+      "step": 2422
+    },
+    {
+      "epoch": 0.10250444200016921,
+      "grad_norm": 0.3757915198802948,
+      "learning_rate": 0.001,
+      "loss": 3.0923,
+      "step": 2423
+    },
+    {
+      "epoch": 0.1025467467636856,
+      "grad_norm": 0.3471343219280243,
+      "learning_rate": 0.001,
+      "loss": 2.3881,
+      "step": 2424
+    },
+    {
+      "epoch": 0.10258905152720196,
+      "grad_norm": 0.2516452372074127,
+      "learning_rate": 0.001,
+      "loss": 2.0774,
+      "step": 2425
+    },
+    {
+      "epoch": 0.10263135629071833,
+      "grad_norm": 0.23940257728099823,
+      "learning_rate": 0.001,
+      "loss": 2.3271,
+      "step": 2426
+    },
+    {
+      "epoch": 0.10267366105423471,
+      "grad_norm": 0.8426430821418762,
+      "learning_rate": 0.001,
+      "loss": 2.4435,
+      "step": 2427
+    },
+    {
+      "epoch": 0.10271596581775108,
+      "grad_norm": 0.7691043019294739,
+      "learning_rate": 0.001,
+      "loss": 1.9913,
+      "step": 2428
+    },
+    {
+      "epoch": 0.10275827058126745,
+      "grad_norm": 0.29252296686172485,
+      "learning_rate": 0.001,
+      "loss": 3.3262,
+      "step": 2429
+    },
+    {
+      "epoch": 0.10280057534478382,
+      "grad_norm": 0.25585702061653137,
+      "learning_rate": 0.001,
+      "loss": 2.3882,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1028428801083002,
+      "grad_norm": 2.2944388389587402,
+      "learning_rate": 0.001,
+      "loss": 2.7547,
+      "step": 2431
+    },
+    {
+      "epoch": 0.10288518487181657,
+      "grad_norm": 0.3069431185722351,
+      "learning_rate": 0.001,
+      "loss": 2.5555,
+      "step": 2432
+    },
+    {
+      "epoch": 0.10292748963533294,
+      "grad_norm": 0.6198341250419617,
+      "learning_rate": 0.001,
+      "loss": 2.2501,
+      "step": 2433
+    },
+    {
+      "epoch": 0.1029697943988493,
+      "grad_norm": 0.23452864587306976,
+      "learning_rate": 0.001,
+      "loss": 2.0086,
+      "step": 2434
+    },
+    {
+      "epoch": 0.10301209916236569,
+      "grad_norm": 0.27118051052093506,
+      "learning_rate": 0.001,
+      "loss": 1.6991,
+      "step": 2435
+    },
+    {
+      "epoch": 0.10305440392588205,
+      "grad_norm": 0.2895638048648834,
+      "learning_rate": 0.001,
+      "loss": 2.391,
+      "step": 2436
+    },
+    {
+      "epoch": 0.10309670868939842,
+      "grad_norm": 0.5216059684753418,
+      "learning_rate": 0.001,
+      "loss": 2.0853,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1031390134529148,
+      "grad_norm": 0.32230275869369507,
+      "learning_rate": 0.001,
+      "loss": 2.3947,
+      "step": 2438
+    },
+    {
+      "epoch": 0.10318131821643117,
+      "grad_norm": 0.34388068318367004,
+      "learning_rate": 0.001,
+      "loss": 2.6358,
+      "step": 2439
+    },
+    {
+      "epoch": 0.10322362297994754,
+      "grad_norm": 0.3274713456630707,
+      "learning_rate": 0.001,
+      "loss": 3.2649,
+      "step": 2440
+    },
+    {
+      "epoch": 0.10326592774346391,
+      "grad_norm": 0.24570752680301666,
+      "learning_rate": 0.001,
+      "loss": 2.3323,
+      "step": 2441
+    },
+    {
+      "epoch": 0.10330823250698029,
+      "grad_norm": 0.23743736743927002,
+      "learning_rate": 0.001,
+      "loss": 1.6279,
+      "step": 2442
+    },
+    {
+      "epoch": 0.10335053727049666,
+      "grad_norm": 0.2327599972486496,
+      "learning_rate": 0.001,
+      "loss": 2.0293,
+      "step": 2443
+    },
+    {
+      "epoch": 0.10339284203401303,
+      "grad_norm": 0.24773749709129333,
+      "learning_rate": 0.001,
+      "loss": 2.1824,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1034351467975294,
+      "grad_norm": 0.3534669578075409,
+      "learning_rate": 0.001,
+      "loss": 2.7218,
+      "step": 2445
+    },
+    {
+      "epoch": 0.10347745156104578,
+      "grad_norm": 0.5942555069923401,
+      "learning_rate": 0.001,
+      "loss": 2.3746,
+      "step": 2446
+    },
+    {
+      "epoch": 0.10351975632456215,
+      "grad_norm": 0.26471078395843506,
+      "learning_rate": 0.001,
+      "loss": 2.2947,
+      "step": 2447
+    },
+    {
+      "epoch": 0.10356206108807851,
+      "grad_norm": 0.3464234173297882,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 2448
+    },
+    {
+      "epoch": 0.1036043658515949,
+      "grad_norm": 1.5699352025985718,
+      "learning_rate": 0.001,
+      "loss": 2.5707,
+      "step": 2449
+    },
+    {
+      "epoch": 0.10364667061511126,
+      "grad_norm": 1.2185924053192139,
+      "learning_rate": 0.001,
+      "loss": 2.956,
+      "step": 2450
+    },
+    {
+      "epoch": 0.10368897537862763,
+      "grad_norm": 0.2412140667438507,
+      "learning_rate": 0.001,
+      "loss": 2.1361,
+      "step": 2451
+    },
+    {
+      "epoch": 0.103731280142144,
+      "grad_norm": 1.7203179597854614,
+      "learning_rate": 0.001,
+      "loss": 2.7507,
+      "step": 2452
+    },
+    {
+      "epoch": 0.10377358490566038,
+      "grad_norm": 0.40216708183288574,
+      "learning_rate": 0.001,
+      "loss": 2.3239,
+      "step": 2453
+    },
+    {
+      "epoch": 0.10381588966917675,
+      "grad_norm": 0.41361749172210693,
+      "learning_rate": 0.001,
+      "loss": 2.4953,
+      "step": 2454
+    },
+    {
+      "epoch": 0.10385819443269312,
+      "grad_norm": 2.6895222663879395,
+      "learning_rate": 0.001,
+      "loss": 2.6164,
+      "step": 2455
+    },
+    {
+      "epoch": 0.10390049919620949,
+      "grad_norm": 0.6734895706176758,
+      "learning_rate": 0.001,
+      "loss": 2.5122,
+      "step": 2456
+    },
+    {
+      "epoch": 0.10394280395972587,
+      "grad_norm": 3.203244686126709,
+      "learning_rate": 0.001,
+      "loss": 2.6159,
+      "step": 2457
+    },
+    {
+      "epoch": 0.10398510872324224,
+      "grad_norm": 7.857353687286377,
+      "learning_rate": 0.001,
+      "loss": 2.2798,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1040274134867586,
+      "grad_norm": 0.4832097589969635,
+      "learning_rate": 0.001,
+      "loss": 2.6077,
+      "step": 2459
+    },
+    {
+      "epoch": 0.10406971825027499,
+      "grad_norm": 1.2383735179901123,
+      "learning_rate": 0.001,
+      "loss": 2.2514,
+      "step": 2460
+    },
+    {
+      "epoch": 0.10411202301379135,
+      "grad_norm": 0.27097123861312866,
+      "learning_rate": 0.001,
+      "loss": 1.8764,
+      "step": 2461
+    },
+    {
+      "epoch": 0.10415432777730772,
+      "grad_norm": 0.2842322587966919,
+      "learning_rate": 0.001,
+      "loss": 2.4073,
+      "step": 2462
+    },
+    {
+      "epoch": 0.10419663254082409,
+      "grad_norm": 0.31297391653060913,
+      "learning_rate": 0.001,
+      "loss": 1.8048,
+      "step": 2463
+    },
+    {
+      "epoch": 0.10423893730434047,
+      "grad_norm": 0.4402758479118347,
+      "learning_rate": 0.001,
+      "loss": 3.4671,
+      "step": 2464
+    },
+    {
+      "epoch": 0.10428124206785684,
+      "grad_norm": 0.28556615114212036,
+      "learning_rate": 0.001,
+      "loss": 2.2693,
+      "step": 2465
+    },
+    {
+      "epoch": 0.10432354683137321,
+      "grad_norm": 0.2614104151725769,
+      "learning_rate": 0.001,
+      "loss": 1.8248,
+      "step": 2466
+    },
+    {
+      "epoch": 0.10436585159488958,
+      "grad_norm": 0.24164602160453796,
+      "learning_rate": 0.001,
+      "loss": 2.1898,
+      "step": 2467
+    },
+    {
+      "epoch": 0.10440815635840596,
+      "grad_norm": 0.3009679913520813,
+      "learning_rate": 0.001,
+      "loss": 3.0112,
+      "step": 2468
+    },
+    {
+      "epoch": 0.10445046112192233,
+      "grad_norm": 0.3488333821296692,
+      "learning_rate": 0.001,
+      "loss": 2.0925,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1044927658854387,
+      "grad_norm": 0.6833105087280273,
+      "learning_rate": 0.001,
+      "loss": 2.22,
+      "step": 2470
+    },
+    {
+      "epoch": 0.10453507064895508,
+      "grad_norm": 0.28377044200897217,
+      "learning_rate": 0.001,
+      "loss": 2.2993,
+      "step": 2471
+    },
+    {
+      "epoch": 0.10457737541247145,
+      "grad_norm": 0.3093280792236328,
+      "learning_rate": 0.001,
+      "loss": 2.701,
+      "step": 2472
+    },
+    {
+      "epoch": 0.10461968017598781,
+      "grad_norm": 0.31424522399902344,
+      "learning_rate": 0.001,
+      "loss": 2.6909,
+      "step": 2473
+    },
+    {
+      "epoch": 0.10466198493950418,
+      "grad_norm": 0.23883268237113953,
+      "learning_rate": 0.001,
+      "loss": 3.2359,
+      "step": 2474
+    },
+    {
+      "epoch": 0.10470428970302056,
+      "grad_norm": 0.3218192458152771,
+      "learning_rate": 0.001,
+      "loss": 2.6222,
+      "step": 2475
+    },
+    {
+      "epoch": 0.10474659446653693,
+      "grad_norm": 0.3537428677082062,
+      "learning_rate": 0.001,
+      "loss": 2.6754,
+      "step": 2476
+    },
+    {
+      "epoch": 0.1047888992300533,
+      "grad_norm": 0.2585582137107849,
+      "learning_rate": 0.001,
+      "loss": 2.7918,
+      "step": 2477
+    },
+    {
+      "epoch": 0.10483120399356968,
+      "grad_norm": 0.28491732478141785,
+      "learning_rate": 0.001,
+      "loss": 1.814,
+      "step": 2478
+    },
+    {
+      "epoch": 0.10487350875708605,
+      "grad_norm": 0.31250932812690735,
+      "learning_rate": 0.001,
+      "loss": 2.1781,
+      "step": 2479
+    },
+    {
+      "epoch": 0.10491581352060242,
+      "grad_norm": 0.6643452644348145,
+      "learning_rate": 0.001,
+      "loss": 4.0611,
+      "step": 2480
+    },
+    {
+      "epoch": 0.10495811828411879,
+      "grad_norm": 0.2877008318901062,
+      "learning_rate": 0.001,
+      "loss": 2.3672,
+      "step": 2481
+    },
+    {
+      "epoch": 0.10500042304763517,
+      "grad_norm": 0.23607558012008667,
+      "learning_rate": 0.001,
+      "loss": 1.5298,
+      "step": 2482
+    },
+    {
+      "epoch": 0.10504272781115154,
+      "grad_norm": 3.870473623275757,
+      "learning_rate": 0.001,
+      "loss": 2.4808,
+      "step": 2483
+    },
+    {
+      "epoch": 0.1050850325746679,
+      "grad_norm": 0.29349610209465027,
+      "learning_rate": 0.001,
+      "loss": 1.9366,
+      "step": 2484
+    },
+    {
+      "epoch": 0.10512733733818427,
+      "grad_norm": 0.23734307289123535,
+      "learning_rate": 0.001,
+      "loss": 2.6115,
+      "step": 2485
+    },
+    {
+      "epoch": 0.10516964210170066,
+      "grad_norm": 0.8626511693000793,
+      "learning_rate": 0.001,
+      "loss": 2.4259,
+      "step": 2486
+    },
+    {
+      "epoch": 0.10521194686521702,
+      "grad_norm": 5.853229999542236,
+      "learning_rate": 0.001,
+      "loss": 2.406,
+      "step": 2487
+    },
+    {
+      "epoch": 0.10525425162873339,
+      "grad_norm": 0.23822012543678284,
+      "learning_rate": 0.001,
+      "loss": 3.1322,
+      "step": 2488
+    },
+    {
+      "epoch": 0.10529655639224977,
+      "grad_norm": 0.2993183732032776,
+      "learning_rate": 0.001,
+      "loss": 3.2551,
+      "step": 2489
+    },
+    {
+      "epoch": 0.10533886115576614,
+      "grad_norm": 1.8715354204177856,
+      "learning_rate": 0.001,
+      "loss": 2.2385,
+      "step": 2490
+    },
+    {
+      "epoch": 0.10538116591928251,
+      "grad_norm": 0.3598434329032898,
+      "learning_rate": 0.001,
+      "loss": 2.4187,
+      "step": 2491
+    },
+    {
+      "epoch": 0.10542347068279888,
+      "grad_norm": 0.5928519368171692,
+      "learning_rate": 0.001,
+      "loss": 1.9091,
+      "step": 2492
+    },
+    {
+      "epoch": 0.10546577544631526,
+      "grad_norm": 1.1429487466812134,
+      "learning_rate": 0.001,
+      "loss": 2.5143,
+      "step": 2493
+    },
+    {
+      "epoch": 0.10550808020983163,
+      "grad_norm": 2.955080270767212,
+      "learning_rate": 0.001,
+      "loss": 2.2716,
+      "step": 2494
+    },
+    {
+      "epoch": 0.105550384973348,
+      "grad_norm": 0.3205156922340393,
+      "learning_rate": 0.001,
+      "loss": 3.5673,
+      "step": 2495
+    },
+    {
+      "epoch": 0.10559268973686436,
+      "grad_norm": 0.31520524621009827,
+      "learning_rate": 0.001,
+      "loss": 2.6269,
+      "step": 2496
+    },
+    {
+      "epoch": 0.10563499450038075,
+      "grad_norm": 0.2760033905506134,
+      "learning_rate": 0.001,
+      "loss": 2.1043,
+      "step": 2497
+    },
+    {
+      "epoch": 0.10567729926389711,
+      "grad_norm": 0.23244822025299072,
+      "learning_rate": 0.001,
+      "loss": 2.583,
+      "step": 2498
+    },
+    {
+      "epoch": 0.10571960402741348,
+      "grad_norm": 0.24715182185173035,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 2499
+    },
+    {
+      "epoch": 0.10576190879092986,
+      "grad_norm": 0.3001532256603241,
+      "learning_rate": 0.001,
+      "loss": 2.2312,
+      "step": 2500
+    },
+    {
+      "epoch": 0.10580421355444623,
+      "grad_norm": 0.2582187056541443,
+      "learning_rate": 0.001,
+      "loss": 2.7161,
+      "step": 2501
+    },
+    {
+      "epoch": 0.1058465183179626,
+      "grad_norm": 0.6607329249382019,
+      "learning_rate": 0.001,
+      "loss": 2.7597,
+      "step": 2502
+    },
+    {
+      "epoch": 0.10588882308147897,
+      "grad_norm": 0.2999674677848816,
+      "learning_rate": 0.001,
+      "loss": 2.7306,
+      "step": 2503
+    },
+    {
+      "epoch": 0.10593112784499535,
+      "grad_norm": 0.2406844049692154,
+      "learning_rate": 0.001,
+      "loss": 1.7858,
+      "step": 2504
+    },
+    {
+      "epoch": 0.10597343260851172,
+      "grad_norm": 0.23513177037239075,
+      "learning_rate": 0.001,
+      "loss": 2.072,
+      "step": 2505
+    },
+    {
+      "epoch": 0.10601573737202809,
+      "grad_norm": 0.2586327791213989,
+      "learning_rate": 0.001,
+      "loss": 2.1561,
+      "step": 2506
+    },
+    {
+      "epoch": 0.10605804213554446,
+      "grad_norm": 0.2618561089038849,
+      "learning_rate": 0.001,
+      "loss": 1.7579,
+      "step": 2507
+    },
+    {
+      "epoch": 0.10610034689906084,
+      "grad_norm": 0.34452149271965027,
+      "learning_rate": 0.001,
+      "loss": 2.1415,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1061426516625772,
+      "grad_norm": 0.3686632513999939,
+      "learning_rate": 0.001,
+      "loss": 2.3883,
+      "step": 2509
+    },
+    {
+      "epoch": 0.10618495642609357,
+      "grad_norm": 0.2636253535747528,
+      "learning_rate": 0.001,
+      "loss": 2.2841,
+      "step": 2510
+    },
+    {
+      "epoch": 0.10622726118960996,
+      "grad_norm": 0.2659060060977936,
+      "learning_rate": 0.001,
+      "loss": 2.5267,
+      "step": 2511
+    },
+    {
+      "epoch": 0.10626956595312632,
+      "grad_norm": 0.37761518359184265,
+      "learning_rate": 0.001,
+      "loss": 2.7947,
+      "step": 2512
+    },
+    {
+      "epoch": 0.10631187071664269,
+      "grad_norm": 0.8648855090141296,
+      "learning_rate": 0.001,
+      "loss": 2.7321,
+      "step": 2513
+    },
+    {
+      "epoch": 0.10635417548015906,
+      "grad_norm": 0.26311472058296204,
+      "learning_rate": 0.001,
+      "loss": 1.881,
+      "step": 2514
+    },
+    {
+      "epoch": 0.10639648024367544,
+      "grad_norm": 0.344380259513855,
+      "learning_rate": 0.001,
+      "loss": 2.3485,
+      "step": 2515
+    },
+    {
+      "epoch": 0.10643878500719181,
+      "grad_norm": 0.35995468497276306,
+      "learning_rate": 0.001,
+      "loss": 2.1559,
+      "step": 2516
+    },
+    {
+      "epoch": 0.10648108977070818,
+      "grad_norm": 0.32800522446632385,
+      "learning_rate": 0.001,
+      "loss": 2.0882,
+      "step": 2517
+    },
+    {
+      "epoch": 0.10652339453422455,
+      "grad_norm": 0.3281853497028351,
+      "learning_rate": 0.001,
+      "loss": 2.4298,
+      "step": 2518
+    },
+    {
+      "epoch": 0.10656569929774093,
+      "grad_norm": 0.25847840309143066,
+      "learning_rate": 0.001,
+      "loss": 2.1247,
+      "step": 2519
+    },
+    {
+      "epoch": 0.1066080040612573,
+      "grad_norm": 0.6328796148300171,
+      "learning_rate": 0.001,
+      "loss": 2.7698,
+      "step": 2520
+    },
+    {
+      "epoch": 0.10665030882477367,
+      "grad_norm": 1.8759524822235107,
+      "learning_rate": 0.001,
+      "loss": 2.4454,
+      "step": 2521
+    },
+    {
+      "epoch": 0.10669261358829005,
+      "grad_norm": 0.2672816812992096,
+      "learning_rate": 0.001,
+      "loss": 2.3352,
+      "step": 2522
+    },
+    {
+      "epoch": 0.10673491835180642,
+      "grad_norm": 0.22392208874225616,
+      "learning_rate": 0.001,
+      "loss": 1.6423,
+      "step": 2523
+    },
+    {
+      "epoch": 0.10677722311532278,
+      "grad_norm": 0.43734171986579895,
+      "learning_rate": 0.001,
+      "loss": 2.8971,
+      "step": 2524
+    },
+    {
+      "epoch": 0.10681952787883915,
+      "grad_norm": 0.28167760372161865,
+      "learning_rate": 0.001,
+      "loss": 1.9086,
+      "step": 2525
+    },
+    {
+      "epoch": 0.10686183264235553,
+      "grad_norm": 0.2657049000263214,
+      "learning_rate": 0.001,
+      "loss": 3.812,
+      "step": 2526
+    },
+    {
+      "epoch": 0.1069041374058719,
+      "grad_norm": 0.26134321093559265,
+      "learning_rate": 0.001,
+      "loss": 2.2141,
+      "step": 2527
+    },
+    {
+      "epoch": 0.10694644216938827,
+      "grad_norm": 0.271685928106308,
+      "learning_rate": 0.001,
+      "loss": 2.1261,
+      "step": 2528
+    },
+    {
+      "epoch": 0.10698874693290465,
+      "grad_norm": 0.7208402156829834,
+      "learning_rate": 0.001,
+      "loss": 2.3952,
+      "step": 2529
+    },
+    {
+      "epoch": 0.10703105169642102,
+      "grad_norm": 0.5197535157203674,
+      "learning_rate": 0.001,
+      "loss": 2.6504,
+      "step": 2530
+    },
+    {
+      "epoch": 0.10707335645993739,
+      "grad_norm": 0.27384886145591736,
+      "learning_rate": 0.001,
+      "loss": 3.3872,
+      "step": 2531
+    },
+    {
+      "epoch": 0.10711566122345376,
+      "grad_norm": 4.649167060852051,
+      "learning_rate": 0.001,
+      "loss": 2.3127,
+      "step": 2532
+    },
+    {
+      "epoch": 0.10715796598697014,
+      "grad_norm": 0.5148127675056458,
+      "learning_rate": 0.001,
+      "loss": 3.0018,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1072002707504865,
+      "grad_norm": 1.9632915258407593,
+      "learning_rate": 0.001,
+      "loss": 3.2377,
+      "step": 2534
+    },
+    {
+      "epoch": 0.10724257551400287,
+      "grad_norm": 1.258927583694458,
+      "learning_rate": 0.001,
+      "loss": 1.5741,
+      "step": 2535
+    },
+    {
+      "epoch": 0.10728488027751924,
+      "grad_norm": 0.38562342524528503,
+      "learning_rate": 0.001,
+      "loss": 2.2438,
+      "step": 2536
+    },
+    {
+      "epoch": 0.10732718504103562,
+      "grad_norm": 0.4237939715385437,
+      "learning_rate": 0.001,
+      "loss": 2.8422,
+      "step": 2537
+    },
+    {
+      "epoch": 0.10736948980455199,
+      "grad_norm": 12.856072425842285,
+      "learning_rate": 0.001,
+      "loss": 1.9867,
+      "step": 2538
+    },
+    {
+      "epoch": 0.10741179456806836,
+      "grad_norm": 0.3507222533226013,
+      "learning_rate": 0.001,
+      "loss": 2.4188,
+      "step": 2539
+    },
+    {
+      "epoch": 0.10745409933158474,
+      "grad_norm": 0.29266253113746643,
+      "learning_rate": 0.001,
+      "loss": 2.3492,
+      "step": 2540
+    },
+    {
+      "epoch": 0.10749640409510111,
+      "grad_norm": 0.31082549691200256,
+      "learning_rate": 0.001,
+      "loss": 2.4325,
+      "step": 2541
+    },
+    {
+      "epoch": 0.10753870885861748,
+      "grad_norm": 0.41879016160964966,
+      "learning_rate": 0.001,
+      "loss": 3.5761,
+      "step": 2542
+    },
+    {
+      "epoch": 0.10758101362213385,
+      "grad_norm": 0.2282257229089737,
+      "learning_rate": 0.001,
+      "loss": 2.143,
+      "step": 2543
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 3.6079633235931396,
+      "learning_rate": 0.001,
+      "loss": 2.3143,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1076656231491666,
+      "grad_norm": 0.24503585696220398,
+      "learning_rate": 0.001,
+      "loss": 2.5591,
+      "step": 2545
+    },
+    {
+      "epoch": 0.10770792791268297,
+      "grad_norm": 0.49733033776283264,
+      "learning_rate": 0.001,
+      "loss": 2.3954,
+      "step": 2546
+    },
+    {
+      "epoch": 0.10775023267619933,
+      "grad_norm": 1.210880994796753,
+      "learning_rate": 0.001,
+      "loss": 2.9697,
+      "step": 2547
+    },
+    {
+      "epoch": 0.10779253743971572,
+      "grad_norm": 0.6089571714401245,
+      "learning_rate": 0.001,
+      "loss": 2.2833,
+      "step": 2548
+    },
+    {
+      "epoch": 0.10783484220323208,
+      "grad_norm": 0.2385062873363495,
+      "learning_rate": 0.001,
+      "loss": 3.0483,
+      "step": 2549
+    },
+    {
+      "epoch": 0.10787714696674845,
+      "grad_norm": 0.2744576334953308,
+      "learning_rate": 0.001,
+      "loss": 1.9552,
+      "step": 2550
+    },
+    {
+      "epoch": 0.10791945173026483,
+      "grad_norm": 0.32292699813842773,
+      "learning_rate": 0.001,
+      "loss": 2.3587,
+      "step": 2551
+    },
+    {
+      "epoch": 0.1079617564937812,
+      "grad_norm": 0.4315103590488434,
+      "learning_rate": 0.001,
+      "loss": 3.1126,
+      "step": 2552
+    },
+    {
+      "epoch": 0.10800406125729757,
+      "grad_norm": 0.38236674666404724,
+      "learning_rate": 0.001,
+      "loss": 2.787,
+      "step": 2553
+    },
+    {
+      "epoch": 0.10804636602081394,
+      "grad_norm": 0.25783267617225647,
+      "learning_rate": 0.001,
+      "loss": 1.9616,
+      "step": 2554
+    },
+    {
+      "epoch": 0.10808867078433032,
+      "grad_norm": 0.6072126030921936,
+      "learning_rate": 0.001,
+      "loss": 2.2998,
+      "step": 2555
+    },
+    {
+      "epoch": 0.10813097554784669,
+      "grad_norm": 1.3105621337890625,
+      "learning_rate": 0.001,
+      "loss": 2.144,
+      "step": 2556
+    },
+    {
+      "epoch": 0.10817328031136306,
+      "grad_norm": 0.3103470504283905,
+      "learning_rate": 0.001,
+      "loss": 2.4171,
+      "step": 2557
+    },
+    {
+      "epoch": 0.10821558507487943,
+      "grad_norm": 2.017469644546509,
+      "learning_rate": 0.001,
+      "loss": 1.9626,
+      "step": 2558
+    },
+    {
+      "epoch": 0.10825788983839581,
+      "grad_norm": 0.6327214241027832,
+      "learning_rate": 0.001,
+      "loss": 3.576,
+      "step": 2559
+    },
+    {
+      "epoch": 0.10830019460191218,
+      "grad_norm": 0.3792698383331299,
+      "learning_rate": 0.001,
+      "loss": 2.7368,
+      "step": 2560
+    },
+    {
+      "epoch": 0.10834249936542854,
+      "grad_norm": 0.3013581931591034,
+      "learning_rate": 0.001,
+      "loss": 2.3676,
+      "step": 2561
+    },
+    {
+      "epoch": 0.10838480412894493,
+      "grad_norm": 0.2713608145713806,
+      "learning_rate": 0.001,
+      "loss": 2.6759,
+      "step": 2562
+    },
+    {
+      "epoch": 0.1084271088924613,
+      "grad_norm": 0.32632705569267273,
+      "learning_rate": 0.001,
+      "loss": 3.0297,
+      "step": 2563
+    },
+    {
+      "epoch": 0.10846941365597766,
+      "grad_norm": 0.48675236105918884,
+      "learning_rate": 0.001,
+      "loss": 2.3548,
+      "step": 2564
+    },
+    {
+      "epoch": 0.10851171841949403,
+      "grad_norm": 0.3283641040325165,
+      "learning_rate": 0.001,
+      "loss": 3.2577,
+      "step": 2565
+    },
+    {
+      "epoch": 0.10855402318301041,
+      "grad_norm": 0.3809754550457001,
+      "learning_rate": 0.001,
+      "loss": 1.8483,
+      "step": 2566
+    },
+    {
+      "epoch": 0.10859632794652678,
+      "grad_norm": 0.5347718596458435,
+      "learning_rate": 0.001,
+      "loss": 2.3094,
+      "step": 2567
+    },
+    {
+      "epoch": 0.10863863271004315,
+      "grad_norm": 0.28041812777519226,
+      "learning_rate": 0.001,
+      "loss": 2.4324,
+      "step": 2568
+    },
+    {
+      "epoch": 0.10868093747355952,
+      "grad_norm": 0.33430221676826477,
+      "learning_rate": 0.001,
+      "loss": 2.1342,
+      "step": 2569
+    },
+    {
+      "epoch": 0.1087232422370759,
+      "grad_norm": 0.23147587478160858,
+      "learning_rate": 0.001,
+      "loss": 2.2938,
+      "step": 2570
+    },
+    {
+      "epoch": 0.10876554700059227,
+      "grad_norm": 0.25908035039901733,
+      "learning_rate": 0.001,
+      "loss": 3.265,
+      "step": 2571
+    },
+    {
+      "epoch": 0.10880785176410863,
+      "grad_norm": 0.3302173316478729,
+      "learning_rate": 0.001,
+      "loss": 3.0818,
+      "step": 2572
+    },
+    {
+      "epoch": 0.10885015652762502,
+      "grad_norm": 0.3298220634460449,
+      "learning_rate": 0.001,
+      "loss": 2.6535,
+      "step": 2573
+    },
+    {
+      "epoch": 0.10889246129114138,
+      "grad_norm": 0.9503449201583862,
+      "learning_rate": 0.001,
+      "loss": 1.846,
+      "step": 2574
+    },
+    {
+      "epoch": 0.10893476605465775,
+      "grad_norm": 0.4278430938720703,
+      "learning_rate": 0.001,
+      "loss": 2.8229,
+      "step": 2575
+    },
+    {
+      "epoch": 0.10897707081817412,
+      "grad_norm": 0.35355350375175476,
+      "learning_rate": 0.001,
+      "loss": 2.7936,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1090193755816905,
+      "grad_norm": 2.9743497371673584,
+      "learning_rate": 0.001,
+      "loss": 1.9395,
+      "step": 2577
+    },
+    {
+      "epoch": 0.10906168034520687,
+      "grad_norm": 1.5361433029174805,
+      "learning_rate": 0.001,
+      "loss": 2.2915,
+      "step": 2578
+    },
+    {
+      "epoch": 0.10910398510872324,
+      "grad_norm": 0.35586532950401306,
+      "learning_rate": 0.001,
+      "loss": 2.2418,
+      "step": 2579
+    },
+    {
+      "epoch": 0.10914628987223961,
+      "grad_norm": 0.7446362972259521,
+      "learning_rate": 0.001,
+      "loss": 2.3212,
+      "step": 2580
+    },
+    {
+      "epoch": 0.10918859463575599,
+      "grad_norm": 0.3713662922382355,
+      "learning_rate": 0.001,
+      "loss": 2.0863,
+      "step": 2581
+    },
+    {
+      "epoch": 0.10923089939927236,
+      "grad_norm": 0.4044400155544281,
+      "learning_rate": 0.001,
+      "loss": 2.1731,
+      "step": 2582
+    },
+    {
+      "epoch": 0.10927320416278873,
+      "grad_norm": 2.1356008052825928,
+      "learning_rate": 0.001,
+      "loss": 2.3904,
+      "step": 2583
+    },
+    {
+      "epoch": 0.10931550892630511,
+      "grad_norm": 0.9785255193710327,
+      "learning_rate": 0.001,
+      "loss": 1.529,
+      "step": 2584
+    },
+    {
+      "epoch": 0.10935781368982148,
+      "grad_norm": 1.0335931777954102,
+      "learning_rate": 0.001,
+      "loss": 3.1518,
+      "step": 2585
+    },
+    {
+      "epoch": 0.10940011845333784,
+      "grad_norm": 0.8856411576271057,
+      "learning_rate": 0.001,
+      "loss": 2.0349,
+      "step": 2586
+    },
+    {
+      "epoch": 0.10944242321685421,
+      "grad_norm": 0.8047211766242981,
+      "learning_rate": 0.001,
+      "loss": 2.1286,
+      "step": 2587
+    },
+    {
+      "epoch": 0.1094847279803706,
+      "grad_norm": 0.5640533566474915,
+      "learning_rate": 0.001,
+      "loss": 2.8253,
+      "step": 2588
+    },
+    {
+      "epoch": 0.10952703274388696,
+      "grad_norm": 0.3256753087043762,
+      "learning_rate": 0.001,
+      "loss": 2.0331,
+      "step": 2589
+    },
+    {
+      "epoch": 0.10956933750740333,
+      "grad_norm": 1.4646278619766235,
+      "learning_rate": 0.001,
+      "loss": 2.928,
+      "step": 2590
+    },
+    {
+      "epoch": 0.10961164227091971,
+      "grad_norm": 5.041200637817383,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 2591
+    },
+    {
+      "epoch": 0.10965394703443608,
+      "grad_norm": 0.8993614912033081,
+      "learning_rate": 0.001,
+      "loss": 2.4207,
+      "step": 2592
+    },
+    {
+      "epoch": 0.10969625179795245,
+      "grad_norm": 0.47238776087760925,
+      "learning_rate": 0.001,
+      "loss": 2.7745,
+      "step": 2593
+    },
+    {
+      "epoch": 0.10973855656146882,
+      "grad_norm": 0.5236493945121765,
+      "learning_rate": 0.001,
+      "loss": 2.6803,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1097808613249852,
+      "grad_norm": 0.49597102403640747,
+      "learning_rate": 0.001,
+      "loss": 2.4384,
+      "step": 2595
+    },
+    {
+      "epoch": 0.10982316608850157,
+      "grad_norm": 0.7567354440689087,
+      "learning_rate": 0.001,
+      "loss": 2.0341,
+      "step": 2596
+    },
+    {
+      "epoch": 0.10986547085201794,
+      "grad_norm": 43.87346649169922,
+      "learning_rate": 0.001,
+      "loss": 2.3568,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1099077756155343,
+      "grad_norm": 0.926131546497345,
+      "learning_rate": 0.001,
+      "loss": 3.707,
+      "step": 2598
+    },
+    {
+      "epoch": 0.10995008037905069,
+      "grad_norm": 0.2967435419559479,
+      "learning_rate": 0.001,
+      "loss": 2.0035,
+      "step": 2599
+    },
+    {
+      "epoch": 0.10999238514256705,
+      "grad_norm": 0.4026656150817871,
+      "learning_rate": 0.001,
+      "loss": 2.9419,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11003468990608342,
+      "grad_norm": 1.9945727586746216,
+      "learning_rate": 0.001,
+      "loss": 2.478,
+      "step": 2601
+    },
+    {
+      "epoch": 0.1100769946695998,
+      "grad_norm": 8.51455307006836,
+      "learning_rate": 0.001,
+      "loss": 2.4987,
+      "step": 2602
+    },
+    {
+      "epoch": 0.11011929943311617,
+      "grad_norm": 7.822841167449951,
+      "learning_rate": 0.001,
+      "loss": 2.9194,
+      "step": 2603
+    },
+    {
+      "epoch": 0.11016160419663254,
+      "grad_norm": 0.6975441575050354,
+      "learning_rate": 0.001,
+      "loss": 4.2243,
+      "step": 2604
+    },
+    {
+      "epoch": 0.11020390896014891,
+      "grad_norm": 9.386311531066895,
+      "learning_rate": 0.001,
+      "loss": 2.6628,
+      "step": 2605
+    },
+    {
+      "epoch": 0.11024621372366529,
+      "grad_norm": 0.5597333908081055,
+      "learning_rate": 0.001,
+      "loss": 3.464,
+      "step": 2606
+    },
+    {
+      "epoch": 0.11028851848718166,
+      "grad_norm": 3.522331714630127,
+      "learning_rate": 0.001,
+      "loss": 2.7945,
+      "step": 2607
+    },
+    {
+      "epoch": 0.11033082325069803,
+      "grad_norm": 0.4466964602470398,
+      "learning_rate": 0.001,
+      "loss": 3.2322,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1103731280142144,
+      "grad_norm": 12.962498664855957,
+      "learning_rate": 0.001,
+      "loss": 2.4416,
+      "step": 2609
+    },
+    {
+      "epoch": 0.11041543277773078,
+      "grad_norm": 0.35099491477012634,
+      "learning_rate": 0.001,
+      "loss": 3.0072,
+      "step": 2610
+    },
+    {
+      "epoch": 0.11045773754124714,
+      "grad_norm": 0.5855743885040283,
+      "learning_rate": 0.001,
+      "loss": 3.3016,
+      "step": 2611
+    },
+    {
+      "epoch": 0.11050004230476351,
+      "grad_norm": 1.2006386518478394,
+      "learning_rate": 0.001,
+      "loss": 4.1105,
+      "step": 2612
+    },
+    {
+      "epoch": 0.1105423470682799,
+      "grad_norm": 0.3867046535015106,
+      "learning_rate": 0.001,
+      "loss": 2.4014,
+      "step": 2613
+    },
+    {
+      "epoch": 0.11058465183179626,
+      "grad_norm": 0.6155429482460022,
+      "learning_rate": 0.001,
+      "loss": 3.2606,
+      "step": 2614
+    },
+    {
+      "epoch": 0.11062695659531263,
+      "grad_norm": 1.17522394657135,
+      "learning_rate": 0.001,
+      "loss": 2.7431,
+      "step": 2615
+    },
+    {
+      "epoch": 0.110669261358829,
+      "grad_norm": 1.3711291551589966,
+      "learning_rate": 0.001,
+      "loss": 2.6056,
+      "step": 2616
+    },
+    {
+      "epoch": 0.11071156612234538,
+      "grad_norm": 0.33851131796836853,
+      "learning_rate": 0.001,
+      "loss": 2.9142,
+      "step": 2617
+    },
+    {
+      "epoch": 0.11075387088586175,
+      "grad_norm": 0.32981690764427185,
+      "learning_rate": 0.001,
+      "loss": 2.4273,
+      "step": 2618
+    },
+    {
+      "epoch": 0.11079617564937812,
+      "grad_norm": 0.3835381269454956,
+      "learning_rate": 0.001,
+      "loss": 3.0215,
+      "step": 2619
+    },
+    {
+      "epoch": 0.11083848041289449,
+      "grad_norm": 0.2677971422672272,
+      "learning_rate": 0.001,
+      "loss": 2.0736,
+      "step": 2620
+    },
+    {
+      "epoch": 0.11088078517641087,
+      "grad_norm": 0.33958640694618225,
+      "learning_rate": 0.001,
+      "loss": 3.4959,
+      "step": 2621
+    },
+    {
+      "epoch": 0.11092308993992724,
+      "grad_norm": 0.281795859336853,
+      "learning_rate": 0.001,
+      "loss": 2.2221,
+      "step": 2622
+    },
+    {
+      "epoch": 0.1109653947034436,
+      "grad_norm": 0.3385681211948395,
+      "learning_rate": 0.001,
+      "loss": 1.5377,
+      "step": 2623
+    },
+    {
+      "epoch": 0.11100769946695999,
+      "grad_norm": 0.466964453458786,
+      "learning_rate": 0.001,
+      "loss": 2.3652,
+      "step": 2624
+    },
+    {
+      "epoch": 0.11105000423047635,
+      "grad_norm": 0.25589802861213684,
+      "learning_rate": 0.001,
+      "loss": 1.903,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11109230899399272,
+      "grad_norm": 0.3010057508945465,
+      "learning_rate": 0.001,
+      "loss": 3.1568,
+      "step": 2626
+    },
+    {
+      "epoch": 0.11113461375750909,
+      "grad_norm": 6.012112140655518,
+      "learning_rate": 0.001,
+      "loss": 2.411,
+      "step": 2627
+    },
+    {
+      "epoch": 0.11117691852102547,
+      "grad_norm": 0.4896494448184967,
+      "learning_rate": 0.001,
+      "loss": 3.1489,
+      "step": 2628
+    },
+    {
+      "epoch": 0.11121922328454184,
+      "grad_norm": 0.33814510703086853,
+      "learning_rate": 0.001,
+      "loss": 2.1617,
+      "step": 2629
+    },
+    {
+      "epoch": 0.11126152804805821,
+      "grad_norm": 0.23706887662410736,
+      "learning_rate": 0.001,
+      "loss": 2.3078,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11130383281157458,
+      "grad_norm": 0.280916690826416,
+      "learning_rate": 0.001,
+      "loss": 2.4966,
+      "step": 2631
+    },
+    {
+      "epoch": 0.11134613757509096,
+      "grad_norm": 1.4228007793426514,
+      "learning_rate": 0.001,
+      "loss": 1.9013,
+      "step": 2632
+    },
+    {
+      "epoch": 0.11138844233860733,
+      "grad_norm": 0.3774404525756836,
+      "learning_rate": 0.001,
+      "loss": 2.2246,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1114307471021237,
+      "grad_norm": 13.576204299926758,
+      "learning_rate": 0.001,
+      "loss": 2.8407,
+      "step": 2634
+    },
+    {
+      "epoch": 0.11147305186564008,
+      "grad_norm": 1.4729946851730347,
+      "learning_rate": 0.001,
+      "loss": 2.5873,
+      "step": 2635
+    },
+    {
+      "epoch": 0.11151535662915645,
+      "grad_norm": 0.2995510697364807,
+      "learning_rate": 0.001,
+      "loss": 2.0534,
+      "step": 2636
+    },
+    {
+      "epoch": 0.11155766139267281,
+      "grad_norm": 0.389316201210022,
+      "learning_rate": 0.001,
+      "loss": 3.0107,
+      "step": 2637
+    },
+    {
+      "epoch": 0.11159996615618918,
+      "grad_norm": 0.375186562538147,
+      "learning_rate": 0.001,
+      "loss": 3.2694,
+      "step": 2638
+    },
+    {
+      "epoch": 0.11164227091970556,
+      "grad_norm": 0.24365709722042084,
+      "learning_rate": 0.001,
+      "loss": 2.6694,
+      "step": 2639
+    },
+    {
+      "epoch": 0.11168457568322193,
+      "grad_norm": 0.35557329654693604,
+      "learning_rate": 0.001,
+      "loss": 2.1832,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1117268804467383,
+      "grad_norm": 0.2812917232513428,
+      "learning_rate": 0.001,
+      "loss": 2.6873,
+      "step": 2641
+    },
+    {
+      "epoch": 0.11176918521025467,
+      "grad_norm": 0.3376398980617523,
+      "learning_rate": 0.001,
+      "loss": 2.1201,
+      "step": 2642
+    },
+    {
+      "epoch": 0.11181148997377105,
+      "grad_norm": 1.3907296657562256,
+      "learning_rate": 0.001,
+      "loss": 2.5576,
+      "step": 2643
+    },
+    {
+      "epoch": 0.11185379473728742,
+      "grad_norm": 3.7346065044403076,
+      "learning_rate": 0.001,
+      "loss": 2.812,
+      "step": 2644
+    },
+    {
+      "epoch": 0.11189609950080379,
+      "grad_norm": 15.240631103515625,
+      "learning_rate": 0.001,
+      "loss": 2.7597,
+      "step": 2645
+    },
+    {
+      "epoch": 0.11193840426432017,
+      "grad_norm": 0.24908733367919922,
+      "learning_rate": 0.001,
+      "loss": 2.6942,
+      "step": 2646
+    },
+    {
+      "epoch": 0.11198070902783654,
+      "grad_norm": 0.43838199973106384,
+      "learning_rate": 0.001,
+      "loss": 2.216,
+      "step": 2647
+    },
+    {
+      "epoch": 0.1120230137913529,
+      "grad_norm": 1.0071830749511719,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 2648
+    },
+    {
+      "epoch": 0.11206531855486927,
+      "grad_norm": 1.4726154804229736,
+      "learning_rate": 0.001,
+      "loss": 2.3125,
+      "step": 2649
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 0.4552990198135376,
+      "learning_rate": 0.001,
+      "loss": 2.402,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11214992808190202,
+      "grad_norm": 0.4046100378036499,
+      "learning_rate": 0.001,
+      "loss": 2.8129,
+      "step": 2651
+    },
+    {
+      "epoch": 0.11219223284541839,
+      "grad_norm": 0.3966389000415802,
+      "learning_rate": 0.001,
+      "loss": 3.0564,
+      "step": 2652
+    },
+    {
+      "epoch": 0.11223453760893477,
+      "grad_norm": 0.25914904475212097,
+      "learning_rate": 0.001,
+      "loss": 2.0611,
+      "step": 2653
+    },
+    {
+      "epoch": 0.11227684237245114,
+      "grad_norm": 0.2761140763759613,
+      "learning_rate": 0.001,
+      "loss": 2.1811,
+      "step": 2654
+    },
+    {
+      "epoch": 0.11231914713596751,
+      "grad_norm": 0.3019154965877533,
+      "learning_rate": 0.001,
+      "loss": 2.5075,
+      "step": 2655
+    },
+    {
+      "epoch": 0.11236145189948388,
+      "grad_norm": 0.22981767356395721,
+      "learning_rate": 0.001,
+      "loss": 1.7464,
+      "step": 2656
+    },
+    {
+      "epoch": 0.11240375666300026,
+      "grad_norm": 0.32616347074508667,
+      "learning_rate": 0.001,
+      "loss": 3.1906,
+      "step": 2657
+    },
+    {
+      "epoch": 0.11244606142651663,
+      "grad_norm": 0.2503935992717743,
+      "learning_rate": 0.001,
+      "loss": 2.4619,
+      "step": 2658
+    },
+    {
+      "epoch": 0.112488366190033,
+      "grad_norm": 0.27525201439857483,
+      "learning_rate": 0.001,
+      "loss": 2.0142,
+      "step": 2659
+    },
+    {
+      "epoch": 0.11253067095354936,
+      "grad_norm": 0.23904815316200256,
+      "learning_rate": 0.001,
+      "loss": 1.9905,
+      "step": 2660
+    },
+    {
+      "epoch": 0.11257297571706575,
+      "grad_norm": 1.022966980934143,
+      "learning_rate": 0.001,
+      "loss": 2.2188,
+      "step": 2661
+    },
+    {
+      "epoch": 0.11261528048058211,
+      "grad_norm": 0.39735960960388184,
+      "learning_rate": 0.001,
+      "loss": 2.3742,
+      "step": 2662
+    },
+    {
+      "epoch": 0.11265758524409848,
+      "grad_norm": 0.22529076039791107,
+      "learning_rate": 0.001,
+      "loss": 2.3343,
+      "step": 2663
+    },
+    {
+      "epoch": 0.11269989000761486,
+      "grad_norm": 0.6422929167747498,
+      "learning_rate": 0.001,
+      "loss": 2.8202,
+      "step": 2664
+    },
+    {
+      "epoch": 0.11274219477113123,
+      "grad_norm": 0.32373046875,
+      "learning_rate": 0.001,
+      "loss": 3.1001,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1127844995346476,
+      "grad_norm": 0.4991035759449005,
+      "learning_rate": 0.001,
+      "loss": 1.8736,
+      "step": 2666
+    },
+    {
+      "epoch": 0.11282680429816397,
+      "grad_norm": 0.2527635097503662,
+      "learning_rate": 0.001,
+      "loss": 2.5017,
+      "step": 2667
+    },
+    {
+      "epoch": 0.11286910906168035,
+      "grad_norm": 7.938605785369873,
+      "learning_rate": 0.001,
+      "loss": 1.5331,
+      "step": 2668
+    },
+    {
+      "epoch": 0.11291141382519672,
+      "grad_norm": 0.4624551832675934,
+      "learning_rate": 0.001,
+      "loss": 3.5508,
+      "step": 2669
+    },
+    {
+      "epoch": 0.11295371858871309,
+      "grad_norm": 0.28487929701805115,
+      "learning_rate": 0.001,
+      "loss": 2.2229,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11299602335222945,
+      "grad_norm": 1.6842297315597534,
+      "learning_rate": 0.001,
+      "loss": 2.3192,
+      "step": 2671
+    },
+    {
+      "epoch": 0.11303832811574584,
+      "grad_norm": 1.5760769844055176,
+      "learning_rate": 0.001,
+      "loss": 2.3325,
+      "step": 2672
+    },
+    {
+      "epoch": 0.1130806328792622,
+      "grad_norm": 1.3597266674041748,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 2673
+    },
+    {
+      "epoch": 0.11312293764277857,
+      "grad_norm": 0.3656711280345917,
+      "learning_rate": 0.001,
+      "loss": 3.6342,
+      "step": 2674
+    },
+    {
+      "epoch": 0.11316524240629496,
+      "grad_norm": 0.4814014136791229,
+      "learning_rate": 0.001,
+      "loss": 2.4345,
+      "step": 2675
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.24970856308937073,
+      "learning_rate": 0.001,
+      "loss": 1.9132,
+      "step": 2676
+    },
+    {
+      "epoch": 0.11324985193332769,
+      "grad_norm": 0.31391432881355286,
+      "learning_rate": 0.001,
+      "loss": 2.6646,
+      "step": 2677
+    },
+    {
+      "epoch": 0.11329215669684406,
+      "grad_norm": 0.4391164481639862,
+      "learning_rate": 0.001,
+      "loss": 2.9287,
+      "step": 2678
+    },
+    {
+      "epoch": 0.11333446146036044,
+      "grad_norm": 0.3026899993419647,
+      "learning_rate": 0.001,
+      "loss": 2.882,
+      "step": 2679
+    },
+    {
+      "epoch": 0.11337676622387681,
+      "grad_norm": 0.755450963973999,
+      "learning_rate": 0.001,
+      "loss": 3.3115,
+      "step": 2680
+    },
+    {
+      "epoch": 0.11341907098739318,
+      "grad_norm": 0.3200805187225342,
+      "learning_rate": 0.001,
+      "loss": 2.3836,
+      "step": 2681
+    },
+    {
+      "epoch": 0.11346137575090955,
+      "grad_norm": 0.25980228185653687,
+      "learning_rate": 0.001,
+      "loss": 1.8265,
+      "step": 2682
+    },
+    {
+      "epoch": 0.11350368051442593,
+      "grad_norm": 0.4624451696872711,
+      "learning_rate": 0.001,
+      "loss": 2.7016,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1135459852779423,
+      "grad_norm": 0.7080914378166199,
+      "learning_rate": 0.001,
+      "loss": 2.4309,
+      "step": 2684
+    },
+    {
+      "epoch": 0.11358829004145866,
+      "grad_norm": 0.45167163014411926,
+      "learning_rate": 0.001,
+      "loss": 3.1995,
+      "step": 2685
+    },
+    {
+      "epoch": 0.11363059480497505,
+      "grad_norm": 0.21926790475845337,
+      "learning_rate": 0.001,
+      "loss": 2.0448,
+      "step": 2686
+    },
+    {
+      "epoch": 0.11367289956849141,
+      "grad_norm": 1.0838067531585693,
+      "learning_rate": 0.001,
+      "loss": 1.8552,
+      "step": 2687
+    },
+    {
+      "epoch": 0.11371520433200778,
+      "grad_norm": 1.3477392196655273,
+      "learning_rate": 0.001,
+      "loss": 2.7008,
+      "step": 2688
+    },
+    {
+      "epoch": 0.11375750909552415,
+      "grad_norm": 0.41002964973449707,
+      "learning_rate": 0.001,
+      "loss": 3.4506,
+      "step": 2689
+    },
+    {
+      "epoch": 0.11379981385904053,
+      "grad_norm": 0.2277841717004776,
+      "learning_rate": 0.001,
+      "loss": 2.1868,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1138421186225569,
+      "grad_norm": 0.6149588227272034,
+      "learning_rate": 0.001,
+      "loss": 2.6013,
+      "step": 2691
+    },
+    {
+      "epoch": 0.11388442338607327,
+      "grad_norm": 0.36279934644699097,
+      "learning_rate": 0.001,
+      "loss": 2.4312,
+      "step": 2692
+    },
+    {
+      "epoch": 0.11392672814958964,
+      "grad_norm": 0.47520431876182556,
+      "learning_rate": 0.001,
+      "loss": 3.0366,
+      "step": 2693
+    },
+    {
+      "epoch": 0.11396903291310602,
+      "grad_norm": 0.31242018938064575,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 2694
+    },
+    {
+      "epoch": 0.11401133767662239,
+      "grad_norm": 1.7739008665084839,
+      "learning_rate": 0.001,
+      "loss": 1.895,
+      "step": 2695
+    },
+    {
+      "epoch": 0.11405364244013876,
+      "grad_norm": 0.5066441297531128,
+      "learning_rate": 0.001,
+      "loss": 2.6342,
+      "step": 2696
+    },
+    {
+      "epoch": 0.11409594720365514,
+      "grad_norm": 0.24368147552013397,
+      "learning_rate": 0.001,
+      "loss": 2.7584,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1141382519671715,
+      "grad_norm": 0.44066333770751953,
+      "learning_rate": 0.001,
+      "loss": 2.3854,
+      "step": 2698
+    },
+    {
+      "epoch": 0.11418055673068787,
+      "grad_norm": 0.26902955770492554,
+      "learning_rate": 0.001,
+      "loss": 2.7149,
+      "step": 2699
+    },
+    {
+      "epoch": 0.11422286149420424,
+      "grad_norm": 4.966346263885498,
+      "learning_rate": 0.001,
+      "loss": 2.2739,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11426516625772062,
+      "grad_norm": 0.82923823595047,
+      "learning_rate": 0.001,
+      "loss": 2.4333,
+      "step": 2701
+    },
+    {
+      "epoch": 0.11430747102123699,
+      "grad_norm": 0.37718573212623596,
+      "learning_rate": 0.001,
+      "loss": 2.846,
+      "step": 2702
+    },
+    {
+      "epoch": 0.11434977578475336,
+      "grad_norm": 1.1402643918991089,
+      "learning_rate": 0.001,
+      "loss": 2.2025,
+      "step": 2703
+    },
+    {
+      "epoch": 0.11439208054826973,
+      "grad_norm": 0.5933586359024048,
+      "learning_rate": 0.001,
+      "loss": 4.4132,
+      "step": 2704
+    },
+    {
+      "epoch": 0.11443438531178611,
+      "grad_norm": 0.9966050386428833,
+      "learning_rate": 0.001,
+      "loss": 3.1715,
+      "step": 2705
+    },
+    {
+      "epoch": 0.11447669007530248,
+      "grad_norm": 0.3818608820438385,
+      "learning_rate": 0.001,
+      "loss": 2.9614,
+      "step": 2706
+    },
+    {
+      "epoch": 0.11451899483881885,
+      "grad_norm": 0.355037122964859,
+      "learning_rate": 0.001,
+      "loss": 2.7848,
+      "step": 2707
+    },
+    {
+      "epoch": 0.11456129960233523,
+      "grad_norm": 0.4167128801345825,
+      "learning_rate": 0.001,
+      "loss": 2.1713,
+      "step": 2708
+    },
+    {
+      "epoch": 0.1146036043658516,
+      "grad_norm": 0.3600790202617645,
+      "learning_rate": 0.001,
+      "loss": 2.0852,
+      "step": 2709
+    },
+    {
+      "epoch": 0.11464590912936796,
+      "grad_norm": 0.7185443639755249,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11468821389288433,
+      "grad_norm": 0.5277695655822754,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 2711
+    },
+    {
+      "epoch": 0.11473051865640072,
+      "grad_norm": 2.362328052520752,
+      "learning_rate": 0.001,
+      "loss": 2.655,
+      "step": 2712
+    },
+    {
+      "epoch": 0.11477282341991708,
+      "grad_norm": 1.576801061630249,
+      "learning_rate": 0.001,
+      "loss": 3.0455,
+      "step": 2713
+    },
+    {
+      "epoch": 0.11481512818343345,
+      "grad_norm": 0.5034794807434082,
+      "learning_rate": 0.001,
+      "loss": 1.9612,
+      "step": 2714
+    },
+    {
+      "epoch": 0.11485743294694983,
+      "grad_norm": 0.5581480860710144,
+      "learning_rate": 0.001,
+      "loss": 2.1294,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1148997377104662,
+      "grad_norm": 1.2885347604751587,
+      "learning_rate": 0.001,
+      "loss": 3.0082,
+      "step": 2716
+    },
+    {
+      "epoch": 0.11494204247398257,
+      "grad_norm": 1.621717929840088,
+      "learning_rate": 0.001,
+      "loss": 2.4047,
+      "step": 2717
+    },
+    {
+      "epoch": 0.11498434723749894,
+      "grad_norm": 22.007076263427734,
+      "learning_rate": 0.001,
+      "loss": 3.3206,
+      "step": 2718
+    },
+    {
+      "epoch": 0.11502665200101532,
+      "grad_norm": 3.476715087890625,
+      "learning_rate": 0.001,
+      "loss": 2.8062,
+      "step": 2719
+    },
+    {
+      "epoch": 0.11506895676453169,
+      "grad_norm": 0.24378205835819244,
+      "learning_rate": 0.001,
+      "loss": 1.6405,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11511126152804806,
+      "grad_norm": 3.1240763664245605,
+      "learning_rate": 0.001,
+      "loss": 2.9476,
+      "step": 2721
+    },
+    {
+      "epoch": 0.11515356629156442,
+      "grad_norm": 0.27668654918670654,
+      "learning_rate": 0.001,
+      "loss": 1.6878,
+      "step": 2722
+    },
+    {
+      "epoch": 0.1151958710550808,
+      "grad_norm": 33.68696975708008,
+      "learning_rate": 0.001,
+      "loss": 2.2438,
+      "step": 2723
+    },
+    {
+      "epoch": 0.11523817581859717,
+      "grad_norm": 1.328370213508606,
+      "learning_rate": 0.001,
+      "loss": 2.6715,
+      "step": 2724
+    },
+    {
+      "epoch": 0.11528048058211354,
+      "grad_norm": 22.9268741607666,
+      "learning_rate": 0.001,
+      "loss": 3.4431,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11532278534562992,
+      "grad_norm": 0.6958709955215454,
+      "learning_rate": 0.001,
+      "loss": 2.9032,
+      "step": 2726
+    },
+    {
+      "epoch": 0.11536509010914629,
+      "grad_norm": 0.41907453536987305,
+      "learning_rate": 0.001,
+      "loss": 2.925,
+      "step": 2727
+    },
+    {
+      "epoch": 0.11540739487266266,
+      "grad_norm": 0.2991430461406708,
+      "learning_rate": 0.001,
+      "loss": 2.5263,
+      "step": 2728
+    },
+    {
+      "epoch": 0.11544969963617903,
+      "grad_norm": 0.7529126405715942,
+      "learning_rate": 0.001,
+      "loss": 2.8932,
+      "step": 2729
+    },
+    {
+      "epoch": 0.11549200439969541,
+      "grad_norm": 0.8318426609039307,
+      "learning_rate": 0.001,
+      "loss": 2.7539,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11553430916321178,
+      "grad_norm": 1.7126951217651367,
+      "learning_rate": 0.001,
+      "loss": 2.2141,
+      "step": 2731
+    },
+    {
+      "epoch": 0.11557661392672815,
+      "grad_norm": 0.28930333256721497,
+      "learning_rate": 0.001,
+      "loss": 2.4101,
+      "step": 2732
+    },
+    {
+      "epoch": 0.11561891869024452,
+      "grad_norm": 0.3181939423084259,
+      "learning_rate": 0.001,
+      "loss": 2.2034,
+      "step": 2733
+    },
+    {
+      "epoch": 0.1156612234537609,
+      "grad_norm": 0.34087949991226196,
+      "learning_rate": 0.001,
+      "loss": 2.6263,
+      "step": 2734
+    },
+    {
+      "epoch": 0.11570352821727727,
+      "grad_norm": 1.4377021789550781,
+      "learning_rate": 0.001,
+      "loss": 2.2172,
+      "step": 2735
+    },
+    {
+      "epoch": 0.11574583298079363,
+      "grad_norm": 0.3622700572013855,
+      "learning_rate": 0.001,
+      "loss": 2.5439,
+      "step": 2736
+    },
+    {
+      "epoch": 0.11578813774431002,
+      "grad_norm": 0.27332210540771484,
+      "learning_rate": 0.001,
+      "loss": 2.9426,
+      "step": 2737
+    },
+    {
+      "epoch": 0.11583044250782638,
+      "grad_norm": 32.977970123291016,
+      "learning_rate": 0.001,
+      "loss": 2.6618,
+      "step": 2738
+    },
+    {
+      "epoch": 0.11587274727134275,
+      "grad_norm": 1.1440379619598389,
+      "learning_rate": 0.001,
+      "loss": 2.1755,
+      "step": 2739
+    },
+    {
+      "epoch": 0.11591505203485912,
+      "grad_norm": 0.3366926610469818,
+      "learning_rate": 0.001,
+      "loss": 2.1376,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1159573567983755,
+      "grad_norm": 0.3079652786254883,
+      "learning_rate": 0.001,
+      "loss": 2.873,
+      "step": 2741
+    },
+    {
+      "epoch": 0.11599966156189187,
+      "grad_norm": 0.3998541533946991,
+      "learning_rate": 0.001,
+      "loss": 2.7169,
+      "step": 2742
+    },
+    {
+      "epoch": 0.11604196632540824,
+      "grad_norm": 0.33957117795944214,
+      "learning_rate": 0.001,
+      "loss": 2.496,
+      "step": 2743
+    },
+    {
+      "epoch": 0.1160842710889246,
+      "grad_norm": 0.9444336891174316,
+      "learning_rate": 0.001,
+      "loss": 2.6322,
+      "step": 2744
+    },
+    {
+      "epoch": 0.11612657585244099,
+      "grad_norm": 0.8139225840568542,
+      "learning_rate": 0.001,
+      "loss": 2.0574,
+      "step": 2745
+    },
+    {
+      "epoch": 0.11616888061595736,
+      "grad_norm": 0.2738553583621979,
+      "learning_rate": 0.001,
+      "loss": 3.0332,
+      "step": 2746
+    },
+    {
+      "epoch": 0.11621118537947372,
+      "grad_norm": 0.539919912815094,
+      "learning_rate": 0.001,
+      "loss": 2.1122,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1162534901429901,
+      "grad_norm": 0.29261156916618347,
+      "learning_rate": 0.001,
+      "loss": 2.6415,
+      "step": 2748
+    },
+    {
+      "epoch": 0.11629579490650647,
+      "grad_norm": 0.36019593477249146,
+      "learning_rate": 0.001,
+      "loss": 2.0997,
+      "step": 2749
+    },
+    {
+      "epoch": 0.11633809967002284,
+      "grad_norm": 1.0619240999221802,
+      "learning_rate": 0.001,
+      "loss": 2.7637,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11638040443353921,
+      "grad_norm": 0.32508841156959534,
+      "learning_rate": 0.001,
+      "loss": 3.1838,
+      "step": 2751
+    },
+    {
+      "epoch": 0.1164227091970556,
+      "grad_norm": 0.37493109703063965,
+      "learning_rate": 0.001,
+      "loss": 2.8842,
+      "step": 2752
+    },
+    {
+      "epoch": 0.11646501396057196,
+      "grad_norm": 0.2882426381111145,
+      "learning_rate": 0.001,
+      "loss": 2.1448,
+      "step": 2753
+    },
+    {
+      "epoch": 0.11650731872408833,
+      "grad_norm": 0.30877891182899475,
+      "learning_rate": 0.001,
+      "loss": 2.4269,
+      "step": 2754
+    },
+    {
+      "epoch": 0.1165496234876047,
+      "grad_norm": 0.2244696319103241,
+      "learning_rate": 0.001,
+      "loss": 1.6318,
+      "step": 2755
+    },
+    {
+      "epoch": 0.11659192825112108,
+      "grad_norm": 0.2662592828273773,
+      "learning_rate": 0.001,
+      "loss": 2.2368,
+      "step": 2756
+    },
+    {
+      "epoch": 0.11663423301463745,
+      "grad_norm": 0.7968965172767639,
+      "learning_rate": 0.001,
+      "loss": 2.4511,
+      "step": 2757
+    },
+    {
+      "epoch": 0.11667653777815382,
+      "grad_norm": 0.3993340730667114,
+      "learning_rate": 0.001,
+      "loss": 2.2677,
+      "step": 2758
+    },
+    {
+      "epoch": 0.1167188425416702,
+      "grad_norm": 7.428627014160156,
+      "learning_rate": 0.001,
+      "loss": 2.8915,
+      "step": 2759
+    },
+    {
+      "epoch": 0.11676114730518657,
+      "grad_norm": 0.35218873620033264,
+      "learning_rate": 0.001,
+      "loss": 3.2628,
+      "step": 2760
+    },
+    {
+      "epoch": 0.11680345206870293,
+      "grad_norm": 0.33024418354034424,
+      "learning_rate": 0.001,
+      "loss": 2.3512,
+      "step": 2761
+    },
+    {
+      "epoch": 0.1168457568322193,
+      "grad_norm": 2.4879989624023438,
+      "learning_rate": 0.001,
+      "loss": 2.9775,
+      "step": 2762
+    },
+    {
+      "epoch": 0.11688806159573568,
+      "grad_norm": 0.2612648904323578,
+      "learning_rate": 0.001,
+      "loss": 2.8071,
+      "step": 2763
+    },
+    {
+      "epoch": 0.11693036635925205,
+      "grad_norm": 0.2603396475315094,
+      "learning_rate": 0.001,
+      "loss": 2.2835,
+      "step": 2764
+    },
+    {
+      "epoch": 0.11697267112276842,
+      "grad_norm": 1.3054839372634888,
+      "learning_rate": 0.001,
+      "loss": 2.3514,
+      "step": 2765
+    },
+    {
+      "epoch": 0.11701497588628479,
+      "grad_norm": 0.46820470690727234,
+      "learning_rate": 0.001,
+      "loss": 2.3354,
+      "step": 2766
+    },
+    {
+      "epoch": 0.11705728064980117,
+      "grad_norm": 0.2680201530456543,
+      "learning_rate": 0.001,
+      "loss": 2.6435,
+      "step": 2767
+    },
+    {
+      "epoch": 0.11709958541331754,
+      "grad_norm": 0.2581261694431305,
+      "learning_rate": 0.001,
+      "loss": 2.5269,
+      "step": 2768
+    },
+    {
+      "epoch": 0.11714189017683391,
+      "grad_norm": 0.2201681286096573,
+      "learning_rate": 0.001,
+      "loss": 1.9714,
+      "step": 2769
+    },
+    {
+      "epoch": 0.11718419494035029,
+      "grad_norm": 0.31604522466659546,
+      "learning_rate": 0.001,
+      "loss": 2.4044,
+      "step": 2770
+    },
+    {
+      "epoch": 0.11722649970386666,
+      "grad_norm": 0.23219691216945648,
+      "learning_rate": 0.001,
+      "loss": 2.4686,
+      "step": 2771
+    },
+    {
+      "epoch": 0.11726880446738303,
+      "grad_norm": 0.24609123170375824,
+      "learning_rate": 0.001,
+      "loss": 1.7137,
+      "step": 2772
+    },
+    {
+      "epoch": 0.1173111092308994,
+      "grad_norm": 0.3166695535182953,
+      "learning_rate": 0.001,
+      "loss": 2.44,
+      "step": 2773
+    },
+    {
+      "epoch": 0.11735341399441578,
+      "grad_norm": 0.6801291108131409,
+      "learning_rate": 0.001,
+      "loss": 2.8316,
+      "step": 2774
+    },
+    {
+      "epoch": 0.11739571875793214,
+      "grad_norm": 0.25939130783081055,
+      "learning_rate": 0.001,
+      "loss": 2.155,
+      "step": 2775
+    },
+    {
+      "epoch": 0.11743802352144851,
+      "grad_norm": 0.2404932826757431,
+      "learning_rate": 0.001,
+      "loss": 1.9066,
+      "step": 2776
+    },
+    {
+      "epoch": 0.1174803282849649,
+      "grad_norm": 0.32322371006011963,
+      "learning_rate": 0.001,
+      "loss": 3.1624,
+      "step": 2777
+    },
+    {
+      "epoch": 0.11752263304848126,
+      "grad_norm": 0.29162564873695374,
+      "learning_rate": 0.001,
+      "loss": 2.1677,
+      "step": 2778
+    },
+    {
+      "epoch": 0.11756493781199763,
+      "grad_norm": 0.4201529324054718,
+      "learning_rate": 0.001,
+      "loss": 3.65,
+      "step": 2779
+    },
+    {
+      "epoch": 0.117607242575514,
+      "grad_norm": 0.3757007420063019,
+      "learning_rate": 0.001,
+      "loss": 2.7946,
+      "step": 2780
+    },
+    {
+      "epoch": 0.11764954733903038,
+      "grad_norm": 0.22242939472198486,
+      "learning_rate": 0.001,
+      "loss": 2.1342,
+      "step": 2781
+    },
+    {
+      "epoch": 0.11769185210254675,
+      "grad_norm": 0.7142042517662048,
+      "learning_rate": 0.001,
+      "loss": 2.9385,
+      "step": 2782
+    },
+    {
+      "epoch": 0.11773415686606312,
+      "grad_norm": 0.4159119129180908,
+      "learning_rate": 0.001,
+      "loss": 2.2336,
+      "step": 2783
+    },
+    {
+      "epoch": 0.11777646162957948,
+      "grad_norm": 38.654056549072266,
+      "learning_rate": 0.001,
+      "loss": 3.0095,
+      "step": 2784
+    },
+    {
+      "epoch": 0.11781876639309587,
+      "grad_norm": 0.2876143157482147,
+      "learning_rate": 0.001,
+      "loss": 2.1954,
+      "step": 2785
+    },
+    {
+      "epoch": 0.11786107115661223,
+      "grad_norm": 2.3631954193115234,
+      "learning_rate": 0.001,
+      "loss": 3.1529,
+      "step": 2786
+    },
+    {
+      "epoch": 0.1179033759201286,
+      "grad_norm": 0.2558928430080414,
+      "learning_rate": 0.001,
+      "loss": 2.0726,
+      "step": 2787
+    },
+    {
+      "epoch": 0.11794568068364499,
+      "grad_norm": 2.708045721054077,
+      "learning_rate": 0.001,
+      "loss": 2.3912,
+      "step": 2788
+    },
+    {
+      "epoch": 0.11798798544716135,
+      "grad_norm": 0.44729793071746826,
+      "learning_rate": 0.001,
+      "loss": 1.9636,
+      "step": 2789
+    },
+    {
+      "epoch": 0.11803029021067772,
+      "grad_norm": 0.5945959091186523,
+      "learning_rate": 0.001,
+      "loss": 2.6161,
+      "step": 2790
+    },
+    {
+      "epoch": 0.11807259497419409,
+      "grad_norm": 0.35763904452323914,
+      "learning_rate": 0.001,
+      "loss": 3.1271,
+      "step": 2791
+    },
+    {
+      "epoch": 0.11811489973771047,
+      "grad_norm": 0.24619075655937195,
+      "learning_rate": 0.001,
+      "loss": 2.6635,
+      "step": 2792
+    },
+    {
+      "epoch": 0.11815720450122684,
+      "grad_norm": 0.323395311832428,
+      "learning_rate": 0.001,
+      "loss": 3.0143,
+      "step": 2793
+    },
+    {
+      "epoch": 0.11819950926474321,
+      "grad_norm": 0.2660515010356903,
+      "learning_rate": 0.001,
+      "loss": 1.9591,
+      "step": 2794
+    },
+    {
+      "epoch": 0.11824181402825958,
+      "grad_norm": 0.5746325850486755,
+      "learning_rate": 0.001,
+      "loss": 2.3898,
+      "step": 2795
+    },
+    {
+      "epoch": 0.11828411879177596,
+      "grad_norm": 0.2861972153186798,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 2796
+    },
+    {
+      "epoch": 0.11832642355529233,
+      "grad_norm": 0.2711239457130432,
+      "learning_rate": 0.001,
+      "loss": 2.5983,
+      "step": 2797
+    },
+    {
+      "epoch": 0.1183687283188087,
+      "grad_norm": 0.22106711566448212,
+      "learning_rate": 0.001,
+      "loss": 1.6675,
+      "step": 2798
+    },
+    {
+      "epoch": 0.11841103308232508,
+      "grad_norm": 0.2644370198249817,
+      "learning_rate": 0.001,
+      "loss": 2.7655,
+      "step": 2799
+    },
+    {
+      "epoch": 0.11845333784584144,
+      "grad_norm": 0.24654194712638855,
+      "learning_rate": 0.001,
+      "loss": 3.2642,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11849564260935781,
+      "grad_norm": 0.7044499516487122,
+      "learning_rate": 0.001,
+      "loss": 2.9117,
+      "step": 2801
+    },
+    {
+      "epoch": 0.11853794737287418,
+      "grad_norm": 0.3446405827999115,
+      "learning_rate": 0.001,
+      "loss": 2.4988,
+      "step": 2802
+    },
+    {
+      "epoch": 0.11858025213639056,
+      "grad_norm": 0.27378132939338684,
+      "learning_rate": 0.001,
+      "loss": 2.2514,
+      "step": 2803
+    },
+    {
+      "epoch": 0.11862255689990693,
+      "grad_norm": 0.20993216335773468,
+      "learning_rate": 0.001,
+      "loss": 2.4,
+      "step": 2804
+    },
+    {
+      "epoch": 0.1186648616634233,
+      "grad_norm": 0.5672168731689453,
+      "learning_rate": 0.001,
+      "loss": 2.1295,
+      "step": 2805
+    },
+    {
+      "epoch": 0.11870716642693967,
+      "grad_norm": 0.2943876385688782,
+      "learning_rate": 0.001,
+      "loss": 3.7169,
+      "step": 2806
+    },
+    {
+      "epoch": 0.11874947119045605,
+      "grad_norm": 0.6475502252578735,
+      "learning_rate": 0.001,
+      "loss": 2.3737,
+      "step": 2807
+    },
+    {
+      "epoch": 0.11879177595397242,
+      "grad_norm": 0.3072225749492645,
+      "learning_rate": 0.001,
+      "loss": 1.9827,
+      "step": 2808
+    },
+    {
+      "epoch": 0.11883408071748879,
+      "grad_norm": 0.31037554144859314,
+      "learning_rate": 0.001,
+      "loss": 1.558,
+      "step": 2809
+    },
+    {
+      "epoch": 0.11887638548100517,
+      "grad_norm": 0.23181147873401642,
+      "learning_rate": 0.001,
+      "loss": 2.0697,
+      "step": 2810
+    },
+    {
+      "epoch": 0.11891869024452154,
+      "grad_norm": 10.96942138671875,
+      "learning_rate": 0.001,
+      "loss": 1.6794,
+      "step": 2811
+    },
+    {
+      "epoch": 0.1189609950080379,
+      "grad_norm": 0.42705076932907104,
+      "learning_rate": 0.001,
+      "loss": 2.3502,
+      "step": 2812
+    },
+    {
+      "epoch": 0.11900329977155427,
+      "grad_norm": 0.26363325119018555,
+      "learning_rate": 0.001,
+      "loss": 3.0325,
+      "step": 2813
+    },
+    {
+      "epoch": 0.11904560453507065,
+      "grad_norm": 0.26286640763282776,
+      "learning_rate": 0.001,
+      "loss": 2.1994,
+      "step": 2814
+    },
+    {
+      "epoch": 0.11908790929858702,
+      "grad_norm": 0.2361530065536499,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 2815
+    },
+    {
+      "epoch": 0.11913021406210339,
+      "grad_norm": 0.8257160186767578,
+      "learning_rate": 0.001,
+      "loss": 3.0286,
+      "step": 2816
+    },
+    {
+      "epoch": 0.11917251882561976,
+      "grad_norm": 0.2814808487892151,
+      "learning_rate": 0.001,
+      "loss": 2.8173,
+      "step": 2817
+    },
+    {
+      "epoch": 0.11921482358913614,
+      "grad_norm": 0.23943133652210236,
+      "learning_rate": 0.001,
+      "loss": 3.0642,
+      "step": 2818
+    },
+    {
+      "epoch": 0.11925712835265251,
+      "grad_norm": 0.206394761800766,
+      "learning_rate": 0.001,
+      "loss": 1.8372,
+      "step": 2819
+    },
+    {
+      "epoch": 0.11929943311616888,
+      "grad_norm": 0.29365062713623047,
+      "learning_rate": 0.001,
+      "loss": 2.7685,
+      "step": 2820
+    },
+    {
+      "epoch": 0.11934173787968526,
+      "grad_norm": 0.2639111280441284,
+      "learning_rate": 0.001,
+      "loss": 4.2966,
+      "step": 2821
+    },
+    {
+      "epoch": 0.11938404264320163,
+      "grad_norm": 0.24708105623722076,
+      "learning_rate": 0.001,
+      "loss": 2.184,
+      "step": 2822
+    },
+    {
+      "epoch": 0.119426347406718,
+      "grad_norm": 0.3220023512840271,
+      "learning_rate": 0.001,
+      "loss": 2.0879,
+      "step": 2823
+    },
+    {
+      "epoch": 0.11946865217023436,
+      "grad_norm": 0.30507785081863403,
+      "learning_rate": 0.001,
+      "loss": 2.2369,
+      "step": 2824
+    },
+    {
+      "epoch": 0.11951095693375074,
+      "grad_norm": 0.23936758935451508,
+      "learning_rate": 0.001,
+      "loss": 3.1638,
+      "step": 2825
+    },
+    {
+      "epoch": 0.11955326169726711,
+      "grad_norm": 0.5877100825309753,
+      "learning_rate": 0.001,
+      "loss": 2.9414,
+      "step": 2826
+    },
+    {
+      "epoch": 0.11959556646078348,
+      "grad_norm": 0.21661652624607086,
+      "learning_rate": 0.001,
+      "loss": 2.3561,
+      "step": 2827
+    },
+    {
+      "epoch": 0.11963787122429985,
+      "grad_norm": 0.2344318926334381,
+      "learning_rate": 0.001,
+      "loss": 2.9277,
+      "step": 2828
+    },
+    {
+      "epoch": 0.11968017598781623,
+      "grad_norm": 0.3025057017803192,
+      "learning_rate": 0.001,
+      "loss": 2.3685,
+      "step": 2829
+    },
+    {
+      "epoch": 0.1197224807513326,
+      "grad_norm": 0.24070441722869873,
+      "learning_rate": 0.001,
+      "loss": 3.1727,
+      "step": 2830
+    },
+    {
+      "epoch": 0.11976478551484897,
+      "grad_norm": 0.25270789861679077,
+      "learning_rate": 0.001,
+      "loss": 2.9425,
+      "step": 2831
+    },
+    {
+      "epoch": 0.11980709027836535,
+      "grad_norm": 0.23945005238056183,
+      "learning_rate": 0.001,
+      "loss": 3.2619,
+      "step": 2832
+    },
+    {
+      "epoch": 0.11984939504188172,
+      "grad_norm": 0.257040798664093,
+      "learning_rate": 0.001,
+      "loss": 2.4056,
+      "step": 2833
+    },
+    {
+      "epoch": 0.11989169980539809,
+      "grad_norm": 0.3537012040615082,
+      "learning_rate": 0.001,
+      "loss": 2.5206,
+      "step": 2834
+    },
+    {
+      "epoch": 0.11993400456891445,
+      "grad_norm": 0.9324373006820679,
+      "learning_rate": 0.001,
+      "loss": 3.0553,
+      "step": 2835
+    },
+    {
+      "epoch": 0.11997630933243084,
+      "grad_norm": 0.32961177825927734,
+      "learning_rate": 0.001,
+      "loss": 1.9009,
+      "step": 2836
+    },
+    {
+      "epoch": 0.1200186140959472,
+      "grad_norm": 0.21902523934841156,
+      "learning_rate": 0.001,
+      "loss": 2.2892,
+      "step": 2837
+    },
+    {
+      "epoch": 0.12006091885946357,
+      "grad_norm": 0.35956844687461853,
+      "learning_rate": 0.001,
+      "loss": 2.2198,
+      "step": 2838
+    },
+    {
+      "epoch": 0.12010322362297995,
+      "grad_norm": 0.31855231523513794,
+      "learning_rate": 0.001,
+      "loss": 2.6221,
+      "step": 2839
+    },
+    {
+      "epoch": 0.12014552838649632,
+      "grad_norm": 1.0669035911560059,
+      "learning_rate": 0.001,
+      "loss": 2.6779,
+      "step": 2840
+    },
+    {
+      "epoch": 0.12018783315001269,
+      "grad_norm": 0.5651441812515259,
+      "learning_rate": 0.001,
+      "loss": 2.3371,
+      "step": 2841
+    },
+    {
+      "epoch": 0.12023013791352906,
+      "grad_norm": 0.2813262939453125,
+      "learning_rate": 0.001,
+      "loss": 2.847,
+      "step": 2842
+    },
+    {
+      "epoch": 0.12027244267704544,
+      "grad_norm": 0.21474647521972656,
+      "learning_rate": 0.001,
+      "loss": 2.262,
+      "step": 2843
+    },
+    {
+      "epoch": 0.12031474744056181,
+      "grad_norm": 0.9822518229484558,
+      "learning_rate": 0.001,
+      "loss": 2.1823,
+      "step": 2844
+    },
+    {
+      "epoch": 0.12035705220407818,
+      "grad_norm": 0.4417326748371124,
+      "learning_rate": 0.001,
+      "loss": 1.6555,
+      "step": 2845
+    },
+    {
+      "epoch": 0.12039935696759455,
+      "grad_norm": 0.31064194440841675,
+      "learning_rate": 0.001,
+      "loss": 1.981,
+      "step": 2846
+    },
+    {
+      "epoch": 0.12044166173111093,
+      "grad_norm": 1.0444011688232422,
+      "learning_rate": 0.001,
+      "loss": 2.0207,
+      "step": 2847
+    },
+    {
+      "epoch": 0.1204839664946273,
+      "grad_norm": 0.25042444467544556,
+      "learning_rate": 0.001,
+      "loss": 2.2386,
+      "step": 2848
+    },
+    {
+      "epoch": 0.12052627125814366,
+      "grad_norm": 0.26854464411735535,
+      "learning_rate": 0.001,
+      "loss": 2.7635,
+      "step": 2849
+    },
+    {
+      "epoch": 0.12056857602166005,
+      "grad_norm": 0.28399658203125,
+      "learning_rate": 0.001,
+      "loss": 2.6783,
+      "step": 2850
+    },
+    {
+      "epoch": 0.12061088078517641,
+      "grad_norm": 0.3274761736392975,
+      "learning_rate": 0.001,
+      "loss": 2.8354,
+      "step": 2851
+    },
+    {
+      "epoch": 0.12065318554869278,
+      "grad_norm": 0.44863566756248474,
+      "learning_rate": 0.001,
+      "loss": 3.583,
+      "step": 2852
+    },
+    {
+      "epoch": 0.12069549031220915,
+      "grad_norm": 0.2759772539138794,
+      "learning_rate": 0.001,
+      "loss": 2.4853,
+      "step": 2853
+    },
+    {
+      "epoch": 0.12073779507572553,
+      "grad_norm": 0.3000403344631195,
+      "learning_rate": 0.001,
+      "loss": 2.8492,
+      "step": 2854
+    },
+    {
+      "epoch": 0.1207800998392419,
+      "grad_norm": 0.2891789376735687,
+      "learning_rate": 0.001,
+      "loss": 2.5121,
+      "step": 2855
+    },
+    {
+      "epoch": 0.12082240460275827,
+      "grad_norm": 0.26525014638900757,
+      "learning_rate": 0.001,
+      "loss": 3.1642,
+      "step": 2856
+    },
+    {
+      "epoch": 0.12086470936627464,
+      "grad_norm": 2.695129156112671,
+      "learning_rate": 0.001,
+      "loss": 3.6222,
+      "step": 2857
+    },
+    {
+      "epoch": 0.12090701412979102,
+      "grad_norm": 0.27255064249038696,
+      "learning_rate": 0.001,
+      "loss": 2.6515,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12094931889330739,
+      "grad_norm": 1.2651582956314087,
+      "learning_rate": 0.001,
+      "loss": 3.0903,
+      "step": 2859
+    },
+    {
+      "epoch": 0.12099162365682375,
+      "grad_norm": 0.42988303303718567,
+      "learning_rate": 0.001,
+      "loss": 2.2843,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12103392842034014,
+      "grad_norm": 0.27339550852775574,
+      "learning_rate": 0.001,
+      "loss": 2.0482,
+      "step": 2861
+    },
+    {
+      "epoch": 0.1210762331838565,
+      "grad_norm": 0.43721139430999756,
+      "learning_rate": 0.001,
+      "loss": 3.0085,
+      "step": 2862
+    },
+    {
+      "epoch": 0.12111853794737287,
+      "grad_norm": 0.5759567022323608,
+      "learning_rate": 0.001,
+      "loss": 3.0561,
+      "step": 2863
+    },
+    {
+      "epoch": 0.12116084271088924,
+      "grad_norm": 1.3344426155090332,
+      "learning_rate": 0.001,
+      "loss": 2.0301,
+      "step": 2864
+    },
+    {
+      "epoch": 0.12120314747440562,
+      "grad_norm": 0.325006902217865,
+      "learning_rate": 0.001,
+      "loss": 2.0085,
+      "step": 2865
+    },
+    {
+      "epoch": 0.12124545223792199,
+      "grad_norm": 0.27443817257881165,
+      "learning_rate": 0.001,
+      "loss": 1.91,
+      "step": 2866
+    },
+    {
+      "epoch": 0.12128775700143836,
+      "grad_norm": 0.5229665637016296,
+      "learning_rate": 0.001,
+      "loss": 2.2958,
+      "step": 2867
+    },
+    {
+      "epoch": 0.12133006176495473,
+      "grad_norm": 0.22957605123519897,
+      "learning_rate": 0.001,
+      "loss": 1.9397,
+      "step": 2868
+    },
+    {
+      "epoch": 0.12137236652847111,
+      "grad_norm": 0.23909400403499603,
+      "learning_rate": 0.001,
+      "loss": 2.603,
+      "step": 2869
+    },
+    {
+      "epoch": 0.12141467129198748,
+      "grad_norm": 0.9538590908050537,
+      "learning_rate": 0.001,
+      "loss": 1.8949,
+      "step": 2870
+    },
+    {
+      "epoch": 0.12145697605550385,
+      "grad_norm": 0.25031131505966187,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 2871
+    },
+    {
+      "epoch": 0.12149928081902023,
+      "grad_norm": 0.30926719307899475,
+      "learning_rate": 0.001,
+      "loss": 2.2511,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1215415855825366,
+      "grad_norm": 0.2400166243314743,
+      "learning_rate": 0.001,
+      "loss": 2.2112,
+      "step": 2873
+    },
+    {
+      "epoch": 0.12158389034605296,
+      "grad_norm": 0.5655799508094788,
+      "learning_rate": 0.001,
+      "loss": 2.8809,
+      "step": 2874
+    },
+    {
+      "epoch": 0.12162619510956933,
+      "grad_norm": 0.495080828666687,
+      "learning_rate": 0.001,
+      "loss": 2.6534,
+      "step": 2875
+    },
+    {
+      "epoch": 0.12166849987308571,
+      "grad_norm": 0.46104100346565247,
+      "learning_rate": 0.001,
+      "loss": 2.5455,
+      "step": 2876
+    },
+    {
+      "epoch": 0.12171080463660208,
+      "grad_norm": 3.711001396179199,
+      "learning_rate": 0.001,
+      "loss": 2.2507,
+      "step": 2877
+    },
+    {
+      "epoch": 0.12175310940011845,
+      "grad_norm": 0.30323144793510437,
+      "learning_rate": 0.001,
+      "loss": 2.8051,
+      "step": 2878
+    },
+    {
+      "epoch": 0.12179541416363482,
+      "grad_norm": 0.5401155948638916,
+      "learning_rate": 0.001,
+      "loss": 1.596,
+      "step": 2879
+    },
+    {
+      "epoch": 0.1218377189271512,
+      "grad_norm": 0.28242409229278564,
+      "learning_rate": 0.001,
+      "loss": 2.9103,
+      "step": 2880
+    },
+    {
+      "epoch": 0.12188002369066757,
+      "grad_norm": 0.5616193413734436,
+      "learning_rate": 0.001,
+      "loss": 2.0185,
+      "step": 2881
+    },
+    {
+      "epoch": 0.12192232845418394,
+      "grad_norm": 0.48654183745384216,
+      "learning_rate": 0.001,
+      "loss": 1.996,
+      "step": 2882
+    },
+    {
+      "epoch": 0.12196463321770032,
+      "grad_norm": 0.2724916338920593,
+      "learning_rate": 0.001,
+      "loss": 2.6126,
+      "step": 2883
+    },
+    {
+      "epoch": 0.12200693798121669,
+      "grad_norm": 0.3358596861362457,
+      "learning_rate": 0.001,
+      "loss": 2.2338,
+      "step": 2884
+    },
+    {
+      "epoch": 0.12204924274473306,
+      "grad_norm": 0.28303262591362,
+      "learning_rate": 0.001,
+      "loss": 3.4625,
+      "step": 2885
+    },
+    {
+      "epoch": 0.12209154750824942,
+      "grad_norm": 1.536812663078308,
+      "learning_rate": 0.001,
+      "loss": 2.7973,
+      "step": 2886
+    },
+    {
+      "epoch": 0.1221338522717658,
+      "grad_norm": 0.2855709195137024,
+      "learning_rate": 0.001,
+      "loss": 2.583,
+      "step": 2887
+    },
+    {
+      "epoch": 0.12217615703528217,
+      "grad_norm": 3.504598379135132,
+      "learning_rate": 0.001,
+      "loss": 2.1699,
+      "step": 2888
+    },
+    {
+      "epoch": 0.12221846179879854,
+      "grad_norm": 1.3819570541381836,
+      "learning_rate": 0.001,
+      "loss": 3.72,
+      "step": 2889
+    },
+    {
+      "epoch": 0.12226076656231491,
+      "grad_norm": 0.21111451089382172,
+      "learning_rate": 0.001,
+      "loss": 1.495,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12230307132583129,
+      "grad_norm": 0.4661988317966461,
+      "learning_rate": 0.001,
+      "loss": 2.697,
+      "step": 2891
+    },
+    {
+      "epoch": 0.12234537608934766,
+      "grad_norm": 0.47962015867233276,
+      "learning_rate": 0.001,
+      "loss": 1.6233,
+      "step": 2892
+    },
+    {
+      "epoch": 0.12238768085286403,
+      "grad_norm": 0.33557215332984924,
+      "learning_rate": 0.001,
+      "loss": 3.1164,
+      "step": 2893
+    },
+    {
+      "epoch": 0.12242998561638041,
+      "grad_norm": 0.25374147295951843,
+      "learning_rate": 0.001,
+      "loss": 2.2069,
+      "step": 2894
+    },
+    {
+      "epoch": 0.12247229037989678,
+      "grad_norm": 0.397442489862442,
+      "learning_rate": 0.001,
+      "loss": 2.6315,
+      "step": 2895
+    },
+    {
+      "epoch": 0.12251459514341315,
+      "grad_norm": 0.2583683729171753,
+      "learning_rate": 0.001,
+      "loss": 2.516,
+      "step": 2896
+    },
+    {
+      "epoch": 0.12255689990692951,
+      "grad_norm": 0.3760106861591339,
+      "learning_rate": 0.001,
+      "loss": 1.859,
+      "step": 2897
+    },
+    {
+      "epoch": 0.1225992046704459,
+      "grad_norm": 0.3791460394859314,
+      "learning_rate": 0.001,
+      "loss": 2.9503,
+      "step": 2898
+    },
+    {
+      "epoch": 0.12264150943396226,
+      "grad_norm": 0.5899031162261963,
+      "learning_rate": 0.001,
+      "loss": 2.4752,
+      "step": 2899
+    },
+    {
+      "epoch": 0.12268381419747863,
+      "grad_norm": 0.29078209400177,
+      "learning_rate": 0.001,
+      "loss": 2.3481,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12272611896099501,
+      "grad_norm": 6.856618881225586,
+      "learning_rate": 0.001,
+      "loss": 3.0669,
+      "step": 2901
+    },
+    {
+      "epoch": 0.12276842372451138,
+      "grad_norm": 1.2403521537780762,
+      "learning_rate": 0.001,
+      "loss": 1.9102,
+      "step": 2902
+    },
+    {
+      "epoch": 0.12281072848802775,
+      "grad_norm": 0.6264525055885315,
+      "learning_rate": 0.001,
+      "loss": 1.9421,
+      "step": 2903
+    },
+    {
+      "epoch": 0.12285303325154412,
+      "grad_norm": 0.9500356912612915,
+      "learning_rate": 0.001,
+      "loss": 2.146,
+      "step": 2904
+    },
+    {
+      "epoch": 0.1228953380150605,
+      "grad_norm": 0.27822932600975037,
+      "learning_rate": 0.001,
+      "loss": 2.3955,
+      "step": 2905
+    },
+    {
+      "epoch": 0.12293764277857687,
+      "grad_norm": 0.27553531527519226,
+      "learning_rate": 0.001,
+      "loss": 2.3293,
+      "step": 2906
+    },
+    {
+      "epoch": 0.12297994754209324,
+      "grad_norm": 0.2871011197566986,
+      "learning_rate": 0.001,
+      "loss": 2.0411,
+      "step": 2907
+    },
+    {
+      "epoch": 0.1230222523056096,
+      "grad_norm": 0.3595285713672638,
+      "learning_rate": 0.001,
+      "loss": 2.316,
+      "step": 2908
+    },
+    {
+      "epoch": 0.12306455706912599,
+      "grad_norm": 0.32838475704193115,
+      "learning_rate": 0.001,
+      "loss": 3.3189,
+      "step": 2909
+    },
+    {
+      "epoch": 0.12310686183264236,
+      "grad_norm": 0.2596978545188904,
+      "learning_rate": 0.001,
+      "loss": 2.4232,
+      "step": 2910
+    },
+    {
+      "epoch": 0.12314916659615872,
+      "grad_norm": 0.2212909609079361,
+      "learning_rate": 0.001,
+      "loss": 2.0995,
+      "step": 2911
+    },
+    {
+      "epoch": 0.1231914713596751,
+      "grad_norm": 0.2778913974761963,
+      "learning_rate": 0.001,
+      "loss": 2.1786,
+      "step": 2912
+    },
+    {
+      "epoch": 0.12323377612319147,
+      "grad_norm": 0.23343005776405334,
+      "learning_rate": 0.001,
+      "loss": 2.1327,
+      "step": 2913
+    },
+    {
+      "epoch": 0.12327608088670784,
+      "grad_norm": 0.3342663049697876,
+      "learning_rate": 0.001,
+      "loss": 2.605,
+      "step": 2914
+    },
+    {
+      "epoch": 0.12331838565022421,
+      "grad_norm": 0.23159651458263397,
+      "learning_rate": 0.001,
+      "loss": 2.4356,
+      "step": 2915
+    },
+    {
+      "epoch": 0.12336069041374059,
+      "grad_norm": 0.36005109548568726,
+      "learning_rate": 0.001,
+      "loss": 1.9138,
+      "step": 2916
+    },
+    {
+      "epoch": 0.12340299517725696,
+      "grad_norm": 0.21737025678157806,
+      "learning_rate": 0.001,
+      "loss": 1.9291,
+      "step": 2917
+    },
+    {
+      "epoch": 0.12344529994077333,
+      "grad_norm": 0.9176145195960999,
+      "learning_rate": 0.001,
+      "loss": 1.9523,
+      "step": 2918
+    },
+    {
+      "epoch": 0.1234876047042897,
+      "grad_norm": 0.4270826280117035,
+      "learning_rate": 0.001,
+      "loss": 2.6282,
+      "step": 2919
+    },
+    {
+      "epoch": 0.12352990946780608,
+      "grad_norm": 0.23092937469482422,
+      "learning_rate": 0.001,
+      "loss": 2.1475,
+      "step": 2920
+    },
+    {
+      "epoch": 0.12357221423132245,
+      "grad_norm": 0.32089483737945557,
+      "learning_rate": 0.001,
+      "loss": 2.8787,
+      "step": 2921
+    },
+    {
+      "epoch": 0.12361451899483882,
+      "grad_norm": 0.5019152164459229,
+      "learning_rate": 0.001,
+      "loss": 2.4339,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1236568237583552,
+      "grad_norm": 8.176496505737305,
+      "learning_rate": 0.001,
+      "loss": 2.9616,
+      "step": 2923
+    },
+    {
+      "epoch": 0.12369912852187157,
+      "grad_norm": 0.42185178399086,
+      "learning_rate": 0.001,
+      "loss": 3.3083,
+      "step": 2924
+    },
+    {
+      "epoch": 0.12374143328538793,
+      "grad_norm": 0.6926589608192444,
+      "learning_rate": 0.001,
+      "loss": 4.3947,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1237837380489043,
+      "grad_norm": 2.2707536220550537,
+      "learning_rate": 0.001,
+      "loss": 1.8298,
+      "step": 2926
+    },
+    {
+      "epoch": 0.12382604281242068,
+      "grad_norm": 0.2851990759372711,
+      "learning_rate": 0.001,
+      "loss": 2.0727,
+      "step": 2927
+    },
+    {
+      "epoch": 0.12386834757593705,
+      "grad_norm": 0.7518361210823059,
+      "learning_rate": 0.001,
+      "loss": 1.6419,
+      "step": 2928
+    },
+    {
+      "epoch": 0.12391065233945342,
+      "grad_norm": 0.2845136523246765,
+      "learning_rate": 0.001,
+      "loss": 2.6422,
+      "step": 2929
+    },
+    {
+      "epoch": 0.12395295710296979,
+      "grad_norm": 0.32894209027290344,
+      "learning_rate": 0.001,
+      "loss": 2.7414,
+      "step": 2930
+    },
+    {
+      "epoch": 0.12399526186648617,
+      "grad_norm": 1.5604556798934937,
+      "learning_rate": 0.001,
+      "loss": 2.3227,
+      "step": 2931
+    },
+    {
+      "epoch": 0.12403756663000254,
+      "grad_norm": 0.49824783205986023,
+      "learning_rate": 0.001,
+      "loss": 1.9455,
+      "step": 2932
+    },
+    {
+      "epoch": 0.1240798713935189,
+      "grad_norm": 0.40257930755615234,
+      "learning_rate": 0.001,
+      "loss": 2.6387,
+      "step": 2933
+    },
+    {
+      "epoch": 0.12412217615703529,
+      "grad_norm": 0.27476415038108826,
+      "learning_rate": 0.001,
+      "loss": 2.3756,
+      "step": 2934
+    },
+    {
+      "epoch": 0.12416448092055166,
+      "grad_norm": 0.3869374096393585,
+      "learning_rate": 0.001,
+      "loss": 2.8796,
+      "step": 2935
+    },
+    {
+      "epoch": 0.12420678568406802,
+      "grad_norm": 0.47996222972869873,
+      "learning_rate": 0.001,
+      "loss": 2.8597,
+      "step": 2936
+    },
+    {
+      "epoch": 0.12424909044758439,
+      "grad_norm": 0.2530654966831207,
+      "learning_rate": 0.001,
+      "loss": 2.6784,
+      "step": 2937
+    },
+    {
+      "epoch": 0.12429139521110077,
+      "grad_norm": 0.8970469832420349,
+      "learning_rate": 0.001,
+      "loss": 2.6644,
+      "step": 2938
+    },
+    {
+      "epoch": 0.12433369997461714,
+      "grad_norm": 0.24477270245552063,
+      "learning_rate": 0.001,
+      "loss": 2.4868,
+      "step": 2939
+    },
+    {
+      "epoch": 0.12437600473813351,
+      "grad_norm": 1.4595485925674438,
+      "learning_rate": 0.001,
+      "loss": 2.1721,
+      "step": 2940
+    },
+    {
+      "epoch": 0.12441830950164988,
+      "grad_norm": 0.31102073192596436,
+      "learning_rate": 0.001,
+      "loss": 2.7839,
+      "step": 2941
+    },
+    {
+      "epoch": 0.12446061426516626,
+      "grad_norm": 1.8330522775650024,
+      "learning_rate": 0.001,
+      "loss": 3.1596,
+      "step": 2942
+    },
+    {
+      "epoch": 0.12450291902868263,
+      "grad_norm": 0.30926570296287537,
+      "learning_rate": 0.001,
+      "loss": 3.4155,
+      "step": 2943
+    },
+    {
+      "epoch": 0.124545223792199,
+      "grad_norm": 0.3584842085838318,
+      "learning_rate": 0.001,
+      "loss": 5.0056,
+      "step": 2944
+    },
+    {
+      "epoch": 0.12458752855571538,
+      "grad_norm": 0.24263957142829895,
+      "learning_rate": 0.001,
+      "loss": 1.7688,
+      "step": 2945
+    },
+    {
+      "epoch": 0.12462983331923175,
+      "grad_norm": 0.7257891893386841,
+      "learning_rate": 0.001,
+      "loss": 2.9134,
+      "step": 2946
+    },
+    {
+      "epoch": 0.12467213808274812,
+      "grad_norm": 0.2676026225090027,
+      "learning_rate": 0.001,
+      "loss": 3.1154,
+      "step": 2947
+    },
+    {
+      "epoch": 0.12471444284626448,
+      "grad_norm": 1.5698000192642212,
+      "learning_rate": 0.001,
+      "loss": 2.7397,
+      "step": 2948
+    },
+    {
+      "epoch": 0.12475674760978087,
+      "grad_norm": 0.2666419744491577,
+      "learning_rate": 0.001,
+      "loss": 2.6645,
+      "step": 2949
+    },
+    {
+      "epoch": 0.12479905237329723,
+      "grad_norm": 0.9079283475875854,
+      "learning_rate": 0.001,
+      "loss": 1.6547,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1248413571368136,
+      "grad_norm": 0.27443641424179077,
+      "learning_rate": 0.001,
+      "loss": 2.2036,
+      "step": 2951
+    },
+    {
+      "epoch": 0.12488366190032997,
+      "grad_norm": 0.27384769916534424,
+      "learning_rate": 0.001,
+      "loss": 2.2574,
+      "step": 2952
+    },
+    {
+      "epoch": 0.12492596666384635,
+      "grad_norm": 0.37685632705688477,
+      "learning_rate": 0.001,
+      "loss": 2.1753,
+      "step": 2953
+    },
+    {
+      "epoch": 0.12496827142736272,
+      "grad_norm": 0.5168205499649048,
+      "learning_rate": 0.001,
+      "loss": 3.1164,
+      "step": 2954
+    },
+    {
+      "epoch": 0.1250105761908791,
+      "grad_norm": 0.22727070748806,
+      "learning_rate": 0.001,
+      "loss": 1.8189,
+      "step": 2955
+    },
+    {
+      "epoch": 0.12505288095439546,
+      "grad_norm": 2.466627836227417,
+      "learning_rate": 0.001,
+      "loss": 1.925,
+      "step": 2956
+    },
+    {
+      "epoch": 0.12509518571791184,
+      "grad_norm": 0.37915435433387756,
+      "learning_rate": 0.001,
+      "loss": 3.6095,
+      "step": 2957
+    },
+    {
+      "epoch": 0.12513749048142822,
+      "grad_norm": 0.3377191424369812,
+      "learning_rate": 0.001,
+      "loss": 3.52,
+      "step": 2958
+    },
+    {
+      "epoch": 0.12517979524494457,
+      "grad_norm": 0.5984206199645996,
+      "learning_rate": 0.001,
+      "loss": 2.0392,
+      "step": 2959
+    },
+    {
+      "epoch": 0.12522210000846096,
+      "grad_norm": 0.2659249007701874,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1252644047719773,
+      "grad_norm": 0.25528255105018616,
+      "learning_rate": 0.001,
+      "loss": 2.2558,
+      "step": 2961
+    },
+    {
+      "epoch": 0.1253067095354937,
+      "grad_norm": 0.3032079339027405,
+      "learning_rate": 0.001,
+      "loss": 2.3852,
+      "step": 2962
+    },
+    {
+      "epoch": 0.12534901429901008,
+      "grad_norm": 0.29467710852622986,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 2963
+    },
+    {
+      "epoch": 0.12539131906252643,
+      "grad_norm": 0.5475602746009827,
+      "learning_rate": 0.001,
+      "loss": 2.0179,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1254336238260428,
+      "grad_norm": 0.2524794340133667,
+      "learning_rate": 0.001,
+      "loss": 3.1053,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1254759285895592,
+      "grad_norm": 0.26027101278305054,
+      "learning_rate": 0.001,
+      "loss": 2.4225,
+      "step": 2966
+    },
+    {
+      "epoch": 0.12551823335307555,
+      "grad_norm": 0.2270876169204712,
+      "learning_rate": 0.001,
+      "loss": 1.8372,
+      "step": 2967
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 1.1829967498779297,
+      "learning_rate": 0.001,
+      "loss": 2.1506,
+      "step": 2968
+    },
+    {
+      "epoch": 0.1256028428801083,
+      "grad_norm": 0.423272967338562,
+      "learning_rate": 0.001,
+      "loss": 2.3921,
+      "step": 2969
+    },
+    {
+      "epoch": 0.12564514764362467,
+      "grad_norm": 0.8556601405143738,
+      "learning_rate": 0.001,
+      "loss": 2.0293,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12568745240714105,
+      "grad_norm": 0.28667885065078735,
+      "learning_rate": 0.001,
+      "loss": 2.2469,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1257297571706574,
+      "grad_norm": 0.3176821768283844,
+      "learning_rate": 0.001,
+      "loss": 2.315,
+      "step": 2972
+    },
+    {
+      "epoch": 0.12577206193417378,
+      "grad_norm": 0.3208604156970978,
+      "learning_rate": 0.001,
+      "loss": 3.3225,
+      "step": 2973
+    },
+    {
+      "epoch": 0.12581436669769017,
+      "grad_norm": 0.3094156086444855,
+      "learning_rate": 0.001,
+      "loss": 1.8009,
+      "step": 2974
+    },
+    {
+      "epoch": 0.12585667146120652,
+      "grad_norm": 0.2794618606567383,
+      "learning_rate": 0.001,
+      "loss": 2.4301,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1258989762247229,
+      "grad_norm": 1.7522642612457275,
+      "learning_rate": 0.001,
+      "loss": 2.5346,
+      "step": 2976
+    },
+    {
+      "epoch": 0.12594128098823928,
+      "grad_norm": 0.7548322081565857,
+      "learning_rate": 0.001,
+      "loss": 2.3579,
+      "step": 2977
+    },
+    {
+      "epoch": 0.12598358575175564,
+      "grad_norm": 0.31399986147880554,
+      "learning_rate": 0.001,
+      "loss": 2.8131,
+      "step": 2978
+    },
+    {
+      "epoch": 0.12602589051527202,
+      "grad_norm": 0.5363809466362,
+      "learning_rate": 0.001,
+      "loss": 2.5607,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1260681952787884,
+      "grad_norm": 0.495072603225708,
+      "learning_rate": 0.001,
+      "loss": 4.0278,
+      "step": 2980
+    },
+    {
+      "epoch": 0.12611050004230476,
+      "grad_norm": 0.36730262637138367,
+      "learning_rate": 0.001,
+      "loss": 2.1536,
+      "step": 2981
+    },
+    {
+      "epoch": 0.12615280480582114,
+      "grad_norm": 0.31238043308258057,
+      "learning_rate": 0.001,
+      "loss": 2.0246,
+      "step": 2982
+    },
+    {
+      "epoch": 0.1261951095693375,
+      "grad_norm": 0.2928439974784851,
+      "learning_rate": 0.001,
+      "loss": 3.6699,
+      "step": 2983
+    },
+    {
+      "epoch": 0.12623741433285388,
+      "grad_norm": 0.31920191645622253,
+      "learning_rate": 0.001,
+      "loss": 3.32,
+      "step": 2984
+    },
+    {
+      "epoch": 0.12627971909637026,
+      "grad_norm": 0.27571341395378113,
+      "learning_rate": 0.001,
+      "loss": 1.9494,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1263220238598866,
+      "grad_norm": 0.3193123936653137,
+      "learning_rate": 0.001,
+      "loss": 2.1162,
+      "step": 2986
+    },
+    {
+      "epoch": 0.126364328623403,
+      "grad_norm": 0.24550597369670868,
+      "learning_rate": 0.001,
+      "loss": 3.5763,
+      "step": 2987
+    },
+    {
+      "epoch": 0.12640663338691938,
+      "grad_norm": 0.2659911513328552,
+      "learning_rate": 0.001,
+      "loss": 2.2886,
+      "step": 2988
+    },
+    {
+      "epoch": 0.12644893815043573,
+      "grad_norm": 0.3616361916065216,
+      "learning_rate": 0.001,
+      "loss": 2.4602,
+      "step": 2989
+    },
+    {
+      "epoch": 0.1264912429139521,
+      "grad_norm": 1.0882079601287842,
+      "learning_rate": 0.001,
+      "loss": 1.6782,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1265335476774685,
+      "grad_norm": 0.1997089982032776,
+      "learning_rate": 0.001,
+      "loss": 2.1267,
+      "step": 2991
+    },
+    {
+      "epoch": 0.12657585244098485,
+      "grad_norm": 0.2554531991481781,
+      "learning_rate": 0.001,
+      "loss": 2.2105,
+      "step": 2992
+    },
+    {
+      "epoch": 0.12661815720450123,
+      "grad_norm": 1.5472227334976196,
+      "learning_rate": 0.001,
+      "loss": 2.6065,
+      "step": 2993
+    },
+    {
+      "epoch": 0.1266604619680176,
+      "grad_norm": 0.4191298186779022,
+      "learning_rate": 0.001,
+      "loss": 3.8673,
+      "step": 2994
+    },
+    {
+      "epoch": 0.12670276673153397,
+      "grad_norm": 0.3299647569656372,
+      "learning_rate": 0.001,
+      "loss": 2.6935,
+      "step": 2995
+    },
+    {
+      "epoch": 0.12674507149505035,
+      "grad_norm": 0.26964640617370605,
+      "learning_rate": 0.001,
+      "loss": 2.0866,
+      "step": 2996
+    },
+    {
+      "epoch": 0.1267873762585667,
+      "grad_norm": 0.22724339365959167,
+      "learning_rate": 0.001,
+      "loss": 1.6615,
+      "step": 2997
+    },
+    {
+      "epoch": 0.12682968102208309,
+      "grad_norm": 0.28157809376716614,
+      "learning_rate": 0.001,
+      "loss": 2.4699,
+      "step": 2998
+    },
+    {
+      "epoch": 0.12687198578559947,
+      "grad_norm": 0.3091438412666321,
+      "learning_rate": 0.001,
+      "loss": 2.359,
+      "step": 2999
+    },
+    {
+      "epoch": 0.12691429054911582,
+      "grad_norm": 0.2963161766529083,
+      "learning_rate": 0.001,
+      "loss": 2.3396,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1269565953126322,
+      "grad_norm": 0.3389158546924591,
+      "learning_rate": 0.001,
+      "loss": 3.0763,
+      "step": 3001
+    },
+    {
+      "epoch": 0.12699890007614859,
+      "grad_norm": 1.0668145418167114,
+      "learning_rate": 0.001,
+      "loss": 2.6003,
+      "step": 3002
+    },
+    {
+      "epoch": 0.12704120483966494,
+      "grad_norm": 0.28413814306259155,
+      "learning_rate": 0.001,
+      "loss": 2.9304,
+      "step": 3003
+    },
+    {
+      "epoch": 0.12708350960318132,
+      "grad_norm": 0.5639888048171997,
+      "learning_rate": 0.001,
+      "loss": 2.1774,
+      "step": 3004
+    },
+    {
+      "epoch": 0.1271258143666977,
+      "grad_norm": 0.24791397154331207,
+      "learning_rate": 0.001,
+      "loss": 2.9347,
+      "step": 3005
+    },
+    {
+      "epoch": 0.12716811913021406,
+      "grad_norm": 0.23885676264762878,
+      "learning_rate": 0.001,
+      "loss": 2.189,
+      "step": 3006
+    },
+    {
+      "epoch": 0.12721042389373044,
+      "grad_norm": 0.2213631123304367,
+      "learning_rate": 0.001,
+      "loss": 2.3451,
+      "step": 3007
+    },
+    {
+      "epoch": 0.1272527286572468,
+      "grad_norm": 0.29912659525871277,
+      "learning_rate": 0.001,
+      "loss": 2.8164,
+      "step": 3008
+    },
+    {
+      "epoch": 0.12729503342076318,
+      "grad_norm": 0.7422712445259094,
+      "learning_rate": 0.001,
+      "loss": 2.4031,
+      "step": 3009
+    },
+    {
+      "epoch": 0.12733733818427956,
+      "grad_norm": 0.2729654014110565,
+      "learning_rate": 0.001,
+      "loss": 2.3894,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1273796429477959,
+      "grad_norm": 0.2171071618795395,
+      "learning_rate": 0.001,
+      "loss": 2.0108,
+      "step": 3011
+    },
+    {
+      "epoch": 0.1274219477113123,
+      "grad_norm": 0.2428562194108963,
+      "learning_rate": 0.001,
+      "loss": 2.9788,
+      "step": 3012
+    },
+    {
+      "epoch": 0.12746425247482868,
+      "grad_norm": 0.19264620542526245,
+      "learning_rate": 0.001,
+      "loss": 1.949,
+      "step": 3013
+    },
+    {
+      "epoch": 0.12750655723834503,
+      "grad_norm": 0.2231709510087967,
+      "learning_rate": 0.001,
+      "loss": 2.2655,
+      "step": 3014
+    },
+    {
+      "epoch": 0.1275488620018614,
+      "grad_norm": 0.6086738705635071,
+      "learning_rate": 0.001,
+      "loss": 3.2293,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1275911667653778,
+      "grad_norm": 0.20442450046539307,
+      "learning_rate": 0.001,
+      "loss": 1.6902,
+      "step": 3016
+    },
+    {
+      "epoch": 0.12763347152889415,
+      "grad_norm": 0.2568240463733673,
+      "learning_rate": 0.001,
+      "loss": 2.3126,
+      "step": 3017
+    },
+    {
+      "epoch": 0.12767577629241053,
+      "grad_norm": 0.6941277384757996,
+      "learning_rate": 0.001,
+      "loss": 2.5573,
+      "step": 3018
+    },
+    {
+      "epoch": 0.12771808105592689,
+      "grad_norm": 0.2820114493370056,
+      "learning_rate": 0.001,
+      "loss": 3.45,
+      "step": 3019
+    },
+    {
+      "epoch": 0.12776038581944327,
+      "grad_norm": 0.27247634530067444,
+      "learning_rate": 0.001,
+      "loss": 3.0103,
+      "step": 3020
+    },
+    {
+      "epoch": 0.12780269058295965,
+      "grad_norm": 0.30561110377311707,
+      "learning_rate": 0.001,
+      "loss": 2.9789,
+      "step": 3021
+    },
+    {
+      "epoch": 0.127844995346476,
+      "grad_norm": 0.7641719579696655,
+      "learning_rate": 0.001,
+      "loss": 1.9984,
+      "step": 3022
+    },
+    {
+      "epoch": 0.12788730010999239,
+      "grad_norm": 1.2558095455169678,
+      "learning_rate": 0.001,
+      "loss": 2.3797,
+      "step": 3023
+    },
+    {
+      "epoch": 0.12792960487350877,
+      "grad_norm": 0.24033454060554504,
+      "learning_rate": 0.001,
+      "loss": 2.5231,
+      "step": 3024
+    },
+    {
+      "epoch": 0.12797190963702512,
+      "grad_norm": 0.9906525611877441,
+      "learning_rate": 0.001,
+      "loss": 3.2656,
+      "step": 3025
+    },
+    {
+      "epoch": 0.1280142144005415,
+      "grad_norm": 0.35158273577690125,
+      "learning_rate": 0.001,
+      "loss": 2.5469,
+      "step": 3026
+    },
+    {
+      "epoch": 0.12805651916405789,
+      "grad_norm": 0.313698947429657,
+      "learning_rate": 0.001,
+      "loss": 2.2092,
+      "step": 3027
+    },
+    {
+      "epoch": 0.12809882392757424,
+      "grad_norm": 0.2757653594017029,
+      "learning_rate": 0.001,
+      "loss": 1.7885,
+      "step": 3028
+    },
+    {
+      "epoch": 0.12814112869109062,
+      "grad_norm": 0.23092950880527496,
+      "learning_rate": 0.001,
+      "loss": 1.9807,
+      "step": 3029
+    },
+    {
+      "epoch": 0.12818343345460698,
+      "grad_norm": 0.7450085282325745,
+      "learning_rate": 0.001,
+      "loss": 1.9991,
+      "step": 3030
+    },
+    {
+      "epoch": 0.12822573821812336,
+      "grad_norm": 0.362114816904068,
+      "learning_rate": 0.001,
+      "loss": 4.5653,
+      "step": 3031
+    },
+    {
+      "epoch": 0.12826804298163974,
+      "grad_norm": 0.23789288103580475,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 3032
+    },
+    {
+      "epoch": 0.1283103477451561,
+      "grad_norm": 0.27310964465141296,
+      "learning_rate": 0.001,
+      "loss": 3.2499,
+      "step": 3033
+    },
+    {
+      "epoch": 0.12835265250867248,
+      "grad_norm": 0.3834547996520996,
+      "learning_rate": 0.001,
+      "loss": 2.3692,
+      "step": 3034
+    },
+    {
+      "epoch": 0.12839495727218886,
+      "grad_norm": 0.28623712062835693,
+      "learning_rate": 0.001,
+      "loss": 2.875,
+      "step": 3035
+    },
+    {
+      "epoch": 0.1284372620357052,
+      "grad_norm": 0.35939669609069824,
+      "learning_rate": 0.001,
+      "loss": 2.4295,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1284795667992216,
+      "grad_norm": 0.403884619474411,
+      "learning_rate": 0.001,
+      "loss": 2.8118,
+      "step": 3037
+    },
+    {
+      "epoch": 0.12852187156273798,
+      "grad_norm": 0.7596374154090881,
+      "learning_rate": 0.001,
+      "loss": 2.5978,
+      "step": 3038
+    },
+    {
+      "epoch": 0.12856417632625433,
+      "grad_norm": 0.24331355094909668,
+      "learning_rate": 0.001,
+      "loss": 2.3842,
+      "step": 3039
+    },
+    {
+      "epoch": 0.1286064810897707,
+      "grad_norm": 0.23005782067775726,
+      "learning_rate": 0.001,
+      "loss": 2.1265,
+      "step": 3040
+    },
+    {
+      "epoch": 0.12864878585328707,
+      "grad_norm": 0.23082035779953003,
+      "learning_rate": 0.001,
+      "loss": 2.273,
+      "step": 3041
+    },
+    {
+      "epoch": 0.12869109061680345,
+      "grad_norm": 0.26438039541244507,
+      "learning_rate": 0.001,
+      "loss": 2.3408,
+      "step": 3042
+    },
+    {
+      "epoch": 0.12873339538031983,
+      "grad_norm": 0.22107988595962524,
+      "learning_rate": 0.001,
+      "loss": 1.9021,
+      "step": 3043
+    },
+    {
+      "epoch": 0.12877570014383619,
+      "grad_norm": 0.21788084506988525,
+      "learning_rate": 0.001,
+      "loss": 3.0715,
+      "step": 3044
+    },
+    {
+      "epoch": 0.12881800490735257,
+      "grad_norm": 0.23471501469612122,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 3045
+    },
+    {
+      "epoch": 0.12886030967086895,
+      "grad_norm": 0.28260231018066406,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1289026144343853,
+      "grad_norm": 0.26574090123176575,
+      "learning_rate": 0.001,
+      "loss": 2.0225,
+      "step": 3047
+    },
+    {
+      "epoch": 0.1289449191979017,
+      "grad_norm": 0.34297651052474976,
+      "learning_rate": 0.001,
+      "loss": 3.0373,
+      "step": 3048
+    },
+    {
+      "epoch": 0.12898722396141807,
+      "grad_norm": 0.586717426776886,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 3049
+    },
+    {
+      "epoch": 0.12902952872493442,
+      "grad_norm": 0.21093139052391052,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1290718334884508,
+      "grad_norm": 0.29111140966415405,
+      "learning_rate": 0.001,
+      "loss": 2.1633,
+      "step": 3051
+    },
+    {
+      "epoch": 0.12911413825196716,
+      "grad_norm": 0.30666249990463257,
+      "learning_rate": 0.001,
+      "loss": 3.0201,
+      "step": 3052
+    },
+    {
+      "epoch": 0.12915644301548354,
+      "grad_norm": 0.24410761892795563,
+      "learning_rate": 0.001,
+      "loss": 2.8614,
+      "step": 3053
+    },
+    {
+      "epoch": 0.12919874777899992,
+      "grad_norm": 0.2161339968442917,
+      "learning_rate": 0.001,
+      "loss": 1.7318,
+      "step": 3054
+    },
+    {
+      "epoch": 0.12924105254251628,
+      "grad_norm": 0.19016364216804504,
+      "learning_rate": 0.001,
+      "loss": 2.0828,
+      "step": 3055
+    },
+    {
+      "epoch": 0.12928335730603266,
+      "grad_norm": 0.23141834139823914,
+      "learning_rate": 0.001,
+      "loss": 2.8,
+      "step": 3056
+    },
+    {
+      "epoch": 0.12932566206954904,
+      "grad_norm": 1.5664108991622925,
+      "learning_rate": 0.001,
+      "loss": 2.1793,
+      "step": 3057
+    },
+    {
+      "epoch": 0.1293679668330654,
+      "grad_norm": 2.544483184814453,
+      "learning_rate": 0.001,
+      "loss": 2.9984,
+      "step": 3058
+    },
+    {
+      "epoch": 0.12941027159658178,
+      "grad_norm": 0.30240291357040405,
+      "learning_rate": 0.001,
+      "loss": 2.197,
+      "step": 3059
+    },
+    {
+      "epoch": 0.12945257636009816,
+      "grad_norm": 0.24638395011425018,
+      "learning_rate": 0.001,
+      "loss": 2.2356,
+      "step": 3060
+    },
+    {
+      "epoch": 0.1294948811236145,
+      "grad_norm": 0.9346217513084412,
+      "learning_rate": 0.001,
+      "loss": 2.1051,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1295371858871309,
+      "grad_norm": 0.256610244512558,
+      "learning_rate": 0.001,
+      "loss": 2.3009,
+      "step": 3062
+    },
+    {
+      "epoch": 0.12957949065064725,
+      "grad_norm": 0.40396857261657715,
+      "learning_rate": 0.001,
+      "loss": 2.5668,
+      "step": 3063
+    },
+    {
+      "epoch": 0.12962179541416363,
+      "grad_norm": 10.963554382324219,
+      "learning_rate": 0.001,
+      "loss": 2.1443,
+      "step": 3064
+    },
+    {
+      "epoch": 0.12966410017768001,
+      "grad_norm": 2.6234822273254395,
+      "learning_rate": 0.001,
+      "loss": 2.6211,
+      "step": 3065
+    },
+    {
+      "epoch": 0.12970640494119637,
+      "grad_norm": 0.3858783543109894,
+      "learning_rate": 0.001,
+      "loss": 1.9484,
+      "step": 3066
+    },
+    {
+      "epoch": 0.12974870970471275,
+      "grad_norm": 4.675527572631836,
+      "learning_rate": 0.001,
+      "loss": 2.551,
+      "step": 3067
+    },
+    {
+      "epoch": 0.12979101446822913,
+      "grad_norm": 0.3626082241535187,
+      "learning_rate": 0.001,
+      "loss": 1.7012,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1298333192317455,
+      "grad_norm": 0.605455756187439,
+      "learning_rate": 0.001,
+      "loss": 1.6341,
+      "step": 3069
+    },
+    {
+      "epoch": 0.12987562399526187,
+      "grad_norm": 0.38578104972839355,
+      "learning_rate": 0.001,
+      "loss": 2.0653,
+      "step": 3070
+    },
+    {
+      "epoch": 0.12991792875877825,
+      "grad_norm": 0.3008464574813843,
+      "learning_rate": 0.001,
+      "loss": 2.3254,
+      "step": 3071
+    },
+    {
+      "epoch": 0.1299602335222946,
+      "grad_norm": 0.31806275248527527,
+      "learning_rate": 0.001,
+      "loss": 2.4444,
+      "step": 3072
+    },
+    {
+      "epoch": 0.130002538285811,
+      "grad_norm": 0.3251250386238098,
+      "learning_rate": 0.001,
+      "loss": 2.173,
+      "step": 3073
+    },
+    {
+      "epoch": 0.13004484304932734,
+      "grad_norm": 0.3980516791343689,
+      "learning_rate": 0.001,
+      "loss": 2.8316,
+      "step": 3074
+    },
+    {
+      "epoch": 0.13008714781284372,
+      "grad_norm": 0.3091077506542206,
+      "learning_rate": 0.001,
+      "loss": 2.9065,
+      "step": 3075
+    },
+    {
+      "epoch": 0.1301294525763601,
+      "grad_norm": 0.6228027939796448,
+      "learning_rate": 0.001,
+      "loss": 2.3924,
+      "step": 3076
+    },
+    {
+      "epoch": 0.13017175733987646,
+      "grad_norm": 0.2532864212989807,
+      "learning_rate": 0.001,
+      "loss": 2.3496,
+      "step": 3077
+    },
+    {
+      "epoch": 0.13021406210339284,
+      "grad_norm": 0.32938069105148315,
+      "learning_rate": 0.001,
+      "loss": 2.9943,
+      "step": 3078
+    },
+    {
+      "epoch": 0.13025636686690922,
+      "grad_norm": 0.7084699273109436,
+      "learning_rate": 0.001,
+      "loss": 2.3694,
+      "step": 3079
+    },
+    {
+      "epoch": 0.13029867163042558,
+      "grad_norm": 0.2795187830924988,
+      "learning_rate": 0.001,
+      "loss": 2.4931,
+      "step": 3080
+    },
+    {
+      "epoch": 0.13034097639394196,
+      "grad_norm": 2.5224292278289795,
+      "learning_rate": 0.001,
+      "loss": 2.2264,
+      "step": 3081
+    },
+    {
+      "epoch": 0.13038328115745834,
+      "grad_norm": 0.25982218980789185,
+      "learning_rate": 0.001,
+      "loss": 2.2179,
+      "step": 3082
+    },
+    {
+      "epoch": 0.1304255859209747,
+      "grad_norm": 0.35819700360298157,
+      "learning_rate": 0.001,
+      "loss": 2.5449,
+      "step": 3083
+    },
+    {
+      "epoch": 0.13046789068449108,
+      "grad_norm": 0.7076107263565063,
+      "learning_rate": 0.001,
+      "loss": 2.2326,
+      "step": 3084
+    },
+    {
+      "epoch": 0.13051019544800743,
+      "grad_norm": 0.4517766237258911,
+      "learning_rate": 0.001,
+      "loss": 3.3274,
+      "step": 3085
+    },
+    {
+      "epoch": 0.13055250021152381,
+      "grad_norm": 0.3324654698371887,
+      "learning_rate": 0.001,
+      "loss": 2.1643,
+      "step": 3086
+    },
+    {
+      "epoch": 0.1305948049750402,
+      "grad_norm": 0.2703245282173157,
+      "learning_rate": 0.001,
+      "loss": 2.2756,
+      "step": 3087
+    },
+    {
+      "epoch": 0.13063710973855655,
+      "grad_norm": 0.24688389897346497,
+      "learning_rate": 0.001,
+      "loss": 1.4894,
+      "step": 3088
+    },
+    {
+      "epoch": 0.13067941450207293,
+      "grad_norm": 0.2690267860889435,
+      "learning_rate": 0.001,
+      "loss": 2.0199,
+      "step": 3089
+    },
+    {
+      "epoch": 0.13072171926558931,
+      "grad_norm": 0.2187577337026596,
+      "learning_rate": 0.001,
+      "loss": 1.8486,
+      "step": 3090
+    },
+    {
+      "epoch": 0.13076402402910567,
+      "grad_norm": 0.3903275728225708,
+      "learning_rate": 0.001,
+      "loss": 1.9344,
+      "step": 3091
+    },
+    {
+      "epoch": 0.13080632879262205,
+      "grad_norm": 1.0155338048934937,
+      "learning_rate": 0.001,
+      "loss": 2.6314,
+      "step": 3092
+    },
+    {
+      "epoch": 0.13084863355613843,
+      "grad_norm": 2.3899335861206055,
+      "learning_rate": 0.001,
+      "loss": 2.6188,
+      "step": 3093
+    },
+    {
+      "epoch": 0.1308909383196548,
+      "grad_norm": 0.25962334871292114,
+      "learning_rate": 0.001,
+      "loss": 2.4554,
+      "step": 3094
+    },
+    {
+      "epoch": 0.13093324308317117,
+      "grad_norm": 0.4966904819011688,
+      "learning_rate": 0.001,
+      "loss": 2.3855,
+      "step": 3095
+    },
+    {
+      "epoch": 0.13097554784668752,
+      "grad_norm": 0.26274529099464417,
+      "learning_rate": 0.001,
+      "loss": 2.2674,
+      "step": 3096
+    },
+    {
+      "epoch": 0.1310178526102039,
+      "grad_norm": 0.7305838465690613,
+      "learning_rate": 0.001,
+      "loss": 3.0309,
+      "step": 3097
+    },
+    {
+      "epoch": 0.1310601573737203,
+      "grad_norm": 0.2599664330482483,
+      "learning_rate": 0.001,
+      "loss": 3.2799,
+      "step": 3098
+    },
+    {
+      "epoch": 0.13110246213723664,
+      "grad_norm": 0.3204602301120758,
+      "learning_rate": 0.001,
+      "loss": 1.795,
+      "step": 3099
+    },
+    {
+      "epoch": 0.13114476690075302,
+      "grad_norm": 0.4395667612552643,
+      "learning_rate": 0.001,
+      "loss": 2.2309,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1311870716642694,
+      "grad_norm": 0.5608950257301331,
+      "learning_rate": 0.001,
+      "loss": 3.5101,
+      "step": 3101
+    },
+    {
+      "epoch": 0.13122937642778576,
+      "grad_norm": 0.28046101331710815,
+      "learning_rate": 0.001,
+      "loss": 2.8559,
+      "step": 3102
+    },
+    {
+      "epoch": 0.13127168119130214,
+      "grad_norm": 0.7474309206008911,
+      "learning_rate": 0.001,
+      "loss": 2.1403,
+      "step": 3103
+    },
+    {
+      "epoch": 0.13131398595481852,
+      "grad_norm": 0.3261275291442871,
+      "learning_rate": 0.001,
+      "loss": 2.4168,
+      "step": 3104
+    },
+    {
+      "epoch": 0.13135629071833488,
+      "grad_norm": 0.29234766960144043,
+      "learning_rate": 0.001,
+      "loss": 1.9736,
+      "step": 3105
+    },
+    {
+      "epoch": 0.13139859548185126,
+      "grad_norm": 0.32281428575515747,
+      "learning_rate": 0.001,
+      "loss": 2.9004,
+      "step": 3106
+    },
+    {
+      "epoch": 0.13144090024536761,
+      "grad_norm": 0.3510703146457672,
+      "learning_rate": 0.001,
+      "loss": 3.3073,
+      "step": 3107
+    },
+    {
+      "epoch": 0.131483205008884,
+      "grad_norm": 0.25277993083000183,
+      "learning_rate": 0.001,
+      "loss": 1.9087,
+      "step": 3108
+    },
+    {
+      "epoch": 0.13152550977240038,
+      "grad_norm": 0.5728982090950012,
+      "learning_rate": 0.001,
+      "loss": 2.3717,
+      "step": 3109
+    },
+    {
+      "epoch": 0.13156781453591673,
+      "grad_norm": 0.2837635278701782,
+      "learning_rate": 0.001,
+      "loss": 3.2256,
+      "step": 3110
+    },
+    {
+      "epoch": 0.13161011929943311,
+      "grad_norm": 0.9751304984092712,
+      "learning_rate": 0.001,
+      "loss": 1.9881,
+      "step": 3111
+    },
+    {
+      "epoch": 0.1316524240629495,
+      "grad_norm": 0.26259645819664,
+      "learning_rate": 0.001,
+      "loss": 2.9728,
+      "step": 3112
+    },
+    {
+      "epoch": 0.13169472882646585,
+      "grad_norm": 0.28536665439605713,
+      "learning_rate": 0.001,
+      "loss": 2.4389,
+      "step": 3113
+    },
+    {
+      "epoch": 0.13173703358998223,
+      "grad_norm": 0.8838117718696594,
+      "learning_rate": 0.001,
+      "loss": 2.7865,
+      "step": 3114
+    },
+    {
+      "epoch": 0.13177933835349862,
+      "grad_norm": 0.351610392332077,
+      "learning_rate": 0.001,
+      "loss": 4.3358,
+      "step": 3115
+    },
+    {
+      "epoch": 0.13182164311701497,
+      "grad_norm": 0.6937295198440552,
+      "learning_rate": 0.001,
+      "loss": 3.2516,
+      "step": 3116
+    },
+    {
+      "epoch": 0.13186394788053135,
+      "grad_norm": 0.24456751346588135,
+      "learning_rate": 0.001,
+      "loss": 3.5256,
+      "step": 3117
+    },
+    {
+      "epoch": 0.13190625264404773,
+      "grad_norm": 1.5360091924667358,
+      "learning_rate": 0.001,
+      "loss": 2.8469,
+      "step": 3118
+    },
+    {
+      "epoch": 0.1319485574075641,
+      "grad_norm": 0.3377660810947418,
+      "learning_rate": 0.001,
+      "loss": 2.5228,
+      "step": 3119
+    },
+    {
+      "epoch": 0.13199086217108047,
+      "grad_norm": 0.3961987793445587,
+      "learning_rate": 0.001,
+      "loss": 2.7447,
+      "step": 3120
+    },
+    {
+      "epoch": 0.13203316693459682,
+      "grad_norm": 0.32120051980018616,
+      "learning_rate": 0.001,
+      "loss": 2.3102,
+      "step": 3121
+    },
+    {
+      "epoch": 0.1320754716981132,
+      "grad_norm": 0.3845050036907196,
+      "learning_rate": 0.001,
+      "loss": 3.7081,
+      "step": 3122
+    },
+    {
+      "epoch": 0.1321177764616296,
+      "grad_norm": 0.2517256736755371,
+      "learning_rate": 0.001,
+      "loss": 2.6708,
+      "step": 3123
+    },
+    {
+      "epoch": 0.13216008122514594,
+      "grad_norm": 0.25273680686950684,
+      "learning_rate": 0.001,
+      "loss": 2.748,
+      "step": 3124
+    },
+    {
+      "epoch": 0.13220238598866232,
+      "grad_norm": 0.3618360161781311,
+      "learning_rate": 0.001,
+      "loss": 3.418,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1322446907521787,
+      "grad_norm": 0.4369431436061859,
+      "learning_rate": 0.001,
+      "loss": 3.6098,
+      "step": 3126
+    },
+    {
+      "epoch": 0.13228699551569506,
+      "grad_norm": 0.28894707560539246,
+      "learning_rate": 0.001,
+      "loss": 1.7968,
+      "step": 3127
+    },
+    {
+      "epoch": 0.13232930027921144,
+      "grad_norm": 0.29500776529312134,
+      "learning_rate": 0.001,
+      "loss": 2.6331,
+      "step": 3128
+    },
+    {
+      "epoch": 0.13237160504272782,
+      "grad_norm": 0.8470263481140137,
+      "learning_rate": 0.001,
+      "loss": 2.9434,
+      "step": 3129
+    },
+    {
+      "epoch": 0.13241390980624418,
+      "grad_norm": 0.20783580839633942,
+      "learning_rate": 0.001,
+      "loss": 1.635,
+      "step": 3130
+    },
+    {
+      "epoch": 0.13245621456976056,
+      "grad_norm": 0.4847624897956848,
+      "learning_rate": 0.001,
+      "loss": 2.6969,
+      "step": 3131
+    },
+    {
+      "epoch": 0.13249851933327692,
+      "grad_norm": 0.2904376983642578,
+      "learning_rate": 0.001,
+      "loss": 3.1772,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1325408240967933,
+      "grad_norm": 0.29514384269714355,
+      "learning_rate": 0.001,
+      "loss": 2.2651,
+      "step": 3133
+    },
+    {
+      "epoch": 0.13258312886030968,
+      "grad_norm": 0.3028735816478729,
+      "learning_rate": 0.001,
+      "loss": 2.2932,
+      "step": 3134
+    },
+    {
+      "epoch": 0.13262543362382603,
+      "grad_norm": 0.2190832495689392,
+      "learning_rate": 0.001,
+      "loss": 2.291,
+      "step": 3135
+    },
+    {
+      "epoch": 0.13266773838734242,
+      "grad_norm": 0.4608692526817322,
+      "learning_rate": 0.001,
+      "loss": 2.9313,
+      "step": 3136
+    },
+    {
+      "epoch": 0.1327100431508588,
+      "grad_norm": 0.5700821280479431,
+      "learning_rate": 0.001,
+      "loss": 2.6597,
+      "step": 3137
+    },
+    {
+      "epoch": 0.13275234791437515,
+      "grad_norm": 0.41855689883232117,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 3138
+    },
+    {
+      "epoch": 0.13279465267789153,
+      "grad_norm": 0.2460172474384308,
+      "learning_rate": 0.001,
+      "loss": 2.0043,
+      "step": 3139
+    },
+    {
+      "epoch": 0.13283695744140792,
+      "grad_norm": 0.26085442304611206,
+      "learning_rate": 0.001,
+      "loss": 2.8132,
+      "step": 3140
+    },
+    {
+      "epoch": 0.13287926220492427,
+      "grad_norm": 0.27267521619796753,
+      "learning_rate": 0.001,
+      "loss": 1.9265,
+      "step": 3141
+    },
+    {
+      "epoch": 0.13292156696844065,
+      "grad_norm": 0.4825087785720825,
+      "learning_rate": 0.001,
+      "loss": 2.5328,
+      "step": 3142
+    },
+    {
+      "epoch": 0.132963871731957,
+      "grad_norm": 0.7955046892166138,
+      "learning_rate": 0.001,
+      "loss": 2.4661,
+      "step": 3143
+    },
+    {
+      "epoch": 0.1330061764954734,
+      "grad_norm": 0.22219106554985046,
+      "learning_rate": 0.001,
+      "loss": 2.8113,
+      "step": 3144
+    },
+    {
+      "epoch": 0.13304848125898977,
+      "grad_norm": 0.20542827248573303,
+      "learning_rate": 0.001,
+      "loss": 1.5933,
+      "step": 3145
+    },
+    {
+      "epoch": 0.13309078602250612,
+      "grad_norm": 1.0127766132354736,
+      "learning_rate": 0.001,
+      "loss": 2.4128,
+      "step": 3146
+    },
+    {
+      "epoch": 0.1331330907860225,
+      "grad_norm": 0.7383486032485962,
+      "learning_rate": 0.001,
+      "loss": 2.2777,
+      "step": 3147
+    },
+    {
+      "epoch": 0.1331753955495389,
+      "grad_norm": 0.714616596698761,
+      "learning_rate": 0.001,
+      "loss": 2.5478,
+      "step": 3148
+    },
+    {
+      "epoch": 0.13321770031305524,
+      "grad_norm": 0.4222099781036377,
+      "learning_rate": 0.001,
+      "loss": 2.4327,
+      "step": 3149
+    },
+    {
+      "epoch": 0.13326000507657162,
+      "grad_norm": 0.3047798275947571,
+      "learning_rate": 0.001,
+      "loss": 1.8279,
+      "step": 3150
+    },
+    {
+      "epoch": 0.133302309840088,
+      "grad_norm": 32.82894515991211,
+      "learning_rate": 0.001,
+      "loss": 2.1241,
+      "step": 3151
+    },
+    {
+      "epoch": 0.13334461460360436,
+      "grad_norm": 4.569951057434082,
+      "learning_rate": 0.001,
+      "loss": 2.4177,
+      "step": 3152
+    },
+    {
+      "epoch": 0.13338691936712074,
+      "grad_norm": 0.2539486587047577,
+      "learning_rate": 0.001,
+      "loss": 2.6896,
+      "step": 3153
+    },
+    {
+      "epoch": 0.1334292241306371,
+      "grad_norm": 0.3113529086112976,
+      "learning_rate": 0.001,
+      "loss": 2.3834,
+      "step": 3154
+    },
+    {
+      "epoch": 0.13347152889415348,
+      "grad_norm": 0.5559782981872559,
+      "learning_rate": 0.001,
+      "loss": 3.521,
+      "step": 3155
+    },
+    {
+      "epoch": 0.13351383365766986,
+      "grad_norm": 0.3125370144844055,
+      "learning_rate": 0.001,
+      "loss": 2.5982,
+      "step": 3156
+    },
+    {
+      "epoch": 0.13355613842118622,
+      "grad_norm": 0.2650197744369507,
+      "learning_rate": 0.001,
+      "loss": 2.1565,
+      "step": 3157
+    },
+    {
+      "epoch": 0.1335984431847026,
+      "grad_norm": 0.3328724503517151,
+      "learning_rate": 0.001,
+      "loss": 2.0653,
+      "step": 3158
+    },
+    {
+      "epoch": 0.13364074794821898,
+      "grad_norm": 0.2736479341983795,
+      "learning_rate": 0.001,
+      "loss": 3.6324,
+      "step": 3159
+    },
+    {
+      "epoch": 0.13368305271173533,
+      "grad_norm": 0.3234842121601105,
+      "learning_rate": 0.001,
+      "loss": 2.1484,
+      "step": 3160
+    },
+    {
+      "epoch": 0.13372535747525172,
+      "grad_norm": 0.2782366871833801,
+      "learning_rate": 0.001,
+      "loss": 4.0075,
+      "step": 3161
+    },
+    {
+      "epoch": 0.1337676622387681,
+      "grad_norm": 0.2664390504360199,
+      "learning_rate": 0.001,
+      "loss": 1.7258,
+      "step": 3162
+    },
+    {
+      "epoch": 0.13380996700228445,
+      "grad_norm": 0.23455047607421875,
+      "learning_rate": 0.001,
+      "loss": 1.9677,
+      "step": 3163
+    },
+    {
+      "epoch": 0.13385227176580083,
+      "grad_norm": 0.22314786911010742,
+      "learning_rate": 0.001,
+      "loss": 2.3173,
+      "step": 3164
+    },
+    {
+      "epoch": 0.1338945765293172,
+      "grad_norm": 0.2761704921722412,
+      "learning_rate": 0.001,
+      "loss": 1.8873,
+      "step": 3165
+    },
+    {
+      "epoch": 0.13393688129283357,
+      "grad_norm": 0.6037988066673279,
+      "learning_rate": 0.001,
+      "loss": 2.6251,
+      "step": 3166
+    },
+    {
+      "epoch": 0.13397918605634995,
+      "grad_norm": 0.42490071058273315,
+      "learning_rate": 0.001,
+      "loss": 2.1205,
+      "step": 3167
+    },
+    {
+      "epoch": 0.1340214908198663,
+      "grad_norm": 0.27945172786712646,
+      "learning_rate": 0.001,
+      "loss": 2.0086,
+      "step": 3168
+    },
+    {
+      "epoch": 0.1340637955833827,
+      "grad_norm": 0.2539001703262329,
+      "learning_rate": 0.001,
+      "loss": 2.1604,
+      "step": 3169
+    },
+    {
+      "epoch": 0.13410610034689907,
+      "grad_norm": 0.23439648747444153,
+      "learning_rate": 0.001,
+      "loss": 2.4407,
+      "step": 3170
+    },
+    {
+      "epoch": 0.13414840511041543,
+      "grad_norm": 0.3301600515842438,
+      "learning_rate": 0.001,
+      "loss": 2.0621,
+      "step": 3171
+    },
+    {
+      "epoch": 0.1341907098739318,
+      "grad_norm": 2.253829002380371,
+      "learning_rate": 0.001,
+      "loss": 2.1485,
+      "step": 3172
+    },
+    {
+      "epoch": 0.1342330146374482,
+      "grad_norm": 0.34564846754074097,
+      "learning_rate": 0.001,
+      "loss": 2.2515,
+      "step": 3173
+    },
+    {
+      "epoch": 0.13427531940096454,
+      "grad_norm": 0.269344300031662,
+      "learning_rate": 0.001,
+      "loss": 3.2911,
+      "step": 3174
+    },
+    {
+      "epoch": 0.13431762416448093,
+      "grad_norm": 0.2677054703235626,
+      "learning_rate": 0.001,
+      "loss": 1.9107,
+      "step": 3175
+    },
+    {
+      "epoch": 0.13435992892799728,
+      "grad_norm": 1.6993762254714966,
+      "learning_rate": 0.001,
+      "loss": 2.3414,
+      "step": 3176
+    },
+    {
+      "epoch": 0.13440223369151366,
+      "grad_norm": 0.29692259430885315,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 3177
+    },
+    {
+      "epoch": 0.13444453845503004,
+      "grad_norm": 0.24042674899101257,
+      "learning_rate": 0.001,
+      "loss": 2.9594,
+      "step": 3178
+    },
+    {
+      "epoch": 0.1344868432185464,
+      "grad_norm": 0.7225860357284546,
+      "learning_rate": 0.001,
+      "loss": 2.6567,
+      "step": 3179
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 0.24381719529628754,
+      "learning_rate": 0.001,
+      "loss": 2.0548,
+      "step": 3180
+    },
+    {
+      "epoch": 0.13457145274557916,
+      "grad_norm": 0.2055826634168625,
+      "learning_rate": 0.001,
+      "loss": 2.3845,
+      "step": 3181
+    },
+    {
+      "epoch": 0.13461375750909552,
+      "grad_norm": 0.2573340833187103,
+      "learning_rate": 0.001,
+      "loss": 2.0984,
+      "step": 3182
+    },
+    {
+      "epoch": 0.1346560622726119,
+      "grad_norm": 0.2962018847465515,
+      "learning_rate": 0.001,
+      "loss": 2.5636,
+      "step": 3183
+    },
+    {
+      "epoch": 0.13469836703612828,
+      "grad_norm": 0.40545910596847534,
+      "learning_rate": 0.001,
+      "loss": 2.3188,
+      "step": 3184
+    },
+    {
+      "epoch": 0.13474067179964463,
+      "grad_norm": 0.6234009861946106,
+      "learning_rate": 0.001,
+      "loss": 1.9832,
+      "step": 3185
+    },
+    {
+      "epoch": 0.13478297656316102,
+      "grad_norm": 0.9001304507255554,
+      "learning_rate": 0.001,
+      "loss": 2.0331,
+      "step": 3186
+    },
+    {
+      "epoch": 0.13482528132667737,
+      "grad_norm": 3.3280558586120605,
+      "learning_rate": 0.001,
+      "loss": 2.8476,
+      "step": 3187
+    },
+    {
+      "epoch": 0.13486758609019375,
+      "grad_norm": 0.25616398453712463,
+      "learning_rate": 0.001,
+      "loss": 2.8065,
+      "step": 3188
+    },
+    {
+      "epoch": 0.13490989085371013,
+      "grad_norm": 4.6577467918396,
+      "learning_rate": 0.001,
+      "loss": 2.244,
+      "step": 3189
+    },
+    {
+      "epoch": 0.1349521956172265,
+      "grad_norm": 0.37131521105766296,
+      "learning_rate": 0.001,
+      "loss": 2.2248,
+      "step": 3190
+    },
+    {
+      "epoch": 0.13499450038074287,
+      "grad_norm": 0.4129955768585205,
+      "learning_rate": 0.001,
+      "loss": 2.701,
+      "step": 3191
+    },
+    {
+      "epoch": 0.13503680514425925,
+      "grad_norm": 0.24256403744220734,
+      "learning_rate": 0.001,
+      "loss": 2.0321,
+      "step": 3192
+    },
+    {
+      "epoch": 0.1350791099077756,
+      "grad_norm": 0.4478329122066498,
+      "learning_rate": 0.001,
+      "loss": 2.888,
+      "step": 3193
+    },
+    {
+      "epoch": 0.135121414671292,
+      "grad_norm": 0.24205787479877472,
+      "learning_rate": 0.001,
+      "loss": 2.9618,
+      "step": 3194
+    },
+    {
+      "epoch": 0.13516371943480837,
+      "grad_norm": 0.2591802179813385,
+      "learning_rate": 0.001,
+      "loss": 2.3314,
+      "step": 3195
+    },
+    {
+      "epoch": 0.13520602419832473,
+      "grad_norm": 0.26308709383010864,
+      "learning_rate": 0.001,
+      "loss": 1.5799,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1352483289618411,
+      "grad_norm": 2.3362512588500977,
+      "learning_rate": 0.001,
+      "loss": 2.72,
+      "step": 3197
+    },
+    {
+      "epoch": 0.13529063372535746,
+      "grad_norm": 0.25793129205703735,
+      "learning_rate": 0.001,
+      "loss": 3.2309,
+      "step": 3198
+    },
+    {
+      "epoch": 0.13533293848887384,
+      "grad_norm": 0.24646157026290894,
+      "learning_rate": 0.001,
+      "loss": 2.5992,
+      "step": 3199
+    },
+    {
+      "epoch": 0.13537524325239023,
+      "grad_norm": 1.71589195728302,
+      "learning_rate": 0.001,
+      "loss": 2.0522,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13541754801590658,
+      "grad_norm": 0.24351108074188232,
+      "learning_rate": 0.001,
+      "loss": 1.7597,
+      "step": 3201
+    },
+    {
+      "epoch": 0.13545985277942296,
+      "grad_norm": 0.23399876058101654,
+      "learning_rate": 0.001,
+      "loss": 3.4004,
+      "step": 3202
+    },
+    {
+      "epoch": 0.13550215754293934,
+      "grad_norm": 0.2488681972026825,
+      "learning_rate": 0.001,
+      "loss": 3.0523,
+      "step": 3203
+    },
+    {
+      "epoch": 0.1355444623064557,
+      "grad_norm": 0.276132732629776,
+      "learning_rate": 0.001,
+      "loss": 3.5311,
+      "step": 3204
+    },
+    {
+      "epoch": 0.13558676706997208,
+      "grad_norm": 0.2650410830974579,
+      "learning_rate": 0.001,
+      "loss": 2.3133,
+      "step": 3205
+    },
+    {
+      "epoch": 0.13562907183348846,
+      "grad_norm": 8.522130966186523,
+      "learning_rate": 0.001,
+      "loss": 2.3636,
+      "step": 3206
+    },
+    {
+      "epoch": 0.13567137659700482,
+      "grad_norm": 0.26246312260627747,
+      "learning_rate": 0.001,
+      "loss": 2.2168,
+      "step": 3207
+    },
+    {
+      "epoch": 0.1357136813605212,
+      "grad_norm": 0.29150646924972534,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 3208
+    },
+    {
+      "epoch": 0.13575598612403755,
+      "grad_norm": 0.2513675093650818,
+      "learning_rate": 0.001,
+      "loss": 2.0227,
+      "step": 3209
+    },
+    {
+      "epoch": 0.13579829088755394,
+      "grad_norm": 0.274081289768219,
+      "learning_rate": 0.001,
+      "loss": 3.2165,
+      "step": 3210
+    },
+    {
+      "epoch": 0.13584059565107032,
+      "grad_norm": 0.7095363736152649,
+      "learning_rate": 0.001,
+      "loss": 2.1727,
+      "step": 3211
+    },
+    {
+      "epoch": 0.13588290041458667,
+      "grad_norm": 0.24826420843601227,
+      "learning_rate": 0.001,
+      "loss": 2.9115,
+      "step": 3212
+    },
+    {
+      "epoch": 0.13592520517810305,
+      "grad_norm": 0.2485196441411972,
+      "learning_rate": 0.001,
+      "loss": 2.705,
+      "step": 3213
+    },
+    {
+      "epoch": 0.13596750994161944,
+      "grad_norm": 1.0916763544082642,
+      "learning_rate": 0.001,
+      "loss": 2.5535,
+      "step": 3214
+    },
+    {
+      "epoch": 0.1360098147051358,
+      "grad_norm": 0.29727303981781006,
+      "learning_rate": 0.001,
+      "loss": 3.2918,
+      "step": 3215
+    },
+    {
+      "epoch": 0.13605211946865217,
+      "grad_norm": 0.24911724030971527,
+      "learning_rate": 0.001,
+      "loss": 2.0456,
+      "step": 3216
+    },
+    {
+      "epoch": 0.13609442423216855,
+      "grad_norm": 0.2424779087305069,
+      "learning_rate": 0.001,
+      "loss": 2.466,
+      "step": 3217
+    },
+    {
+      "epoch": 0.1361367289956849,
+      "grad_norm": 0.23129363358020782,
+      "learning_rate": 0.001,
+      "loss": 2.5406,
+      "step": 3218
+    },
+    {
+      "epoch": 0.1361790337592013,
+      "grad_norm": 0.2544875741004944,
+      "learning_rate": 0.001,
+      "loss": 2.7152,
+      "step": 3219
+    },
+    {
+      "epoch": 0.13622133852271764,
+      "grad_norm": 0.28286901116371155,
+      "learning_rate": 0.001,
+      "loss": 2.7993,
+      "step": 3220
+    },
+    {
+      "epoch": 0.13626364328623403,
+      "grad_norm": 0.3165169060230255,
+      "learning_rate": 0.001,
+      "loss": 1.8458,
+      "step": 3221
+    },
+    {
+      "epoch": 0.1363059480497504,
+      "grad_norm": 0.2771192491054535,
+      "learning_rate": 0.001,
+      "loss": 2.2885,
+      "step": 3222
+    },
+    {
+      "epoch": 0.13634825281326676,
+      "grad_norm": 0.2702016830444336,
+      "learning_rate": 0.001,
+      "loss": 2.1035,
+      "step": 3223
+    },
+    {
+      "epoch": 0.13639055757678314,
+      "grad_norm": 0.20894509553909302,
+      "learning_rate": 0.001,
+      "loss": 2.4901,
+      "step": 3224
+    },
+    {
+      "epoch": 0.13643286234029953,
+      "grad_norm": 0.4794827997684479,
+      "learning_rate": 0.001,
+      "loss": 2.3976,
+      "step": 3225
+    },
+    {
+      "epoch": 0.13647516710381588,
+      "grad_norm": 0.3762035667896271,
+      "learning_rate": 0.001,
+      "loss": 2.3773,
+      "step": 3226
+    },
+    {
+      "epoch": 0.13651747186733226,
+      "grad_norm": 0.9282992482185364,
+      "learning_rate": 0.001,
+      "loss": 2.3559,
+      "step": 3227
+    },
+    {
+      "epoch": 0.13655977663084864,
+      "grad_norm": 0.24388918280601501,
+      "learning_rate": 0.001,
+      "loss": 2.6357,
+      "step": 3228
+    },
+    {
+      "epoch": 0.136602081394365,
+      "grad_norm": 0.22450682520866394,
+      "learning_rate": 0.001,
+      "loss": 3.0694,
+      "step": 3229
+    },
+    {
+      "epoch": 0.13664438615788138,
+      "grad_norm": 0.22949005663394928,
+      "learning_rate": 0.001,
+      "loss": 2.4709,
+      "step": 3230
+    },
+    {
+      "epoch": 0.13668669092139776,
+      "grad_norm": 0.2965823709964752,
+      "learning_rate": 0.001,
+      "loss": 2.0617,
+      "step": 3231
+    },
+    {
+      "epoch": 0.13672899568491412,
+      "grad_norm": 0.6691174507141113,
+      "learning_rate": 0.001,
+      "loss": 2.4596,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1367713004484305,
+      "grad_norm": 0.21386466920375824,
+      "learning_rate": 0.001,
+      "loss": 2.3095,
+      "step": 3233
+    },
+    {
+      "epoch": 0.13681360521194685,
+      "grad_norm": 0.3088574707508087,
+      "learning_rate": 0.001,
+      "loss": 2.1195,
+      "step": 3234
+    },
+    {
+      "epoch": 0.13685590997546324,
+      "grad_norm": 0.23426178097724915,
+      "learning_rate": 0.001,
+      "loss": 2.036,
+      "step": 3235
+    },
+    {
+      "epoch": 0.13689821473897962,
+      "grad_norm": 0.2373049259185791,
+      "learning_rate": 0.001,
+      "loss": 2.3527,
+      "step": 3236
+    },
+    {
+      "epoch": 0.13694051950249597,
+      "grad_norm": 0.253162682056427,
+      "learning_rate": 0.001,
+      "loss": 1.8557,
+      "step": 3237
+    },
+    {
+      "epoch": 0.13698282426601235,
+      "grad_norm": 0.23162192106246948,
+      "learning_rate": 0.001,
+      "loss": 2.3313,
+      "step": 3238
+    },
+    {
+      "epoch": 0.13702512902952874,
+      "grad_norm": 0.20582619309425354,
+      "learning_rate": 0.001,
+      "loss": 2.1614,
+      "step": 3239
+    },
+    {
+      "epoch": 0.1370674337930451,
+      "grad_norm": 0.23828250169754028,
+      "learning_rate": 0.001,
+      "loss": 2.4971,
+      "step": 3240
+    },
+    {
+      "epoch": 0.13710973855656147,
+      "grad_norm": 0.2543167471885681,
+      "learning_rate": 0.001,
+      "loss": 2.4651,
+      "step": 3241
+    },
+    {
+      "epoch": 0.13715204332007785,
+      "grad_norm": 1.886708378791809,
+      "learning_rate": 0.001,
+      "loss": 2.7207,
+      "step": 3242
+    },
+    {
+      "epoch": 0.1371943480835942,
+      "grad_norm": 0.2185901701450348,
+      "learning_rate": 0.001,
+      "loss": 1.9575,
+      "step": 3243
+    },
+    {
+      "epoch": 0.1372366528471106,
+      "grad_norm": 0.25897449254989624,
+      "learning_rate": 0.001,
+      "loss": 3.6101,
+      "step": 3244
+    },
+    {
+      "epoch": 0.13727895761062694,
+      "grad_norm": 0.22902828454971313,
+      "learning_rate": 0.001,
+      "loss": 2.5504,
+      "step": 3245
+    },
+    {
+      "epoch": 0.13732126237414333,
+      "grad_norm": 0.9940201044082642,
+      "learning_rate": 0.001,
+      "loss": 2.4984,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1373635671376597,
+      "grad_norm": 0.5756583213806152,
+      "learning_rate": 0.001,
+      "loss": 2.0633,
+      "step": 3247
+    },
+    {
+      "epoch": 0.13740587190117606,
+      "grad_norm": 0.24829471111297607,
+      "learning_rate": 0.001,
+      "loss": 2.2558,
+      "step": 3248
+    },
+    {
+      "epoch": 0.13744817666469245,
+      "grad_norm": 0.2848198413848877,
+      "learning_rate": 0.001,
+      "loss": 2.2559,
+      "step": 3249
+    },
+    {
+      "epoch": 0.13749048142820883,
+      "grad_norm": 0.2882290184497833,
+      "learning_rate": 0.001,
+      "loss": 2.2497,
+      "step": 3250
+    },
+    {
+      "epoch": 0.13753278619172518,
+      "grad_norm": 0.3670541048049927,
+      "learning_rate": 0.001,
+      "loss": 2.8339,
+      "step": 3251
+    },
+    {
+      "epoch": 0.13757509095524156,
+      "grad_norm": 0.2615547180175781,
+      "learning_rate": 0.001,
+      "loss": 2.7191,
+      "step": 3252
+    },
+    {
+      "epoch": 0.13761739571875795,
+      "grad_norm": 0.2824948728084564,
+      "learning_rate": 0.001,
+      "loss": 2.4394,
+      "step": 3253
+    },
+    {
+      "epoch": 0.1376597004822743,
+      "grad_norm": 0.2855933904647827,
+      "learning_rate": 0.001,
+      "loss": 1.7882,
+      "step": 3254
+    },
+    {
+      "epoch": 0.13770200524579068,
+      "grad_norm": 0.22275911271572113,
+      "learning_rate": 0.001,
+      "loss": 2.8593,
+      "step": 3255
+    },
+    {
+      "epoch": 0.13774431000930704,
+      "grad_norm": 0.29241636395454407,
+      "learning_rate": 0.001,
+      "loss": 2.4057,
+      "step": 3256
+    },
+    {
+      "epoch": 0.13778661477282342,
+      "grad_norm": 0.2222166359424591,
+      "learning_rate": 0.001,
+      "loss": 2.4944,
+      "step": 3257
+    },
+    {
+      "epoch": 0.1378289195363398,
+      "grad_norm": 0.23610618710517883,
+      "learning_rate": 0.001,
+      "loss": 2.5925,
+      "step": 3258
+    },
+    {
+      "epoch": 0.13787122429985615,
+      "grad_norm": 0.25960981845855713,
+      "learning_rate": 0.001,
+      "loss": 2.085,
+      "step": 3259
+    },
+    {
+      "epoch": 0.13791352906337254,
+      "grad_norm": 0.23140114545822144,
+      "learning_rate": 0.001,
+      "loss": 2.1499,
+      "step": 3260
+    },
+    {
+      "epoch": 0.13795583382688892,
+      "grad_norm": 1.9852933883666992,
+      "learning_rate": 0.001,
+      "loss": 2.7648,
+      "step": 3261
+    },
+    {
+      "epoch": 0.13799813859040527,
+      "grad_norm": 0.27138063311576843,
+      "learning_rate": 0.001,
+      "loss": 2.131,
+      "step": 3262
+    },
+    {
+      "epoch": 0.13804044335392165,
+      "grad_norm": 0.23058192431926727,
+      "learning_rate": 0.001,
+      "loss": 1.905,
+      "step": 3263
+    },
+    {
+      "epoch": 0.13808274811743804,
+      "grad_norm": 0.360061377286911,
+      "learning_rate": 0.001,
+      "loss": 1.958,
+      "step": 3264
+    },
+    {
+      "epoch": 0.1381250528809544,
+      "grad_norm": 0.2711131274700165,
+      "learning_rate": 0.001,
+      "loss": 2.4258,
+      "step": 3265
+    },
+    {
+      "epoch": 0.13816735764447077,
+      "grad_norm": 0.30782273411750793,
+      "learning_rate": 0.001,
+      "loss": 3.0677,
+      "step": 3266
+    },
+    {
+      "epoch": 0.13820966240798713,
+      "grad_norm": 0.2656923234462738,
+      "learning_rate": 0.001,
+      "loss": 1.8376,
+      "step": 3267
+    },
+    {
+      "epoch": 0.1382519671715035,
+      "grad_norm": 0.24168632924556732,
+      "learning_rate": 0.001,
+      "loss": 1.9629,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1382942719350199,
+      "grad_norm": 0.34883561730384827,
+      "learning_rate": 0.001,
+      "loss": 3.1204,
+      "step": 3269
+    },
+    {
+      "epoch": 0.13833657669853625,
+      "grad_norm": 0.25927019119262695,
+      "learning_rate": 0.001,
+      "loss": 2.9234,
+      "step": 3270
+    },
+    {
+      "epoch": 0.13837888146205263,
+      "grad_norm": 0.38157638907432556,
+      "learning_rate": 0.001,
+      "loss": 2.7691,
+      "step": 3271
+    },
+    {
+      "epoch": 0.138421186225569,
+      "grad_norm": 0.4185699224472046,
+      "learning_rate": 0.001,
+      "loss": 2.306,
+      "step": 3272
+    },
+    {
+      "epoch": 0.13846349098908536,
+      "grad_norm": 0.22628718614578247,
+      "learning_rate": 0.001,
+      "loss": 2.0184,
+      "step": 3273
+    },
+    {
+      "epoch": 0.13850579575260175,
+      "grad_norm": 0.25233402848243713,
+      "learning_rate": 0.001,
+      "loss": 2.845,
+      "step": 3274
+    },
+    {
+      "epoch": 0.13854810051611813,
+      "grad_norm": 0.26197144389152527,
+      "learning_rate": 0.001,
+      "loss": 1.7947,
+      "step": 3275
+    },
+    {
+      "epoch": 0.13859040527963448,
+      "grad_norm": 0.38328817486763,
+      "learning_rate": 0.001,
+      "loss": 2.4887,
+      "step": 3276
+    },
+    {
+      "epoch": 0.13863271004315086,
+      "grad_norm": 0.2760562300682068,
+      "learning_rate": 0.001,
+      "loss": 2.0466,
+      "step": 3277
+    },
+    {
+      "epoch": 0.13867501480666722,
+      "grad_norm": 0.2207472026348114,
+      "learning_rate": 0.001,
+      "loss": 2.2844,
+      "step": 3278
+    },
+    {
+      "epoch": 0.1387173195701836,
+      "grad_norm": 0.24221129715442657,
+      "learning_rate": 0.001,
+      "loss": 2.7634,
+      "step": 3279
+    },
+    {
+      "epoch": 0.13875962433369998,
+      "grad_norm": 0.23639145493507385,
+      "learning_rate": 0.001,
+      "loss": 2.3057,
+      "step": 3280
+    },
+    {
+      "epoch": 0.13880192909721634,
+      "grad_norm": 0.6643745303153992,
+      "learning_rate": 0.001,
+      "loss": 2.1484,
+      "step": 3281
+    },
+    {
+      "epoch": 0.13884423386073272,
+      "grad_norm": 0.29646036028862,
+      "learning_rate": 0.001,
+      "loss": 1.7328,
+      "step": 3282
+    },
+    {
+      "epoch": 0.1388865386242491,
+      "grad_norm": 0.7454001307487488,
+      "learning_rate": 0.001,
+      "loss": 2.0942,
+      "step": 3283
+    },
+    {
+      "epoch": 0.13892884338776545,
+      "grad_norm": 1.1968051195144653,
+      "learning_rate": 0.001,
+      "loss": 2.9441,
+      "step": 3284
+    },
+    {
+      "epoch": 0.13897114815128184,
+      "grad_norm": 0.23698095977306366,
+      "learning_rate": 0.001,
+      "loss": 2.6655,
+      "step": 3285
+    },
+    {
+      "epoch": 0.13901345291479822,
+      "grad_norm": 0.6837499141693115,
+      "learning_rate": 0.001,
+      "loss": 2.2513,
+      "step": 3286
+    },
+    {
+      "epoch": 0.13905575767831457,
+      "grad_norm": 0.48883894085884094,
+      "learning_rate": 0.001,
+      "loss": 2.6241,
+      "step": 3287
+    },
+    {
+      "epoch": 0.13909806244183096,
+      "grad_norm": 0.5206930637359619,
+      "learning_rate": 0.001,
+      "loss": 2.2263,
+      "step": 3288
+    },
+    {
+      "epoch": 0.1391403672053473,
+      "grad_norm": 0.40440839529037476,
+      "learning_rate": 0.001,
+      "loss": 3.109,
+      "step": 3289
+    },
+    {
+      "epoch": 0.1391826719688637,
+      "grad_norm": 1.5446665287017822,
+      "learning_rate": 0.001,
+      "loss": 2.3798,
+      "step": 3290
+    },
+    {
+      "epoch": 0.13922497673238007,
+      "grad_norm": 0.6700052618980408,
+      "learning_rate": 0.001,
+      "loss": 1.8315,
+      "step": 3291
+    },
+    {
+      "epoch": 0.13926728149589643,
+      "grad_norm": 0.5063325762748718,
+      "learning_rate": 0.001,
+      "loss": 2.6604,
+      "step": 3292
+    },
+    {
+      "epoch": 0.1393095862594128,
+      "grad_norm": 0.38244378566741943,
+      "learning_rate": 0.001,
+      "loss": 2.4225,
+      "step": 3293
+    },
+    {
+      "epoch": 0.1393518910229292,
+      "grad_norm": 0.3094440996646881,
+      "learning_rate": 0.001,
+      "loss": 2.6557,
+      "step": 3294
+    },
+    {
+      "epoch": 0.13939419578644555,
+      "grad_norm": 0.2953416109085083,
+      "learning_rate": 0.001,
+      "loss": 2.1801,
+      "step": 3295
+    },
+    {
+      "epoch": 0.13943650054996193,
+      "grad_norm": 0.2831237018108368,
+      "learning_rate": 0.001,
+      "loss": 2.5021,
+      "step": 3296
+    },
+    {
+      "epoch": 0.1394788053134783,
+      "grad_norm": 14.991360664367676,
+      "learning_rate": 0.001,
+      "loss": 3.954,
+      "step": 3297
+    },
+    {
+      "epoch": 0.13952111007699466,
+      "grad_norm": 0.26766881346702576,
+      "learning_rate": 0.001,
+      "loss": 2.3572,
+      "step": 3298
+    },
+    {
+      "epoch": 0.13956341484051105,
+      "grad_norm": 0.4463373124599457,
+      "learning_rate": 0.001,
+      "loss": 2.2735,
+      "step": 3299
+    },
+    {
+      "epoch": 0.1396057196040274,
+      "grad_norm": 0.30204683542251587,
+      "learning_rate": 0.001,
+      "loss": 2.7257,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13964802436754378,
+      "grad_norm": 0.34700825810432434,
+      "learning_rate": 0.001,
+      "loss": 3.2625,
+      "step": 3301
+    },
+    {
+      "epoch": 0.13969032913106016,
+      "grad_norm": 0.348042756319046,
+      "learning_rate": 0.001,
+      "loss": 2.5979,
+      "step": 3302
+    },
+    {
+      "epoch": 0.13973263389457652,
+      "grad_norm": 0.38788822293281555,
+      "learning_rate": 0.001,
+      "loss": 2.5037,
+      "step": 3303
+    },
+    {
+      "epoch": 0.1397749386580929,
+      "grad_norm": 0.24049419164657593,
+      "learning_rate": 0.001,
+      "loss": 2.2218,
+      "step": 3304
+    },
+    {
+      "epoch": 0.13981724342160928,
+      "grad_norm": 0.2679419219493866,
+      "learning_rate": 0.001,
+      "loss": 2.8018,
+      "step": 3305
+    },
+    {
+      "epoch": 0.13985954818512564,
+      "grad_norm": 0.2681131064891815,
+      "learning_rate": 0.001,
+      "loss": 3.2434,
+      "step": 3306
+    },
+    {
+      "epoch": 0.13990185294864202,
+      "grad_norm": 0.25279781222343445,
+      "learning_rate": 0.001,
+      "loss": 2.6153,
+      "step": 3307
+    },
+    {
+      "epoch": 0.1399441577121584,
+      "grad_norm": 0.23080165684223175,
+      "learning_rate": 0.001,
+      "loss": 2.5528,
+      "step": 3308
+    },
+    {
+      "epoch": 0.13998646247567476,
+      "grad_norm": 0.21056526899337769,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 3309
+    },
+    {
+      "epoch": 0.14002876723919114,
+      "grad_norm": 1.0948318243026733,
+      "learning_rate": 0.001,
+      "loss": 1.8427,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1400710720027075,
+      "grad_norm": 0.27404388785362244,
+      "learning_rate": 0.001,
+      "loss": 1.971,
+      "step": 3311
+    },
+    {
+      "epoch": 0.14011337676622387,
+      "grad_norm": 0.23093107342720032,
+      "learning_rate": 0.001,
+      "loss": 2.145,
+      "step": 3312
+    },
+    {
+      "epoch": 0.14015568152974026,
+      "grad_norm": 0.8817729949951172,
+      "learning_rate": 0.001,
+      "loss": 2.4664,
+      "step": 3313
+    },
+    {
+      "epoch": 0.1401979862932566,
+      "grad_norm": 0.3034093677997589,
+      "learning_rate": 0.001,
+      "loss": 2.2661,
+      "step": 3314
+    },
+    {
+      "epoch": 0.140240291056773,
+      "grad_norm": 0.2352837473154068,
+      "learning_rate": 0.001,
+      "loss": 1.6646,
+      "step": 3315
+    },
+    {
+      "epoch": 0.14028259582028937,
+      "grad_norm": 0.2697775661945343,
+      "learning_rate": 0.001,
+      "loss": 2.322,
+      "step": 3316
+    },
+    {
+      "epoch": 0.14032490058380573,
+      "grad_norm": 0.889599621295929,
+      "learning_rate": 0.001,
+      "loss": 2.4923,
+      "step": 3317
+    },
+    {
+      "epoch": 0.1403672053473221,
+      "grad_norm": 0.3509933352470398,
+      "learning_rate": 0.001,
+      "loss": 2.4785,
+      "step": 3318
+    },
+    {
+      "epoch": 0.1404095101108385,
+      "grad_norm": 0.2816075086593628,
+      "learning_rate": 0.001,
+      "loss": 2.1828,
+      "step": 3319
+    },
+    {
+      "epoch": 0.14045181487435485,
+      "grad_norm": 0.4183788299560547,
+      "learning_rate": 0.001,
+      "loss": 2.2104,
+      "step": 3320
+    },
+    {
+      "epoch": 0.14049411963787123,
+      "grad_norm": 31.530492782592773,
+      "learning_rate": 0.001,
+      "loss": 2.0174,
+      "step": 3321
+    },
+    {
+      "epoch": 0.14053642440138758,
+      "grad_norm": 0.3534470796585083,
+      "learning_rate": 0.001,
+      "loss": 2.234,
+      "step": 3322
+    },
+    {
+      "epoch": 0.14057872916490396,
+      "grad_norm": 0.3242388367652893,
+      "learning_rate": 0.001,
+      "loss": 2.3006,
+      "step": 3323
+    },
+    {
+      "epoch": 0.14062103392842035,
+      "grad_norm": 1.5292270183563232,
+      "learning_rate": 0.001,
+      "loss": 2.3741,
+      "step": 3324
+    },
+    {
+      "epoch": 0.1406633386919367,
+      "grad_norm": 0.5664166212081909,
+      "learning_rate": 0.001,
+      "loss": 2.2073,
+      "step": 3325
+    },
+    {
+      "epoch": 0.14070564345545308,
+      "grad_norm": 0.4291297495365143,
+      "learning_rate": 0.001,
+      "loss": 2.9416,
+      "step": 3326
+    },
+    {
+      "epoch": 0.14074794821896947,
+      "grad_norm": 0.3085836172103882,
+      "learning_rate": 0.001,
+      "loss": 3.2857,
+      "step": 3327
+    },
+    {
+      "epoch": 0.14079025298248582,
+      "grad_norm": 0.3337254524230957,
+      "learning_rate": 0.001,
+      "loss": 2.8056,
+      "step": 3328
+    },
+    {
+      "epoch": 0.1408325577460022,
+      "grad_norm": 0.267048180103302,
+      "learning_rate": 0.001,
+      "loss": 2.1272,
+      "step": 3329
+    },
+    {
+      "epoch": 0.14087486250951858,
+      "grad_norm": 0.8167191743850708,
+      "learning_rate": 0.001,
+      "loss": 2.2273,
+      "step": 3330
+    },
+    {
+      "epoch": 0.14091716727303494,
+      "grad_norm": 0.426624596118927,
+      "learning_rate": 0.001,
+      "loss": 3.7321,
+      "step": 3331
+    },
+    {
+      "epoch": 0.14095947203655132,
+      "grad_norm": 0.2855152189731598,
+      "learning_rate": 0.001,
+      "loss": 3.1353,
+      "step": 3332
+    },
+    {
+      "epoch": 0.14100177680006767,
+      "grad_norm": 0.647718071937561,
+      "learning_rate": 0.001,
+      "loss": 2.2319,
+      "step": 3333
+    },
+    {
+      "epoch": 0.14104408156358406,
+      "grad_norm": 0.7934921383857727,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 3334
+    },
+    {
+      "epoch": 0.14108638632710044,
+      "grad_norm": 0.22644270956516266,
+      "learning_rate": 0.001,
+      "loss": 2.802,
+      "step": 3335
+    },
+    {
+      "epoch": 0.1411286910906168,
+      "grad_norm": 0.5594651699066162,
+      "learning_rate": 0.001,
+      "loss": 2.4036,
+      "step": 3336
+    },
+    {
+      "epoch": 0.14117099585413317,
+      "grad_norm": 0.34176865220069885,
+      "learning_rate": 0.001,
+      "loss": 3.0433,
+      "step": 3337
+    },
+    {
+      "epoch": 0.14121330061764956,
+      "grad_norm": 0.9747275114059448,
+      "learning_rate": 0.001,
+      "loss": 2.6015,
+      "step": 3338
+    },
+    {
+      "epoch": 0.1412556053811659,
+      "grad_norm": 0.4771696925163269,
+      "learning_rate": 0.001,
+      "loss": 2.0621,
+      "step": 3339
+    },
+    {
+      "epoch": 0.1412979101446823,
+      "grad_norm": 0.2995487451553345,
+      "learning_rate": 0.001,
+      "loss": 2.6303,
+      "step": 3340
+    },
+    {
+      "epoch": 0.14134021490819867,
+      "grad_norm": 0.3293488919734955,
+      "learning_rate": 0.001,
+      "loss": 3.3868,
+      "step": 3341
+    },
+    {
+      "epoch": 0.14138251967171503,
+      "grad_norm": 0.2893940508365631,
+      "learning_rate": 0.001,
+      "loss": 3.0442,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1414248244352314,
+      "grad_norm": 0.31001016497612,
+      "learning_rate": 0.001,
+      "loss": 2.3384,
+      "step": 3343
+    },
+    {
+      "epoch": 0.14146712919874777,
+      "grad_norm": 0.31601953506469727,
+      "learning_rate": 0.001,
+      "loss": 2.9162,
+      "step": 3344
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.27529218792915344,
+      "learning_rate": 0.001,
+      "loss": 2.4267,
+      "step": 3345
+    },
+    {
+      "epoch": 0.14155173872578053,
+      "grad_norm": 0.6147814393043518,
+      "learning_rate": 0.001,
+      "loss": 1.7324,
+      "step": 3346
+    },
+    {
+      "epoch": 0.14159404348929688,
+      "grad_norm": 0.25061115622520447,
+      "learning_rate": 0.001,
+      "loss": 2.4223,
+      "step": 3347
+    },
+    {
+      "epoch": 0.14163634825281327,
+      "grad_norm": 7.874605178833008,
+      "learning_rate": 0.001,
+      "loss": 2.3057,
+      "step": 3348
+    },
+    {
+      "epoch": 0.14167865301632965,
+      "grad_norm": 0.23804105818271637,
+      "learning_rate": 0.001,
+      "loss": 1.9634,
+      "step": 3349
+    },
+    {
+      "epoch": 0.141720957779846,
+      "grad_norm": 1.9372200965881348,
+      "learning_rate": 0.001,
+      "loss": 2.3494,
+      "step": 3350
+    },
+    {
+      "epoch": 0.14176326254336238,
+      "grad_norm": 0.3472435772418976,
+      "learning_rate": 0.001,
+      "loss": 2.9078,
+      "step": 3351
+    },
+    {
+      "epoch": 0.14180556730687877,
+      "grad_norm": 0.35763394832611084,
+      "learning_rate": 0.001,
+      "loss": 3.6208,
+      "step": 3352
+    },
+    {
+      "epoch": 0.14184787207039512,
+      "grad_norm": 0.3070499300956726,
+      "learning_rate": 0.001,
+      "loss": 1.5918,
+      "step": 3353
+    },
+    {
+      "epoch": 0.1418901768339115,
+      "grad_norm": 0.4274042248725891,
+      "learning_rate": 0.001,
+      "loss": 2.3483,
+      "step": 3354
+    },
+    {
+      "epoch": 0.14193248159742788,
+      "grad_norm": 0.2639428675174713,
+      "learning_rate": 0.001,
+      "loss": 2.1166,
+      "step": 3355
+    },
+    {
+      "epoch": 0.14197478636094424,
+      "grad_norm": 0.23096594214439392,
+      "learning_rate": 0.001,
+      "loss": 1.9444,
+      "step": 3356
+    },
+    {
+      "epoch": 0.14201709112446062,
+      "grad_norm": 0.2586401402950287,
+      "learning_rate": 0.001,
+      "loss": 2.3109,
+      "step": 3357
+    },
+    {
+      "epoch": 0.14205939588797697,
+      "grad_norm": 0.26827335357666016,
+      "learning_rate": 0.001,
+      "loss": 2.2147,
+      "step": 3358
+    },
+    {
+      "epoch": 0.14210170065149336,
+      "grad_norm": 0.26217904686927795,
+      "learning_rate": 0.001,
+      "loss": 2.1678,
+      "step": 3359
+    },
+    {
+      "epoch": 0.14214400541500974,
+      "grad_norm": 0.45599186420440674,
+      "learning_rate": 0.001,
+      "loss": 2.3742,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1421863101785261,
+      "grad_norm": 0.24741707742214203,
+      "learning_rate": 0.001,
+      "loss": 1.727,
+      "step": 3361
+    },
+    {
+      "epoch": 0.14222861494204248,
+      "grad_norm": 0.300525963306427,
+      "learning_rate": 0.001,
+      "loss": 2.6092,
+      "step": 3362
+    },
+    {
+      "epoch": 0.14227091970555886,
+      "grad_norm": 0.27571022510528564,
+      "learning_rate": 0.001,
+      "loss": 2.3276,
+      "step": 3363
+    },
+    {
+      "epoch": 0.1423132244690752,
+      "grad_norm": 0.42578956484794617,
+      "learning_rate": 0.001,
+      "loss": 1.5469,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1423555292325916,
+      "grad_norm": 0.2815234661102295,
+      "learning_rate": 0.001,
+      "loss": 3.2236,
+      "step": 3365
+    },
+    {
+      "epoch": 0.14239783399610798,
+      "grad_norm": 3.376448392868042,
+      "learning_rate": 0.001,
+      "loss": 2.2013,
+      "step": 3366
+    },
+    {
+      "epoch": 0.14244013875962433,
+      "grad_norm": 1.9560842514038086,
+      "learning_rate": 0.001,
+      "loss": 2.0186,
+      "step": 3367
+    },
+    {
+      "epoch": 0.1424824435231407,
+      "grad_norm": 0.23897530138492584,
+      "learning_rate": 0.001,
+      "loss": 2.2577,
+      "step": 3368
+    },
+    {
+      "epoch": 0.14252474828665707,
+      "grad_norm": 0.25748342275619507,
+      "learning_rate": 0.001,
+      "loss": 2.3233,
+      "step": 3369
+    },
+    {
+      "epoch": 0.14256705305017345,
+      "grad_norm": 0.2940260171890259,
+      "learning_rate": 0.001,
+      "loss": 1.8611,
+      "step": 3370
+    },
+    {
+      "epoch": 0.14260935781368983,
+      "grad_norm": 0.5368818640708923,
+      "learning_rate": 0.001,
+      "loss": 5.5408,
+      "step": 3371
+    },
+    {
+      "epoch": 0.14265166257720618,
+      "grad_norm": 0.2916557788848877,
+      "learning_rate": 0.001,
+      "loss": 2.0839,
+      "step": 3372
+    },
+    {
+      "epoch": 0.14269396734072257,
+      "grad_norm": 0.21450082957744598,
+      "learning_rate": 0.001,
+      "loss": 3.1867,
+      "step": 3373
+    },
+    {
+      "epoch": 0.14273627210423895,
+      "grad_norm": 0.43757933378219604,
+      "learning_rate": 0.001,
+      "loss": 3.901,
+      "step": 3374
+    },
+    {
+      "epoch": 0.1427785768677553,
+      "grad_norm": 0.2997872829437256,
+      "learning_rate": 0.001,
+      "loss": 2.2917,
+      "step": 3375
+    },
+    {
+      "epoch": 0.14282088163127168,
+      "grad_norm": 0.26807701587677,
+      "learning_rate": 0.001,
+      "loss": 1.9812,
+      "step": 3376
+    },
+    {
+      "epoch": 0.14286318639478807,
+      "grad_norm": 0.7264923453330994,
+      "learning_rate": 0.001,
+      "loss": 1.7461,
+      "step": 3377
+    },
+    {
+      "epoch": 0.14290549115830442,
+      "grad_norm": 0.23869849741458893,
+      "learning_rate": 0.001,
+      "loss": 2.7253,
+      "step": 3378
+    },
+    {
+      "epoch": 0.1429477959218208,
+      "grad_norm": 2.684951066970825,
+      "learning_rate": 0.001,
+      "loss": 1.8465,
+      "step": 3379
+    },
+    {
+      "epoch": 0.14299010068533716,
+      "grad_norm": 0.33340945839881897,
+      "learning_rate": 0.001,
+      "loss": 2.7263,
+      "step": 3380
+    },
+    {
+      "epoch": 0.14303240544885354,
+      "grad_norm": 0.8364661335945129,
+      "learning_rate": 0.001,
+      "loss": 3.4191,
+      "step": 3381
+    },
+    {
+      "epoch": 0.14307471021236992,
+      "grad_norm": 0.2745528221130371,
+      "learning_rate": 0.001,
+      "loss": 2.9753,
+      "step": 3382
+    },
+    {
+      "epoch": 0.14311701497588628,
+      "grad_norm": 0.5543707609176636,
+      "learning_rate": 0.001,
+      "loss": 2.9777,
+      "step": 3383
+    },
+    {
+      "epoch": 0.14315931973940266,
+      "grad_norm": 0.31856632232666016,
+      "learning_rate": 0.001,
+      "loss": 3.4294,
+      "step": 3384
+    },
+    {
+      "epoch": 0.14320162450291904,
+      "grad_norm": 0.2563943862915039,
+      "learning_rate": 0.001,
+      "loss": 2.3104,
+      "step": 3385
+    },
+    {
+      "epoch": 0.1432439292664354,
+      "grad_norm": 1.2910809516906738,
+      "learning_rate": 0.001,
+      "loss": 2.1959,
+      "step": 3386
+    },
+    {
+      "epoch": 0.14328623402995178,
+      "grad_norm": 0.261447548866272,
+      "learning_rate": 0.001,
+      "loss": 1.9312,
+      "step": 3387
+    },
+    {
+      "epoch": 0.14332853879346816,
+      "grad_norm": 1.013630747795105,
+      "learning_rate": 0.001,
+      "loss": 1.6115,
+      "step": 3388
+    },
+    {
+      "epoch": 0.1433708435569845,
+      "grad_norm": 2.35341477394104,
+      "learning_rate": 0.001,
+      "loss": 2.19,
+      "step": 3389
+    },
+    {
+      "epoch": 0.1434131483205009,
+      "grad_norm": 0.24885472655296326,
+      "learning_rate": 0.001,
+      "loss": 2.1222,
+      "step": 3390
+    },
+    {
+      "epoch": 0.14345545308401725,
+      "grad_norm": 0.45557206869125366,
+      "learning_rate": 0.001,
+      "loss": 2.2925,
+      "step": 3391
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.7096180319786072,
+      "learning_rate": 0.001,
+      "loss": 1.7682,
+      "step": 3392
+    },
+    {
+      "epoch": 0.14354006261105,
+      "grad_norm": 0.6223247051239014,
+      "learning_rate": 0.001,
+      "loss": 2.476,
+      "step": 3393
+    },
+    {
+      "epoch": 0.14358236737456637,
+      "grad_norm": 0.3025702238082886,
+      "learning_rate": 0.001,
+      "loss": 2.1303,
+      "step": 3394
+    },
+    {
+      "epoch": 0.14362467213808275,
+      "grad_norm": 0.6526556015014648,
+      "learning_rate": 0.001,
+      "loss": 2.4046,
+      "step": 3395
+    },
+    {
+      "epoch": 0.14366697690159913,
+      "grad_norm": 0.40125930309295654,
+      "learning_rate": 0.001,
+      "loss": 3.7139,
+      "step": 3396
+    },
+    {
+      "epoch": 0.14370928166511548,
+      "grad_norm": 0.26466187834739685,
+      "learning_rate": 0.001,
+      "loss": 2.4826,
+      "step": 3397
+    },
+    {
+      "epoch": 0.14375158642863187,
+      "grad_norm": 0.26691725850105286,
+      "learning_rate": 0.001,
+      "loss": 1.7453,
+      "step": 3398
+    },
+    {
+      "epoch": 0.14379389119214825,
+      "grad_norm": 0.2618395686149597,
+      "learning_rate": 0.001,
+      "loss": 2.1459,
+      "step": 3399
+    },
+    {
+      "epoch": 0.1438361959556646,
+      "grad_norm": 0.7489890456199646,
+      "learning_rate": 0.001,
+      "loss": 2.8917,
+      "step": 3400
+    },
+    {
+      "epoch": 0.14387850071918099,
+      "grad_norm": 0.24983298778533936,
+      "learning_rate": 0.001,
+      "loss": 1.8924,
+      "step": 3401
+    },
+    {
+      "epoch": 0.14392080548269734,
+      "grad_norm": 0.210697740316391,
+      "learning_rate": 0.001,
+      "loss": 2.1125,
+      "step": 3402
+    },
+    {
+      "epoch": 0.14396311024621372,
+      "grad_norm": 0.25680264830589294,
+      "learning_rate": 0.001,
+      "loss": 2.1817,
+      "step": 3403
+    },
+    {
+      "epoch": 0.1440054150097301,
+      "grad_norm": 0.2333562821149826,
+      "learning_rate": 0.001,
+      "loss": 2.1268,
+      "step": 3404
+    },
+    {
+      "epoch": 0.14404771977324646,
+      "grad_norm": 0.26086026430130005,
+      "learning_rate": 0.001,
+      "loss": 2.647,
+      "step": 3405
+    },
+    {
+      "epoch": 0.14409002453676284,
+      "grad_norm": 0.24561172723770142,
+      "learning_rate": 0.001,
+      "loss": 1.7931,
+      "step": 3406
+    },
+    {
+      "epoch": 0.14413232930027922,
+      "grad_norm": 0.2474185675382614,
+      "learning_rate": 0.001,
+      "loss": 2.4043,
+      "step": 3407
+    },
+    {
+      "epoch": 0.14417463406379558,
+      "grad_norm": 0.2294658124446869,
+      "learning_rate": 0.001,
+      "loss": 2.243,
+      "step": 3408
+    },
+    {
+      "epoch": 0.14421693882731196,
+      "grad_norm": 0.2385028749704361,
+      "learning_rate": 0.001,
+      "loss": 2.5765,
+      "step": 3409
+    },
+    {
+      "epoch": 0.14425924359082834,
+      "grad_norm": 0.2432500123977661,
+      "learning_rate": 0.001,
+      "loss": 1.8027,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1443015483543447,
+      "grad_norm": 0.2768508791923523,
+      "learning_rate": 0.001,
+      "loss": 2.4532,
+      "step": 3411
+    },
+    {
+      "epoch": 0.14434385311786108,
+      "grad_norm": 0.892453134059906,
+      "learning_rate": 0.001,
+      "loss": 2.4542,
+      "step": 3412
+    },
+    {
+      "epoch": 0.14438615788137743,
+      "grad_norm": 0.2457326054573059,
+      "learning_rate": 0.001,
+      "loss": 2.1699,
+      "step": 3413
+    },
+    {
+      "epoch": 0.1444284626448938,
+      "grad_norm": 0.26426583528518677,
+      "learning_rate": 0.001,
+      "loss": 2.1309,
+      "step": 3414
+    },
+    {
+      "epoch": 0.1444707674084102,
+      "grad_norm": 0.2871556282043457,
+      "learning_rate": 0.001,
+      "loss": 3.3215,
+      "step": 3415
+    },
+    {
+      "epoch": 0.14451307217192655,
+      "grad_norm": 0.2520632743835449,
+      "learning_rate": 0.001,
+      "loss": 1.9709,
+      "step": 3416
+    },
+    {
+      "epoch": 0.14455537693544293,
+      "grad_norm": 0.38204339146614075,
+      "learning_rate": 0.001,
+      "loss": 3.6021,
+      "step": 3417
+    },
+    {
+      "epoch": 0.1445976816989593,
+      "grad_norm": 0.26071059703826904,
+      "learning_rate": 0.001,
+      "loss": 2.2053,
+      "step": 3418
+    },
+    {
+      "epoch": 0.14463998646247567,
+      "grad_norm": 0.2733023762702942,
+      "learning_rate": 0.001,
+      "loss": 3.47,
+      "step": 3419
+    },
+    {
+      "epoch": 0.14468229122599205,
+      "grad_norm": 1.002029538154602,
+      "learning_rate": 0.001,
+      "loss": 3.2734,
+      "step": 3420
+    },
+    {
+      "epoch": 0.14472459598950843,
+      "grad_norm": 0.3164158761501312,
+      "learning_rate": 0.001,
+      "loss": 2.763,
+      "step": 3421
+    },
+    {
+      "epoch": 0.14476690075302479,
+      "grad_norm": 0.2513934075832367,
+      "learning_rate": 0.001,
+      "loss": 2.0143,
+      "step": 3422
+    },
+    {
+      "epoch": 0.14480920551654117,
+      "grad_norm": 0.24494628608226776,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 3423
+    },
+    {
+      "epoch": 0.14485151028005752,
+      "grad_norm": 0.27729180455207825,
+      "learning_rate": 0.001,
+      "loss": 2.6749,
+      "step": 3424
+    },
+    {
+      "epoch": 0.1448938150435739,
+      "grad_norm": 0.26897814869880676,
+      "learning_rate": 0.001,
+      "loss": 3.574,
+      "step": 3425
+    },
+    {
+      "epoch": 0.14493611980709029,
+      "grad_norm": 0.22974863648414612,
+      "learning_rate": 0.001,
+      "loss": 3.4897,
+      "step": 3426
+    },
+    {
+      "epoch": 0.14497842457060664,
+      "grad_norm": 0.22176092863082886,
+      "learning_rate": 0.001,
+      "loss": 1.9926,
+      "step": 3427
+    },
+    {
+      "epoch": 0.14502072933412302,
+      "grad_norm": 0.38495469093322754,
+      "learning_rate": 0.001,
+      "loss": 2.5689,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1450630340976394,
+      "grad_norm": 0.22981159389019012,
+      "learning_rate": 0.001,
+      "loss": 2.2184,
+      "step": 3429
+    },
+    {
+      "epoch": 0.14510533886115576,
+      "grad_norm": 0.2774735987186432,
+      "learning_rate": 0.001,
+      "loss": 3.0189,
+      "step": 3430
+    },
+    {
+      "epoch": 0.14514764362467214,
+      "grad_norm": 0.2698344886302948,
+      "learning_rate": 0.001,
+      "loss": 2.4688,
+      "step": 3431
+    },
+    {
+      "epoch": 0.14518994838818852,
+      "grad_norm": 0.6816653609275818,
+      "learning_rate": 0.001,
+      "loss": 2.0268,
+      "step": 3432
+    },
+    {
+      "epoch": 0.14523225315170488,
+      "grad_norm": 0.29013943672180176,
+      "learning_rate": 0.001,
+      "loss": 3.2067,
+      "step": 3433
+    },
+    {
+      "epoch": 0.14527455791522126,
+      "grad_norm": 0.22570838034152985,
+      "learning_rate": 0.001,
+      "loss": 3.0135,
+      "step": 3434
+    },
+    {
+      "epoch": 0.1453168626787376,
+      "grad_norm": 0.21075336635112762,
+      "learning_rate": 0.001,
+      "loss": 2.242,
+      "step": 3435
+    },
+    {
+      "epoch": 0.145359167442254,
+      "grad_norm": 0.2940874993801117,
+      "learning_rate": 0.001,
+      "loss": 3.109,
+      "step": 3436
+    },
+    {
+      "epoch": 0.14540147220577038,
+      "grad_norm": 0.21364165842533112,
+      "learning_rate": 0.001,
+      "loss": 1.3964,
+      "step": 3437
+    },
+    {
+      "epoch": 0.14544377696928673,
+      "grad_norm": 0.2253018319606781,
+      "learning_rate": 0.001,
+      "loss": 2.2411,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1454860817328031,
+      "grad_norm": 0.2131890505552292,
+      "learning_rate": 0.001,
+      "loss": 1.7277,
+      "step": 3439
+    },
+    {
+      "epoch": 0.1455283864963195,
+      "grad_norm": 0.2189098596572876,
+      "learning_rate": 0.001,
+      "loss": 1.7839,
+      "step": 3440
+    },
+    {
+      "epoch": 0.14557069125983585,
+      "grad_norm": 2.004612922668457,
+      "learning_rate": 0.001,
+      "loss": 1.8655,
+      "step": 3441
+    },
+    {
+      "epoch": 0.14561299602335223,
+      "grad_norm": 0.2156573235988617,
+      "learning_rate": 0.001,
+      "loss": 2.3993,
+      "step": 3442
+    },
+    {
+      "epoch": 0.1456553007868686,
+      "grad_norm": 0.30679431557655334,
+      "learning_rate": 0.001,
+      "loss": 2.738,
+      "step": 3443
+    },
+    {
+      "epoch": 0.14569760555038497,
+      "grad_norm": 0.2574104070663452,
+      "learning_rate": 0.001,
+      "loss": 2.4518,
+      "step": 3444
+    },
+    {
+      "epoch": 0.14573991031390135,
+      "grad_norm": 0.2621307075023651,
+      "learning_rate": 0.001,
+      "loss": 2.3622,
+      "step": 3445
+    },
+    {
+      "epoch": 0.1457822150774177,
+      "grad_norm": 0.30141088366508484,
+      "learning_rate": 0.001,
+      "loss": 2.494,
+      "step": 3446
+    },
+    {
+      "epoch": 0.14582451984093409,
+      "grad_norm": 0.19590161740779877,
+      "learning_rate": 0.001,
+      "loss": 1.6406,
+      "step": 3447
+    },
+    {
+      "epoch": 0.14586682460445047,
+      "grad_norm": 0.21325886249542236,
+      "learning_rate": 0.001,
+      "loss": 1.6452,
+      "step": 3448
+    },
+    {
+      "epoch": 0.14590912936796682,
+      "grad_norm": 0.22410674393177032,
+      "learning_rate": 0.001,
+      "loss": 2.4961,
+      "step": 3449
+    },
+    {
+      "epoch": 0.1459514341314832,
+      "grad_norm": 0.2333114743232727,
+      "learning_rate": 0.001,
+      "loss": 1.9583,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1459937388949996,
+      "grad_norm": 0.43659529089927673,
+      "learning_rate": 0.001,
+      "loss": 2.4177,
+      "step": 3451
+    },
+    {
+      "epoch": 0.14603604365851594,
+      "grad_norm": 0.23108899593353271,
+      "learning_rate": 0.001,
+      "loss": 1.9033,
+      "step": 3452
+    },
+    {
+      "epoch": 0.14607834842203232,
+      "grad_norm": 0.2105552852153778,
+      "learning_rate": 0.001,
+      "loss": 1.7711,
+      "step": 3453
+    },
+    {
+      "epoch": 0.1461206531855487,
+      "grad_norm": 0.22220350801944733,
+      "learning_rate": 0.001,
+      "loss": 3.0378,
+      "step": 3454
+    },
+    {
+      "epoch": 0.14616295794906506,
+      "grad_norm": 0.2945927679538727,
+      "learning_rate": 0.001,
+      "loss": 2.0337,
+      "step": 3455
+    },
+    {
+      "epoch": 0.14620526271258144,
+      "grad_norm": 0.26347851753234863,
+      "learning_rate": 0.001,
+      "loss": 2.1827,
+      "step": 3456
+    },
+    {
+      "epoch": 0.1462475674760978,
+      "grad_norm": 1.3712072372436523,
+      "learning_rate": 0.001,
+      "loss": 2.5637,
+      "step": 3457
+    },
+    {
+      "epoch": 0.14628987223961418,
+      "grad_norm": 0.33388158679008484,
+      "learning_rate": 0.001,
+      "loss": 1.8558,
+      "step": 3458
+    },
+    {
+      "epoch": 0.14633217700313056,
+      "grad_norm": 0.23212599754333496,
+      "learning_rate": 0.001,
+      "loss": 2.7854,
+      "step": 3459
+    },
+    {
+      "epoch": 0.1463744817666469,
+      "grad_norm": 0.2477710247039795,
+      "learning_rate": 0.001,
+      "loss": 2.2062,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1464167865301633,
+      "grad_norm": 0.5189692974090576,
+      "learning_rate": 0.001,
+      "loss": 2.5085,
+      "step": 3461
+    },
+    {
+      "epoch": 0.14645909129367968,
+      "grad_norm": 0.39620745182037354,
+      "learning_rate": 0.001,
+      "loss": 2.913,
+      "step": 3462
+    },
+    {
+      "epoch": 0.14650139605719603,
+      "grad_norm": 0.23096932470798492,
+      "learning_rate": 0.001,
+      "loss": 2.7137,
+      "step": 3463
+    },
+    {
+      "epoch": 0.1465437008207124,
+      "grad_norm": 0.2657645642757416,
+      "learning_rate": 0.001,
+      "loss": 2.3191,
+      "step": 3464
+    },
+    {
+      "epoch": 0.1465860055842288,
+      "grad_norm": 0.2581949532032013,
+      "learning_rate": 0.001,
+      "loss": 2.4491,
+      "step": 3465
+    },
+    {
+      "epoch": 0.14662831034774515,
+      "grad_norm": 0.21660561859607697,
+      "learning_rate": 0.001,
+      "loss": 2.4903,
+      "step": 3466
+    },
+    {
+      "epoch": 0.14667061511126153,
+      "grad_norm": 0.2140873819589615,
+      "learning_rate": 0.001,
+      "loss": 2.8028,
+      "step": 3467
+    },
+    {
+      "epoch": 0.1467129198747779,
+      "grad_norm": 0.23401807248592377,
+      "learning_rate": 0.001,
+      "loss": 1.9589,
+      "step": 3468
+    },
+    {
+      "epoch": 0.14675522463829427,
+      "grad_norm": 0.43440452218055725,
+      "learning_rate": 0.001,
+      "loss": 2.3189,
+      "step": 3469
+    },
+    {
+      "epoch": 0.14679752940181065,
+      "grad_norm": 0.2246183156967163,
+      "learning_rate": 0.001,
+      "loss": 2.8871,
+      "step": 3470
+    },
+    {
+      "epoch": 0.146839834165327,
+      "grad_norm": 0.25875502824783325,
+      "learning_rate": 0.001,
+      "loss": 2.123,
+      "step": 3471
+    },
+    {
+      "epoch": 0.1468821389288434,
+      "grad_norm": 0.2841735780239105,
+      "learning_rate": 0.001,
+      "loss": 2.4137,
+      "step": 3472
+    },
+    {
+      "epoch": 0.14692444369235977,
+      "grad_norm": 0.3783358037471771,
+      "learning_rate": 0.001,
+      "loss": 2.2655,
+      "step": 3473
+    },
+    {
+      "epoch": 0.14696674845587612,
+      "grad_norm": 0.37711983919143677,
+      "learning_rate": 0.001,
+      "loss": 2.7385,
+      "step": 3474
+    },
+    {
+      "epoch": 0.1470090532193925,
+      "grad_norm": 0.2421020269393921,
+      "learning_rate": 0.001,
+      "loss": 3.2295,
+      "step": 3475
+    },
+    {
+      "epoch": 0.1470513579829089,
+      "grad_norm": 0.3066774308681488,
+      "learning_rate": 0.001,
+      "loss": 2.2816,
+      "step": 3476
+    },
+    {
+      "epoch": 0.14709366274642524,
+      "grad_norm": 0.29278188943862915,
+      "learning_rate": 0.001,
+      "loss": 2.3431,
+      "step": 3477
+    },
+    {
+      "epoch": 0.14713596750994162,
+      "grad_norm": 0.259764701128006,
+      "learning_rate": 0.001,
+      "loss": 2.7806,
+      "step": 3478
+    },
+    {
+      "epoch": 0.147178272273458,
+      "grad_norm": 0.2844379246234894,
+      "learning_rate": 0.001,
+      "loss": 2.9955,
+      "step": 3479
+    },
+    {
+      "epoch": 0.14722057703697436,
+      "grad_norm": 0.46466735005378723,
+      "learning_rate": 0.001,
+      "loss": 2.212,
+      "step": 3480
+    },
+    {
+      "epoch": 0.14726288180049074,
+      "grad_norm": 0.2156391441822052,
+      "learning_rate": 0.001,
+      "loss": 1.6904,
+      "step": 3481
+    },
+    {
+      "epoch": 0.1473051865640071,
+      "grad_norm": 0.3095730245113373,
+      "learning_rate": 0.001,
+      "loss": 2.1532,
+      "step": 3482
+    },
+    {
+      "epoch": 0.14734749132752348,
+      "grad_norm": 0.23583956062793732,
+      "learning_rate": 0.001,
+      "loss": 1.8892,
+      "step": 3483
+    },
+    {
+      "epoch": 0.14738979609103986,
+      "grad_norm": 0.19967339932918549,
+      "learning_rate": 0.001,
+      "loss": 2.1184,
+      "step": 3484
+    },
+    {
+      "epoch": 0.14743210085455621,
+      "grad_norm": 0.22221580147743225,
+      "learning_rate": 0.001,
+      "loss": 2.1233,
+      "step": 3485
+    },
+    {
+      "epoch": 0.1474744056180726,
+      "grad_norm": 0.21620295941829681,
+      "learning_rate": 0.001,
+      "loss": 1.8619,
+      "step": 3486
+    },
+    {
+      "epoch": 0.14751671038158898,
+      "grad_norm": 0.2938464283943176,
+      "learning_rate": 0.001,
+      "loss": 2.0775,
+      "step": 3487
+    },
+    {
+      "epoch": 0.14755901514510533,
+      "grad_norm": 0.4123593270778656,
+      "learning_rate": 0.001,
+      "loss": 2.9428,
+      "step": 3488
+    },
+    {
+      "epoch": 0.14760131990862171,
+      "grad_norm": 0.3647308647632599,
+      "learning_rate": 0.001,
+      "loss": 2.6945,
+      "step": 3489
+    },
+    {
+      "epoch": 0.1476436246721381,
+      "grad_norm": 0.9250668287277222,
+      "learning_rate": 0.001,
+      "loss": 2.283,
+      "step": 3490
+    },
+    {
+      "epoch": 0.14768592943565445,
+      "grad_norm": 0.19745373725891113,
+      "learning_rate": 0.001,
+      "loss": 1.9975,
+      "step": 3491
+    },
+    {
+      "epoch": 0.14772823419917083,
+      "grad_norm": 0.30542847514152527,
+      "learning_rate": 0.001,
+      "loss": 2.3923,
+      "step": 3492
+    },
+    {
+      "epoch": 0.1477705389626872,
+      "grad_norm": 0.23577959835529327,
+      "learning_rate": 0.001,
+      "loss": 1.8087,
+      "step": 3493
+    },
+    {
+      "epoch": 0.14781284372620357,
+      "grad_norm": 1.1520733833312988,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 3494
+    },
+    {
+      "epoch": 0.14785514848971995,
+      "grad_norm": 0.4492546617984772,
+      "learning_rate": 0.001,
+      "loss": 2.4425,
+      "step": 3495
+    },
+    {
+      "epoch": 0.1478974532532363,
+      "grad_norm": 0.2272578924894333,
+      "learning_rate": 0.001,
+      "loss": 2.0611,
+      "step": 3496
+    },
+    {
+      "epoch": 0.1479397580167527,
+      "grad_norm": 1.9856147766113281,
+      "learning_rate": 0.001,
+      "loss": 2.5383,
+      "step": 3497
+    },
+    {
+      "epoch": 0.14798206278026907,
+      "grad_norm": 0.32524406909942627,
+      "learning_rate": 0.001,
+      "loss": 2.6931,
+      "step": 3498
+    },
+    {
+      "epoch": 0.14802436754378542,
+      "grad_norm": 0.3067854046821594,
+      "learning_rate": 0.001,
+      "loss": 2.6123,
+      "step": 3499
+    },
+    {
+      "epoch": 0.1480666723073018,
+      "grad_norm": 0.2817644476890564,
+      "learning_rate": 0.001,
+      "loss": 2.0649,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1481089770708182,
+      "grad_norm": 1.7400952577590942,
+      "learning_rate": 0.001,
+      "loss": 2.2053,
+      "step": 3501
+    },
+    {
+      "epoch": 0.14815128183433454,
+      "grad_norm": 0.2977108061313629,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 3502
+    },
+    {
+      "epoch": 0.14819358659785092,
+      "grad_norm": 0.3165165185928345,
+      "learning_rate": 0.001,
+      "loss": 3.425,
+      "step": 3503
+    },
+    {
+      "epoch": 0.14823589136136728,
+      "grad_norm": 0.25336670875549316,
+      "learning_rate": 0.001,
+      "loss": 1.8021,
+      "step": 3504
+    },
+    {
+      "epoch": 0.14827819612488366,
+      "grad_norm": 1.0275665521621704,
+      "learning_rate": 0.001,
+      "loss": 2.5291,
+      "step": 3505
+    },
+    {
+      "epoch": 0.14832050088840004,
+      "grad_norm": 0.291814923286438,
+      "learning_rate": 0.001,
+      "loss": 1.9457,
+      "step": 3506
+    },
+    {
+      "epoch": 0.1483628056519164,
+      "grad_norm": 0.25686556100845337,
+      "learning_rate": 0.001,
+      "loss": 2.059,
+      "step": 3507
+    },
+    {
+      "epoch": 0.14840511041543278,
+      "grad_norm": 0.2979572117328644,
+      "learning_rate": 0.001,
+      "loss": 1.9924,
+      "step": 3508
+    },
+    {
+      "epoch": 0.14844741517894916,
+      "grad_norm": 0.2901727259159088,
+      "learning_rate": 0.001,
+      "loss": 2.0276,
+      "step": 3509
+    },
+    {
+      "epoch": 0.14848971994246551,
+      "grad_norm": 0.34327811002731323,
+      "learning_rate": 0.001,
+      "loss": 2.7836,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1485320247059819,
+      "grad_norm": 0.38972118496894836,
+      "learning_rate": 0.001,
+      "loss": 3.0989,
+      "step": 3511
+    },
+    {
+      "epoch": 0.14857432946949828,
+      "grad_norm": 11.712198257446289,
+      "learning_rate": 0.001,
+      "loss": 3.3529,
+      "step": 3512
+    },
+    {
+      "epoch": 0.14861663423301463,
+      "grad_norm": 5.345020771026611,
+      "learning_rate": 0.001,
+      "loss": 3.9772,
+      "step": 3513
+    },
+    {
+      "epoch": 0.14865893899653101,
+      "grad_norm": 0.3382362425327301,
+      "learning_rate": 0.001,
+      "loss": 2.0698,
+      "step": 3514
+    },
+    {
+      "epoch": 0.14870124376004737,
+      "grad_norm": 0.24800531566143036,
+      "learning_rate": 0.001,
+      "loss": 2.5415,
+      "step": 3515
+    },
+    {
+      "epoch": 0.14874354852356375,
+      "grad_norm": 0.2699640393257141,
+      "learning_rate": 0.001,
+      "loss": 3.1028,
+      "step": 3516
+    },
+    {
+      "epoch": 0.14878585328708013,
+      "grad_norm": 0.273624062538147,
+      "learning_rate": 0.001,
+      "loss": 2.3766,
+      "step": 3517
+    },
+    {
+      "epoch": 0.1488281580505965,
+      "grad_norm": 0.2503005266189575,
+      "learning_rate": 0.001,
+      "loss": 2.6137,
+      "step": 3518
+    },
+    {
+      "epoch": 0.14887046281411287,
+      "grad_norm": 0.8663057684898376,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 3519
+    },
+    {
+      "epoch": 0.14891276757762925,
+      "grad_norm": 0.25076824426651,
+      "learning_rate": 0.001,
+      "loss": 2.2382,
+      "step": 3520
+    },
+    {
+      "epoch": 0.1489550723411456,
+      "grad_norm": 0.34838005900382996,
+      "learning_rate": 0.001,
+      "loss": 2.0448,
+      "step": 3521
+    },
+    {
+      "epoch": 0.148997377104662,
+      "grad_norm": 0.2928178012371063,
+      "learning_rate": 0.001,
+      "loss": 3.3827,
+      "step": 3522
+    },
+    {
+      "epoch": 0.14903968186817837,
+      "grad_norm": 1.0668600797653198,
+      "learning_rate": 0.001,
+      "loss": 2.3127,
+      "step": 3523
+    },
+    {
+      "epoch": 0.14908198663169472,
+      "grad_norm": 0.32312437891960144,
+      "learning_rate": 0.001,
+      "loss": 2.3694,
+      "step": 3524
+    },
+    {
+      "epoch": 0.1491242913952111,
+      "grad_norm": 19.220008850097656,
+      "learning_rate": 0.001,
+      "loss": 1.897,
+      "step": 3525
+    },
+    {
+      "epoch": 0.14916659615872746,
+      "grad_norm": 0.2430810034275055,
+      "learning_rate": 0.001,
+      "loss": 2.16,
+      "step": 3526
+    },
+    {
+      "epoch": 0.14920890092224384,
+      "grad_norm": 0.3711296319961548,
+      "learning_rate": 0.001,
+      "loss": 2.4911,
+      "step": 3527
+    },
+    {
+      "epoch": 0.14925120568576022,
+      "grad_norm": 0.4155080020427704,
+      "learning_rate": 0.001,
+      "loss": 3.8939,
+      "step": 3528
+    },
+    {
+      "epoch": 0.14929351044927658,
+      "grad_norm": 0.3918618857860565,
+      "learning_rate": 0.001,
+      "loss": 2.4925,
+      "step": 3529
+    },
+    {
+      "epoch": 0.14933581521279296,
+      "grad_norm": 0.3465852439403534,
+      "learning_rate": 0.001,
+      "loss": 2.4602,
+      "step": 3530
+    },
+    {
+      "epoch": 0.14937811997630934,
+      "grad_norm": 0.2546325623989105,
+      "learning_rate": 0.001,
+      "loss": 3.3572,
+      "step": 3531
+    },
+    {
+      "epoch": 0.1494204247398257,
+      "grad_norm": 0.24824251234531403,
+      "learning_rate": 0.001,
+      "loss": 2.3866,
+      "step": 3532
+    },
+    {
+      "epoch": 0.14946272950334208,
+      "grad_norm": 0.783733606338501,
+      "learning_rate": 0.001,
+      "loss": 2.2685,
+      "step": 3533
+    },
+    {
+      "epoch": 0.14950503426685846,
+      "grad_norm": 0.27685683965682983,
+      "learning_rate": 0.001,
+      "loss": 2.6839,
+      "step": 3534
+    },
+    {
+      "epoch": 0.14954733903037482,
+      "grad_norm": 0.9122802019119263,
+      "learning_rate": 0.001,
+      "loss": 2.1993,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1495896437938912,
+      "grad_norm": 0.20734846591949463,
+      "learning_rate": 0.001,
+      "loss": 2.7781,
+      "step": 3536
+    },
+    {
+      "epoch": 0.14963194855740755,
+      "grad_norm": 0.303323894739151,
+      "learning_rate": 0.001,
+      "loss": 3.147,
+      "step": 3537
+    },
+    {
+      "epoch": 0.14967425332092393,
+      "grad_norm": 3.6282622814178467,
+      "learning_rate": 0.001,
+      "loss": 3.1942,
+      "step": 3538
+    },
+    {
+      "epoch": 0.14971655808444032,
+      "grad_norm": 0.5021495223045349,
+      "learning_rate": 0.001,
+      "loss": 2.3829,
+      "step": 3539
+    },
+    {
+      "epoch": 0.14975886284795667,
+      "grad_norm": 0.3432963788509369,
+      "learning_rate": 0.001,
+      "loss": 2.3058,
+      "step": 3540
+    },
+    {
+      "epoch": 0.14980116761147305,
+      "grad_norm": 0.46659693121910095,
+      "learning_rate": 0.001,
+      "loss": 2.5437,
+      "step": 3541
+    },
+    {
+      "epoch": 0.14984347237498943,
+      "grad_norm": 0.2588510513305664,
+      "learning_rate": 0.001,
+      "loss": 1.9825,
+      "step": 3542
+    },
+    {
+      "epoch": 0.1498857771385058,
+      "grad_norm": 0.8514876961708069,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 3543
+    },
+    {
+      "epoch": 0.14992808190202217,
+      "grad_norm": 1.0069010257720947,
+      "learning_rate": 0.001,
+      "loss": 1.9367,
+      "step": 3544
+    },
+    {
+      "epoch": 0.14997038666553855,
+      "grad_norm": 0.3433607220649719,
+      "learning_rate": 0.001,
+      "loss": 1.9445,
+      "step": 3545
+    },
+    {
+      "epoch": 0.1500126914290549,
+      "grad_norm": 0.9383977055549622,
+      "learning_rate": 0.001,
+      "loss": 3.1781,
+      "step": 3546
+    },
+    {
+      "epoch": 0.1500549961925713,
+      "grad_norm": 0.6198011040687561,
+      "learning_rate": 0.001,
+      "loss": 2.755,
+      "step": 3547
+    },
+    {
+      "epoch": 0.15009730095608764,
+      "grad_norm": 0.7712514400482178,
+      "learning_rate": 0.001,
+      "loss": 2.3735,
+      "step": 3548
+    },
+    {
+      "epoch": 0.15013960571960402,
+      "grad_norm": 0.8419329524040222,
+      "learning_rate": 0.001,
+      "loss": 1.8685,
+      "step": 3549
+    },
+    {
+      "epoch": 0.1501819104831204,
+      "grad_norm": 0.38785937428474426,
+      "learning_rate": 0.001,
+      "loss": 2.7159,
+      "step": 3550
+    },
+    {
+      "epoch": 0.15022421524663676,
+      "grad_norm": 0.25905489921569824,
+      "learning_rate": 0.001,
+      "loss": 1.916,
+      "step": 3551
+    },
+    {
+      "epoch": 0.15026652001015314,
+      "grad_norm": 0.28637632727622986,
+      "learning_rate": 0.001,
+      "loss": 2.152,
+      "step": 3552
+    },
+    {
+      "epoch": 0.15030882477366952,
+      "grad_norm": 0.4733045697212219,
+      "learning_rate": 0.001,
+      "loss": 2.2419,
+      "step": 3553
+    },
+    {
+      "epoch": 0.15035112953718588,
+      "grad_norm": 0.3638868033885956,
+      "learning_rate": 0.001,
+      "loss": 2.0374,
+      "step": 3554
+    },
+    {
+      "epoch": 0.15039343430070226,
+      "grad_norm": 0.4152970612049103,
+      "learning_rate": 0.001,
+      "loss": 2.3609,
+      "step": 3555
+    },
+    {
+      "epoch": 0.15043573906421864,
+      "grad_norm": 1.5988473892211914,
+      "learning_rate": 0.001,
+      "loss": 1.8426,
+      "step": 3556
+    },
+    {
+      "epoch": 0.150478043827735,
+      "grad_norm": 0.24985834956169128,
+      "learning_rate": 0.001,
+      "loss": 2.2193,
+      "step": 3557
+    },
+    {
+      "epoch": 0.15052034859125138,
+      "grad_norm": 7.093243598937988,
+      "learning_rate": 0.001,
+      "loss": 1.9872,
+      "step": 3558
+    },
+    {
+      "epoch": 0.15056265335476773,
+      "grad_norm": 0.5130526423454285,
+      "learning_rate": 0.001,
+      "loss": 3.1589,
+      "step": 3559
+    },
+    {
+      "epoch": 0.15060495811828412,
+      "grad_norm": 0.9955002665519714,
+      "learning_rate": 0.001,
+      "loss": 2.2292,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1506472628818005,
+      "grad_norm": 0.4683253765106201,
+      "learning_rate": 0.001,
+      "loss": 1.9829,
+      "step": 3561
+    },
+    {
+      "epoch": 0.15068956764531685,
+      "grad_norm": 0.24029046297073364,
+      "learning_rate": 0.001,
+      "loss": 2.5349,
+      "step": 3562
+    },
+    {
+      "epoch": 0.15073187240883323,
+      "grad_norm": 0.24601727724075317,
+      "learning_rate": 0.001,
+      "loss": 2.589,
+      "step": 3563
+    },
+    {
+      "epoch": 0.15077417717234962,
+      "grad_norm": 0.2925271987915039,
+      "learning_rate": 0.001,
+      "loss": 2.6538,
+      "step": 3564
+    },
+    {
+      "epoch": 0.15081648193586597,
+      "grad_norm": 0.6500786542892456,
+      "learning_rate": 0.001,
+      "loss": 2.3506,
+      "step": 3565
+    },
+    {
+      "epoch": 0.15085878669938235,
+      "grad_norm": 0.3403443396091461,
+      "learning_rate": 0.001,
+      "loss": 3.4838,
+      "step": 3566
+    },
+    {
+      "epoch": 0.15090109146289873,
+      "grad_norm": 0.2680628299713135,
+      "learning_rate": 0.001,
+      "loss": 3.188,
+      "step": 3567
+    },
+    {
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.25424718856811523,
+      "learning_rate": 0.001,
+      "loss": 2.5081,
+      "step": 3568
+    },
+    {
+      "epoch": 0.15098570098993147,
+      "grad_norm": 0.24566307663917542,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 3569
+    },
+    {
+      "epoch": 0.15102800575344782,
+      "grad_norm": 2.2985541820526123,
+      "learning_rate": 0.001,
+      "loss": 2.8511,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1510703105169642,
+      "grad_norm": 0.2778152823448181,
+      "learning_rate": 0.001,
+      "loss": 2.8095,
+      "step": 3571
+    },
+    {
+      "epoch": 0.1511126152804806,
+      "grad_norm": 1.3614428043365479,
+      "learning_rate": 0.001,
+      "loss": 2.2643,
+      "step": 3572
+    },
+    {
+      "epoch": 0.15115492004399694,
+      "grad_norm": 4.856557846069336,
+      "learning_rate": 0.001,
+      "loss": 2.3387,
+      "step": 3573
+    },
+    {
+      "epoch": 0.15119722480751333,
+      "grad_norm": 0.2886848449707031,
+      "learning_rate": 0.001,
+      "loss": 3.1371,
+      "step": 3574
+    },
+    {
+      "epoch": 0.1512395295710297,
+      "grad_norm": 0.2378971427679062,
+      "learning_rate": 0.001,
+      "loss": 2.4346,
+      "step": 3575
+    },
+    {
+      "epoch": 0.15128183433454606,
+      "grad_norm": 0.28027504682540894,
+      "learning_rate": 0.001,
+      "loss": 2.2992,
+      "step": 3576
+    },
+    {
+      "epoch": 0.15132413909806244,
+      "grad_norm": 0.9018852114677429,
+      "learning_rate": 0.001,
+      "loss": 1.6448,
+      "step": 3577
+    },
+    {
+      "epoch": 0.15136644386157883,
+      "grad_norm": 0.28057655692100525,
+      "learning_rate": 0.001,
+      "loss": 1.584,
+      "step": 3578
+    },
+    {
+      "epoch": 0.15140874862509518,
+      "grad_norm": 0.6362797021865845,
+      "learning_rate": 0.001,
+      "loss": 3.2339,
+      "step": 3579
+    },
+    {
+      "epoch": 0.15145105338861156,
+      "grad_norm": 2.48857045173645,
+      "learning_rate": 0.001,
+      "loss": 2.0265,
+      "step": 3580
+    },
+    {
+      "epoch": 0.15149335815212792,
+      "grad_norm": 0.341279536485672,
+      "learning_rate": 0.001,
+      "loss": 2.7349,
+      "step": 3581
+    },
+    {
+      "epoch": 0.1515356629156443,
+      "grad_norm": 0.39645954966545105,
+      "learning_rate": 0.001,
+      "loss": 2.335,
+      "step": 3582
+    },
+    {
+      "epoch": 0.15157796767916068,
+      "grad_norm": 2.092052698135376,
+      "learning_rate": 0.001,
+      "loss": 1.5818,
+      "step": 3583
+    },
+    {
+      "epoch": 0.15162027244267703,
+      "grad_norm": 5.003619194030762,
+      "learning_rate": 0.001,
+      "loss": 2.547,
+      "step": 3584
+    },
+    {
+      "epoch": 0.15166257720619342,
+      "grad_norm": 0.3023916482925415,
+      "learning_rate": 0.001,
+      "loss": 1.8312,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1517048819697098,
+      "grad_norm": 0.5600648522377014,
+      "learning_rate": 0.001,
+      "loss": 2.258,
+      "step": 3586
+    },
+    {
+      "epoch": 0.15174718673322615,
+      "grad_norm": 0.38749417662620544,
+      "learning_rate": 0.001,
+      "loss": 2.2861,
+      "step": 3587
+    },
+    {
+      "epoch": 0.15178949149674253,
+      "grad_norm": 1.0356155633926392,
+      "learning_rate": 0.001,
+      "loss": 3.6475,
+      "step": 3588
+    },
+    {
+      "epoch": 0.15183179626025892,
+      "grad_norm": 0.27420303225517273,
+      "learning_rate": 0.001,
+      "loss": 2.0222,
+      "step": 3589
+    },
+    {
+      "epoch": 0.15187410102377527,
+      "grad_norm": 1.7488365173339844,
+      "learning_rate": 0.001,
+      "loss": 2.0034,
+      "step": 3590
+    },
+    {
+      "epoch": 0.15191640578729165,
+      "grad_norm": 0.3303074538707733,
+      "learning_rate": 0.001,
+      "loss": 2.4819,
+      "step": 3591
+    },
+    {
+      "epoch": 0.151958710550808,
+      "grad_norm": 0.8918514847755432,
+      "learning_rate": 0.001,
+      "loss": 2.3028,
+      "step": 3592
+    },
+    {
+      "epoch": 0.1520010153143244,
+      "grad_norm": 0.40100759267807007,
+      "learning_rate": 0.001,
+      "loss": 2.3375,
+      "step": 3593
+    },
+    {
+      "epoch": 0.15204332007784077,
+      "grad_norm": 0.4550624489784241,
+      "learning_rate": 0.001,
+      "loss": 4.096,
+      "step": 3594
+    },
+    {
+      "epoch": 0.15208562484135713,
+      "grad_norm": 0.7799673080444336,
+      "learning_rate": 0.001,
+      "loss": 2.6198,
+      "step": 3595
+    },
+    {
+      "epoch": 0.1521279296048735,
+      "grad_norm": 0.34559065103530884,
+      "learning_rate": 0.001,
+      "loss": 2.9821,
+      "step": 3596
+    },
+    {
+      "epoch": 0.1521702343683899,
+      "grad_norm": 0.3571106493473053,
+      "learning_rate": 0.001,
+      "loss": 2.503,
+      "step": 3597
+    },
+    {
+      "epoch": 0.15221253913190624,
+      "grad_norm": 0.3727911710739136,
+      "learning_rate": 0.001,
+      "loss": 2.5119,
+      "step": 3598
+    },
+    {
+      "epoch": 0.15225484389542263,
+      "grad_norm": 4.5436015129089355,
+      "learning_rate": 0.001,
+      "loss": 3.7948,
+      "step": 3599
+    },
+    {
+      "epoch": 0.152297148658939,
+      "grad_norm": 0.3875895142555237,
+      "learning_rate": 0.001,
+      "loss": 2.6795,
+      "step": 3600
+    },
+    {
+      "epoch": 0.15233945342245536,
+      "grad_norm": 0.3254016935825348,
+      "learning_rate": 0.001,
+      "loss": 2.9257,
+      "step": 3601
+    },
+    {
+      "epoch": 0.15238175818597174,
+      "grad_norm": 0.27709123492240906,
+      "learning_rate": 0.001,
+      "loss": 2.9524,
+      "step": 3602
+    },
+    {
+      "epoch": 0.15242406294948813,
+      "grad_norm": 0.236286923289299,
+      "learning_rate": 0.001,
+      "loss": 2.0965,
+      "step": 3603
+    },
+    {
+      "epoch": 0.15246636771300448,
+      "grad_norm": 0.29797059297561646,
+      "learning_rate": 0.001,
+      "loss": 2.974,
+      "step": 3604
+    },
+    {
+      "epoch": 0.15250867247652086,
+      "grad_norm": 0.9622611999511719,
+      "learning_rate": 0.001,
+      "loss": 1.8431,
+      "step": 3605
+    },
+    {
+      "epoch": 0.15255097724003722,
+      "grad_norm": 0.32603511214256287,
+      "learning_rate": 0.001,
+      "loss": 3.1559,
+      "step": 3606
+    },
+    {
+      "epoch": 0.1525932820035536,
+      "grad_norm": 0.965201199054718,
+      "learning_rate": 0.001,
+      "loss": 3.2636,
+      "step": 3607
+    },
+    {
+      "epoch": 0.15263558676706998,
+      "grad_norm": 0.28725409507751465,
+      "learning_rate": 0.001,
+      "loss": 1.9986,
+      "step": 3608
+    },
+    {
+      "epoch": 0.15267789153058633,
+      "grad_norm": 0.28865811228752136,
+      "learning_rate": 0.001,
+      "loss": 3.1487,
+      "step": 3609
+    },
+    {
+      "epoch": 0.15272019629410272,
+      "grad_norm": 0.22873492538928986,
+      "learning_rate": 0.001,
+      "loss": 1.9311,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1527625010576191,
+      "grad_norm": 0.4197172522544861,
+      "learning_rate": 0.001,
+      "loss": 2.2263,
+      "step": 3611
+    },
+    {
+      "epoch": 0.15280480582113545,
+      "grad_norm": 0.3066697120666504,
+      "learning_rate": 0.001,
+      "loss": 2.3186,
+      "step": 3612
+    },
+    {
+      "epoch": 0.15284711058465184,
+      "grad_norm": 0.23598486185073853,
+      "learning_rate": 0.001,
+      "loss": 1.9821,
+      "step": 3613
+    },
+    {
+      "epoch": 0.15288941534816822,
+      "grad_norm": 0.21602579951286316,
+      "learning_rate": 0.001,
+      "loss": 1.9219,
+      "step": 3614
+    },
+    {
+      "epoch": 0.15293172011168457,
+      "grad_norm": 0.28833669424057007,
+      "learning_rate": 0.001,
+      "loss": 2.1563,
+      "step": 3615
+    },
+    {
+      "epoch": 0.15297402487520095,
+      "grad_norm": 0.6677043437957764,
+      "learning_rate": 0.001,
+      "loss": 2.4993,
+      "step": 3616
+    },
+    {
+      "epoch": 0.1530163296387173,
+      "grad_norm": 0.29518651962280273,
+      "learning_rate": 0.001,
+      "loss": 3.194,
+      "step": 3617
+    },
+    {
+      "epoch": 0.1530586344022337,
+      "grad_norm": 4.215554237365723,
+      "learning_rate": 0.001,
+      "loss": 2.5791,
+      "step": 3618
+    },
+    {
+      "epoch": 0.15310093916575007,
+      "grad_norm": 0.27064064145088196,
+      "learning_rate": 0.001,
+      "loss": 3.0186,
+      "step": 3619
+    },
+    {
+      "epoch": 0.15314324392926643,
+      "grad_norm": 0.22626028954982758,
+      "learning_rate": 0.001,
+      "loss": 1.7413,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1531855486927828,
+      "grad_norm": 0.5641341805458069,
+      "learning_rate": 0.001,
+      "loss": 2.2514,
+      "step": 3621
+    },
+    {
+      "epoch": 0.1532278534562992,
+      "grad_norm": 0.6790673136711121,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 3622
+    },
+    {
+      "epoch": 0.15327015821981554,
+      "grad_norm": 1.0468883514404297,
+      "learning_rate": 0.001,
+      "loss": 2.5064,
+      "step": 3623
+    },
+    {
+      "epoch": 0.15331246298333193,
+      "grad_norm": 0.710141658782959,
+      "learning_rate": 0.001,
+      "loss": 2.2824,
+      "step": 3624
+    },
+    {
+      "epoch": 0.1533547677468483,
+      "grad_norm": 0.46319958567619324,
+      "learning_rate": 0.001,
+      "loss": 2.4836,
+      "step": 3625
+    },
+    {
+      "epoch": 0.15339707251036466,
+      "grad_norm": 0.24517352879047394,
+      "learning_rate": 0.001,
+      "loss": 2.353,
+      "step": 3626
+    },
+    {
+      "epoch": 0.15343937727388104,
+      "grad_norm": 0.25180742144584656,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 3627
+    },
+    {
+      "epoch": 0.1534816820373974,
+      "grad_norm": 0.8973338603973389,
+      "learning_rate": 0.001,
+      "loss": 2.0089,
+      "step": 3628
+    },
+    {
+      "epoch": 0.15352398680091378,
+      "grad_norm": 0.2948894500732422,
+      "learning_rate": 0.001,
+      "loss": 1.8217,
+      "step": 3629
+    },
+    {
+      "epoch": 0.15356629156443016,
+      "grad_norm": 0.4694260358810425,
+      "learning_rate": 0.001,
+      "loss": 1.9615,
+      "step": 3630
+    },
+    {
+      "epoch": 0.15360859632794652,
+      "grad_norm": 0.2402891218662262,
+      "learning_rate": 0.001,
+      "loss": 2.4652,
+      "step": 3631
+    },
+    {
+      "epoch": 0.1536509010914629,
+      "grad_norm": 2.8009684085845947,
+      "learning_rate": 0.001,
+      "loss": 2.6741,
+      "step": 3632
+    },
+    {
+      "epoch": 0.15369320585497928,
+      "grad_norm": 0.3112724721431732,
+      "learning_rate": 0.001,
+      "loss": 2.6238,
+      "step": 3633
+    },
+    {
+      "epoch": 0.15373551061849564,
+      "grad_norm": 0.29018670320510864,
+      "learning_rate": 0.001,
+      "loss": 2.4207,
+      "step": 3634
+    },
+    {
+      "epoch": 0.15377781538201202,
+      "grad_norm": 0.8226271867752075,
+      "learning_rate": 0.001,
+      "loss": 2.0828,
+      "step": 3635
+    },
+    {
+      "epoch": 0.1538201201455284,
+      "grad_norm": 0.26996588706970215,
+      "learning_rate": 0.001,
+      "loss": 2.2658,
+      "step": 3636
+    },
+    {
+      "epoch": 0.15386242490904475,
+      "grad_norm": 0.33914855122566223,
+      "learning_rate": 0.001,
+      "loss": 2.4518,
+      "step": 3637
+    },
+    {
+      "epoch": 0.15390472967256114,
+      "grad_norm": 0.37279003858566284,
+      "learning_rate": 0.001,
+      "loss": 2.0381,
+      "step": 3638
+    },
+    {
+      "epoch": 0.1539470344360775,
+      "grad_norm": 0.4848242700099945,
+      "learning_rate": 0.001,
+      "loss": 2.8453,
+      "step": 3639
+    },
+    {
+      "epoch": 0.15398933919959387,
+      "grad_norm": 0.3430381119251251,
+      "learning_rate": 0.001,
+      "loss": 1.9578,
+      "step": 3640
+    },
+    {
+      "epoch": 0.15403164396311025,
+      "grad_norm": 0.2790931165218353,
+      "learning_rate": 0.001,
+      "loss": 1.5198,
+      "step": 3641
+    },
+    {
+      "epoch": 0.1540739487266266,
+      "grad_norm": 0.23120807111263275,
+      "learning_rate": 0.001,
+      "loss": 2.044,
+      "step": 3642
+    },
+    {
+      "epoch": 0.154116253490143,
+      "grad_norm": 0.26103150844573975,
+      "learning_rate": 0.001,
+      "loss": 2.3744,
+      "step": 3643
+    },
+    {
+      "epoch": 0.15415855825365937,
+      "grad_norm": 0.21912577748298645,
+      "learning_rate": 0.001,
+      "loss": 1.8251,
+      "step": 3644
+    },
+    {
+      "epoch": 0.15420086301717573,
+      "grad_norm": 0.268587201833725,
+      "learning_rate": 0.001,
+      "loss": 4.6103,
+      "step": 3645
+    },
+    {
+      "epoch": 0.1542431677806921,
+      "grad_norm": 0.24096794426441193,
+      "learning_rate": 0.001,
+      "loss": 1.8308,
+      "step": 3646
+    },
+    {
+      "epoch": 0.1542854725442085,
+      "grad_norm": 0.5062342882156372,
+      "learning_rate": 0.001,
+      "loss": 1.7605,
+      "step": 3647
+    },
+    {
+      "epoch": 0.15432777730772484,
+      "grad_norm": 1.528459072113037,
+      "learning_rate": 0.001,
+      "loss": 3.088,
+      "step": 3648
+    },
+    {
+      "epoch": 0.15437008207124123,
+      "grad_norm": 0.27189287543296814,
+      "learning_rate": 0.001,
+      "loss": 1.9692,
+      "step": 3649
+    },
+    {
+      "epoch": 0.15441238683475758,
+      "grad_norm": 0.22782421112060547,
+      "learning_rate": 0.001,
+      "loss": 1.9446,
+      "step": 3650
+    },
+    {
+      "epoch": 0.15445469159827396,
+      "grad_norm": 0.24183671176433563,
+      "learning_rate": 0.001,
+      "loss": 2.9768,
+      "step": 3651
+    },
+    {
+      "epoch": 0.15449699636179035,
+      "grad_norm": 0.37327948212623596,
+      "learning_rate": 0.001,
+      "loss": 2.5958,
+      "step": 3652
+    },
+    {
+      "epoch": 0.1545393011253067,
+      "grad_norm": 0.2614731788635254,
+      "learning_rate": 0.001,
+      "loss": 2.8753,
+      "step": 3653
+    },
+    {
+      "epoch": 0.15458160588882308,
+      "grad_norm": 0.30423954129219055,
+      "learning_rate": 0.001,
+      "loss": 2.9914,
+      "step": 3654
+    },
+    {
+      "epoch": 0.15462391065233946,
+      "grad_norm": 0.2651650607585907,
+      "learning_rate": 0.001,
+      "loss": 2.2149,
+      "step": 3655
+    },
+    {
+      "epoch": 0.15466621541585582,
+      "grad_norm": 0.29150527715682983,
+      "learning_rate": 0.001,
+      "loss": 3.1139,
+      "step": 3656
+    },
+    {
+      "epoch": 0.1547085201793722,
+      "grad_norm": 0.26949405670166016,
+      "learning_rate": 0.001,
+      "loss": 3.0504,
+      "step": 3657
+    },
+    {
+      "epoch": 0.15475082494288858,
+      "grad_norm": 0.277124285697937,
+      "learning_rate": 0.001,
+      "loss": 2.7155,
+      "step": 3658
+    },
+    {
+      "epoch": 0.15479312970640494,
+      "grad_norm": 0.24489636719226837,
+      "learning_rate": 0.001,
+      "loss": 2.4702,
+      "step": 3659
+    },
+    {
+      "epoch": 0.15483543446992132,
+      "grad_norm": 0.18855103850364685,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 3660
+    },
+    {
+      "epoch": 0.15487773923343767,
+      "grad_norm": 0.311081200838089,
+      "learning_rate": 0.001,
+      "loss": 2.6019,
+      "step": 3661
+    },
+    {
+      "epoch": 0.15492004399695405,
+      "grad_norm": 0.22517842054367065,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 3662
+    },
+    {
+      "epoch": 0.15496234876047044,
+      "grad_norm": 0.20022818446159363,
+      "learning_rate": 0.001,
+      "loss": 3.0709,
+      "step": 3663
+    },
+    {
+      "epoch": 0.1550046535239868,
+      "grad_norm": 0.3839741349220276,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 3664
+    },
+    {
+      "epoch": 0.15504695828750317,
+      "grad_norm": 0.3007325828075409,
+      "learning_rate": 0.001,
+      "loss": 3.0195,
+      "step": 3665
+    },
+    {
+      "epoch": 0.15508926305101955,
+      "grad_norm": 0.2808815538883209,
+      "learning_rate": 0.001,
+      "loss": 2.1549,
+      "step": 3666
+    },
+    {
+      "epoch": 0.1551315678145359,
+      "grad_norm": 0.23041900992393494,
+      "learning_rate": 0.001,
+      "loss": 1.8212,
+      "step": 3667
+    },
+    {
+      "epoch": 0.1551738725780523,
+      "grad_norm": 0.2662278711795807,
+      "learning_rate": 0.001,
+      "loss": 2.8651,
+      "step": 3668
+    },
+    {
+      "epoch": 0.15521617734156867,
+      "grad_norm": 1.5377527475357056,
+      "learning_rate": 0.001,
+      "loss": 2.4422,
+      "step": 3669
+    },
+    {
+      "epoch": 0.15525848210508503,
+      "grad_norm": 0.22316411137580872,
+      "learning_rate": 0.001,
+      "loss": 2.1897,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1553007868686014,
+      "grad_norm": 0.26653704047203064,
+      "learning_rate": 0.001,
+      "loss": 2.4008,
+      "step": 3671
+    },
+    {
+      "epoch": 0.15534309163211776,
+      "grad_norm": 0.2589184045791626,
+      "learning_rate": 0.001,
+      "loss": 2.8367,
+      "step": 3672
+    },
+    {
+      "epoch": 0.15538539639563415,
+      "grad_norm": 0.26263269782066345,
+      "learning_rate": 0.001,
+      "loss": 1.9108,
+      "step": 3673
+    },
+    {
+      "epoch": 0.15542770115915053,
+      "grad_norm": 0.3323366343975067,
+      "learning_rate": 0.001,
+      "loss": 2.1237,
+      "step": 3674
+    },
+    {
+      "epoch": 0.15547000592266688,
+      "grad_norm": 0.22809675335884094,
+      "learning_rate": 0.001,
+      "loss": 1.8055,
+      "step": 3675
+    },
+    {
+      "epoch": 0.15551231068618326,
+      "grad_norm": 0.2392113357782364,
+      "learning_rate": 0.001,
+      "loss": 3.3006,
+      "step": 3676
+    },
+    {
+      "epoch": 0.15555461544969965,
+      "grad_norm": 0.24501052498817444,
+      "learning_rate": 0.001,
+      "loss": 2.4065,
+      "step": 3677
+    },
+    {
+      "epoch": 0.155596920213216,
+      "grad_norm": 1.7419697046279907,
+      "learning_rate": 0.001,
+      "loss": 2.0626,
+      "step": 3678
+    },
+    {
+      "epoch": 0.15563922497673238,
+      "grad_norm": 0.2538470923900604,
+      "learning_rate": 0.001,
+      "loss": 2.548,
+      "step": 3679
+    },
+    {
+      "epoch": 0.15568152974024876,
+      "grad_norm": 0.27344977855682373,
+      "learning_rate": 0.001,
+      "loss": 1.8455,
+      "step": 3680
+    },
+    {
+      "epoch": 0.15572383450376512,
+      "grad_norm": 5.969301700592041,
+      "learning_rate": 0.001,
+      "loss": 1.987,
+      "step": 3681
+    },
+    {
+      "epoch": 0.1557661392672815,
+      "grad_norm": 0.21255825459957123,
+      "learning_rate": 0.001,
+      "loss": 2.1096,
+      "step": 3682
+    },
+    {
+      "epoch": 0.15580844403079785,
+      "grad_norm": 0.25273945927619934,
+      "learning_rate": 0.001,
+      "loss": 2.4582,
+      "step": 3683
+    },
+    {
+      "epoch": 0.15585074879431424,
+      "grad_norm": 0.2562407851219177,
+      "learning_rate": 0.001,
+      "loss": 1.9308,
+      "step": 3684
+    },
+    {
+      "epoch": 0.15589305355783062,
+      "grad_norm": 0.596889853477478,
+      "learning_rate": 0.001,
+      "loss": 2.817,
+      "step": 3685
+    },
+    {
+      "epoch": 0.15593535832134697,
+      "grad_norm": 0.55907142162323,
+      "learning_rate": 0.001,
+      "loss": 2.8084,
+      "step": 3686
+    },
+    {
+      "epoch": 0.15597766308486335,
+      "grad_norm": 2.2937467098236084,
+      "learning_rate": 0.001,
+      "loss": 2.1206,
+      "step": 3687
+    },
+    {
+      "epoch": 0.15601996784837974,
+      "grad_norm": 0.2151869237422943,
+      "learning_rate": 0.001,
+      "loss": 2.1484,
+      "step": 3688
+    },
+    {
+      "epoch": 0.1560622726118961,
+      "grad_norm": 0.2432985007762909,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 3689
+    },
+    {
+      "epoch": 0.15610457737541247,
+      "grad_norm": 0.26198938488960266,
+      "learning_rate": 0.001,
+      "loss": 2.4545,
+      "step": 3690
+    },
+    {
+      "epoch": 0.15614688213892886,
+      "grad_norm": 0.2777709364891052,
+      "learning_rate": 0.001,
+      "loss": 2.664,
+      "step": 3691
+    },
+    {
+      "epoch": 0.1561891869024452,
+      "grad_norm": 0.28329089283943176,
+      "learning_rate": 0.001,
+      "loss": 2.5311,
+      "step": 3692
+    },
+    {
+      "epoch": 0.1562314916659616,
+      "grad_norm": 0.3065381944179535,
+      "learning_rate": 0.001,
+      "loss": 2.1369,
+      "step": 3693
+    },
+    {
+      "epoch": 0.15627379642947795,
+      "grad_norm": 0.28511857986450195,
+      "learning_rate": 0.001,
+      "loss": 2.6129,
+      "step": 3694
+    },
+    {
+      "epoch": 0.15631610119299433,
+      "grad_norm": 0.24408607184886932,
+      "learning_rate": 0.001,
+      "loss": 2.0128,
+      "step": 3695
+    },
+    {
+      "epoch": 0.1563584059565107,
+      "grad_norm": 0.8473067879676819,
+      "learning_rate": 0.001,
+      "loss": 2.2145,
+      "step": 3696
+    },
+    {
+      "epoch": 0.15640071072002706,
+      "grad_norm": 0.42082521319389343,
+      "learning_rate": 0.001,
+      "loss": 2.1128,
+      "step": 3697
+    },
+    {
+      "epoch": 0.15644301548354345,
+      "grad_norm": 0.42995426058769226,
+      "learning_rate": 0.001,
+      "loss": 2.3262,
+      "step": 3698
+    },
+    {
+      "epoch": 0.15648532024705983,
+      "grad_norm": 0.3102007210254669,
+      "learning_rate": 0.001,
+      "loss": 2.7998,
+      "step": 3699
+    },
+    {
+      "epoch": 0.15652762501057618,
+      "grad_norm": 0.3263281285762787,
+      "learning_rate": 0.001,
+      "loss": 2.2744,
+      "step": 3700
+    },
+    {
+      "epoch": 0.15656992977409256,
+      "grad_norm": 2.9346940517425537,
+      "learning_rate": 0.001,
+      "loss": 2.1367,
+      "step": 3701
+    },
+    {
+      "epoch": 0.15661223453760895,
+      "grad_norm": 0.298091858625412,
+      "learning_rate": 0.001,
+      "loss": 3.0486,
+      "step": 3702
+    },
+    {
+      "epoch": 0.1566545393011253,
+      "grad_norm": 0.5562085509300232,
+      "learning_rate": 0.001,
+      "loss": 2.0666,
+      "step": 3703
+    },
+    {
+      "epoch": 0.15669684406464168,
+      "grad_norm": 0.26188915967941284,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 3704
+    },
+    {
+      "epoch": 0.15673914882815804,
+      "grad_norm": 0.3531795144081116,
+      "learning_rate": 0.001,
+      "loss": 2.6941,
+      "step": 3705
+    },
+    {
+      "epoch": 0.15678145359167442,
+      "grad_norm": 0.32536250352859497,
+      "learning_rate": 0.001,
+      "loss": 2.1217,
+      "step": 3706
+    },
+    {
+      "epoch": 0.1568237583551908,
+      "grad_norm": 0.2858486771583557,
+      "learning_rate": 0.001,
+      "loss": 2.2795,
+      "step": 3707
+    },
+    {
+      "epoch": 0.15686606311870716,
+      "grad_norm": 0.2625608742237091,
+      "learning_rate": 0.001,
+      "loss": 3.3786,
+      "step": 3708
+    },
+    {
+      "epoch": 0.15690836788222354,
+      "grad_norm": 0.2354310154914856,
+      "learning_rate": 0.001,
+      "loss": 2.0803,
+      "step": 3709
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 0.25832563638687134,
+      "learning_rate": 0.001,
+      "loss": 3.1712,
+      "step": 3710
+    },
+    {
+      "epoch": 0.15699297740925627,
+      "grad_norm": 2.9774281978607178,
+      "learning_rate": 0.001,
+      "loss": 2.6699,
+      "step": 3711
+    },
+    {
+      "epoch": 0.15703528217277266,
+      "grad_norm": 0.945013165473938,
+      "learning_rate": 0.001,
+      "loss": 3.3384,
+      "step": 3712
+    },
+    {
+      "epoch": 0.15707758693628904,
+      "grad_norm": 0.7904821634292603,
+      "learning_rate": 0.001,
+      "loss": 2.3788,
+      "step": 3713
+    },
+    {
+      "epoch": 0.1571198916998054,
+      "grad_norm": 0.2616831958293915,
+      "learning_rate": 0.001,
+      "loss": 2.8228,
+      "step": 3714
+    },
+    {
+      "epoch": 0.15716219646332177,
+      "grad_norm": 0.5680962204933167,
+      "learning_rate": 0.001,
+      "loss": 2.8964,
+      "step": 3715
+    },
+    {
+      "epoch": 0.15720450122683813,
+      "grad_norm": 1.2796063423156738,
+      "learning_rate": 0.001,
+      "loss": 2.6141,
+      "step": 3716
+    },
+    {
+      "epoch": 0.1572468059903545,
+      "grad_norm": 0.34902986884117126,
+      "learning_rate": 0.001,
+      "loss": 3.4248,
+      "step": 3717
+    },
+    {
+      "epoch": 0.1572891107538709,
+      "grad_norm": 0.30092769861221313,
+      "learning_rate": 0.001,
+      "loss": 2.7998,
+      "step": 3718
+    },
+    {
+      "epoch": 0.15733141551738725,
+      "grad_norm": 0.2545045018196106,
+      "learning_rate": 0.001,
+      "loss": 2.4546,
+      "step": 3719
+    },
+    {
+      "epoch": 0.15737372028090363,
+      "grad_norm": 0.30985644459724426,
+      "learning_rate": 0.001,
+      "loss": 2.1622,
+      "step": 3720
+    },
+    {
+      "epoch": 0.15741602504442,
+      "grad_norm": 0.21358482539653778,
+      "learning_rate": 0.001,
+      "loss": 1.9671,
+      "step": 3721
+    },
+    {
+      "epoch": 0.15745832980793636,
+      "grad_norm": 0.2403787076473236,
+      "learning_rate": 0.001,
+      "loss": 1.8596,
+      "step": 3722
+    },
+    {
+      "epoch": 0.15750063457145275,
+      "grad_norm": 0.25962314009666443,
+      "learning_rate": 0.001,
+      "loss": 2.595,
+      "step": 3723
+    },
+    {
+      "epoch": 0.15754293933496913,
+      "grad_norm": 0.9243412613868713,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 3724
+    },
+    {
+      "epoch": 0.15758524409848548,
+      "grad_norm": 0.2106955498456955,
+      "learning_rate": 0.001,
+      "loss": 3.3038,
+      "step": 3725
+    },
+    {
+      "epoch": 0.15762754886200186,
+      "grad_norm": 0.26728370785713196,
+      "learning_rate": 0.001,
+      "loss": 2.3843,
+      "step": 3726
+    },
+    {
+      "epoch": 0.15766985362551825,
+      "grad_norm": 0.26427310705184937,
+      "learning_rate": 0.001,
+      "loss": 2.1108,
+      "step": 3727
+    },
+    {
+      "epoch": 0.1577121583890346,
+      "grad_norm": 0.28308454155921936,
+      "learning_rate": 0.001,
+      "loss": 2.4893,
+      "step": 3728
+    },
+    {
+      "epoch": 0.15775446315255098,
+      "grad_norm": 0.266253262758255,
+      "learning_rate": 0.001,
+      "loss": 2.521,
+      "step": 3729
+    },
+    {
+      "epoch": 0.15779676791606734,
+      "grad_norm": 0.4611959159374237,
+      "learning_rate": 0.001,
+      "loss": 2.2577,
+      "step": 3730
+    },
+    {
+      "epoch": 0.15783907267958372,
+      "grad_norm": 0.3216572701931,
+      "learning_rate": 0.001,
+      "loss": 3.0579,
+      "step": 3731
+    },
+    {
+      "epoch": 0.1578813774431001,
+      "grad_norm": 0.2995474934577942,
+      "learning_rate": 0.001,
+      "loss": 2.8257,
+      "step": 3732
+    },
+    {
+      "epoch": 0.15792368220661646,
+      "grad_norm": 0.704836368560791,
+      "learning_rate": 0.001,
+      "loss": 2.7877,
+      "step": 3733
+    },
+    {
+      "epoch": 0.15796598697013284,
+      "grad_norm": 0.4456735849380493,
+      "learning_rate": 0.001,
+      "loss": 2.1902,
+      "step": 3734
+    },
+    {
+      "epoch": 0.15800829173364922,
+      "grad_norm": 0.21803617477416992,
+      "learning_rate": 0.001,
+      "loss": 2.4335,
+      "step": 3735
+    },
+    {
+      "epoch": 0.15805059649716557,
+      "grad_norm": 0.24227195978164673,
+      "learning_rate": 0.001,
+      "loss": 1.9287,
+      "step": 3736
+    },
+    {
+      "epoch": 0.15809290126068196,
+      "grad_norm": 0.2500062882900238,
+      "learning_rate": 0.001,
+      "loss": 2.7866,
+      "step": 3737
+    },
+    {
+      "epoch": 0.15813520602419834,
+      "grad_norm": 0.24755047261714935,
+      "learning_rate": 0.001,
+      "loss": 2.8725,
+      "step": 3738
+    },
+    {
+      "epoch": 0.1581775107877147,
+      "grad_norm": 0.26385197043418884,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 3739
+    },
+    {
+      "epoch": 0.15821981555123107,
+      "grad_norm": 0.3932577967643738,
+      "learning_rate": 0.001,
+      "loss": 2.5877,
+      "step": 3740
+    },
+    {
+      "epoch": 0.15826212031474743,
+      "grad_norm": 0.21887458860874176,
+      "learning_rate": 0.001,
+      "loss": 2.3305,
+      "step": 3741
+    },
+    {
+      "epoch": 0.1583044250782638,
+      "grad_norm": 0.2982005774974823,
+      "learning_rate": 0.001,
+      "loss": 3.3444,
+      "step": 3742
+    },
+    {
+      "epoch": 0.1583467298417802,
+      "grad_norm": 0.4082737863063812,
+      "learning_rate": 0.001,
+      "loss": 3.0144,
+      "step": 3743
+    },
+    {
+      "epoch": 0.15838903460529655,
+      "grad_norm": 0.4862198829650879,
+      "learning_rate": 0.001,
+      "loss": 2.7298,
+      "step": 3744
+    },
+    {
+      "epoch": 0.15843133936881293,
+      "grad_norm": 0.2732762098312378,
+      "learning_rate": 0.001,
+      "loss": 2.9059,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1584736441323293,
+      "grad_norm": 0.2768647372722626,
+      "learning_rate": 0.001,
+      "loss": 2.005,
+      "step": 3746
+    },
+    {
+      "epoch": 0.15851594889584567,
+      "grad_norm": 0.2584654688835144,
+      "learning_rate": 0.001,
+      "loss": 2.2505,
+      "step": 3747
+    },
+    {
+      "epoch": 0.15855825365936205,
+      "grad_norm": 0.2902711033821106,
+      "learning_rate": 0.001,
+      "loss": 2.4176,
+      "step": 3748
+    },
+    {
+      "epoch": 0.15860055842287843,
+      "grad_norm": 0.2150590866804123,
+      "learning_rate": 0.001,
+      "loss": 1.8724,
+      "step": 3749
+    },
+    {
+      "epoch": 0.15864286318639478,
+      "grad_norm": 1.3695828914642334,
+      "learning_rate": 0.001,
+      "loss": 1.9712,
+      "step": 3750
+    },
+    {
+      "epoch": 0.15868516794991117,
+      "grad_norm": 0.3454398810863495,
+      "learning_rate": 0.001,
+      "loss": 2.6138,
+      "step": 3751
+    },
+    {
+      "epoch": 0.15872747271342752,
+      "grad_norm": 0.7297479510307312,
+      "learning_rate": 0.001,
+      "loss": 3.5123,
+      "step": 3752
+    },
+    {
+      "epoch": 0.1587697774769439,
+      "grad_norm": 0.43115225434303284,
+      "learning_rate": 0.001,
+      "loss": 1.886,
+      "step": 3753
+    },
+    {
+      "epoch": 0.15881208224046028,
+      "grad_norm": 0.2626439929008484,
+      "learning_rate": 0.001,
+      "loss": 1.6985,
+      "step": 3754
+    },
+    {
+      "epoch": 0.15885438700397664,
+      "grad_norm": 1.067133903503418,
+      "learning_rate": 0.001,
+      "loss": 2.0347,
+      "step": 3755
+    },
+    {
+      "epoch": 0.15889669176749302,
+      "grad_norm": 0.28213247656822205,
+      "learning_rate": 0.001,
+      "loss": 2.6482,
+      "step": 3756
+    },
+    {
+      "epoch": 0.1589389965310094,
+      "grad_norm": 0.3096356689929962,
+      "learning_rate": 0.001,
+      "loss": 2.4303,
+      "step": 3757
+    },
+    {
+      "epoch": 0.15898130129452576,
+      "grad_norm": 1.1172544956207275,
+      "learning_rate": 0.001,
+      "loss": 1.4872,
+      "step": 3758
+    },
+    {
+      "epoch": 0.15902360605804214,
+      "grad_norm": 0.5289722681045532,
+      "learning_rate": 0.001,
+      "loss": 2.2845,
+      "step": 3759
+    },
+    {
+      "epoch": 0.15906591082155852,
+      "grad_norm": 0.41261282563209534,
+      "learning_rate": 0.001,
+      "loss": 2.7228,
+      "step": 3760
+    },
+    {
+      "epoch": 0.15910821558507487,
+      "grad_norm": 0.2809741199016571,
+      "learning_rate": 0.001,
+      "loss": 2.4155,
+      "step": 3761
+    },
+    {
+      "epoch": 0.15915052034859126,
+      "grad_norm": 0.6320580840110779,
+      "learning_rate": 0.001,
+      "loss": 1.6108,
+      "step": 3762
+    },
+    {
+      "epoch": 0.1591928251121076,
+      "grad_norm": 0.2594701051712036,
+      "learning_rate": 0.001,
+      "loss": 1.7907,
+      "step": 3763
+    },
+    {
+      "epoch": 0.159235129875624,
+      "grad_norm": 0.266510933637619,
+      "learning_rate": 0.001,
+      "loss": 1.9833,
+      "step": 3764
+    },
+    {
+      "epoch": 0.15927743463914038,
+      "grad_norm": 0.24247369170188904,
+      "learning_rate": 0.001,
+      "loss": 2.0623,
+      "step": 3765
+    },
+    {
+      "epoch": 0.15931973940265673,
+      "grad_norm": 0.23818761110305786,
+      "learning_rate": 0.001,
+      "loss": 2.2446,
+      "step": 3766
+    },
+    {
+      "epoch": 0.1593620441661731,
+      "grad_norm": 0.27693408727645874,
+      "learning_rate": 0.001,
+      "loss": 2.7074,
+      "step": 3767
+    },
+    {
+      "epoch": 0.1594043489296895,
+      "grad_norm": 0.6779216527938843,
+      "learning_rate": 0.001,
+      "loss": 2.1112,
+      "step": 3768
+    },
+    {
+      "epoch": 0.15944665369320585,
+      "grad_norm": 0.5291603207588196,
+      "learning_rate": 0.001,
+      "loss": 2.0643,
+      "step": 3769
+    },
+    {
+      "epoch": 0.15948895845672223,
+      "grad_norm": 0.2355789989233017,
+      "learning_rate": 0.001,
+      "loss": 2.9922,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1595312632202386,
+      "grad_norm": 0.334757536649704,
+      "learning_rate": 0.001,
+      "loss": 3.6081,
+      "step": 3771
+    },
+    {
+      "epoch": 0.15957356798375497,
+      "grad_norm": 0.2403586208820343,
+      "learning_rate": 0.001,
+      "loss": 1.8171,
+      "step": 3772
+    },
+    {
+      "epoch": 0.15961587274727135,
+      "grad_norm": 0.697832465171814,
+      "learning_rate": 0.001,
+      "loss": 2.3378,
+      "step": 3773
+    },
+    {
+      "epoch": 0.1596581775107877,
+      "grad_norm": 0.7586553692817688,
+      "learning_rate": 0.001,
+      "loss": 2.3106,
+      "step": 3774
+    },
+    {
+      "epoch": 0.15970048227430408,
+      "grad_norm": 0.365491658449173,
+      "learning_rate": 0.001,
+      "loss": 3.1585,
+      "step": 3775
+    },
+    {
+      "epoch": 0.15974278703782047,
+      "grad_norm": 1.0467041730880737,
+      "learning_rate": 0.001,
+      "loss": 3.8686,
+      "step": 3776
+    },
+    {
+      "epoch": 0.15978509180133682,
+      "grad_norm": 0.23686069250106812,
+      "learning_rate": 0.001,
+      "loss": 3.3134,
+      "step": 3777
+    },
+    {
+      "epoch": 0.1598273965648532,
+      "grad_norm": 0.2806885838508606,
+      "learning_rate": 0.001,
+      "loss": 2.4206,
+      "step": 3778
+    },
+    {
+      "epoch": 0.15986970132836958,
+      "grad_norm": 0.25082680583000183,
+      "learning_rate": 0.001,
+      "loss": 2.3114,
+      "step": 3779
+    },
+    {
+      "epoch": 0.15991200609188594,
+      "grad_norm": 0.22670866549015045,
+      "learning_rate": 0.001,
+      "loss": 2.852,
+      "step": 3780
+    },
+    {
+      "epoch": 0.15995431085540232,
+      "grad_norm": 0.2668560743331909,
+      "learning_rate": 0.001,
+      "loss": 2.3196,
+      "step": 3781
+    },
+    {
+      "epoch": 0.1599966156189187,
+      "grad_norm": 1.557098150253296,
+      "learning_rate": 0.001,
+      "loss": 1.6908,
+      "step": 3782
+    },
+    {
+      "epoch": 0.16003892038243506,
+      "grad_norm": 0.22779476642608643,
+      "learning_rate": 0.001,
+      "loss": 3.2161,
+      "step": 3783
+    },
+    {
+      "epoch": 0.16008122514595144,
+      "grad_norm": 0.777952253818512,
+      "learning_rate": 0.001,
+      "loss": 1.5536,
+      "step": 3784
+    },
+    {
+      "epoch": 0.1601235299094678,
+      "grad_norm": 1.5574365854263306,
+      "learning_rate": 0.001,
+      "loss": 2.5072,
+      "step": 3785
+    },
+    {
+      "epoch": 0.16016583467298418,
+      "grad_norm": 0.299005925655365,
+      "learning_rate": 0.001,
+      "loss": 3.0707,
+      "step": 3786
+    },
+    {
+      "epoch": 0.16020813943650056,
+      "grad_norm": 0.3228248953819275,
+      "learning_rate": 0.001,
+      "loss": 2.3951,
+      "step": 3787
+    },
+    {
+      "epoch": 0.1602504442000169,
+      "grad_norm": 1.0663105249404907,
+      "learning_rate": 0.001,
+      "loss": 2.2717,
+      "step": 3788
+    },
+    {
+      "epoch": 0.1602927489635333,
+      "grad_norm": 3.4064865112304688,
+      "learning_rate": 0.001,
+      "loss": 2.4347,
+      "step": 3789
+    },
+    {
+      "epoch": 0.16033505372704968,
+      "grad_norm": 0.3657093048095703,
+      "learning_rate": 0.001,
+      "loss": 2.6096,
+      "step": 3790
+    },
+    {
+      "epoch": 0.16037735849056603,
+      "grad_norm": 0.3097366690635681,
+      "learning_rate": 0.001,
+      "loss": 2.0905,
+      "step": 3791
+    },
+    {
+      "epoch": 0.1604196632540824,
+      "grad_norm": 0.35716721415519714,
+      "learning_rate": 0.001,
+      "loss": 2.5812,
+      "step": 3792
+    },
+    {
+      "epoch": 0.1604619680175988,
+      "grad_norm": 0.32534101605415344,
+      "learning_rate": 0.001,
+      "loss": 2.9844,
+      "step": 3793
+    },
+    {
+      "epoch": 0.16050427278111515,
+      "grad_norm": 0.36417415738105774,
+      "learning_rate": 0.001,
+      "loss": 2.5366,
+      "step": 3794
+    },
+    {
+      "epoch": 0.16054657754463153,
+      "grad_norm": 0.26519954204559326,
+      "learning_rate": 0.001,
+      "loss": 2.7738,
+      "step": 3795
+    },
+    {
+      "epoch": 0.16058888230814788,
+      "grad_norm": 0.473812073469162,
+      "learning_rate": 0.001,
+      "loss": 2.2917,
+      "step": 3796
+    },
+    {
+      "epoch": 0.16063118707166427,
+      "grad_norm": 0.3712599277496338,
+      "learning_rate": 0.001,
+      "loss": 2.2126,
+      "step": 3797
+    },
+    {
+      "epoch": 0.16067349183518065,
+      "grad_norm": 0.30291634798049927,
+      "learning_rate": 0.001,
+      "loss": 2.1419,
+      "step": 3798
+    },
+    {
+      "epoch": 0.160715796598697,
+      "grad_norm": 0.30859270691871643,
+      "learning_rate": 0.001,
+      "loss": 1.9372,
+      "step": 3799
+    },
+    {
+      "epoch": 0.16075810136221338,
+      "grad_norm": 0.24209967255592346,
+      "learning_rate": 0.001,
+      "loss": 1.9439,
+      "step": 3800
+    },
+    {
+      "epoch": 0.16080040612572977,
+      "grad_norm": 0.30465030670166016,
+      "learning_rate": 0.001,
+      "loss": 3.1223,
+      "step": 3801
+    },
+    {
+      "epoch": 0.16084271088924612,
+      "grad_norm": 0.29678988456726074,
+      "learning_rate": 0.001,
+      "loss": 2.2062,
+      "step": 3802
+    },
+    {
+      "epoch": 0.1608850156527625,
+      "grad_norm": 0.21192920207977295,
+      "learning_rate": 0.001,
+      "loss": 2.1279,
+      "step": 3803
+    },
+    {
+      "epoch": 0.16092732041627889,
+      "grad_norm": 2.0357847213745117,
+      "learning_rate": 0.001,
+      "loss": 3.1239,
+      "step": 3804
+    },
+    {
+      "epoch": 0.16096962517979524,
+      "grad_norm": 0.31351718306541443,
+      "learning_rate": 0.001,
+      "loss": 2.176,
+      "step": 3805
+    },
+    {
+      "epoch": 0.16101192994331162,
+      "grad_norm": 0.3054656684398651,
+      "learning_rate": 0.001,
+      "loss": 2.5238,
+      "step": 3806
+    },
+    {
+      "epoch": 0.16105423470682798,
+      "grad_norm": 0.5278902053833008,
+      "learning_rate": 0.001,
+      "loss": 1.6492,
+      "step": 3807
+    },
+    {
+      "epoch": 0.16109653947034436,
+      "grad_norm": 0.300001323223114,
+      "learning_rate": 0.001,
+      "loss": 2.562,
+      "step": 3808
+    },
+    {
+      "epoch": 0.16113884423386074,
+      "grad_norm": 0.34436389803886414,
+      "learning_rate": 0.001,
+      "loss": 3.3102,
+      "step": 3809
+    },
+    {
+      "epoch": 0.1611811489973771,
+      "grad_norm": 0.33348026871681213,
+      "learning_rate": 0.001,
+      "loss": 3.1311,
+      "step": 3810
+    },
+    {
+      "epoch": 0.16122345376089348,
+      "grad_norm": 0.24410240352153778,
+      "learning_rate": 0.001,
+      "loss": 3.0828,
+      "step": 3811
+    },
+    {
+      "epoch": 0.16126575852440986,
+      "grad_norm": 0.21113960444927216,
+      "learning_rate": 0.001,
+      "loss": 1.9945,
+      "step": 3812
+    },
+    {
+      "epoch": 0.1613080632879262,
+      "grad_norm": 0.2360842376947403,
+      "learning_rate": 0.001,
+      "loss": 1.9214,
+      "step": 3813
+    },
+    {
+      "epoch": 0.1613503680514426,
+      "grad_norm": 0.3132474720478058,
+      "learning_rate": 0.001,
+      "loss": 3.3039,
+      "step": 3814
+    },
+    {
+      "epoch": 0.16139267281495898,
+      "grad_norm": 0.23220689594745636,
+      "learning_rate": 0.001,
+      "loss": 2.3121,
+      "step": 3815
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.3871288001537323,
+      "learning_rate": 0.001,
+      "loss": 3.1045,
+      "step": 3816
+    },
+    {
+      "epoch": 0.1614772823419917,
+      "grad_norm": 0.2573937773704529,
+      "learning_rate": 0.001,
+      "loss": 2.2547,
+      "step": 3817
+    },
+    {
+      "epoch": 0.16151958710550807,
+      "grad_norm": 0.23230381309986115,
+      "learning_rate": 0.001,
+      "loss": 1.9308,
+      "step": 3818
+    },
+    {
+      "epoch": 0.16156189186902445,
+      "grad_norm": 0.2752211093902588,
+      "learning_rate": 0.001,
+      "loss": 3.6371,
+      "step": 3819
+    },
+    {
+      "epoch": 0.16160419663254083,
+      "grad_norm": 0.2206379473209381,
+      "learning_rate": 0.001,
+      "loss": 3.6659,
+      "step": 3820
+    },
+    {
+      "epoch": 0.16164650139605719,
+      "grad_norm": 0.5794355869293213,
+      "learning_rate": 0.001,
+      "loss": 2.3093,
+      "step": 3821
+    },
+    {
+      "epoch": 0.16168880615957357,
+      "grad_norm": 0.23683209717273712,
+      "learning_rate": 0.001,
+      "loss": 2.5634,
+      "step": 3822
+    },
+    {
+      "epoch": 0.16173111092308995,
+      "grad_norm": 0.2306908518075943,
+      "learning_rate": 0.001,
+      "loss": 1.9803,
+      "step": 3823
+    },
+    {
+      "epoch": 0.1617734156866063,
+      "grad_norm": 0.22796887159347534,
+      "learning_rate": 0.001,
+      "loss": 2.3285,
+      "step": 3824
+    },
+    {
+      "epoch": 0.16181572045012269,
+      "grad_norm": 0.27235713601112366,
+      "learning_rate": 0.001,
+      "loss": 2.6457,
+      "step": 3825
+    },
+    {
+      "epoch": 0.16185802521363907,
+      "grad_norm": 1.1337499618530273,
+      "learning_rate": 0.001,
+      "loss": 2.7345,
+      "step": 3826
+    },
+    {
+      "epoch": 0.16190032997715542,
+      "grad_norm": 0.22080039978027344,
+      "learning_rate": 0.001,
+      "loss": 2.0913,
+      "step": 3827
+    },
+    {
+      "epoch": 0.1619426347406718,
+      "grad_norm": 0.24313177168369293,
+      "learning_rate": 0.001,
+      "loss": 2.4784,
+      "step": 3828
+    },
+    {
+      "epoch": 0.16198493950418816,
+      "grad_norm": 0.28508949279785156,
+      "learning_rate": 0.001,
+      "loss": 2.8932,
+      "step": 3829
+    },
+    {
+      "epoch": 0.16202724426770454,
+      "grad_norm": 0.37626540660858154,
+      "learning_rate": 0.001,
+      "loss": 2.596,
+      "step": 3830
+    },
+    {
+      "epoch": 0.16206954903122092,
+      "grad_norm": 0.22318224608898163,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 3831
+    },
+    {
+      "epoch": 0.16211185379473728,
+      "grad_norm": 0.2062063217163086,
+      "learning_rate": 0.001,
+      "loss": 1.5958,
+      "step": 3832
+    },
+    {
+      "epoch": 0.16215415855825366,
+      "grad_norm": 0.2109755128622055,
+      "learning_rate": 0.001,
+      "loss": 2.1923,
+      "step": 3833
+    },
+    {
+      "epoch": 0.16219646332177004,
+      "grad_norm": 0.21446409821510315,
+      "learning_rate": 0.001,
+      "loss": 2.1621,
+      "step": 3834
+    },
+    {
+      "epoch": 0.1622387680852864,
+      "grad_norm": 0.21408231556415558,
+      "learning_rate": 0.001,
+      "loss": 2.683,
+      "step": 3835
+    },
+    {
+      "epoch": 0.16228107284880278,
+      "grad_norm": 0.2532133460044861,
+      "learning_rate": 0.001,
+      "loss": 2.2158,
+      "step": 3836
+    },
+    {
+      "epoch": 0.16232337761231916,
+      "grad_norm": 0.27173325419425964,
+      "learning_rate": 0.001,
+      "loss": 2.2614,
+      "step": 3837
+    },
+    {
+      "epoch": 0.1623656823758355,
+      "grad_norm": 0.2316475659608841,
+      "learning_rate": 0.001,
+      "loss": 2.1505,
+      "step": 3838
+    },
+    {
+      "epoch": 0.1624079871393519,
+      "grad_norm": 0.23597820103168488,
+      "learning_rate": 0.001,
+      "loss": 2.2652,
+      "step": 3839
+    },
+    {
+      "epoch": 0.16245029190286828,
+      "grad_norm": 0.23943273723125458,
+      "learning_rate": 0.001,
+      "loss": 2.0347,
+      "step": 3840
+    },
+    {
+      "epoch": 0.16249259666638463,
+      "grad_norm": 0.8113381266593933,
+      "learning_rate": 0.001,
+      "loss": 2.0472,
+      "step": 3841
+    },
+    {
+      "epoch": 0.162534901429901,
+      "grad_norm": 2.3139660358428955,
+      "learning_rate": 0.001,
+      "loss": 1.8397,
+      "step": 3842
+    },
+    {
+      "epoch": 0.16257720619341737,
+      "grad_norm": 0.42972567677497864,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 3843
+    },
+    {
+      "epoch": 0.16261951095693375,
+      "grad_norm": 0.2802884578704834,
+      "learning_rate": 0.001,
+      "loss": 1.9931,
+      "step": 3844
+    },
+    {
+      "epoch": 0.16266181572045013,
+      "grad_norm": 0.3165445625782013,
+      "learning_rate": 0.001,
+      "loss": 2.3039,
+      "step": 3845
+    },
+    {
+      "epoch": 0.16270412048396649,
+      "grad_norm": 0.5963391661643982,
+      "learning_rate": 0.001,
+      "loss": 1.952,
+      "step": 3846
+    },
+    {
+      "epoch": 0.16274642524748287,
+      "grad_norm": 0.25218188762664795,
+      "learning_rate": 0.001,
+      "loss": 2.3469,
+      "step": 3847
+    },
+    {
+      "epoch": 0.16278873001099925,
+      "grad_norm": 0.31111615896224976,
+      "learning_rate": 0.001,
+      "loss": 2.0224,
+      "step": 3848
+    },
+    {
+      "epoch": 0.1628310347745156,
+      "grad_norm": 0.3019029200077057,
+      "learning_rate": 0.001,
+      "loss": 2.4928,
+      "step": 3849
+    },
+    {
+      "epoch": 0.16287333953803199,
+      "grad_norm": 0.22773532569408417,
+      "learning_rate": 0.001,
+      "loss": 2.9932,
+      "step": 3850
+    },
+    {
+      "epoch": 0.16291564430154837,
+      "grad_norm": 0.2483637034893036,
+      "learning_rate": 0.001,
+      "loss": 2.8444,
+      "step": 3851
+    },
+    {
+      "epoch": 0.16295794906506472,
+      "grad_norm": 0.2056424915790558,
+      "learning_rate": 0.001,
+      "loss": 1.9696,
+      "step": 3852
+    },
+    {
+      "epoch": 0.1630002538285811,
+      "grad_norm": 0.22549600899219513,
+      "learning_rate": 0.001,
+      "loss": 1.7296,
+      "step": 3853
+    },
+    {
+      "epoch": 0.16304255859209746,
+      "grad_norm": 0.2797006368637085,
+      "learning_rate": 0.001,
+      "loss": 2.0571,
+      "step": 3854
+    },
+    {
+      "epoch": 0.16308486335561384,
+      "grad_norm": 0.2678796947002411,
+      "learning_rate": 0.001,
+      "loss": 1.9062,
+      "step": 3855
+    },
+    {
+      "epoch": 0.16312716811913022,
+      "grad_norm": 0.24409295618534088,
+      "learning_rate": 0.001,
+      "loss": 2.4397,
+      "step": 3856
+    },
+    {
+      "epoch": 0.16316947288264658,
+      "grad_norm": 0.25241202116012573,
+      "learning_rate": 0.001,
+      "loss": 2.1752,
+      "step": 3857
+    },
+    {
+      "epoch": 0.16321177764616296,
+      "grad_norm": 0.23605115711688995,
+      "learning_rate": 0.001,
+      "loss": 2.9556,
+      "step": 3858
+    },
+    {
+      "epoch": 0.16325408240967934,
+      "grad_norm": 0.26487863063812256,
+      "learning_rate": 0.001,
+      "loss": 1.9874,
+      "step": 3859
+    },
+    {
+      "epoch": 0.1632963871731957,
+      "grad_norm": 0.24517613649368286,
+      "learning_rate": 0.001,
+      "loss": 2.321,
+      "step": 3860
+    },
+    {
+      "epoch": 0.16333869193671208,
+      "grad_norm": 0.20442011952400208,
+      "learning_rate": 0.001,
+      "loss": 1.8018,
+      "step": 3861
+    },
+    {
+      "epoch": 0.16338099670022846,
+      "grad_norm": 0.24029582738876343,
+      "learning_rate": 0.001,
+      "loss": 2.722,
+      "step": 3862
+    },
+    {
+      "epoch": 0.1634233014637448,
+      "grad_norm": 0.2389087826013565,
+      "learning_rate": 0.001,
+      "loss": 1.9895,
+      "step": 3863
+    },
+    {
+      "epoch": 0.1634656062272612,
+      "grad_norm": 0.2552791237831116,
+      "learning_rate": 0.001,
+      "loss": 2.4092,
+      "step": 3864
+    },
+    {
+      "epoch": 0.16350791099077755,
+      "grad_norm": 0.20444203913211823,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 3865
+    },
+    {
+      "epoch": 0.16355021575429393,
+      "grad_norm": 1.1959320306777954,
+      "learning_rate": 0.001,
+      "loss": 2.6259,
+      "step": 3866
+    },
+    {
+      "epoch": 0.1635925205178103,
+      "grad_norm": 0.20237359404563904,
+      "learning_rate": 0.001,
+      "loss": 1.7948,
+      "step": 3867
+    },
+    {
+      "epoch": 0.16363482528132667,
+      "grad_norm": 0.21189026534557343,
+      "learning_rate": 0.001,
+      "loss": 2.6693,
+      "step": 3868
+    },
+    {
+      "epoch": 0.16367713004484305,
+      "grad_norm": 0.25462326407432556,
+      "learning_rate": 0.001,
+      "loss": 1.9829,
+      "step": 3869
+    },
+    {
+      "epoch": 0.16371943480835943,
+      "grad_norm": 0.24201983213424683,
+      "learning_rate": 0.001,
+      "loss": 2.1617,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1637617395718758,
+      "grad_norm": 0.22964642941951752,
+      "learning_rate": 0.001,
+      "loss": 2.9446,
+      "step": 3871
+    },
+    {
+      "epoch": 0.16380404433539217,
+      "grad_norm": 0.2601517140865326,
+      "learning_rate": 0.001,
+      "loss": 2.3314,
+      "step": 3872
+    },
+    {
+      "epoch": 0.16384634909890855,
+      "grad_norm": 0.19910211861133575,
+      "learning_rate": 0.001,
+      "loss": 2.236,
+      "step": 3873
+    },
+    {
+      "epoch": 0.1638886538624249,
+      "grad_norm": 0.2370014637708664,
+      "learning_rate": 0.001,
+      "loss": 2.767,
+      "step": 3874
+    },
+    {
+      "epoch": 0.1639309586259413,
+      "grad_norm": 0.273014098405838,
+      "learning_rate": 0.001,
+      "loss": 2.8735,
+      "step": 3875
+    },
+    {
+      "epoch": 0.16397326338945764,
+      "grad_norm": 0.2136770635843277,
+      "learning_rate": 0.001,
+      "loss": 2.2062,
+      "step": 3876
+    },
+    {
+      "epoch": 0.16401556815297402,
+      "grad_norm": 0.196532741189003,
+      "learning_rate": 0.001,
+      "loss": 2.1683,
+      "step": 3877
+    },
+    {
+      "epoch": 0.1640578729164904,
+      "grad_norm": 1.7746950387954712,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 3878
+    },
+    {
+      "epoch": 0.16410017768000676,
+      "grad_norm": 0.29451486468315125,
+      "learning_rate": 0.001,
+      "loss": 2.2666,
+      "step": 3879
+    },
+    {
+      "epoch": 0.16414248244352314,
+      "grad_norm": 0.6255790591239929,
+      "learning_rate": 0.001,
+      "loss": 2.7286,
+      "step": 3880
+    },
+    {
+      "epoch": 0.16418478720703952,
+      "grad_norm": 0.49284324049949646,
+      "learning_rate": 0.001,
+      "loss": 2.0154,
+      "step": 3881
+    },
+    {
+      "epoch": 0.16422709197055588,
+      "grad_norm": 0.27464428544044495,
+      "learning_rate": 0.001,
+      "loss": 2.347,
+      "step": 3882
+    },
+    {
+      "epoch": 0.16426939673407226,
+      "grad_norm": 0.2803431451320648,
+      "learning_rate": 0.001,
+      "loss": 2.6582,
+      "step": 3883
+    },
+    {
+      "epoch": 0.16431170149758864,
+      "grad_norm": 0.2514314353466034,
+      "learning_rate": 0.001,
+      "loss": 2.6842,
+      "step": 3884
+    },
+    {
+      "epoch": 0.164354006261105,
+      "grad_norm": 0.30920740962028503,
+      "learning_rate": 0.001,
+      "loss": 2.5913,
+      "step": 3885
+    },
+    {
+      "epoch": 0.16439631102462138,
+      "grad_norm": 0.6570300459861755,
+      "learning_rate": 0.001,
+      "loss": 2.4439,
+      "step": 3886
+    },
+    {
+      "epoch": 0.16443861578813773,
+      "grad_norm": 0.24634666740894318,
+      "learning_rate": 0.001,
+      "loss": 2.1094,
+      "step": 3887
+    },
+    {
+      "epoch": 0.16448092055165411,
+      "grad_norm": 0.2316695600748062,
+      "learning_rate": 0.001,
+      "loss": 2.3292,
+      "step": 3888
+    },
+    {
+      "epoch": 0.1645232253151705,
+      "grad_norm": 0.6251176595687866,
+      "learning_rate": 0.001,
+      "loss": 2.9064,
+      "step": 3889
+    },
+    {
+      "epoch": 0.16456553007868685,
+      "grad_norm": 0.28041884303092957,
+      "learning_rate": 0.001,
+      "loss": 2.4278,
+      "step": 3890
+    },
+    {
+      "epoch": 0.16460783484220323,
+      "grad_norm": 0.6248710751533508,
+      "learning_rate": 0.001,
+      "loss": 3.7971,
+      "step": 3891
+    },
+    {
+      "epoch": 0.16465013960571961,
+      "grad_norm": 0.24960578978061676,
+      "learning_rate": 0.001,
+      "loss": 3.1562,
+      "step": 3892
+    },
+    {
+      "epoch": 0.16469244436923597,
+      "grad_norm": 0.24922244250774384,
+      "learning_rate": 0.001,
+      "loss": 3.1838,
+      "step": 3893
+    },
+    {
+      "epoch": 0.16473474913275235,
+      "grad_norm": 0.21565461158752441,
+      "learning_rate": 0.001,
+      "loss": 2.5373,
+      "step": 3894
+    },
+    {
+      "epoch": 0.16477705389626873,
+      "grad_norm": 0.30223679542541504,
+      "learning_rate": 0.001,
+      "loss": 1.9657,
+      "step": 3895
+    },
+    {
+      "epoch": 0.1648193586597851,
+      "grad_norm": 0.23021796345710754,
+      "learning_rate": 0.001,
+      "loss": 2.0286,
+      "step": 3896
+    },
+    {
+      "epoch": 0.16486166342330147,
+      "grad_norm": 0.22594904899597168,
+      "learning_rate": 0.001,
+      "loss": 2.0249,
+      "step": 3897
+    },
+    {
+      "epoch": 0.16490396818681782,
+      "grad_norm": 0.20173117518424988,
+      "learning_rate": 0.001,
+      "loss": 2.8041,
+      "step": 3898
+    },
+    {
+      "epoch": 0.1649462729503342,
+      "grad_norm": 0.2069264054298401,
+      "learning_rate": 0.001,
+      "loss": 1.9167,
+      "step": 3899
+    },
+    {
+      "epoch": 0.1649885777138506,
+      "grad_norm": 0.4964730739593506,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 3900
+    },
+    {
+      "epoch": 0.16503088247736694,
+      "grad_norm": 0.2436501681804657,
+      "learning_rate": 0.001,
+      "loss": 1.9212,
+      "step": 3901
+    },
+    {
+      "epoch": 0.16507318724088332,
+      "grad_norm": 0.27425655722618103,
+      "learning_rate": 0.001,
+      "loss": 2.2874,
+      "step": 3902
+    },
+    {
+      "epoch": 0.1651154920043997,
+      "grad_norm": 0.18915478885173798,
+      "learning_rate": 0.001,
+      "loss": 1.7265,
+      "step": 3903
+    },
+    {
+      "epoch": 0.16515779676791606,
+      "grad_norm": 0.34025347232818604,
+      "learning_rate": 0.001,
+      "loss": 2.0228,
+      "step": 3904
+    },
+    {
+      "epoch": 0.16520010153143244,
+      "grad_norm": 0.3264094889163971,
+      "learning_rate": 0.001,
+      "loss": 2.3233,
+      "step": 3905
+    },
+    {
+      "epoch": 0.16524240629494882,
+      "grad_norm": 0.2108362913131714,
+      "learning_rate": 0.001,
+      "loss": 2.6092,
+      "step": 3906
+    },
+    {
+      "epoch": 0.16528471105846518,
+      "grad_norm": 0.27986034750938416,
+      "learning_rate": 0.001,
+      "loss": 1.6611,
+      "step": 3907
+    },
+    {
+      "epoch": 0.16532701582198156,
+      "grad_norm": 0.22576811909675598,
+      "learning_rate": 0.001,
+      "loss": 1.8204,
+      "step": 3908
+    },
+    {
+      "epoch": 0.16536932058549791,
+      "grad_norm": 0.5198656916618347,
+      "learning_rate": 0.001,
+      "loss": 2.44,
+      "step": 3909
+    },
+    {
+      "epoch": 0.1654116253490143,
+      "grad_norm": 0.24308837950229645,
+      "learning_rate": 0.001,
+      "loss": 2.2157,
+      "step": 3910
+    },
+    {
+      "epoch": 0.16545393011253068,
+      "grad_norm": 0.2985725402832031,
+      "learning_rate": 0.001,
+      "loss": 2.6967,
+      "step": 3911
+    },
+    {
+      "epoch": 0.16549623487604703,
+      "grad_norm": 0.2176986038684845,
+      "learning_rate": 0.001,
+      "loss": 1.9572,
+      "step": 3912
+    },
+    {
+      "epoch": 0.16553853963956341,
+      "grad_norm": 0.49889811873435974,
+      "learning_rate": 0.001,
+      "loss": 2.2016,
+      "step": 3913
+    },
+    {
+      "epoch": 0.1655808444030798,
+      "grad_norm": 0.23154956102371216,
+      "learning_rate": 0.001,
+      "loss": 2.0423,
+      "step": 3914
+    },
+    {
+      "epoch": 0.16562314916659615,
+      "grad_norm": 0.2753046751022339,
+      "learning_rate": 0.001,
+      "loss": 1.9889,
+      "step": 3915
+    },
+    {
+      "epoch": 0.16566545393011253,
+      "grad_norm": 0.19035251438617706,
+      "learning_rate": 0.001,
+      "loss": 2.0711,
+      "step": 3916
+    },
+    {
+      "epoch": 0.16570775869362891,
+      "grad_norm": 0.23184069991111755,
+      "learning_rate": 0.001,
+      "loss": 1.9129,
+      "step": 3917
+    },
+    {
+      "epoch": 0.16575006345714527,
+      "grad_norm": 0.20346291363239288,
+      "learning_rate": 0.001,
+      "loss": 2.411,
+      "step": 3918
+    },
+    {
+      "epoch": 0.16579236822066165,
+      "grad_norm": 0.24841812252998352,
+      "learning_rate": 0.001,
+      "loss": 1.8754,
+      "step": 3919
+    },
+    {
+      "epoch": 0.165834672984178,
+      "grad_norm": 0.3259766697883606,
+      "learning_rate": 0.001,
+      "loss": 2.4933,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1658769777476944,
+      "grad_norm": 0.27737030386924744,
+      "learning_rate": 0.001,
+      "loss": 2.1435,
+      "step": 3921
+    },
+    {
+      "epoch": 0.16591928251121077,
+      "grad_norm": 1.7625619173049927,
+      "learning_rate": 0.001,
+      "loss": 2.05,
+      "step": 3922
+    },
+    {
+      "epoch": 0.16596158727472712,
+      "grad_norm": 0.24628551304340363,
+      "learning_rate": 0.001,
+      "loss": 2.3493,
+      "step": 3923
+    },
+    {
+      "epoch": 0.1660038920382435,
+      "grad_norm": 0.29456469416618347,
+      "learning_rate": 0.001,
+      "loss": 2.2105,
+      "step": 3924
+    },
+    {
+      "epoch": 0.1660461968017599,
+      "grad_norm": 0.20997002720832825,
+      "learning_rate": 0.001,
+      "loss": 1.8444,
+      "step": 3925
+    },
+    {
+      "epoch": 0.16608850156527624,
+      "grad_norm": 2.3074915409088135,
+      "learning_rate": 0.001,
+      "loss": 2.2948,
+      "step": 3926
+    },
+    {
+      "epoch": 0.16613080632879262,
+      "grad_norm": 0.3753257691860199,
+      "learning_rate": 0.001,
+      "loss": 3.4608,
+      "step": 3927
+    },
+    {
+      "epoch": 0.166173111092309,
+      "grad_norm": 0.32698577642440796,
+      "learning_rate": 0.001,
+      "loss": 2.5853,
+      "step": 3928
+    },
+    {
+      "epoch": 0.16621541585582536,
+      "grad_norm": 0.22866463661193848,
+      "learning_rate": 0.001,
+      "loss": 2.7564,
+      "step": 3929
+    },
+    {
+      "epoch": 0.16625772061934174,
+      "grad_norm": 0.30471017956733704,
+      "learning_rate": 0.001,
+      "loss": 3.0827,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1663000253828581,
+      "grad_norm": 0.2387181669473648,
+      "learning_rate": 0.001,
+      "loss": 2.8088,
+      "step": 3931
+    },
+    {
+      "epoch": 0.16634233014637448,
+      "grad_norm": 0.2087489813566208,
+      "learning_rate": 0.001,
+      "loss": 1.9849,
+      "step": 3932
+    },
+    {
+      "epoch": 0.16638463490989086,
+      "grad_norm": 0.22151561081409454,
+      "learning_rate": 0.001,
+      "loss": 2.8208,
+      "step": 3933
+    },
+    {
+      "epoch": 0.16642693967340721,
+      "grad_norm": 0.21167586743831635,
+      "learning_rate": 0.001,
+      "loss": 2.9183,
+      "step": 3934
+    },
+    {
+      "epoch": 0.1664692444369236,
+      "grad_norm": 0.2525525391101837,
+      "learning_rate": 0.001,
+      "loss": 3.3236,
+      "step": 3935
+    },
+    {
+      "epoch": 0.16651154920043998,
+      "grad_norm": 0.26379460096359253,
+      "learning_rate": 0.001,
+      "loss": 2.4491,
+      "step": 3936
+    },
+    {
+      "epoch": 0.16655385396395633,
+      "grad_norm": 0.2644009590148926,
+      "learning_rate": 0.001,
+      "loss": 2.4049,
+      "step": 3937
+    },
+    {
+      "epoch": 0.16659615872747272,
+      "grad_norm": 9.060093879699707,
+      "learning_rate": 0.001,
+      "loss": 2.6395,
+      "step": 3938
+    },
+    {
+      "epoch": 0.1666384634909891,
+      "grad_norm": 0.22047647833824158,
+      "learning_rate": 0.001,
+      "loss": 1.9525,
+      "step": 3939
+    },
+    {
+      "epoch": 0.16668076825450545,
+      "grad_norm": 0.5126127004623413,
+      "learning_rate": 0.001,
+      "loss": 1.6991,
+      "step": 3940
+    },
+    {
+      "epoch": 0.16672307301802183,
+      "grad_norm": 0.25247883796691895,
+      "learning_rate": 0.001,
+      "loss": 2.1926,
+      "step": 3941
+    },
+    {
+      "epoch": 0.1667653777815382,
+      "grad_norm": 0.25880908966064453,
+      "learning_rate": 0.001,
+      "loss": 2.1134,
+      "step": 3942
+    },
+    {
+      "epoch": 0.16680768254505457,
+      "grad_norm": 0.23852358758449554,
+      "learning_rate": 0.001,
+      "loss": 2.3002,
+      "step": 3943
+    },
+    {
+      "epoch": 0.16684998730857095,
+      "grad_norm": 0.20607604086399078,
+      "learning_rate": 0.001,
+      "loss": 1.8629,
+      "step": 3944
+    },
+    {
+      "epoch": 0.1668922920720873,
+      "grad_norm": 0.279757559299469,
+      "learning_rate": 0.001,
+      "loss": 3.2208,
+      "step": 3945
+    },
+    {
+      "epoch": 0.1669345968356037,
+      "grad_norm": 0.2418821156024933,
+      "learning_rate": 0.001,
+      "loss": 2.4583,
+      "step": 3946
+    },
+    {
+      "epoch": 0.16697690159912007,
+      "grad_norm": 0.19842597842216492,
+      "learning_rate": 0.001,
+      "loss": 2.4909,
+      "step": 3947
+    },
+    {
+      "epoch": 0.16701920636263642,
+      "grad_norm": 0.6748121976852417,
+      "learning_rate": 0.001,
+      "loss": 2.0367,
+      "step": 3948
+    },
+    {
+      "epoch": 0.1670615111261528,
+      "grad_norm": 0.3548758625984192,
+      "learning_rate": 0.001,
+      "loss": 2.2625,
+      "step": 3949
+    },
+    {
+      "epoch": 0.1671038158896692,
+      "grad_norm": 0.23522132635116577,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 3950
+    },
+    {
+      "epoch": 0.16714612065318554,
+      "grad_norm": 1.6471339464187622,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 3951
+    },
+    {
+      "epoch": 0.16718842541670192,
+      "grad_norm": 0.9741820693016052,
+      "learning_rate": 0.001,
+      "loss": 2.5284,
+      "step": 3952
+    },
+    {
+      "epoch": 0.16723073018021828,
+      "grad_norm": 0.2638908624649048,
+      "learning_rate": 0.001,
+      "loss": 3.4217,
+      "step": 3953
+    },
+    {
+      "epoch": 0.16727303494373466,
+      "grad_norm": 0.2375737726688385,
+      "learning_rate": 0.001,
+      "loss": 2.1506,
+      "step": 3954
+    },
+    {
+      "epoch": 0.16731533970725104,
+      "grad_norm": 0.24627238512039185,
+      "learning_rate": 0.001,
+      "loss": 2.8165,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1673576444707674,
+      "grad_norm": 0.5549525618553162,
+      "learning_rate": 0.001,
+      "loss": 2.7948,
+      "step": 3956
+    },
+    {
+      "epoch": 0.16739994923428378,
+      "grad_norm": 9.49412727355957,
+      "learning_rate": 0.001,
+      "loss": 2.6595,
+      "step": 3957
+    },
+    {
+      "epoch": 0.16744225399780016,
+      "grad_norm": 0.2718313932418823,
+      "learning_rate": 0.001,
+      "loss": 1.5429,
+      "step": 3958
+    },
+    {
+      "epoch": 0.16748455876131652,
+      "grad_norm": 0.2949593663215637,
+      "learning_rate": 0.001,
+      "loss": 2.1059,
+      "step": 3959
+    },
+    {
+      "epoch": 0.1675268635248329,
+      "grad_norm": 0.4094192385673523,
+      "learning_rate": 0.001,
+      "loss": 2.1578,
+      "step": 3960
+    },
+    {
+      "epoch": 0.16756916828834928,
+      "grad_norm": 0.23077575862407684,
+      "learning_rate": 0.001,
+      "loss": 1.7893,
+      "step": 3961
+    },
+    {
+      "epoch": 0.16761147305186563,
+      "grad_norm": 3.148542881011963,
+      "learning_rate": 0.001,
+      "loss": 2.0162,
+      "step": 3962
+    },
+    {
+      "epoch": 0.16765377781538202,
+      "grad_norm": 0.2932436764240265,
+      "learning_rate": 0.001,
+      "loss": 3.187,
+      "step": 3963
+    },
+    {
+      "epoch": 0.1676960825788984,
+      "grad_norm": 4.55126953125,
+      "learning_rate": 0.001,
+      "loss": 4.0645,
+      "step": 3964
+    },
+    {
+      "epoch": 0.16773838734241475,
+      "grad_norm": 0.31805071234703064,
+      "learning_rate": 0.001,
+      "loss": 1.9577,
+      "step": 3965
+    },
+    {
+      "epoch": 0.16778069210593113,
+      "grad_norm": 0.6054160594940186,
+      "learning_rate": 0.001,
+      "loss": 2.4387,
+      "step": 3966
+    },
+    {
+      "epoch": 0.1678229968694475,
+      "grad_norm": 0.34546828269958496,
+      "learning_rate": 0.001,
+      "loss": 1.7153,
+      "step": 3967
+    },
+    {
+      "epoch": 0.16786530163296387,
+      "grad_norm": 0.29277992248535156,
+      "learning_rate": 0.001,
+      "loss": 2.0135,
+      "step": 3968
+    },
+    {
+      "epoch": 0.16790760639648025,
+      "grad_norm": 0.25595077872276306,
+      "learning_rate": 0.001,
+      "loss": 3.5568,
+      "step": 3969
+    },
+    {
+      "epoch": 0.1679499111599966,
+      "grad_norm": 0.38943490386009216,
+      "learning_rate": 0.001,
+      "loss": 2.8329,
+      "step": 3970
+    },
+    {
+      "epoch": 0.167992215923513,
+      "grad_norm": 0.23987014591693878,
+      "learning_rate": 0.001,
+      "loss": 2.4131,
+      "step": 3971
+    },
+    {
+      "epoch": 0.16803452068702937,
+      "grad_norm": 0.3031393587589264,
+      "learning_rate": 0.001,
+      "loss": 2.5168,
+      "step": 3972
+    },
+    {
+      "epoch": 0.16807682545054572,
+      "grad_norm": 0.27220430970191956,
+      "learning_rate": 0.001,
+      "loss": 3.0367,
+      "step": 3973
+    },
+    {
+      "epoch": 0.1681191302140621,
+      "grad_norm": 0.3574322462081909,
+      "learning_rate": 0.001,
+      "loss": 1.6484,
+      "step": 3974
+    },
+    {
+      "epoch": 0.1681614349775785,
+      "grad_norm": 0.6500173211097717,
+      "learning_rate": 0.001,
+      "loss": 2.2295,
+      "step": 3975
+    },
+    {
+      "epoch": 0.16820373974109484,
+      "grad_norm": 0.3387952744960785,
+      "learning_rate": 0.001,
+      "loss": 2.0314,
+      "step": 3976
+    },
+    {
+      "epoch": 0.16824604450461123,
+      "grad_norm": 0.21707701683044434,
+      "learning_rate": 0.001,
+      "loss": 2.2662,
+      "step": 3977
+    },
+    {
+      "epoch": 0.16828834926812758,
+      "grad_norm": 0.2807854115962982,
+      "learning_rate": 0.001,
+      "loss": 3.0303,
+      "step": 3978
+    },
+    {
+      "epoch": 0.16833065403164396,
+      "grad_norm": 0.19229240715503693,
+      "learning_rate": 0.001,
+      "loss": 2.87,
+      "step": 3979
+    },
+    {
+      "epoch": 0.16837295879516034,
+      "grad_norm": 0.24821369349956512,
+      "learning_rate": 0.001,
+      "loss": 3.3929,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1684152635586767,
+      "grad_norm": 0.2526116967201233,
+      "learning_rate": 0.001,
+      "loss": 2.3234,
+      "step": 3981
+    },
+    {
+      "epoch": 0.16845756832219308,
+      "grad_norm": 0.3005029857158661,
+      "learning_rate": 0.001,
+      "loss": 1.7588,
+      "step": 3982
+    },
+    {
+      "epoch": 0.16849987308570946,
+      "grad_norm": 2.6109414100646973,
+      "learning_rate": 0.001,
+      "loss": 2.1013,
+      "step": 3983
+    },
+    {
+      "epoch": 0.16854217784922582,
+      "grad_norm": 0.605918824672699,
+      "learning_rate": 0.001,
+      "loss": 3.1133,
+      "step": 3984
+    },
+    {
+      "epoch": 0.1685844826127422,
+      "grad_norm": 0.21746625006198883,
+      "learning_rate": 0.001,
+      "loss": 2.9431,
+      "step": 3985
+    },
+    {
+      "epoch": 0.16862678737625858,
+      "grad_norm": 0.260303795337677,
+      "learning_rate": 0.001,
+      "loss": 2.0652,
+      "step": 3986
+    },
+    {
+      "epoch": 0.16866909213977493,
+      "grad_norm": 0.2302677035331726,
+      "learning_rate": 0.001,
+      "loss": 3.3315,
+      "step": 3987
+    },
+    {
+      "epoch": 0.16871139690329132,
+      "grad_norm": 0.34791648387908936,
+      "learning_rate": 0.001,
+      "loss": 2.44,
+      "step": 3988
+    },
+    {
+      "epoch": 0.16875370166680767,
+      "grad_norm": 1.0129145383834839,
+      "learning_rate": 0.001,
+      "loss": 1.8525,
+      "step": 3989
+    },
+    {
+      "epoch": 0.16879600643032405,
+      "grad_norm": 0.2503008246421814,
+      "learning_rate": 0.001,
+      "loss": 1.666,
+      "step": 3990
+    },
+    {
+      "epoch": 0.16883831119384043,
+      "grad_norm": 0.2816570997238159,
+      "learning_rate": 0.001,
+      "loss": 2.4679,
+      "step": 3991
+    },
+    {
+      "epoch": 0.1688806159573568,
+      "grad_norm": 0.24317888915538788,
+      "learning_rate": 0.001,
+      "loss": 2.6195,
+      "step": 3992
+    },
+    {
+      "epoch": 0.16892292072087317,
+      "grad_norm": 0.3003460466861725,
+      "learning_rate": 0.001,
+      "loss": 2.6692,
+      "step": 3993
+    },
+    {
+      "epoch": 0.16896522548438955,
+      "grad_norm": 0.2870309054851532,
+      "learning_rate": 0.001,
+      "loss": 2.37,
+      "step": 3994
+    },
+    {
+      "epoch": 0.1690075302479059,
+      "grad_norm": 0.35769209265708923,
+      "learning_rate": 0.001,
+      "loss": 2.6095,
+      "step": 3995
+    },
+    {
+      "epoch": 0.1690498350114223,
+      "grad_norm": 0.647505521774292,
+      "learning_rate": 0.001,
+      "loss": 2.4609,
+      "step": 3996
+    },
+    {
+      "epoch": 0.16909213977493867,
+      "grad_norm": 0.2910658121109009,
+      "learning_rate": 0.001,
+      "loss": 2.1234,
+      "step": 3997
+    },
+    {
+      "epoch": 0.16913444453845503,
+      "grad_norm": 0.4279431402683258,
+      "learning_rate": 0.001,
+      "loss": 2.5119,
+      "step": 3998
+    },
+    {
+      "epoch": 0.1691767493019714,
+      "grad_norm": 0.7346844673156738,
+      "learning_rate": 0.001,
+      "loss": 2.1146,
+      "step": 3999
+    },
+    {
+      "epoch": 0.16921905406548776,
+      "grad_norm": 0.24350988864898682,
+      "learning_rate": 0.001,
+      "loss": 2.28,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16926135882900414,
+      "grad_norm": 0.2387346774339676,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 4001
+    },
+    {
+      "epoch": 0.16930366359252053,
+      "grad_norm": 0.7333387732505798,
+      "learning_rate": 0.001,
+      "loss": 2.0075,
+      "step": 4002
+    },
+    {
+      "epoch": 0.16934596835603688,
+      "grad_norm": 0.289885938167572,
+      "learning_rate": 0.001,
+      "loss": 1.9186,
+      "step": 4003
+    },
+    {
+      "epoch": 0.16938827311955326,
+      "grad_norm": 0.3716363310813904,
+      "learning_rate": 0.001,
+      "loss": 3.237,
+      "step": 4004
+    },
+    {
+      "epoch": 0.16943057788306964,
+      "grad_norm": 0.29184696078300476,
+      "learning_rate": 0.001,
+      "loss": 1.9219,
+      "step": 4005
+    },
+    {
+      "epoch": 0.169472882646586,
+      "grad_norm": 2.09316349029541,
+      "learning_rate": 0.001,
+      "loss": 2.8952,
+      "step": 4006
+    },
+    {
+      "epoch": 0.16951518741010238,
+      "grad_norm": 0.30977028608322144,
+      "learning_rate": 0.001,
+      "loss": 1.6384,
+      "step": 4007
+    },
+    {
+      "epoch": 0.16955749217361876,
+      "grad_norm": 0.2913471460342407,
+      "learning_rate": 0.001,
+      "loss": 2.3069,
+      "step": 4008
+    },
+    {
+      "epoch": 0.16959979693713512,
+      "grad_norm": 0.2933041751384735,
+      "learning_rate": 0.001,
+      "loss": 2.115,
+      "step": 4009
+    },
+    {
+      "epoch": 0.1696421017006515,
+      "grad_norm": 0.29783979058265686,
+      "learning_rate": 0.001,
+      "loss": 3.0471,
+      "step": 4010
+    },
+    {
+      "epoch": 0.16968440646416785,
+      "grad_norm": 0.46605968475341797,
+      "learning_rate": 0.001,
+      "loss": 2.4636,
+      "step": 4011
+    },
+    {
+      "epoch": 0.16972671122768423,
+      "grad_norm": 0.346618115901947,
+      "learning_rate": 0.001,
+      "loss": 2.6208,
+      "step": 4012
+    },
+    {
+      "epoch": 0.16976901599120062,
+      "grad_norm": 0.279699444770813,
+      "learning_rate": 0.001,
+      "loss": 2.4847,
+      "step": 4013
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.5628467798233032,
+      "learning_rate": 0.001,
+      "loss": 2.4294,
+      "step": 4014
+    },
+    {
+      "epoch": 0.16985362551823335,
+      "grad_norm": 0.30765336751937866,
+      "learning_rate": 0.001,
+      "loss": 2.5887,
+      "step": 4015
+    },
+    {
+      "epoch": 0.16989593028174974,
+      "grad_norm": 0.43252626061439514,
+      "learning_rate": 0.001,
+      "loss": 2.6351,
+      "step": 4016
+    },
+    {
+      "epoch": 0.1699382350452661,
+      "grad_norm": 0.7735976576805115,
+      "learning_rate": 0.001,
+      "loss": 2.4618,
+      "step": 4017
+    },
+    {
+      "epoch": 0.16998053980878247,
+      "grad_norm": 0.4775107204914093,
+      "learning_rate": 0.001,
+      "loss": 2.93,
+      "step": 4018
+    },
+    {
+      "epoch": 0.17002284457229885,
+      "grad_norm": 0.2542901933193207,
+      "learning_rate": 0.001,
+      "loss": 1.9901,
+      "step": 4019
+    },
+    {
+      "epoch": 0.1700651493358152,
+      "grad_norm": 0.7986733317375183,
+      "learning_rate": 0.001,
+      "loss": 1.8875,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1701074540993316,
+      "grad_norm": 0.24276584386825562,
+      "learning_rate": 0.001,
+      "loss": 2.8431,
+      "step": 4021
+    },
+    {
+      "epoch": 0.17014975886284794,
+      "grad_norm": 0.29002276062965393,
+      "learning_rate": 0.001,
+      "loss": 2.1358,
+      "step": 4022
+    },
+    {
+      "epoch": 0.17019206362636433,
+      "grad_norm": 0.7131394743919373,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 4023
+    },
+    {
+      "epoch": 0.1702343683898807,
+      "grad_norm": 0.22828415036201477,
+      "learning_rate": 0.001,
+      "loss": 1.7578,
+      "step": 4024
+    },
+    {
+      "epoch": 0.17027667315339706,
+      "grad_norm": 0.2842770218849182,
+      "learning_rate": 0.001,
+      "loss": 2.5149,
+      "step": 4025
+    },
+    {
+      "epoch": 0.17031897791691344,
+      "grad_norm": 0.24497929215431213,
+      "learning_rate": 0.001,
+      "loss": 2.7276,
+      "step": 4026
+    },
+    {
+      "epoch": 0.17036128268042983,
+      "grad_norm": 0.2938736379146576,
+      "learning_rate": 0.001,
+      "loss": 3.0422,
+      "step": 4027
+    },
+    {
+      "epoch": 0.17040358744394618,
+      "grad_norm": 0.2794887125492096,
+      "learning_rate": 0.001,
+      "loss": 1.9868,
+      "step": 4028
+    },
+    {
+      "epoch": 0.17044589220746256,
+      "grad_norm": 0.23071792721748352,
+      "learning_rate": 0.001,
+      "loss": 3.0668,
+      "step": 4029
+    },
+    {
+      "epoch": 0.17048819697097894,
+      "grad_norm": 0.2662751376628876,
+      "learning_rate": 0.001,
+      "loss": 2.5487,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1705305017344953,
+      "grad_norm": 0.21133390069007874,
+      "learning_rate": 0.001,
+      "loss": 2.2983,
+      "step": 4031
+    },
+    {
+      "epoch": 0.17057280649801168,
+      "grad_norm": 0.22290173172950745,
+      "learning_rate": 0.001,
+      "loss": 2.4885,
+      "step": 4032
+    },
+    {
+      "epoch": 0.17061511126152804,
+      "grad_norm": 0.29108884930610657,
+      "learning_rate": 0.001,
+      "loss": 1.848,
+      "step": 4033
+    },
+    {
+      "epoch": 0.17065741602504442,
+      "grad_norm": 0.28918585181236267,
+      "learning_rate": 0.001,
+      "loss": 2.7061,
+      "step": 4034
+    },
+    {
+      "epoch": 0.1706997207885608,
+      "grad_norm": 1.4359923601150513,
+      "learning_rate": 0.001,
+      "loss": 2.7202,
+      "step": 4035
+    },
+    {
+      "epoch": 0.17074202555207715,
+      "grad_norm": 0.6065989136695862,
+      "learning_rate": 0.001,
+      "loss": 2.978,
+      "step": 4036
+    },
+    {
+      "epoch": 0.17078433031559354,
+      "grad_norm": 1.2570323944091797,
+      "learning_rate": 0.001,
+      "loss": 2.1911,
+      "step": 4037
+    },
+    {
+      "epoch": 0.17082663507910992,
+      "grad_norm": 0.2336377501487732,
+      "learning_rate": 0.001,
+      "loss": 2.2424,
+      "step": 4038
+    },
+    {
+      "epoch": 0.17086893984262627,
+      "grad_norm": 3.624565601348877,
+      "learning_rate": 0.001,
+      "loss": 1.9867,
+      "step": 4039
+    },
+    {
+      "epoch": 0.17091124460614265,
+      "grad_norm": 0.21247118711471558,
+      "learning_rate": 0.001,
+      "loss": 2.1356,
+      "step": 4040
+    },
+    {
+      "epoch": 0.17095354936965904,
+      "grad_norm": 0.7525285482406616,
+      "learning_rate": 0.001,
+      "loss": 2.7077,
+      "step": 4041
+    },
+    {
+      "epoch": 0.1709958541331754,
+      "grad_norm": 0.3459639549255371,
+      "learning_rate": 0.001,
+      "loss": 2.9617,
+      "step": 4042
+    },
+    {
+      "epoch": 0.17103815889669177,
+      "grad_norm": 0.28101351857185364,
+      "learning_rate": 0.001,
+      "loss": 2.4219,
+      "step": 4043
+    },
+    {
+      "epoch": 0.17108046366020813,
+      "grad_norm": 0.262946218252182,
+      "learning_rate": 0.001,
+      "loss": 2.1022,
+      "step": 4044
+    },
+    {
+      "epoch": 0.1711227684237245,
+      "grad_norm": 134.08250427246094,
+      "learning_rate": 0.001,
+      "loss": 2.2692,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1711650731872409,
+      "grad_norm": 0.24002955853939056,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 4046
+    },
+    {
+      "epoch": 0.17120737795075724,
+      "grad_norm": 0.29694628715515137,
+      "learning_rate": 0.001,
+      "loss": 2.6118,
+      "step": 4047
+    },
+    {
+      "epoch": 0.17124968271427363,
+      "grad_norm": 12.005644798278809,
+      "learning_rate": 0.001,
+      "loss": 2.6029,
+      "step": 4048
+    },
+    {
+      "epoch": 0.17129198747779,
+      "grad_norm": 0.2015325427055359,
+      "learning_rate": 0.001,
+      "loss": 1.6005,
+      "step": 4049
+    },
+    {
+      "epoch": 0.17133429224130636,
+      "grad_norm": 0.18817874789237976,
+      "learning_rate": 0.001,
+      "loss": 2.2318,
+      "step": 4050
+    },
+    {
+      "epoch": 0.17137659700482274,
+      "grad_norm": 0.22923152148723602,
+      "learning_rate": 0.001,
+      "loss": 1.8762,
+      "step": 4051
+    },
+    {
+      "epoch": 0.17141890176833913,
+      "grad_norm": 0.2749441862106323,
+      "learning_rate": 0.001,
+      "loss": 2.4774,
+      "step": 4052
+    },
+    {
+      "epoch": 0.17146120653185548,
+      "grad_norm": 0.23520220816135406,
+      "learning_rate": 0.001,
+      "loss": 1.9863,
+      "step": 4053
+    },
+    {
+      "epoch": 0.17150351129537186,
+      "grad_norm": 0.21112807095050812,
+      "learning_rate": 0.001,
+      "loss": 1.7663,
+      "step": 4054
+    },
+    {
+      "epoch": 0.17154581605888822,
+      "grad_norm": 0.23605698347091675,
+      "learning_rate": 0.001,
+      "loss": 2.1954,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1715881208224046,
+      "grad_norm": 0.19001781940460205,
+      "learning_rate": 0.001,
+      "loss": 2.1111,
+      "step": 4056
+    },
+    {
+      "epoch": 0.17163042558592098,
+      "grad_norm": 0.34026920795440674,
+      "learning_rate": 0.001,
+      "loss": 2.5831,
+      "step": 4057
+    },
+    {
+      "epoch": 0.17167273034943734,
+      "grad_norm": 3.0105605125427246,
+      "learning_rate": 0.001,
+      "loss": 3.2851,
+      "step": 4058
+    },
+    {
+      "epoch": 0.17171503511295372,
+      "grad_norm": 0.2173168957233429,
+      "learning_rate": 0.001,
+      "loss": 1.7838,
+      "step": 4059
+    },
+    {
+      "epoch": 0.1717573398764701,
+      "grad_norm": 0.23681773245334625,
+      "learning_rate": 0.001,
+      "loss": 2.5279,
+      "step": 4060
+    },
+    {
+      "epoch": 0.17179964463998645,
+      "grad_norm": 0.4054577648639679,
+      "learning_rate": 0.001,
+      "loss": 1.787,
+      "step": 4061
+    },
+    {
+      "epoch": 0.17184194940350284,
+      "grad_norm": 0.6441399455070496,
+      "learning_rate": 0.001,
+      "loss": 2.2311,
+      "step": 4062
+    },
+    {
+      "epoch": 0.17188425416701922,
+      "grad_norm": 1.3562344312667847,
+      "learning_rate": 0.001,
+      "loss": 2.7153,
+      "step": 4063
+    },
+    {
+      "epoch": 0.17192655893053557,
+      "grad_norm": 0.26166918873786926,
+      "learning_rate": 0.001,
+      "loss": 3.0784,
+      "step": 4064
+    },
+    {
+      "epoch": 0.17196886369405195,
+      "grad_norm": 0.22632955014705658,
+      "learning_rate": 0.001,
+      "loss": 1.6698,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1720111684575683,
+      "grad_norm": 0.7687202095985413,
+      "learning_rate": 0.001,
+      "loss": 2.1554,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1720534732210847,
+      "grad_norm": 0.6964684128761292,
+      "learning_rate": 0.001,
+      "loss": 2.7002,
+      "step": 4067
+    },
+    {
+      "epoch": 0.17209577798460107,
+      "grad_norm": 0.4354396164417267,
+      "learning_rate": 0.001,
+      "loss": 3.5678,
+      "step": 4068
+    },
+    {
+      "epoch": 0.17213808274811743,
+      "grad_norm": 0.2912690341472626,
+      "learning_rate": 0.001,
+      "loss": 2.0832,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1721803875116338,
+      "grad_norm": 1.5569998025894165,
+      "learning_rate": 0.001,
+      "loss": 2.1359,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1722226922751502,
+      "grad_norm": 0.893632709980011,
+      "learning_rate": 0.001,
+      "loss": 3.0861,
+      "step": 4071
+    },
+    {
+      "epoch": 0.17226499703866655,
+      "grad_norm": 0.7286151051521301,
+      "learning_rate": 0.001,
+      "loss": 2.3538,
+      "step": 4072
+    },
+    {
+      "epoch": 0.17230730180218293,
+      "grad_norm": 0.29030993580818176,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1723496065656993,
+      "grad_norm": 0.28612467646598816,
+      "learning_rate": 0.001,
+      "loss": 2.9171,
+      "step": 4074
+    },
+    {
+      "epoch": 0.17239191132921566,
+      "grad_norm": 3.892707109451294,
+      "learning_rate": 0.001,
+      "loss": 2.1065,
+      "step": 4075
+    },
+    {
+      "epoch": 0.17243421609273205,
+      "grad_norm": 0.24053117632865906,
+      "learning_rate": 0.001,
+      "loss": 2.6832,
+      "step": 4076
+    },
+    {
+      "epoch": 0.1724765208562484,
+      "grad_norm": 0.2183643877506256,
+      "learning_rate": 0.001,
+      "loss": 3.0531,
+      "step": 4077
+    },
+    {
+      "epoch": 0.17251882561976478,
+      "grad_norm": 1.0080100297927856,
+      "learning_rate": 0.001,
+      "loss": 2.9435,
+      "step": 4078
+    },
+    {
+      "epoch": 0.17256113038328116,
+      "grad_norm": 0.3189499080181122,
+      "learning_rate": 0.001,
+      "loss": 2.0975,
+      "step": 4079
+    },
+    {
+      "epoch": 0.17260343514679752,
+      "grad_norm": 0.2897562086582184,
+      "learning_rate": 0.001,
+      "loss": 2.5301,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1726457399103139,
+      "grad_norm": 0.2769043743610382,
+      "learning_rate": 0.001,
+      "loss": 1.8059,
+      "step": 4081
+    },
+    {
+      "epoch": 0.17268804467383028,
+      "grad_norm": 0.8298352360725403,
+      "learning_rate": 0.001,
+      "loss": 2.0551,
+      "step": 4082
+    },
+    {
+      "epoch": 0.17273034943734664,
+      "grad_norm": 0.26442840695381165,
+      "learning_rate": 0.001,
+      "loss": 1.9421,
+      "step": 4083
+    },
+    {
+      "epoch": 0.17277265420086302,
+      "grad_norm": 0.27219903469085693,
+      "learning_rate": 0.001,
+      "loss": 2.6075,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1728149589643794,
+      "grad_norm": 0.21034714579582214,
+      "learning_rate": 0.001,
+      "loss": 2.0325,
+      "step": 4085
+    },
+    {
+      "epoch": 0.17285726372789575,
+      "grad_norm": 3.3212363719940186,
+      "learning_rate": 0.001,
+      "loss": 1.9906,
+      "step": 4086
+    },
+    {
+      "epoch": 0.17289956849141214,
+      "grad_norm": 0.9745553135871887,
+      "learning_rate": 0.001,
+      "loss": 2.8193,
+      "step": 4087
+    },
+    {
+      "epoch": 0.17294187325492852,
+      "grad_norm": 0.2681751847267151,
+      "learning_rate": 0.001,
+      "loss": 2.3555,
+      "step": 4088
+    },
+    {
+      "epoch": 0.17298417801844487,
+      "grad_norm": 0.31693801283836365,
+      "learning_rate": 0.001,
+      "loss": 2.0349,
+      "step": 4089
+    },
+    {
+      "epoch": 0.17302648278196125,
+      "grad_norm": 0.32253557443618774,
+      "learning_rate": 0.001,
+      "loss": 1.9938,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1730687875454776,
+      "grad_norm": 0.25818389654159546,
+      "learning_rate": 0.001,
+      "loss": 2.6496,
+      "step": 4091
+    },
+    {
+      "epoch": 0.173111092308994,
+      "grad_norm": 0.24890708923339844,
+      "learning_rate": 0.001,
+      "loss": 1.7513,
+      "step": 4092
+    },
+    {
+      "epoch": 0.17315339707251037,
+      "grad_norm": 0.22379307448863983,
+      "learning_rate": 0.001,
+      "loss": 1.7251,
+      "step": 4093
+    },
+    {
+      "epoch": 0.17319570183602673,
+      "grad_norm": 0.24759642779827118,
+      "learning_rate": 0.001,
+      "loss": 3.0331,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1732380065995431,
+      "grad_norm": 0.3050745129585266,
+      "learning_rate": 0.001,
+      "loss": 3.0863,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1732803113630595,
+      "grad_norm": 0.29657745361328125,
+      "learning_rate": 0.001,
+      "loss": 3.5831,
+      "step": 4096
+    },
+    {
+      "epoch": 0.17332261612657585,
+      "grad_norm": 0.254731684923172,
+      "learning_rate": 0.001,
+      "loss": 1.9884,
+      "step": 4097
+    },
+    {
+      "epoch": 0.17336492089009223,
+      "grad_norm": 0.24810083210468292,
+      "learning_rate": 0.001,
+      "loss": 1.6098,
+      "step": 4098
+    },
+    {
+      "epoch": 0.1734072256536086,
+      "grad_norm": 0.7149272561073303,
+      "learning_rate": 0.001,
+      "loss": 1.6882,
+      "step": 4099
+    },
+    {
+      "epoch": 0.17344953041712496,
+      "grad_norm": 0.34439778327941895,
+      "learning_rate": 0.001,
+      "loss": 3.468,
+      "step": 4100
+    },
+    {
+      "epoch": 0.17349183518064135,
+      "grad_norm": 0.27477285265922546,
+      "learning_rate": 0.001,
+      "loss": 3.7453,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1735341399441577,
+      "grad_norm": 0.2810986042022705,
+      "learning_rate": 0.001,
+      "loss": 2.4957,
+      "step": 4102
+    },
+    {
+      "epoch": 0.17357644470767408,
+      "grad_norm": 0.24247132241725922,
+      "learning_rate": 0.001,
+      "loss": 1.6966,
+      "step": 4103
+    },
+    {
+      "epoch": 0.17361874947119046,
+      "grad_norm": 0.1760856807231903,
+      "learning_rate": 0.001,
+      "loss": 2.591,
+      "step": 4104
+    },
+    {
+      "epoch": 0.17366105423470682,
+      "grad_norm": 0.25415733456611633,
+      "learning_rate": 0.001,
+      "loss": 2.1011,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1737033589982232,
+      "grad_norm": 2.5164103507995605,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 4106
+    },
+    {
+      "epoch": 0.17374566376173958,
+      "grad_norm": 0.5804241895675659,
+      "learning_rate": 0.001,
+      "loss": 2.2759,
+      "step": 4107
+    },
+    {
+      "epoch": 0.17378796852525594,
+      "grad_norm": 0.24741019308567047,
+      "learning_rate": 0.001,
+      "loss": 2.2455,
+      "step": 4108
+    },
+    {
+      "epoch": 0.17383027328877232,
+      "grad_norm": 0.230664923787117,
+      "learning_rate": 0.001,
+      "loss": 1.8416,
+      "step": 4109
+    },
+    {
+      "epoch": 0.1738725780522887,
+      "grad_norm": 0.3088621497154236,
+      "learning_rate": 0.001,
+      "loss": 2.0536,
+      "step": 4110
+    },
+    {
+      "epoch": 0.17391488281580506,
+      "grad_norm": 0.21473775804042816,
+      "learning_rate": 0.001,
+      "loss": 1.4643,
+      "step": 4111
+    },
+    {
+      "epoch": 0.17395718757932144,
+      "grad_norm": 0.23546308279037476,
+      "learning_rate": 0.001,
+      "loss": 2.4978,
+      "step": 4112
+    },
+    {
+      "epoch": 0.1739994923428378,
+      "grad_norm": 0.24598172307014465,
+      "learning_rate": 0.001,
+      "loss": 3.8651,
+      "step": 4113
+    },
+    {
+      "epoch": 0.17404179710635417,
+      "grad_norm": 0.24466127157211304,
+      "learning_rate": 0.001,
+      "loss": 3.8869,
+      "step": 4114
+    },
+    {
+      "epoch": 0.17408410186987056,
+      "grad_norm": 0.21379978954792023,
+      "learning_rate": 0.001,
+      "loss": 3.1837,
+      "step": 4115
+    },
+    {
+      "epoch": 0.1741264066333869,
+      "grad_norm": 3.490262031555176,
+      "learning_rate": 0.001,
+      "loss": 1.6441,
+      "step": 4116
+    },
+    {
+      "epoch": 0.1741687113969033,
+      "grad_norm": 0.30612412095069885,
+      "learning_rate": 0.001,
+      "loss": 2.0432,
+      "step": 4117
+    },
+    {
+      "epoch": 0.17421101616041967,
+      "grad_norm": 0.32725802063941956,
+      "learning_rate": 0.001,
+      "loss": 2.1969,
+      "step": 4118
+    },
+    {
+      "epoch": 0.17425332092393603,
+      "grad_norm": 0.34883803129196167,
+      "learning_rate": 0.001,
+      "loss": 2.9586,
+      "step": 4119
+    },
+    {
+      "epoch": 0.1742956256874524,
+      "grad_norm": 0.2720988690853119,
+      "learning_rate": 0.001,
+      "loss": 2.6285,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1743379304509688,
+      "grad_norm": 0.47883912920951843,
+      "learning_rate": 0.001,
+      "loss": 2.1746,
+      "step": 4121
+    },
+    {
+      "epoch": 0.17438023521448515,
+      "grad_norm": 0.3157065212726593,
+      "learning_rate": 0.001,
+      "loss": 2.5907,
+      "step": 4122
+    },
+    {
+      "epoch": 0.17442253997800153,
+      "grad_norm": 0.22530221939086914,
+      "learning_rate": 0.001,
+      "loss": 1.8267,
+      "step": 4123
+    },
+    {
+      "epoch": 0.17446484474151788,
+      "grad_norm": 0.6682106852531433,
+      "learning_rate": 0.001,
+      "loss": 1.8648,
+      "step": 4124
+    },
+    {
+      "epoch": 0.17450714950503426,
+      "grad_norm": 0.24948126077651978,
+      "learning_rate": 0.001,
+      "loss": 3.0028,
+      "step": 4125
+    },
+    {
+      "epoch": 0.17454945426855065,
+      "grad_norm": 0.23554354906082153,
+      "learning_rate": 0.001,
+      "loss": 2.8481,
+      "step": 4126
+    },
+    {
+      "epoch": 0.174591759032067,
+      "grad_norm": 0.2798170745372772,
+      "learning_rate": 0.001,
+      "loss": 2.5993,
+      "step": 4127
+    },
+    {
+      "epoch": 0.17463406379558338,
+      "grad_norm": 0.1892385184764862,
+      "learning_rate": 0.001,
+      "loss": 2.373,
+      "step": 4128
+    },
+    {
+      "epoch": 0.17467636855909977,
+      "grad_norm": 4.151578426361084,
+      "learning_rate": 0.001,
+      "loss": 2.2752,
+      "step": 4129
+    },
+    {
+      "epoch": 0.17471867332261612,
+      "grad_norm": 0.2719863951206207,
+      "learning_rate": 0.001,
+      "loss": 2.4036,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1747609780861325,
+      "grad_norm": 0.27087652683258057,
+      "learning_rate": 0.001,
+      "loss": 2.1397,
+      "step": 4131
+    },
+    {
+      "epoch": 0.17480328284964888,
+      "grad_norm": 0.27612924575805664,
+      "learning_rate": 0.001,
+      "loss": 3.2754,
+      "step": 4132
+    },
+    {
+      "epoch": 0.17484558761316524,
+      "grad_norm": 0.32560476660728455,
+      "learning_rate": 0.001,
+      "loss": 2.3888,
+      "step": 4133
+    },
+    {
+      "epoch": 0.17488789237668162,
+      "grad_norm": 0.5404102802276611,
+      "learning_rate": 0.001,
+      "loss": 2.7755,
+      "step": 4134
+    },
+    {
+      "epoch": 0.17493019714019797,
+      "grad_norm": 0.28054457902908325,
+      "learning_rate": 0.001,
+      "loss": 2.4661,
+      "step": 4135
+    },
+    {
+      "epoch": 0.17497250190371436,
+      "grad_norm": 7.133134365081787,
+      "learning_rate": 0.001,
+      "loss": 2.3996,
+      "step": 4136
+    },
+    {
+      "epoch": 0.17501480666723074,
+      "grad_norm": 0.4292699098587036,
+      "learning_rate": 0.001,
+      "loss": 2.2355,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1750571114307471,
+      "grad_norm": 0.2766590416431427,
+      "learning_rate": 0.001,
+      "loss": 2.556,
+      "step": 4138
+    },
+    {
+      "epoch": 0.17509941619426347,
+      "grad_norm": 0.26670682430267334,
+      "learning_rate": 0.001,
+      "loss": 2.2956,
+      "step": 4139
+    },
+    {
+      "epoch": 0.17514172095777986,
+      "grad_norm": 0.2878996729850769,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1751840257212962,
+      "grad_norm": 0.4456641376018524,
+      "learning_rate": 0.001,
+      "loss": 2.4915,
+      "step": 4141
+    },
+    {
+      "epoch": 0.1752263304848126,
+      "grad_norm": 0.28615471720695496,
+      "learning_rate": 0.001,
+      "loss": 2.2347,
+      "step": 4142
+    },
+    {
+      "epoch": 0.17526863524832897,
+      "grad_norm": 0.28259655833244324,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 4143
+    },
+    {
+      "epoch": 0.17531094001184533,
+      "grad_norm": 0.22979553043842316,
+      "learning_rate": 0.001,
+      "loss": 2.1944,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1753532447753617,
+      "grad_norm": 0.6153604984283447,
+      "learning_rate": 0.001,
+      "loss": 2.1908,
+      "step": 4145
+    },
+    {
+      "epoch": 0.17539554953887806,
+      "grad_norm": 1.9100154638290405,
+      "learning_rate": 0.001,
+      "loss": 2.154,
+      "step": 4146
+    },
+    {
+      "epoch": 0.17543785430239445,
+      "grad_norm": 0.23111866414546967,
+      "learning_rate": 0.001,
+      "loss": 1.4795,
+      "step": 4147
+    },
+    {
+      "epoch": 0.17548015906591083,
+      "grad_norm": 0.21934235095977783,
+      "learning_rate": 0.001,
+      "loss": 2.0412,
+      "step": 4148
+    },
+    {
+      "epoch": 0.17552246382942718,
+      "grad_norm": 0.31571948528289795,
+      "learning_rate": 0.001,
+      "loss": 1.7655,
+      "step": 4149
+    },
+    {
+      "epoch": 0.17556476859294357,
+      "grad_norm": 0.35099247097969055,
+      "learning_rate": 0.001,
+      "loss": 2.8888,
+      "step": 4150
+    },
+    {
+      "epoch": 0.17560707335645995,
+      "grad_norm": 0.25058233737945557,
+      "learning_rate": 0.001,
+      "loss": 1.6825,
+      "step": 4151
+    },
+    {
+      "epoch": 0.1756493781199763,
+      "grad_norm": 0.35002222657203674,
+      "learning_rate": 0.001,
+      "loss": 2.3041,
+      "step": 4152
+    },
+    {
+      "epoch": 0.17569168288349268,
+      "grad_norm": 0.2953129708766937,
+      "learning_rate": 0.001,
+      "loss": 2.1666,
+      "step": 4153
+    },
+    {
+      "epoch": 0.17573398764700907,
+      "grad_norm": 0.2621290981769562,
+      "learning_rate": 0.001,
+      "loss": 3.3843,
+      "step": 4154
+    },
+    {
+      "epoch": 0.17577629241052542,
+      "grad_norm": 0.2586938738822937,
+      "learning_rate": 0.001,
+      "loss": 1.8781,
+      "step": 4155
+    },
+    {
+      "epoch": 0.1758185971740418,
+      "grad_norm": 0.29512864351272583,
+      "learning_rate": 0.001,
+      "loss": 2.2121,
+      "step": 4156
+    },
+    {
+      "epoch": 0.17586090193755816,
+      "grad_norm": 0.2333596795797348,
+      "learning_rate": 0.001,
+      "loss": 2.6819,
+      "step": 4157
+    },
+    {
+      "epoch": 0.17590320670107454,
+      "grad_norm": 0.2719203531742096,
+      "learning_rate": 0.001,
+      "loss": 2.4844,
+      "step": 4158
+    },
+    {
+      "epoch": 0.17594551146459092,
+      "grad_norm": 0.9604827761650085,
+      "learning_rate": 0.001,
+      "loss": 2.4097,
+      "step": 4159
+    },
+    {
+      "epoch": 0.17598781622810727,
+      "grad_norm": 0.29836225509643555,
+      "learning_rate": 0.001,
+      "loss": 1.9671,
+      "step": 4160
+    },
+    {
+      "epoch": 0.17603012099162366,
+      "grad_norm": 0.2874457538127899,
+      "learning_rate": 0.001,
+      "loss": 2.8141,
+      "step": 4161
+    },
+    {
+      "epoch": 0.17607242575514004,
+      "grad_norm": 0.3310123085975647,
+      "learning_rate": 0.001,
+      "loss": 2.1315,
+      "step": 4162
+    },
+    {
+      "epoch": 0.1761147305186564,
+      "grad_norm": 0.26700732111930847,
+      "learning_rate": 0.001,
+      "loss": 2.7584,
+      "step": 4163
+    },
+    {
+      "epoch": 0.17615703528217277,
+      "grad_norm": 0.2958766520023346,
+      "learning_rate": 0.001,
+      "loss": 2.1226,
+      "step": 4164
+    },
+    {
+      "epoch": 0.17619934004568916,
+      "grad_norm": 0.30483001470565796,
+      "learning_rate": 0.001,
+      "loss": 2.3202,
+      "step": 4165
+    },
+    {
+      "epoch": 0.1762416448092055,
+      "grad_norm": 0.792739987373352,
+      "learning_rate": 0.001,
+      "loss": 2.3611,
+      "step": 4166
+    },
+    {
+      "epoch": 0.1762839495727219,
+      "grad_norm": 0.39795559644699097,
+      "learning_rate": 0.001,
+      "loss": 3.8292,
+      "step": 4167
+    },
+    {
+      "epoch": 0.17632625433623825,
+      "grad_norm": 0.5967978239059448,
+      "learning_rate": 0.001,
+      "loss": 2.3902,
+      "step": 4168
+    },
+    {
+      "epoch": 0.17636855909975463,
+      "grad_norm": 0.25690481066703796,
+      "learning_rate": 0.001,
+      "loss": 3.1277,
+      "step": 4169
+    },
+    {
+      "epoch": 0.176410863863271,
+      "grad_norm": 0.3368457555770874,
+      "learning_rate": 0.001,
+      "loss": 2.4135,
+      "step": 4170
+    },
+    {
+      "epoch": 0.17645316862678737,
+      "grad_norm": 0.5564723014831543,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 4171
+    },
+    {
+      "epoch": 0.17649547339030375,
+      "grad_norm": 0.28822070360183716,
+      "learning_rate": 0.001,
+      "loss": 2.0568,
+      "step": 4172
+    },
+    {
+      "epoch": 0.17653777815382013,
+      "grad_norm": 0.2540615200996399,
+      "learning_rate": 0.001,
+      "loss": 2.7371,
+      "step": 4173
+    },
+    {
+      "epoch": 0.17658008291733648,
+      "grad_norm": 0.30011865496635437,
+      "learning_rate": 0.001,
+      "loss": 2.0551,
+      "step": 4174
+    },
+    {
+      "epoch": 0.17662238768085287,
+      "grad_norm": 0.28831255435943604,
+      "learning_rate": 0.001,
+      "loss": 2.3959,
+      "step": 4175
+    },
+    {
+      "epoch": 0.17666469244436925,
+      "grad_norm": 0.2988402843475342,
+      "learning_rate": 0.001,
+      "loss": 1.8196,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1767069972078856,
+      "grad_norm": 1.3660717010498047,
+      "learning_rate": 0.001,
+      "loss": 1.853,
+      "step": 4177
+    },
+    {
+      "epoch": 0.17674930197140198,
+      "grad_norm": 0.25615715980529785,
+      "learning_rate": 0.001,
+      "loss": 2.7053,
+      "step": 4178
+    },
+    {
+      "epoch": 0.17679160673491834,
+      "grad_norm": 0.25990021228790283,
+      "learning_rate": 0.001,
+      "loss": 3.0701,
+      "step": 4179
+    },
+    {
+      "epoch": 0.17683391149843472,
+      "grad_norm": 0.24227409064769745,
+      "learning_rate": 0.001,
+      "loss": 2.085,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1768762162619511,
+      "grad_norm": 0.2668752670288086,
+      "learning_rate": 0.001,
+      "loss": 2.9824,
+      "step": 4181
+    },
+    {
+      "epoch": 0.17691852102546746,
+      "grad_norm": 0.2560098469257355,
+      "learning_rate": 0.001,
+      "loss": 1.9853,
+      "step": 4182
+    },
+    {
+      "epoch": 0.17696082578898384,
+      "grad_norm": 26.13444709777832,
+      "learning_rate": 0.001,
+      "loss": 1.7282,
+      "step": 4183
+    },
+    {
+      "epoch": 0.17700313055250022,
+      "grad_norm": 0.28741776943206787,
+      "learning_rate": 0.001,
+      "loss": 1.9061,
+      "step": 4184
+    },
+    {
+      "epoch": 0.17704543531601658,
+      "grad_norm": 0.2741623520851135,
+      "learning_rate": 0.001,
+      "loss": 2.4942,
+      "step": 4185
+    },
+    {
+      "epoch": 0.17708774007953296,
+      "grad_norm": 0.24903661012649536,
+      "learning_rate": 0.001,
+      "loss": 3.1056,
+      "step": 4186
+    },
+    {
+      "epoch": 0.17713004484304934,
+      "grad_norm": 0.21220599114894867,
+      "learning_rate": 0.001,
+      "loss": 1.4461,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1771723496065657,
+      "grad_norm": 0.3349112570285797,
+      "learning_rate": 0.001,
+      "loss": 2.0917,
+      "step": 4188
+    },
+    {
+      "epoch": 0.17721465437008208,
+      "grad_norm": 0.7024485468864441,
+      "learning_rate": 0.001,
+      "loss": 2.4401,
+      "step": 4189
+    },
+    {
+      "epoch": 0.17725695913359843,
+      "grad_norm": 0.2594164311885834,
+      "learning_rate": 0.001,
+      "loss": 2.3743,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1772992638971148,
+      "grad_norm": 0.26141050457954407,
+      "learning_rate": 0.001,
+      "loss": 3.0745,
+      "step": 4191
+    },
+    {
+      "epoch": 0.1773415686606312,
+      "grad_norm": 0.25073671340942383,
+      "learning_rate": 0.001,
+      "loss": 2.8072,
+      "step": 4192
+    },
+    {
+      "epoch": 0.17738387342414755,
+      "grad_norm": 0.25238800048828125,
+      "learning_rate": 0.001,
+      "loss": 2.5714,
+      "step": 4193
+    },
+    {
+      "epoch": 0.17742617818766393,
+      "grad_norm": 2.378894805908203,
+      "learning_rate": 0.001,
+      "loss": 3.8264,
+      "step": 4194
+    },
+    {
+      "epoch": 0.1774684829511803,
+      "grad_norm": 0.2004442662000656,
+      "learning_rate": 0.001,
+      "loss": 3.0773,
+      "step": 4195
+    },
+    {
+      "epoch": 0.17751078771469667,
+      "grad_norm": 0.48582643270492554,
+      "learning_rate": 0.001,
+      "loss": 2.3928,
+      "step": 4196
+    },
+    {
+      "epoch": 0.17755309247821305,
+      "grad_norm": 0.3151031732559204,
+      "learning_rate": 0.001,
+      "loss": 2.8011,
+      "step": 4197
+    },
+    {
+      "epoch": 0.17759539724172943,
+      "grad_norm": 0.257080078125,
+      "learning_rate": 0.001,
+      "loss": 2.5786,
+      "step": 4198
+    },
+    {
+      "epoch": 0.17763770200524578,
+      "grad_norm": 0.42748337984085083,
+      "learning_rate": 0.001,
+      "loss": 2.5647,
+      "step": 4199
+    },
+    {
+      "epoch": 0.17768000676876217,
+      "grad_norm": 0.3146316409111023,
+      "learning_rate": 0.001,
+      "loss": 2.9859,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17772231153227852,
+      "grad_norm": 0.2928094267845154,
+      "learning_rate": 0.001,
+      "loss": 2.3672,
+      "step": 4201
+    },
+    {
+      "epoch": 0.1777646162957949,
+      "grad_norm": 0.31037336587905884,
+      "learning_rate": 0.001,
+      "loss": 2.9569,
+      "step": 4202
+    },
+    {
+      "epoch": 0.17780692105931128,
+      "grad_norm": 0.2981893718242645,
+      "learning_rate": 0.001,
+      "loss": 2.2674,
+      "step": 4203
+    },
+    {
+      "epoch": 0.17784922582282764,
+      "grad_norm": 0.26700565218925476,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 4204
+    },
+    {
+      "epoch": 0.17789153058634402,
+      "grad_norm": 0.3305959701538086,
+      "learning_rate": 0.001,
+      "loss": 1.9121,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1779338353498604,
+      "grad_norm": 0.323095440864563,
+      "learning_rate": 0.001,
+      "loss": 2.6363,
+      "step": 4206
+    },
+    {
+      "epoch": 0.17797614011337676,
+      "grad_norm": 0.5207691788673401,
+      "learning_rate": 0.001,
+      "loss": 2.3411,
+      "step": 4207
+    },
+    {
+      "epoch": 0.17801844487689314,
+      "grad_norm": 0.5256071090698242,
+      "learning_rate": 0.001,
+      "loss": 2.0633,
+      "step": 4208
+    },
+    {
+      "epoch": 0.17806074964040952,
+      "grad_norm": 0.20449815690517426,
+      "learning_rate": 0.001,
+      "loss": 1.5104,
+      "step": 4209
+    },
+    {
+      "epoch": 0.17810305440392588,
+      "grad_norm": 0.3280058801174164,
+      "learning_rate": 0.001,
+      "loss": 2.0113,
+      "step": 4210
+    },
+    {
+      "epoch": 0.17814535916744226,
+      "grad_norm": 0.20661139488220215,
+      "learning_rate": 0.001,
+      "loss": 1.8537,
+      "step": 4211
+    },
+    {
+      "epoch": 0.17818766393095864,
+      "grad_norm": 0.22364257276058197,
+      "learning_rate": 0.001,
+      "loss": 2.2209,
+      "step": 4212
+    },
+    {
+      "epoch": 0.178229968694475,
+      "grad_norm": 0.3003136217594147,
+      "learning_rate": 0.001,
+      "loss": 2.3106,
+      "step": 4213
+    },
+    {
+      "epoch": 0.17827227345799138,
+      "grad_norm": 0.39940145611763,
+      "learning_rate": 0.001,
+      "loss": 2.6153,
+      "step": 4214
+    },
+    {
+      "epoch": 0.17831457822150773,
+      "grad_norm": 0.2537018954753876,
+      "learning_rate": 0.001,
+      "loss": 2.6277,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1783568829850241,
+      "grad_norm": 0.25565820932388306,
+      "learning_rate": 0.001,
+      "loss": 2.7391,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1783991877485405,
+      "grad_norm": 0.3583895266056061,
+      "learning_rate": 0.001,
+      "loss": 2.151,
+      "step": 4217
+    },
+    {
+      "epoch": 0.17844149251205685,
+      "grad_norm": 0.21725155413150787,
+      "learning_rate": 0.001,
+      "loss": 2.0606,
+      "step": 4218
+    },
+    {
+      "epoch": 0.17848379727557323,
+      "grad_norm": 0.21374504268169403,
+      "learning_rate": 0.001,
+      "loss": 2.0371,
+      "step": 4219
+    },
+    {
+      "epoch": 0.1785261020390896,
+      "grad_norm": 0.21078385412693024,
+      "learning_rate": 0.001,
+      "loss": 1.8766,
+      "step": 4220
+    },
+    {
+      "epoch": 0.17856840680260597,
+      "grad_norm": 0.2687901258468628,
+      "learning_rate": 0.001,
+      "loss": 2.4474,
+      "step": 4221
+    },
+    {
+      "epoch": 0.17861071156612235,
+      "grad_norm": 0.3007321357727051,
+      "learning_rate": 0.001,
+      "loss": 2.4974,
+      "step": 4222
+    },
+    {
+      "epoch": 0.17865301632963873,
+      "grad_norm": 0.35999658703804016,
+      "learning_rate": 0.001,
+      "loss": 2.2235,
+      "step": 4223
+    },
+    {
+      "epoch": 0.17869532109315509,
+      "grad_norm": 0.5664570331573486,
+      "learning_rate": 0.001,
+      "loss": 2.6284,
+      "step": 4224
+    },
+    {
+      "epoch": 0.17873762585667147,
+      "grad_norm": 0.28979048132896423,
+      "learning_rate": 0.001,
+      "loss": 2.0292,
+      "step": 4225
+    },
+    {
+      "epoch": 0.17877993062018782,
+      "grad_norm": 0.3224925398826599,
+      "learning_rate": 0.001,
+      "loss": 1.9622,
+      "step": 4226
+    },
+    {
+      "epoch": 0.1788222353837042,
+      "grad_norm": 4.19497537612915,
+      "learning_rate": 0.001,
+      "loss": 2.2633,
+      "step": 4227
+    },
+    {
+      "epoch": 0.17886454014722059,
+      "grad_norm": 0.252350389957428,
+      "learning_rate": 0.001,
+      "loss": 3.7678,
+      "step": 4228
+    },
+    {
+      "epoch": 0.17890684491073694,
+      "grad_norm": 0.26426586508750916,
+      "learning_rate": 0.001,
+      "loss": 2.3954,
+      "step": 4229
+    },
+    {
+      "epoch": 0.17894914967425332,
+      "grad_norm": 0.24827632308006287,
+      "learning_rate": 0.001,
+      "loss": 1.9709,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1789914544377697,
+      "grad_norm": 4.474344253540039,
+      "learning_rate": 0.001,
+      "loss": 2.9707,
+      "step": 4231
+    },
+    {
+      "epoch": 0.17903375920128606,
+      "grad_norm": 0.4154488444328308,
+      "learning_rate": 0.001,
+      "loss": 2.6213,
+      "step": 4232
+    },
+    {
+      "epoch": 0.17907606396480244,
+      "grad_norm": 0.2699248492717743,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 4233
+    },
+    {
+      "epoch": 0.17911836872831882,
+      "grad_norm": 0.2923453152179718,
+      "learning_rate": 0.001,
+      "loss": 1.9264,
+      "step": 4234
+    },
+    {
+      "epoch": 0.17916067349183518,
+      "grad_norm": 0.31689560413360596,
+      "learning_rate": 0.001,
+      "loss": 2.5336,
+      "step": 4235
+    },
+    {
+      "epoch": 0.17920297825535156,
+      "grad_norm": 0.2745906412601471,
+      "learning_rate": 0.001,
+      "loss": 1.8758,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1792452830188679,
+      "grad_norm": 1.625520944595337,
+      "learning_rate": 0.001,
+      "loss": 4.2199,
+      "step": 4237
+    },
+    {
+      "epoch": 0.1792875877823843,
+      "grad_norm": 0.32071831822395325,
+      "learning_rate": 0.001,
+      "loss": 2.993,
+      "step": 4238
+    },
+    {
+      "epoch": 0.17932989254590068,
+      "grad_norm": 0.2907838225364685,
+      "learning_rate": 0.001,
+      "loss": 2.3436,
+      "step": 4239
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.25917503237724304,
+      "learning_rate": 0.001,
+      "loss": 1.7643,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1794145020729334,
+      "grad_norm": 0.35071495175361633,
+      "learning_rate": 0.001,
+      "loss": 2.6049,
+      "step": 4241
+    },
+    {
+      "epoch": 0.1794568068364498,
+      "grad_norm": 0.402885377407074,
+      "learning_rate": 0.001,
+      "loss": 3.151,
+      "step": 4242
+    },
+    {
+      "epoch": 0.17949911159996615,
+      "grad_norm": 0.2622643709182739,
+      "learning_rate": 0.001,
+      "loss": 1.8995,
+      "step": 4243
+    },
+    {
+      "epoch": 0.17954141636348253,
+      "grad_norm": 0.2709498107433319,
+      "learning_rate": 0.001,
+      "loss": 2.5256,
+      "step": 4244
+    },
+    {
+      "epoch": 0.1795837211269989,
+      "grad_norm": 0.27462783455848694,
+      "learning_rate": 0.001,
+      "loss": 2.714,
+      "step": 4245
+    },
+    {
+      "epoch": 0.17962602589051527,
+      "grad_norm": 0.20611165463924408,
+      "learning_rate": 0.001,
+      "loss": 1.6344,
+      "step": 4246
+    },
+    {
+      "epoch": 0.17966833065403165,
+      "grad_norm": 0.32255131006240845,
+      "learning_rate": 0.001,
+      "loss": 2.2145,
+      "step": 4247
+    },
+    {
+      "epoch": 0.179710635417548,
+      "grad_norm": 0.21528469026088715,
+      "learning_rate": 0.001,
+      "loss": 1.8803,
+      "step": 4248
+    },
+    {
+      "epoch": 0.17975294018106439,
+      "grad_norm": 0.21335670351982117,
+      "learning_rate": 0.001,
+      "loss": 2.8012,
+      "step": 4249
+    },
+    {
+      "epoch": 0.17979524494458077,
+      "grad_norm": 1.517720341682434,
+      "learning_rate": 0.001,
+      "loss": 2.1899,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17983754970809712,
+      "grad_norm": 0.762283205986023,
+      "learning_rate": 0.001,
+      "loss": 3.0798,
+      "step": 4251
+    },
+    {
+      "epoch": 0.1798798544716135,
+      "grad_norm": 0.23315021395683289,
+      "learning_rate": 0.001,
+      "loss": 2.9799,
+      "step": 4252
+    },
+    {
+      "epoch": 0.17992215923512989,
+      "grad_norm": 1.0861098766326904,
+      "learning_rate": 0.001,
+      "loss": 3.0077,
+      "step": 4253
+    },
+    {
+      "epoch": 0.17996446399864624,
+      "grad_norm": 0.4819357991218567,
+      "learning_rate": 0.001,
+      "loss": 3.5156,
+      "step": 4254
+    },
+    {
+      "epoch": 0.18000676876216262,
+      "grad_norm": 0.5548194646835327,
+      "learning_rate": 0.001,
+      "loss": 2.9003,
+      "step": 4255
+    },
+    {
+      "epoch": 0.180049073525679,
+      "grad_norm": 8.51793384552002,
+      "learning_rate": 0.001,
+      "loss": 2.8978,
+      "step": 4256
+    },
+    {
+      "epoch": 0.18009137828919536,
+      "grad_norm": 0.37549543380737305,
+      "learning_rate": 0.001,
+      "loss": 2.1621,
+      "step": 4257
+    },
+    {
+      "epoch": 0.18013368305271174,
+      "grad_norm": 1.0575660467147827,
+      "learning_rate": 0.001,
+      "loss": 2.7155,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1801759878162281,
+      "grad_norm": 0.26176881790161133,
+      "learning_rate": 0.001,
+      "loss": 1.717,
+      "step": 4259
+    },
+    {
+      "epoch": 0.18021829257974448,
+      "grad_norm": 0.418062299489975,
+      "learning_rate": 0.001,
+      "loss": 3.2709,
+      "step": 4260
+    },
+    {
+      "epoch": 0.18026059734326086,
+      "grad_norm": 0.3230735957622528,
+      "learning_rate": 0.001,
+      "loss": 2.752,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1803029021067772,
+      "grad_norm": 0.3032633662223816,
+      "learning_rate": 0.001,
+      "loss": 1.8807,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1803452068702936,
+      "grad_norm": 0.37252533435821533,
+      "learning_rate": 0.001,
+      "loss": 2.2327,
+      "step": 4263
+    },
+    {
+      "epoch": 0.18038751163380998,
+      "grad_norm": 0.8877015709877014,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 4264
+    },
+    {
+      "epoch": 0.18042981639732633,
+      "grad_norm": 0.6750360131263733,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1804721211608427,
+      "grad_norm": 0.626649796962738,
+      "learning_rate": 0.001,
+      "loss": 3.1486,
+      "step": 4266
+    },
+    {
+      "epoch": 0.1805144259243591,
+      "grad_norm": 0.3944655656814575,
+      "learning_rate": 0.001,
+      "loss": 2.4488,
+      "step": 4267
+    },
+    {
+      "epoch": 0.18055673068787545,
+      "grad_norm": 0.28062745928764343,
+      "learning_rate": 0.001,
+      "loss": 1.9655,
+      "step": 4268
+    },
+    {
+      "epoch": 0.18059903545139183,
+      "grad_norm": 2.311063051223755,
+      "learning_rate": 0.001,
+      "loss": 2.6447,
+      "step": 4269
+    },
+    {
+      "epoch": 0.18064134021490819,
+      "grad_norm": 0.30582699179649353,
+      "learning_rate": 0.001,
+      "loss": 2.7865,
+      "step": 4270
+    },
+    {
+      "epoch": 0.18068364497842457,
+      "grad_norm": 0.6388780474662781,
+      "learning_rate": 0.001,
+      "loss": 3.422,
+      "step": 4271
+    },
+    {
+      "epoch": 0.18072594974194095,
+      "grad_norm": 0.24511829018592834,
+      "learning_rate": 0.001,
+      "loss": 1.925,
+      "step": 4272
+    },
+    {
+      "epoch": 0.1807682545054573,
+      "grad_norm": 10.821248054504395,
+      "learning_rate": 0.001,
+      "loss": 2.6423,
+      "step": 4273
+    },
+    {
+      "epoch": 0.1808105592689737,
+      "grad_norm": 0.9741492867469788,
+      "learning_rate": 0.001,
+      "loss": 2.614,
+      "step": 4274
+    },
+    {
+      "epoch": 0.18085286403249007,
+      "grad_norm": 0.4937375485897064,
+      "learning_rate": 0.001,
+      "loss": 2.9155,
+      "step": 4275
+    },
+    {
+      "epoch": 0.18089516879600642,
+      "grad_norm": 0.32273703813552856,
+      "learning_rate": 0.001,
+      "loss": 3.4425,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1809374735595228,
+      "grad_norm": 0.3622698187828064,
+      "learning_rate": 0.001,
+      "loss": 2.5883,
+      "step": 4277
+    },
+    {
+      "epoch": 0.1809797783230392,
+      "grad_norm": 0.2949691116809845,
+      "learning_rate": 0.001,
+      "loss": 3.5752,
+      "step": 4278
+    },
+    {
+      "epoch": 0.18102208308655554,
+      "grad_norm": 0.2598804533481598,
+      "learning_rate": 0.001,
+      "loss": 1.8536,
+      "step": 4279
+    },
+    {
+      "epoch": 0.18106438785007192,
+      "grad_norm": 1.1611905097961426,
+      "learning_rate": 0.001,
+      "loss": 3.0881,
+      "step": 4280
+    },
+    {
+      "epoch": 0.18110669261358828,
+      "grad_norm": 1.2100458145141602,
+      "learning_rate": 0.001,
+      "loss": 2.2188,
+      "step": 4281
+    },
+    {
+      "epoch": 0.18114899737710466,
+      "grad_norm": 0.5973420739173889,
+      "learning_rate": 0.001,
+      "loss": 2.6307,
+      "step": 4282
+    },
+    {
+      "epoch": 0.18119130214062104,
+      "grad_norm": 0.29785507917404175,
+      "learning_rate": 0.001,
+      "loss": 2.6614,
+      "step": 4283
+    },
+    {
+      "epoch": 0.1812336069041374,
+      "grad_norm": 0.2885024845600128,
+      "learning_rate": 0.001,
+      "loss": 2.4603,
+      "step": 4284
+    },
+    {
+      "epoch": 0.18127591166765378,
+      "grad_norm": 5.69395112991333,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 4285
+    },
+    {
+      "epoch": 0.18131821643117016,
+      "grad_norm": 0.9888959527015686,
+      "learning_rate": 0.001,
+      "loss": 3.3646,
+      "step": 4286
+    },
+    {
+      "epoch": 0.1813605211946865,
+      "grad_norm": 1.5482476949691772,
+      "learning_rate": 0.001,
+      "loss": 1.983,
+      "step": 4287
+    },
+    {
+      "epoch": 0.1814028259582029,
+      "grad_norm": 1.1126073598861694,
+      "learning_rate": 0.001,
+      "loss": 2.3665,
+      "step": 4288
+    },
+    {
+      "epoch": 0.18144513072171928,
+      "grad_norm": 3.9008285999298096,
+      "learning_rate": 0.001,
+      "loss": 2.3922,
+      "step": 4289
+    },
+    {
+      "epoch": 0.18148743548523563,
+      "grad_norm": 0.30987706780433655,
+      "learning_rate": 0.001,
+      "loss": 2.7724,
+      "step": 4290
+    },
+    {
+      "epoch": 0.18152974024875201,
+      "grad_norm": 0.5834580659866333,
+      "learning_rate": 0.001,
+      "loss": 3.4511,
+      "step": 4291
+    },
+    {
+      "epoch": 0.18157204501226837,
+      "grad_norm": 0.37197399139404297,
+      "learning_rate": 0.001,
+      "loss": 2.7004,
+      "step": 4292
+    },
+    {
+      "epoch": 0.18161434977578475,
+      "grad_norm": 0.2617497742176056,
+      "learning_rate": 0.001,
+      "loss": 2.2616,
+      "step": 4293
+    },
+    {
+      "epoch": 0.18165665453930113,
+      "grad_norm": 0.25308120250701904,
+      "learning_rate": 0.001,
+      "loss": 2.5587,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1816989593028175,
+      "grad_norm": 0.22817359864711761,
+      "learning_rate": 0.001,
+      "loss": 2.3229,
+      "step": 4295
+    },
+    {
+      "epoch": 0.18174126406633387,
+      "grad_norm": 0.27048158645629883,
+      "learning_rate": 0.001,
+      "loss": 2.0208,
+      "step": 4296
+    },
+    {
+      "epoch": 0.18178356882985025,
+      "grad_norm": 0.7957993149757385,
+      "learning_rate": 0.001,
+      "loss": 2.4789,
+      "step": 4297
+    },
+    {
+      "epoch": 0.1818258735933666,
+      "grad_norm": 0.2634366452693939,
+      "learning_rate": 0.001,
+      "loss": 2.9582,
+      "step": 4298
+    },
+    {
+      "epoch": 0.181868178356883,
+      "grad_norm": 0.2717626094818115,
+      "learning_rate": 0.001,
+      "loss": 2.3598,
+      "step": 4299
+    },
+    {
+      "epoch": 0.18191048312039937,
+      "grad_norm": 1.2515925168991089,
+      "learning_rate": 0.001,
+      "loss": 2.4262,
+      "step": 4300
+    },
+    {
+      "epoch": 0.18195278788391572,
+      "grad_norm": 0.307389497756958,
+      "learning_rate": 0.001,
+      "loss": 2.2833,
+      "step": 4301
+    },
+    {
+      "epoch": 0.1819950926474321,
+      "grad_norm": 0.6592715978622437,
+      "learning_rate": 0.001,
+      "loss": 2.8568,
+      "step": 4302
+    },
+    {
+      "epoch": 0.18203739741094846,
+      "grad_norm": 0.24115906655788422,
+      "learning_rate": 0.001,
+      "loss": 2.0772,
+      "step": 4303
+    },
+    {
+      "epoch": 0.18207970217446484,
+      "grad_norm": 0.3468441963195801,
+      "learning_rate": 0.001,
+      "loss": 3.1513,
+      "step": 4304
+    },
+    {
+      "epoch": 0.18212200693798122,
+      "grad_norm": 0.3268192708492279,
+      "learning_rate": 0.001,
+      "loss": 2.2592,
+      "step": 4305
+    },
+    {
+      "epoch": 0.18216431170149758,
+      "grad_norm": 0.4834609925746918,
+      "learning_rate": 0.001,
+      "loss": 1.9173,
+      "step": 4306
+    },
+    {
+      "epoch": 0.18220661646501396,
+      "grad_norm": 0.29343941807746887,
+      "learning_rate": 0.001,
+      "loss": 4.07,
+      "step": 4307
+    },
+    {
+      "epoch": 0.18224892122853034,
+      "grad_norm": 0.5635595321655273,
+      "learning_rate": 0.001,
+      "loss": 3.5071,
+      "step": 4308
+    },
+    {
+      "epoch": 0.1822912259920467,
+      "grad_norm": 6.358903884887695,
+      "learning_rate": 0.001,
+      "loss": 2.6644,
+      "step": 4309
+    },
+    {
+      "epoch": 0.18233353075556308,
+      "grad_norm": 1.2210367918014526,
+      "learning_rate": 0.001,
+      "loss": 3.2344,
+      "step": 4310
+    },
+    {
+      "epoch": 0.18237583551907946,
+      "grad_norm": 0.3168500065803528,
+      "learning_rate": 0.001,
+      "loss": 1.8108,
+      "step": 4311
+    },
+    {
+      "epoch": 0.18241814028259581,
+      "grad_norm": 0.2486400455236435,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 4312
+    },
+    {
+      "epoch": 0.1824604450461122,
+      "grad_norm": 0.2820669114589691,
+      "learning_rate": 0.001,
+      "loss": 2.4824,
+      "step": 4313
+    },
+    {
+      "epoch": 0.18250274980962855,
+      "grad_norm": 0.23513592779636383,
+      "learning_rate": 0.001,
+      "loss": 2.2421,
+      "step": 4314
+    },
+    {
+      "epoch": 0.18254505457314493,
+      "grad_norm": 0.33237895369529724,
+      "learning_rate": 0.001,
+      "loss": 2.6718,
+      "step": 4315
+    },
+    {
+      "epoch": 0.18258735933666131,
+      "grad_norm": 1.0040096044540405,
+      "learning_rate": 0.001,
+      "loss": 2.4138,
+      "step": 4316
+    },
+    {
+      "epoch": 0.18262966410017767,
+      "grad_norm": 2.4072158336639404,
+      "learning_rate": 0.001,
+      "loss": 2.3787,
+      "step": 4317
+    },
+    {
+      "epoch": 0.18267196886369405,
+      "grad_norm": 0.2276010513305664,
+      "learning_rate": 0.001,
+      "loss": 2.5066,
+      "step": 4318
+    },
+    {
+      "epoch": 0.18271427362721043,
+      "grad_norm": 0.2673986256122589,
+      "learning_rate": 0.001,
+      "loss": 2.4379,
+      "step": 4319
+    },
+    {
+      "epoch": 0.1827565783907268,
+      "grad_norm": 0.3459979295730591,
+      "learning_rate": 0.001,
+      "loss": 2.8818,
+      "step": 4320
+    },
+    {
+      "epoch": 0.18279888315424317,
+      "grad_norm": 0.30909305810928345,
+      "learning_rate": 0.001,
+      "loss": 2.359,
+      "step": 4321
+    },
+    {
+      "epoch": 0.18284118791775955,
+      "grad_norm": 0.29360231757164,
+      "learning_rate": 0.001,
+      "loss": 2.3982,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1828834926812759,
+      "grad_norm": 0.3940381407737732,
+      "learning_rate": 0.001,
+      "loss": 2.7807,
+      "step": 4323
+    },
+    {
+      "epoch": 0.1829257974447923,
+      "grad_norm": 0.7100034356117249,
+      "learning_rate": 0.001,
+      "loss": 3.6592,
+      "step": 4324
+    },
+    {
+      "epoch": 0.18296810220830864,
+      "grad_norm": 4.326420307159424,
+      "learning_rate": 0.001,
+      "loss": 2.7129,
+      "step": 4325
+    },
+    {
+      "epoch": 0.18301040697182502,
+      "grad_norm": 0.2793787717819214,
+      "learning_rate": 0.001,
+      "loss": 3.1424,
+      "step": 4326
+    },
+    {
+      "epoch": 0.1830527117353414,
+      "grad_norm": 0.9387155175209045,
+      "learning_rate": 0.001,
+      "loss": 3.1617,
+      "step": 4327
+    },
+    {
+      "epoch": 0.18309501649885776,
+      "grad_norm": 0.3149893581867218,
+      "learning_rate": 0.001,
+      "loss": 2.914,
+      "step": 4328
+    },
+    {
+      "epoch": 0.18313732126237414,
+      "grad_norm": 1.0925941467285156,
+      "learning_rate": 0.001,
+      "loss": 2.902,
+      "step": 4329
+    },
+    {
+      "epoch": 0.18317962602589052,
+      "grad_norm": 0.3763814866542816,
+      "learning_rate": 0.001,
+      "loss": 3.5513,
+      "step": 4330
+    },
+    {
+      "epoch": 0.18322193078940688,
+      "grad_norm": 0.2654663622379303,
+      "learning_rate": 0.001,
+      "loss": 3.1993,
+      "step": 4331
+    },
+    {
+      "epoch": 0.18326423555292326,
+      "grad_norm": 0.34242555499076843,
+      "learning_rate": 0.001,
+      "loss": 2.4377,
+      "step": 4332
+    },
+    {
+      "epoch": 0.18330654031643964,
+      "grad_norm": 0.37130022048950195,
+      "learning_rate": 0.001,
+      "loss": 3.27,
+      "step": 4333
+    },
+    {
+      "epoch": 0.183348845079956,
+      "grad_norm": 0.4086019694805145,
+      "learning_rate": 0.001,
+      "loss": 2.3096,
+      "step": 4334
+    },
+    {
+      "epoch": 0.18339114984347238,
+      "grad_norm": 0.2771133482456207,
+      "learning_rate": 0.001,
+      "loss": 3.351,
+      "step": 4335
+    },
+    {
+      "epoch": 0.18343345460698876,
+      "grad_norm": 0.3115023970603943,
+      "learning_rate": 0.001,
+      "loss": 1.9335,
+      "step": 4336
+    },
+    {
+      "epoch": 0.18347575937050511,
+      "grad_norm": 0.513750433921814,
+      "learning_rate": 0.001,
+      "loss": 3.1363,
+      "step": 4337
+    },
+    {
+      "epoch": 0.1835180641340215,
+      "grad_norm": 0.3518810570240021,
+      "learning_rate": 0.001,
+      "loss": 3.5947,
+      "step": 4338
+    },
+    {
+      "epoch": 0.18356036889753785,
+      "grad_norm": 0.2949376106262207,
+      "learning_rate": 0.001,
+      "loss": 2.2923,
+      "step": 4339
+    },
+    {
+      "epoch": 0.18360267366105423,
+      "grad_norm": 0.2582836449146271,
+      "learning_rate": 0.001,
+      "loss": 3.0664,
+      "step": 4340
+    },
+    {
+      "epoch": 0.18364497842457062,
+      "grad_norm": 0.19488857686519623,
+      "learning_rate": 0.001,
+      "loss": 2.7243,
+      "step": 4341
+    },
+    {
+      "epoch": 0.18368728318808697,
+      "grad_norm": 0.26442408561706543,
+      "learning_rate": 0.001,
+      "loss": 3.4023,
+      "step": 4342
+    },
+    {
+      "epoch": 0.18372958795160335,
+      "grad_norm": 0.22754669189453125,
+      "learning_rate": 0.001,
+      "loss": 2.4041,
+      "step": 4343
+    },
+    {
+      "epoch": 0.18377189271511973,
+      "grad_norm": 0.22837084531784058,
+      "learning_rate": 0.001,
+      "loss": 3.132,
+      "step": 4344
+    },
+    {
+      "epoch": 0.1838141974786361,
+      "grad_norm": 6.180835723876953,
+      "learning_rate": 0.001,
+      "loss": 1.6405,
+      "step": 4345
+    },
+    {
+      "epoch": 0.18385650224215247,
+      "grad_norm": 0.21038007736206055,
+      "learning_rate": 0.001,
+      "loss": 2.0711,
+      "step": 4346
+    },
+    {
+      "epoch": 0.18389880700566885,
+      "grad_norm": 0.2657643258571625,
+      "learning_rate": 0.001,
+      "loss": 2.2425,
+      "step": 4347
+    },
+    {
+      "epoch": 0.1839411117691852,
+      "grad_norm": 1.5151898860931396,
+      "learning_rate": 0.001,
+      "loss": 2.9717,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1839834165327016,
+      "grad_norm": 0.2588333189487457,
+      "learning_rate": 0.001,
+      "loss": 2.7555,
+      "step": 4349
+    },
+    {
+      "epoch": 0.18402572129621794,
+      "grad_norm": 0.28635910153388977,
+      "learning_rate": 0.001,
+      "loss": 2.7952,
+      "step": 4350
+    },
+    {
+      "epoch": 0.18406802605973432,
+      "grad_norm": 0.32209694385528564,
+      "learning_rate": 0.001,
+      "loss": 2.6064,
+      "step": 4351
+    },
+    {
+      "epoch": 0.1841103308232507,
+      "grad_norm": 2.678985357284546,
+      "learning_rate": 0.001,
+      "loss": 2.9109,
+      "step": 4352
+    },
+    {
+      "epoch": 0.18415263558676706,
+      "grad_norm": 0.25267353653907776,
+      "learning_rate": 0.001,
+      "loss": 2.4621,
+      "step": 4353
+    },
+    {
+      "epoch": 0.18419494035028344,
+      "grad_norm": 0.24416348338127136,
+      "learning_rate": 0.001,
+      "loss": 1.9122,
+      "step": 4354
+    },
+    {
+      "epoch": 0.18423724511379982,
+      "grad_norm": 0.24533572793006897,
+      "learning_rate": 0.001,
+      "loss": 2.3288,
+      "step": 4355
+    },
+    {
+      "epoch": 0.18427954987731618,
+      "grad_norm": 0.2123999446630478,
+      "learning_rate": 0.001,
+      "loss": 2.8654,
+      "step": 4356
+    },
+    {
+      "epoch": 0.18432185464083256,
+      "grad_norm": 0.5792009234428406,
+      "learning_rate": 0.001,
+      "loss": 2.8559,
+      "step": 4357
+    },
+    {
+      "epoch": 0.18436415940434894,
+      "grad_norm": 0.30343350768089294,
+      "learning_rate": 0.001,
+      "loss": 2.3648,
+      "step": 4358
+    },
+    {
+      "epoch": 0.1844064641678653,
+      "grad_norm": 0.23615819215774536,
+      "learning_rate": 0.001,
+      "loss": 1.7143,
+      "step": 4359
+    },
+    {
+      "epoch": 0.18444876893138168,
+      "grad_norm": 0.2771995961666107,
+      "learning_rate": 0.001,
+      "loss": 2.4038,
+      "step": 4360
+    },
+    {
+      "epoch": 0.18449107369489803,
+      "grad_norm": 0.7265622615814209,
+      "learning_rate": 0.001,
+      "loss": 2.8253,
+      "step": 4361
+    },
+    {
+      "epoch": 0.18453337845841442,
+      "grad_norm": 0.1909170150756836,
+      "learning_rate": 0.001,
+      "loss": 2.1125,
+      "step": 4362
+    },
+    {
+      "epoch": 0.1845756832219308,
+      "grad_norm": 0.2617343068122864,
+      "learning_rate": 0.001,
+      "loss": 2.8421,
+      "step": 4363
+    },
+    {
+      "epoch": 0.18461798798544715,
+      "grad_norm": 0.6342441439628601,
+      "learning_rate": 0.001,
+      "loss": 2.0168,
+      "step": 4364
+    },
+    {
+      "epoch": 0.18466029274896353,
+      "grad_norm": 0.3092588484287262,
+      "learning_rate": 0.001,
+      "loss": 2.8462,
+      "step": 4365
+    },
+    {
+      "epoch": 0.18470259751247992,
+      "grad_norm": 0.27178463339805603,
+      "learning_rate": 0.001,
+      "loss": 2.1236,
+      "step": 4366
+    },
+    {
+      "epoch": 0.18474490227599627,
+      "grad_norm": 10.186301231384277,
+      "learning_rate": 0.001,
+      "loss": 2.949,
+      "step": 4367
+    },
+    {
+      "epoch": 0.18478720703951265,
+      "grad_norm": 0.23955731093883514,
+      "learning_rate": 0.001,
+      "loss": 2.6389,
+      "step": 4368
+    },
+    {
+      "epoch": 0.18482951180302903,
+      "grad_norm": 1.5058904886245728,
+      "learning_rate": 0.001,
+      "loss": 2.7837,
+      "step": 4369
+    },
+    {
+      "epoch": 0.1848718165665454,
+      "grad_norm": 2.1568853855133057,
+      "learning_rate": 0.001,
+      "loss": 3.1977,
+      "step": 4370
+    },
+    {
+      "epoch": 0.18491412133006177,
+      "grad_norm": 0.23633380234241486,
+      "learning_rate": 0.001,
+      "loss": 1.8045,
+      "step": 4371
+    },
+    {
+      "epoch": 0.18495642609357812,
+      "grad_norm": 0.31348666548728943,
+      "learning_rate": 0.001,
+      "loss": 2.6933,
+      "step": 4372
+    },
+    {
+      "epoch": 0.1849987308570945,
+      "grad_norm": 0.3345177471637726,
+      "learning_rate": 0.001,
+      "loss": 2.535,
+      "step": 4373
+    },
+    {
+      "epoch": 0.1850410356206109,
+      "grad_norm": 0.26245883107185364,
+      "learning_rate": 0.001,
+      "loss": 1.6552,
+      "step": 4374
+    },
+    {
+      "epoch": 0.18508334038412724,
+      "grad_norm": 0.5210462808609009,
+      "learning_rate": 0.001,
+      "loss": 2.2235,
+      "step": 4375
+    },
+    {
+      "epoch": 0.18512564514764362,
+      "grad_norm": 1.8073543310165405,
+      "learning_rate": 0.001,
+      "loss": 2.5607,
+      "step": 4376
+    },
+    {
+      "epoch": 0.18516794991116,
+      "grad_norm": 1.0985215902328491,
+      "learning_rate": 0.001,
+      "loss": 2.2265,
+      "step": 4377
+    },
+    {
+      "epoch": 0.18521025467467636,
+      "grad_norm": 0.4432186484336853,
+      "learning_rate": 0.001,
+      "loss": 3.1894,
+      "step": 4378
+    },
+    {
+      "epoch": 0.18525255943819274,
+      "grad_norm": 10.145984649658203,
+      "learning_rate": 0.001,
+      "loss": 2.348,
+      "step": 4379
+    },
+    {
+      "epoch": 0.18529486420170913,
+      "grad_norm": 0.7287303805351257,
+      "learning_rate": 0.001,
+      "loss": 3.4078,
+      "step": 4380
+    },
+    {
+      "epoch": 0.18533716896522548,
+      "grad_norm": 0.4048196077346802,
+      "learning_rate": 0.001,
+      "loss": 2.3973,
+      "step": 4381
+    },
+    {
+      "epoch": 0.18537947372874186,
+      "grad_norm": 0.3356500566005707,
+      "learning_rate": 0.001,
+      "loss": 1.8402,
+      "step": 4382
+    },
+    {
+      "epoch": 0.18542177849225822,
+      "grad_norm": 0.3641221225261688,
+      "learning_rate": 0.001,
+      "loss": 2.937,
+      "step": 4383
+    },
+    {
+      "epoch": 0.1854640832557746,
+      "grad_norm": 0.3666855990886688,
+      "learning_rate": 0.001,
+      "loss": 2.2261,
+      "step": 4384
+    },
+    {
+      "epoch": 0.18550638801929098,
+      "grad_norm": 0.34403887391090393,
+      "learning_rate": 0.001,
+      "loss": 2.4207,
+      "step": 4385
+    },
+    {
+      "epoch": 0.18554869278280733,
+      "grad_norm": 0.3421815037727356,
+      "learning_rate": 0.001,
+      "loss": 2.913,
+      "step": 4386
+    },
+    {
+      "epoch": 0.18559099754632372,
+      "grad_norm": 0.22695082426071167,
+      "learning_rate": 0.001,
+      "loss": 1.8004,
+      "step": 4387
+    },
+    {
+      "epoch": 0.1856333023098401,
+      "grad_norm": 0.24995842576026917,
+      "learning_rate": 0.001,
+      "loss": 1.5226,
+      "step": 4388
+    },
+    {
+      "epoch": 0.18567560707335645,
+      "grad_norm": 0.23015795648097992,
+      "learning_rate": 0.001,
+      "loss": 2.1822,
+      "step": 4389
+    },
+    {
+      "epoch": 0.18571791183687283,
+      "grad_norm": 0.26571375131607056,
+      "learning_rate": 0.001,
+      "loss": 2.0279,
+      "step": 4390
+    },
+    {
+      "epoch": 0.18576021660038922,
+      "grad_norm": 0.22029922902584076,
+      "learning_rate": 0.001,
+      "loss": 2.4631,
+      "step": 4391
+    },
+    {
+      "epoch": 0.18580252136390557,
+      "grad_norm": 0.24884076416492462,
+      "learning_rate": 0.001,
+      "loss": 2.1439,
+      "step": 4392
+    },
+    {
+      "epoch": 0.18584482612742195,
+      "grad_norm": 0.27514705061912537,
+      "learning_rate": 0.001,
+      "loss": 2.0805,
+      "step": 4393
+    },
+    {
+      "epoch": 0.1858871308909383,
+      "grad_norm": 0.23706066608428955,
+      "learning_rate": 0.001,
+      "loss": 2.3177,
+      "step": 4394
+    },
+    {
+      "epoch": 0.1859294356544547,
+      "grad_norm": 0.21739627420902252,
+      "learning_rate": 0.001,
+      "loss": 1.9557,
+      "step": 4395
+    },
+    {
+      "epoch": 0.18597174041797107,
+      "grad_norm": 0.3604746460914612,
+      "learning_rate": 0.001,
+      "loss": 1.6202,
+      "step": 4396
+    },
+    {
+      "epoch": 0.18601404518148743,
+      "grad_norm": 0.25236776471138,
+      "learning_rate": 0.001,
+      "loss": 2.6437,
+      "step": 4397
+    },
+    {
+      "epoch": 0.1860563499450038,
+      "grad_norm": 0.23242443799972534,
+      "learning_rate": 0.001,
+      "loss": 3.186,
+      "step": 4398
+    },
+    {
+      "epoch": 0.1860986547085202,
+      "grad_norm": 0.3794444799423218,
+      "learning_rate": 0.001,
+      "loss": 2.4195,
+      "step": 4399
+    },
+    {
+      "epoch": 0.18614095947203654,
+      "grad_norm": 0.23218633234500885,
+      "learning_rate": 0.001,
+      "loss": 2.6508,
+      "step": 4400
+    },
+    {
+      "epoch": 0.18618326423555293,
+      "grad_norm": 0.20233222842216492,
+      "learning_rate": 0.001,
+      "loss": 1.5472,
+      "step": 4401
+    },
+    {
+      "epoch": 0.1862255689990693,
+      "grad_norm": 0.6316521167755127,
+      "learning_rate": 0.001,
+      "loss": 3.3565,
+      "step": 4402
+    },
+    {
+      "epoch": 0.18626787376258566,
+      "grad_norm": 0.40583282709121704,
+      "learning_rate": 0.001,
+      "loss": 2.4887,
+      "step": 4403
+    },
+    {
+      "epoch": 0.18631017852610204,
+      "grad_norm": 0.35350820422172546,
+      "learning_rate": 0.001,
+      "loss": 2.5153,
+      "step": 4404
+    },
+    {
+      "epoch": 0.1863524832896184,
+      "grad_norm": 0.2667691707611084,
+      "learning_rate": 0.001,
+      "loss": 3.2235,
+      "step": 4405
+    },
+    {
+      "epoch": 0.18639478805313478,
+      "grad_norm": 0.7131944894790649,
+      "learning_rate": 0.001,
+      "loss": 3.317,
+      "step": 4406
+    },
+    {
+      "epoch": 0.18643709281665116,
+      "grad_norm": 0.2557836174964905,
+      "learning_rate": 0.001,
+      "loss": 2.8428,
+      "step": 4407
+    },
+    {
+      "epoch": 0.18647939758016752,
+      "grad_norm": 0.18979831039905548,
+      "learning_rate": 0.001,
+      "loss": 2.1681,
+      "step": 4408
+    },
+    {
+      "epoch": 0.1865217023436839,
+      "grad_norm": 0.8334029316902161,
+      "learning_rate": 0.001,
+      "loss": 1.9816,
+      "step": 4409
+    },
+    {
+      "epoch": 0.18656400710720028,
+      "grad_norm": 0.2034035176038742,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 4410
+    },
+    {
+      "epoch": 0.18660631187071663,
+      "grad_norm": 0.25661614537239075,
+      "learning_rate": 0.001,
+      "loss": 2.1606,
+      "step": 4411
+    },
+    {
+      "epoch": 0.18664861663423302,
+      "grad_norm": 0.21977266669273376,
+      "learning_rate": 0.001,
+      "loss": 2.2903,
+      "step": 4412
+    },
+    {
+      "epoch": 0.1866909213977494,
+      "grad_norm": 0.5622522234916687,
+      "learning_rate": 0.001,
+      "loss": 3.0276,
+      "step": 4413
+    },
+    {
+      "epoch": 0.18673322616126575,
+      "grad_norm": 0.2287818193435669,
+      "learning_rate": 0.001,
+      "loss": 3.0729,
+      "step": 4414
+    },
+    {
+      "epoch": 0.18677553092478213,
+      "grad_norm": 0.2905156910419464,
+      "learning_rate": 0.001,
+      "loss": 3.1699,
+      "step": 4415
+    },
+    {
+      "epoch": 0.1868178356882985,
+      "grad_norm": 0.23308047652244568,
+      "learning_rate": 0.001,
+      "loss": 2.4299,
+      "step": 4416
+    },
+    {
+      "epoch": 0.18686014045181487,
+      "grad_norm": 0.23100613057613373,
+      "learning_rate": 0.001,
+      "loss": 2.6107,
+      "step": 4417
+    },
+    {
+      "epoch": 0.18690244521533125,
+      "grad_norm": 0.2881180942058563,
+      "learning_rate": 0.001,
+      "loss": 2.3265,
+      "step": 4418
+    },
+    {
+      "epoch": 0.1869447499788476,
+      "grad_norm": 0.26562678813934326,
+      "learning_rate": 0.001,
+      "loss": 2.6836,
+      "step": 4419
+    },
+    {
+      "epoch": 0.186987054742364,
+      "grad_norm": 0.21036438643932343,
+      "learning_rate": 0.001,
+      "loss": 1.9917,
+      "step": 4420
+    },
+    {
+      "epoch": 0.18702935950588037,
+      "grad_norm": 0.27544263005256653,
+      "learning_rate": 0.001,
+      "loss": 3.4425,
+      "step": 4421
+    },
+    {
+      "epoch": 0.18707166426939673,
+      "grad_norm": 0.2592555284500122,
+      "learning_rate": 0.001,
+      "loss": 2.09,
+      "step": 4422
+    },
+    {
+      "epoch": 0.1871139690329131,
+      "grad_norm": 0.5881548523902893,
+      "learning_rate": 0.001,
+      "loss": 2.1932,
+      "step": 4423
+    },
+    {
+      "epoch": 0.1871562737964295,
+      "grad_norm": 0.20673885941505432,
+      "learning_rate": 0.001,
+      "loss": 2.2809,
+      "step": 4424
+    },
+    {
+      "epoch": 0.18719857855994584,
+      "grad_norm": 0.20291772484779358,
+      "learning_rate": 0.001,
+      "loss": 2.0593,
+      "step": 4425
+    },
+    {
+      "epoch": 0.18724088332346223,
+      "grad_norm": 0.270967960357666,
+      "learning_rate": 0.001,
+      "loss": 1.6903,
+      "step": 4426
+    },
+    {
+      "epoch": 0.18728318808697858,
+      "grad_norm": 0.35140615701675415,
+      "learning_rate": 0.001,
+      "loss": 2.364,
+      "step": 4427
+    },
+    {
+      "epoch": 0.18732549285049496,
+      "grad_norm": 0.20926691591739655,
+      "learning_rate": 0.001,
+      "loss": 1.9464,
+      "step": 4428
+    },
+    {
+      "epoch": 0.18736779761401134,
+      "grad_norm": 0.28219670057296753,
+      "learning_rate": 0.001,
+      "loss": 3.2835,
+      "step": 4429
+    },
+    {
+      "epoch": 0.1874101023775277,
+      "grad_norm": 3.9769115447998047,
+      "learning_rate": 0.001,
+      "loss": 2.0957,
+      "step": 4430
+    },
+    {
+      "epoch": 0.18745240714104408,
+      "grad_norm": 0.24822956323623657,
+      "learning_rate": 0.001,
+      "loss": 2.3957,
+      "step": 4431
+    },
+    {
+      "epoch": 0.18749471190456046,
+      "grad_norm": 0.22063516080379486,
+      "learning_rate": 0.001,
+      "loss": 2.1515,
+      "step": 4432
+    },
+    {
+      "epoch": 0.18753701666807682,
+      "grad_norm": 0.25612160563468933,
+      "learning_rate": 0.001,
+      "loss": 2.687,
+      "step": 4433
+    },
+    {
+      "epoch": 0.1875793214315932,
+      "grad_norm": 0.2492561936378479,
+      "learning_rate": 0.001,
+      "loss": 2.4043,
+      "step": 4434
+    },
+    {
+      "epoch": 0.18762162619510958,
+      "grad_norm": 0.28341639041900635,
+      "learning_rate": 0.001,
+      "loss": 2.0796,
+      "step": 4435
+    },
+    {
+      "epoch": 0.18766393095862594,
+      "grad_norm": 0.2945077121257782,
+      "learning_rate": 0.001,
+      "loss": 2.1302,
+      "step": 4436
+    },
+    {
+      "epoch": 0.18770623572214232,
+      "grad_norm": 0.3373602330684662,
+      "learning_rate": 0.001,
+      "loss": 2.0283,
+      "step": 4437
+    },
+    {
+      "epoch": 0.18774854048565867,
+      "grad_norm": 0.24458883702754974,
+      "learning_rate": 0.001,
+      "loss": 2.6162,
+      "step": 4438
+    },
+    {
+      "epoch": 0.18779084524917505,
+      "grad_norm": 0.21525456011295319,
+      "learning_rate": 0.001,
+      "loss": 2.0196,
+      "step": 4439
+    },
+    {
+      "epoch": 0.18783315001269144,
+      "grad_norm": 0.22924627363681793,
+      "learning_rate": 0.001,
+      "loss": 2.1364,
+      "step": 4440
+    },
+    {
+      "epoch": 0.1878754547762078,
+      "grad_norm": 0.22823546826839447,
+      "learning_rate": 0.001,
+      "loss": 3.0261,
+      "step": 4441
+    },
+    {
+      "epoch": 0.18791775953972417,
+      "grad_norm": 0.37336304783821106,
+      "learning_rate": 0.001,
+      "loss": 2.0595,
+      "step": 4442
+    },
+    {
+      "epoch": 0.18796006430324055,
+      "grad_norm": 2.6956934928894043,
+      "learning_rate": 0.001,
+      "loss": 1.9668,
+      "step": 4443
+    },
+    {
+      "epoch": 0.1880023690667569,
+      "grad_norm": 0.7836053371429443,
+      "learning_rate": 0.001,
+      "loss": 2.3485,
+      "step": 4444
+    },
+    {
+      "epoch": 0.1880446738302733,
+      "grad_norm": 1.6367077827453613,
+      "learning_rate": 0.001,
+      "loss": 3.084,
+      "step": 4445
+    },
+    {
+      "epoch": 0.18808697859378967,
+      "grad_norm": 0.26571542024612427,
+      "learning_rate": 0.001,
+      "loss": 2.4098,
+      "step": 4446
+    },
+    {
+      "epoch": 0.18812928335730603,
+      "grad_norm": 0.2825041711330414,
+      "learning_rate": 0.001,
+      "loss": 2.2918,
+      "step": 4447
+    },
+    {
+      "epoch": 0.1881715881208224,
+      "grad_norm": 0.246236190199852,
+      "learning_rate": 0.001,
+      "loss": 2.4611,
+      "step": 4448
+    },
+    {
+      "epoch": 0.1882138928843388,
+      "grad_norm": 0.5740474462509155,
+      "learning_rate": 0.001,
+      "loss": 2.9142,
+      "step": 4449
+    },
+    {
+      "epoch": 0.18825619764785514,
+      "grad_norm": 0.37695226073265076,
+      "learning_rate": 0.001,
+      "loss": 1.6922,
+      "step": 4450
+    },
+    {
+      "epoch": 0.18829850241137153,
+      "grad_norm": 1.8497018814086914,
+      "learning_rate": 0.001,
+      "loss": 2.2615,
+      "step": 4451
+    },
+    {
+      "epoch": 0.18834080717488788,
+      "grad_norm": 0.2767746150493622,
+      "learning_rate": 0.001,
+      "loss": 2.1198,
+      "step": 4452
+    },
+    {
+      "epoch": 0.18838311193840426,
+      "grad_norm": 1.5770310163497925,
+      "learning_rate": 0.001,
+      "loss": 2.4928,
+      "step": 4453
+    },
+    {
+      "epoch": 0.18842541670192064,
+      "grad_norm": 0.23543912172317505,
+      "learning_rate": 0.001,
+      "loss": 1.8614,
+      "step": 4454
+    },
+    {
+      "epoch": 0.188467721465437,
+      "grad_norm": 0.26487410068511963,
+      "learning_rate": 0.001,
+      "loss": 2.2644,
+      "step": 4455
+    },
+    {
+      "epoch": 0.18851002622895338,
+      "grad_norm": 0.26625239849090576,
+      "learning_rate": 0.001,
+      "loss": 2.4482,
+      "step": 4456
+    },
+    {
+      "epoch": 0.18855233099246976,
+      "grad_norm": 0.21155591309070587,
+      "learning_rate": 0.001,
+      "loss": 1.929,
+      "step": 4457
+    },
+    {
+      "epoch": 0.18859463575598612,
+      "grad_norm": 0.25420722365379333,
+      "learning_rate": 0.001,
+      "loss": 2.1863,
+      "step": 4458
+    },
+    {
+      "epoch": 0.1886369405195025,
+      "grad_norm": 0.27316761016845703,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 4459
+    },
+    {
+      "epoch": 0.18867924528301888,
+      "grad_norm": 3.036153793334961,
+      "learning_rate": 0.001,
+      "loss": 3.9133,
+      "step": 4460
+    },
+    {
+      "epoch": 0.18872155004653524,
+      "grad_norm": 0.22309522330760956,
+      "learning_rate": 0.001,
+      "loss": 2.348,
+      "step": 4461
+    },
+    {
+      "epoch": 0.18876385481005162,
+      "grad_norm": 0.21288998425006866,
+      "learning_rate": 0.001,
+      "loss": 1.5457,
+      "step": 4462
+    },
+    {
+      "epoch": 0.18880615957356797,
+      "grad_norm": 3.6792640686035156,
+      "learning_rate": 0.001,
+      "loss": 2.3953,
+      "step": 4463
+    },
+    {
+      "epoch": 0.18884846433708435,
+      "grad_norm": 0.3118633031845093,
+      "learning_rate": 0.001,
+      "loss": 3.1294,
+      "step": 4464
+    },
+    {
+      "epoch": 0.18889076910060074,
+      "grad_norm": 2.057178497314453,
+      "learning_rate": 0.001,
+      "loss": 1.7783,
+      "step": 4465
+    },
+    {
+      "epoch": 0.1889330738641171,
+      "grad_norm": 0.6279289722442627,
+      "learning_rate": 0.001,
+      "loss": 1.8288,
+      "step": 4466
+    },
+    {
+      "epoch": 0.18897537862763347,
+      "grad_norm": 0.3405144214630127,
+      "learning_rate": 0.001,
+      "loss": 2.3726,
+      "step": 4467
+    },
+    {
+      "epoch": 0.18901768339114985,
+      "grad_norm": 0.2901994585990906,
+      "learning_rate": 0.001,
+      "loss": 1.8338,
+      "step": 4468
+    },
+    {
+      "epoch": 0.1890599881546662,
+      "grad_norm": 0.3155980706214905,
+      "learning_rate": 0.001,
+      "loss": 3.4189,
+      "step": 4469
+    },
+    {
+      "epoch": 0.1891022929181826,
+      "grad_norm": 0.5617590546607971,
+      "learning_rate": 0.001,
+      "loss": 2.9148,
+      "step": 4470
+    },
+    {
+      "epoch": 0.18914459768169897,
+      "grad_norm": 0.3109961450099945,
+      "learning_rate": 0.001,
+      "loss": 2.4199,
+      "step": 4471
+    },
+    {
+      "epoch": 0.18918690244521533,
+      "grad_norm": 0.23848043382167816,
+      "learning_rate": 0.001,
+      "loss": 2.1432,
+      "step": 4472
+    },
+    {
+      "epoch": 0.1892292072087317,
+      "grad_norm": 0.2614109218120575,
+      "learning_rate": 0.001,
+      "loss": 1.9615,
+      "step": 4473
+    },
+    {
+      "epoch": 0.18927151197224806,
+      "grad_norm": 0.30078810453414917,
+      "learning_rate": 0.001,
+      "loss": 2.6696,
+      "step": 4474
+    },
+    {
+      "epoch": 0.18931381673576445,
+      "grad_norm": 0.25650012493133545,
+      "learning_rate": 0.001,
+      "loss": 1.8638,
+      "step": 4475
+    },
+    {
+      "epoch": 0.18935612149928083,
+      "grad_norm": 0.21869732439517975,
+      "learning_rate": 0.001,
+      "loss": 1.5891,
+      "step": 4476
+    },
+    {
+      "epoch": 0.18939842626279718,
+      "grad_norm": 8.205790519714355,
+      "learning_rate": 0.001,
+      "loss": 1.5669,
+      "step": 4477
+    },
+    {
+      "epoch": 0.18944073102631356,
+      "grad_norm": 0.2447681427001953,
+      "learning_rate": 0.001,
+      "loss": 2.9438,
+      "step": 4478
+    },
+    {
+      "epoch": 0.18948303578982995,
+      "grad_norm": 1.446384072303772,
+      "learning_rate": 0.001,
+      "loss": 2.7589,
+      "step": 4479
+    },
+    {
+      "epoch": 0.1895253405533463,
+      "grad_norm": 0.8409281373023987,
+      "learning_rate": 0.001,
+      "loss": 1.9919,
+      "step": 4480
+    },
+    {
+      "epoch": 0.18956764531686268,
+      "grad_norm": 0.27768608927726746,
+      "learning_rate": 0.001,
+      "loss": 2.4846,
+      "step": 4481
+    },
+    {
+      "epoch": 0.18960995008037906,
+      "grad_norm": 0.4152839779853821,
+      "learning_rate": 0.001,
+      "loss": 3.0854,
+      "step": 4482
+    },
+    {
+      "epoch": 0.18965225484389542,
+      "grad_norm": 0.2601878046989441,
+      "learning_rate": 0.001,
+      "loss": 2.7528,
+      "step": 4483
+    },
+    {
+      "epoch": 0.1896945596074118,
+      "grad_norm": 0.2510074973106384,
+      "learning_rate": 0.001,
+      "loss": 2.1918,
+      "step": 4484
+    },
+    {
+      "epoch": 0.18973686437092815,
+      "grad_norm": 0.22711610794067383,
+      "learning_rate": 0.001,
+      "loss": 2.3735,
+      "step": 4485
+    },
+    {
+      "epoch": 0.18977916913444454,
+      "grad_norm": 0.35560375452041626,
+      "learning_rate": 0.001,
+      "loss": 2.2848,
+      "step": 4486
+    },
+    {
+      "epoch": 0.18982147389796092,
+      "grad_norm": 0.27158334851264954,
+      "learning_rate": 0.001,
+      "loss": 2.6733,
+      "step": 4487
+    },
+    {
+      "epoch": 0.18986377866147727,
+      "grad_norm": 0.2838427424430847,
+      "learning_rate": 0.001,
+      "loss": 2.8558,
+      "step": 4488
+    },
+    {
+      "epoch": 0.18990608342499365,
+      "grad_norm": 0.2345813810825348,
+      "learning_rate": 0.001,
+      "loss": 3.6895,
+      "step": 4489
+    },
+    {
+      "epoch": 0.18994838818851004,
+      "grad_norm": 0.2404671013355255,
+      "learning_rate": 0.001,
+      "loss": 2.1784,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1899906929520264,
+      "grad_norm": 0.2407745122909546,
+      "learning_rate": 0.001,
+      "loss": 2.8477,
+      "step": 4491
+    },
+    {
+      "epoch": 0.19003299771554277,
+      "grad_norm": 0.2435643970966339,
+      "learning_rate": 0.001,
+      "loss": 2.7741,
+      "step": 4492
+    },
+    {
+      "epoch": 0.19007530247905915,
+      "grad_norm": 0.20882461965084076,
+      "learning_rate": 0.001,
+      "loss": 2.0404,
+      "step": 4493
+    },
+    {
+      "epoch": 0.1901176072425755,
+      "grad_norm": 0.20815284550189972,
+      "learning_rate": 0.001,
+      "loss": 1.4838,
+      "step": 4494
+    },
+    {
+      "epoch": 0.1901599120060919,
+      "grad_norm": 0.5364993810653687,
+      "learning_rate": 0.001,
+      "loss": 2.5901,
+      "step": 4495
+    },
+    {
+      "epoch": 0.19020221676960825,
+      "grad_norm": 0.18857216835021973,
+      "learning_rate": 0.001,
+      "loss": 2.0745,
+      "step": 4496
+    },
+    {
+      "epoch": 0.19024452153312463,
+      "grad_norm": 0.22083306312561035,
+      "learning_rate": 0.001,
+      "loss": 2.2613,
+      "step": 4497
+    },
+    {
+      "epoch": 0.190286826296641,
+      "grad_norm": 0.21830333769321442,
+      "learning_rate": 0.001,
+      "loss": 2.4491,
+      "step": 4498
+    },
+    {
+      "epoch": 0.19032913106015736,
+      "grad_norm": 0.918835461139679,
+      "learning_rate": 0.001,
+      "loss": 2.8763,
+      "step": 4499
+    },
+    {
+      "epoch": 0.19037143582367375,
+      "grad_norm": 0.2459859997034073,
+      "learning_rate": 0.001,
+      "loss": 2.3936,
+      "step": 4500
+    },
+    {
+      "epoch": 0.19041374058719013,
+      "grad_norm": 0.2378329634666443,
+      "learning_rate": 0.001,
+      "loss": 1.8399,
+      "step": 4501
+    },
+    {
+      "epoch": 0.19045604535070648,
+      "grad_norm": 0.3564843237400055,
+      "learning_rate": 0.001,
+      "loss": 3.5808,
+      "step": 4502
+    },
+    {
+      "epoch": 0.19049835011422286,
+      "grad_norm": 0.5608360171318054,
+      "learning_rate": 0.001,
+      "loss": 2.1496,
+      "step": 4503
+    },
+    {
+      "epoch": 0.19054065487773925,
+      "grad_norm": 0.3090474009513855,
+      "learning_rate": 0.001,
+      "loss": 2.3587,
+      "step": 4504
+    },
+    {
+      "epoch": 0.1905829596412556,
+      "grad_norm": 0.2552463114261627,
+      "learning_rate": 0.001,
+      "loss": 1.866,
+      "step": 4505
+    },
+    {
+      "epoch": 0.19062526440477198,
+      "grad_norm": 0.41491034626960754,
+      "learning_rate": 0.001,
+      "loss": 2.1764,
+      "step": 4506
+    },
+    {
+      "epoch": 0.19066756916828834,
+      "grad_norm": 0.2689819633960724,
+      "learning_rate": 0.001,
+      "loss": 1.6736,
+      "step": 4507
+    },
+    {
+      "epoch": 0.19070987393180472,
+      "grad_norm": 0.2358677089214325,
+      "learning_rate": 0.001,
+      "loss": 1.9717,
+      "step": 4508
+    },
+    {
+      "epoch": 0.1907521786953211,
+      "grad_norm": 0.2064763307571411,
+      "learning_rate": 0.001,
+      "loss": 2.6779,
+      "step": 4509
+    },
+    {
+      "epoch": 0.19079448345883745,
+      "grad_norm": 0.24154812097549438,
+      "learning_rate": 0.001,
+      "loss": 1.7884,
+      "step": 4510
+    },
+    {
+      "epoch": 0.19083678822235384,
+      "grad_norm": 0.22353674471378326,
+      "learning_rate": 0.001,
+      "loss": 1.9496,
+      "step": 4511
+    },
+    {
+      "epoch": 0.19087909298587022,
+      "grad_norm": 0.2455575168132782,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 4512
+    },
+    {
+      "epoch": 0.19092139774938657,
+      "grad_norm": 0.22272691130638123,
+      "learning_rate": 0.001,
+      "loss": 2.6156,
+      "step": 4513
+    },
+    {
+      "epoch": 0.19096370251290296,
+      "grad_norm": 0.21342520415782928,
+      "learning_rate": 0.001,
+      "loss": 2.2985,
+      "step": 4514
+    },
+    {
+      "epoch": 0.19100600727641934,
+      "grad_norm": 0.21540196239948273,
+      "learning_rate": 0.001,
+      "loss": 2.5549,
+      "step": 4515
+    },
+    {
+      "epoch": 0.1910483120399357,
+      "grad_norm": 0.20063146948814392,
+      "learning_rate": 0.001,
+      "loss": 3.4714,
+      "step": 4516
+    },
+    {
+      "epoch": 0.19109061680345207,
+      "grad_norm": 0.1966932862997055,
+      "learning_rate": 0.001,
+      "loss": 2.1782,
+      "step": 4517
+    },
+    {
+      "epoch": 0.19113292156696843,
+      "grad_norm": 1.2073802947998047,
+      "learning_rate": 0.001,
+      "loss": 2.668,
+      "step": 4518
+    },
+    {
+      "epoch": 0.1911752263304848,
+      "grad_norm": 0.3695375323295593,
+      "learning_rate": 0.001,
+      "loss": 3.0143,
+      "step": 4519
+    },
+    {
+      "epoch": 0.1912175310940012,
+      "grad_norm": 0.21040910482406616,
+      "learning_rate": 0.001,
+      "loss": 2.2652,
+      "step": 4520
+    },
+    {
+      "epoch": 0.19125983585751755,
+      "grad_norm": 0.2612590491771698,
+      "learning_rate": 0.001,
+      "loss": 2.0682,
+      "step": 4521
+    },
+    {
+      "epoch": 0.19130214062103393,
+      "grad_norm": 1.0337492227554321,
+      "learning_rate": 0.001,
+      "loss": 1.7501,
+      "step": 4522
+    },
+    {
+      "epoch": 0.1913444453845503,
+      "grad_norm": 0.2554808259010315,
+      "learning_rate": 0.001,
+      "loss": 2.4155,
+      "step": 4523
+    },
+    {
+      "epoch": 0.19138675014806666,
+      "grad_norm": 2.867785692214966,
+      "learning_rate": 0.001,
+      "loss": 1.7372,
+      "step": 4524
+    },
+    {
+      "epoch": 0.19142905491158305,
+      "grad_norm": 3.803913116455078,
+      "learning_rate": 0.001,
+      "loss": 2.8557,
+      "step": 4525
+    },
+    {
+      "epoch": 0.19147135967509943,
+      "grad_norm": 0.671362578868866,
+      "learning_rate": 0.001,
+      "loss": 1.9373,
+      "step": 4526
+    },
+    {
+      "epoch": 0.19151366443861578,
+      "grad_norm": 0.33404502272605896,
+      "learning_rate": 0.001,
+      "loss": 2.3405,
+      "step": 4527
+    },
+    {
+      "epoch": 0.19155596920213216,
+      "grad_norm": 7.813088893890381,
+      "learning_rate": 0.001,
+      "loss": 3.0833,
+      "step": 4528
+    },
+    {
+      "epoch": 0.19159827396564852,
+      "grad_norm": 0.33136120438575745,
+      "learning_rate": 0.001,
+      "loss": 3.0437,
+      "step": 4529
+    },
+    {
+      "epoch": 0.1916405787291649,
+      "grad_norm": 0.3208834230899811,
+      "learning_rate": 0.001,
+      "loss": 2.8585,
+      "step": 4530
+    },
+    {
+      "epoch": 0.19168288349268128,
+      "grad_norm": 0.46846067905426025,
+      "learning_rate": 0.001,
+      "loss": 3.273,
+      "step": 4531
+    },
+    {
+      "epoch": 0.19172518825619764,
+      "grad_norm": 0.4278513789176941,
+      "learning_rate": 0.001,
+      "loss": 2.7628,
+      "step": 4532
+    },
+    {
+      "epoch": 0.19176749301971402,
+      "grad_norm": 0.3164735734462738,
+      "learning_rate": 0.001,
+      "loss": 3.0253,
+      "step": 4533
+    },
+    {
+      "epoch": 0.1918097977832304,
+      "grad_norm": 0.22496308386325836,
+      "learning_rate": 0.001,
+      "loss": 2.1171,
+      "step": 4534
+    },
+    {
+      "epoch": 0.19185210254674676,
+      "grad_norm": 0.2776702344417572,
+      "learning_rate": 0.001,
+      "loss": 2.7838,
+      "step": 4535
+    },
+    {
+      "epoch": 0.19189440731026314,
+      "grad_norm": 0.6228007674217224,
+      "learning_rate": 0.001,
+      "loss": 2.2344,
+      "step": 4536
+    },
+    {
+      "epoch": 0.19193671207377952,
+      "grad_norm": 0.48833128809928894,
+      "learning_rate": 0.001,
+      "loss": 2.986,
+      "step": 4537
+    },
+    {
+      "epoch": 0.19197901683729587,
+      "grad_norm": 0.3756866753101349,
+      "learning_rate": 0.001,
+      "loss": 1.9291,
+      "step": 4538
+    },
+    {
+      "epoch": 0.19202132160081226,
+      "grad_norm": 0.278314471244812,
+      "learning_rate": 0.001,
+      "loss": 2.1759,
+      "step": 4539
+    },
+    {
+      "epoch": 0.1920636263643286,
+      "grad_norm": 1.2533029317855835,
+      "learning_rate": 0.001,
+      "loss": 2.107,
+      "step": 4540
+    },
+    {
+      "epoch": 0.192105931127845,
+      "grad_norm": 0.2423766851425171,
+      "learning_rate": 0.001,
+      "loss": 1.7797,
+      "step": 4541
+    },
+    {
+      "epoch": 0.19214823589136137,
+      "grad_norm": 0.3214251399040222,
+      "learning_rate": 0.001,
+      "loss": 2.94,
+      "step": 4542
+    },
+    {
+      "epoch": 0.19219054065487773,
+      "grad_norm": 0.4931315779685974,
+      "learning_rate": 0.001,
+      "loss": 1.9467,
+      "step": 4543
+    },
+    {
+      "epoch": 0.1922328454183941,
+      "grad_norm": 2.458348035812378,
+      "learning_rate": 0.001,
+      "loss": 2.5617,
+      "step": 4544
+    },
+    {
+      "epoch": 0.1922751501819105,
+      "grad_norm": 1.2763197422027588,
+      "learning_rate": 0.001,
+      "loss": 2.5777,
+      "step": 4545
+    },
+    {
+      "epoch": 0.19231745494542685,
+      "grad_norm": 0.2197532057762146,
+      "learning_rate": 0.001,
+      "loss": 1.8814,
+      "step": 4546
+    },
+    {
+      "epoch": 0.19235975970894323,
+      "grad_norm": 0.335099458694458,
+      "learning_rate": 0.001,
+      "loss": 1.9736,
+      "step": 4547
+    },
+    {
+      "epoch": 0.1924020644724596,
+      "grad_norm": 0.24733540415763855,
+      "learning_rate": 0.001,
+      "loss": 1.7578,
+      "step": 4548
+    },
+    {
+      "epoch": 0.19244436923597597,
+      "grad_norm": 0.2788896858692169,
+      "learning_rate": 0.001,
+      "loss": 3.6972,
+      "step": 4549
+    },
+    {
+      "epoch": 0.19248667399949235,
+      "grad_norm": 0.250205397605896,
+      "learning_rate": 0.001,
+      "loss": 2.5927,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1925289787630087,
+      "grad_norm": 0.9215936660766602,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 4551
+    },
+    {
+      "epoch": 0.19257128352652508,
+      "grad_norm": 0.2189418524503708,
+      "learning_rate": 0.001,
+      "loss": 2.3595,
+      "step": 4552
+    },
+    {
+      "epoch": 0.19261358829004147,
+      "grad_norm": 0.2297450602054596,
+      "learning_rate": 0.001,
+      "loss": 2.0018,
+      "step": 4553
+    },
+    {
+      "epoch": 0.19265589305355782,
+      "grad_norm": 0.8725435137748718,
+      "learning_rate": 0.001,
+      "loss": 2.3299,
+      "step": 4554
+    },
+    {
+      "epoch": 0.1926981978170742,
+      "grad_norm": 0.2853243350982666,
+      "learning_rate": 0.001,
+      "loss": 3.1252,
+      "step": 4555
+    },
+    {
+      "epoch": 0.19274050258059058,
+      "grad_norm": 0.26800212264060974,
+      "learning_rate": 0.001,
+      "loss": 2.5468,
+      "step": 4556
+    },
+    {
+      "epoch": 0.19278280734410694,
+      "grad_norm": 0.4624558091163635,
+      "learning_rate": 0.001,
+      "loss": 2.0941,
+      "step": 4557
+    },
+    {
+      "epoch": 0.19282511210762332,
+      "grad_norm": 0.2232256382703781,
+      "learning_rate": 0.001,
+      "loss": 2.4683,
+      "step": 4558
+    },
+    {
+      "epoch": 0.1928674168711397,
+      "grad_norm": 0.2688036859035492,
+      "learning_rate": 0.001,
+      "loss": 2.223,
+      "step": 4559
+    },
+    {
+      "epoch": 0.19290972163465606,
+      "grad_norm": 0.2234850972890854,
+      "learning_rate": 0.001,
+      "loss": 2.6623,
+      "step": 4560
+    },
+    {
+      "epoch": 0.19295202639817244,
+      "grad_norm": 0.3354385495185852,
+      "learning_rate": 0.001,
+      "loss": 3.2651,
+      "step": 4561
+    },
+    {
+      "epoch": 0.1929943311616888,
+      "grad_norm": 0.26778674125671387,
+      "learning_rate": 0.001,
+      "loss": 2.6781,
+      "step": 4562
+    },
+    {
+      "epoch": 0.19303663592520517,
+      "grad_norm": 0.2896103858947754,
+      "learning_rate": 0.001,
+      "loss": 2.2987,
+      "step": 4563
+    },
+    {
+      "epoch": 0.19307894068872156,
+      "grad_norm": 0.21023137867450714,
+      "learning_rate": 0.001,
+      "loss": 1.6636,
+      "step": 4564
+    },
+    {
+      "epoch": 0.1931212454522379,
+      "grad_norm": 0.2824184000492096,
+      "learning_rate": 0.001,
+      "loss": 2.5228,
+      "step": 4565
+    },
+    {
+      "epoch": 0.1931635502157543,
+      "grad_norm": 0.22193598747253418,
+      "learning_rate": 0.001,
+      "loss": 1.9797,
+      "step": 4566
+    },
+    {
+      "epoch": 0.19320585497927067,
+      "grad_norm": 0.26808369159698486,
+      "learning_rate": 0.001,
+      "loss": 2.533,
+      "step": 4567
+    },
+    {
+      "epoch": 0.19324815974278703,
+      "grad_norm": 0.2427438348531723,
+      "learning_rate": 0.001,
+      "loss": 2.6625,
+      "step": 4568
+    },
+    {
+      "epoch": 0.1932904645063034,
+      "grad_norm": 0.2184421867132187,
+      "learning_rate": 0.001,
+      "loss": 2.5565,
+      "step": 4569
+    },
+    {
+      "epoch": 0.1933327692698198,
+      "grad_norm": 0.4565901756286621,
+      "learning_rate": 0.001,
+      "loss": 2.0758,
+      "step": 4570
+    },
+    {
+      "epoch": 0.19337507403333615,
+      "grad_norm": 0.9183976650238037,
+      "learning_rate": 0.001,
+      "loss": 2.1397,
+      "step": 4571
+    },
+    {
+      "epoch": 0.19341737879685253,
+      "grad_norm": 0.20792505145072937,
+      "learning_rate": 0.001,
+      "loss": 2.7359,
+      "step": 4572
+    },
+    {
+      "epoch": 0.1934596835603689,
+      "grad_norm": 0.7904244661331177,
+      "learning_rate": 0.001,
+      "loss": 3.3685,
+      "step": 4573
+    },
+    {
+      "epoch": 0.19350198832388527,
+      "grad_norm": 0.2975200116634369,
+      "learning_rate": 0.001,
+      "loss": 3.3311,
+      "step": 4574
+    },
+    {
+      "epoch": 0.19354429308740165,
+      "grad_norm": 4.709847450256348,
+      "learning_rate": 0.001,
+      "loss": 2.6445,
+      "step": 4575
+    },
+    {
+      "epoch": 0.193586597850918,
+      "grad_norm": 0.23912140727043152,
+      "learning_rate": 0.001,
+      "loss": 2.272,
+      "step": 4576
+    },
+    {
+      "epoch": 0.19362890261443438,
+      "grad_norm": 0.3137105405330658,
+      "learning_rate": 0.001,
+      "loss": 2.5401,
+      "step": 4577
+    },
+    {
+      "epoch": 0.19367120737795077,
+      "grad_norm": 0.27049148082733154,
+      "learning_rate": 0.001,
+      "loss": 2.8427,
+      "step": 4578
+    },
+    {
+      "epoch": 0.19371351214146712,
+      "grad_norm": 0.3511445224285126,
+      "learning_rate": 0.001,
+      "loss": 1.8104,
+      "step": 4579
+    },
+    {
+      "epoch": 0.1937558169049835,
+      "grad_norm": 0.280909925699234,
+      "learning_rate": 0.001,
+      "loss": 1.8538,
+      "step": 4580
+    },
+    {
+      "epoch": 0.19379812166849988,
+      "grad_norm": 0.2330571413040161,
+      "learning_rate": 0.001,
+      "loss": 1.9844,
+      "step": 4581
+    },
+    {
+      "epoch": 0.19384042643201624,
+      "grad_norm": 0.2649915814399719,
+      "learning_rate": 0.001,
+      "loss": 3.0022,
+      "step": 4582
+    },
+    {
+      "epoch": 0.19388273119553262,
+      "grad_norm": 0.28710460662841797,
+      "learning_rate": 0.001,
+      "loss": 2.5591,
+      "step": 4583
+    },
+    {
+      "epoch": 0.193925035959049,
+      "grad_norm": 0.22108560800552368,
+      "learning_rate": 0.001,
+      "loss": 3.5141,
+      "step": 4584
+    },
+    {
+      "epoch": 0.19396734072256536,
+      "grad_norm": 0.21730417013168335,
+      "learning_rate": 0.001,
+      "loss": 2.453,
+      "step": 4585
+    },
+    {
+      "epoch": 0.19400964548608174,
+      "grad_norm": 0.4755695164203644,
+      "learning_rate": 0.001,
+      "loss": 2.3401,
+      "step": 4586
+    },
+    {
+      "epoch": 0.1940519502495981,
+      "grad_norm": 0.2430569976568222,
+      "learning_rate": 0.001,
+      "loss": 2.5329,
+      "step": 4587
+    },
+    {
+      "epoch": 0.19409425501311448,
+      "grad_norm": 3.246774435043335,
+      "learning_rate": 0.001,
+      "loss": 2.1003,
+      "step": 4588
+    },
+    {
+      "epoch": 0.19413655977663086,
+      "grad_norm": 0.22115357220172882,
+      "learning_rate": 0.001,
+      "loss": 2.791,
+      "step": 4589
+    },
+    {
+      "epoch": 0.1941788645401472,
+      "grad_norm": 0.3158797025680542,
+      "learning_rate": 0.001,
+      "loss": 2.5248,
+      "step": 4590
+    },
+    {
+      "epoch": 0.1942211693036636,
+      "grad_norm": 0.21238943934440613,
+      "learning_rate": 0.001,
+      "loss": 2.0581,
+      "step": 4591
+    },
+    {
+      "epoch": 0.19426347406717998,
+      "grad_norm": 0.2093350887298584,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 4592
+    },
+    {
+      "epoch": 0.19430577883069633,
+      "grad_norm": 0.22821201384067535,
+      "learning_rate": 0.001,
+      "loss": 3.2084,
+      "step": 4593
+    },
+    {
+      "epoch": 0.1943480835942127,
+      "grad_norm": 0.26408034563064575,
+      "learning_rate": 0.001,
+      "loss": 2.2652,
+      "step": 4594
+    },
+    {
+      "epoch": 0.1943903883577291,
+      "grad_norm": 0.24591988325119019,
+      "learning_rate": 0.001,
+      "loss": 2.1064,
+      "step": 4595
+    },
+    {
+      "epoch": 0.19443269312124545,
+      "grad_norm": 5.068770408630371,
+      "learning_rate": 0.001,
+      "loss": 1.804,
+      "step": 4596
+    },
+    {
+      "epoch": 0.19447499788476183,
+      "grad_norm": 0.22931112349033356,
+      "learning_rate": 0.001,
+      "loss": 1.7943,
+      "step": 4597
+    },
+    {
+      "epoch": 0.19451730264827818,
+      "grad_norm": 0.5310946702957153,
+      "learning_rate": 0.001,
+      "loss": 2.0597,
+      "step": 4598
+    },
+    {
+      "epoch": 0.19455960741179457,
+      "grad_norm": 0.30972105264663696,
+      "learning_rate": 0.001,
+      "loss": 3.2119,
+      "step": 4599
+    },
+    {
+      "epoch": 0.19460191217531095,
+      "grad_norm": 0.2220851629972458,
+      "learning_rate": 0.001,
+      "loss": 2.2495,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1946442169388273,
+      "grad_norm": 0.871143639087677,
+      "learning_rate": 0.001,
+      "loss": 2.6081,
+      "step": 4601
+    },
+    {
+      "epoch": 0.19468652170234368,
+      "grad_norm": 0.2440234124660492,
+      "learning_rate": 0.001,
+      "loss": 3.2444,
+      "step": 4602
+    },
+    {
+      "epoch": 0.19472882646586007,
+      "grad_norm": 15.243217468261719,
+      "learning_rate": 0.001,
+      "loss": 2.0989,
+      "step": 4603
+    },
+    {
+      "epoch": 0.19477113122937642,
+      "grad_norm": 0.22179925441741943,
+      "learning_rate": 0.001,
+      "loss": 2.6825,
+      "step": 4604
+    },
+    {
+      "epoch": 0.1948134359928928,
+      "grad_norm": 0.2920258641242981,
+      "learning_rate": 0.001,
+      "loss": 1.8067,
+      "step": 4605
+    },
+    {
+      "epoch": 0.19485574075640918,
+      "grad_norm": 0.45571228861808777,
+      "learning_rate": 0.001,
+      "loss": 2.4332,
+      "step": 4606
+    },
+    {
+      "epoch": 0.19489804551992554,
+      "grad_norm": 0.21809923648834229,
+      "learning_rate": 0.001,
+      "loss": 1.9698,
+      "step": 4607
+    },
+    {
+      "epoch": 0.19494035028344192,
+      "grad_norm": 0.22986778616905212,
+      "learning_rate": 0.001,
+      "loss": 2.3261,
+      "step": 4608
+    },
+    {
+      "epoch": 0.19498265504695828,
+      "grad_norm": 0.2756772041320801,
+      "learning_rate": 0.001,
+      "loss": 3.1784,
+      "step": 4609
+    },
+    {
+      "epoch": 0.19502495981047466,
+      "grad_norm": 0.20755811035633087,
+      "learning_rate": 0.001,
+      "loss": 2.1669,
+      "step": 4610
+    },
+    {
+      "epoch": 0.19506726457399104,
+      "grad_norm": 0.33731144666671753,
+      "learning_rate": 0.001,
+      "loss": 2.3201,
+      "step": 4611
+    },
+    {
+      "epoch": 0.1951095693375074,
+      "grad_norm": 0.35516980290412903,
+      "learning_rate": 0.001,
+      "loss": 2.119,
+      "step": 4612
+    },
+    {
+      "epoch": 0.19515187410102378,
+      "grad_norm": 0.21821679174900055,
+      "learning_rate": 0.001,
+      "loss": 3.3628,
+      "step": 4613
+    },
+    {
+      "epoch": 0.19519417886454016,
+      "grad_norm": 0.23565897345542908,
+      "learning_rate": 0.001,
+      "loss": 2.407,
+      "step": 4614
+    },
+    {
+      "epoch": 0.1952364836280565,
+      "grad_norm": 0.19071324169635773,
+      "learning_rate": 0.001,
+      "loss": 2.0236,
+      "step": 4615
+    },
+    {
+      "epoch": 0.1952787883915729,
+      "grad_norm": 0.21241721510887146,
+      "learning_rate": 0.001,
+      "loss": 1.9463,
+      "step": 4616
+    },
+    {
+      "epoch": 0.19532109315508928,
+      "grad_norm": 0.21527446806430817,
+      "learning_rate": 0.001,
+      "loss": 2.137,
+      "step": 4617
+    },
+    {
+      "epoch": 0.19536339791860563,
+      "grad_norm": 0.21654748916625977,
+      "learning_rate": 0.001,
+      "loss": 2.9063,
+      "step": 4618
+    },
+    {
+      "epoch": 0.195405702682122,
+      "grad_norm": 0.26138582825660706,
+      "learning_rate": 0.001,
+      "loss": 2.4052,
+      "step": 4619
+    },
+    {
+      "epoch": 0.19544800744563837,
+      "grad_norm": 0.2170308381319046,
+      "learning_rate": 0.001,
+      "loss": 1.9937,
+      "step": 4620
+    },
+    {
+      "epoch": 0.19549031220915475,
+      "grad_norm": 1.4368832111358643,
+      "learning_rate": 0.001,
+      "loss": 2.1986,
+      "step": 4621
+    },
+    {
+      "epoch": 0.19553261697267113,
+      "grad_norm": 0.21207021176815033,
+      "learning_rate": 0.001,
+      "loss": 1.8493,
+      "step": 4622
+    },
+    {
+      "epoch": 0.19557492173618748,
+      "grad_norm": 0.3097752034664154,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 4623
+    },
+    {
+      "epoch": 0.19561722649970387,
+      "grad_norm": 0.2724337875843048,
+      "learning_rate": 0.001,
+      "loss": 2.3367,
+      "step": 4624
+    },
+    {
+      "epoch": 0.19565953126322025,
+      "grad_norm": 0.21333764493465424,
+      "learning_rate": 0.001,
+      "loss": 1.6781,
+      "step": 4625
+    },
+    {
+      "epoch": 0.1957018360267366,
+      "grad_norm": 0.30645281076431274,
+      "learning_rate": 0.001,
+      "loss": 2.288,
+      "step": 4626
+    },
+    {
+      "epoch": 0.19574414079025299,
+      "grad_norm": 0.2575608193874359,
+      "learning_rate": 0.001,
+      "loss": 2.4528,
+      "step": 4627
+    },
+    {
+      "epoch": 0.19578644555376937,
+      "grad_norm": 0.21486923098564148,
+      "learning_rate": 0.001,
+      "loss": 1.9437,
+      "step": 4628
+    },
+    {
+      "epoch": 0.19582875031728572,
+      "grad_norm": 0.24129249155521393,
+      "learning_rate": 0.001,
+      "loss": 3.5545,
+      "step": 4629
+    },
+    {
+      "epoch": 0.1958710550808021,
+      "grad_norm": 0.3026060461997986,
+      "learning_rate": 0.001,
+      "loss": 2.1121,
+      "step": 4630
+    },
+    {
+      "epoch": 0.19591335984431846,
+      "grad_norm": 0.2635842561721802,
+      "learning_rate": 0.001,
+      "loss": 2.1426,
+      "step": 4631
+    },
+    {
+      "epoch": 0.19595566460783484,
+      "grad_norm": 0.38324952125549316,
+      "learning_rate": 0.001,
+      "loss": 2.0874,
+      "step": 4632
+    },
+    {
+      "epoch": 0.19599796937135122,
+      "grad_norm": 0.6430294513702393,
+      "learning_rate": 0.001,
+      "loss": 2.474,
+      "step": 4633
+    },
+    {
+      "epoch": 0.19604027413486758,
+      "grad_norm": 0.2072354555130005,
+      "learning_rate": 0.001,
+      "loss": 3.0056,
+      "step": 4634
+    },
+    {
+      "epoch": 0.19608257889838396,
+      "grad_norm": 0.22297710180282593,
+      "learning_rate": 0.001,
+      "loss": 2.1925,
+      "step": 4635
+    },
+    {
+      "epoch": 0.19612488366190034,
+      "grad_norm": 2.166142702102661,
+      "learning_rate": 0.001,
+      "loss": 2.1752,
+      "step": 4636
+    },
+    {
+      "epoch": 0.1961671884254167,
+      "grad_norm": 0.21612171828746796,
+      "learning_rate": 0.001,
+      "loss": 3.9448,
+      "step": 4637
+    },
+    {
+      "epoch": 0.19620949318893308,
+      "grad_norm": 0.2798874080181122,
+      "learning_rate": 0.001,
+      "loss": 2.152,
+      "step": 4638
+    },
+    {
+      "epoch": 0.19625179795244946,
+      "grad_norm": 0.29378223419189453,
+      "learning_rate": 0.001,
+      "loss": 1.913,
+      "step": 4639
+    },
+    {
+      "epoch": 0.1962941027159658,
+      "grad_norm": 0.37038248777389526,
+      "learning_rate": 0.001,
+      "loss": 2.5038,
+      "step": 4640
+    },
+    {
+      "epoch": 0.1963364074794822,
+      "grad_norm": 0.2652187645435333,
+      "learning_rate": 0.001,
+      "loss": 2.1267,
+      "step": 4641
+    },
+    {
+      "epoch": 0.19637871224299855,
+      "grad_norm": 0.27471134066581726,
+      "learning_rate": 0.001,
+      "loss": 2.1653,
+      "step": 4642
+    },
+    {
+      "epoch": 0.19642101700651493,
+      "grad_norm": 0.19736327230930328,
+      "learning_rate": 0.001,
+      "loss": 2.9415,
+      "step": 4643
+    },
+    {
+      "epoch": 0.1964633217700313,
+      "grad_norm": 0.22357088327407837,
+      "learning_rate": 0.001,
+      "loss": 3.3843,
+      "step": 4644
+    },
+    {
+      "epoch": 0.19650562653354767,
+      "grad_norm": 0.22968696057796478,
+      "learning_rate": 0.001,
+      "loss": 2.0118,
+      "step": 4645
+    },
+    {
+      "epoch": 0.19654793129706405,
+      "grad_norm": 1.8765842914581299,
+      "learning_rate": 0.001,
+      "loss": 2.4737,
+      "step": 4646
+    },
+    {
+      "epoch": 0.19659023606058043,
+      "grad_norm": 0.41798344254493713,
+      "learning_rate": 0.001,
+      "loss": 2.6721,
+      "step": 4647
+    },
+    {
+      "epoch": 0.19663254082409679,
+      "grad_norm": 0.20289246737957,
+      "learning_rate": 0.001,
+      "loss": 3.4892,
+      "step": 4648
+    },
+    {
+      "epoch": 0.19667484558761317,
+      "grad_norm": 0.2355775684118271,
+      "learning_rate": 0.001,
+      "loss": 1.8723,
+      "step": 4649
+    },
+    {
+      "epoch": 0.19671715035112955,
+      "grad_norm": 0.2144707590341568,
+      "learning_rate": 0.001,
+      "loss": 2.787,
+      "step": 4650
+    },
+    {
+      "epoch": 0.1967594551146459,
+      "grad_norm": 0.23321262001991272,
+      "learning_rate": 0.001,
+      "loss": 1.7408,
+      "step": 4651
+    },
+    {
+      "epoch": 0.19680175987816229,
+      "grad_norm": 0.21460668742656708,
+      "learning_rate": 0.001,
+      "loss": 1.8128,
+      "step": 4652
+    },
+    {
+      "epoch": 0.19684406464167864,
+      "grad_norm": 0.6912943124771118,
+      "learning_rate": 0.001,
+      "loss": 2.8737,
+      "step": 4653
+    },
+    {
+      "epoch": 0.19688636940519502,
+      "grad_norm": 0.24109399318695068,
+      "learning_rate": 0.001,
+      "loss": 2.2734,
+      "step": 4654
+    },
+    {
+      "epoch": 0.1969286741687114,
+      "grad_norm": 0.47939634323120117,
+      "learning_rate": 0.001,
+      "loss": 1.6569,
+      "step": 4655
+    },
+    {
+      "epoch": 0.19697097893222776,
+      "grad_norm": 2.933166742324829,
+      "learning_rate": 0.001,
+      "loss": 2.6041,
+      "step": 4656
+    },
+    {
+      "epoch": 0.19701328369574414,
+      "grad_norm": 0.8214706778526306,
+      "learning_rate": 0.001,
+      "loss": 2.518,
+      "step": 4657
+    },
+    {
+      "epoch": 0.19705558845926052,
+      "grad_norm": 0.2619834542274475,
+      "learning_rate": 0.001,
+      "loss": 3.2975,
+      "step": 4658
+    },
+    {
+      "epoch": 0.19709789322277688,
+      "grad_norm": 0.3553191125392914,
+      "learning_rate": 0.001,
+      "loss": 3.8384,
+      "step": 4659
+    },
+    {
+      "epoch": 0.19714019798629326,
+      "grad_norm": 0.1962711215019226,
+      "learning_rate": 0.001,
+      "loss": 1.984,
+      "step": 4660
+    },
+    {
+      "epoch": 0.19718250274980964,
+      "grad_norm": 0.2404264360666275,
+      "learning_rate": 0.001,
+      "loss": 2.4286,
+      "step": 4661
+    },
+    {
+      "epoch": 0.197224807513326,
+      "grad_norm": 0.24695684015750885,
+      "learning_rate": 0.001,
+      "loss": 2.8653,
+      "step": 4662
+    },
+    {
+      "epoch": 0.19726711227684238,
+      "grad_norm": 20.713294982910156,
+      "learning_rate": 0.001,
+      "loss": 2.2609,
+      "step": 4663
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.25775420665740967,
+      "learning_rate": 0.001,
+      "loss": 2.2937,
+      "step": 4664
+    },
+    {
+      "epoch": 0.1973517218038751,
+      "grad_norm": 0.24397379159927368,
+      "learning_rate": 0.001,
+      "loss": 2.0571,
+      "step": 4665
+    },
+    {
+      "epoch": 0.1973940265673915,
+      "grad_norm": 0.38282090425491333,
+      "learning_rate": 0.001,
+      "loss": 2.2702,
+      "step": 4666
+    },
+    {
+      "epoch": 0.19743633133090785,
+      "grad_norm": 0.23603591322898865,
+      "learning_rate": 0.001,
+      "loss": 2.01,
+      "step": 4667
+    },
+    {
+      "epoch": 0.19747863609442423,
+      "grad_norm": 0.2705064117908478,
+      "learning_rate": 0.001,
+      "loss": 2.6685,
+      "step": 4668
+    },
+    {
+      "epoch": 0.1975209408579406,
+      "grad_norm": 3.485954999923706,
+      "learning_rate": 0.001,
+      "loss": 3.6868,
+      "step": 4669
+    },
+    {
+      "epoch": 0.19756324562145697,
+      "grad_norm": 0.3059000074863434,
+      "learning_rate": 0.001,
+      "loss": 2.5588,
+      "step": 4670
+    },
+    {
+      "epoch": 0.19760555038497335,
+      "grad_norm": 2.187786102294922,
+      "learning_rate": 0.001,
+      "loss": 3.4677,
+      "step": 4671
+    },
+    {
+      "epoch": 0.19764785514848973,
+      "grad_norm": 0.25730597972869873,
+      "learning_rate": 0.001,
+      "loss": 2.3108,
+      "step": 4672
+    },
+    {
+      "epoch": 0.19769015991200609,
+      "grad_norm": 0.2296760380268097,
+      "learning_rate": 0.001,
+      "loss": 1.5381,
+      "step": 4673
+    },
+    {
+      "epoch": 0.19773246467552247,
+      "grad_norm": 0.31700170040130615,
+      "learning_rate": 0.001,
+      "loss": 3.6578,
+      "step": 4674
+    },
+    {
+      "epoch": 0.19777476943903882,
+      "grad_norm": 0.43023067712783813,
+      "learning_rate": 0.001,
+      "loss": 3.3044,
+      "step": 4675
+    },
+    {
+      "epoch": 0.1978170742025552,
+      "grad_norm": 1.113132357597351,
+      "learning_rate": 0.001,
+      "loss": 3.9424,
+      "step": 4676
+    },
+    {
+      "epoch": 0.1978593789660716,
+      "grad_norm": 0.2837026119232178,
+      "learning_rate": 0.001,
+      "loss": 2.3631,
+      "step": 4677
+    },
+    {
+      "epoch": 0.19790168372958794,
+      "grad_norm": 0.29237067699432373,
+      "learning_rate": 0.001,
+      "loss": 2.443,
+      "step": 4678
+    },
+    {
+      "epoch": 0.19794398849310432,
+      "grad_norm": 0.3563275635242462,
+      "learning_rate": 0.001,
+      "loss": 2.2434,
+      "step": 4679
+    },
+    {
+      "epoch": 0.1979862932566207,
+      "grad_norm": 1.2392581701278687,
+      "learning_rate": 0.001,
+      "loss": 3.6908,
+      "step": 4680
+    },
+    {
+      "epoch": 0.19802859802013706,
+      "grad_norm": 2.4038748741149902,
+      "learning_rate": 0.001,
+      "loss": 1.984,
+      "step": 4681
+    },
+    {
+      "epoch": 0.19807090278365344,
+      "grad_norm": 0.44923508167266846,
+      "learning_rate": 0.001,
+      "loss": 2.7697,
+      "step": 4682
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.3151043951511383,
+      "learning_rate": 0.001,
+      "loss": 2.9882,
+      "step": 4683
+    },
+    {
+      "epoch": 0.19815551231068618,
+      "grad_norm": 0.46017053723335266,
+      "learning_rate": 0.001,
+      "loss": 1.9041,
+      "step": 4684
+    },
+    {
+      "epoch": 0.19819781707420256,
+      "grad_norm": 0.3234018385410309,
+      "learning_rate": 0.001,
+      "loss": 1.8249,
+      "step": 4685
+    },
+    {
+      "epoch": 0.1982401218377189,
+      "grad_norm": 0.3335753083229065,
+      "learning_rate": 0.001,
+      "loss": 2.3614,
+      "step": 4686
+    },
+    {
+      "epoch": 0.1982824266012353,
+      "grad_norm": 0.382220059633255,
+      "learning_rate": 0.001,
+      "loss": 2.4,
+      "step": 4687
+    },
+    {
+      "epoch": 0.19832473136475168,
+      "grad_norm": 0.9957152605056763,
+      "learning_rate": 0.001,
+      "loss": 3.0074,
+      "step": 4688
+    },
+    {
+      "epoch": 0.19836703612826803,
+      "grad_norm": 0.2819012701511383,
+      "learning_rate": 0.001,
+      "loss": 3.5457,
+      "step": 4689
+    },
+    {
+      "epoch": 0.1984093408917844,
+      "grad_norm": 2.3886940479278564,
+      "learning_rate": 0.001,
+      "loss": 3.4807,
+      "step": 4690
+    },
+    {
+      "epoch": 0.1984516456553008,
+      "grad_norm": 0.35311177372932434,
+      "learning_rate": 0.001,
+      "loss": 3.1578,
+      "step": 4691
+    },
+    {
+      "epoch": 0.19849395041881715,
+      "grad_norm": 1.2206075191497803,
+      "learning_rate": 0.001,
+      "loss": 2.128,
+      "step": 4692
+    },
+    {
+      "epoch": 0.19853625518233353,
+      "grad_norm": 5.357852458953857,
+      "learning_rate": 0.001,
+      "loss": 2.2252,
+      "step": 4693
+    },
+    {
+      "epoch": 0.19857855994584991,
+      "grad_norm": 0.4290313422679901,
+      "learning_rate": 0.001,
+      "loss": 2.0768,
+      "step": 4694
+    },
+    {
+      "epoch": 0.19862086470936627,
+      "grad_norm": 0.22791223227977753,
+      "learning_rate": 0.001,
+      "loss": 1.7422,
+      "step": 4695
+    },
+    {
+      "epoch": 0.19866316947288265,
+      "grad_norm": 0.24464336037635803,
+      "learning_rate": 0.001,
+      "loss": 2.6629,
+      "step": 4696
+    },
+    {
+      "epoch": 0.19870547423639903,
+      "grad_norm": 1.3040649890899658,
+      "learning_rate": 0.001,
+      "loss": 2.1108,
+      "step": 4697
+    },
+    {
+      "epoch": 0.1987477789999154,
+      "grad_norm": 0.5187906622886658,
+      "learning_rate": 0.001,
+      "loss": 2.7388,
+      "step": 4698
+    },
+    {
+      "epoch": 0.19879008376343177,
+      "grad_norm": 0.2980749011039734,
+      "learning_rate": 0.001,
+      "loss": 3.4038,
+      "step": 4699
+    },
+    {
+      "epoch": 0.19883238852694812,
+      "grad_norm": 0.3226117193698883,
+      "learning_rate": 0.001,
+      "loss": 2.8376,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1988746932904645,
+      "grad_norm": 0.2342327982187271,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 4701
+    },
+    {
+      "epoch": 0.1989169980539809,
+      "grad_norm": 0.23881351947784424,
+      "learning_rate": 0.001,
+      "loss": 3.1819,
+      "step": 4702
+    },
+    {
+      "epoch": 0.19895930281749724,
+      "grad_norm": 0.1929883062839508,
+      "learning_rate": 0.001,
+      "loss": 2.0568,
+      "step": 4703
+    },
+    {
+      "epoch": 0.19900160758101362,
+      "grad_norm": 0.25098490715026855,
+      "learning_rate": 0.001,
+      "loss": 2.0162,
+      "step": 4704
+    },
+    {
+      "epoch": 0.19904391234453,
+      "grad_norm": 0.25516989827156067,
+      "learning_rate": 0.001,
+      "loss": 2.6734,
+      "step": 4705
+    },
+    {
+      "epoch": 0.19908621710804636,
+      "grad_norm": 0.2264355570077896,
+      "learning_rate": 0.001,
+      "loss": 2.4311,
+      "step": 4706
+    },
+    {
+      "epoch": 0.19912852187156274,
+      "grad_norm": 0.18645042181015015,
+      "learning_rate": 0.001,
+      "loss": 1.6494,
+      "step": 4707
+    },
+    {
+      "epoch": 0.19917082663507912,
+      "grad_norm": 9.609095573425293,
+      "learning_rate": 0.001,
+      "loss": 2.3706,
+      "step": 4708
+    },
+    {
+      "epoch": 0.19921313139859548,
+      "grad_norm": 0.2301723212003708,
+      "learning_rate": 0.001,
+      "loss": 2.3175,
+      "step": 4709
+    },
+    {
+      "epoch": 0.19925543616211186,
+      "grad_norm": 0.30773094296455383,
+      "learning_rate": 0.001,
+      "loss": 3.0642,
+      "step": 4710
+    },
+    {
+      "epoch": 0.19929774092562821,
+      "grad_norm": 0.48115605115890503,
+      "learning_rate": 0.001,
+      "loss": 2.8732,
+      "step": 4711
+    },
+    {
+      "epoch": 0.1993400456891446,
+      "grad_norm": 0.20981243252754211,
+      "learning_rate": 0.001,
+      "loss": 2.5917,
+      "step": 4712
+    },
+    {
+      "epoch": 0.19938235045266098,
+      "grad_norm": 0.22592686116695404,
+      "learning_rate": 0.001,
+      "loss": 2.8297,
+      "step": 4713
+    },
+    {
+      "epoch": 0.19942465521617733,
+      "grad_norm": 0.27815452218055725,
+      "learning_rate": 0.001,
+      "loss": 2.3364,
+      "step": 4714
+    },
+    {
+      "epoch": 0.19946695997969371,
+      "grad_norm": 0.23436151444911957,
+      "learning_rate": 0.001,
+      "loss": 2.8687,
+      "step": 4715
+    },
+    {
+      "epoch": 0.1995092647432101,
+      "grad_norm": 1.096976637840271,
+      "learning_rate": 0.001,
+      "loss": 2.2385,
+      "step": 4716
+    },
+    {
+      "epoch": 0.19955156950672645,
+      "grad_norm": 0.22871260344982147,
+      "learning_rate": 0.001,
+      "loss": 2.3228,
+      "step": 4717
+    },
+    {
+      "epoch": 0.19959387427024283,
+      "grad_norm": 0.7824380397796631,
+      "learning_rate": 0.001,
+      "loss": 2.2027,
+      "step": 4718
+    },
+    {
+      "epoch": 0.19963617903375921,
+      "grad_norm": 0.20783767104148865,
+      "learning_rate": 0.001,
+      "loss": 1.4938,
+      "step": 4719
+    },
+    {
+      "epoch": 0.19967848379727557,
+      "grad_norm": 0.25429844856262207,
+      "learning_rate": 0.001,
+      "loss": 1.8029,
+      "step": 4720
+    },
+    {
+      "epoch": 0.19972078856079195,
+      "grad_norm": 0.37199491262435913,
+      "learning_rate": 0.001,
+      "loss": 2.5307,
+      "step": 4721
+    },
+    {
+      "epoch": 0.1997630933243083,
+      "grad_norm": 0.258245587348938,
+      "learning_rate": 0.001,
+      "loss": 2.2374,
+      "step": 4722
+    },
+    {
+      "epoch": 0.1998053980878247,
+      "grad_norm": 0.2625525891780853,
+      "learning_rate": 0.001,
+      "loss": 3.0043,
+      "step": 4723
+    },
+    {
+      "epoch": 0.19984770285134107,
+      "grad_norm": 6.895899772644043,
+      "learning_rate": 0.001,
+      "loss": 2.5071,
+      "step": 4724
+    },
+    {
+      "epoch": 0.19989000761485742,
+      "grad_norm": 0.5165795087814331,
+      "learning_rate": 0.001,
+      "loss": 2.563,
+      "step": 4725
+    },
+    {
+      "epoch": 0.1999323123783738,
+      "grad_norm": 1.534584879875183,
+      "learning_rate": 0.001,
+      "loss": 2.2309,
+      "step": 4726
+    },
+    {
+      "epoch": 0.1999746171418902,
+      "grad_norm": 3.5470550060272217,
+      "learning_rate": 0.001,
+      "loss": 2.8135,
+      "step": 4727
+    },
+    {
+      "epoch": 0.20001692190540654,
+      "grad_norm": 0.46171459555625916,
+      "learning_rate": 0.001,
+      "loss": 2.4945,
+      "step": 4728
+    },
+    {
+      "epoch": 0.20005922666892292,
+      "grad_norm": 0.2912849485874176,
+      "learning_rate": 0.001,
+      "loss": 2.1541,
+      "step": 4729
+    },
+    {
+      "epoch": 0.2001015314324393,
+      "grad_norm": 0.2877223491668701,
+      "learning_rate": 0.001,
+      "loss": 2.4918,
+      "step": 4730
+    },
+    {
+      "epoch": 0.20014383619595566,
+      "grad_norm": 0.28247973322868347,
+      "learning_rate": 0.001,
+      "loss": 3.4501,
+      "step": 4731
+    },
+    {
+      "epoch": 0.20018614095947204,
+      "grad_norm": 0.24483604729175568,
+      "learning_rate": 0.001,
+      "loss": 3.568,
+      "step": 4732
+    },
+    {
+      "epoch": 0.2002284457229884,
+      "grad_norm": 0.2491256296634674,
+      "learning_rate": 0.001,
+      "loss": 2.456,
+      "step": 4733
+    },
+    {
+      "epoch": 0.20027075048650478,
+      "grad_norm": 0.21148993074893951,
+      "learning_rate": 0.001,
+      "loss": 1.9497,
+      "step": 4734
+    },
+    {
+      "epoch": 0.20031305525002116,
+      "grad_norm": 0.24428227543830872,
+      "learning_rate": 0.001,
+      "loss": 2.6689,
+      "step": 4735
+    },
+    {
+      "epoch": 0.20035536001353751,
+      "grad_norm": 0.21488066017627716,
+      "learning_rate": 0.001,
+      "loss": 2.897,
+      "step": 4736
+    },
+    {
+      "epoch": 0.2003976647770539,
+      "grad_norm": 0.2354801744222641,
+      "learning_rate": 0.001,
+      "loss": 2.4959,
+      "step": 4737
+    },
+    {
+      "epoch": 0.20043996954057028,
+      "grad_norm": 0.3543696403503418,
+      "learning_rate": 0.001,
+      "loss": 2.3337,
+      "step": 4738
+    },
+    {
+      "epoch": 0.20048227430408663,
+      "grad_norm": 0.21295201778411865,
+      "learning_rate": 0.001,
+      "loss": 2.2182,
+      "step": 4739
+    },
+    {
+      "epoch": 0.20052457906760301,
+      "grad_norm": 0.2649983763694763,
+      "learning_rate": 0.001,
+      "loss": 2.0836,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2005668838311194,
+      "grad_norm": 0.4929693341255188,
+      "learning_rate": 0.001,
+      "loss": 2.1557,
+      "step": 4741
+    },
+    {
+      "epoch": 0.20060918859463575,
+      "grad_norm": 0.24373185634613037,
+      "learning_rate": 0.001,
+      "loss": 2.2373,
+      "step": 4742
+    },
+    {
+      "epoch": 0.20065149335815213,
+      "grad_norm": 0.2782137095928192,
+      "learning_rate": 0.001,
+      "loss": 2.0663,
+      "step": 4743
+    },
+    {
+      "epoch": 0.2006937981216685,
+      "grad_norm": 0.9865705370903015,
+      "learning_rate": 0.001,
+      "loss": 2.4,
+      "step": 4744
+    },
+    {
+      "epoch": 0.20073610288518487,
+      "grad_norm": 0.24436363577842712,
+      "learning_rate": 0.001,
+      "loss": 3.0016,
+      "step": 4745
+    },
+    {
+      "epoch": 0.20077840764870125,
+      "grad_norm": 0.22285853326320648,
+      "learning_rate": 0.001,
+      "loss": 2.4025,
+      "step": 4746
+    },
+    {
+      "epoch": 0.2008207124122176,
+      "grad_norm": 0.2801361083984375,
+      "learning_rate": 0.001,
+      "loss": 3.0465,
+      "step": 4747
+    },
+    {
+      "epoch": 0.200863017175734,
+      "grad_norm": 1.4615576267242432,
+      "learning_rate": 0.001,
+      "loss": 1.813,
+      "step": 4748
+    },
+    {
+      "epoch": 0.20090532193925037,
+      "grad_norm": 0.21313884854316711,
+      "learning_rate": 0.001,
+      "loss": 1.858,
+      "step": 4749
+    },
+    {
+      "epoch": 0.20094762670276672,
+      "grad_norm": 0.7325378060340881,
+      "learning_rate": 0.001,
+      "loss": 3.6507,
+      "step": 4750
+    },
+    {
+      "epoch": 0.2009899314662831,
+      "grad_norm": 102.32513427734375,
+      "learning_rate": 0.001,
+      "loss": 3.3369,
+      "step": 4751
+    },
+    {
+      "epoch": 0.2010322362297995,
+      "grad_norm": 0.3230404555797577,
+      "learning_rate": 0.001,
+      "loss": 3.2636,
+      "step": 4752
+    },
+    {
+      "epoch": 0.20107454099331584,
+      "grad_norm": 0.27479425072669983,
+      "learning_rate": 0.001,
+      "loss": 2.8126,
+      "step": 4753
+    },
+    {
+      "epoch": 0.20111684575683222,
+      "grad_norm": 1.0016082525253296,
+      "learning_rate": 0.001,
+      "loss": 3.3368,
+      "step": 4754
+    },
+    {
+      "epoch": 0.20115915052034858,
+      "grad_norm": 2.433990478515625,
+      "learning_rate": 0.001,
+      "loss": 2.5722,
+      "step": 4755
+    },
+    {
+      "epoch": 0.20120145528386496,
+      "grad_norm": 0.29447877407073975,
+      "learning_rate": 0.001,
+      "loss": 2.011,
+      "step": 4756
+    },
+    {
+      "epoch": 0.20124376004738134,
+      "grad_norm": 0.23931360244750977,
+      "learning_rate": 0.001,
+      "loss": 3.1154,
+      "step": 4757
+    },
+    {
+      "epoch": 0.2012860648108977,
+      "grad_norm": 0.24119767546653748,
+      "learning_rate": 0.001,
+      "loss": 2.6907,
+      "step": 4758
+    },
+    {
+      "epoch": 0.20132836957441408,
+      "grad_norm": 0.2806399166584015,
+      "learning_rate": 0.001,
+      "loss": 2.435,
+      "step": 4759
+    },
+    {
+      "epoch": 0.20137067433793046,
+      "grad_norm": 0.2589947581291199,
+      "learning_rate": 0.001,
+      "loss": 3.0918,
+      "step": 4760
+    },
+    {
+      "epoch": 0.20141297910144682,
+      "grad_norm": 0.2585409879684448,
+      "learning_rate": 0.001,
+      "loss": 1.6908,
+      "step": 4761
+    },
+    {
+      "epoch": 0.2014552838649632,
+      "grad_norm": 0.4902653992176056,
+      "learning_rate": 0.001,
+      "loss": 2.5211,
+      "step": 4762
+    },
+    {
+      "epoch": 0.20149758862847958,
+      "grad_norm": 0.2097211331129074,
+      "learning_rate": 0.001,
+      "loss": 1.7957,
+      "step": 4763
+    },
+    {
+      "epoch": 0.20153989339199593,
+      "grad_norm": 0.26205360889434814,
+      "learning_rate": 0.001,
+      "loss": 1.7048,
+      "step": 4764
+    },
+    {
+      "epoch": 0.20158219815551232,
+      "grad_norm": 1.476196527481079,
+      "learning_rate": 0.001,
+      "loss": 2.1979,
+      "step": 4765
+    },
+    {
+      "epoch": 0.20162450291902867,
+      "grad_norm": 0.26606282591819763,
+      "learning_rate": 0.001,
+      "loss": 1.9978,
+      "step": 4766
+    },
+    {
+      "epoch": 0.20166680768254505,
+      "grad_norm": 0.7488316297531128,
+      "learning_rate": 0.001,
+      "loss": 2.2828,
+      "step": 4767
+    },
+    {
+      "epoch": 0.20170911244606143,
+      "grad_norm": 0.2467147260904312,
+      "learning_rate": 0.001,
+      "loss": 2.2932,
+      "step": 4768
+    },
+    {
+      "epoch": 0.2017514172095778,
+      "grad_norm": 0.3320417106151581,
+      "learning_rate": 0.001,
+      "loss": 2.5512,
+      "step": 4769
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 0.827613115310669,
+      "learning_rate": 0.001,
+      "loss": 2.9978,
+      "step": 4770
+    },
+    {
+      "epoch": 0.20183602673661055,
+      "grad_norm": 0.2728056311607361,
+      "learning_rate": 0.001,
+      "loss": 2.3317,
+      "step": 4771
+    },
+    {
+      "epoch": 0.2018783315001269,
+      "grad_norm": 12.807820320129395,
+      "learning_rate": 0.001,
+      "loss": 3.0231,
+      "step": 4772
+    },
+    {
+      "epoch": 0.2019206362636433,
+      "grad_norm": 0.25953209400177,
+      "learning_rate": 0.001,
+      "loss": 2.5744,
+      "step": 4773
+    },
+    {
+      "epoch": 0.20196294102715967,
+      "grad_norm": 0.2514730393886566,
+      "learning_rate": 0.001,
+      "loss": 1.816,
+      "step": 4774
+    },
+    {
+      "epoch": 0.20200524579067602,
+      "grad_norm": 0.3252697288990021,
+      "learning_rate": 0.001,
+      "loss": 1.7334,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2020475505541924,
+      "grad_norm": 0.24858559668064117,
+      "learning_rate": 0.001,
+      "loss": 1.9274,
+      "step": 4776
+    },
+    {
+      "epoch": 0.20208985531770876,
+      "grad_norm": 0.24600395560264587,
+      "learning_rate": 0.001,
+      "loss": 2.0192,
+      "step": 4777
+    },
+    {
+      "epoch": 0.20213216008122514,
+      "grad_norm": 0.2425483614206314,
+      "learning_rate": 0.001,
+      "loss": 1.9709,
+      "step": 4778
+    },
+    {
+      "epoch": 0.20217446484474152,
+      "grad_norm": 0.30808165669441223,
+      "learning_rate": 0.001,
+      "loss": 3.5698,
+      "step": 4779
+    },
+    {
+      "epoch": 0.20221676960825788,
+      "grad_norm": 0.23074059188365936,
+      "learning_rate": 0.001,
+      "loss": 2.4182,
+      "step": 4780
+    },
+    {
+      "epoch": 0.20225907437177426,
+      "grad_norm": 0.22748015820980072,
+      "learning_rate": 0.001,
+      "loss": 2.067,
+      "step": 4781
+    },
+    {
+      "epoch": 0.20230137913529064,
+      "grad_norm": 0.23154856264591217,
+      "learning_rate": 0.001,
+      "loss": 1.9049,
+      "step": 4782
+    },
+    {
+      "epoch": 0.202343683898807,
+      "grad_norm": 0.23665709793567657,
+      "learning_rate": 0.001,
+      "loss": 2.272,
+      "step": 4783
+    },
+    {
+      "epoch": 0.20238598866232338,
+      "grad_norm": 0.20455607771873474,
+      "learning_rate": 0.001,
+      "loss": 2.1121,
+      "step": 4784
+    },
+    {
+      "epoch": 0.20242829342583976,
+      "grad_norm": 0.4246728718280792,
+      "learning_rate": 0.001,
+      "loss": 2.6021,
+      "step": 4785
+    },
+    {
+      "epoch": 0.20247059818935612,
+      "grad_norm": 0.235957533121109,
+      "learning_rate": 0.001,
+      "loss": 3.1363,
+      "step": 4786
+    },
+    {
+      "epoch": 0.2025129029528725,
+      "grad_norm": 0.20879578590393066,
+      "learning_rate": 0.001,
+      "loss": 2.4947,
+      "step": 4787
+    },
+    {
+      "epoch": 0.20255520771638885,
+      "grad_norm": 0.22662273049354553,
+      "learning_rate": 0.001,
+      "loss": 2.2168,
+      "step": 4788
+    },
+    {
+      "epoch": 0.20259751247990523,
+      "grad_norm": 0.39572110772132874,
+      "learning_rate": 0.001,
+      "loss": 2.7348,
+      "step": 4789
+    },
+    {
+      "epoch": 0.20263981724342162,
+      "grad_norm": 0.19888943433761597,
+      "learning_rate": 0.001,
+      "loss": 2.4301,
+      "step": 4790
+    },
+    {
+      "epoch": 0.20268212200693797,
+      "grad_norm": 0.2209424376487732,
+      "learning_rate": 0.001,
+      "loss": 2.1574,
+      "step": 4791
+    },
+    {
+      "epoch": 0.20272442677045435,
+      "grad_norm": 0.25813013315200806,
+      "learning_rate": 0.001,
+      "loss": 2.2174,
+      "step": 4792
+    },
+    {
+      "epoch": 0.20276673153397073,
+      "grad_norm": 0.28615349531173706,
+      "learning_rate": 0.001,
+      "loss": 2.5578,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2028090362974871,
+      "grad_norm": 0.25967228412628174,
+      "learning_rate": 0.001,
+      "loss": 2.1363,
+      "step": 4794
+    },
+    {
+      "epoch": 0.20285134106100347,
+      "grad_norm": 0.6237789988517761,
+      "learning_rate": 0.001,
+      "loss": 2.4102,
+      "step": 4795
+    },
+    {
+      "epoch": 0.20289364582451985,
+      "grad_norm": 0.20853352546691895,
+      "learning_rate": 0.001,
+      "loss": 2.2166,
+      "step": 4796
+    },
+    {
+      "epoch": 0.2029359505880362,
+      "grad_norm": 0.24653717875480652,
+      "learning_rate": 0.001,
+      "loss": 2.9338,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2029782553515526,
+      "grad_norm": 0.4108681082725525,
+      "learning_rate": 0.001,
+      "loss": 1.4424,
+      "step": 4798
+    },
+    {
+      "epoch": 0.20302056011506894,
+      "grad_norm": 0.40089255571365356,
+      "learning_rate": 0.001,
+      "loss": 4.6068,
+      "step": 4799
+    },
+    {
+      "epoch": 0.20306286487858533,
+      "grad_norm": 0.24011704325675964,
+      "learning_rate": 0.001,
+      "loss": 2.9857,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2031051696421017,
+      "grad_norm": 0.27986645698547363,
+      "learning_rate": 0.001,
+      "loss": 2.6674,
+      "step": 4801
+    },
+    {
+      "epoch": 0.20314747440561806,
+      "grad_norm": 0.23206351697444916,
+      "learning_rate": 0.001,
+      "loss": 2.4333,
+      "step": 4802
+    },
+    {
+      "epoch": 0.20318977916913444,
+      "grad_norm": 0.5023967623710632,
+      "learning_rate": 0.001,
+      "loss": 2.2802,
+      "step": 4803
+    },
+    {
+      "epoch": 0.20323208393265083,
+      "grad_norm": 0.4342386722564697,
+      "learning_rate": 0.001,
+      "loss": 2.8626,
+      "step": 4804
+    },
+    {
+      "epoch": 0.20327438869616718,
+      "grad_norm": 0.21476680040359497,
+      "learning_rate": 0.001,
+      "loss": 1.9667,
+      "step": 4805
+    },
+    {
+      "epoch": 0.20331669345968356,
+      "grad_norm": 0.22180813550949097,
+      "learning_rate": 0.001,
+      "loss": 1.9456,
+      "step": 4806
+    },
+    {
+      "epoch": 0.20335899822319994,
+      "grad_norm": 0.21526440978050232,
+      "learning_rate": 0.001,
+      "loss": 2.1147,
+      "step": 4807
+    },
+    {
+      "epoch": 0.2034013029867163,
+      "grad_norm": 0.22872960567474365,
+      "learning_rate": 0.001,
+      "loss": 2.3991,
+      "step": 4808
+    },
+    {
+      "epoch": 0.20344360775023268,
+      "grad_norm": 0.21979276835918427,
+      "learning_rate": 0.001,
+      "loss": 1.7426,
+      "step": 4809
+    },
+    {
+      "epoch": 0.20348591251374903,
+      "grad_norm": 0.20522305369377136,
+      "learning_rate": 0.001,
+      "loss": 2.61,
+      "step": 4810
+    },
+    {
+      "epoch": 0.20352821727726542,
+      "grad_norm": 0.24340516328811646,
+      "learning_rate": 0.001,
+      "loss": 3.118,
+      "step": 4811
+    },
+    {
+      "epoch": 0.2035705220407818,
+      "grad_norm": 0.7604936361312866,
+      "learning_rate": 0.001,
+      "loss": 2.2295,
+      "step": 4812
+    },
+    {
+      "epoch": 0.20361282680429815,
+      "grad_norm": 0.20879897475242615,
+      "learning_rate": 0.001,
+      "loss": 3.0593,
+      "step": 4813
+    },
+    {
+      "epoch": 0.20365513156781453,
+      "grad_norm": 0.22496330738067627,
+      "learning_rate": 0.001,
+      "loss": 1.967,
+      "step": 4814
+    },
+    {
+      "epoch": 0.20369743633133092,
+      "grad_norm": 0.6500147581100464,
+      "learning_rate": 0.001,
+      "loss": 2.5665,
+      "step": 4815
+    },
+    {
+      "epoch": 0.20373974109484727,
+      "grad_norm": 0.2536025941371918,
+      "learning_rate": 0.001,
+      "loss": 3.4292,
+      "step": 4816
+    },
+    {
+      "epoch": 0.20378204585836365,
+      "grad_norm": 0.25820499658584595,
+      "learning_rate": 0.001,
+      "loss": 2.0012,
+      "step": 4817
+    },
+    {
+      "epoch": 0.20382435062188003,
+      "grad_norm": 0.886145830154419,
+      "learning_rate": 0.001,
+      "loss": 1.7369,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2038666553853964,
+      "grad_norm": 0.19858545064926147,
+      "learning_rate": 0.001,
+      "loss": 1.86,
+      "step": 4819
+    },
+    {
+      "epoch": 0.20390896014891277,
+      "grad_norm": 0.6188323497772217,
+      "learning_rate": 0.001,
+      "loss": 1.8931,
+      "step": 4820
+    },
+    {
+      "epoch": 0.20395126491242915,
+      "grad_norm": 0.2777750790119171,
+      "learning_rate": 0.001,
+      "loss": 3.6006,
+      "step": 4821
+    },
+    {
+      "epoch": 0.2039935696759455,
+      "grad_norm": 0.26666122674942017,
+      "learning_rate": 0.001,
+      "loss": 2.4877,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2040358744394619,
+      "grad_norm": 0.2691092789173126,
+      "learning_rate": 0.001,
+      "loss": 2.1836,
+      "step": 4823
+    },
+    {
+      "epoch": 0.20407817920297824,
+      "grad_norm": 0.24200117588043213,
+      "learning_rate": 0.001,
+      "loss": 2.5001,
+      "step": 4824
+    },
+    {
+      "epoch": 0.20412048396649463,
+      "grad_norm": 0.2400779128074646,
+      "learning_rate": 0.001,
+      "loss": 1.6834,
+      "step": 4825
+    },
+    {
+      "epoch": 0.204162788730011,
+      "grad_norm": 0.23831325769424438,
+      "learning_rate": 0.001,
+      "loss": 1.8685,
+      "step": 4826
+    },
+    {
+      "epoch": 0.20420509349352736,
+      "grad_norm": 0.2239019125699997,
+      "learning_rate": 0.001,
+      "loss": 2.894,
+      "step": 4827
+    },
+    {
+      "epoch": 0.20424739825704374,
+      "grad_norm": 0.23331467807292938,
+      "learning_rate": 0.001,
+      "loss": 2.893,
+      "step": 4828
+    },
+    {
+      "epoch": 0.20428970302056013,
+      "grad_norm": 0.2696641981601715,
+      "learning_rate": 0.001,
+      "loss": 3.3553,
+      "step": 4829
+    },
+    {
+      "epoch": 0.20433200778407648,
+      "grad_norm": 0.45643630623817444,
+      "learning_rate": 0.001,
+      "loss": 2.6336,
+      "step": 4830
+    },
+    {
+      "epoch": 0.20437431254759286,
+      "grad_norm": 0.5224469900131226,
+      "learning_rate": 0.001,
+      "loss": 2.7963,
+      "step": 4831
+    },
+    {
+      "epoch": 0.20441661731110924,
+      "grad_norm": 0.20430107414722443,
+      "learning_rate": 0.001,
+      "loss": 2.001,
+      "step": 4832
+    },
+    {
+      "epoch": 0.2044589220746256,
+      "grad_norm": 0.47029760479927063,
+      "learning_rate": 0.001,
+      "loss": 1.7341,
+      "step": 4833
+    },
+    {
+      "epoch": 0.20450122683814198,
+      "grad_norm": 0.20866377651691437,
+      "learning_rate": 0.001,
+      "loss": 2.1228,
+      "step": 4834
+    },
+    {
+      "epoch": 0.20454353160165833,
+      "grad_norm": 0.23199744522571564,
+      "learning_rate": 0.001,
+      "loss": 1.6639,
+      "step": 4835
+    },
+    {
+      "epoch": 0.20458583636517472,
+      "grad_norm": 0.23351909220218658,
+      "learning_rate": 0.001,
+      "loss": 2.9277,
+      "step": 4836
+    },
+    {
+      "epoch": 0.2046281411286911,
+      "grad_norm": 0.2393893003463745,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 4837
+    },
+    {
+      "epoch": 0.20467044589220745,
+      "grad_norm": 0.19761449098587036,
+      "learning_rate": 0.001,
+      "loss": 1.9734,
+      "step": 4838
+    },
+    {
+      "epoch": 0.20471275065572384,
+      "grad_norm": 0.45381811261177063,
+      "learning_rate": 0.001,
+      "loss": 2.3705,
+      "step": 4839
+    },
+    {
+      "epoch": 0.20475505541924022,
+      "grad_norm": 0.25361669063568115,
+      "learning_rate": 0.001,
+      "loss": 2.7643,
+      "step": 4840
+    },
+    {
+      "epoch": 0.20479736018275657,
+      "grad_norm": 0.23903019726276398,
+      "learning_rate": 0.001,
+      "loss": 2.1605,
+      "step": 4841
+    },
+    {
+      "epoch": 0.20483966494627295,
+      "grad_norm": 1.0079925060272217,
+      "learning_rate": 0.001,
+      "loss": 2.5728,
+      "step": 4842
+    },
+    {
+      "epoch": 0.20488196970978934,
+      "grad_norm": 0.22258150577545166,
+      "learning_rate": 0.001,
+      "loss": 2.7058,
+      "step": 4843
+    },
+    {
+      "epoch": 0.2049242744733057,
+      "grad_norm": 0.22821152210235596,
+      "learning_rate": 0.001,
+      "loss": 1.7562,
+      "step": 4844
+    },
+    {
+      "epoch": 0.20496657923682207,
+      "grad_norm": 0.29148948192596436,
+      "learning_rate": 0.001,
+      "loss": 2.5153,
+      "step": 4845
+    },
+    {
+      "epoch": 0.20500888400033843,
+      "grad_norm": 0.37664884328842163,
+      "learning_rate": 0.001,
+      "loss": 2.9711,
+      "step": 4846
+    },
+    {
+      "epoch": 0.2050511887638548,
+      "grad_norm": 0.2692352533340454,
+      "learning_rate": 0.001,
+      "loss": 2.3962,
+      "step": 4847
+    },
+    {
+      "epoch": 0.2050934935273712,
+      "grad_norm": 0.2488355189561844,
+      "learning_rate": 0.001,
+      "loss": 2.6436,
+      "step": 4848
+    },
+    {
+      "epoch": 0.20513579829088754,
+      "grad_norm": 33.88499069213867,
+      "learning_rate": 0.001,
+      "loss": 1.9808,
+      "step": 4849
+    },
+    {
+      "epoch": 0.20517810305440393,
+      "grad_norm": 0.24988500773906708,
+      "learning_rate": 0.001,
+      "loss": 3.2433,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2052204078179203,
+      "grad_norm": 0.2517193853855133,
+      "learning_rate": 0.001,
+      "loss": 2.9665,
+      "step": 4851
+    },
+    {
+      "epoch": 0.20526271258143666,
+      "grad_norm": 0.38323888182640076,
+      "learning_rate": 0.001,
+      "loss": 1.9108,
+      "step": 4852
+    },
+    {
+      "epoch": 0.20530501734495304,
+      "grad_norm": 0.21491578221321106,
+      "learning_rate": 0.001,
+      "loss": 1.8844,
+      "step": 4853
+    },
+    {
+      "epoch": 0.20534732210846943,
+      "grad_norm": 0.23121948540210724,
+      "learning_rate": 0.001,
+      "loss": 2.4052,
+      "step": 4854
+    },
+    {
+      "epoch": 0.20538962687198578,
+      "grad_norm": 1.7590092420578003,
+      "learning_rate": 0.001,
+      "loss": 2.9853,
+      "step": 4855
+    },
+    {
+      "epoch": 0.20543193163550216,
+      "grad_norm": 0.3250569701194763,
+      "learning_rate": 0.001,
+      "loss": 2.9398,
+      "step": 4856
+    },
+    {
+      "epoch": 0.20547423639901852,
+      "grad_norm": 0.5734913349151611,
+      "learning_rate": 0.001,
+      "loss": 1.8469,
+      "step": 4857
+    },
+    {
+      "epoch": 0.2055165411625349,
+      "grad_norm": 0.2421615719795227,
+      "learning_rate": 0.001,
+      "loss": 2.0313,
+      "step": 4858
+    },
+    {
+      "epoch": 0.20555884592605128,
+      "grad_norm": 0.2648211121559143,
+      "learning_rate": 0.001,
+      "loss": 2.3516,
+      "step": 4859
+    },
+    {
+      "epoch": 0.20560115068956764,
+      "grad_norm": 0.5653280019760132,
+      "learning_rate": 0.001,
+      "loss": 1.7878,
+      "step": 4860
+    },
+    {
+      "epoch": 0.20564345545308402,
+      "grad_norm": 0.4042348861694336,
+      "learning_rate": 0.001,
+      "loss": 2.8081,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2056857602166004,
+      "grad_norm": 0.25248414278030396,
+      "learning_rate": 0.001,
+      "loss": 3.5081,
+      "step": 4862
+    },
+    {
+      "epoch": 0.20572806498011675,
+      "grad_norm": 0.9863622784614563,
+      "learning_rate": 0.001,
+      "loss": 3.1157,
+      "step": 4863
+    },
+    {
+      "epoch": 0.20577036974363314,
+      "grad_norm": 0.2207251340150833,
+      "learning_rate": 0.001,
+      "loss": 2.1768,
+      "step": 4864
+    },
+    {
+      "epoch": 0.20581267450714952,
+      "grad_norm": 0.2811149060726166,
+      "learning_rate": 0.001,
+      "loss": 3.3907,
+      "step": 4865
+    },
+    {
+      "epoch": 0.20585497927066587,
+      "grad_norm": 0.9956902265548706,
+      "learning_rate": 0.001,
+      "loss": 2.2663,
+      "step": 4866
+    },
+    {
+      "epoch": 0.20589728403418225,
+      "grad_norm": 0.38394954800605774,
+      "learning_rate": 0.001,
+      "loss": 2.641,
+      "step": 4867
+    },
+    {
+      "epoch": 0.2059395887976986,
+      "grad_norm": 0.1891484558582306,
+      "learning_rate": 0.001,
+      "loss": 2.0367,
+      "step": 4868
+    },
+    {
+      "epoch": 0.205981893561215,
+      "grad_norm": 0.21494343876838684,
+      "learning_rate": 0.001,
+      "loss": 3.1822,
+      "step": 4869
+    },
+    {
+      "epoch": 0.20602419832473137,
+      "grad_norm": 0.27545684576034546,
+      "learning_rate": 0.001,
+      "loss": 3.4021,
+      "step": 4870
+    },
+    {
+      "epoch": 0.20606650308824773,
+      "grad_norm": 0.23246289789676666,
+      "learning_rate": 0.001,
+      "loss": 2.2404,
+      "step": 4871
+    },
+    {
+      "epoch": 0.2061088078517641,
+      "grad_norm": 0.826824963092804,
+      "learning_rate": 0.001,
+      "loss": 3.1329,
+      "step": 4872
+    },
+    {
+      "epoch": 0.2061511126152805,
+      "grad_norm": 0.38859671354293823,
+      "learning_rate": 0.001,
+      "loss": 2.1988,
+      "step": 4873
+    },
+    {
+      "epoch": 0.20619341737879684,
+      "grad_norm": 0.22136381268501282,
+      "learning_rate": 0.001,
+      "loss": 2.243,
+      "step": 4874
+    },
+    {
+      "epoch": 0.20623572214231323,
+      "grad_norm": 0.24336136877536774,
+      "learning_rate": 0.001,
+      "loss": 3.3275,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2062780269058296,
+      "grad_norm": 0.24462832510471344,
+      "learning_rate": 0.001,
+      "loss": 2.9589,
+      "step": 4876
+    },
+    {
+      "epoch": 0.20632033166934596,
+      "grad_norm": 7.4848175048828125,
+      "learning_rate": 0.001,
+      "loss": 1.9259,
+      "step": 4877
+    },
+    {
+      "epoch": 0.20636263643286235,
+      "grad_norm": 0.25238528847694397,
+      "learning_rate": 0.001,
+      "loss": 1.9616,
+      "step": 4878
+    },
+    {
+      "epoch": 0.2064049411963787,
+      "grad_norm": 1.0370349884033203,
+      "learning_rate": 0.001,
+      "loss": 2.0602,
+      "step": 4879
+    },
+    {
+      "epoch": 0.20644724595989508,
+      "grad_norm": 0.2667924165725708,
+      "learning_rate": 0.001,
+      "loss": 2.1587,
+      "step": 4880
+    },
+    {
+      "epoch": 0.20648955072341146,
+      "grad_norm": 5.384062767028809,
+      "learning_rate": 0.001,
+      "loss": 2.1591,
+      "step": 4881
+    },
+    {
+      "epoch": 0.20653185548692782,
+      "grad_norm": 0.653923511505127,
+      "learning_rate": 0.001,
+      "loss": 2.1794,
+      "step": 4882
+    },
+    {
+      "epoch": 0.2065741602504442,
+      "grad_norm": 6.2755937576293945,
+      "learning_rate": 0.001,
+      "loss": 2.7158,
+      "step": 4883
+    },
+    {
+      "epoch": 0.20661646501396058,
+      "grad_norm": 0.5690131187438965,
+      "learning_rate": 0.001,
+      "loss": 2.2172,
+      "step": 4884
+    },
+    {
+      "epoch": 0.20665876977747694,
+      "grad_norm": 0.6004669070243835,
+      "learning_rate": 0.001,
+      "loss": 2.4962,
+      "step": 4885
+    },
+    {
+      "epoch": 0.20670107454099332,
+      "grad_norm": 0.5099846124649048,
+      "learning_rate": 0.001,
+      "loss": 2.7323,
+      "step": 4886
+    },
+    {
+      "epoch": 0.2067433793045097,
+      "grad_norm": 0.8244269490242004,
+      "learning_rate": 0.001,
+      "loss": 2.5037,
+      "step": 4887
+    },
+    {
+      "epoch": 0.20678568406802605,
+      "grad_norm": 0.7311514616012573,
+      "learning_rate": 0.001,
+      "loss": 2.8543,
+      "step": 4888
+    },
+    {
+      "epoch": 0.20682798883154244,
+      "grad_norm": 1.6396245956420898,
+      "learning_rate": 0.001,
+      "loss": 3.5585,
+      "step": 4889
+    },
+    {
+      "epoch": 0.2068702935950588,
+      "grad_norm": 0.7014549374580383,
+      "learning_rate": 0.001,
+      "loss": 2.8487,
+      "step": 4890
+    },
+    {
+      "epoch": 0.20691259835857517,
+      "grad_norm": 0.7353118062019348,
+      "learning_rate": 0.001,
+      "loss": 2.7439,
+      "step": 4891
+    },
+    {
+      "epoch": 0.20695490312209155,
+      "grad_norm": 0.3765983283519745,
+      "learning_rate": 0.001,
+      "loss": 4.2012,
+      "step": 4892
+    },
+    {
+      "epoch": 0.2069972078856079,
+      "grad_norm": 1.587646245956421,
+      "learning_rate": 0.001,
+      "loss": 2.1776,
+      "step": 4893
+    },
+    {
+      "epoch": 0.2070395126491243,
+      "grad_norm": 0.45445260405540466,
+      "learning_rate": 0.001,
+      "loss": 2.7538,
+      "step": 4894
+    },
+    {
+      "epoch": 0.20708181741264067,
+      "grad_norm": 0.5824608206748962,
+      "learning_rate": 0.001,
+      "loss": 2.4091,
+      "step": 4895
+    },
+    {
+      "epoch": 0.20712412217615703,
+      "grad_norm": 0.29046350717544556,
+      "learning_rate": 0.001,
+      "loss": 2.3128,
+      "step": 4896
+    },
+    {
+      "epoch": 0.2071664269396734,
+      "grad_norm": 0.2888144552707672,
+      "learning_rate": 0.001,
+      "loss": 2.5121,
+      "step": 4897
+    },
+    {
+      "epoch": 0.2072087317031898,
+      "grad_norm": 0.4128365218639374,
+      "learning_rate": 0.001,
+      "loss": 2.8964,
+      "step": 4898
+    },
+    {
+      "epoch": 0.20725103646670615,
+      "grad_norm": 2.1219639778137207,
+      "learning_rate": 0.001,
+      "loss": 2.9168,
+      "step": 4899
+    },
+    {
+      "epoch": 0.20729334123022253,
+      "grad_norm": 0.910344660282135,
+      "learning_rate": 0.001,
+      "loss": 4.5335,
+      "step": 4900
+    },
+    {
+      "epoch": 0.20733564599373888,
+      "grad_norm": 0.2882691025733948,
+      "learning_rate": 0.001,
+      "loss": 2.8958,
+      "step": 4901
+    },
+    {
+      "epoch": 0.20737795075725526,
+      "grad_norm": 0.2687327563762665,
+      "learning_rate": 0.001,
+      "loss": 2.1742,
+      "step": 4902
+    },
+    {
+      "epoch": 0.20742025552077165,
+      "grad_norm": 0.8853581547737122,
+      "learning_rate": 0.001,
+      "loss": 1.9429,
+      "step": 4903
+    },
+    {
+      "epoch": 0.207462560284288,
+      "grad_norm": 1.759624719619751,
+      "learning_rate": 0.001,
+      "loss": 2.8563,
+      "step": 4904
+    },
+    {
+      "epoch": 0.20750486504780438,
+      "grad_norm": 3.137960910797119,
+      "learning_rate": 0.001,
+      "loss": 3.4388,
+      "step": 4905
+    },
+    {
+      "epoch": 0.20754716981132076,
+      "grad_norm": 3.9103238582611084,
+      "learning_rate": 0.001,
+      "loss": 2.4237,
+      "step": 4906
+    },
+    {
+      "epoch": 0.20758947457483712,
+      "grad_norm": 0.40916767716407776,
+      "learning_rate": 0.001,
+      "loss": 2.2436,
+      "step": 4907
+    },
+    {
+      "epoch": 0.2076317793383535,
+      "grad_norm": 0.30772218108177185,
+      "learning_rate": 0.001,
+      "loss": 2.3738,
+      "step": 4908
+    },
+    {
+      "epoch": 0.20767408410186988,
+      "grad_norm": 1.7794135808944702,
+      "learning_rate": 0.001,
+      "loss": 3.5238,
+      "step": 4909
+    },
+    {
+      "epoch": 0.20771638886538624,
+      "grad_norm": 0.23665037751197815,
+      "learning_rate": 0.001,
+      "loss": 1.9228,
+      "step": 4910
+    },
+    {
+      "epoch": 0.20775869362890262,
+      "grad_norm": 0.2872195541858673,
+      "learning_rate": 0.001,
+      "loss": 2.6802,
+      "step": 4911
+    },
+    {
+      "epoch": 0.20780099839241897,
+      "grad_norm": 6.244613170623779,
+      "learning_rate": 0.001,
+      "loss": 2.8313,
+      "step": 4912
+    },
+    {
+      "epoch": 0.20784330315593535,
+      "grad_norm": 0.2514772117137909,
+      "learning_rate": 0.001,
+      "loss": 2.0803,
+      "step": 4913
+    },
+    {
+      "epoch": 0.20788560791945174,
+      "grad_norm": 11.249589920043945,
+      "learning_rate": 0.001,
+      "loss": 2.0849,
+      "step": 4914
+    },
+    {
+      "epoch": 0.2079279126829681,
+      "grad_norm": 0.3307616114616394,
+      "learning_rate": 0.001,
+      "loss": 2.9033,
+      "step": 4915
+    },
+    {
+      "epoch": 0.20797021744648447,
+      "grad_norm": 0.3316502869129181,
+      "learning_rate": 0.001,
+      "loss": 2.3176,
+      "step": 4916
+    },
+    {
+      "epoch": 0.20801252221000086,
+      "grad_norm": 0.4694264829158783,
+      "learning_rate": 0.001,
+      "loss": 3.0293,
+      "step": 4917
+    },
+    {
+      "epoch": 0.2080548269735172,
+      "grad_norm": 0.2822750210762024,
+      "learning_rate": 0.001,
+      "loss": 2.4169,
+      "step": 4918
+    },
+    {
+      "epoch": 0.2080971317370336,
+      "grad_norm": 0.604796826839447,
+      "learning_rate": 0.001,
+      "loss": 2.7267,
+      "step": 4919
+    },
+    {
+      "epoch": 0.20813943650054997,
+      "grad_norm": 0.23968219757080078,
+      "learning_rate": 0.001,
+      "loss": 2.7295,
+      "step": 4920
+    },
+    {
+      "epoch": 0.20818174126406633,
+      "grad_norm": 0.2379380464553833,
+      "learning_rate": 0.001,
+      "loss": 2.2191,
+      "step": 4921
+    },
+    {
+      "epoch": 0.2082240460275827,
+      "grad_norm": 0.2360192835330963,
+      "learning_rate": 0.001,
+      "loss": 1.8814,
+      "step": 4922
+    },
+    {
+      "epoch": 0.20826635079109906,
+      "grad_norm": 0.40133988857269287,
+      "learning_rate": 0.001,
+      "loss": 1.9194,
+      "step": 4923
+    },
+    {
+      "epoch": 0.20830865555461545,
+      "grad_norm": 0.314656525850296,
+      "learning_rate": 0.001,
+      "loss": 2.3935,
+      "step": 4924
+    },
+    {
+      "epoch": 0.20835096031813183,
+      "grad_norm": 1.1727277040481567,
+      "learning_rate": 0.001,
+      "loss": 2.8302,
+      "step": 4925
+    },
+    {
+      "epoch": 0.20839326508164818,
+      "grad_norm": 0.22926130890846252,
+      "learning_rate": 0.001,
+      "loss": 2.7978,
+      "step": 4926
+    },
+    {
+      "epoch": 0.20843556984516456,
+      "grad_norm": 0.4540160596370697,
+      "learning_rate": 0.001,
+      "loss": 1.8122,
+      "step": 4927
+    },
+    {
+      "epoch": 0.20847787460868095,
+      "grad_norm": 23.994537353515625,
+      "learning_rate": 0.001,
+      "loss": 2.6313,
+      "step": 4928
+    },
+    {
+      "epoch": 0.2085201793721973,
+      "grad_norm": 0.6194539070129395,
+      "learning_rate": 0.001,
+      "loss": 2.9595,
+      "step": 4929
+    },
+    {
+      "epoch": 0.20856248413571368,
+      "grad_norm": 13.033817291259766,
+      "learning_rate": 0.001,
+      "loss": 2.1685,
+      "step": 4930
+    },
+    {
+      "epoch": 0.20860478889923006,
+      "grad_norm": 0.323756605386734,
+      "learning_rate": 0.001,
+      "loss": 2.9983,
+      "step": 4931
+    },
+    {
+      "epoch": 0.20864709366274642,
+      "grad_norm": 4.008752346038818,
+      "learning_rate": 0.001,
+      "loss": 3.0625,
+      "step": 4932
+    },
+    {
+      "epoch": 0.2086893984262628,
+      "grad_norm": 0.28806641697883606,
+      "learning_rate": 0.001,
+      "loss": 2.3397,
+      "step": 4933
+    },
+    {
+      "epoch": 0.20873170318977916,
+      "grad_norm": 0.49707770347595215,
+      "learning_rate": 0.001,
+      "loss": 2.1648,
+      "step": 4934
+    },
+    {
+      "epoch": 0.20877400795329554,
+      "grad_norm": 0.29439014196395874,
+      "learning_rate": 0.001,
+      "loss": 2.0093,
+      "step": 4935
+    },
+    {
+      "epoch": 0.20881631271681192,
+      "grad_norm": 0.7795244455337524,
+      "learning_rate": 0.001,
+      "loss": 2.36,
+      "step": 4936
+    },
+    {
+      "epoch": 0.20885861748032827,
+      "grad_norm": 0.2714845836162567,
+      "learning_rate": 0.001,
+      "loss": 1.7267,
+      "step": 4937
+    },
+    {
+      "epoch": 0.20890092224384466,
+      "grad_norm": 0.4351987838745117,
+      "learning_rate": 0.001,
+      "loss": 2.605,
+      "step": 4938
+    },
+    {
+      "epoch": 0.20894322700736104,
+      "grad_norm": 0.39439719915390015,
+      "learning_rate": 0.001,
+      "loss": 2.6728,
+      "step": 4939
+    },
+    {
+      "epoch": 0.2089855317708774,
+      "grad_norm": 0.8820605874061584,
+      "learning_rate": 0.001,
+      "loss": 3.4102,
+      "step": 4940
+    },
+    {
+      "epoch": 0.20902783653439377,
+      "grad_norm": 0.3229409158229828,
+      "learning_rate": 0.001,
+      "loss": 2.9257,
+      "step": 4941
+    },
+    {
+      "epoch": 0.20907014129791016,
+      "grad_norm": 0.3107205629348755,
+      "learning_rate": 0.001,
+      "loss": 2.3553,
+      "step": 4942
+    },
+    {
+      "epoch": 0.2091124460614265,
+      "grad_norm": 0.36764320731163025,
+      "learning_rate": 0.001,
+      "loss": 3.036,
+      "step": 4943
+    },
+    {
+      "epoch": 0.2091547508249429,
+      "grad_norm": 0.2639576494693756,
+      "learning_rate": 0.001,
+      "loss": 2.8147,
+      "step": 4944
+    },
+    {
+      "epoch": 0.20919705558845927,
+      "grad_norm": 0.24890625476837158,
+      "learning_rate": 0.001,
+      "loss": 1.888,
+      "step": 4945
+    },
+    {
+      "epoch": 0.20923936035197563,
+      "grad_norm": 0.28734108805656433,
+      "learning_rate": 0.001,
+      "loss": 2.3685,
+      "step": 4946
+    },
+    {
+      "epoch": 0.209281665115492,
+      "grad_norm": 1.45154869556427,
+      "learning_rate": 0.001,
+      "loss": 2.8914,
+      "step": 4947
+    },
+    {
+      "epoch": 0.20932396987900836,
+      "grad_norm": 0.24972866475582123,
+      "learning_rate": 0.001,
+      "loss": 2.1973,
+      "step": 4948
+    },
+    {
+      "epoch": 0.20936627464252475,
+      "grad_norm": 0.7616045475006104,
+      "learning_rate": 0.001,
+      "loss": 2.4037,
+      "step": 4949
+    },
+    {
+      "epoch": 0.20940857940604113,
+      "grad_norm": 0.3818029463291168,
+      "learning_rate": 0.001,
+      "loss": 2.3215,
+      "step": 4950
+    },
+    {
+      "epoch": 0.20945088416955748,
+      "grad_norm": 32.00923156738281,
+      "learning_rate": 0.001,
+      "loss": 2.4591,
+      "step": 4951
+    },
+    {
+      "epoch": 0.20949318893307387,
+      "grad_norm": 0.31729358434677124,
+      "learning_rate": 0.001,
+      "loss": 2.4413,
+      "step": 4952
+    },
+    {
+      "epoch": 0.20953549369659025,
+      "grad_norm": 2.464691638946533,
+      "learning_rate": 0.001,
+      "loss": 2.4398,
+      "step": 4953
+    },
+    {
+      "epoch": 0.2095777984601066,
+      "grad_norm": 35.59858322143555,
+      "learning_rate": 0.001,
+      "loss": 2.887,
+      "step": 4954
+    },
+    {
+      "epoch": 0.20962010322362298,
+      "grad_norm": 0.5536134243011475,
+      "learning_rate": 0.001,
+      "loss": 2.9387,
+      "step": 4955
+    },
+    {
+      "epoch": 0.20966240798713937,
+      "grad_norm": 1.5458406209945679,
+      "learning_rate": 0.001,
+      "loss": 2.1783,
+      "step": 4956
+    },
+    {
+      "epoch": 0.20970471275065572,
+      "grad_norm": 0.8525711894035339,
+      "learning_rate": 0.001,
+      "loss": 3.0125,
+      "step": 4957
+    },
+    {
+      "epoch": 0.2097470175141721,
+      "grad_norm": 3.157151460647583,
+      "learning_rate": 0.001,
+      "loss": 2.9392,
+      "step": 4958
+    },
+    {
+      "epoch": 0.20978932227768846,
+      "grad_norm": 0.42130786180496216,
+      "learning_rate": 0.001,
+      "loss": 2.071,
+      "step": 4959
+    },
+    {
+      "epoch": 0.20983162704120484,
+      "grad_norm": 2.541234016418457,
+      "learning_rate": 0.001,
+      "loss": 4.1087,
+      "step": 4960
+    },
+    {
+      "epoch": 0.20987393180472122,
+      "grad_norm": 0.24158483743667603,
+      "learning_rate": 0.001,
+      "loss": 1.9446,
+      "step": 4961
+    },
+    {
+      "epoch": 0.20991623656823757,
+      "grad_norm": 0.45275890827178955,
+      "learning_rate": 0.001,
+      "loss": 2.1538,
+      "step": 4962
+    },
+    {
+      "epoch": 0.20995854133175396,
+      "grad_norm": 1.040623664855957,
+      "learning_rate": 0.001,
+      "loss": 2.0899,
+      "step": 4963
+    },
+    {
+      "epoch": 0.21000084609527034,
+      "grad_norm": 0.2312449812889099,
+      "learning_rate": 0.001,
+      "loss": 1.9906,
+      "step": 4964
+    },
+    {
+      "epoch": 0.2100431508587867,
+      "grad_norm": 0.33807173371315,
+      "learning_rate": 0.001,
+      "loss": 2.1848,
+      "step": 4965
+    },
+    {
+      "epoch": 0.21008545562230307,
+      "grad_norm": 0.256266325712204,
+      "learning_rate": 0.001,
+      "loss": 2.1486,
+      "step": 4966
+    },
+    {
+      "epoch": 0.21012776038581946,
+      "grad_norm": 0.4030238687992096,
+      "learning_rate": 0.001,
+      "loss": 1.9343,
+      "step": 4967
+    },
+    {
+      "epoch": 0.2101700651493358,
+      "grad_norm": 0.33248621225357056,
+      "learning_rate": 0.001,
+      "loss": 2.2524,
+      "step": 4968
+    },
+    {
+      "epoch": 0.2102123699128522,
+      "grad_norm": 0.3359518349170685,
+      "learning_rate": 0.001,
+      "loss": 2.6882,
+      "step": 4969
+    },
+    {
+      "epoch": 0.21025467467636855,
+      "grad_norm": 0.3091052174568176,
+      "learning_rate": 0.001,
+      "loss": 3.1879,
+      "step": 4970
+    },
+    {
+      "epoch": 0.21029697943988493,
+      "grad_norm": 0.21481722593307495,
+      "learning_rate": 0.001,
+      "loss": 1.8638,
+      "step": 4971
+    },
+    {
+      "epoch": 0.2103392842034013,
+      "grad_norm": 0.21660932898521423,
+      "learning_rate": 0.001,
+      "loss": 2.0017,
+      "step": 4972
+    },
+    {
+      "epoch": 0.21038158896691767,
+      "grad_norm": 0.4549258351325989,
+      "learning_rate": 0.001,
+      "loss": 3.1636,
+      "step": 4973
+    },
+    {
+      "epoch": 0.21042389373043405,
+      "grad_norm": 1.2132890224456787,
+      "learning_rate": 0.001,
+      "loss": 3.037,
+      "step": 4974
+    },
+    {
+      "epoch": 0.21046619849395043,
+      "grad_norm": 0.24998739361763,
+      "learning_rate": 0.001,
+      "loss": 2.3893,
+      "step": 4975
+    },
+    {
+      "epoch": 0.21050850325746678,
+      "grad_norm": 0.2769694924354553,
+      "learning_rate": 0.001,
+      "loss": 2.0534,
+      "step": 4976
+    },
+    {
+      "epoch": 0.21055080802098317,
+      "grad_norm": 8.817028999328613,
+      "learning_rate": 0.001,
+      "loss": 2.5478,
+      "step": 4977
+    },
+    {
+      "epoch": 0.21059311278449955,
+      "grad_norm": 0.840341329574585,
+      "learning_rate": 0.001,
+      "loss": 2.5962,
+      "step": 4978
+    },
+    {
+      "epoch": 0.2106354175480159,
+      "grad_norm": 0.33536335825920105,
+      "learning_rate": 0.001,
+      "loss": 1.973,
+      "step": 4979
+    },
+    {
+      "epoch": 0.21067772231153228,
+      "grad_norm": 0.34550049901008606,
+      "learning_rate": 0.001,
+      "loss": 2.3634,
+      "step": 4980
+    },
+    {
+      "epoch": 0.21072002707504864,
+      "grad_norm": 0.4002279043197632,
+      "learning_rate": 0.001,
+      "loss": 2.7841,
+      "step": 4981
+    },
+    {
+      "epoch": 0.21076233183856502,
+      "grad_norm": 0.3583948314189911,
+      "learning_rate": 0.001,
+      "loss": 3.248,
+      "step": 4982
+    },
+    {
+      "epoch": 0.2108046366020814,
+      "grad_norm": 0.2348349541425705,
+      "learning_rate": 0.001,
+      "loss": 2.4177,
+      "step": 4983
+    },
+    {
+      "epoch": 0.21084694136559776,
+      "grad_norm": 0.2678260803222656,
+      "learning_rate": 0.001,
+      "loss": 2.775,
+      "step": 4984
+    },
+    {
+      "epoch": 0.21088924612911414,
+      "grad_norm": 0.24458220601081848,
+      "learning_rate": 0.001,
+      "loss": 2.3764,
+      "step": 4985
+    },
+    {
+      "epoch": 0.21093155089263052,
+      "grad_norm": 0.38108593225479126,
+      "learning_rate": 0.001,
+      "loss": 2.247,
+      "step": 4986
+    },
+    {
+      "epoch": 0.21097385565614687,
+      "grad_norm": 0.21886327862739563,
+      "learning_rate": 0.001,
+      "loss": 2.7944,
+      "step": 4987
+    },
+    {
+      "epoch": 0.21101616041966326,
+      "grad_norm": 7.095550537109375,
+      "learning_rate": 0.001,
+      "loss": 1.5504,
+      "step": 4988
+    },
+    {
+      "epoch": 0.21105846518317964,
+      "grad_norm": 0.2182372361421585,
+      "learning_rate": 0.001,
+      "loss": 2.7729,
+      "step": 4989
+    },
+    {
+      "epoch": 0.211100769946696,
+      "grad_norm": 0.24648889899253845,
+      "learning_rate": 0.001,
+      "loss": 2.2145,
+      "step": 4990
+    },
+    {
+      "epoch": 0.21114307471021238,
+      "grad_norm": 0.22735580801963806,
+      "learning_rate": 0.001,
+      "loss": 1.9403,
+      "step": 4991
+    },
+    {
+      "epoch": 0.21118537947372873,
+      "grad_norm": 2.2606966495513916,
+      "learning_rate": 0.001,
+      "loss": 2.5648,
+      "step": 4992
+    },
+    {
+      "epoch": 0.2112276842372451,
+      "grad_norm": 1.0740227699279785,
+      "learning_rate": 0.001,
+      "loss": 3.2116,
+      "step": 4993
+    },
+    {
+      "epoch": 0.2112699890007615,
+      "grad_norm": 1.1305190324783325,
+      "learning_rate": 0.001,
+      "loss": 3.5133,
+      "step": 4994
+    },
+    {
+      "epoch": 0.21131229376427785,
+      "grad_norm": 0.8290959000587463,
+      "learning_rate": 0.001,
+      "loss": 2.4756,
+      "step": 4995
+    },
+    {
+      "epoch": 0.21135459852779423,
+      "grad_norm": 0.30517685413360596,
+      "learning_rate": 0.001,
+      "loss": 2.2226,
+      "step": 4996
+    },
+    {
+      "epoch": 0.2113969032913106,
+      "grad_norm": 0.25863179564476013,
+      "learning_rate": 0.001,
+      "loss": 2.1738,
+      "step": 4997
+    },
+    {
+      "epoch": 0.21143920805482697,
+      "grad_norm": 0.33867284655570984,
+      "learning_rate": 0.001,
+      "loss": 2.6182,
+      "step": 4998
+    },
+    {
+      "epoch": 0.21148151281834335,
+      "grad_norm": 0.3169691264629364,
+      "learning_rate": 0.001,
+      "loss": 2.8752,
+      "step": 4999
+    },
+    {
+      "epoch": 0.21152381758185973,
+      "grad_norm": 0.2934655249118805,
+      "learning_rate": 0.001,
+      "loss": 3.4324,
+      "step": 5000
+    },
+    {
+      "epoch": 0.21156612234537608,
+      "grad_norm": 1.9539580345153809,
+      "learning_rate": 0.001,
+      "loss": 2.0974,
+      "step": 5001
+    },
+    {
+      "epoch": 0.21160842710889247,
+      "grad_norm": 0.28352606296539307,
+      "learning_rate": 0.001,
+      "loss": 2.3892,
+      "step": 5002
+    },
+    {
+      "epoch": 0.21165073187240882,
+      "grad_norm": 0.36468642950057983,
+      "learning_rate": 0.001,
+      "loss": 2.67,
+      "step": 5003
+    },
+    {
+      "epoch": 0.2116930366359252,
+      "grad_norm": 0.28954243659973145,
+      "learning_rate": 0.001,
+      "loss": 2.4779,
+      "step": 5004
+    },
+    {
+      "epoch": 0.21173534139944158,
+      "grad_norm": 0.2951817810535431,
+      "learning_rate": 0.001,
+      "loss": 2.6193,
+      "step": 5005
+    },
+    {
+      "epoch": 0.21177764616295794,
+      "grad_norm": 0.24036556482315063,
+      "learning_rate": 0.001,
+      "loss": 2.362,
+      "step": 5006
+    },
+    {
+      "epoch": 0.21181995092647432,
+      "grad_norm": 0.23176927864551544,
+      "learning_rate": 0.001,
+      "loss": 2.4304,
+      "step": 5007
+    },
+    {
+      "epoch": 0.2118622556899907,
+      "grad_norm": 0.2653713822364807,
+      "learning_rate": 0.001,
+      "loss": 2.4899,
+      "step": 5008
+    },
+    {
+      "epoch": 0.21190456045350706,
+      "grad_norm": 0.22608830034732819,
+      "learning_rate": 0.001,
+      "loss": 1.915,
+      "step": 5009
+    },
+    {
+      "epoch": 0.21194686521702344,
+      "grad_norm": 0.26712745428085327,
+      "learning_rate": 0.001,
+      "loss": 2.8369,
+      "step": 5010
+    },
+    {
+      "epoch": 0.21198916998053982,
+      "grad_norm": 0.23458565771579742,
+      "learning_rate": 0.001,
+      "loss": 2.713,
+      "step": 5011
+    },
+    {
+      "epoch": 0.21203147474405618,
+      "grad_norm": 0.3184855282306671,
+      "learning_rate": 0.001,
+      "loss": 3.2585,
+      "step": 5012
+    },
+    {
+      "epoch": 0.21207377950757256,
+      "grad_norm": 0.29982033371925354,
+      "learning_rate": 0.001,
+      "loss": 2.6947,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2121160842710889,
+      "grad_norm": 0.22614946961402893,
+      "learning_rate": 0.001,
+      "loss": 1.8255,
+      "step": 5014
+    },
+    {
+      "epoch": 0.2121583890346053,
+      "grad_norm": 0.22984737157821655,
+      "learning_rate": 0.001,
+      "loss": 1.9238,
+      "step": 5015
+    },
+    {
+      "epoch": 0.21220069379812168,
+      "grad_norm": 0.2999686598777771,
+      "learning_rate": 0.001,
+      "loss": 2.102,
+      "step": 5016
+    },
+    {
+      "epoch": 0.21224299856163803,
+      "grad_norm": 0.21866334974765778,
+      "learning_rate": 0.001,
+      "loss": 2.5355,
+      "step": 5017
+    },
+    {
+      "epoch": 0.2122853033251544,
+      "grad_norm": 0.30054131150245667,
+      "learning_rate": 0.001,
+      "loss": 1.7776,
+      "step": 5018
+    },
+    {
+      "epoch": 0.2123276080886708,
+      "grad_norm": 0.28100132942199707,
+      "learning_rate": 0.001,
+      "loss": 3.4669,
+      "step": 5019
+    },
+    {
+      "epoch": 0.21236991285218715,
+      "grad_norm": 0.2859184443950653,
+      "learning_rate": 0.001,
+      "loss": 1.9491,
+      "step": 5020
+    },
+    {
+      "epoch": 0.21241221761570353,
+      "grad_norm": 1.7632389068603516,
+      "learning_rate": 0.001,
+      "loss": 2.4757,
+      "step": 5021
+    },
+    {
+      "epoch": 0.2124545223792199,
+      "grad_norm": 0.4322301149368286,
+      "learning_rate": 0.001,
+      "loss": 1.9597,
+      "step": 5022
+    },
+    {
+      "epoch": 0.21249682714273627,
+      "grad_norm": 0.27014920115470886,
+      "learning_rate": 0.001,
+      "loss": 2.5075,
+      "step": 5023
+    },
+    {
+      "epoch": 0.21253913190625265,
+      "grad_norm": 0.21609550714492798,
+      "learning_rate": 0.001,
+      "loss": 2.2271,
+      "step": 5024
+    },
+    {
+      "epoch": 0.212581436669769,
+      "grad_norm": 0.2416316717863083,
+      "learning_rate": 0.001,
+      "loss": 2.0308,
+      "step": 5025
+    },
+    {
+      "epoch": 0.21262374143328538,
+      "grad_norm": 0.6305524110794067,
+      "learning_rate": 0.001,
+      "loss": 2.16,
+      "step": 5026
+    },
+    {
+      "epoch": 0.21266604619680177,
+      "grad_norm": 0.21432656049728394,
+      "learning_rate": 0.001,
+      "loss": 2.6648,
+      "step": 5027
+    },
+    {
+      "epoch": 0.21270835096031812,
+      "grad_norm": 0.854989767074585,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 5028
+    },
+    {
+      "epoch": 0.2127506557238345,
+      "grad_norm": 0.2650429308414459,
+      "learning_rate": 0.001,
+      "loss": 2.1294,
+      "step": 5029
+    },
+    {
+      "epoch": 0.21279296048735089,
+      "grad_norm": 0.1896660178899765,
+      "learning_rate": 0.001,
+      "loss": 1.7712,
+      "step": 5030
+    },
+    {
+      "epoch": 0.21283526525086724,
+      "grad_norm": 0.4741089940071106,
+      "learning_rate": 0.001,
+      "loss": 2.5271,
+      "step": 5031
+    },
+    {
+      "epoch": 0.21287757001438362,
+      "grad_norm": 0.22527028620243073,
+      "learning_rate": 0.001,
+      "loss": 2.7449,
+      "step": 5032
+    },
+    {
+      "epoch": 0.2129198747779,
+      "grad_norm": 0.2307746261358261,
+      "learning_rate": 0.001,
+      "loss": 2.8616,
+      "step": 5033
+    },
+    {
+      "epoch": 0.21296217954141636,
+      "grad_norm": 0.4615713655948639,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 5034
+    },
+    {
+      "epoch": 0.21300448430493274,
+      "grad_norm": 23.890384674072266,
+      "learning_rate": 0.001,
+      "loss": 1.4897,
+      "step": 5035
+    },
+    {
+      "epoch": 0.2130467890684491,
+      "grad_norm": 0.497530072927475,
+      "learning_rate": 0.001,
+      "loss": 2.1057,
+      "step": 5036
+    },
+    {
+      "epoch": 0.21308909383196548,
+      "grad_norm": 4.099823474884033,
+      "learning_rate": 0.001,
+      "loss": 2.2474,
+      "step": 5037
+    },
+    {
+      "epoch": 0.21313139859548186,
+      "grad_norm": 0.23300927877426147,
+      "learning_rate": 0.001,
+      "loss": 2.6447,
+      "step": 5038
+    },
+    {
+      "epoch": 0.2131737033589982,
+      "grad_norm": 1.2628371715545654,
+      "learning_rate": 0.001,
+      "loss": 2.3729,
+      "step": 5039
+    },
+    {
+      "epoch": 0.2132160081225146,
+      "grad_norm": 3.1556689739227295,
+      "learning_rate": 0.001,
+      "loss": 2.0734,
+      "step": 5040
+    },
+    {
+      "epoch": 0.21325831288603098,
+      "grad_norm": 0.23001788556575775,
+      "learning_rate": 0.001,
+      "loss": 2.0797,
+      "step": 5041
+    },
+    {
+      "epoch": 0.21330061764954733,
+      "grad_norm": 0.3631579577922821,
+      "learning_rate": 0.001,
+      "loss": 2.1848,
+      "step": 5042
+    },
+    {
+      "epoch": 0.2133429224130637,
+      "grad_norm": 0.2023169994354248,
+      "learning_rate": 0.001,
+      "loss": 2.0585,
+      "step": 5043
+    },
+    {
+      "epoch": 0.2133852271765801,
+      "grad_norm": 0.25876331329345703,
+      "learning_rate": 0.001,
+      "loss": 2.2102,
+      "step": 5044
+    },
+    {
+      "epoch": 0.21342753194009645,
+      "grad_norm": 0.22901210188865662,
+      "learning_rate": 0.001,
+      "loss": 2.1429,
+      "step": 5045
+    },
+    {
+      "epoch": 0.21346983670361283,
+      "grad_norm": 0.472971647977829,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 5046
+    },
+    {
+      "epoch": 0.21351214146712919,
+      "grad_norm": 0.47565358877182007,
+      "learning_rate": 0.001,
+      "loss": 2.3727,
+      "step": 5047
+    },
+    {
+      "epoch": 0.21355444623064557,
+      "grad_norm": 3.025669813156128,
+      "learning_rate": 0.001,
+      "loss": 1.6324,
+      "step": 5048
+    },
+    {
+      "epoch": 0.21359675099416195,
+      "grad_norm": 0.7364016175270081,
+      "learning_rate": 0.001,
+      "loss": 2.7439,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2136390557576783,
+      "grad_norm": 0.25837811827659607,
+      "learning_rate": 0.001,
+      "loss": 2.591,
+      "step": 5050
+    },
+    {
+      "epoch": 0.21368136052119469,
+      "grad_norm": 0.556201159954071,
+      "learning_rate": 0.001,
+      "loss": 2.0559,
+      "step": 5051
+    },
+    {
+      "epoch": 0.21372366528471107,
+      "grad_norm": 0.23484918475151062,
+      "learning_rate": 0.001,
+      "loss": 2.0903,
+      "step": 5052
+    },
+    {
+      "epoch": 0.21376597004822742,
+      "grad_norm": 0.20582491159439087,
+      "learning_rate": 0.001,
+      "loss": 2.5129,
+      "step": 5053
+    },
+    {
+      "epoch": 0.2138082748117438,
+      "grad_norm": 0.4761744439601898,
+      "learning_rate": 0.001,
+      "loss": 2.3142,
+      "step": 5054
+    },
+    {
+      "epoch": 0.21385057957526019,
+      "grad_norm": 0.5703874230384827,
+      "learning_rate": 0.001,
+      "loss": 2.5662,
+      "step": 5055
+    },
+    {
+      "epoch": 0.21389288433877654,
+      "grad_norm": 0.19592377543449402,
+      "learning_rate": 0.001,
+      "loss": 2.497,
+      "step": 5056
+    },
+    {
+      "epoch": 0.21393518910229292,
+      "grad_norm": 0.21815036237239838,
+      "learning_rate": 0.001,
+      "loss": 2.4716,
+      "step": 5057
+    },
+    {
+      "epoch": 0.2139774938658093,
+      "grad_norm": 0.976707398891449,
+      "learning_rate": 0.001,
+      "loss": 2.0707,
+      "step": 5058
+    },
+    {
+      "epoch": 0.21401979862932566,
+      "grad_norm": 1.2479373216629028,
+      "learning_rate": 0.001,
+      "loss": 1.7482,
+      "step": 5059
+    },
+    {
+      "epoch": 0.21406210339284204,
+      "grad_norm": 0.31325292587280273,
+      "learning_rate": 0.001,
+      "loss": 2.5769,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2141044081563584,
+      "grad_norm": 3.2416415214538574,
+      "learning_rate": 0.001,
+      "loss": 2.0277,
+      "step": 5061
+    },
+    {
+      "epoch": 0.21414671291987478,
+      "grad_norm": 0.21428047120571136,
+      "learning_rate": 0.001,
+      "loss": 1.7499,
+      "step": 5062
+    },
+    {
+      "epoch": 0.21418901768339116,
+      "grad_norm": 0.7413913607597351,
+      "learning_rate": 0.001,
+      "loss": 3.0148,
+      "step": 5063
+    },
+    {
+      "epoch": 0.2142313224469075,
+      "grad_norm": 3.3833765983581543,
+      "learning_rate": 0.001,
+      "loss": 2.9668,
+      "step": 5064
+    },
+    {
+      "epoch": 0.2142736272104239,
+      "grad_norm": 0.934876024723053,
+      "learning_rate": 0.001,
+      "loss": 2.7919,
+      "step": 5065
+    },
+    {
+      "epoch": 0.21431593197394028,
+      "grad_norm": 0.4856380224227905,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 5066
+    },
+    {
+      "epoch": 0.21435823673745663,
+      "grad_norm": 0.5465160012245178,
+      "learning_rate": 0.001,
+      "loss": 2.8415,
+      "step": 5067
+    },
+    {
+      "epoch": 0.214400541500973,
+      "grad_norm": 0.22289830446243286,
+      "learning_rate": 0.001,
+      "loss": 2.0494,
+      "step": 5068
+    },
+    {
+      "epoch": 0.2144428462644894,
+      "grad_norm": 2.262659788131714,
+      "learning_rate": 0.001,
+      "loss": 2.1616,
+      "step": 5069
+    },
+    {
+      "epoch": 0.21448515102800575,
+      "grad_norm": 0.2547045052051544,
+      "learning_rate": 0.001,
+      "loss": 2.714,
+      "step": 5070
+    },
+    {
+      "epoch": 0.21452745579152213,
+      "grad_norm": 0.2590230703353882,
+      "learning_rate": 0.001,
+      "loss": 2.928,
+      "step": 5071
+    },
+    {
+      "epoch": 0.21456976055503849,
+      "grad_norm": 0.42084890604019165,
+      "learning_rate": 0.001,
+      "loss": 3.0684,
+      "step": 5072
+    },
+    {
+      "epoch": 0.21461206531855487,
+      "grad_norm": 0.7488611340522766,
+      "learning_rate": 0.001,
+      "loss": 2.8074,
+      "step": 5073
+    },
+    {
+      "epoch": 0.21465437008207125,
+      "grad_norm": 0.32366877794265747,
+      "learning_rate": 0.001,
+      "loss": 3.2808,
+      "step": 5074
+    },
+    {
+      "epoch": 0.2146966748455876,
+      "grad_norm": 0.20374761521816254,
+      "learning_rate": 0.001,
+      "loss": 2.5365,
+      "step": 5075
+    },
+    {
+      "epoch": 0.21473897960910399,
+      "grad_norm": 0.23549343645572662,
+      "learning_rate": 0.001,
+      "loss": 3.1777,
+      "step": 5076
+    },
+    {
+      "epoch": 0.21478128437262037,
+      "grad_norm": 17.203941345214844,
+      "learning_rate": 0.001,
+      "loss": 2.5717,
+      "step": 5077
+    },
+    {
+      "epoch": 0.21482358913613672,
+      "grad_norm": 0.25522732734680176,
+      "learning_rate": 0.001,
+      "loss": 2.9367,
+      "step": 5078
+    },
+    {
+      "epoch": 0.2148658938996531,
+      "grad_norm": 0.3394809365272522,
+      "learning_rate": 0.001,
+      "loss": 2.3561,
+      "step": 5079
+    },
+    {
+      "epoch": 0.2149081986631695,
+      "grad_norm": 0.21299435198307037,
+      "learning_rate": 0.001,
+      "loss": 2.7499,
+      "step": 5080
+    },
+    {
+      "epoch": 0.21495050342668584,
+      "grad_norm": 0.1964573711156845,
+      "learning_rate": 0.001,
+      "loss": 2.3468,
+      "step": 5081
+    },
+    {
+      "epoch": 0.21499280819020222,
+      "grad_norm": 0.3184814155101776,
+      "learning_rate": 0.001,
+      "loss": 2.4803,
+      "step": 5082
+    },
+    {
+      "epoch": 0.21503511295371858,
+      "grad_norm": 0.34336036443710327,
+      "learning_rate": 0.001,
+      "loss": 3.3502,
+      "step": 5083
+    },
+    {
+      "epoch": 0.21507741771723496,
+      "grad_norm": 0.25828707218170166,
+      "learning_rate": 0.001,
+      "loss": 3.5646,
+      "step": 5084
+    },
+    {
+      "epoch": 0.21511972248075134,
+      "grad_norm": 0.4242289066314697,
+      "learning_rate": 0.001,
+      "loss": 2.8577,
+      "step": 5085
+    },
+    {
+      "epoch": 0.2151620272442677,
+      "grad_norm": 0.18986240029335022,
+      "learning_rate": 0.001,
+      "loss": 1.9367,
+      "step": 5086
+    },
+    {
+      "epoch": 0.21520433200778408,
+      "grad_norm": 0.19585952162742615,
+      "learning_rate": 0.001,
+      "loss": 2.1596,
+      "step": 5087
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.3815001845359802,
+      "learning_rate": 0.001,
+      "loss": 2.9462,
+      "step": 5088
+    },
+    {
+      "epoch": 0.2152889415348168,
+      "grad_norm": 0.26725849509239197,
+      "learning_rate": 0.001,
+      "loss": 3.1378,
+      "step": 5089
+    },
+    {
+      "epoch": 0.2153312462983332,
+      "grad_norm": 0.4309972822666168,
+      "learning_rate": 0.001,
+      "loss": 3.3749,
+      "step": 5090
+    },
+    {
+      "epoch": 0.21537355106184958,
+      "grad_norm": 0.19142082333564758,
+      "learning_rate": 0.001,
+      "loss": 2.2686,
+      "step": 5091
+    },
+    {
+      "epoch": 0.21541585582536593,
+      "grad_norm": 0.22003363072872162,
+      "learning_rate": 0.001,
+      "loss": 1.996,
+      "step": 5092
+    },
+    {
+      "epoch": 0.2154581605888823,
+      "grad_norm": 0.30616286396980286,
+      "learning_rate": 0.001,
+      "loss": 3.3171,
+      "step": 5093
+    },
+    {
+      "epoch": 0.21550046535239867,
+      "grad_norm": 8.013957977294922,
+      "learning_rate": 0.001,
+      "loss": 2.2265,
+      "step": 5094
+    },
+    {
+      "epoch": 0.21554277011591505,
+      "grad_norm": 0.3616850972175598,
+      "learning_rate": 0.001,
+      "loss": 3.1465,
+      "step": 5095
+    },
+    {
+      "epoch": 0.21558507487943143,
+      "grad_norm": 1.7922744750976562,
+      "learning_rate": 0.001,
+      "loss": 2.7905,
+      "step": 5096
+    },
+    {
+      "epoch": 0.2156273796429478,
+      "grad_norm": 0.18036864697933197,
+      "learning_rate": 0.001,
+      "loss": 1.8923,
+      "step": 5097
+    },
+    {
+      "epoch": 0.21566968440646417,
+      "grad_norm": 0.5982964038848877,
+      "learning_rate": 0.001,
+      "loss": 2.0202,
+      "step": 5098
+    },
+    {
+      "epoch": 0.21571198916998055,
+      "grad_norm": 0.37752029299736023,
+      "learning_rate": 0.001,
+      "loss": 1.8078,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2157542939334969,
+      "grad_norm": 0.28953149914741516,
+      "learning_rate": 0.001,
+      "loss": 2.2722,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2157965986970133,
+      "grad_norm": 0.275361031293869,
+      "learning_rate": 0.001,
+      "loss": 1.8862,
+      "step": 5101
+    },
+    {
+      "epoch": 0.21583890346052967,
+      "grad_norm": 2.51975679397583,
+      "learning_rate": 0.001,
+      "loss": 2.4179,
+      "step": 5102
+    },
+    {
+      "epoch": 0.21588120822404602,
+      "grad_norm": 3.455864191055298,
+      "learning_rate": 0.001,
+      "loss": 2.9148,
+      "step": 5103
+    },
+    {
+      "epoch": 0.2159235129875624,
+      "grad_norm": 0.20818361639976501,
+      "learning_rate": 0.001,
+      "loss": 1.5796,
+      "step": 5104
+    },
+    {
+      "epoch": 0.21596581775107876,
+      "grad_norm": 0.2759215831756592,
+      "learning_rate": 0.001,
+      "loss": 3.4658,
+      "step": 5105
+    },
+    {
+      "epoch": 0.21600812251459514,
+      "grad_norm": 0.2338385432958603,
+      "learning_rate": 0.001,
+      "loss": 2.5726,
+      "step": 5106
+    },
+    {
+      "epoch": 0.21605042727811152,
+      "grad_norm": 0.2205616980791092,
+      "learning_rate": 0.001,
+      "loss": 4.0847,
+      "step": 5107
+    },
+    {
+      "epoch": 0.21609273204162788,
+      "grad_norm": 0.8315731287002563,
+      "learning_rate": 0.001,
+      "loss": 2.2934,
+      "step": 5108
+    },
+    {
+      "epoch": 0.21613503680514426,
+      "grad_norm": 0.22039955854415894,
+      "learning_rate": 0.001,
+      "loss": 2.0346,
+      "step": 5109
+    },
+    {
+      "epoch": 0.21617734156866064,
+      "grad_norm": 0.89706951379776,
+      "learning_rate": 0.001,
+      "loss": 2.0214,
+      "step": 5110
+    },
+    {
+      "epoch": 0.216219646332177,
+      "grad_norm": 13.432646751403809,
+      "learning_rate": 0.001,
+      "loss": 2.9813,
+      "step": 5111
+    },
+    {
+      "epoch": 0.21626195109569338,
+      "grad_norm": 0.20482416450977325,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 5112
+    },
+    {
+      "epoch": 0.21630425585920976,
+      "grad_norm": 0.4937507212162018,
+      "learning_rate": 0.001,
+      "loss": 1.7228,
+      "step": 5113
+    },
+    {
+      "epoch": 0.21634656062272611,
+      "grad_norm": 0.3154240548610687,
+      "learning_rate": 0.001,
+      "loss": 2.3277,
+      "step": 5114
+    },
+    {
+      "epoch": 0.2163888653862425,
+      "grad_norm": 0.4587913453578949,
+      "learning_rate": 0.001,
+      "loss": 2.2894,
+      "step": 5115
+    },
+    {
+      "epoch": 0.21643117014975885,
+      "grad_norm": 0.39989644289016724,
+      "learning_rate": 0.001,
+      "loss": 2.5237,
+      "step": 5116
+    },
+    {
+      "epoch": 0.21647347491327523,
+      "grad_norm": 0.21547210216522217,
+      "learning_rate": 0.001,
+      "loss": 2.0295,
+      "step": 5117
+    },
+    {
+      "epoch": 0.21651577967679161,
+      "grad_norm": 2.3156797885894775,
+      "learning_rate": 0.001,
+      "loss": 2.5201,
+      "step": 5118
+    },
+    {
+      "epoch": 0.21655808444030797,
+      "grad_norm": 0.19337671995162964,
+      "learning_rate": 0.001,
+      "loss": 1.8627,
+      "step": 5119
+    },
+    {
+      "epoch": 0.21660038920382435,
+      "grad_norm": 0.24067647755146027,
+      "learning_rate": 0.001,
+      "loss": 2.2175,
+      "step": 5120
+    },
+    {
+      "epoch": 0.21664269396734073,
+      "grad_norm": 10.653289794921875,
+      "learning_rate": 0.001,
+      "loss": 3.6184,
+      "step": 5121
+    },
+    {
+      "epoch": 0.2166849987308571,
+      "grad_norm": 0.22319035232067108,
+      "learning_rate": 0.001,
+      "loss": 2.0583,
+      "step": 5122
+    },
+    {
+      "epoch": 0.21672730349437347,
+      "grad_norm": 13.307822227478027,
+      "learning_rate": 0.001,
+      "loss": 1.885,
+      "step": 5123
+    },
+    {
+      "epoch": 0.21676960825788985,
+      "grad_norm": 41.82732009887695,
+      "learning_rate": 0.001,
+      "loss": 3.325,
+      "step": 5124
+    },
+    {
+      "epoch": 0.2168119130214062,
+      "grad_norm": 0.20223648846149445,
+      "learning_rate": 0.001,
+      "loss": 1.657,
+      "step": 5125
+    },
+    {
+      "epoch": 0.2168542177849226,
+      "grad_norm": 0.708741307258606,
+      "learning_rate": 0.001,
+      "loss": 2.2723,
+      "step": 5126
+    },
+    {
+      "epoch": 0.21689652254843894,
+      "grad_norm": 0.2860840857028961,
+      "learning_rate": 0.001,
+      "loss": 2.3074,
+      "step": 5127
+    },
+    {
+      "epoch": 0.21693882731195532,
+      "grad_norm": 0.4499269425868988,
+      "learning_rate": 0.001,
+      "loss": 2.5004,
+      "step": 5128
+    },
+    {
+      "epoch": 0.2169811320754717,
+      "grad_norm": 0.3753020763397217,
+      "learning_rate": 0.001,
+      "loss": 2.7281,
+      "step": 5129
+    },
+    {
+      "epoch": 0.21702343683898806,
+      "grad_norm": 6.676267623901367,
+      "learning_rate": 0.001,
+      "loss": 2.5515,
+      "step": 5130
+    },
+    {
+      "epoch": 0.21706574160250444,
+      "grad_norm": 2.6914873123168945,
+      "learning_rate": 0.001,
+      "loss": 2.5111,
+      "step": 5131
+    },
+    {
+      "epoch": 0.21710804636602082,
+      "grad_norm": 0.24604924023151398,
+      "learning_rate": 0.001,
+      "loss": 2.1823,
+      "step": 5132
+    },
+    {
+      "epoch": 0.21715035112953718,
+      "grad_norm": 0.2534256875514984,
+      "learning_rate": 0.001,
+      "loss": 2.6668,
+      "step": 5133
+    },
+    {
+      "epoch": 0.21719265589305356,
+      "grad_norm": 0.32457131147384644,
+      "learning_rate": 0.001,
+      "loss": 3.2796,
+      "step": 5134
+    },
+    {
+      "epoch": 0.21723496065656994,
+      "grad_norm": 0.3132564425468445,
+      "learning_rate": 0.001,
+      "loss": 3.2073,
+      "step": 5135
+    },
+    {
+      "epoch": 0.2172772654200863,
+      "grad_norm": 0.2281041145324707,
+      "learning_rate": 0.001,
+      "loss": 2.6315,
+      "step": 5136
+    },
+    {
+      "epoch": 0.21731957018360268,
+      "grad_norm": 0.7160950303077698,
+      "learning_rate": 0.001,
+      "loss": 2.3004,
+      "step": 5137
+    },
+    {
+      "epoch": 0.21736187494711903,
+      "grad_norm": 0.2322395145893097,
+      "learning_rate": 0.001,
+      "loss": 2.7504,
+      "step": 5138
+    },
+    {
+      "epoch": 0.21740417971063541,
+      "grad_norm": 0.3613683879375458,
+      "learning_rate": 0.001,
+      "loss": 2.5574,
+      "step": 5139
+    },
+    {
+      "epoch": 0.2174464844741518,
+      "grad_norm": 0.24612249433994293,
+      "learning_rate": 0.001,
+      "loss": 2.8064,
+      "step": 5140
+    },
+    {
+      "epoch": 0.21748878923766815,
+      "grad_norm": 0.2229098081588745,
+      "learning_rate": 0.001,
+      "loss": 2.07,
+      "step": 5141
+    },
+    {
+      "epoch": 0.21753109400118453,
+      "grad_norm": 0.19228215515613556,
+      "learning_rate": 0.001,
+      "loss": 1.859,
+      "step": 5142
+    },
+    {
+      "epoch": 0.21757339876470091,
+      "grad_norm": 0.3609655499458313,
+      "learning_rate": 0.001,
+      "loss": 3.1708,
+      "step": 5143
+    },
+    {
+      "epoch": 0.21761570352821727,
+      "grad_norm": 0.3540642261505127,
+      "learning_rate": 0.001,
+      "loss": 2.7362,
+      "step": 5144
+    },
+    {
+      "epoch": 0.21765800829173365,
+      "grad_norm": 0.37258389592170715,
+      "learning_rate": 0.001,
+      "loss": 2.3635,
+      "step": 5145
+    },
+    {
+      "epoch": 0.21770031305525003,
+      "grad_norm": 0.2843034863471985,
+      "learning_rate": 0.001,
+      "loss": 2.8098,
+      "step": 5146
+    },
+    {
+      "epoch": 0.2177426178187664,
+      "grad_norm": 0.2767336368560791,
+      "learning_rate": 0.001,
+      "loss": 2.7574,
+      "step": 5147
+    },
+    {
+      "epoch": 0.21778492258228277,
+      "grad_norm": 0.21004167199134827,
+      "learning_rate": 0.001,
+      "loss": 2.2051,
+      "step": 5148
+    },
+    {
+      "epoch": 0.21782722734579912,
+      "grad_norm": 0.23345938324928284,
+      "learning_rate": 0.001,
+      "loss": 3.7967,
+      "step": 5149
+    },
+    {
+      "epoch": 0.2178695321093155,
+      "grad_norm": 0.44973114132881165,
+      "learning_rate": 0.001,
+      "loss": 3.6378,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2179118368728319,
+      "grad_norm": 0.26152345538139343,
+      "learning_rate": 0.001,
+      "loss": 2.7861,
+      "step": 5151
+    },
+    {
+      "epoch": 0.21795414163634824,
+      "grad_norm": 0.3166126012802124,
+      "learning_rate": 0.001,
+      "loss": 2.9621,
+      "step": 5152
+    },
+    {
+      "epoch": 0.21799644639986462,
+      "grad_norm": 0.2214854210615158,
+      "learning_rate": 0.001,
+      "loss": 2.0622,
+      "step": 5153
+    },
+    {
+      "epoch": 0.218038751163381,
+      "grad_norm": 0.21160806715488434,
+      "learning_rate": 0.001,
+      "loss": 2.9396,
+      "step": 5154
+    },
+    {
+      "epoch": 0.21808105592689736,
+      "grad_norm": 0.21767565608024597,
+      "learning_rate": 0.001,
+      "loss": 1.7393,
+      "step": 5155
+    },
+    {
+      "epoch": 0.21812336069041374,
+      "grad_norm": 18.17046356201172,
+      "learning_rate": 0.001,
+      "loss": 2.2801,
+      "step": 5156
+    },
+    {
+      "epoch": 0.21816566545393012,
+      "grad_norm": 0.4204169511795044,
+      "learning_rate": 0.001,
+      "loss": 2.1432,
+      "step": 5157
+    },
+    {
+      "epoch": 0.21820797021744648,
+      "grad_norm": 0.42446818947792053,
+      "learning_rate": 0.001,
+      "loss": 1.7413,
+      "step": 5158
+    },
+    {
+      "epoch": 0.21825027498096286,
+      "grad_norm": 0.2123391479253769,
+      "learning_rate": 0.001,
+      "loss": 2.2634,
+      "step": 5159
+    },
+    {
+      "epoch": 0.21829257974447921,
+      "grad_norm": 0.2388673722743988,
+      "learning_rate": 0.001,
+      "loss": 2.2383,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2183348845079956,
+      "grad_norm": 0.870901882648468,
+      "learning_rate": 0.001,
+      "loss": 3.157,
+      "step": 5161
+    },
+    {
+      "epoch": 0.21837718927151198,
+      "grad_norm": 0.3035809397697449,
+      "learning_rate": 0.001,
+      "loss": 3.1643,
+      "step": 5162
+    },
+    {
+      "epoch": 0.21841949403502833,
+      "grad_norm": 0.40469470620155334,
+      "learning_rate": 0.001,
+      "loss": 2.211,
+      "step": 5163
+    },
+    {
+      "epoch": 0.21846179879854472,
+      "grad_norm": 0.22423990070819855,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2185041035620611,
+      "grad_norm": 0.20097650587558746,
+      "learning_rate": 0.001,
+      "loss": 2.2083,
+      "step": 5165
+    },
+    {
+      "epoch": 0.21854640832557745,
+      "grad_norm": 0.2772468626499176,
+      "learning_rate": 0.001,
+      "loss": 2.8581,
+      "step": 5166
+    },
+    {
+      "epoch": 0.21858871308909383,
+      "grad_norm": 0.20195744931697845,
+      "learning_rate": 0.001,
+      "loss": 1.8894,
+      "step": 5167
+    },
+    {
+      "epoch": 0.21863101785261022,
+      "grad_norm": 0.21510477364063263,
+      "learning_rate": 0.001,
+      "loss": 2.1195,
+      "step": 5168
+    },
+    {
+      "epoch": 0.21867332261612657,
+      "grad_norm": 0.2677229344844818,
+      "learning_rate": 0.001,
+      "loss": 2.9506,
+      "step": 5169
+    },
+    {
+      "epoch": 0.21871562737964295,
+      "grad_norm": 1.1444584131240845,
+      "learning_rate": 0.001,
+      "loss": 1.6645,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2187579321431593,
+      "grad_norm": 0.30999642610549927,
+      "learning_rate": 0.001,
+      "loss": 2.7087,
+      "step": 5171
+    },
+    {
+      "epoch": 0.2188002369066757,
+      "grad_norm": 0.20951640605926514,
+      "learning_rate": 0.001,
+      "loss": 2.8629,
+      "step": 5172
+    },
+    {
+      "epoch": 0.21884254167019207,
+      "grad_norm": 0.2823593020439148,
+      "learning_rate": 0.001,
+      "loss": 1.8454,
+      "step": 5173
+    },
+    {
+      "epoch": 0.21888484643370842,
+      "grad_norm": 0.19759036600589752,
+      "learning_rate": 0.001,
+      "loss": 1.6832,
+      "step": 5174
+    },
+    {
+      "epoch": 0.2189271511972248,
+      "grad_norm": 1.2790015935897827,
+      "learning_rate": 0.001,
+      "loss": 2.2598,
+      "step": 5175
+    },
+    {
+      "epoch": 0.2189694559607412,
+      "grad_norm": 0.3004913926124573,
+      "learning_rate": 0.001,
+      "loss": 2.8445,
+      "step": 5176
+    },
+    {
+      "epoch": 0.21901176072425754,
+      "grad_norm": 0.1984034925699234,
+      "learning_rate": 0.001,
+      "loss": 2.0588,
+      "step": 5177
+    },
+    {
+      "epoch": 0.21905406548777392,
+      "grad_norm": 0.6654098033905029,
+      "learning_rate": 0.001,
+      "loss": 2.5635,
+      "step": 5178
+    },
+    {
+      "epoch": 0.2190963702512903,
+      "grad_norm": 0.21716728806495667,
+      "learning_rate": 0.001,
+      "loss": 1.8904,
+      "step": 5179
+    },
+    {
+      "epoch": 0.21913867501480666,
+      "grad_norm": 0.2529633939266205,
+      "learning_rate": 0.001,
+      "loss": 3.3847,
+      "step": 5180
+    },
+    {
+      "epoch": 0.21918097977832304,
+      "grad_norm": 0.24113009870052338,
+      "learning_rate": 0.001,
+      "loss": 2.8504,
+      "step": 5181
+    },
+    {
+      "epoch": 0.21922328454183942,
+      "grad_norm": 0.23012512922286987,
+      "learning_rate": 0.001,
+      "loss": 1.8835,
+      "step": 5182
+    },
+    {
+      "epoch": 0.21926558930535578,
+      "grad_norm": 0.2334389090538025,
+      "learning_rate": 0.001,
+      "loss": 2.9091,
+      "step": 5183
+    },
+    {
+      "epoch": 0.21930789406887216,
+      "grad_norm": 3.1486589908599854,
+      "learning_rate": 0.001,
+      "loss": 3.0328,
+      "step": 5184
+    },
+    {
+      "epoch": 0.21935019883238852,
+      "grad_norm": 0.18506896495819092,
+      "learning_rate": 0.001,
+      "loss": 3.1343,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2193925035959049,
+      "grad_norm": 0.24055512249469757,
+      "learning_rate": 0.001,
+      "loss": 1.9749,
+      "step": 5186
+    },
+    {
+      "epoch": 0.21943480835942128,
+      "grad_norm": 0.3492937684059143,
+      "learning_rate": 0.001,
+      "loss": 2.3837,
+      "step": 5187
+    },
+    {
+      "epoch": 0.21947711312293763,
+      "grad_norm": 0.2223334014415741,
+      "learning_rate": 0.001,
+      "loss": 2.7189,
+      "step": 5188
+    },
+    {
+      "epoch": 0.21951941788645402,
+      "grad_norm": 2.0929033756256104,
+      "learning_rate": 0.001,
+      "loss": 2.3026,
+      "step": 5189
+    },
+    {
+      "epoch": 0.2195617226499704,
+      "grad_norm": 3.1999146938323975,
+      "learning_rate": 0.001,
+      "loss": 1.7523,
+      "step": 5190
+    },
+    {
+      "epoch": 0.21960402741348675,
+      "grad_norm": 0.19789402186870575,
+      "learning_rate": 0.001,
+      "loss": 1.6299,
+      "step": 5191
+    },
+    {
+      "epoch": 0.21964633217700313,
+      "grad_norm": 0.23743316531181335,
+      "learning_rate": 0.001,
+      "loss": 3.3135,
+      "step": 5192
+    },
+    {
+      "epoch": 0.21968863694051952,
+      "grad_norm": 0.2639296054840088,
+      "learning_rate": 0.001,
+      "loss": 2.3337,
+      "step": 5193
+    },
+    {
+      "epoch": 0.21973094170403587,
+      "grad_norm": 0.24590055644512177,
+      "learning_rate": 0.001,
+      "loss": 2.4579,
+      "step": 5194
+    },
+    {
+      "epoch": 0.21977324646755225,
+      "grad_norm": 0.2431252896785736,
+      "learning_rate": 0.001,
+      "loss": 2.1513,
+      "step": 5195
+    },
+    {
+      "epoch": 0.2198155512310686,
+      "grad_norm": 0.22125622630119324,
+      "learning_rate": 0.001,
+      "loss": 2.3895,
+      "step": 5196
+    },
+    {
+      "epoch": 0.219857855994585,
+      "grad_norm": 0.19629360735416412,
+      "learning_rate": 0.001,
+      "loss": 2.045,
+      "step": 5197
+    },
+    {
+      "epoch": 0.21990016075810137,
+      "grad_norm": 0.23924358189105988,
+      "learning_rate": 0.001,
+      "loss": 2.2106,
+      "step": 5198
+    },
+    {
+      "epoch": 0.21994246552161772,
+      "grad_norm": 0.21963438391685486,
+      "learning_rate": 0.001,
+      "loss": 2.2913,
+      "step": 5199
+    },
+    {
+      "epoch": 0.2199847702851341,
+      "grad_norm": 0.2901560366153717,
+      "learning_rate": 0.001,
+      "loss": 2.8481,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2200270750486505,
+      "grad_norm": 0.2538588345050812,
+      "learning_rate": 0.001,
+      "loss": 3.4287,
+      "step": 5201
+    },
+    {
+      "epoch": 0.22006937981216684,
+      "grad_norm": 0.23842386901378632,
+      "learning_rate": 0.001,
+      "loss": 3.3795,
+      "step": 5202
+    },
+    {
+      "epoch": 0.22011168457568323,
+      "grad_norm": 0.255536288022995,
+      "learning_rate": 0.001,
+      "loss": 3.2765,
+      "step": 5203
+    },
+    {
+      "epoch": 0.2201539893391996,
+      "grad_norm": 4.658047199249268,
+      "learning_rate": 0.001,
+      "loss": 2.3231,
+      "step": 5204
+    },
+    {
+      "epoch": 0.22019629410271596,
+      "grad_norm": 0.20021434128284454,
+      "learning_rate": 0.001,
+      "loss": 2.3354,
+      "step": 5205
+    },
+    {
+      "epoch": 0.22023859886623234,
+      "grad_norm": 0.19822002947330475,
+      "learning_rate": 0.001,
+      "loss": 1.7009,
+      "step": 5206
+    },
+    {
+      "epoch": 0.2202809036297487,
+      "grad_norm": 0.21478721499443054,
+      "learning_rate": 0.001,
+      "loss": 2.0951,
+      "step": 5207
+    },
+    {
+      "epoch": 0.22032320839326508,
+      "grad_norm": 0.21483932435512543,
+      "learning_rate": 0.001,
+      "loss": 2.7571,
+      "step": 5208
+    },
+    {
+      "epoch": 0.22036551315678146,
+      "grad_norm": 0.23699632287025452,
+      "learning_rate": 0.001,
+      "loss": 2.2913,
+      "step": 5209
+    },
+    {
+      "epoch": 0.22040781792029782,
+      "grad_norm": 0.26198896765708923,
+      "learning_rate": 0.001,
+      "loss": 4.5366,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2204501226838142,
+      "grad_norm": 0.21379588544368744,
+      "learning_rate": 0.001,
+      "loss": 1.9979,
+      "step": 5211
+    },
+    {
+      "epoch": 0.22049242744733058,
+      "grad_norm": 0.2645533084869385,
+      "learning_rate": 0.001,
+      "loss": 2.2221,
+      "step": 5212
+    },
+    {
+      "epoch": 0.22053473221084693,
+      "grad_norm": 0.19984637200832367,
+      "learning_rate": 0.001,
+      "loss": 2.0598,
+      "step": 5213
+    },
+    {
+      "epoch": 0.22057703697436332,
+      "grad_norm": 0.2022748589515686,
+      "learning_rate": 0.001,
+      "loss": 2.3214,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2206193417378797,
+      "grad_norm": 0.20694908499717712,
+      "learning_rate": 0.001,
+      "loss": 2.0205,
+      "step": 5215
+    },
+    {
+      "epoch": 0.22066164650139605,
+      "grad_norm": 0.2556994557380676,
+      "learning_rate": 0.001,
+      "loss": 2.4558,
+      "step": 5216
+    },
+    {
+      "epoch": 0.22070395126491243,
+      "grad_norm": 0.7074611186981201,
+      "learning_rate": 0.001,
+      "loss": 1.6551,
+      "step": 5217
+    },
+    {
+      "epoch": 0.2207462560284288,
+      "grad_norm": 0.19164715707302094,
+      "learning_rate": 0.001,
+      "loss": 2.1213,
+      "step": 5218
+    },
+    {
+      "epoch": 0.22078856079194517,
+      "grad_norm": 0.32563894987106323,
+      "learning_rate": 0.001,
+      "loss": 2.366,
+      "step": 5219
+    },
+    {
+      "epoch": 0.22083086555546155,
+      "grad_norm": 0.21705080568790436,
+      "learning_rate": 0.001,
+      "loss": 2.4052,
+      "step": 5220
+    },
+    {
+      "epoch": 0.2208731703189779,
+      "grad_norm": 0.20345237851142883,
+      "learning_rate": 0.001,
+      "loss": 2.2577,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2209154750824943,
+      "grad_norm": 0.2837914824485779,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 5222
+    },
+    {
+      "epoch": 0.22095777984601067,
+      "grad_norm": 1.1149832010269165,
+      "learning_rate": 0.001,
+      "loss": 2.357,
+      "step": 5223
+    },
+    {
+      "epoch": 0.22100008460952703,
+      "grad_norm": 0.23472867906093597,
+      "learning_rate": 0.001,
+      "loss": 2.6525,
+      "step": 5224
+    },
+    {
+      "epoch": 0.2210423893730434,
+      "grad_norm": 0.6274850964546204,
+      "learning_rate": 0.001,
+      "loss": 3.9738,
+      "step": 5225
+    },
+    {
+      "epoch": 0.2210846941365598,
+      "grad_norm": 0.21737419068813324,
+      "learning_rate": 0.001,
+      "loss": 2.4552,
+      "step": 5226
+    },
+    {
+      "epoch": 0.22112699890007614,
+      "grad_norm": 0.2056940793991089,
+      "learning_rate": 0.001,
+      "loss": 2.8104,
+      "step": 5227
+    },
+    {
+      "epoch": 0.22116930366359253,
+      "grad_norm": 0.2175094336271286,
+      "learning_rate": 0.001,
+      "loss": 2.1255,
+      "step": 5228
+    },
+    {
+      "epoch": 0.22121160842710888,
+      "grad_norm": 0.39655065536499023,
+      "learning_rate": 0.001,
+      "loss": 1.9519,
+      "step": 5229
+    },
+    {
+      "epoch": 0.22125391319062526,
+      "grad_norm": 0.20697876811027527,
+      "learning_rate": 0.001,
+      "loss": 2.4179,
+      "step": 5230
+    },
+    {
+      "epoch": 0.22129621795414164,
+      "grad_norm": 0.217100128531456,
+      "learning_rate": 0.001,
+      "loss": 2.7213,
+      "step": 5231
+    },
+    {
+      "epoch": 0.221338522717658,
+      "grad_norm": 0.21442492306232452,
+      "learning_rate": 0.001,
+      "loss": 2.8483,
+      "step": 5232
+    },
+    {
+      "epoch": 0.22138082748117438,
+      "grad_norm": 0.21602658927440643,
+      "learning_rate": 0.001,
+      "loss": 1.6262,
+      "step": 5233
+    },
+    {
+      "epoch": 0.22142313224469076,
+      "grad_norm": 0.22940513491630554,
+      "learning_rate": 0.001,
+      "loss": 1.9484,
+      "step": 5234
+    },
+    {
+      "epoch": 0.22146543700820712,
+      "grad_norm": 3.8185322284698486,
+      "learning_rate": 0.001,
+      "loss": 2.8154,
+      "step": 5235
+    },
+    {
+      "epoch": 0.2215077417717235,
+      "grad_norm": 0.28936222195625305,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 5236
+    },
+    {
+      "epoch": 0.22155004653523988,
+      "grad_norm": 0.24605505168437958,
+      "learning_rate": 0.001,
+      "loss": 3.03,
+      "step": 5237
+    },
+    {
+      "epoch": 0.22159235129875623,
+      "grad_norm": 0.20465493202209473,
+      "learning_rate": 0.001,
+      "loss": 2.3525,
+      "step": 5238
+    },
+    {
+      "epoch": 0.22163465606227262,
+      "grad_norm": 0.20358186960220337,
+      "learning_rate": 0.001,
+      "loss": 2.1494,
+      "step": 5239
+    },
+    {
+      "epoch": 0.22167696082578897,
+      "grad_norm": 0.2088003009557724,
+      "learning_rate": 0.001,
+      "loss": 1.9404,
+      "step": 5240
+    },
+    {
+      "epoch": 0.22171926558930535,
+      "grad_norm": 0.20697228610515594,
+      "learning_rate": 0.001,
+      "loss": 3.8656,
+      "step": 5241
+    },
+    {
+      "epoch": 0.22176157035282174,
+      "grad_norm": 0.22819316387176514,
+      "learning_rate": 0.001,
+      "loss": 2.1,
+      "step": 5242
+    },
+    {
+      "epoch": 0.2218038751163381,
+      "grad_norm": 0.32436373829841614,
+      "learning_rate": 0.001,
+      "loss": 3.2291,
+      "step": 5243
+    },
+    {
+      "epoch": 0.22184617987985447,
+      "grad_norm": 0.3842344880104065,
+      "learning_rate": 0.001,
+      "loss": 1.9864,
+      "step": 5244
+    },
+    {
+      "epoch": 0.22188848464337085,
+      "grad_norm": 0.22547200322151184,
+      "learning_rate": 0.001,
+      "loss": 2.1806,
+      "step": 5245
+    },
+    {
+      "epoch": 0.2219307894068872,
+      "grad_norm": 0.201228067278862,
+      "learning_rate": 0.001,
+      "loss": 2.6097,
+      "step": 5246
+    },
+    {
+      "epoch": 0.2219730941704036,
+      "grad_norm": 0.21469905972480774,
+      "learning_rate": 0.001,
+      "loss": 1.884,
+      "step": 5247
+    },
+    {
+      "epoch": 0.22201539893391997,
+      "grad_norm": 17.149986267089844,
+      "learning_rate": 0.001,
+      "loss": 2.0795,
+      "step": 5248
+    },
+    {
+      "epoch": 0.22205770369743633,
+      "grad_norm": 2.927840232849121,
+      "learning_rate": 0.001,
+      "loss": 2.0508,
+      "step": 5249
+    },
+    {
+      "epoch": 0.2221000084609527,
+      "grad_norm": 0.20757484436035156,
+      "learning_rate": 0.001,
+      "loss": 3.9396,
+      "step": 5250
+    },
+    {
+      "epoch": 0.22214231322446906,
+      "grad_norm": 0.22015203535556793,
+      "learning_rate": 0.001,
+      "loss": 2.512,
+      "step": 5251
+    },
+    {
+      "epoch": 0.22218461798798544,
+      "grad_norm": 0.22938333451747894,
+      "learning_rate": 0.001,
+      "loss": 2.2927,
+      "step": 5252
+    },
+    {
+      "epoch": 0.22222692275150183,
+      "grad_norm": 0.24902082979679108,
+      "learning_rate": 0.001,
+      "loss": 2.9212,
+      "step": 5253
+    },
+    {
+      "epoch": 0.22226922751501818,
+      "grad_norm": 0.23875977098941803,
+      "learning_rate": 0.001,
+      "loss": 2.0939,
+      "step": 5254
+    },
+    {
+      "epoch": 0.22231153227853456,
+      "grad_norm": 0.21278776228427887,
+      "learning_rate": 0.001,
+      "loss": 1.9339,
+      "step": 5255
+    },
+    {
+      "epoch": 0.22235383704205094,
+      "grad_norm": 0.2972838282585144,
+      "learning_rate": 0.001,
+      "loss": 3.0246,
+      "step": 5256
+    },
+    {
+      "epoch": 0.2223961418055673,
+      "grad_norm": 0.20768025517463684,
+      "learning_rate": 0.001,
+      "loss": 1.9835,
+      "step": 5257
+    },
+    {
+      "epoch": 0.22243844656908368,
+      "grad_norm": 5.894577980041504,
+      "learning_rate": 0.001,
+      "loss": 2.6385,
+      "step": 5258
+    },
+    {
+      "epoch": 0.22248075133260006,
+      "grad_norm": 0.2140832394361496,
+      "learning_rate": 0.001,
+      "loss": 2.7344,
+      "step": 5259
+    },
+    {
+      "epoch": 0.22252305609611642,
+      "grad_norm": 0.20773886144161224,
+      "learning_rate": 0.001,
+      "loss": 2.4667,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2225653608596328,
+      "grad_norm": 0.21243473887443542,
+      "learning_rate": 0.001,
+      "loss": 2.1209,
+      "step": 5261
+    },
+    {
+      "epoch": 0.22260766562314915,
+      "grad_norm": 0.2120569497346878,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 5262
+    },
+    {
+      "epoch": 0.22264997038666554,
+      "grad_norm": 0.25889766216278076,
+      "learning_rate": 0.001,
+      "loss": 2.5132,
+      "step": 5263
+    },
+    {
+      "epoch": 0.22269227515018192,
+      "grad_norm": 0.2106718271970749,
+      "learning_rate": 0.001,
+      "loss": 1.8965,
+      "step": 5264
+    },
+    {
+      "epoch": 0.22273457991369827,
+      "grad_norm": 0.2909772992134094,
+      "learning_rate": 0.001,
+      "loss": 2.1466,
+      "step": 5265
+    },
+    {
+      "epoch": 0.22277688467721465,
+      "grad_norm": 1.4828362464904785,
+      "learning_rate": 0.001,
+      "loss": 2.8965,
+      "step": 5266
+    },
+    {
+      "epoch": 0.22281918944073104,
+      "grad_norm": 0.23213310539722443,
+      "learning_rate": 0.001,
+      "loss": 2.0563,
+      "step": 5267
+    },
+    {
+      "epoch": 0.2228614942042474,
+      "grad_norm": 0.2343592345714569,
+      "learning_rate": 0.001,
+      "loss": 2.6482,
+      "step": 5268
+    },
+    {
+      "epoch": 0.22290379896776377,
+      "grad_norm": 0.2614687979221344,
+      "learning_rate": 0.001,
+      "loss": 2.7576,
+      "step": 5269
+    },
+    {
+      "epoch": 0.22294610373128015,
+      "grad_norm": 4.1902008056640625,
+      "learning_rate": 0.001,
+      "loss": 2.9158,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2229884084947965,
+      "grad_norm": 0.20279575884342194,
+      "learning_rate": 0.001,
+      "loss": 2.205,
+      "step": 5271
+    },
+    {
+      "epoch": 0.2230307132583129,
+      "grad_norm": 0.373855322599411,
+      "learning_rate": 0.001,
+      "loss": 2.6883,
+      "step": 5272
+    },
+    {
+      "epoch": 0.22307301802182924,
+      "grad_norm": 0.3078487813472748,
+      "learning_rate": 0.001,
+      "loss": 1.9063,
+      "step": 5273
+    },
+    {
+      "epoch": 0.22311532278534563,
+      "grad_norm": 0.3240393102169037,
+      "learning_rate": 0.001,
+      "loss": 2.6232,
+      "step": 5274
+    },
+    {
+      "epoch": 0.223157627548862,
+      "grad_norm": 0.3263351321220398,
+      "learning_rate": 0.001,
+      "loss": 2.3034,
+      "step": 5275
+    },
+    {
+      "epoch": 0.22319993231237836,
+      "grad_norm": 0.7479755282402039,
+      "learning_rate": 0.001,
+      "loss": 1.8465,
+      "step": 5276
+    },
+    {
+      "epoch": 0.22324223707589474,
+      "grad_norm": 2.0020651817321777,
+      "learning_rate": 0.001,
+      "loss": 1.859,
+      "step": 5277
+    },
+    {
+      "epoch": 0.22328454183941113,
+      "grad_norm": 0.2206108123064041,
+      "learning_rate": 0.001,
+      "loss": 1.5864,
+      "step": 5278
+    },
+    {
+      "epoch": 0.22332684660292748,
+      "grad_norm": 3.234041452407837,
+      "learning_rate": 0.001,
+      "loss": 2.1619,
+      "step": 5279
+    },
+    {
+      "epoch": 0.22336915136644386,
+      "grad_norm": 0.21610672771930695,
+      "learning_rate": 0.001,
+      "loss": 1.948,
+      "step": 5280
+    },
+    {
+      "epoch": 0.22341145612996025,
+      "grad_norm": 1.4125299453735352,
+      "learning_rate": 0.001,
+      "loss": 2.7824,
+      "step": 5281
+    },
+    {
+      "epoch": 0.2234537608934766,
+      "grad_norm": 0.19420327246189117,
+      "learning_rate": 0.001,
+      "loss": 1.985,
+      "step": 5282
+    },
+    {
+      "epoch": 0.22349606565699298,
+      "grad_norm": 0.26585274934768677,
+      "learning_rate": 0.001,
+      "loss": 2.6315,
+      "step": 5283
+    },
+    {
+      "epoch": 0.22353837042050934,
+      "grad_norm": 1.4780011177062988,
+      "learning_rate": 0.001,
+      "loss": 2.5902,
+      "step": 5284
+    },
+    {
+      "epoch": 0.22358067518402572,
+      "grad_norm": 0.29642167687416077,
+      "learning_rate": 0.001,
+      "loss": 1.7713,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2236229799475421,
+      "grad_norm": 0.192737877368927,
+      "learning_rate": 0.001,
+      "loss": 2.7865,
+      "step": 5286
+    },
+    {
+      "epoch": 0.22366528471105845,
+      "grad_norm": 2.1191747188568115,
+      "learning_rate": 0.001,
+      "loss": 2.4322,
+      "step": 5287
+    },
+    {
+      "epoch": 0.22370758947457484,
+      "grad_norm": 1.5505908727645874,
+      "learning_rate": 0.001,
+      "loss": 2.4076,
+      "step": 5288
+    },
+    {
+      "epoch": 0.22374989423809122,
+      "grad_norm": 0.2791816294193268,
+      "learning_rate": 0.001,
+      "loss": 2.6396,
+      "step": 5289
+    },
+    {
+      "epoch": 0.22379219900160757,
+      "grad_norm": 0.5878250598907471,
+      "learning_rate": 0.001,
+      "loss": 2.4715,
+      "step": 5290
+    },
+    {
+      "epoch": 0.22383450376512395,
+      "grad_norm": 0.22260555624961853,
+      "learning_rate": 0.001,
+      "loss": 1.734,
+      "step": 5291
+    },
+    {
+      "epoch": 0.22387680852864034,
+      "grad_norm": 0.27154138684272766,
+      "learning_rate": 0.001,
+      "loss": 2.9405,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2239191132921567,
+      "grad_norm": 0.24614472687244415,
+      "learning_rate": 0.001,
+      "loss": 3.2545,
+      "step": 5293
+    },
+    {
+      "epoch": 0.22396141805567307,
+      "grad_norm": 0.26551756262779236,
+      "learning_rate": 0.001,
+      "loss": 2.62,
+      "step": 5294
+    },
+    {
+      "epoch": 0.22400372281918943,
+      "grad_norm": 0.23047612607479095,
+      "learning_rate": 0.001,
+      "loss": 1.9018,
+      "step": 5295
+    },
+    {
+      "epoch": 0.2240460275827058,
+      "grad_norm": 0.2711014151573181,
+      "learning_rate": 0.001,
+      "loss": 2.3669,
+      "step": 5296
+    },
+    {
+      "epoch": 0.2240883323462222,
+      "grad_norm": 0.2241722196340561,
+      "learning_rate": 0.001,
+      "loss": 2.2924,
+      "step": 5297
+    },
+    {
+      "epoch": 0.22413063710973855,
+      "grad_norm": 0.36074158549308777,
+      "learning_rate": 0.001,
+      "loss": 2.4914,
+      "step": 5298
+    },
+    {
+      "epoch": 0.22417294187325493,
+      "grad_norm": 0.21040941774845123,
+      "learning_rate": 0.001,
+      "loss": 2.2778,
+      "step": 5299
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 0.17329736053943634,
+      "learning_rate": 0.001,
+      "loss": 2.7799,
+      "step": 5300
+    },
+    {
+      "epoch": 0.22425755140028766,
+      "grad_norm": 1.4445072412490845,
+      "learning_rate": 0.001,
+      "loss": 2.6139,
+      "step": 5301
+    },
+    {
+      "epoch": 0.22429985616380405,
+      "grad_norm": 0.20153391361236572,
+      "learning_rate": 0.001,
+      "loss": 2.4595,
+      "step": 5302
+    },
+    {
+      "epoch": 0.22434216092732043,
+      "grad_norm": 0.3603563606739044,
+      "learning_rate": 0.001,
+      "loss": 3.4425,
+      "step": 5303
+    },
+    {
+      "epoch": 0.22438446569083678,
+      "grad_norm": 0.2601318955421448,
+      "learning_rate": 0.001,
+      "loss": 3.1128,
+      "step": 5304
+    },
+    {
+      "epoch": 0.22442677045435316,
+      "grad_norm": 0.1929735243320465,
+      "learning_rate": 0.001,
+      "loss": 1.8997,
+      "step": 5305
+    },
+    {
+      "epoch": 0.22446907521786955,
+      "grad_norm": 0.26424339413642883,
+      "learning_rate": 0.001,
+      "loss": 2.3221,
+      "step": 5306
+    },
+    {
+      "epoch": 0.2245113799813859,
+      "grad_norm": 0.396243155002594,
+      "learning_rate": 0.001,
+      "loss": 2.9599,
+      "step": 5307
+    },
+    {
+      "epoch": 0.22455368474490228,
+      "grad_norm": 1.7157014608383179,
+      "learning_rate": 0.001,
+      "loss": 2.1441,
+      "step": 5308
+    },
+    {
+      "epoch": 0.22459598950841864,
+      "grad_norm": 0.33265525102615356,
+      "learning_rate": 0.001,
+      "loss": 1.9819,
+      "step": 5309
+    },
+    {
+      "epoch": 0.22463829427193502,
+      "grad_norm": 0.6093931198120117,
+      "learning_rate": 0.001,
+      "loss": 2.9678,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2246805990354514,
+      "grad_norm": 0.2151605635881424,
+      "learning_rate": 0.001,
+      "loss": 2.5349,
+      "step": 5311
+    },
+    {
+      "epoch": 0.22472290379896775,
+      "grad_norm": 8.028112411499023,
+      "learning_rate": 0.001,
+      "loss": 2.6196,
+      "step": 5312
+    },
+    {
+      "epoch": 0.22476520856248414,
+      "grad_norm": 1.0721027851104736,
+      "learning_rate": 0.001,
+      "loss": 3.4655,
+      "step": 5313
+    },
+    {
+      "epoch": 0.22480751332600052,
+      "grad_norm": 0.20032766461372375,
+      "learning_rate": 0.001,
+      "loss": 2.1955,
+      "step": 5314
+    },
+    {
+      "epoch": 0.22484981808951687,
+      "grad_norm": 0.20679010450839996,
+      "learning_rate": 0.001,
+      "loss": 1.743,
+      "step": 5315
+    },
+    {
+      "epoch": 0.22489212285303326,
+      "grad_norm": 0.23269832134246826,
+      "learning_rate": 0.001,
+      "loss": 1.9682,
+      "step": 5316
+    },
+    {
+      "epoch": 0.22493442761654964,
+      "grad_norm": 0.2920213043689728,
+      "learning_rate": 0.001,
+      "loss": 2.6719,
+      "step": 5317
+    },
+    {
+      "epoch": 0.224976732380066,
+      "grad_norm": 0.31730595231056213,
+      "learning_rate": 0.001,
+      "loss": 2.7025,
+      "step": 5318
+    },
+    {
+      "epoch": 0.22501903714358237,
+      "grad_norm": 0.20346704125404358,
+      "learning_rate": 0.001,
+      "loss": 2.528,
+      "step": 5319
+    },
+    {
+      "epoch": 0.22506134190709873,
+      "grad_norm": 0.9651232957839966,
+      "learning_rate": 0.001,
+      "loss": 2.3393,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2251036466706151,
+      "grad_norm": 3.8393514156341553,
+      "learning_rate": 0.001,
+      "loss": 1.8216,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2251459514341315,
+      "grad_norm": 0.1834096908569336,
+      "learning_rate": 0.001,
+      "loss": 2.0548,
+      "step": 5322
+    },
+    {
+      "epoch": 0.22518825619764785,
+      "grad_norm": 0.4497200846672058,
+      "learning_rate": 0.001,
+      "loss": 2.9522,
+      "step": 5323
+    },
+    {
+      "epoch": 0.22523056096116423,
+      "grad_norm": 0.21612843871116638,
+      "learning_rate": 0.001,
+      "loss": 2.9417,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2252728657246806,
+      "grad_norm": 0.23549750447273254,
+      "learning_rate": 0.001,
+      "loss": 1.8287,
+      "step": 5325
+    },
+    {
+      "epoch": 0.22531517048819696,
+      "grad_norm": 0.25662145018577576,
+      "learning_rate": 0.001,
+      "loss": 3.0902,
+      "step": 5326
+    },
+    {
+      "epoch": 0.22535747525171335,
+      "grad_norm": 0.24648882448673248,
+      "learning_rate": 0.001,
+      "loss": 2.5687,
+      "step": 5327
+    },
+    {
+      "epoch": 0.22539978001522973,
+      "grad_norm": 0.23536236584186554,
+      "learning_rate": 0.001,
+      "loss": 2.0515,
+      "step": 5328
+    },
+    {
+      "epoch": 0.22544208477874608,
+      "grad_norm": 0.25583964586257935,
+      "learning_rate": 0.001,
+      "loss": 2.2743,
+      "step": 5329
+    },
+    {
+      "epoch": 0.22548438954226246,
+      "grad_norm": 6.257376194000244,
+      "learning_rate": 0.001,
+      "loss": 1.6795,
+      "step": 5330
+    },
+    {
+      "epoch": 0.22552669430577882,
+      "grad_norm": 1.3603640794754028,
+      "learning_rate": 0.001,
+      "loss": 3.1865,
+      "step": 5331
+    },
+    {
+      "epoch": 0.2255689990692952,
+      "grad_norm": 0.2052699625492096,
+      "learning_rate": 0.001,
+      "loss": 2.1218,
+      "step": 5332
+    },
+    {
+      "epoch": 0.22561130383281158,
+      "grad_norm": 0.2767829895019531,
+      "learning_rate": 0.001,
+      "loss": 2.3987,
+      "step": 5333
+    },
+    {
+      "epoch": 0.22565360859632794,
+      "grad_norm": 0.22107641398906708,
+      "learning_rate": 0.001,
+      "loss": 1.91,
+      "step": 5334
+    },
+    {
+      "epoch": 0.22569591335984432,
+      "grad_norm": 0.5406429767608643,
+      "learning_rate": 0.001,
+      "loss": 2.1801,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2257382181233607,
+      "grad_norm": 0.3691091239452362,
+      "learning_rate": 0.001,
+      "loss": 2.8873,
+      "step": 5336
+    },
+    {
+      "epoch": 0.22578052288687706,
+      "grad_norm": 0.25709986686706543,
+      "learning_rate": 0.001,
+      "loss": 3.6867,
+      "step": 5337
+    },
+    {
+      "epoch": 0.22582282765039344,
+      "grad_norm": 1.0610352754592896,
+      "learning_rate": 0.001,
+      "loss": 2.0219,
+      "step": 5338
+    },
+    {
+      "epoch": 0.22586513241390982,
+      "grad_norm": 0.2490183711051941,
+      "learning_rate": 0.001,
+      "loss": 3.2717,
+      "step": 5339
+    },
+    {
+      "epoch": 0.22590743717742617,
+      "grad_norm": 0.3832576870918274,
+      "learning_rate": 0.001,
+      "loss": 2.6887,
+      "step": 5340
+    },
+    {
+      "epoch": 0.22594974194094256,
+      "grad_norm": 0.20792420208454132,
+      "learning_rate": 0.001,
+      "loss": 1.4877,
+      "step": 5341
+    },
+    {
+      "epoch": 0.2259920467044589,
+      "grad_norm": 0.2878759503364563,
+      "learning_rate": 0.001,
+      "loss": 2.5409,
+      "step": 5342
+    },
+    {
+      "epoch": 0.2260343514679753,
+      "grad_norm": 0.20281533896923065,
+      "learning_rate": 0.001,
+      "loss": 1.8222,
+      "step": 5343
+    },
+    {
+      "epoch": 0.22607665623149167,
+      "grad_norm": 0.3520990014076233,
+      "learning_rate": 0.001,
+      "loss": 2.7768,
+      "step": 5344
+    },
+    {
+      "epoch": 0.22611896099500803,
+      "grad_norm": 0.25387975573539734,
+      "learning_rate": 0.001,
+      "loss": 2.5626,
+      "step": 5345
+    },
+    {
+      "epoch": 0.2261612657585244,
+      "grad_norm": 0.2082633674144745,
+      "learning_rate": 0.001,
+      "loss": 1.8055,
+      "step": 5346
+    },
+    {
+      "epoch": 0.2262035705220408,
+      "grad_norm": 0.24384428560733795,
+      "learning_rate": 0.001,
+      "loss": 3.0511,
+      "step": 5347
+    },
+    {
+      "epoch": 0.22624587528555715,
+      "grad_norm": 0.2108755111694336,
+      "learning_rate": 0.001,
+      "loss": 2.4456,
+      "step": 5348
+    },
+    {
+      "epoch": 0.22628818004907353,
+      "grad_norm": 0.23625120520591736,
+      "learning_rate": 0.001,
+      "loss": 2.0264,
+      "step": 5349
+    },
+    {
+      "epoch": 0.2263304848125899,
+      "grad_norm": 0.19359584152698517,
+      "learning_rate": 0.001,
+      "loss": 2.1669,
+      "step": 5350
+    },
+    {
+      "epoch": 0.22637278957610626,
+      "grad_norm": 0.23231805860996246,
+      "learning_rate": 0.001,
+      "loss": 2.3967,
+      "step": 5351
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.2224705070257187,
+      "learning_rate": 0.001,
+      "loss": 3.0155,
+      "step": 5352
+    },
+    {
+      "epoch": 0.226457399103139,
+      "grad_norm": 0.18623998761177063,
+      "learning_rate": 0.001,
+      "loss": 2.7007,
+      "step": 5353
+    },
+    {
+      "epoch": 0.22649970386665538,
+      "grad_norm": 0.190797358751297,
+      "learning_rate": 0.001,
+      "loss": 1.8014,
+      "step": 5354
+    },
+    {
+      "epoch": 0.22654200863017177,
+      "grad_norm": 0.2000328004360199,
+      "learning_rate": 0.001,
+      "loss": 2.1544,
+      "step": 5355
+    },
+    {
+      "epoch": 0.22658431339368812,
+      "grad_norm": 0.26492246985435486,
+      "learning_rate": 0.001,
+      "loss": 2.5089,
+      "step": 5356
+    },
+    {
+      "epoch": 0.2266266181572045,
+      "grad_norm": 0.22106075286865234,
+      "learning_rate": 0.001,
+      "loss": 2.0971,
+      "step": 5357
+    },
+    {
+      "epoch": 0.22666892292072088,
+      "grad_norm": 0.20364297926425934,
+      "learning_rate": 0.001,
+      "loss": 2.6095,
+      "step": 5358
+    },
+    {
+      "epoch": 0.22671122768423724,
+      "grad_norm": 0.30333849787712097,
+      "learning_rate": 0.001,
+      "loss": 1.6666,
+      "step": 5359
+    },
+    {
+      "epoch": 0.22675353244775362,
+      "grad_norm": 0.20348311960697174,
+      "learning_rate": 0.001,
+      "loss": 1.8586,
+      "step": 5360
+    },
+    {
+      "epoch": 0.22679583721127,
+      "grad_norm": 0.265525221824646,
+      "learning_rate": 0.001,
+      "loss": 2.3425,
+      "step": 5361
+    },
+    {
+      "epoch": 0.22683814197478636,
+      "grad_norm": 0.2070625275373459,
+      "learning_rate": 0.001,
+      "loss": 1.855,
+      "step": 5362
+    },
+    {
+      "epoch": 0.22688044673830274,
+      "grad_norm": 0.24439737200737,
+      "learning_rate": 0.001,
+      "loss": 2.8247,
+      "step": 5363
+    },
+    {
+      "epoch": 0.2269227515018191,
+      "grad_norm": 0.18254782259464264,
+      "learning_rate": 0.001,
+      "loss": 2.1448,
+      "step": 5364
+    },
+    {
+      "epoch": 0.22696505626533547,
+      "grad_norm": 0.2069178968667984,
+      "learning_rate": 0.001,
+      "loss": 2.2075,
+      "step": 5365
+    },
+    {
+      "epoch": 0.22700736102885186,
+      "grad_norm": 0.2451086789369583,
+      "learning_rate": 0.001,
+      "loss": 2.7744,
+      "step": 5366
+    },
+    {
+      "epoch": 0.2270496657923682,
+      "grad_norm": 0.2338859587907791,
+      "learning_rate": 0.001,
+      "loss": 2.2202,
+      "step": 5367
+    },
+    {
+      "epoch": 0.2270919705558846,
+      "grad_norm": 0.17298074066638947,
+      "learning_rate": 0.001,
+      "loss": 2.8615,
+      "step": 5368
+    },
+    {
+      "epoch": 0.22713427531940097,
+      "grad_norm": 0.20199984312057495,
+      "learning_rate": 0.001,
+      "loss": 2.1147,
+      "step": 5369
+    },
+    {
+      "epoch": 0.22717658008291733,
+      "grad_norm": 0.18716251850128174,
+      "learning_rate": 0.001,
+      "loss": 2.4151,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2272188848464337,
+      "grad_norm": 0.22359853982925415,
+      "learning_rate": 0.001,
+      "loss": 2.2439,
+      "step": 5371
+    },
+    {
+      "epoch": 0.2272611896099501,
+      "grad_norm": 0.1923590749502182,
+      "learning_rate": 0.001,
+      "loss": 1.878,
+      "step": 5372
+    },
+    {
+      "epoch": 0.22730349437346645,
+      "grad_norm": 0.20053477585315704,
+      "learning_rate": 0.001,
+      "loss": 2.8747,
+      "step": 5373
+    },
+    {
+      "epoch": 0.22734579913698283,
+      "grad_norm": 0.20686081051826477,
+      "learning_rate": 0.001,
+      "loss": 1.8966,
+      "step": 5374
+    },
+    {
+      "epoch": 0.22738810390049918,
+      "grad_norm": 0.2037486433982849,
+      "learning_rate": 0.001,
+      "loss": 2.1247,
+      "step": 5375
+    },
+    {
+      "epoch": 0.22743040866401557,
+      "grad_norm": 0.2048976719379425,
+      "learning_rate": 0.001,
+      "loss": 2.5233,
+      "step": 5376
+    },
+    {
+      "epoch": 0.22747271342753195,
+      "grad_norm": 0.22234363853931427,
+      "learning_rate": 0.001,
+      "loss": 3.5596,
+      "step": 5377
+    },
+    {
+      "epoch": 0.2275150181910483,
+      "grad_norm": 0.17984293401241302,
+      "learning_rate": 0.001,
+      "loss": 1.7908,
+      "step": 5378
+    },
+    {
+      "epoch": 0.22755732295456468,
+      "grad_norm": 0.19727393984794617,
+      "learning_rate": 0.001,
+      "loss": 1.9901,
+      "step": 5379
+    },
+    {
+      "epoch": 0.22759962771808107,
+      "grad_norm": 0.6455583572387695,
+      "learning_rate": 0.001,
+      "loss": 2.2072,
+      "step": 5380
+    },
+    {
+      "epoch": 0.22764193248159742,
+      "grad_norm": 0.1867457777261734,
+      "learning_rate": 0.001,
+      "loss": 2.5531,
+      "step": 5381
+    },
+    {
+      "epoch": 0.2276842372451138,
+      "grad_norm": 0.23731602728366852,
+      "learning_rate": 0.001,
+      "loss": 2.3504,
+      "step": 5382
+    },
+    {
+      "epoch": 0.22772654200863018,
+      "grad_norm": 0.1865188479423523,
+      "learning_rate": 0.001,
+      "loss": 2.1391,
+      "step": 5383
+    },
+    {
+      "epoch": 0.22776884677214654,
+      "grad_norm": 0.26208534836769104,
+      "learning_rate": 0.001,
+      "loss": 2.0536,
+      "step": 5384
+    },
+    {
+      "epoch": 0.22781115153566292,
+      "grad_norm": 0.16341978311538696,
+      "learning_rate": 0.001,
+      "loss": 2.8151,
+      "step": 5385
+    },
+    {
+      "epoch": 0.22785345629917927,
+      "grad_norm": 0.2644388675689697,
+      "learning_rate": 0.001,
+      "loss": 2.5974,
+      "step": 5386
+    },
+    {
+      "epoch": 0.22789576106269566,
+      "grad_norm": 0.18066078424453735,
+      "learning_rate": 0.001,
+      "loss": 1.6,
+      "step": 5387
+    },
+    {
+      "epoch": 0.22793806582621204,
+      "grad_norm": 0.34033629298210144,
+      "learning_rate": 0.001,
+      "loss": 1.4947,
+      "step": 5388
+    },
+    {
+      "epoch": 0.2279803705897284,
+      "grad_norm": 0.19720013439655304,
+      "learning_rate": 0.001,
+      "loss": 1.9336,
+      "step": 5389
+    },
+    {
+      "epoch": 0.22802267535324477,
+      "grad_norm": 0.20138509571552277,
+      "learning_rate": 0.001,
+      "loss": 2.4402,
+      "step": 5390
+    },
+    {
+      "epoch": 0.22806498011676116,
+      "grad_norm": 0.163251593708992,
+      "learning_rate": 0.001,
+      "loss": 1.7993,
+      "step": 5391
+    },
+    {
+      "epoch": 0.2281072848802775,
+      "grad_norm": 0.20354165136814117,
+      "learning_rate": 0.001,
+      "loss": 2.5252,
+      "step": 5392
+    },
+    {
+      "epoch": 0.2281495896437939,
+      "grad_norm": 6.216144561767578,
+      "learning_rate": 0.001,
+      "loss": 2.3808,
+      "step": 5393
+    },
+    {
+      "epoch": 0.22819189440731028,
+      "grad_norm": 0.21125631034374237,
+      "learning_rate": 0.001,
+      "loss": 2.29,
+      "step": 5394
+    },
+    {
+      "epoch": 0.22823419917082663,
+      "grad_norm": 0.19609850645065308,
+      "learning_rate": 0.001,
+      "loss": 1.9172,
+      "step": 5395
+    },
+    {
+      "epoch": 0.228276503934343,
+      "grad_norm": 0.22020597755908966,
+      "learning_rate": 0.001,
+      "loss": 3.1209,
+      "step": 5396
+    },
+    {
+      "epoch": 0.22831880869785937,
+      "grad_norm": 0.21470773220062256,
+      "learning_rate": 0.001,
+      "loss": 2.4758,
+      "step": 5397
+    },
+    {
+      "epoch": 0.22836111346137575,
+      "grad_norm": 0.24168118834495544,
+      "learning_rate": 0.001,
+      "loss": 2.8874,
+      "step": 5398
+    },
+    {
+      "epoch": 0.22840341822489213,
+      "grad_norm": 0.2095698118209839,
+      "learning_rate": 0.001,
+      "loss": 2.1628,
+      "step": 5399
+    },
+    {
+      "epoch": 0.22844572298840848,
+      "grad_norm": 0.2221812754869461,
+      "learning_rate": 0.001,
+      "loss": 2.1466,
+      "step": 5400
+    },
+    {
+      "epoch": 0.22848802775192487,
+      "grad_norm": 0.17952899634838104,
+      "learning_rate": 0.001,
+      "loss": 2.3154,
+      "step": 5401
+    },
+    {
+      "epoch": 0.22853033251544125,
+      "grad_norm": 0.1853719800710678,
+      "learning_rate": 0.001,
+      "loss": 1.7711,
+      "step": 5402
+    },
+    {
+      "epoch": 0.2285726372789576,
+      "grad_norm": 0.2924804985523224,
+      "learning_rate": 0.001,
+      "loss": 1.6892,
+      "step": 5403
+    },
+    {
+      "epoch": 0.22861494204247398,
+      "grad_norm": 0.20289921760559082,
+      "learning_rate": 0.001,
+      "loss": 2.1891,
+      "step": 5404
+    },
+    {
+      "epoch": 0.22865724680599037,
+      "grad_norm": 0.2075899988412857,
+      "learning_rate": 0.001,
+      "loss": 2.1022,
+      "step": 5405
+    },
+    {
+      "epoch": 0.22869955156950672,
+      "grad_norm": 0.19035397469997406,
+      "learning_rate": 0.001,
+      "loss": 1.5325,
+      "step": 5406
+    },
+    {
+      "epoch": 0.2287418563330231,
+      "grad_norm": 9.618221282958984,
+      "learning_rate": 0.001,
+      "loss": 2.1168,
+      "step": 5407
+    },
+    {
+      "epoch": 0.22878416109653946,
+      "grad_norm": 0.20112530887126923,
+      "learning_rate": 0.001,
+      "loss": 2.3236,
+      "step": 5408
+    },
+    {
+      "epoch": 0.22882646586005584,
+      "grad_norm": 0.24566976726055145,
+      "learning_rate": 0.001,
+      "loss": 2.5824,
+      "step": 5409
+    },
+    {
+      "epoch": 0.22886877062357222,
+      "grad_norm": 0.42258378863334656,
+      "learning_rate": 0.001,
+      "loss": 2.3928,
+      "step": 5410
+    },
+    {
+      "epoch": 0.22891107538708858,
+      "grad_norm": 0.17424896359443665,
+      "learning_rate": 0.001,
+      "loss": 2.8459,
+      "step": 5411
+    },
+    {
+      "epoch": 0.22895338015060496,
+      "grad_norm": 0.23018798232078552,
+      "learning_rate": 0.001,
+      "loss": 2.3751,
+      "step": 5412
+    },
+    {
+      "epoch": 0.22899568491412134,
+      "grad_norm": 16.62144660949707,
+      "learning_rate": 0.001,
+      "loss": 1.8728,
+      "step": 5413
+    },
+    {
+      "epoch": 0.2290379896776377,
+      "grad_norm": 0.19507117569446564,
+      "learning_rate": 0.001,
+      "loss": 2.0108,
+      "step": 5414
+    },
+    {
+      "epoch": 0.22908029444115408,
+      "grad_norm": 0.18236559629440308,
+      "learning_rate": 0.001,
+      "loss": 1.877,
+      "step": 5415
+    },
+    {
+      "epoch": 0.22912259920467046,
+      "grad_norm": 0.18296857178211212,
+      "learning_rate": 0.001,
+      "loss": 2.3096,
+      "step": 5416
+    },
+    {
+      "epoch": 0.2291649039681868,
+      "grad_norm": 0.22751617431640625,
+      "learning_rate": 0.001,
+      "loss": 2.0818,
+      "step": 5417
+    },
+    {
+      "epoch": 0.2292072087317032,
+      "grad_norm": 0.22728577256202698,
+      "learning_rate": 0.001,
+      "loss": 2.0094,
+      "step": 5418
+    },
+    {
+      "epoch": 0.22924951349521955,
+      "grad_norm": 0.20583270490169525,
+      "learning_rate": 0.001,
+      "loss": 1.8739,
+      "step": 5419
+    },
+    {
+      "epoch": 0.22929181825873593,
+      "grad_norm": 0.2076723724603653,
+      "learning_rate": 0.001,
+      "loss": 1.8793,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2293341230222523,
+      "grad_norm": 0.23529089987277985,
+      "learning_rate": 0.001,
+      "loss": 1.4192,
+      "step": 5421
+    },
+    {
+      "epoch": 0.22937642778576867,
+      "grad_norm": 1.0901966094970703,
+      "learning_rate": 0.001,
+      "loss": 2.0787,
+      "step": 5422
+    },
+    {
+      "epoch": 0.22941873254928505,
+      "grad_norm": 0.21631963551044464,
+      "learning_rate": 0.001,
+      "loss": 2.0501,
+      "step": 5423
+    },
+    {
+      "epoch": 0.22946103731280143,
+      "grad_norm": 0.2734993100166321,
+      "learning_rate": 0.001,
+      "loss": 3.2331,
+      "step": 5424
+    },
+    {
+      "epoch": 0.22950334207631778,
+      "grad_norm": 0.23572872579097748,
+      "learning_rate": 0.001,
+      "loss": 2.8713,
+      "step": 5425
+    },
+    {
+      "epoch": 0.22954564683983417,
+      "grad_norm": 0.19519785046577454,
+      "learning_rate": 0.001,
+      "loss": 2.4311,
+      "step": 5426
+    },
+    {
+      "epoch": 0.22958795160335055,
+      "grad_norm": 0.19540639221668243,
+      "learning_rate": 0.001,
+      "loss": 2.0317,
+      "step": 5427
+    },
+    {
+      "epoch": 0.2296302563668669,
+      "grad_norm": 0.28334343433380127,
+      "learning_rate": 0.001,
+      "loss": 3.7543,
+      "step": 5428
+    },
+    {
+      "epoch": 0.22967256113038328,
+      "grad_norm": 0.18688379228115082,
+      "learning_rate": 0.001,
+      "loss": 1.9003,
+      "step": 5429
+    },
+    {
+      "epoch": 0.22971486589389967,
+      "grad_norm": 1.3220264911651611,
+      "learning_rate": 0.001,
+      "loss": 1.7229,
+      "step": 5430
+    },
+    {
+      "epoch": 0.22975717065741602,
+      "grad_norm": 0.23925921320915222,
+      "learning_rate": 0.001,
+      "loss": 2.1974,
+      "step": 5431
+    },
+    {
+      "epoch": 0.2297994754209324,
+      "grad_norm": 0.25730374455451965,
+      "learning_rate": 0.001,
+      "loss": 3.2443,
+      "step": 5432
+    },
+    {
+      "epoch": 0.22984178018444876,
+      "grad_norm": 1.0473192930221558,
+      "learning_rate": 0.001,
+      "loss": 2.2703,
+      "step": 5433
+    },
+    {
+      "epoch": 0.22988408494796514,
+      "grad_norm": 0.20525427162647247,
+      "learning_rate": 0.001,
+      "loss": 2.2747,
+      "step": 5434
+    },
+    {
+      "epoch": 0.22992638971148152,
+      "grad_norm": 0.24831177294254303,
+      "learning_rate": 0.001,
+      "loss": 2.3047,
+      "step": 5435
+    },
+    {
+      "epoch": 0.22996869447499788,
+      "grad_norm": 0.16241610050201416,
+      "learning_rate": 0.001,
+      "loss": 1.5626,
+      "step": 5436
+    },
+    {
+      "epoch": 0.23001099923851426,
+      "grad_norm": 0.23656898736953735,
+      "learning_rate": 0.001,
+      "loss": 1.8401,
+      "step": 5437
+    },
+    {
+      "epoch": 0.23005330400203064,
+      "grad_norm": 0.21393883228302002,
+      "learning_rate": 0.001,
+      "loss": 2.2871,
+      "step": 5438
+    },
+    {
+      "epoch": 0.230095608765547,
+      "grad_norm": 0.471842497587204,
+      "learning_rate": 0.001,
+      "loss": 2.9926,
+      "step": 5439
+    },
+    {
+      "epoch": 0.23013791352906338,
+      "grad_norm": 0.20878411829471588,
+      "learning_rate": 0.001,
+      "loss": 1.7661,
+      "step": 5440
+    },
+    {
+      "epoch": 0.23018021829257976,
+      "grad_norm": 0.21290957927703857,
+      "learning_rate": 0.001,
+      "loss": 2.2779,
+      "step": 5441
+    },
+    {
+      "epoch": 0.2302225230560961,
+      "grad_norm": 0.2180158644914627,
+      "learning_rate": 0.001,
+      "loss": 2.9909,
+      "step": 5442
+    },
+    {
+      "epoch": 0.2302648278196125,
+      "grad_norm": 0.22294098138809204,
+      "learning_rate": 0.001,
+      "loss": 2.7471,
+      "step": 5443
+    },
+    {
+      "epoch": 0.23030713258312885,
+      "grad_norm": 0.20272763073444366,
+      "learning_rate": 0.001,
+      "loss": 2.2887,
+      "step": 5444
+    },
+    {
+      "epoch": 0.23034943734664523,
+      "grad_norm": 0.5208730697631836,
+      "learning_rate": 0.001,
+      "loss": 1.786,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2303917421101616,
+      "grad_norm": 0.23959921300411224,
+      "learning_rate": 0.001,
+      "loss": 2.5374,
+      "step": 5446
+    },
+    {
+      "epoch": 0.23043404687367797,
+      "grad_norm": 0.22310377657413483,
+      "learning_rate": 0.001,
+      "loss": 2.9862,
+      "step": 5447
+    },
+    {
+      "epoch": 0.23047635163719435,
+      "grad_norm": 0.27970150113105774,
+      "learning_rate": 0.001,
+      "loss": 2.2442,
+      "step": 5448
+    },
+    {
+      "epoch": 0.23051865640071073,
+      "grad_norm": 0.230902299284935,
+      "learning_rate": 0.001,
+      "loss": 1.6546,
+      "step": 5449
+    },
+    {
+      "epoch": 0.23056096116422709,
+      "grad_norm": 0.22653928399085999,
+      "learning_rate": 0.001,
+      "loss": 2.9463,
+      "step": 5450
+    },
+    {
+      "epoch": 0.23060326592774347,
+      "grad_norm": 0.25787416100502014,
+      "learning_rate": 0.001,
+      "loss": 2.4843,
+      "step": 5451
+    },
+    {
+      "epoch": 0.23064557069125985,
+      "grad_norm": 0.44561514258384705,
+      "learning_rate": 0.001,
+      "loss": 2.011,
+      "step": 5452
+    },
+    {
+      "epoch": 0.2306878754547762,
+      "grad_norm": 0.27553293108940125,
+      "learning_rate": 0.001,
+      "loss": 2.7563,
+      "step": 5453
+    },
+    {
+      "epoch": 0.23073018021829259,
+      "grad_norm": 0.2613121569156647,
+      "learning_rate": 0.001,
+      "loss": 2.2846,
+      "step": 5454
+    },
+    {
+      "epoch": 0.23077248498180894,
+      "grad_norm": 0.36091411113739014,
+      "learning_rate": 0.001,
+      "loss": 3.0635,
+      "step": 5455
+    },
+    {
+      "epoch": 0.23081478974532532,
+      "grad_norm": 0.19998879730701447,
+      "learning_rate": 0.001,
+      "loss": 2.578,
+      "step": 5456
+    },
+    {
+      "epoch": 0.2308570945088417,
+      "grad_norm": 0.235711470246315,
+      "learning_rate": 0.001,
+      "loss": 2.1896,
+      "step": 5457
+    },
+    {
+      "epoch": 0.23089939927235806,
+      "grad_norm": 0.7797324061393738,
+      "learning_rate": 0.001,
+      "loss": 2.4483,
+      "step": 5458
+    },
+    {
+      "epoch": 0.23094170403587444,
+      "grad_norm": 0.17661023139953613,
+      "learning_rate": 0.001,
+      "loss": 2.4288,
+      "step": 5459
+    },
+    {
+      "epoch": 0.23098400879939082,
+      "grad_norm": 0.18640023469924927,
+      "learning_rate": 0.001,
+      "loss": 2.1796,
+      "step": 5460
+    },
+    {
+      "epoch": 0.23102631356290718,
+      "grad_norm": 7.270983695983887,
+      "learning_rate": 0.001,
+      "loss": 2.1537,
+      "step": 5461
+    },
+    {
+      "epoch": 0.23106861832642356,
+      "grad_norm": 0.5781380534172058,
+      "learning_rate": 0.001,
+      "loss": 2.9149,
+      "step": 5462
+    },
+    {
+      "epoch": 0.23111092308993994,
+      "grad_norm": 0.21426080167293549,
+      "learning_rate": 0.001,
+      "loss": 2.6075,
+      "step": 5463
+    },
+    {
+      "epoch": 0.2311532278534563,
+      "grad_norm": 0.24475079774856567,
+      "learning_rate": 0.001,
+      "loss": 2.4721,
+      "step": 5464
+    },
+    {
+      "epoch": 0.23119553261697268,
+      "grad_norm": 1.0542508363723755,
+      "learning_rate": 0.001,
+      "loss": 2.0507,
+      "step": 5465
+    },
+    {
+      "epoch": 0.23123783738048903,
+      "grad_norm": 0.2142096906900406,
+      "learning_rate": 0.001,
+      "loss": 1.8177,
+      "step": 5466
+    },
+    {
+      "epoch": 0.2312801421440054,
+      "grad_norm": 0.25511589646339417,
+      "learning_rate": 0.001,
+      "loss": 2.9547,
+      "step": 5467
+    },
+    {
+      "epoch": 0.2313224469075218,
+      "grad_norm": 4.037811279296875,
+      "learning_rate": 0.001,
+      "loss": 2.0729,
+      "step": 5468
+    },
+    {
+      "epoch": 0.23136475167103815,
+      "grad_norm": 0.42371609807014465,
+      "learning_rate": 0.001,
+      "loss": 2.684,
+      "step": 5469
+    },
+    {
+      "epoch": 0.23140705643455453,
+      "grad_norm": 0.25769296288490295,
+      "learning_rate": 0.001,
+      "loss": 1.8104,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2314493611980709,
+      "grad_norm": 0.9849298596382141,
+      "learning_rate": 0.001,
+      "loss": 3.1677,
+      "step": 5471
+    },
+    {
+      "epoch": 0.23149166596158727,
+      "grad_norm": 10.10745620727539,
+      "learning_rate": 0.001,
+      "loss": 2.3539,
+      "step": 5472
+    },
+    {
+      "epoch": 0.23153397072510365,
+      "grad_norm": 0.28518936038017273,
+      "learning_rate": 0.001,
+      "loss": 3.5426,
+      "step": 5473
+    },
+    {
+      "epoch": 0.23157627548862003,
+      "grad_norm": 0.40552154183387756,
+      "learning_rate": 0.001,
+      "loss": 2.9542,
+      "step": 5474
+    },
+    {
+      "epoch": 0.23161858025213639,
+      "grad_norm": 0.26154863834381104,
+      "learning_rate": 0.001,
+      "loss": 1.9963,
+      "step": 5475
+    },
+    {
+      "epoch": 0.23166088501565277,
+      "grad_norm": 0.2571628987789154,
+      "learning_rate": 0.001,
+      "loss": 1.95,
+      "step": 5476
+    },
+    {
+      "epoch": 0.23170318977916912,
+      "grad_norm": 0.2619466185569763,
+      "learning_rate": 0.001,
+      "loss": 2.0637,
+      "step": 5477
+    },
+    {
+      "epoch": 0.2317454945426855,
+      "grad_norm": 0.28519323468208313,
+      "learning_rate": 0.001,
+      "loss": 2.5762,
+      "step": 5478
+    },
+    {
+      "epoch": 0.23178779930620189,
+      "grad_norm": 1.2424720525741577,
+      "learning_rate": 0.001,
+      "loss": 3.0967,
+      "step": 5479
+    },
+    {
+      "epoch": 0.23183010406971824,
+      "grad_norm": 0.4895488917827606,
+      "learning_rate": 0.001,
+      "loss": 3.3258,
+      "step": 5480
+    },
+    {
+      "epoch": 0.23187240883323462,
+      "grad_norm": 0.39501601457595825,
+      "learning_rate": 0.001,
+      "loss": 2.6346,
+      "step": 5481
+    },
+    {
+      "epoch": 0.231914713596751,
+      "grad_norm": 0.2693521976470947,
+      "learning_rate": 0.001,
+      "loss": 2.1174,
+      "step": 5482
+    },
+    {
+      "epoch": 0.23195701836026736,
+      "grad_norm": 0.2762487828731537,
+      "learning_rate": 0.001,
+      "loss": 2.0984,
+      "step": 5483
+    },
+    {
+      "epoch": 0.23199932312378374,
+      "grad_norm": 0.2225947082042694,
+      "learning_rate": 0.001,
+      "loss": 1.9871,
+      "step": 5484
+    },
+    {
+      "epoch": 0.23204162788730012,
+      "grad_norm": 0.38060519099235535,
+      "learning_rate": 0.001,
+      "loss": 2.3824,
+      "step": 5485
+    },
+    {
+      "epoch": 0.23208393265081648,
+      "grad_norm": 0.21459133923053741,
+      "learning_rate": 0.001,
+      "loss": 2.0364,
+      "step": 5486
+    },
+    {
+      "epoch": 0.23212623741433286,
+      "grad_norm": 0.5198157429695129,
+      "learning_rate": 0.001,
+      "loss": 3.1656,
+      "step": 5487
+    },
+    {
+      "epoch": 0.2321685421778492,
+      "grad_norm": 0.675767183303833,
+      "learning_rate": 0.001,
+      "loss": 3.456,
+      "step": 5488
+    },
+    {
+      "epoch": 0.2322108469413656,
+      "grad_norm": 0.22795476019382477,
+      "learning_rate": 0.001,
+      "loss": 3.3561,
+      "step": 5489
+    },
+    {
+      "epoch": 0.23225315170488198,
+      "grad_norm": 0.9936857223510742,
+      "learning_rate": 0.001,
+      "loss": 2.4905,
+      "step": 5490
+    },
+    {
+      "epoch": 0.23229545646839833,
+      "grad_norm": 0.3095644414424896,
+      "learning_rate": 0.001,
+      "loss": 2.0372,
+      "step": 5491
+    },
+    {
+      "epoch": 0.2323377612319147,
+      "grad_norm": 0.18799656629562378,
+      "learning_rate": 0.001,
+      "loss": 1.9836,
+      "step": 5492
+    },
+    {
+      "epoch": 0.2323800659954311,
+      "grad_norm": 0.21074536442756653,
+      "learning_rate": 0.001,
+      "loss": 2.2601,
+      "step": 5493
+    },
+    {
+      "epoch": 0.23242237075894745,
+      "grad_norm": 0.32169005274772644,
+      "learning_rate": 0.001,
+      "loss": 3.0188,
+      "step": 5494
+    },
+    {
+      "epoch": 0.23246467552246383,
+      "grad_norm": 0.21272505819797516,
+      "learning_rate": 0.001,
+      "loss": 2.1103,
+      "step": 5495
+    },
+    {
+      "epoch": 0.2325069802859802,
+      "grad_norm": 0.4344709515571594,
+      "learning_rate": 0.001,
+      "loss": 3.0217,
+      "step": 5496
+    },
+    {
+      "epoch": 0.23254928504949657,
+      "grad_norm": 1.6699187755584717,
+      "learning_rate": 0.001,
+      "loss": 2.5814,
+      "step": 5497
+    },
+    {
+      "epoch": 0.23259158981301295,
+      "grad_norm": 0.22509506344795227,
+      "learning_rate": 0.001,
+      "loss": 2.8708,
+      "step": 5498
+    },
+    {
+      "epoch": 0.2326338945765293,
+      "grad_norm": 0.24898000061511993,
+      "learning_rate": 0.001,
+      "loss": 2.8715,
+      "step": 5499
+    },
+    {
+      "epoch": 0.2326761993400457,
+      "grad_norm": 0.7220833897590637,
+      "learning_rate": 0.001,
+      "loss": 2.1767,
+      "step": 5500
+    },
+    {
+      "epoch": 0.23271850410356207,
+      "grad_norm": 0.25471919775009155,
+      "learning_rate": 0.001,
+      "loss": 3.0134,
+      "step": 5501
+    },
+    {
+      "epoch": 0.23276080886707842,
+      "grad_norm": 0.2646358907222748,
+      "learning_rate": 0.001,
+      "loss": 2.0253,
+      "step": 5502
+    },
+    {
+      "epoch": 0.2328031136305948,
+      "grad_norm": 0.20259220898151398,
+      "learning_rate": 0.001,
+      "loss": 2.053,
+      "step": 5503
+    },
+    {
+      "epoch": 0.2328454183941112,
+      "grad_norm": 0.49306970834732056,
+      "learning_rate": 0.001,
+      "loss": 2.2513,
+      "step": 5504
+    },
+    {
+      "epoch": 0.23288772315762754,
+      "grad_norm": 0.2066371738910675,
+      "learning_rate": 0.001,
+      "loss": 2.7308,
+      "step": 5505
+    },
+    {
+      "epoch": 0.23293002792114392,
+      "grad_norm": 0.20641134679317474,
+      "learning_rate": 0.001,
+      "loss": 2.5732,
+      "step": 5506
+    },
+    {
+      "epoch": 0.2329723326846603,
+      "grad_norm": 0.21827174723148346,
+      "learning_rate": 0.001,
+      "loss": 2.6558,
+      "step": 5507
+    },
+    {
+      "epoch": 0.23301463744817666,
+      "grad_norm": 0.2925966680049896,
+      "learning_rate": 0.001,
+      "loss": 2.6788,
+      "step": 5508
+    },
+    {
+      "epoch": 0.23305694221169304,
+      "grad_norm": 0.2853943109512329,
+      "learning_rate": 0.001,
+      "loss": 2.5224,
+      "step": 5509
+    },
+    {
+      "epoch": 0.2330992469752094,
+      "grad_norm": 0.20848669111728668,
+      "learning_rate": 0.001,
+      "loss": 2.3025,
+      "step": 5510
+    },
+    {
+      "epoch": 0.23314155173872578,
+      "grad_norm": 1.7982761859893799,
+      "learning_rate": 0.001,
+      "loss": 2.1487,
+      "step": 5511
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.2628767490386963,
+      "learning_rate": 0.001,
+      "loss": 2.7454,
+      "step": 5512
+    },
+    {
+      "epoch": 0.2332261612657585,
+      "grad_norm": 0.21487084031105042,
+      "learning_rate": 0.001,
+      "loss": 1.7897,
+      "step": 5513
+    },
+    {
+      "epoch": 0.2332684660292749,
+      "grad_norm": 0.2565120756626129,
+      "learning_rate": 0.001,
+      "loss": 2.3647,
+      "step": 5514
+    },
+    {
+      "epoch": 0.23331077079279128,
+      "grad_norm": 0.2229587584733963,
+      "learning_rate": 0.001,
+      "loss": 1.7196,
+      "step": 5515
+    },
+    {
+      "epoch": 0.23335307555630763,
+      "grad_norm": 0.23122352361679077,
+      "learning_rate": 0.001,
+      "loss": 3.264,
+      "step": 5516
+    },
+    {
+      "epoch": 0.23339538031982401,
+      "grad_norm": 0.2294238656759262,
+      "learning_rate": 0.001,
+      "loss": 2.1468,
+      "step": 5517
+    },
+    {
+      "epoch": 0.2334376850833404,
+      "grad_norm": 1.3923436403274536,
+      "learning_rate": 0.001,
+      "loss": 2.8618,
+      "step": 5518
+    },
+    {
+      "epoch": 0.23347998984685675,
+      "grad_norm": 4.941551685333252,
+      "learning_rate": 0.001,
+      "loss": 2.7653,
+      "step": 5519
+    },
+    {
+      "epoch": 0.23352229461037313,
+      "grad_norm": 0.23686207830905914,
+      "learning_rate": 0.001,
+      "loss": 1.6694,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2335645993738895,
+      "grad_norm": 0.38587716221809387,
+      "learning_rate": 0.001,
+      "loss": 2.2298,
+      "step": 5521
+    },
+    {
+      "epoch": 0.23360690413740587,
+      "grad_norm": 0.2625442147254944,
+      "learning_rate": 0.001,
+      "loss": 2.0646,
+      "step": 5522
+    },
+    {
+      "epoch": 0.23364920890092225,
+      "grad_norm": 0.2582058906555176,
+      "learning_rate": 0.001,
+      "loss": 2.6844,
+      "step": 5523
+    },
+    {
+      "epoch": 0.2336915136644386,
+      "grad_norm": 2.8828814029693604,
+      "learning_rate": 0.001,
+      "loss": 2.1936,
+      "step": 5524
+    },
+    {
+      "epoch": 0.233733818427955,
+      "grad_norm": 0.2290409654378891,
+      "learning_rate": 0.001,
+      "loss": 1.9075,
+      "step": 5525
+    },
+    {
+      "epoch": 0.23377612319147137,
+      "grad_norm": 0.1932823210954666,
+      "learning_rate": 0.001,
+      "loss": 3.3037,
+      "step": 5526
+    },
+    {
+      "epoch": 0.23381842795498772,
+      "grad_norm": 0.20359812676906586,
+      "learning_rate": 0.001,
+      "loss": 2.209,
+      "step": 5527
+    },
+    {
+      "epoch": 0.2338607327185041,
+      "grad_norm": 0.3113921582698822,
+      "learning_rate": 0.001,
+      "loss": 1.766,
+      "step": 5528
+    },
+    {
+      "epoch": 0.2339030374820205,
+      "grad_norm": 20.93682098388672,
+      "learning_rate": 0.001,
+      "loss": 2.5909,
+      "step": 5529
+    },
+    {
+      "epoch": 0.23394534224553684,
+      "grad_norm": 0.2424592673778534,
+      "learning_rate": 0.001,
+      "loss": 2.3092,
+      "step": 5530
+    },
+    {
+      "epoch": 0.23398764700905322,
+      "grad_norm": 0.22494091093540192,
+      "learning_rate": 0.001,
+      "loss": 2.2942,
+      "step": 5531
+    },
+    {
+      "epoch": 0.23402995177256958,
+      "grad_norm": 0.20457018911838531,
+      "learning_rate": 0.001,
+      "loss": 2.31,
+      "step": 5532
+    },
+    {
+      "epoch": 0.23407225653608596,
+      "grad_norm": 0.32877421379089355,
+      "learning_rate": 0.001,
+      "loss": 1.8861,
+      "step": 5533
+    },
+    {
+      "epoch": 0.23411456129960234,
+      "grad_norm": 0.42228415608406067,
+      "learning_rate": 0.001,
+      "loss": 2.7353,
+      "step": 5534
+    },
+    {
+      "epoch": 0.2341568660631187,
+      "grad_norm": 0.47094154357910156,
+      "learning_rate": 0.001,
+      "loss": 3.8558,
+      "step": 5535
+    },
+    {
+      "epoch": 0.23419917082663508,
+      "grad_norm": 0.2175481915473938,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 5536
+    },
+    {
+      "epoch": 0.23424147559015146,
+      "grad_norm": 3.194316864013672,
+      "learning_rate": 0.001,
+      "loss": 2.5488,
+      "step": 5537
+    },
+    {
+      "epoch": 0.23428378035366781,
+      "grad_norm": 0.3026067912578583,
+      "learning_rate": 0.001,
+      "loss": 2.4307,
+      "step": 5538
+    },
+    {
+      "epoch": 0.2343260851171842,
+      "grad_norm": 0.8778800368309021,
+      "learning_rate": 0.001,
+      "loss": 1.6685,
+      "step": 5539
+    },
+    {
+      "epoch": 0.23436838988070058,
+      "grad_norm": 0.1916915625333786,
+      "learning_rate": 0.001,
+      "loss": 2.2532,
+      "step": 5540
+    },
+    {
+      "epoch": 0.23441069464421693,
+      "grad_norm": 0.28568512201309204,
+      "learning_rate": 0.001,
+      "loss": 2.0453,
+      "step": 5541
+    },
+    {
+      "epoch": 0.23445299940773331,
+      "grad_norm": 0.5260047316551208,
+      "learning_rate": 0.001,
+      "loss": 3.2691,
+      "step": 5542
+    },
+    {
+      "epoch": 0.23449530417124967,
+      "grad_norm": 0.2400904893875122,
+      "learning_rate": 0.001,
+      "loss": 2.483,
+      "step": 5543
+    },
+    {
+      "epoch": 0.23453760893476605,
+      "grad_norm": 0.232883483171463,
+      "learning_rate": 0.001,
+      "loss": 3.0007,
+      "step": 5544
+    },
+    {
+      "epoch": 0.23457991369828243,
+      "grad_norm": 0.21904799342155457,
+      "learning_rate": 0.001,
+      "loss": 2.0221,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2346222184617988,
+      "grad_norm": 0.27808424830436707,
+      "learning_rate": 0.001,
+      "loss": 1.7361,
+      "step": 5546
+    },
+    {
+      "epoch": 0.23466452322531517,
+      "grad_norm": 0.33674612641334534,
+      "learning_rate": 0.001,
+      "loss": 2.5554,
+      "step": 5547
+    },
+    {
+      "epoch": 0.23470682798883155,
+      "grad_norm": 0.21071641147136688,
+      "learning_rate": 0.001,
+      "loss": 2.0207,
+      "step": 5548
+    },
+    {
+      "epoch": 0.2347491327523479,
+      "grad_norm": 0.20552270114421844,
+      "learning_rate": 0.001,
+      "loss": 2.2293,
+      "step": 5549
+    },
+    {
+      "epoch": 0.2347914375158643,
+      "grad_norm": 0.31367790699005127,
+      "learning_rate": 0.001,
+      "loss": 2.6712,
+      "step": 5550
+    },
+    {
+      "epoch": 0.23483374227938067,
+      "grad_norm": 0.7364101409912109,
+      "learning_rate": 0.001,
+      "loss": 2.0776,
+      "step": 5551
+    },
+    {
+      "epoch": 0.23487604704289702,
+      "grad_norm": 0.24621638655662537,
+      "learning_rate": 0.001,
+      "loss": 2.0778,
+      "step": 5552
+    },
+    {
+      "epoch": 0.2349183518064134,
+      "grad_norm": 0.23604516685009003,
+      "learning_rate": 0.001,
+      "loss": 3.3079,
+      "step": 5553
+    },
+    {
+      "epoch": 0.2349606565699298,
+      "grad_norm": 0.8118863105773926,
+      "learning_rate": 0.001,
+      "loss": 2.4182,
+      "step": 5554
+    },
+    {
+      "epoch": 0.23500296133344614,
+      "grad_norm": 0.21892663836479187,
+      "learning_rate": 0.001,
+      "loss": 2.2327,
+      "step": 5555
+    },
+    {
+      "epoch": 0.23504526609696252,
+      "grad_norm": 0.28527218103408813,
+      "learning_rate": 0.001,
+      "loss": 3.7358,
+      "step": 5556
+    },
+    {
+      "epoch": 0.23508757086047888,
+      "grad_norm": 0.2317507416009903,
+      "learning_rate": 0.001,
+      "loss": 3.2241,
+      "step": 5557
+    },
+    {
+      "epoch": 0.23512987562399526,
+      "grad_norm": 0.23719051480293274,
+      "learning_rate": 0.001,
+      "loss": 2.1869,
+      "step": 5558
+    },
+    {
+      "epoch": 0.23517218038751164,
+      "grad_norm": 0.2504566013813019,
+      "learning_rate": 0.001,
+      "loss": 2.2632,
+      "step": 5559
+    },
+    {
+      "epoch": 0.235214485151028,
+      "grad_norm": 0.2797103226184845,
+      "learning_rate": 0.001,
+      "loss": 2.0476,
+      "step": 5560
+    },
+    {
+      "epoch": 0.23525678991454438,
+      "grad_norm": 0.223208948969841,
+      "learning_rate": 0.001,
+      "loss": 2.3264,
+      "step": 5561
+    },
+    {
+      "epoch": 0.23529909467806076,
+      "grad_norm": 2.268007755279541,
+      "learning_rate": 0.001,
+      "loss": 3.4196,
+      "step": 5562
+    },
+    {
+      "epoch": 0.23534139944157711,
+      "grad_norm": 0.20520193874835968,
+      "learning_rate": 0.001,
+      "loss": 2.1592,
+      "step": 5563
+    },
+    {
+      "epoch": 0.2353837042050935,
+      "grad_norm": 0.1952827274799347,
+      "learning_rate": 0.001,
+      "loss": 2.2686,
+      "step": 5564
+    },
+    {
+      "epoch": 0.23542600896860988,
+      "grad_norm": 0.24060063064098358,
+      "learning_rate": 0.001,
+      "loss": 2.6335,
+      "step": 5565
+    },
+    {
+      "epoch": 0.23546831373212623,
+      "grad_norm": 0.18970201909542084,
+      "learning_rate": 0.001,
+      "loss": 2.1595,
+      "step": 5566
+    },
+    {
+      "epoch": 0.23551061849564262,
+      "grad_norm": 0.20681913197040558,
+      "learning_rate": 0.001,
+      "loss": 2.7761,
+      "step": 5567
+    },
+    {
+      "epoch": 0.23555292325915897,
+      "grad_norm": 1.009856104850769,
+      "learning_rate": 0.001,
+      "loss": 3.113,
+      "step": 5568
+    },
+    {
+      "epoch": 0.23559522802267535,
+      "grad_norm": 0.20202496647834778,
+      "learning_rate": 0.001,
+      "loss": 2.7176,
+      "step": 5569
+    },
+    {
+      "epoch": 0.23563753278619173,
+      "grad_norm": 0.2296670824289322,
+      "learning_rate": 0.001,
+      "loss": 1.9414,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2356798375497081,
+      "grad_norm": 2.0302107334136963,
+      "learning_rate": 0.001,
+      "loss": 3.5255,
+      "step": 5571
+    },
+    {
+      "epoch": 0.23572214231322447,
+      "grad_norm": 0.2701435983181,
+      "learning_rate": 0.001,
+      "loss": 2.0376,
+      "step": 5572
+    },
+    {
+      "epoch": 0.23576444707674085,
+      "grad_norm": 1.3323795795440674,
+      "learning_rate": 0.001,
+      "loss": 3.0352,
+      "step": 5573
+    },
+    {
+      "epoch": 0.2358067518402572,
+      "grad_norm": 0.2649135887622833,
+      "learning_rate": 0.001,
+      "loss": 2.5785,
+      "step": 5574
+    },
+    {
+      "epoch": 0.2358490566037736,
+      "grad_norm": 0.21373842656612396,
+      "learning_rate": 0.001,
+      "loss": 1.9535,
+      "step": 5575
+    },
+    {
+      "epoch": 0.23589136136728997,
+      "grad_norm": 0.2295471578836441,
+      "learning_rate": 0.001,
+      "loss": 1.9545,
+      "step": 5576
+    },
+    {
+      "epoch": 0.23593366613080632,
+      "grad_norm": 0.32685786485671997,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 5577
+    },
+    {
+      "epoch": 0.2359759708943227,
+      "grad_norm": 0.22259478271007538,
+      "learning_rate": 0.001,
+      "loss": 2.1198,
+      "step": 5578
+    },
+    {
+      "epoch": 0.23601827565783906,
+      "grad_norm": 0.2274249643087387,
+      "learning_rate": 0.001,
+      "loss": 2.2864,
+      "step": 5579
+    },
+    {
+      "epoch": 0.23606058042135544,
+      "grad_norm": 0.23607851564884186,
+      "learning_rate": 0.001,
+      "loss": 3.7908,
+      "step": 5580
+    },
+    {
+      "epoch": 0.23610288518487182,
+      "grad_norm": 0.22319360077381134,
+      "learning_rate": 0.001,
+      "loss": 1.8991,
+      "step": 5581
+    },
+    {
+      "epoch": 0.23614518994838818,
+      "grad_norm": 0.1910727173089981,
+      "learning_rate": 0.001,
+      "loss": 2.1673,
+      "step": 5582
+    },
+    {
+      "epoch": 0.23618749471190456,
+      "grad_norm": 0.35509058833122253,
+      "learning_rate": 0.001,
+      "loss": 1.9815,
+      "step": 5583
+    },
+    {
+      "epoch": 0.23622979947542094,
+      "grad_norm": 0.5298141837120056,
+      "learning_rate": 0.001,
+      "loss": 2.5921,
+      "step": 5584
+    },
+    {
+      "epoch": 0.2362721042389373,
+      "grad_norm": 0.2098468840122223,
+      "learning_rate": 0.001,
+      "loss": 2.8481,
+      "step": 5585
+    },
+    {
+      "epoch": 0.23631440900245368,
+      "grad_norm": 0.6692335605621338,
+      "learning_rate": 0.001,
+      "loss": 2.6247,
+      "step": 5586
+    },
+    {
+      "epoch": 0.23635671376597006,
+      "grad_norm": 0.2933887541294098,
+      "learning_rate": 0.001,
+      "loss": 2.677,
+      "step": 5587
+    },
+    {
+      "epoch": 0.23639901852948642,
+      "grad_norm": 1.0850200653076172,
+      "learning_rate": 0.001,
+      "loss": 2.6977,
+      "step": 5588
+    },
+    {
+      "epoch": 0.2364413232930028,
+      "grad_norm": 8.362639427185059,
+      "learning_rate": 0.001,
+      "loss": 2.4108,
+      "step": 5589
+    },
+    {
+      "epoch": 0.23648362805651915,
+      "grad_norm": 0.1976199746131897,
+      "learning_rate": 0.001,
+      "loss": 2.2232,
+      "step": 5590
+    },
+    {
+      "epoch": 0.23652593282003553,
+      "grad_norm": 1.5596935749053955,
+      "learning_rate": 0.001,
+      "loss": 2.3909,
+      "step": 5591
+    },
+    {
+      "epoch": 0.23656823758355192,
+      "grad_norm": 0.3760131895542145,
+      "learning_rate": 0.001,
+      "loss": 2.1015,
+      "step": 5592
+    },
+    {
+      "epoch": 0.23661054234706827,
+      "grad_norm": 0.4867786169052124,
+      "learning_rate": 0.001,
+      "loss": 2.3708,
+      "step": 5593
+    },
+    {
+      "epoch": 0.23665284711058465,
+      "grad_norm": 0.3288508355617523,
+      "learning_rate": 0.001,
+      "loss": 3.3302,
+      "step": 5594
+    },
+    {
+      "epoch": 0.23669515187410103,
+      "grad_norm": 0.3380516767501831,
+      "learning_rate": 0.001,
+      "loss": 2.3632,
+      "step": 5595
+    },
+    {
+      "epoch": 0.2367374566376174,
+      "grad_norm": 0.2778700590133667,
+      "learning_rate": 0.001,
+      "loss": 2.2904,
+      "step": 5596
+    },
+    {
+      "epoch": 0.23677976140113377,
+      "grad_norm": 2.0466156005859375,
+      "learning_rate": 0.001,
+      "loss": 2.2768,
+      "step": 5597
+    },
+    {
+      "epoch": 0.23682206616465015,
+      "grad_norm": 0.5960863828659058,
+      "learning_rate": 0.001,
+      "loss": 2.5767,
+      "step": 5598
+    },
+    {
+      "epoch": 0.2368643709281665,
+      "grad_norm": 0.3480601906776428,
+      "learning_rate": 0.001,
+      "loss": 2.7427,
+      "step": 5599
+    },
+    {
+      "epoch": 0.2369066756916829,
+      "grad_norm": 0.22647280991077423,
+      "learning_rate": 0.001,
+      "loss": 4.0171,
+      "step": 5600
+    },
+    {
+      "epoch": 0.23694898045519924,
+      "grad_norm": 0.37474966049194336,
+      "learning_rate": 0.001,
+      "loss": 2.2962,
+      "step": 5601
+    },
+    {
+      "epoch": 0.23699128521871562,
+      "grad_norm": 0.35157617926597595,
+      "learning_rate": 0.001,
+      "loss": 3.5374,
+      "step": 5602
+    },
+    {
+      "epoch": 0.237033589982232,
+      "grad_norm": 1.5529965162277222,
+      "learning_rate": 0.001,
+      "loss": 3.0981,
+      "step": 5603
+    },
+    {
+      "epoch": 0.23707589474574836,
+      "grad_norm": 0.4362848103046417,
+      "learning_rate": 0.001,
+      "loss": 3.007,
+      "step": 5604
+    },
+    {
+      "epoch": 0.23711819950926474,
+      "grad_norm": 0.2729444205760956,
+      "learning_rate": 0.001,
+      "loss": 2.2965,
+      "step": 5605
+    },
+    {
+      "epoch": 0.23716050427278113,
+      "grad_norm": 0.2660609781742096,
+      "learning_rate": 0.001,
+      "loss": 2.2928,
+      "step": 5606
+    },
+    {
+      "epoch": 0.23720280903629748,
+      "grad_norm": 0.23978281021118164,
+      "learning_rate": 0.001,
+      "loss": 2.5681,
+      "step": 5607
+    },
+    {
+      "epoch": 0.23724511379981386,
+      "grad_norm": 0.21553561091423035,
+      "learning_rate": 0.001,
+      "loss": 1.9796,
+      "step": 5608
+    },
+    {
+      "epoch": 0.23728741856333024,
+      "grad_norm": 0.5897154808044434,
+      "learning_rate": 0.001,
+      "loss": 2.986,
+      "step": 5609
+    },
+    {
+      "epoch": 0.2373297233268466,
+      "grad_norm": 0.22807319462299347,
+      "learning_rate": 0.001,
+      "loss": 3.1874,
+      "step": 5610
+    },
+    {
+      "epoch": 0.23737202809036298,
+      "grad_norm": 0.22051000595092773,
+      "learning_rate": 0.001,
+      "loss": 1.8367,
+      "step": 5611
+    },
+    {
+      "epoch": 0.23741433285387933,
+      "grad_norm": 59.09831237792969,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 5612
+    },
+    {
+      "epoch": 0.23745663761739572,
+      "grad_norm": 0.21315006911754608,
+      "learning_rate": 0.001,
+      "loss": 2.1231,
+      "step": 5613
+    },
+    {
+      "epoch": 0.2374989423809121,
+      "grad_norm": 0.45930665731430054,
+      "learning_rate": 0.001,
+      "loss": 3.1726,
+      "step": 5614
+    },
+    {
+      "epoch": 0.23754124714442845,
+      "grad_norm": 0.5922689437866211,
+      "learning_rate": 0.001,
+      "loss": 1.8162,
+      "step": 5615
+    },
+    {
+      "epoch": 0.23758355190794483,
+      "grad_norm": 0.23892509937286377,
+      "learning_rate": 0.001,
+      "loss": 2.2867,
+      "step": 5616
+    },
+    {
+      "epoch": 0.23762585667146122,
+      "grad_norm": 0.2679390013217926,
+      "learning_rate": 0.001,
+      "loss": 2.9999,
+      "step": 5617
+    },
+    {
+      "epoch": 0.23766816143497757,
+      "grad_norm": 0.7704877257347107,
+      "learning_rate": 0.001,
+      "loss": 2.2267,
+      "step": 5618
+    },
+    {
+      "epoch": 0.23771046619849395,
+      "grad_norm": 0.4141584038734436,
+      "learning_rate": 0.001,
+      "loss": 3.1586,
+      "step": 5619
+    },
+    {
+      "epoch": 0.23775277096201033,
+      "grad_norm": 0.6287165284156799,
+      "learning_rate": 0.001,
+      "loss": 2.379,
+      "step": 5620
+    },
+    {
+      "epoch": 0.2377950757255267,
+      "grad_norm": 0.19241921603679657,
+      "learning_rate": 0.001,
+      "loss": 1.791,
+      "step": 5621
+    },
+    {
+      "epoch": 0.23783738048904307,
+      "grad_norm": 0.2662370502948761,
+      "learning_rate": 0.001,
+      "loss": 1.7727,
+      "step": 5622
+    },
+    {
+      "epoch": 0.23787968525255943,
+      "grad_norm": 68.88314819335938,
+      "learning_rate": 0.001,
+      "loss": 2.168,
+      "step": 5623
+    },
+    {
+      "epoch": 0.2379219900160758,
+      "grad_norm": 0.8763318061828613,
+      "learning_rate": 0.001,
+      "loss": 3.0477,
+      "step": 5624
+    },
+    {
+      "epoch": 0.2379642947795922,
+      "grad_norm": 0.878894567489624,
+      "learning_rate": 0.001,
+      "loss": 1.9467,
+      "step": 5625
+    },
+    {
+      "epoch": 0.23800659954310854,
+      "grad_norm": 0.4422621428966522,
+      "learning_rate": 0.001,
+      "loss": 2.7905,
+      "step": 5626
+    },
+    {
+      "epoch": 0.23804890430662493,
+      "grad_norm": 5.424615383148193,
+      "learning_rate": 0.001,
+      "loss": 2.3152,
+      "step": 5627
+    },
+    {
+      "epoch": 0.2380912090701413,
+      "grad_norm": 2.8828539848327637,
+      "learning_rate": 0.001,
+      "loss": 2.4996,
+      "step": 5628
+    },
+    {
+      "epoch": 0.23813351383365766,
+      "grad_norm": 1.1331160068511963,
+      "learning_rate": 0.001,
+      "loss": 2.6441,
+      "step": 5629
+    },
+    {
+      "epoch": 0.23817581859717404,
+      "grad_norm": 1.2550442218780518,
+      "learning_rate": 0.001,
+      "loss": 3.1826,
+      "step": 5630
+    },
+    {
+      "epoch": 0.23821812336069043,
+      "grad_norm": 0.27109989523887634,
+      "learning_rate": 0.001,
+      "loss": 3.46,
+      "step": 5631
+    },
+    {
+      "epoch": 0.23826042812420678,
+      "grad_norm": 0.2896735966205597,
+      "learning_rate": 0.001,
+      "loss": 2.7865,
+      "step": 5632
+    },
+    {
+      "epoch": 0.23830273288772316,
+      "grad_norm": 0.2764877676963806,
+      "learning_rate": 0.001,
+      "loss": 2.0332,
+      "step": 5633
+    },
+    {
+      "epoch": 0.23834503765123952,
+      "grad_norm": 1.915027141571045,
+      "learning_rate": 0.001,
+      "loss": 2.3694,
+      "step": 5634
+    },
+    {
+      "epoch": 0.2383873424147559,
+      "grad_norm": 0.4358770549297333,
+      "learning_rate": 0.001,
+      "loss": 2.3084,
+      "step": 5635
+    },
+    {
+      "epoch": 0.23842964717827228,
+      "grad_norm": 0.2749125063419342,
+      "learning_rate": 0.001,
+      "loss": 2.1625,
+      "step": 5636
+    },
+    {
+      "epoch": 0.23847195194178863,
+      "grad_norm": 0.29137587547302246,
+      "learning_rate": 0.001,
+      "loss": 3.2571,
+      "step": 5637
+    },
+    {
+      "epoch": 0.23851425670530502,
+      "grad_norm": 0.2710089385509491,
+      "learning_rate": 0.001,
+      "loss": 2.8085,
+      "step": 5638
+    },
+    {
+      "epoch": 0.2385565614688214,
+      "grad_norm": 0.3743971288204193,
+      "learning_rate": 0.001,
+      "loss": 3.0382,
+      "step": 5639
+    },
+    {
+      "epoch": 0.23859886623233775,
+      "grad_norm": 0.6103678345680237,
+      "learning_rate": 0.001,
+      "loss": 1.8752,
+      "step": 5640
+    },
+    {
+      "epoch": 0.23864117099585413,
+      "grad_norm": 0.2807563543319702,
+      "learning_rate": 0.001,
+      "loss": 2.6665,
+      "step": 5641
+    },
+    {
+      "epoch": 0.23868347575937052,
+      "grad_norm": 0.4093710780143738,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 5642
+    },
+    {
+      "epoch": 0.23872578052288687,
+      "grad_norm": 0.2517501711845398,
+      "learning_rate": 0.001,
+      "loss": 3.2762,
+      "step": 5643
+    },
+    {
+      "epoch": 0.23876808528640325,
+      "grad_norm": 0.21582388877868652,
+      "learning_rate": 0.001,
+      "loss": 2.8667,
+      "step": 5644
+    },
+    {
+      "epoch": 0.2388103900499196,
+      "grad_norm": 0.22406849265098572,
+      "learning_rate": 0.001,
+      "loss": 2.279,
+      "step": 5645
+    },
+    {
+      "epoch": 0.238852694813436,
+      "grad_norm": 0.17993216216564178,
+      "learning_rate": 0.001,
+      "loss": 1.6104,
+      "step": 5646
+    },
+    {
+      "epoch": 0.23889499957695237,
+      "grad_norm": 0.29950782656669617,
+      "learning_rate": 0.001,
+      "loss": 2.2527,
+      "step": 5647
+    },
+    {
+      "epoch": 0.23893730434046873,
+      "grad_norm": 0.20286992192268372,
+      "learning_rate": 0.001,
+      "loss": 2.5939,
+      "step": 5648
+    },
+    {
+      "epoch": 0.2389796091039851,
+      "grad_norm": 0.5291157960891724,
+      "learning_rate": 0.001,
+      "loss": 2.1778,
+      "step": 5649
+    },
+    {
+      "epoch": 0.2390219138675015,
+      "grad_norm": 0.34574270248413086,
+      "learning_rate": 0.001,
+      "loss": 2.5221,
+      "step": 5650
+    },
+    {
+      "epoch": 0.23906421863101784,
+      "grad_norm": 0.25159138441085815,
+      "learning_rate": 0.001,
+      "loss": 2.4263,
+      "step": 5651
+    },
+    {
+      "epoch": 0.23910652339453423,
+      "grad_norm": 1.9903714656829834,
+      "learning_rate": 0.001,
+      "loss": 2.5326,
+      "step": 5652
+    },
+    {
+      "epoch": 0.2391488281580506,
+      "grad_norm": 0.26104992628097534,
+      "learning_rate": 0.001,
+      "loss": 3.192,
+      "step": 5653
+    },
+    {
+      "epoch": 0.23919113292156696,
+      "grad_norm": 0.21419936418533325,
+      "learning_rate": 0.001,
+      "loss": 2.0116,
+      "step": 5654
+    },
+    {
+      "epoch": 0.23923343768508334,
+      "grad_norm": 0.49943071603775024,
+      "learning_rate": 0.001,
+      "loss": 1.9931,
+      "step": 5655
+    },
+    {
+      "epoch": 0.2392757424485997,
+      "grad_norm": 0.23732519149780273,
+      "learning_rate": 0.001,
+      "loss": 2.1189,
+      "step": 5656
+    },
+    {
+      "epoch": 0.23931804721211608,
+      "grad_norm": 0.2006048560142517,
+      "learning_rate": 0.001,
+      "loss": 2.1357,
+      "step": 5657
+    },
+    {
+      "epoch": 0.23936035197563246,
+      "grad_norm": 0.2843586504459381,
+      "learning_rate": 0.001,
+      "loss": 2.3875,
+      "step": 5658
+    },
+    {
+      "epoch": 0.23940265673914882,
+      "grad_norm": 0.16184455156326294,
+      "learning_rate": 0.001,
+      "loss": 2.2042,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2394449615026652,
+      "grad_norm": 0.1909385770559311,
+      "learning_rate": 0.001,
+      "loss": 2.8616,
+      "step": 5660
+    },
+    {
+      "epoch": 0.23948726626618158,
+      "grad_norm": 0.18999890983104706,
+      "learning_rate": 0.001,
+      "loss": 1.9102,
+      "step": 5661
+    },
+    {
+      "epoch": 0.23952957102969794,
+      "grad_norm": 2.6552658081054688,
+      "learning_rate": 0.001,
+      "loss": 2.8128,
+      "step": 5662
+    },
+    {
+      "epoch": 0.23957187579321432,
+      "grad_norm": 0.21462197601795197,
+      "learning_rate": 0.001,
+      "loss": 2.404,
+      "step": 5663
+    },
+    {
+      "epoch": 0.2396141805567307,
+      "grad_norm": 0.27352675795555115,
+      "learning_rate": 0.001,
+      "loss": 2.4358,
+      "step": 5664
+    },
+    {
+      "epoch": 0.23965648532024705,
+      "grad_norm": 0.22441208362579346,
+      "learning_rate": 0.001,
+      "loss": 2.3097,
+      "step": 5665
+    },
+    {
+      "epoch": 0.23969879008376344,
+      "grad_norm": 0.3627524673938751,
+      "learning_rate": 0.001,
+      "loss": 3.7734,
+      "step": 5666
+    },
+    {
+      "epoch": 0.23974109484727982,
+      "grad_norm": 0.4762495756149292,
+      "learning_rate": 0.001,
+      "loss": 2.1284,
+      "step": 5667
+    },
+    {
+      "epoch": 0.23978339961079617,
+      "grad_norm": 0.2010953575372696,
+      "learning_rate": 0.001,
+      "loss": 2.329,
+      "step": 5668
+    },
+    {
+      "epoch": 0.23982570437431255,
+      "grad_norm": 0.24924032390117645,
+      "learning_rate": 0.001,
+      "loss": 3.1285,
+      "step": 5669
+    },
+    {
+      "epoch": 0.2398680091378289,
+      "grad_norm": 0.6464244723320007,
+      "learning_rate": 0.001,
+      "loss": 2.4777,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2399103139013453,
+      "grad_norm": 0.1977522075176239,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 5671
+    },
+    {
+      "epoch": 0.23995261866486167,
+      "grad_norm": 0.1925719529390335,
+      "learning_rate": 0.001,
+      "loss": 2.2735,
+      "step": 5672
+    },
+    {
+      "epoch": 0.23999492342837803,
+      "grad_norm": 0.18989835679531097,
+      "learning_rate": 0.001,
+      "loss": 2.142,
+      "step": 5673
+    },
+    {
+      "epoch": 0.2400372281918944,
+      "grad_norm": 0.21661527454853058,
+      "learning_rate": 0.001,
+      "loss": 2.0072,
+      "step": 5674
+    },
+    {
+      "epoch": 0.2400795329554108,
+      "grad_norm": 1.2266058921813965,
+      "learning_rate": 0.001,
+      "loss": 2.5989,
+      "step": 5675
+    },
+    {
+      "epoch": 0.24012183771892714,
+      "grad_norm": 0.20844648778438568,
+      "learning_rate": 0.001,
+      "loss": 1.9061,
+      "step": 5676
+    },
+    {
+      "epoch": 0.24016414248244353,
+      "grad_norm": 0.27534279227256775,
+      "learning_rate": 0.001,
+      "loss": 2.631,
+      "step": 5677
+    },
+    {
+      "epoch": 0.2402064472459599,
+      "grad_norm": 0.2036500722169876,
+      "learning_rate": 0.001,
+      "loss": 1.604,
+      "step": 5678
+    },
+    {
+      "epoch": 0.24024875200947626,
+      "grad_norm": 0.18914766609668732,
+      "learning_rate": 0.001,
+      "loss": 1.9546,
+      "step": 5679
+    },
+    {
+      "epoch": 0.24029105677299264,
+      "grad_norm": 0.2073379009962082,
+      "learning_rate": 0.001,
+      "loss": 2.5842,
+      "step": 5680
+    },
+    {
+      "epoch": 0.240333361536509,
+      "grad_norm": 0.32036593556404114,
+      "learning_rate": 0.001,
+      "loss": 2.5532,
+      "step": 5681
+    },
+    {
+      "epoch": 0.24037566630002538,
+      "grad_norm": 0.1830267608165741,
+      "learning_rate": 0.001,
+      "loss": 2.4035,
+      "step": 5682
+    },
+    {
+      "epoch": 0.24041797106354176,
+      "grad_norm": 0.1883174479007721,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 5683
+    },
+    {
+      "epoch": 0.24046027582705812,
+      "grad_norm": 0.2873155176639557,
+      "learning_rate": 0.001,
+      "loss": 2.2658,
+      "step": 5684
+    },
+    {
+      "epoch": 0.2405025805905745,
+      "grad_norm": 0.22399446368217468,
+      "learning_rate": 0.001,
+      "loss": 1.8286,
+      "step": 5685
+    },
+    {
+      "epoch": 0.24054488535409088,
+      "grad_norm": 0.7013282775878906,
+      "learning_rate": 0.001,
+      "loss": 2.4934,
+      "step": 5686
+    },
+    {
+      "epoch": 0.24058719011760724,
+      "grad_norm": 0.21409347653388977,
+      "learning_rate": 0.001,
+      "loss": 2.2924,
+      "step": 5687
+    },
+    {
+      "epoch": 0.24062949488112362,
+      "grad_norm": 2.542304515838623,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 5688
+    },
+    {
+      "epoch": 0.24067179964464,
+      "grad_norm": 0.39416250586509705,
+      "learning_rate": 0.001,
+      "loss": 2.9602,
+      "step": 5689
+    },
+    {
+      "epoch": 0.24071410440815635,
+      "grad_norm": 0.307919979095459,
+      "learning_rate": 0.001,
+      "loss": 1.6349,
+      "step": 5690
+    },
+    {
+      "epoch": 0.24075640917167274,
+      "grad_norm": 0.24086859822273254,
+      "learning_rate": 0.001,
+      "loss": 2.0097,
+      "step": 5691
+    },
+    {
+      "epoch": 0.2407987139351891,
+      "grad_norm": 0.1972251832485199,
+      "learning_rate": 0.001,
+      "loss": 1.7443,
+      "step": 5692
+    },
+    {
+      "epoch": 0.24084101869870547,
+      "grad_norm": 0.5142681002616882,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 5693
+    },
+    {
+      "epoch": 0.24088332346222185,
+      "grad_norm": 0.28559404611587524,
+      "learning_rate": 0.001,
+      "loss": 2.4702,
+      "step": 5694
+    },
+    {
+      "epoch": 0.2409256282257382,
+      "grad_norm": 0.20487889647483826,
+      "learning_rate": 0.001,
+      "loss": 1.9676,
+      "step": 5695
+    },
+    {
+      "epoch": 0.2409679329892546,
+      "grad_norm": 0.20150291919708252,
+      "learning_rate": 0.001,
+      "loss": 1.9557,
+      "step": 5696
+    },
+    {
+      "epoch": 0.24101023775277097,
+      "grad_norm": 0.2249046266078949,
+      "learning_rate": 0.001,
+      "loss": 2.1542,
+      "step": 5697
+    },
+    {
+      "epoch": 0.24105254251628733,
+      "grad_norm": 0.20710447430610657,
+      "learning_rate": 0.001,
+      "loss": 1.6919,
+      "step": 5698
+    },
+    {
+      "epoch": 0.2410948472798037,
+      "grad_norm": 0.2290615290403366,
+      "learning_rate": 0.001,
+      "loss": 2.3559,
+      "step": 5699
+    },
+    {
+      "epoch": 0.2411371520433201,
+      "grad_norm": 0.3252545893192291,
+      "learning_rate": 0.001,
+      "loss": 2.7864,
+      "step": 5700
+    },
+    {
+      "epoch": 0.24117945680683645,
+      "grad_norm": 0.21582552790641785,
+      "learning_rate": 0.001,
+      "loss": 1.7943,
+      "step": 5701
+    },
+    {
+      "epoch": 0.24122176157035283,
+      "grad_norm": 0.2762306332588196,
+      "learning_rate": 0.001,
+      "loss": 2.1902,
+      "step": 5702
+    },
+    {
+      "epoch": 0.24126406633386918,
+      "grad_norm": 0.22092297673225403,
+      "learning_rate": 0.001,
+      "loss": 2.4124,
+      "step": 5703
+    },
+    {
+      "epoch": 0.24130637109738556,
+      "grad_norm": 1.6592704057693481,
+      "learning_rate": 0.001,
+      "loss": 1.9375,
+      "step": 5704
+    },
+    {
+      "epoch": 0.24134867586090195,
+      "grad_norm": 0.2536323666572571,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 5705
+    },
+    {
+      "epoch": 0.2413909806244183,
+      "grad_norm": 0.23811465501785278,
+      "learning_rate": 0.001,
+      "loss": 2.3419,
+      "step": 5706
+    },
+    {
+      "epoch": 0.24143328538793468,
+      "grad_norm": 8.864980697631836,
+      "learning_rate": 0.001,
+      "loss": 3.1004,
+      "step": 5707
+    },
+    {
+      "epoch": 0.24147559015145106,
+      "grad_norm": 0.23312745988368988,
+      "learning_rate": 0.001,
+      "loss": 2.076,
+      "step": 5708
+    },
+    {
+      "epoch": 0.24151789491496742,
+      "grad_norm": 0.2500420808792114,
+      "learning_rate": 0.001,
+      "loss": 1.8015,
+      "step": 5709
+    },
+    {
+      "epoch": 0.2415601996784838,
+      "grad_norm": 0.397246390581131,
+      "learning_rate": 0.001,
+      "loss": 2.0747,
+      "step": 5710
+    },
+    {
+      "epoch": 0.24160250444200018,
+      "grad_norm": 0.256539523601532,
+      "learning_rate": 0.001,
+      "loss": 2.0406,
+      "step": 5711
+    },
+    {
+      "epoch": 0.24164480920551654,
+      "grad_norm": 0.2969229817390442,
+      "learning_rate": 0.001,
+      "loss": 2.8112,
+      "step": 5712
+    },
+    {
+      "epoch": 0.24168711396903292,
+      "grad_norm": 0.35546842217445374,
+      "learning_rate": 0.001,
+      "loss": 3.5403,
+      "step": 5713
+    },
+    {
+      "epoch": 0.24172941873254927,
+      "grad_norm": 0.22269092500209808,
+      "learning_rate": 0.001,
+      "loss": 1.9222,
+      "step": 5714
+    },
+    {
+      "epoch": 0.24177172349606565,
+      "grad_norm": 0.20512692630290985,
+      "learning_rate": 0.001,
+      "loss": 1.9088,
+      "step": 5715
+    },
+    {
+      "epoch": 0.24181402825958204,
+      "grad_norm": 0.26039499044418335,
+      "learning_rate": 0.001,
+      "loss": 2.0708,
+      "step": 5716
+    },
+    {
+      "epoch": 0.2418563330230984,
+      "grad_norm": 38.39719009399414,
+      "learning_rate": 0.001,
+      "loss": 1.9211,
+      "step": 5717
+    },
+    {
+      "epoch": 0.24189863778661477,
+      "grad_norm": 0.3820383548736572,
+      "learning_rate": 0.001,
+      "loss": 3.4881,
+      "step": 5718
+    },
+    {
+      "epoch": 0.24194094255013116,
+      "grad_norm": 0.3408060073852539,
+      "learning_rate": 0.001,
+      "loss": 2.9283,
+      "step": 5719
+    },
+    {
+      "epoch": 0.2419832473136475,
+      "grad_norm": 0.36963334679603577,
+      "learning_rate": 0.001,
+      "loss": 3.4062,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2420255520771639,
+      "grad_norm": 1.3119021654129028,
+      "learning_rate": 0.001,
+      "loss": 3.7547,
+      "step": 5721
+    },
+    {
+      "epoch": 0.24206785684068027,
+      "grad_norm": 0.27805081009864807,
+      "learning_rate": 0.001,
+      "loss": 3.2918,
+      "step": 5722
+    },
+    {
+      "epoch": 0.24211016160419663,
+      "grad_norm": 0.3566412031650543,
+      "learning_rate": 0.001,
+      "loss": 2.1117,
+      "step": 5723
+    },
+    {
+      "epoch": 0.242152466367713,
+      "grad_norm": 0.3360730707645416,
+      "learning_rate": 0.001,
+      "loss": 2.1076,
+      "step": 5724
+    },
+    {
+      "epoch": 0.24219477113122936,
+      "grad_norm": 0.27928465604782104,
+      "learning_rate": 0.001,
+      "loss": 2.2339,
+      "step": 5725
+    },
+    {
+      "epoch": 0.24223707589474575,
+      "grad_norm": 2.314484119415283,
+      "learning_rate": 0.001,
+      "loss": 2.1504,
+      "step": 5726
+    },
+    {
+      "epoch": 0.24227938065826213,
+      "grad_norm": 0.2751530706882477,
+      "learning_rate": 0.001,
+      "loss": 2.5127,
+      "step": 5727
+    },
+    {
+      "epoch": 0.24232168542177848,
+      "grad_norm": 0.20994529128074646,
+      "learning_rate": 0.001,
+      "loss": 2.1564,
+      "step": 5728
+    },
+    {
+      "epoch": 0.24236399018529486,
+      "grad_norm": 0.246437668800354,
+      "learning_rate": 0.001,
+      "loss": 2.3628,
+      "step": 5729
+    },
+    {
+      "epoch": 0.24240629494881125,
+      "grad_norm": 2.577066659927368,
+      "learning_rate": 0.001,
+      "loss": 2.6037,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2424485997123276,
+      "grad_norm": 0.2616536021232605,
+      "learning_rate": 0.001,
+      "loss": 2.1903,
+      "step": 5731
+    },
+    {
+      "epoch": 0.24249090447584398,
+      "grad_norm": 0.24390681087970734,
+      "learning_rate": 0.001,
+      "loss": 2.4864,
+      "step": 5732
+    },
+    {
+      "epoch": 0.24253320923936036,
+      "grad_norm": 0.18465809524059296,
+      "learning_rate": 0.001,
+      "loss": 1.9357,
+      "step": 5733
+    },
+    {
+      "epoch": 0.24257551400287672,
+      "grad_norm": 1.6933213472366333,
+      "learning_rate": 0.001,
+      "loss": 1.7803,
+      "step": 5734
+    },
+    {
+      "epoch": 0.2426178187663931,
+      "grad_norm": 0.25248533487319946,
+      "learning_rate": 0.001,
+      "loss": 2.9161,
+      "step": 5735
+    },
+    {
+      "epoch": 0.24266012352990945,
+      "grad_norm": 0.20103050768375397,
+      "learning_rate": 0.001,
+      "loss": 1.9586,
+      "step": 5736
+    },
+    {
+      "epoch": 0.24270242829342584,
+      "grad_norm": 0.8724969029426575,
+      "learning_rate": 0.001,
+      "loss": 1.979,
+      "step": 5737
+    },
+    {
+      "epoch": 0.24274473305694222,
+      "grad_norm": 0.23048318922519684,
+      "learning_rate": 0.001,
+      "loss": 2.1012,
+      "step": 5738
+    },
+    {
+      "epoch": 0.24278703782045857,
+      "grad_norm": 0.24367755651474,
+      "learning_rate": 0.001,
+      "loss": 2.41,
+      "step": 5739
+    },
+    {
+      "epoch": 0.24282934258397496,
+      "grad_norm": 0.4339103102684021,
+      "learning_rate": 0.001,
+      "loss": 2.6809,
+      "step": 5740
+    },
+    {
+      "epoch": 0.24287164734749134,
+      "grad_norm": 1.0650697946548462,
+      "learning_rate": 0.001,
+      "loss": 2.7325,
+      "step": 5741
+    },
+    {
+      "epoch": 0.2429139521110077,
+      "grad_norm": 2.662815809249878,
+      "learning_rate": 0.001,
+      "loss": 1.9265,
+      "step": 5742
+    },
+    {
+      "epoch": 0.24295625687452407,
+      "grad_norm": 0.1942417025566101,
+      "learning_rate": 0.001,
+      "loss": 1.7619,
+      "step": 5743
+    },
+    {
+      "epoch": 0.24299856163804046,
+      "grad_norm": 0.28779444098472595,
+      "learning_rate": 0.001,
+      "loss": 2.8304,
+      "step": 5744
+    },
+    {
+      "epoch": 0.2430408664015568,
+      "grad_norm": 0.28456562757492065,
+      "learning_rate": 0.001,
+      "loss": 3.3492,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2430831711650732,
+      "grad_norm": 2.139453172683716,
+      "learning_rate": 0.001,
+      "loss": 3.8633,
+      "step": 5746
+    },
+    {
+      "epoch": 0.24312547592858955,
+      "grad_norm": 0.35128921270370483,
+      "learning_rate": 0.001,
+      "loss": 2.3143,
+      "step": 5747
+    },
+    {
+      "epoch": 0.24316778069210593,
+      "grad_norm": 0.3761100172996521,
+      "learning_rate": 0.001,
+      "loss": 2.5602,
+      "step": 5748
+    },
+    {
+      "epoch": 0.2432100854556223,
+      "grad_norm": 0.3312492370605469,
+      "learning_rate": 0.001,
+      "loss": 2.6397,
+      "step": 5749
+    },
+    {
+      "epoch": 0.24325239021913866,
+      "grad_norm": 14.835386276245117,
+      "learning_rate": 0.001,
+      "loss": 2.1131,
+      "step": 5750
+    },
+    {
+      "epoch": 0.24329469498265505,
+      "grad_norm": 0.33980563282966614,
+      "learning_rate": 0.001,
+      "loss": 3.0289,
+      "step": 5751
+    },
+    {
+      "epoch": 0.24333699974617143,
+      "grad_norm": 0.3430834412574768,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 5752
+    },
+    {
+      "epoch": 0.24337930450968778,
+      "grad_norm": 0.2788030207157135,
+      "learning_rate": 0.001,
+      "loss": 2.3714,
+      "step": 5753
+    },
+    {
+      "epoch": 0.24342160927320416,
+      "grad_norm": 0.242362380027771,
+      "learning_rate": 0.001,
+      "loss": 2.3746,
+      "step": 5754
+    },
+    {
+      "epoch": 0.24346391403672055,
+      "grad_norm": 0.4052402079105377,
+      "learning_rate": 0.001,
+      "loss": 3.5768,
+      "step": 5755
+    },
+    {
+      "epoch": 0.2435062188002369,
+      "grad_norm": 0.24132952094078064,
+      "learning_rate": 0.001,
+      "loss": 3.8857,
+      "step": 5756
+    },
+    {
+      "epoch": 0.24354852356375328,
+      "grad_norm": 1.1747184991836548,
+      "learning_rate": 0.001,
+      "loss": 2.5792,
+      "step": 5757
+    },
+    {
+      "epoch": 0.24359082832726964,
+      "grad_norm": 0.3181629776954651,
+      "learning_rate": 0.001,
+      "loss": 2.4274,
+      "step": 5758
+    },
+    {
+      "epoch": 0.24363313309078602,
+      "grad_norm": 0.22457624971866608,
+      "learning_rate": 0.001,
+      "loss": 2.6695,
+      "step": 5759
+    },
+    {
+      "epoch": 0.2436754378543024,
+      "grad_norm": 0.24323022365570068,
+      "learning_rate": 0.001,
+      "loss": 2.0115,
+      "step": 5760
+    },
+    {
+      "epoch": 0.24371774261781876,
+      "grad_norm": 0.2709481120109558,
+      "learning_rate": 0.001,
+      "loss": 2.3548,
+      "step": 5761
+    },
+    {
+      "epoch": 0.24376004738133514,
+      "grad_norm": 0.20824141800403595,
+      "learning_rate": 0.001,
+      "loss": 1.701,
+      "step": 5762
+    },
+    {
+      "epoch": 0.24380235214485152,
+      "grad_norm": 1.6470791101455688,
+      "learning_rate": 0.001,
+      "loss": 2.4078,
+      "step": 5763
+    },
+    {
+      "epoch": 0.24384465690836787,
+      "grad_norm": 0.2733412981033325,
+      "learning_rate": 0.001,
+      "loss": 2.0431,
+      "step": 5764
+    },
+    {
+      "epoch": 0.24388696167188426,
+      "grad_norm": 0.2784044146537781,
+      "learning_rate": 0.001,
+      "loss": 2.6605,
+      "step": 5765
+    },
+    {
+      "epoch": 0.24392926643540064,
+      "grad_norm": 0.661558210849762,
+      "learning_rate": 0.001,
+      "loss": 2.7723,
+      "step": 5766
+    },
+    {
+      "epoch": 0.243971571198917,
+      "grad_norm": 0.24639226496219635,
+      "learning_rate": 0.001,
+      "loss": 2.2698,
+      "step": 5767
+    },
+    {
+      "epoch": 0.24401387596243337,
+      "grad_norm": 0.2241000235080719,
+      "learning_rate": 0.001,
+      "loss": 2.6614,
+      "step": 5768
+    },
+    {
+      "epoch": 0.24405618072594973,
+      "grad_norm": 0.23688313364982605,
+      "learning_rate": 0.001,
+      "loss": 3.4856,
+      "step": 5769
+    },
+    {
+      "epoch": 0.2440984854894661,
+      "grad_norm": 0.22975680232048035,
+      "learning_rate": 0.001,
+      "loss": 2.7368,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2441407902529825,
+      "grad_norm": 0.2196754515171051,
+      "learning_rate": 0.001,
+      "loss": 2.1829,
+      "step": 5771
+    },
+    {
+      "epoch": 0.24418309501649885,
+      "grad_norm": 0.21035362780094147,
+      "learning_rate": 0.001,
+      "loss": 3.0376,
+      "step": 5772
+    },
+    {
+      "epoch": 0.24422539978001523,
+      "grad_norm": 0.19896003603935242,
+      "learning_rate": 0.001,
+      "loss": 1.9874,
+      "step": 5773
+    },
+    {
+      "epoch": 0.2442677045435316,
+      "grad_norm": 0.24322247505187988,
+      "learning_rate": 0.001,
+      "loss": 3.4208,
+      "step": 5774
+    },
+    {
+      "epoch": 0.24431000930704797,
+      "grad_norm": 0.20282556116580963,
+      "learning_rate": 0.001,
+      "loss": 2.0628,
+      "step": 5775
+    },
+    {
+      "epoch": 0.24435231407056435,
+      "grad_norm": 0.19425134360790253,
+      "learning_rate": 0.001,
+      "loss": 2.9513,
+      "step": 5776
+    },
+    {
+      "epoch": 0.24439461883408073,
+      "grad_norm": 0.26466214656829834,
+      "learning_rate": 0.001,
+      "loss": 2.5617,
+      "step": 5777
+    },
+    {
+      "epoch": 0.24443692359759708,
+      "grad_norm": 0.2062908262014389,
+      "learning_rate": 0.001,
+      "loss": 2.3014,
+      "step": 5778
+    },
+    {
+      "epoch": 0.24447922836111347,
+      "grad_norm": 0.17950376868247986,
+      "learning_rate": 0.001,
+      "loss": 2.2402,
+      "step": 5779
+    },
+    {
+      "epoch": 0.24452153312462982,
+      "grad_norm": 0.17766666412353516,
+      "learning_rate": 0.001,
+      "loss": 1.9251,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2445638378881462,
+      "grad_norm": 0.20533357560634613,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 5781
+    },
+    {
+      "epoch": 0.24460614265166258,
+      "grad_norm": 0.21567344665527344,
+      "learning_rate": 0.001,
+      "loss": 2.074,
+      "step": 5782
+    },
+    {
+      "epoch": 0.24464844741517894,
+      "grad_norm": 0.17948715388774872,
+      "learning_rate": 0.001,
+      "loss": 2.2158,
+      "step": 5783
+    },
+    {
+      "epoch": 0.24469075217869532,
+      "grad_norm": 0.2008078396320343,
+      "learning_rate": 0.001,
+      "loss": 1.9838,
+      "step": 5784
+    },
+    {
+      "epoch": 0.2447330569422117,
+      "grad_norm": 0.6070753931999207,
+      "learning_rate": 0.001,
+      "loss": 2.0619,
+      "step": 5785
+    },
+    {
+      "epoch": 0.24477536170572806,
+      "grad_norm": 0.21907269954681396,
+      "learning_rate": 0.001,
+      "loss": 2.5838,
+      "step": 5786
+    },
+    {
+      "epoch": 0.24481766646924444,
+      "grad_norm": 0.46992072463035583,
+      "learning_rate": 0.001,
+      "loss": 3.1865,
+      "step": 5787
+    },
+    {
+      "epoch": 0.24485997123276082,
+      "grad_norm": 0.16599377989768982,
+      "learning_rate": 0.001,
+      "loss": 1.9173,
+      "step": 5788
+    },
+    {
+      "epoch": 0.24490227599627717,
+      "grad_norm": 0.1835167109966278,
+      "learning_rate": 0.001,
+      "loss": 2.0218,
+      "step": 5789
+    },
+    {
+      "epoch": 0.24494458075979356,
+      "grad_norm": 3.3056118488311768,
+      "learning_rate": 0.001,
+      "loss": 2.8621,
+      "step": 5790
+    },
+    {
+      "epoch": 0.24498688552330994,
+      "grad_norm": 0.4746275842189789,
+      "learning_rate": 0.001,
+      "loss": 2.476,
+      "step": 5791
+    },
+    {
+      "epoch": 0.2450291902868263,
+      "grad_norm": 0.48032623529434204,
+      "learning_rate": 0.001,
+      "loss": 2.3223,
+      "step": 5792
+    },
+    {
+      "epoch": 0.24507149505034267,
+      "grad_norm": 0.31363826990127563,
+      "learning_rate": 0.001,
+      "loss": 3.0721,
+      "step": 5793
+    },
+    {
+      "epoch": 0.24511379981385903,
+      "grad_norm": 0.1737576723098755,
+      "learning_rate": 0.001,
+      "loss": 2.0079,
+      "step": 5794
+    },
+    {
+      "epoch": 0.2451561045773754,
+      "grad_norm": 0.8084182143211365,
+      "learning_rate": 0.001,
+      "loss": 2.1507,
+      "step": 5795
+    },
+    {
+      "epoch": 0.2451984093408918,
+      "grad_norm": 0.1980573832988739,
+      "learning_rate": 0.001,
+      "loss": 2.65,
+      "step": 5796
+    },
+    {
+      "epoch": 0.24524071410440815,
+      "grad_norm": 0.32529035210609436,
+      "learning_rate": 0.001,
+      "loss": 2.2491,
+      "step": 5797
+    },
+    {
+      "epoch": 0.24528301886792453,
+      "grad_norm": 0.2029356062412262,
+      "learning_rate": 0.001,
+      "loss": 2.3983,
+      "step": 5798
+    },
+    {
+      "epoch": 0.2453253236314409,
+      "grad_norm": 0.1999257355928421,
+      "learning_rate": 0.001,
+      "loss": 2.4885,
+      "step": 5799
+    },
+    {
+      "epoch": 0.24536762839495727,
+      "grad_norm": 0.34168320894241333,
+      "learning_rate": 0.001,
+      "loss": 2.712,
+      "step": 5800
+    },
+    {
+      "epoch": 0.24540993315847365,
+      "grad_norm": 0.24313102662563324,
+      "learning_rate": 0.001,
+      "loss": 2.064,
+      "step": 5801
+    },
+    {
+      "epoch": 0.24545223792199003,
+      "grad_norm": 0.2077663093805313,
+      "learning_rate": 0.001,
+      "loss": 2.3874,
+      "step": 5802
+    },
+    {
+      "epoch": 0.24549454268550638,
+      "grad_norm": 0.21080611646175385,
+      "learning_rate": 0.001,
+      "loss": 4.459,
+      "step": 5803
+    },
+    {
+      "epoch": 0.24553684744902277,
+      "grad_norm": 0.22150756418704987,
+      "learning_rate": 0.001,
+      "loss": 2.0071,
+      "step": 5804
+    },
+    {
+      "epoch": 0.24557915221253912,
+      "grad_norm": 0.27595874667167664,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 5805
+    },
+    {
+      "epoch": 0.2456214569760555,
+      "grad_norm": 38.239009857177734,
+      "learning_rate": 0.001,
+      "loss": 1.7931,
+      "step": 5806
+    },
+    {
+      "epoch": 0.24566376173957188,
+      "grad_norm": 0.24801005423069,
+      "learning_rate": 0.001,
+      "loss": 2.6051,
+      "step": 5807
+    },
+    {
+      "epoch": 0.24570606650308824,
+      "grad_norm": 0.26121652126312256,
+      "learning_rate": 0.001,
+      "loss": 2.9589,
+      "step": 5808
+    },
+    {
+      "epoch": 0.24574837126660462,
+      "grad_norm": 0.18741196393966675,
+      "learning_rate": 0.001,
+      "loss": 1.4772,
+      "step": 5809
+    },
+    {
+      "epoch": 0.245790676030121,
+      "grad_norm": 0.2526096701622009,
+      "learning_rate": 0.001,
+      "loss": 1.857,
+      "step": 5810
+    },
+    {
+      "epoch": 0.24583298079363736,
+      "grad_norm": 0.28198450803756714,
+      "learning_rate": 0.001,
+      "loss": 2.0914,
+      "step": 5811
+    },
+    {
+      "epoch": 0.24587528555715374,
+      "grad_norm": 0.22132883965969086,
+      "learning_rate": 0.001,
+      "loss": 2.4616,
+      "step": 5812
+    },
+    {
+      "epoch": 0.24591759032067012,
+      "grad_norm": 0.26563504338264465,
+      "learning_rate": 0.001,
+      "loss": 2.2087,
+      "step": 5813
+    },
+    {
+      "epoch": 0.24595989508418648,
+      "grad_norm": 0.35990941524505615,
+      "learning_rate": 0.001,
+      "loss": 2.3027,
+      "step": 5814
+    },
+    {
+      "epoch": 0.24600219984770286,
+      "grad_norm": 0.20027977228164673,
+      "learning_rate": 0.001,
+      "loss": 3.1369,
+      "step": 5815
+    },
+    {
+      "epoch": 0.2460445046112192,
+      "grad_norm": 0.9017431735992432,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 5816
+    },
+    {
+      "epoch": 0.2460868093747356,
+      "grad_norm": 1.3316782712936401,
+      "learning_rate": 0.001,
+      "loss": 2.9245,
+      "step": 5817
+    },
+    {
+      "epoch": 0.24612911413825198,
+      "grad_norm": 0.24641363322734833,
+      "learning_rate": 0.001,
+      "loss": 2.1923,
+      "step": 5818
+    },
+    {
+      "epoch": 0.24617141890176833,
+      "grad_norm": 0.3345772325992584,
+      "learning_rate": 0.001,
+      "loss": 2.3912,
+      "step": 5819
+    },
+    {
+      "epoch": 0.2462137236652847,
+      "grad_norm": 0.23146621882915497,
+      "learning_rate": 0.001,
+      "loss": 2.2022,
+      "step": 5820
+    },
+    {
+      "epoch": 0.2462560284288011,
+      "grad_norm": 0.3035587966442108,
+      "learning_rate": 0.001,
+      "loss": 2.5982,
+      "step": 5821
+    },
+    {
+      "epoch": 0.24629833319231745,
+      "grad_norm": 0.4304903447628021,
+      "learning_rate": 0.001,
+      "loss": 2.566,
+      "step": 5822
+    },
+    {
+      "epoch": 0.24634063795583383,
+      "grad_norm": 0.21231508255004883,
+      "learning_rate": 0.001,
+      "loss": 2.6924,
+      "step": 5823
+    },
+    {
+      "epoch": 0.2463829427193502,
+      "grad_norm": 0.21332456171512604,
+      "learning_rate": 0.001,
+      "loss": 2.458,
+      "step": 5824
+    },
+    {
+      "epoch": 0.24642524748286657,
+      "grad_norm": 0.24201619625091553,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 5825
+    },
+    {
+      "epoch": 0.24646755224638295,
+      "grad_norm": 0.20796683430671692,
+      "learning_rate": 0.001,
+      "loss": 2.2657,
+      "step": 5826
+    },
+    {
+      "epoch": 0.2465098570098993,
+      "grad_norm": 0.3387661278247833,
+      "learning_rate": 0.001,
+      "loss": 1.8694,
+      "step": 5827
+    },
+    {
+      "epoch": 0.24655216177341568,
+      "grad_norm": 0.5283912420272827,
+      "learning_rate": 0.001,
+      "loss": 2.0767,
+      "step": 5828
+    },
+    {
+      "epoch": 0.24659446653693207,
+      "grad_norm": 0.21991148591041565,
+      "learning_rate": 0.001,
+      "loss": 3.0532,
+      "step": 5829
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 0.2817951738834381,
+      "learning_rate": 0.001,
+      "loss": 2.8324,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2466790760639648,
+      "grad_norm": 0.2231799066066742,
+      "learning_rate": 0.001,
+      "loss": 3.0957,
+      "step": 5831
+    },
+    {
+      "epoch": 0.24672138082748118,
+      "grad_norm": 0.23108860850334167,
+      "learning_rate": 0.001,
+      "loss": 2.1957,
+      "step": 5832
+    },
+    {
+      "epoch": 0.24676368559099754,
+      "grad_norm": 0.47794216871261597,
+      "learning_rate": 0.001,
+      "loss": 2.4318,
+      "step": 5833
+    },
+    {
+      "epoch": 0.24680599035451392,
+      "grad_norm": 0.3009633421897888,
+      "learning_rate": 0.001,
+      "loss": 2.801,
+      "step": 5834
+    },
+    {
+      "epoch": 0.2468482951180303,
+      "grad_norm": 0.1829899400472641,
+      "learning_rate": 0.001,
+      "loss": 1.762,
+      "step": 5835
+    },
+    {
+      "epoch": 0.24689059988154666,
+      "grad_norm": 0.25043395161628723,
+      "learning_rate": 0.001,
+      "loss": 2.103,
+      "step": 5836
+    },
+    {
+      "epoch": 0.24693290464506304,
+      "grad_norm": 1.0401873588562012,
+      "learning_rate": 0.001,
+      "loss": 2.5898,
+      "step": 5837
+    },
+    {
+      "epoch": 0.2469752094085794,
+      "grad_norm": 0.22603413462638855,
+      "learning_rate": 0.001,
+      "loss": 3.3657,
+      "step": 5838
+    },
+    {
+      "epoch": 0.24701751417209578,
+      "grad_norm": 0.25399866700172424,
+      "learning_rate": 0.001,
+      "loss": 2.7811,
+      "step": 5839
+    },
+    {
+      "epoch": 0.24705981893561216,
+      "grad_norm": 0.24066664278507233,
+      "learning_rate": 0.001,
+      "loss": 3.1812,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2471021236991285,
+      "grad_norm": 0.27932751178741455,
+      "learning_rate": 0.001,
+      "loss": 2.5269,
+      "step": 5841
+    },
+    {
+      "epoch": 0.2471444284626449,
+      "grad_norm": 0.2165471911430359,
+      "learning_rate": 0.001,
+      "loss": 1.9364,
+      "step": 5842
+    },
+    {
+      "epoch": 0.24718673322616128,
+      "grad_norm": 0.1947283148765564,
+      "learning_rate": 0.001,
+      "loss": 1.9661,
+      "step": 5843
+    },
+    {
+      "epoch": 0.24722903798967763,
+      "grad_norm": 0.245658740401268,
+      "learning_rate": 0.001,
+      "loss": 2.2399,
+      "step": 5844
+    },
+    {
+      "epoch": 0.247271342753194,
+      "grad_norm": 0.21546167135238647,
+      "learning_rate": 0.001,
+      "loss": 2.4518,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2473136475167104,
+      "grad_norm": 0.18316084146499634,
+      "learning_rate": 0.001,
+      "loss": 1.8123,
+      "step": 5846
+    },
+    {
+      "epoch": 0.24735595228022675,
+      "grad_norm": 2.9340388774871826,
+      "learning_rate": 0.001,
+      "loss": 2.8578,
+      "step": 5847
+    },
+    {
+      "epoch": 0.24739825704374313,
+      "grad_norm": 0.16483551263809204,
+      "learning_rate": 0.001,
+      "loss": 1.7181,
+      "step": 5848
+    },
+    {
+      "epoch": 0.24744056180725948,
+      "grad_norm": 0.18957403302192688,
+      "learning_rate": 0.001,
+      "loss": 2.1544,
+      "step": 5849
+    },
+    {
+      "epoch": 0.24748286657077587,
+      "grad_norm": 0.19052253663539886,
+      "learning_rate": 0.001,
+      "loss": 1.8171,
+      "step": 5850
+    },
+    {
+      "epoch": 0.24752517133429225,
+      "grad_norm": 0.19511641561985016,
+      "learning_rate": 0.001,
+      "loss": 1.7506,
+      "step": 5851
+    },
+    {
+      "epoch": 0.2475674760978086,
+      "grad_norm": 0.7128787040710449,
+      "learning_rate": 0.001,
+      "loss": 1.6148,
+      "step": 5852
+    },
+    {
+      "epoch": 0.24760978086132499,
+      "grad_norm": 0.1971040964126587,
+      "learning_rate": 0.001,
+      "loss": 2.5899,
+      "step": 5853
+    },
+    {
+      "epoch": 0.24765208562484137,
+      "grad_norm": 0.3237464129924774,
+      "learning_rate": 0.001,
+      "loss": 1.9698,
+      "step": 5854
+    },
+    {
+      "epoch": 0.24769439038835772,
+      "grad_norm": 0.20334038138389587,
+      "learning_rate": 0.001,
+      "loss": 1.8541,
+      "step": 5855
+    },
+    {
+      "epoch": 0.2477366951518741,
+      "grad_norm": 0.19235415756702423,
+      "learning_rate": 0.001,
+      "loss": 2.1102,
+      "step": 5856
+    },
+    {
+      "epoch": 0.24777899991539049,
+      "grad_norm": 0.18635153770446777,
+      "learning_rate": 0.001,
+      "loss": 1.7164,
+      "step": 5857
+    },
+    {
+      "epoch": 0.24782130467890684,
+      "grad_norm": 0.8393286466598511,
+      "learning_rate": 0.001,
+      "loss": 2.0415,
+      "step": 5858
+    },
+    {
+      "epoch": 0.24786360944242322,
+      "grad_norm": 0.20402997732162476,
+      "learning_rate": 0.001,
+      "loss": 2.5329,
+      "step": 5859
+    },
+    {
+      "epoch": 0.24790591420593958,
+      "grad_norm": 0.20956696569919586,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 5860
+    },
+    {
+      "epoch": 0.24794821896945596,
+      "grad_norm": 0.3180994391441345,
+      "learning_rate": 0.001,
+      "loss": 2.319,
+      "step": 5861
+    },
+    {
+      "epoch": 0.24799052373297234,
+      "grad_norm": 0.4629594087600708,
+      "learning_rate": 0.001,
+      "loss": 2.763,
+      "step": 5862
+    },
+    {
+      "epoch": 0.2480328284964887,
+      "grad_norm": 0.2035917043685913,
+      "learning_rate": 0.001,
+      "loss": 2.7616,
+      "step": 5863
+    },
+    {
+      "epoch": 0.24807513326000508,
+      "grad_norm": 0.20831398665905,
+      "learning_rate": 0.001,
+      "loss": 3.2368,
+      "step": 5864
+    },
+    {
+      "epoch": 0.24811743802352146,
+      "grad_norm": 0.24462851881980896,
+      "learning_rate": 0.001,
+      "loss": 2.8819,
+      "step": 5865
+    },
+    {
+      "epoch": 0.2481597427870378,
+      "grad_norm": 0.72509765625,
+      "learning_rate": 0.001,
+      "loss": 1.4765,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2482020475505542,
+      "grad_norm": 0.2004188746213913,
+      "learning_rate": 0.001,
+      "loss": 2.7114,
+      "step": 5867
+    },
+    {
+      "epoch": 0.24824435231407058,
+      "grad_norm": 0.16426634788513184,
+      "learning_rate": 0.001,
+      "loss": 2.2513,
+      "step": 5868
+    },
+    {
+      "epoch": 0.24828665707758693,
+      "grad_norm": 0.23938223719596863,
+      "learning_rate": 0.001,
+      "loss": 2.5919,
+      "step": 5869
+    },
+    {
+      "epoch": 0.2483289618411033,
+      "grad_norm": 0.20206694304943085,
+      "learning_rate": 0.001,
+      "loss": 1.8613,
+      "step": 5870
+    },
+    {
+      "epoch": 0.24837126660461967,
+      "grad_norm": 0.20341122150421143,
+      "learning_rate": 0.001,
+      "loss": 1.6956,
+      "step": 5871
+    },
+    {
+      "epoch": 0.24841357136813605,
+      "grad_norm": 0.16674411296844482,
+      "learning_rate": 0.001,
+      "loss": 2.3249,
+      "step": 5872
+    },
+    {
+      "epoch": 0.24845587613165243,
+      "grad_norm": 0.17640060186386108,
+      "learning_rate": 0.001,
+      "loss": 1.5507,
+      "step": 5873
+    },
+    {
+      "epoch": 0.24849818089516879,
+      "grad_norm": 0.2345527559518814,
+      "learning_rate": 0.001,
+      "loss": 2.1983,
+      "step": 5874
+    },
+    {
+      "epoch": 0.24854048565868517,
+      "grad_norm": 2.3376219272613525,
+      "learning_rate": 0.001,
+      "loss": 2.1126,
+      "step": 5875
+    },
+    {
+      "epoch": 0.24858279042220155,
+      "grad_norm": 0.24027837812900543,
+      "learning_rate": 0.001,
+      "loss": 1.968,
+      "step": 5876
+    },
+    {
+      "epoch": 0.2486250951857179,
+      "grad_norm": 0.22010159492492676,
+      "learning_rate": 0.001,
+      "loss": 1.8924,
+      "step": 5877
+    },
+    {
+      "epoch": 0.24866739994923429,
+      "grad_norm": 0.1861301064491272,
+      "learning_rate": 0.001,
+      "loss": 2.1468,
+      "step": 5878
+    },
+    {
+      "epoch": 0.24870970471275067,
+      "grad_norm": 0.2004723697900772,
+      "learning_rate": 0.001,
+      "loss": 2.5531,
+      "step": 5879
+    },
+    {
+      "epoch": 0.24875200947626702,
+      "grad_norm": 0.33933591842651367,
+      "learning_rate": 0.001,
+      "loss": 2.7241,
+      "step": 5880
+    },
+    {
+      "epoch": 0.2487943142397834,
+      "grad_norm": 0.26953572034835815,
+      "learning_rate": 0.001,
+      "loss": 2.3623,
+      "step": 5881
+    },
+    {
+      "epoch": 0.24883661900329976,
+      "grad_norm": 0.1680213063955307,
+      "learning_rate": 0.001,
+      "loss": 2.5205,
+      "step": 5882
+    },
+    {
+      "epoch": 0.24887892376681614,
+      "grad_norm": 0.17806370556354523,
+      "learning_rate": 0.001,
+      "loss": 2.0251,
+      "step": 5883
+    },
+    {
+      "epoch": 0.24892122853033252,
+      "grad_norm": 0.44884923100471497,
+      "learning_rate": 0.001,
+      "loss": 1.8082,
+      "step": 5884
+    },
+    {
+      "epoch": 0.24896353329384888,
+      "grad_norm": 0.28582075238227844,
+      "learning_rate": 0.001,
+      "loss": 3.7704,
+      "step": 5885
+    },
+    {
+      "epoch": 0.24900583805736526,
+      "grad_norm": 0.21283996105194092,
+      "learning_rate": 0.001,
+      "loss": 1.7666,
+      "step": 5886
+    },
+    {
+      "epoch": 0.24904814282088164,
+      "grad_norm": 0.22247281670570374,
+      "learning_rate": 0.001,
+      "loss": 2.6567,
+      "step": 5887
+    },
+    {
+      "epoch": 0.249090447584398,
+      "grad_norm": 0.2375125139951706,
+      "learning_rate": 0.001,
+      "loss": 2.3062,
+      "step": 5888
+    },
+    {
+      "epoch": 0.24913275234791438,
+      "grad_norm": 0.3201584219932556,
+      "learning_rate": 0.001,
+      "loss": 3.1853,
+      "step": 5889
+    },
+    {
+      "epoch": 0.24917505711143076,
+      "grad_norm": 0.21060162782669067,
+      "learning_rate": 0.001,
+      "loss": 2.4099,
+      "step": 5890
+    },
+    {
+      "epoch": 0.2492173618749471,
+      "grad_norm": 0.2675623595714569,
+      "learning_rate": 0.001,
+      "loss": 2.5034,
+      "step": 5891
+    },
+    {
+      "epoch": 0.2492596666384635,
+      "grad_norm": 3.282320022583008,
+      "learning_rate": 0.001,
+      "loss": 1.659,
+      "step": 5892
+    },
+    {
+      "epoch": 0.24930197140197985,
+      "grad_norm": 0.28002822399139404,
+      "learning_rate": 0.001,
+      "loss": 2.2949,
+      "step": 5893
+    },
+    {
+      "epoch": 0.24934427616549623,
+      "grad_norm": 0.17490912973880768,
+      "learning_rate": 0.001,
+      "loss": 2.7314,
+      "step": 5894
+    },
+    {
+      "epoch": 0.2493865809290126,
+      "grad_norm": 0.5958810448646545,
+      "learning_rate": 0.001,
+      "loss": 2.6676,
+      "step": 5895
+    },
+    {
+      "epoch": 0.24942888569252897,
+      "grad_norm": 0.18583674728870392,
+      "learning_rate": 0.001,
+      "loss": 2.1634,
+      "step": 5896
+    },
+    {
+      "epoch": 0.24947119045604535,
+      "grad_norm": 0.2628713548183441,
+      "learning_rate": 0.001,
+      "loss": 2.3261,
+      "step": 5897
+    },
+    {
+      "epoch": 0.24951349521956173,
+      "grad_norm": 0.1902725249528885,
+      "learning_rate": 0.001,
+      "loss": 2.037,
+      "step": 5898
+    },
+    {
+      "epoch": 0.24955579998307809,
+      "grad_norm": 0.8447865843772888,
+      "learning_rate": 0.001,
+      "loss": 1.9191,
+      "step": 5899
+    },
+    {
+      "epoch": 0.24959810474659447,
+      "grad_norm": 0.25084027647972107,
+      "learning_rate": 0.001,
+      "loss": 2.6923,
+      "step": 5900
+    },
+    {
+      "epoch": 0.24964040951011085,
+      "grad_norm": 0.2469928115606308,
+      "learning_rate": 0.001,
+      "loss": 2.6025,
+      "step": 5901
+    },
+    {
+      "epoch": 0.2496827142736272,
+      "grad_norm": 0.18557626008987427,
+      "learning_rate": 0.001,
+      "loss": 1.8872,
+      "step": 5902
+    },
+    {
+      "epoch": 0.2497250190371436,
+      "grad_norm": 0.22026236355304718,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 5903
+    },
+    {
+      "epoch": 0.24976732380065994,
+      "grad_norm": 0.21095015108585358,
+      "learning_rate": 0.001,
+      "loss": 2.2376,
+      "step": 5904
+    },
+    {
+      "epoch": 0.24980962856417632,
+      "grad_norm": 0.18303923308849335,
+      "learning_rate": 0.001,
+      "loss": 3.1426,
+      "step": 5905
+    },
+    {
+      "epoch": 0.2498519333276927,
+      "grad_norm": 0.9932180643081665,
+      "learning_rate": 0.001,
+      "loss": 2.3002,
+      "step": 5906
+    },
+    {
+      "epoch": 0.24989423809120906,
+      "grad_norm": 0.2075134962797165,
+      "learning_rate": 0.001,
+      "loss": 2.2085,
+      "step": 5907
+    },
+    {
+      "epoch": 0.24993654285472544,
+      "grad_norm": 0.1895306408405304,
+      "learning_rate": 0.001,
+      "loss": 1.8905,
+      "step": 5908
+    },
+    {
+      "epoch": 0.24997884761824182,
+      "grad_norm": 0.2975296378135681,
+      "learning_rate": 0.001,
+      "loss": 3.0346,
+      "step": 5909
+    },
+    {
+      "epoch": 0.2500211523817582,
+      "grad_norm": 0.19449001550674438,
+      "learning_rate": 0.001,
+      "loss": 2.3004,
+      "step": 5910
+    },
+    {
+      "epoch": 0.25006345714527456,
+      "grad_norm": 0.24991470575332642,
+      "learning_rate": 0.001,
+      "loss": 3.1309,
+      "step": 5911
+    },
+    {
+      "epoch": 0.2501057619087909,
+      "grad_norm": 0.35687020421028137,
+      "learning_rate": 0.001,
+      "loss": 2.2765,
+      "step": 5912
+    },
+    {
+      "epoch": 0.2501480666723073,
+      "grad_norm": 0.17699697613716125,
+      "learning_rate": 0.001,
+      "loss": 2.4377,
+      "step": 5913
+    },
+    {
+      "epoch": 0.2501903714358237,
+      "grad_norm": 0.2586126923561096,
+      "learning_rate": 0.001,
+      "loss": 3.4109,
+      "step": 5914
+    },
+    {
+      "epoch": 0.25023267619934003,
+      "grad_norm": 0.21270103752613068,
+      "learning_rate": 0.001,
+      "loss": 1.8928,
+      "step": 5915
+    },
+    {
+      "epoch": 0.25027498096285644,
+      "grad_norm": 0.225282222032547,
+      "learning_rate": 0.001,
+      "loss": 2.3448,
+      "step": 5916
+    },
+    {
+      "epoch": 0.2503172857263728,
+      "grad_norm": 0.18407504260540009,
+      "learning_rate": 0.001,
+      "loss": 2.0276,
+      "step": 5917
+    },
+    {
+      "epoch": 0.25035959048988915,
+      "grad_norm": 0.22186896204948425,
+      "learning_rate": 0.001,
+      "loss": 2.0891,
+      "step": 5918
+    },
+    {
+      "epoch": 0.25040189525340556,
+      "grad_norm": 2.6034348011016846,
+      "learning_rate": 0.001,
+      "loss": 2.0726,
+      "step": 5919
+    },
+    {
+      "epoch": 0.2504442000169219,
+      "grad_norm": 0.20303837954998016,
+      "learning_rate": 0.001,
+      "loss": 2.3935,
+      "step": 5920
+    },
+    {
+      "epoch": 0.25048650478043827,
+      "grad_norm": 0.18995042145252228,
+      "learning_rate": 0.001,
+      "loss": 2.0791,
+      "step": 5921
+    },
+    {
+      "epoch": 0.2505288095439546,
+      "grad_norm": 0.17359282076358795,
+      "learning_rate": 0.001,
+      "loss": 2.0846,
+      "step": 5922
+    },
+    {
+      "epoch": 0.25057111430747103,
+      "grad_norm": 0.2554149031639099,
+      "learning_rate": 0.001,
+      "loss": 1.8015,
+      "step": 5923
+    },
+    {
+      "epoch": 0.2506134190709874,
+      "grad_norm": 0.4858115315437317,
+      "learning_rate": 0.001,
+      "loss": 2.3795,
+      "step": 5924
+    },
+    {
+      "epoch": 0.25065572383450374,
+      "grad_norm": 0.3626372516155243,
+      "learning_rate": 0.001,
+      "loss": 2.0548,
+      "step": 5925
+    },
+    {
+      "epoch": 0.25069802859802015,
+      "grad_norm": 0.17694664001464844,
+      "learning_rate": 0.001,
+      "loss": 1.7527,
+      "step": 5926
+    },
+    {
+      "epoch": 0.2507403333615365,
+      "grad_norm": 0.5203558206558228,
+      "learning_rate": 0.001,
+      "loss": 3.2289,
+      "step": 5927
+    },
+    {
+      "epoch": 0.25078263812505286,
+      "grad_norm": 0.1976146101951599,
+      "learning_rate": 0.001,
+      "loss": 2.4577,
+      "step": 5928
+    },
+    {
+      "epoch": 0.25082494288856927,
+      "grad_norm": 0.2331826388835907,
+      "learning_rate": 0.001,
+      "loss": 3.3164,
+      "step": 5929
+    },
+    {
+      "epoch": 0.2508672476520856,
+      "grad_norm": 0.4622613787651062,
+      "learning_rate": 0.001,
+      "loss": 1.578,
+      "step": 5930
+    },
+    {
+      "epoch": 0.250909552415602,
+      "grad_norm": 0.20590519905090332,
+      "learning_rate": 0.001,
+      "loss": 2.2907,
+      "step": 5931
+    },
+    {
+      "epoch": 0.2509518571791184,
+      "grad_norm": 0.20743978023529053,
+      "learning_rate": 0.001,
+      "loss": 2.0356,
+      "step": 5932
+    },
+    {
+      "epoch": 0.25099416194263474,
+      "grad_norm": 4.0422892570495605,
+      "learning_rate": 0.001,
+      "loss": 2.584,
+      "step": 5933
+    },
+    {
+      "epoch": 0.2510364667061511,
+      "grad_norm": 0.4900633990764618,
+      "learning_rate": 0.001,
+      "loss": 3.1356,
+      "step": 5934
+    },
+    {
+      "epoch": 0.2510787714696675,
+      "grad_norm": 0.19140180945396423,
+      "learning_rate": 0.001,
+      "loss": 2.1572,
+      "step": 5935
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.2231162041425705,
+      "learning_rate": 0.001,
+      "loss": 2.8386,
+      "step": 5936
+    },
+    {
+      "epoch": 0.2511633809967002,
+      "grad_norm": 1.3613148927688599,
+      "learning_rate": 0.001,
+      "loss": 2.2125,
+      "step": 5937
+    },
+    {
+      "epoch": 0.2512056857602166,
+      "grad_norm": 0.23480670154094696,
+      "learning_rate": 0.001,
+      "loss": 2.3708,
+      "step": 5938
+    },
+    {
+      "epoch": 0.251247990523733,
+      "grad_norm": 0.5639317035675049,
+      "learning_rate": 0.001,
+      "loss": 2.2903,
+      "step": 5939
+    },
+    {
+      "epoch": 0.25129029528724933,
+      "grad_norm": 1.7350952625274658,
+      "learning_rate": 0.001,
+      "loss": 2.4496,
+      "step": 5940
+    },
+    {
+      "epoch": 0.25133260005076574,
+      "grad_norm": 0.1928754597902298,
+      "learning_rate": 0.001,
+      "loss": 2.6528,
+      "step": 5941
+    },
+    {
+      "epoch": 0.2513749048142821,
+      "grad_norm": 0.27500829100608826,
+      "learning_rate": 0.001,
+      "loss": 2.7894,
+      "step": 5942
+    },
+    {
+      "epoch": 0.25141720957779845,
+      "grad_norm": 0.2504851520061493,
+      "learning_rate": 0.001,
+      "loss": 3.2059,
+      "step": 5943
+    },
+    {
+      "epoch": 0.2514595143413148,
+      "grad_norm": 0.45656251907348633,
+      "learning_rate": 0.001,
+      "loss": 2.6401,
+      "step": 5944
+    },
+    {
+      "epoch": 0.2515018191048312,
+      "grad_norm": 0.2430989146232605,
+      "learning_rate": 0.001,
+      "loss": 2.3806,
+      "step": 5945
+    },
+    {
+      "epoch": 0.25154412386834757,
+      "grad_norm": 0.44327855110168457,
+      "learning_rate": 0.001,
+      "loss": 1.9777,
+      "step": 5946
+    },
+    {
+      "epoch": 0.2515864286318639,
+      "grad_norm": 0.2197687029838562,
+      "learning_rate": 0.001,
+      "loss": 2.3825,
+      "step": 5947
+    },
+    {
+      "epoch": 0.25162873339538033,
+      "grad_norm": 0.24564971029758453,
+      "learning_rate": 0.001,
+      "loss": 2.022,
+      "step": 5948
+    },
+    {
+      "epoch": 0.2516710381588967,
+      "grad_norm": 0.41810470819473267,
+      "learning_rate": 0.001,
+      "loss": 2.4189,
+      "step": 5949
+    },
+    {
+      "epoch": 0.25171334292241304,
+      "grad_norm": 0.2302566021680832,
+      "learning_rate": 0.001,
+      "loss": 2.1172,
+      "step": 5950
+    },
+    {
+      "epoch": 0.25175564768592945,
+      "grad_norm": 0.5698367357254028,
+      "learning_rate": 0.001,
+      "loss": 1.8955,
+      "step": 5951
+    },
+    {
+      "epoch": 0.2517979524494458,
+      "grad_norm": 0.20264704525470734,
+      "learning_rate": 0.001,
+      "loss": 2.55,
+      "step": 5952
+    },
+    {
+      "epoch": 0.25184025721296216,
+      "grad_norm": 0.19961802661418915,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 5953
+    },
+    {
+      "epoch": 0.25188256197647857,
+      "grad_norm": 0.24171172082424164,
+      "learning_rate": 0.001,
+      "loss": 2.9635,
+      "step": 5954
+    },
+    {
+      "epoch": 0.2519248667399949,
+      "grad_norm": 0.19584928452968597,
+      "learning_rate": 0.001,
+      "loss": 2.0816,
+      "step": 5955
+    },
+    {
+      "epoch": 0.2519671715035113,
+      "grad_norm": 0.7652029991149902,
+      "learning_rate": 0.001,
+      "loss": 2.8902,
+      "step": 5956
+    },
+    {
+      "epoch": 0.2520094762670277,
+      "grad_norm": 0.22400104999542236,
+      "learning_rate": 0.001,
+      "loss": 2.0901,
+      "step": 5957
+    },
+    {
+      "epoch": 0.25205178103054404,
+      "grad_norm": 0.21903732419013977,
+      "learning_rate": 0.001,
+      "loss": 2.0618,
+      "step": 5958
+    },
+    {
+      "epoch": 0.2520940857940604,
+      "grad_norm": 0.24086321890354156,
+      "learning_rate": 0.001,
+      "loss": 2.9158,
+      "step": 5959
+    },
+    {
+      "epoch": 0.2521363905575768,
+      "grad_norm": 2.2688310146331787,
+      "learning_rate": 0.001,
+      "loss": 2.2415,
+      "step": 5960
+    },
+    {
+      "epoch": 0.25217869532109316,
+      "grad_norm": 0.2769038677215576,
+      "learning_rate": 0.001,
+      "loss": 2.5179,
+      "step": 5961
+    },
+    {
+      "epoch": 0.2522210000846095,
+      "grad_norm": 0.19732657074928284,
+      "learning_rate": 0.001,
+      "loss": 1.5416,
+      "step": 5962
+    },
+    {
+      "epoch": 0.2522633048481259,
+      "grad_norm": 0.21410557627677917,
+      "learning_rate": 0.001,
+      "loss": 1.9668,
+      "step": 5963
+    },
+    {
+      "epoch": 0.2523056096116423,
+      "grad_norm": 0.20264627039432526,
+      "learning_rate": 0.001,
+      "loss": 1.8539,
+      "step": 5964
+    },
+    {
+      "epoch": 0.25234791437515863,
+      "grad_norm": 0.9442425966262817,
+      "learning_rate": 0.001,
+      "loss": 2.2133,
+      "step": 5965
+    },
+    {
+      "epoch": 0.252390219138675,
+      "grad_norm": 0.17984417080879211,
+      "learning_rate": 0.001,
+      "loss": 1.9855,
+      "step": 5966
+    },
+    {
+      "epoch": 0.2524325239021914,
+      "grad_norm": 0.3779272735118866,
+      "learning_rate": 0.001,
+      "loss": 3.2348,
+      "step": 5967
+    },
+    {
+      "epoch": 0.25247482866570775,
+      "grad_norm": 0.23306319117546082,
+      "learning_rate": 0.001,
+      "loss": 2.9164,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2525171334292241,
+      "grad_norm": 0.21737343072891235,
+      "learning_rate": 0.001,
+      "loss": 2.4458,
+      "step": 5969
+    },
+    {
+      "epoch": 0.2525594381927405,
+      "grad_norm": 0.38391655683517456,
+      "learning_rate": 0.001,
+      "loss": 1.901,
+      "step": 5970
+    },
+    {
+      "epoch": 0.25260174295625687,
+      "grad_norm": 0.21682848036289215,
+      "learning_rate": 0.001,
+      "loss": 2.745,
+      "step": 5971
+    },
+    {
+      "epoch": 0.2526440477197732,
+      "grad_norm": 0.1871444135904312,
+      "learning_rate": 0.001,
+      "loss": 2.6208,
+      "step": 5972
+    },
+    {
+      "epoch": 0.25268635248328963,
+      "grad_norm": 0.23921331763267517,
+      "learning_rate": 0.001,
+      "loss": 2.7021,
+      "step": 5973
+    },
+    {
+      "epoch": 0.252728657246806,
+      "grad_norm": 0.17137141525745392,
+      "learning_rate": 0.001,
+      "loss": 2.0486,
+      "step": 5974
+    },
+    {
+      "epoch": 0.25277096201032234,
+      "grad_norm": 0.5864259600639343,
+      "learning_rate": 0.001,
+      "loss": 1.7707,
+      "step": 5975
+    },
+    {
+      "epoch": 0.25281326677383875,
+      "grad_norm": 0.18779319524765015,
+      "learning_rate": 0.001,
+      "loss": 1.8549,
+      "step": 5976
+    },
+    {
+      "epoch": 0.2528555715373551,
+      "grad_norm": 0.2662065625190735,
+      "learning_rate": 0.001,
+      "loss": 2.9272,
+      "step": 5977
+    },
+    {
+      "epoch": 0.25289787630087146,
+      "grad_norm": 0.23567529022693634,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 5978
+    },
+    {
+      "epoch": 0.25294018106438787,
+      "grad_norm": 0.1899128556251526,
+      "learning_rate": 0.001,
+      "loss": 1.992,
+      "step": 5979
+    },
+    {
+      "epoch": 0.2529824858279042,
+      "grad_norm": 0.23404709994792938,
+      "learning_rate": 0.001,
+      "loss": 2.3842,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2530247905914206,
+      "grad_norm": 0.5312284827232361,
+      "learning_rate": 0.001,
+      "loss": 2.2135,
+      "step": 5981
+    },
+    {
+      "epoch": 0.253067095354937,
+      "grad_norm": 0.2125319540500641,
+      "learning_rate": 0.001,
+      "loss": 2.3937,
+      "step": 5982
+    },
+    {
+      "epoch": 0.25310940011845334,
+      "grad_norm": 0.20524631440639496,
+      "learning_rate": 0.001,
+      "loss": 2.5639,
+      "step": 5983
+    },
+    {
+      "epoch": 0.2531517048819697,
+      "grad_norm": 0.20020252466201782,
+      "learning_rate": 0.001,
+      "loss": 2.6334,
+      "step": 5984
+    },
+    {
+      "epoch": 0.2531940096454861,
+      "grad_norm": 0.2192360907793045,
+      "learning_rate": 0.001,
+      "loss": 2.2578,
+      "step": 5985
+    },
+    {
+      "epoch": 0.25323631440900246,
+      "grad_norm": 0.16975408792495728,
+      "learning_rate": 0.001,
+      "loss": 1.8028,
+      "step": 5986
+    },
+    {
+      "epoch": 0.2532786191725188,
+      "grad_norm": 1.4105056524276733,
+      "learning_rate": 0.001,
+      "loss": 3.1986,
+      "step": 5987
+    },
+    {
+      "epoch": 0.2533209239360352,
+      "grad_norm": 1.0389015674591064,
+      "learning_rate": 0.001,
+      "loss": 2.7766,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2533632286995516,
+      "grad_norm": 1.0410923957824707,
+      "learning_rate": 0.001,
+      "loss": 2.4953,
+      "step": 5989
+    },
+    {
+      "epoch": 0.25340553346306793,
+      "grad_norm": 0.24617451429367065,
+      "learning_rate": 0.001,
+      "loss": 1.7532,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2534478382265843,
+      "grad_norm": 0.22218583524227142,
+      "learning_rate": 0.001,
+      "loss": 1.8373,
+      "step": 5991
+    },
+    {
+      "epoch": 0.2534901429901007,
+      "grad_norm": 0.3823182284832001,
+      "learning_rate": 0.001,
+      "loss": 1.8441,
+      "step": 5992
+    },
+    {
+      "epoch": 0.25353244775361705,
+      "grad_norm": 0.2810841202735901,
+      "learning_rate": 0.001,
+      "loss": 1.8284,
+      "step": 5993
+    },
+    {
+      "epoch": 0.2535747525171334,
+      "grad_norm": 0.21739636361598969,
+      "learning_rate": 0.001,
+      "loss": 2.66,
+      "step": 5994
+    },
+    {
+      "epoch": 0.2536170572806498,
+      "grad_norm": 0.8973679542541504,
+      "learning_rate": 0.001,
+      "loss": 2.4392,
+      "step": 5995
+    },
+    {
+      "epoch": 0.25365936204416617,
+      "grad_norm": 0.24056527018547058,
+      "learning_rate": 0.001,
+      "loss": 2.4884,
+      "step": 5996
+    },
+    {
+      "epoch": 0.2537016668076825,
+      "grad_norm": 0.7096660733222961,
+      "learning_rate": 0.001,
+      "loss": 2.4973,
+      "step": 5997
+    },
+    {
+      "epoch": 0.25374397157119893,
+      "grad_norm": 3.935514450073242,
+      "learning_rate": 0.001,
+      "loss": 2.5481,
+      "step": 5998
+    },
+    {
+      "epoch": 0.2537862763347153,
+      "grad_norm": 0.22361649572849274,
+      "learning_rate": 0.001,
+      "loss": 1.8946,
+      "step": 5999
+    },
+    {
+      "epoch": 0.25382858109823164,
+      "grad_norm": 1.974503755569458,
+      "learning_rate": 0.001,
+      "loss": 2.4045,
+      "step": 6000
+    },
+    {
+      "epoch": 0.25387088586174805,
+      "grad_norm": 0.44096559286117554,
+      "learning_rate": 0.001,
+      "loss": 2.4101,
+      "step": 6001
+    },
+    {
+      "epoch": 0.2539131906252644,
+      "grad_norm": 0.4442439079284668,
+      "learning_rate": 0.001,
+      "loss": 2.372,
+      "step": 6002
+    },
+    {
+      "epoch": 0.25395549538878076,
+      "grad_norm": 0.24630646407604218,
+      "learning_rate": 0.001,
+      "loss": 2.3218,
+      "step": 6003
+    },
+    {
+      "epoch": 0.25399780015229717,
+      "grad_norm": 0.32622864842414856,
+      "learning_rate": 0.001,
+      "loss": 3.0714,
+      "step": 6004
+    },
+    {
+      "epoch": 0.2540401049158135,
+      "grad_norm": 0.2776721119880676,
+      "learning_rate": 0.001,
+      "loss": 2.2819,
+      "step": 6005
+    },
+    {
+      "epoch": 0.2540824096793299,
+      "grad_norm": 0.22410140931606293,
+      "learning_rate": 0.001,
+      "loss": 2.2823,
+      "step": 6006
+    },
+    {
+      "epoch": 0.2541247144428463,
+      "grad_norm": 0.7463712096214294,
+      "learning_rate": 0.001,
+      "loss": 2.9199,
+      "step": 6007
+    },
+    {
+      "epoch": 0.25416701920636264,
+      "grad_norm": 0.3214509189128876,
+      "learning_rate": 0.001,
+      "loss": 3.1198,
+      "step": 6008
+    },
+    {
+      "epoch": 0.254209323969879,
+      "grad_norm": 0.2701822817325592,
+      "learning_rate": 0.001,
+      "loss": 1.9467,
+      "step": 6009
+    },
+    {
+      "epoch": 0.2542516287333954,
+      "grad_norm": 0.28340399265289307,
+      "learning_rate": 0.001,
+      "loss": 1.888,
+      "step": 6010
+    },
+    {
+      "epoch": 0.25429393349691176,
+      "grad_norm": 0.3803146779537201,
+      "learning_rate": 0.001,
+      "loss": 3.7349,
+      "step": 6011
+    },
+    {
+      "epoch": 0.2543362382604281,
+      "grad_norm": 0.2079993635416031,
+      "learning_rate": 0.001,
+      "loss": 2.412,
+      "step": 6012
+    },
+    {
+      "epoch": 0.25437854302394447,
+      "grad_norm": 0.2278539538383484,
+      "learning_rate": 0.001,
+      "loss": 2.1716,
+      "step": 6013
+    },
+    {
+      "epoch": 0.2544208477874609,
+      "grad_norm": 0.23987050354480743,
+      "learning_rate": 0.001,
+      "loss": 2.7047,
+      "step": 6014
+    },
+    {
+      "epoch": 0.25446315255097723,
+      "grad_norm": 0.32305946946144104,
+      "learning_rate": 0.001,
+      "loss": 2.1134,
+      "step": 6015
+    },
+    {
+      "epoch": 0.2545054573144936,
+      "grad_norm": 0.3197750449180603,
+      "learning_rate": 0.001,
+      "loss": 2.5648,
+      "step": 6016
+    },
+    {
+      "epoch": 0.25454776207801,
+      "grad_norm": 0.2103656828403473,
+      "learning_rate": 0.001,
+      "loss": 2.1189,
+      "step": 6017
+    },
+    {
+      "epoch": 0.25459006684152635,
+      "grad_norm": 0.6218228936195374,
+      "learning_rate": 0.001,
+      "loss": 1.9693,
+      "step": 6018
+    },
+    {
+      "epoch": 0.2546323716050427,
+      "grad_norm": 0.6921700835227966,
+      "learning_rate": 0.001,
+      "loss": 2.4787,
+      "step": 6019
+    },
+    {
+      "epoch": 0.2546746763685591,
+      "grad_norm": 0.20332172513008118,
+      "learning_rate": 0.001,
+      "loss": 1.9037,
+      "step": 6020
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.5180513262748718,
+      "learning_rate": 0.001,
+      "loss": 2.5619,
+      "step": 6021
+    },
+    {
+      "epoch": 0.2547592858955918,
+      "grad_norm": 0.199415922164917,
+      "learning_rate": 0.001,
+      "loss": 2.1282,
+      "step": 6022
+    },
+    {
+      "epoch": 0.25480159065910823,
+      "grad_norm": 0.20138099789619446,
+      "learning_rate": 0.001,
+      "loss": 1.8072,
+      "step": 6023
+    },
+    {
+      "epoch": 0.2548438954226246,
+      "grad_norm": 0.24892349541187286,
+      "learning_rate": 0.001,
+      "loss": 2.2655,
+      "step": 6024
+    },
+    {
+      "epoch": 0.25488620018614094,
+      "grad_norm": 0.4697567820549011,
+      "learning_rate": 0.001,
+      "loss": 2.9607,
+      "step": 6025
+    },
+    {
+      "epoch": 0.25492850494965735,
+      "grad_norm": 0.21022146940231323,
+      "learning_rate": 0.001,
+      "loss": 2.0065,
+      "step": 6026
+    },
+    {
+      "epoch": 0.2549708097131737,
+      "grad_norm": 2.630333423614502,
+      "learning_rate": 0.001,
+      "loss": 1.862,
+      "step": 6027
+    },
+    {
+      "epoch": 0.25501311447669006,
+      "grad_norm": 0.2536892890930176,
+      "learning_rate": 0.001,
+      "loss": 2.4433,
+      "step": 6028
+    },
+    {
+      "epoch": 0.25505541924020647,
+      "grad_norm": 0.212021142244339,
+      "learning_rate": 0.001,
+      "loss": 2.0887,
+      "step": 6029
+    },
+    {
+      "epoch": 0.2550977240037228,
+      "grad_norm": 0.2172902524471283,
+      "learning_rate": 0.001,
+      "loss": 1.958,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2551400287672392,
+      "grad_norm": 0.22086498141288757,
+      "learning_rate": 0.001,
+      "loss": 2.5838,
+      "step": 6031
+    },
+    {
+      "epoch": 0.2551823335307556,
+      "grad_norm": 0.2510300576686859,
+      "learning_rate": 0.001,
+      "loss": 1.6274,
+      "step": 6032
+    },
+    {
+      "epoch": 0.25522463829427194,
+      "grad_norm": 1.1860932111740112,
+      "learning_rate": 0.001,
+      "loss": 3.0628,
+      "step": 6033
+    },
+    {
+      "epoch": 0.2552669430577883,
+      "grad_norm": 7.115164756774902,
+      "learning_rate": 0.001,
+      "loss": 3.3898,
+      "step": 6034
+    },
+    {
+      "epoch": 0.25530924782130465,
+      "grad_norm": 0.2606843411922455,
+      "learning_rate": 0.001,
+      "loss": 2.2712,
+      "step": 6035
+    },
+    {
+      "epoch": 0.25535155258482106,
+      "grad_norm": 0.25481027364730835,
+      "learning_rate": 0.001,
+      "loss": 2.5696,
+      "step": 6036
+    },
+    {
+      "epoch": 0.2553938573483374,
+      "grad_norm": 0.19426991045475006,
+      "learning_rate": 0.001,
+      "loss": 1.8278,
+      "step": 6037
+    },
+    {
+      "epoch": 0.25543616211185377,
+      "grad_norm": 0.23399263620376587,
+      "learning_rate": 0.001,
+      "loss": 2.0001,
+      "step": 6038
+    },
+    {
+      "epoch": 0.2554784668753702,
+      "grad_norm": 0.2326452136039734,
+      "learning_rate": 0.001,
+      "loss": 1.9532,
+      "step": 6039
+    },
+    {
+      "epoch": 0.25552077163888653,
+      "grad_norm": 0.5606713891029358,
+      "learning_rate": 0.001,
+      "loss": 2.2031,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2555630764024029,
+      "grad_norm": 1.2523069381713867,
+      "learning_rate": 0.001,
+      "loss": 1.9169,
+      "step": 6041
+    },
+    {
+      "epoch": 0.2556053811659193,
+      "grad_norm": 0.21582403779029846,
+      "learning_rate": 0.001,
+      "loss": 2.3978,
+      "step": 6042
+    },
+    {
+      "epoch": 0.25564768592943565,
+      "grad_norm": 0.3164418339729309,
+      "learning_rate": 0.001,
+      "loss": 2.7011,
+      "step": 6043
+    },
+    {
+      "epoch": 0.255689990692952,
+      "grad_norm": 1.014667272567749,
+      "learning_rate": 0.001,
+      "loss": 2.3931,
+      "step": 6044
+    },
+    {
+      "epoch": 0.2557322954564684,
+      "grad_norm": 0.2742740213871002,
+      "learning_rate": 0.001,
+      "loss": 2.705,
+      "step": 6045
+    },
+    {
+      "epoch": 0.25577460021998477,
+      "grad_norm": 0.3064813017845154,
+      "learning_rate": 0.001,
+      "loss": 2.5047,
+      "step": 6046
+    },
+    {
+      "epoch": 0.2558169049835011,
+      "grad_norm": 0.2514945864677429,
+      "learning_rate": 0.001,
+      "loss": 1.743,
+      "step": 6047
+    },
+    {
+      "epoch": 0.25585920974701754,
+      "grad_norm": 0.48261088132858276,
+      "learning_rate": 0.001,
+      "loss": 2.7783,
+      "step": 6048
+    },
+    {
+      "epoch": 0.2559015145105339,
+      "grad_norm": 0.2113606333732605,
+      "learning_rate": 0.001,
+      "loss": 2.6006,
+      "step": 6049
+    },
+    {
+      "epoch": 0.25594381927405024,
+      "grad_norm": 0.21159514784812927,
+      "learning_rate": 0.001,
+      "loss": 1.9746,
+      "step": 6050
+    },
+    {
+      "epoch": 0.25598612403756665,
+      "grad_norm": 0.30281734466552734,
+      "learning_rate": 0.001,
+      "loss": 2.6729,
+      "step": 6051
+    },
+    {
+      "epoch": 0.256028428801083,
+      "grad_norm": 0.19452209770679474,
+      "learning_rate": 0.001,
+      "loss": 1.9967,
+      "step": 6052
+    },
+    {
+      "epoch": 0.25607073356459936,
+      "grad_norm": 0.19158554077148438,
+      "learning_rate": 0.001,
+      "loss": 2.0234,
+      "step": 6053
+    },
+    {
+      "epoch": 0.25611303832811577,
+      "grad_norm": 0.3292371332645416,
+      "learning_rate": 0.001,
+      "loss": 1.8973,
+      "step": 6054
+    },
+    {
+      "epoch": 0.2561553430916321,
+      "grad_norm": 0.1782860904932022,
+      "learning_rate": 0.001,
+      "loss": 1.786,
+      "step": 6055
+    },
+    {
+      "epoch": 0.2561976478551485,
+      "grad_norm": 0.2003399133682251,
+      "learning_rate": 0.001,
+      "loss": 2.1325,
+      "step": 6056
+    },
+    {
+      "epoch": 0.25623995261866483,
+      "grad_norm": 3.0092291831970215,
+      "learning_rate": 0.001,
+      "loss": 2.6524,
+      "step": 6057
+    },
+    {
+      "epoch": 0.25628225738218124,
+      "grad_norm": 0.21355977654457092,
+      "learning_rate": 0.001,
+      "loss": 2.2556,
+      "step": 6058
+    },
+    {
+      "epoch": 0.2563245621456976,
+      "grad_norm": 0.2934946119785309,
+      "learning_rate": 0.001,
+      "loss": 2.8166,
+      "step": 6059
+    },
+    {
+      "epoch": 0.25636686690921395,
+      "grad_norm": 0.1829323172569275,
+      "learning_rate": 0.001,
+      "loss": 1.8922,
+      "step": 6060
+    },
+    {
+      "epoch": 0.25640917167273036,
+      "grad_norm": 0.2807469069957733,
+      "learning_rate": 0.001,
+      "loss": 2.3787,
+      "step": 6061
+    },
+    {
+      "epoch": 0.2564514764362467,
+      "grad_norm": 0.22693084180355072,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 6062
+    },
+    {
+      "epoch": 0.25649378119976307,
+      "grad_norm": 0.1689741611480713,
+      "learning_rate": 0.001,
+      "loss": 2.1221,
+      "step": 6063
+    },
+    {
+      "epoch": 0.2565360859632795,
+      "grad_norm": 1.339165449142456,
+      "learning_rate": 0.001,
+      "loss": 1.8906,
+      "step": 6064
+    },
+    {
+      "epoch": 0.25657839072679584,
+      "grad_norm": 1.1260621547698975,
+      "learning_rate": 0.001,
+      "loss": 2.2244,
+      "step": 6065
+    },
+    {
+      "epoch": 0.2566206954903122,
+      "grad_norm": 0.2071322202682495,
+      "learning_rate": 0.001,
+      "loss": 2.0834,
+      "step": 6066
+    },
+    {
+      "epoch": 0.2566630002538286,
+      "grad_norm": 0.2302442193031311,
+      "learning_rate": 0.001,
+      "loss": 1.8294,
+      "step": 6067
+    },
+    {
+      "epoch": 0.25670530501734495,
+      "grad_norm": 0.4180930554866791,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 6068
+    },
+    {
+      "epoch": 0.2567476097808613,
+      "grad_norm": 0.2579159736633301,
+      "learning_rate": 0.001,
+      "loss": 2.7813,
+      "step": 6069
+    },
+    {
+      "epoch": 0.2567899145443777,
+      "grad_norm": 0.3067325949668884,
+      "learning_rate": 0.001,
+      "loss": 2.3948,
+      "step": 6070
+    },
+    {
+      "epoch": 0.25683221930789407,
+      "grad_norm": 0.2110936939716339,
+      "learning_rate": 0.001,
+      "loss": 2.5168,
+      "step": 6071
+    },
+    {
+      "epoch": 0.2568745240714104,
+      "grad_norm": 0.4867793023586273,
+      "learning_rate": 0.001,
+      "loss": 3.3979,
+      "step": 6072
+    },
+    {
+      "epoch": 0.25691682883492684,
+      "grad_norm": 0.22111758589744568,
+      "learning_rate": 0.001,
+      "loss": 1.9811,
+      "step": 6073
+    },
+    {
+      "epoch": 0.2569591335984432,
+      "grad_norm": 0.7423004508018494,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 6074
+    },
+    {
+      "epoch": 0.25700143836195954,
+      "grad_norm": 0.32565468549728394,
+      "learning_rate": 0.001,
+      "loss": 1.619,
+      "step": 6075
+    },
+    {
+      "epoch": 0.25704374312547595,
+      "grad_norm": 5.436065673828125,
+      "learning_rate": 0.001,
+      "loss": 2.3115,
+      "step": 6076
+    },
+    {
+      "epoch": 0.2570860478889923,
+      "grad_norm": 0.3309813141822815,
+      "learning_rate": 0.001,
+      "loss": 2.1493,
+      "step": 6077
+    },
+    {
+      "epoch": 0.25712835265250866,
+      "grad_norm": 0.7553781270980835,
+      "learning_rate": 0.001,
+      "loss": 3.0408,
+      "step": 6078
+    },
+    {
+      "epoch": 0.257170657416025,
+      "grad_norm": 0.20468690991401672,
+      "learning_rate": 0.001,
+      "loss": 2.1414,
+      "step": 6079
+    },
+    {
+      "epoch": 0.2572129621795414,
+      "grad_norm": 3.904078483581543,
+      "learning_rate": 0.001,
+      "loss": 2.7939,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2572552669430578,
+      "grad_norm": 0.3017195165157318,
+      "learning_rate": 0.001,
+      "loss": 2.5183,
+      "step": 6081
+    },
+    {
+      "epoch": 0.25729757170657414,
+      "grad_norm": 0.3187876343727112,
+      "learning_rate": 0.001,
+      "loss": 3.2043,
+      "step": 6082
+    },
+    {
+      "epoch": 0.25733987647009055,
+      "grad_norm": 0.29611068964004517,
+      "learning_rate": 0.001,
+      "loss": 2.7729,
+      "step": 6083
+    },
+    {
+      "epoch": 0.2573821812336069,
+      "grad_norm": 0.26517346501350403,
+      "learning_rate": 0.001,
+      "loss": 3.1814,
+      "step": 6084
+    },
+    {
+      "epoch": 0.25742448599712325,
+      "grad_norm": 0.22469617426395416,
+      "learning_rate": 0.001,
+      "loss": 2.2341,
+      "step": 6085
+    },
+    {
+      "epoch": 0.25746679076063966,
+      "grad_norm": 0.7284866571426392,
+      "learning_rate": 0.001,
+      "loss": 2.2107,
+      "step": 6086
+    },
+    {
+      "epoch": 0.257509095524156,
+      "grad_norm": 0.35012027621269226,
+      "learning_rate": 0.001,
+      "loss": 2.9378,
+      "step": 6087
+    },
+    {
+      "epoch": 0.25755140028767237,
+      "grad_norm": 0.23448063433170319,
+      "learning_rate": 0.001,
+      "loss": 2.6047,
+      "step": 6088
+    },
+    {
+      "epoch": 0.2575937050511888,
+      "grad_norm": 0.185800239443779,
+      "learning_rate": 0.001,
+      "loss": 3.7792,
+      "step": 6089
+    },
+    {
+      "epoch": 0.25763600981470514,
+      "grad_norm": 0.8078601956367493,
+      "learning_rate": 0.001,
+      "loss": 2.4504,
+      "step": 6090
+    },
+    {
+      "epoch": 0.2576783145782215,
+      "grad_norm": 0.22115640342235565,
+      "learning_rate": 0.001,
+      "loss": 2.1107,
+      "step": 6091
+    },
+    {
+      "epoch": 0.2577206193417379,
+      "grad_norm": 0.2682878375053406,
+      "learning_rate": 0.001,
+      "loss": 2.3305,
+      "step": 6092
+    },
+    {
+      "epoch": 0.25776292410525425,
+      "grad_norm": 1.0689104795455933,
+      "learning_rate": 0.001,
+      "loss": 2.6872,
+      "step": 6093
+    },
+    {
+      "epoch": 0.2578052288687706,
+      "grad_norm": 0.22226876020431519,
+      "learning_rate": 0.001,
+      "loss": 2.9591,
+      "step": 6094
+    },
+    {
+      "epoch": 0.257847533632287,
+      "grad_norm": 0.2324611246585846,
+      "learning_rate": 0.001,
+      "loss": 1.431,
+      "step": 6095
+    },
+    {
+      "epoch": 0.2578898383958034,
+      "grad_norm": 0.2755308449268341,
+      "learning_rate": 0.001,
+      "loss": 2.6437,
+      "step": 6096
+    },
+    {
+      "epoch": 0.2579321431593197,
+      "grad_norm": 0.39930108189582825,
+      "learning_rate": 0.001,
+      "loss": 2.093,
+      "step": 6097
+    },
+    {
+      "epoch": 0.25797444792283614,
+      "grad_norm": 0.5387210845947266,
+      "learning_rate": 0.001,
+      "loss": 4.0456,
+      "step": 6098
+    },
+    {
+      "epoch": 0.2580167526863525,
+      "grad_norm": 0.5283278822898865,
+      "learning_rate": 0.001,
+      "loss": 2.4854,
+      "step": 6099
+    },
+    {
+      "epoch": 0.25805905744986884,
+      "grad_norm": 0.313729465007782,
+      "learning_rate": 0.001,
+      "loss": 2.2843,
+      "step": 6100
+    },
+    {
+      "epoch": 0.25810136221338525,
+      "grad_norm": 0.21948851644992828,
+      "learning_rate": 0.001,
+      "loss": 2.1515,
+      "step": 6101
+    },
+    {
+      "epoch": 0.2581436669769016,
+      "grad_norm": 0.21910326182842255,
+      "learning_rate": 0.001,
+      "loss": 2.5378,
+      "step": 6102
+    },
+    {
+      "epoch": 0.25818597174041796,
+      "grad_norm": 0.22086754441261292,
+      "learning_rate": 0.001,
+      "loss": 2.3505,
+      "step": 6103
+    },
+    {
+      "epoch": 0.2582282765039343,
+      "grad_norm": 0.23505359888076782,
+      "learning_rate": 0.001,
+      "loss": 2.2017,
+      "step": 6104
+    },
+    {
+      "epoch": 0.2582705812674507,
+      "grad_norm": 0.23428547382354736,
+      "learning_rate": 0.001,
+      "loss": 1.8068,
+      "step": 6105
+    },
+    {
+      "epoch": 0.2583128860309671,
+      "grad_norm": 0.16871504485607147,
+      "learning_rate": 0.001,
+      "loss": 1.9511,
+      "step": 6106
+    },
+    {
+      "epoch": 0.25835519079448344,
+      "grad_norm": 0.19622443616390228,
+      "learning_rate": 0.001,
+      "loss": 2.0708,
+      "step": 6107
+    },
+    {
+      "epoch": 0.25839749555799985,
+      "grad_norm": 0.2113541215658188,
+      "learning_rate": 0.001,
+      "loss": 2.1059,
+      "step": 6108
+    },
+    {
+      "epoch": 0.2584398003215162,
+      "grad_norm": 0.2662128806114197,
+      "learning_rate": 0.001,
+      "loss": 2.7546,
+      "step": 6109
+    },
+    {
+      "epoch": 0.25848210508503255,
+      "grad_norm": 0.18668299913406372,
+      "learning_rate": 0.001,
+      "loss": 1.9433,
+      "step": 6110
+    },
+    {
+      "epoch": 0.25852440984854896,
+      "grad_norm": 0.31358543038368225,
+      "learning_rate": 0.001,
+      "loss": 1.6306,
+      "step": 6111
+    },
+    {
+      "epoch": 0.2585667146120653,
+      "grad_norm": 0.23096436262130737,
+      "learning_rate": 0.001,
+      "loss": 1.8918,
+      "step": 6112
+    },
+    {
+      "epoch": 0.2586090193755817,
+      "grad_norm": 0.21152295172214508,
+      "learning_rate": 0.001,
+      "loss": 2.1265,
+      "step": 6113
+    },
+    {
+      "epoch": 0.2586513241390981,
+      "grad_norm": 0.40861940383911133,
+      "learning_rate": 0.001,
+      "loss": 1.994,
+      "step": 6114
+    },
+    {
+      "epoch": 0.25869362890261444,
+      "grad_norm": 0.21992965042591095,
+      "learning_rate": 0.001,
+      "loss": 2.5881,
+      "step": 6115
+    },
+    {
+      "epoch": 0.2587359336661308,
+      "grad_norm": 0.43303796648979187,
+      "learning_rate": 0.001,
+      "loss": 3.6542,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2587782384296472,
+      "grad_norm": 0.2100251019001007,
+      "learning_rate": 0.001,
+      "loss": 2.3924,
+      "step": 6117
+    },
+    {
+      "epoch": 0.25882054319316355,
+      "grad_norm": 0.18890580534934998,
+      "learning_rate": 0.001,
+      "loss": 2.7952,
+      "step": 6118
+    },
+    {
+      "epoch": 0.2588628479566799,
+      "grad_norm": 2.6557726860046387,
+      "learning_rate": 0.001,
+      "loss": 1.9914,
+      "step": 6119
+    },
+    {
+      "epoch": 0.2589051527201963,
+      "grad_norm": 0.16840679943561554,
+      "learning_rate": 0.001,
+      "loss": 3.2511,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2589474574837127,
+      "grad_norm": 0.2200114130973816,
+      "learning_rate": 0.001,
+      "loss": 1.7654,
+      "step": 6121
+    },
+    {
+      "epoch": 0.258989762247229,
+      "grad_norm": 0.2384369820356369,
+      "learning_rate": 0.001,
+      "loss": 2.5738,
+      "step": 6122
+    },
+    {
+      "epoch": 0.25903206701074544,
+      "grad_norm": 0.2304677963256836,
+      "learning_rate": 0.001,
+      "loss": 1.8705,
+      "step": 6123
+    },
+    {
+      "epoch": 0.2590743717742618,
+      "grad_norm": 0.23483799397945404,
+      "learning_rate": 0.001,
+      "loss": 2.1473,
+      "step": 6124
+    },
+    {
+      "epoch": 0.25911667653777815,
+      "grad_norm": 0.3107430040836334,
+      "learning_rate": 0.001,
+      "loss": 2.4566,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2591589813012945,
+      "grad_norm": 0.4102107882499695,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 6126
+    },
+    {
+      "epoch": 0.2592012860648109,
+      "grad_norm": 0.28709524869918823,
+      "learning_rate": 0.001,
+      "loss": 2.6238,
+      "step": 6127
+    },
+    {
+      "epoch": 0.25924359082832726,
+      "grad_norm": 0.23573820292949677,
+      "learning_rate": 0.001,
+      "loss": 1.296,
+      "step": 6128
+    },
+    {
+      "epoch": 0.2592858955918436,
+      "grad_norm": 0.19910579919815063,
+      "learning_rate": 0.001,
+      "loss": 3.5181,
+      "step": 6129
+    },
+    {
+      "epoch": 0.25932820035536003,
+      "grad_norm": 0.2777842879295349,
+      "learning_rate": 0.001,
+      "loss": 2.2883,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2593705051188764,
+      "grad_norm": 0.2618696391582489,
+      "learning_rate": 0.001,
+      "loss": 2.7957,
+      "step": 6131
+    },
+    {
+      "epoch": 0.25941280988239274,
+      "grad_norm": 0.23511484265327454,
+      "learning_rate": 0.001,
+      "loss": 2.1526,
+      "step": 6132
+    },
+    {
+      "epoch": 0.25945511464590915,
+      "grad_norm": 0.36623239517211914,
+      "learning_rate": 0.001,
+      "loss": 1.851,
+      "step": 6133
+    },
+    {
+      "epoch": 0.2594974194094255,
+      "grad_norm": 2.1750786304473877,
+      "learning_rate": 0.001,
+      "loss": 2.3166,
+      "step": 6134
+    },
+    {
+      "epoch": 0.25953972417294185,
+      "grad_norm": 4.039839744567871,
+      "learning_rate": 0.001,
+      "loss": 2.4438,
+      "step": 6135
+    },
+    {
+      "epoch": 0.25958202893645826,
+      "grad_norm": 0.30647915601730347,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 6136
+    },
+    {
+      "epoch": 0.2596243336999746,
+      "grad_norm": 0.23755225539207458,
+      "learning_rate": 0.001,
+      "loss": 2.0286,
+      "step": 6137
+    },
+    {
+      "epoch": 0.259666638463491,
+      "grad_norm": 4.9150919914245605,
+      "learning_rate": 0.001,
+      "loss": 2.212,
+      "step": 6138
+    },
+    {
+      "epoch": 0.2597089432270074,
+      "grad_norm": 0.3311949670314789,
+      "learning_rate": 0.001,
+      "loss": 3.6409,
+      "step": 6139
+    },
+    {
+      "epoch": 0.25975124799052374,
+      "grad_norm": 0.3053383529186249,
+      "learning_rate": 0.001,
+      "loss": 2.218,
+      "step": 6140
+    },
+    {
+      "epoch": 0.2597935527540401,
+      "grad_norm": 0.47039172053337097,
+      "learning_rate": 0.001,
+      "loss": 2.3472,
+      "step": 6141
+    },
+    {
+      "epoch": 0.2598358575175565,
+      "grad_norm": 0.27455034852027893,
+      "learning_rate": 0.001,
+      "loss": 2.3315,
+      "step": 6142
+    },
+    {
+      "epoch": 0.25987816228107286,
+      "grad_norm": 0.34887149930000305,
+      "learning_rate": 0.001,
+      "loss": 2.2812,
+      "step": 6143
+    },
+    {
+      "epoch": 0.2599204670445892,
+      "grad_norm": 0.20073916018009186,
+      "learning_rate": 0.001,
+      "loss": 1.674,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2599627718081056,
+      "grad_norm": 0.5015227794647217,
+      "learning_rate": 0.001,
+      "loss": 3.2171,
+      "step": 6145
+    },
+    {
+      "epoch": 0.260005076571622,
+      "grad_norm": 12.801533699035645,
+      "learning_rate": 0.001,
+      "loss": 2.7793,
+      "step": 6146
+    },
+    {
+      "epoch": 0.26004738133513833,
+      "grad_norm": 0.8890368938446045,
+      "learning_rate": 0.001,
+      "loss": 2.3024,
+      "step": 6147
+    },
+    {
+      "epoch": 0.2600896860986547,
+      "grad_norm": 0.24711109697818756,
+      "learning_rate": 0.001,
+      "loss": 2.2968,
+      "step": 6148
+    },
+    {
+      "epoch": 0.2601319908621711,
+      "grad_norm": 2.7126057147979736,
+      "learning_rate": 0.001,
+      "loss": 1.9963,
+      "step": 6149
+    },
+    {
+      "epoch": 0.26017429562568745,
+      "grad_norm": 0.18857067823410034,
+      "learning_rate": 0.001,
+      "loss": 2.1632,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2602166003892038,
+      "grad_norm": 0.2537596821784973,
+      "learning_rate": 0.001,
+      "loss": 1.9813,
+      "step": 6151
+    },
+    {
+      "epoch": 0.2602589051527202,
+      "grad_norm": 0.2444206178188324,
+      "learning_rate": 0.001,
+      "loss": 3.3214,
+      "step": 6152
+    },
+    {
+      "epoch": 0.26030120991623656,
+      "grad_norm": 0.19725318253040314,
+      "learning_rate": 0.001,
+      "loss": 1.9918,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2603435146797529,
+      "grad_norm": 0.38273319602012634,
+      "learning_rate": 0.001,
+      "loss": 3.036,
+      "step": 6154
+    },
+    {
+      "epoch": 0.26038581944326933,
+      "grad_norm": 1.0224730968475342,
+      "learning_rate": 0.001,
+      "loss": 1.9985,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2604281242067857,
+      "grad_norm": 0.6665830016136169,
+      "learning_rate": 0.001,
+      "loss": 3.0693,
+      "step": 6156
+    },
+    {
+      "epoch": 0.26047042897030204,
+      "grad_norm": 0.20416006445884705,
+      "learning_rate": 0.001,
+      "loss": 2.3206,
+      "step": 6157
+    },
+    {
+      "epoch": 0.26051273373381845,
+      "grad_norm": 0.3051769733428955,
+      "learning_rate": 0.001,
+      "loss": 3.7681,
+      "step": 6158
+    },
+    {
+      "epoch": 0.2605550384973348,
+      "grad_norm": 0.2127608060836792,
+      "learning_rate": 0.001,
+      "loss": 2.4234,
+      "step": 6159
+    },
+    {
+      "epoch": 0.26059734326085116,
+      "grad_norm": 0.28691282868385315,
+      "learning_rate": 0.001,
+      "loss": 1.7813,
+      "step": 6160
+    },
+    {
+      "epoch": 0.26063964802436757,
+      "grad_norm": 2.052556276321411,
+      "learning_rate": 0.001,
+      "loss": 3.2356,
+      "step": 6161
+    },
+    {
+      "epoch": 0.2606819527878839,
+      "grad_norm": 0.20295457541942596,
+      "learning_rate": 0.001,
+      "loss": 2.1718,
+      "step": 6162
+    },
+    {
+      "epoch": 0.2607242575514003,
+      "grad_norm": 0.19287841022014618,
+      "learning_rate": 0.001,
+      "loss": 1.7287,
+      "step": 6163
+    },
+    {
+      "epoch": 0.2607665623149167,
+      "grad_norm": 0.3149377405643463,
+      "learning_rate": 0.001,
+      "loss": 3.1882,
+      "step": 6164
+    },
+    {
+      "epoch": 0.26080886707843304,
+      "grad_norm": 0.21928805112838745,
+      "learning_rate": 0.001,
+      "loss": 2.8839,
+      "step": 6165
+    },
+    {
+      "epoch": 0.2608511718419494,
+      "grad_norm": 0.30192065238952637,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 6166
+    },
+    {
+      "epoch": 0.2608934766054658,
+      "grad_norm": 0.22280140221118927,
+      "learning_rate": 0.001,
+      "loss": 2.7337,
+      "step": 6167
+    },
+    {
+      "epoch": 0.26093578136898216,
+      "grad_norm": 0.19599604606628418,
+      "learning_rate": 0.001,
+      "loss": 2.2517,
+      "step": 6168
+    },
+    {
+      "epoch": 0.2609780861324985,
+      "grad_norm": 0.20323944091796875,
+      "learning_rate": 0.001,
+      "loss": 1.906,
+      "step": 6169
+    },
+    {
+      "epoch": 0.26102039089601486,
+      "grad_norm": 0.5633019804954529,
+      "learning_rate": 0.001,
+      "loss": 2.1333,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2610626956595313,
+      "grad_norm": 0.20369386672973633,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 6171
+    },
+    {
+      "epoch": 0.26110500042304763,
+      "grad_norm": 0.37366142868995667,
+      "learning_rate": 0.001,
+      "loss": 2.2786,
+      "step": 6172
+    },
+    {
+      "epoch": 0.261147305186564,
+      "grad_norm": 0.21656358242034912,
+      "learning_rate": 0.001,
+      "loss": 2.4738,
+      "step": 6173
+    },
+    {
+      "epoch": 0.2611896099500804,
+      "grad_norm": 0.1992012858390808,
+      "learning_rate": 0.001,
+      "loss": 2.1982,
+      "step": 6174
+    },
+    {
+      "epoch": 0.26123191471359675,
+      "grad_norm": 0.32912105321884155,
+      "learning_rate": 0.001,
+      "loss": 2.3972,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2612742194771131,
+      "grad_norm": 0.16803164780139923,
+      "learning_rate": 0.001,
+      "loss": 1.9239,
+      "step": 6176
+    },
+    {
+      "epoch": 0.2613165242406295,
+      "grad_norm": 0.2005642205476761,
+      "learning_rate": 0.001,
+      "loss": 2.1696,
+      "step": 6177
+    },
+    {
+      "epoch": 0.26135882900414587,
+      "grad_norm": 0.22105731070041656,
+      "learning_rate": 0.001,
+      "loss": 3.1925,
+      "step": 6178
+    },
+    {
+      "epoch": 0.2614011337676622,
+      "grad_norm": 0.20076580345630646,
+      "learning_rate": 0.001,
+      "loss": 2.1953,
+      "step": 6179
+    },
+    {
+      "epoch": 0.26144343853117863,
+      "grad_norm": 0.19129331409931183,
+      "learning_rate": 0.001,
+      "loss": 2.6739,
+      "step": 6180
+    },
+    {
+      "epoch": 0.261485743294695,
+      "grad_norm": 0.17908786237239838,
+      "learning_rate": 0.001,
+      "loss": 1.8419,
+      "step": 6181
+    },
+    {
+      "epoch": 0.26152804805821134,
+      "grad_norm": 0.2999635636806488,
+      "learning_rate": 0.001,
+      "loss": 2.5631,
+      "step": 6182
+    },
+    {
+      "epoch": 0.26157035282172775,
+      "grad_norm": 0.1999271810054779,
+      "learning_rate": 0.001,
+      "loss": 2.2973,
+      "step": 6183
+    },
+    {
+      "epoch": 0.2616126575852441,
+      "grad_norm": 0.19798393547534943,
+      "learning_rate": 0.001,
+      "loss": 2.4701,
+      "step": 6184
+    },
+    {
+      "epoch": 0.26165496234876046,
+      "grad_norm": 0.195541650056839,
+      "learning_rate": 0.001,
+      "loss": 3.1255,
+      "step": 6185
+    },
+    {
+      "epoch": 0.26169726711227687,
+      "grad_norm": 0.18188412487506866,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 6186
+    },
+    {
+      "epoch": 0.2617395718757932,
+      "grad_norm": 0.22832369804382324,
+      "learning_rate": 0.001,
+      "loss": 3.4992,
+      "step": 6187
+    },
+    {
+      "epoch": 0.2617818766393096,
+      "grad_norm": 0.20645490288734436,
+      "learning_rate": 0.001,
+      "loss": 2.6053,
+      "step": 6188
+    },
+    {
+      "epoch": 0.261824181402826,
+      "grad_norm": 0.20869730412960052,
+      "learning_rate": 0.001,
+      "loss": 2.9876,
+      "step": 6189
+    },
+    {
+      "epoch": 0.26186648616634234,
+      "grad_norm": 1.8209075927734375,
+      "learning_rate": 0.001,
+      "loss": 1.7739,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2619087909298587,
+      "grad_norm": 0.24617734551429749,
+      "learning_rate": 0.001,
+      "loss": 2.9264,
+      "step": 6191
+    },
+    {
+      "epoch": 0.26195109569337505,
+      "grad_norm": 0.5746178030967712,
+      "learning_rate": 0.001,
+      "loss": 2.4101,
+      "step": 6192
+    },
+    {
+      "epoch": 0.26199340045689146,
+      "grad_norm": 0.3251841366291046,
+      "learning_rate": 0.001,
+      "loss": 1.9637,
+      "step": 6193
+    },
+    {
+      "epoch": 0.2620357052204078,
+      "grad_norm": 0.19260448217391968,
+      "learning_rate": 0.001,
+      "loss": 2.7791,
+      "step": 6194
+    },
+    {
+      "epoch": 0.26207800998392417,
+      "grad_norm": 0.1845528781414032,
+      "learning_rate": 0.001,
+      "loss": 2.2625,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2621203147474406,
+      "grad_norm": 0.230647012591362,
+      "learning_rate": 0.001,
+      "loss": 3.0717,
+      "step": 6196
+    },
+    {
+      "epoch": 0.26216261951095693,
+      "grad_norm": 0.2902037799358368,
+      "learning_rate": 0.001,
+      "loss": 1.7595,
+      "step": 6197
+    },
+    {
+      "epoch": 0.2622049242744733,
+      "grad_norm": 0.20350608229637146,
+      "learning_rate": 0.001,
+      "loss": 2.0974,
+      "step": 6198
+    },
+    {
+      "epoch": 0.2622472290379897,
+      "grad_norm": 0.6259804964065552,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 6199
+    },
+    {
+      "epoch": 0.26228953380150605,
+      "grad_norm": 0.16990482807159424,
+      "learning_rate": 0.001,
+      "loss": 2.5884,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2623318385650224,
+      "grad_norm": 0.2967141568660736,
+      "learning_rate": 0.001,
+      "loss": 2.3306,
+      "step": 6201
+    },
+    {
+      "epoch": 0.2623741433285388,
+      "grad_norm": 0.3053082823753357,
+      "learning_rate": 0.001,
+      "loss": 2.2518,
+      "step": 6202
+    },
+    {
+      "epoch": 0.26241644809205517,
+      "grad_norm": 0.9330534338951111,
+      "learning_rate": 0.001,
+      "loss": 2.6071,
+      "step": 6203
+    },
+    {
+      "epoch": 0.2624587528555715,
+      "grad_norm": 0.19737285375595093,
+      "learning_rate": 0.001,
+      "loss": 2.0726,
+      "step": 6204
+    },
+    {
+      "epoch": 0.26250105761908793,
+      "grad_norm": 0.23269009590148926,
+      "learning_rate": 0.001,
+      "loss": 1.6949,
+      "step": 6205
+    },
+    {
+      "epoch": 0.2625433623826043,
+      "grad_norm": 0.1986543983221054,
+      "learning_rate": 0.001,
+      "loss": 2.53,
+      "step": 6206
+    },
+    {
+      "epoch": 0.26258566714612064,
+      "grad_norm": 1.0258678197860718,
+      "learning_rate": 0.001,
+      "loss": 2.4325,
+      "step": 6207
+    },
+    {
+      "epoch": 0.26262797190963705,
+      "grad_norm": 0.6095445156097412,
+      "learning_rate": 0.001,
+      "loss": 2.5046,
+      "step": 6208
+    },
+    {
+      "epoch": 0.2626702766731534,
+      "grad_norm": 1.8578591346740723,
+      "learning_rate": 0.001,
+      "loss": 2.3447,
+      "step": 6209
+    },
+    {
+      "epoch": 0.26271258143666976,
+      "grad_norm": 5.6780619621276855,
+      "learning_rate": 0.001,
+      "loss": 2.2678,
+      "step": 6210
+    },
+    {
+      "epoch": 0.26275488620018617,
+      "grad_norm": 0.2798073887825012,
+      "learning_rate": 0.001,
+      "loss": 2.0733,
+      "step": 6211
+    },
+    {
+      "epoch": 0.2627971909637025,
+      "grad_norm": 1.4034373760223389,
+      "learning_rate": 0.001,
+      "loss": 1.8523,
+      "step": 6212
+    },
+    {
+      "epoch": 0.2628394957272189,
+      "grad_norm": 0.198062464594841,
+      "learning_rate": 0.001,
+      "loss": 1.4889,
+      "step": 6213
+    },
+    {
+      "epoch": 0.26288180049073523,
+      "grad_norm": 0.2564472556114197,
+      "learning_rate": 0.001,
+      "loss": 2.4133,
+      "step": 6214
+    },
+    {
+      "epoch": 0.26292410525425164,
+      "grad_norm": 2.446117877960205,
+      "learning_rate": 0.001,
+      "loss": 1.8318,
+      "step": 6215
+    },
+    {
+      "epoch": 0.262966410017768,
+      "grad_norm": 0.28336429595947266,
+      "learning_rate": 0.001,
+      "loss": 3.2593,
+      "step": 6216
+    },
+    {
+      "epoch": 0.26300871478128435,
+      "grad_norm": 0.29390016198158264,
+      "learning_rate": 0.001,
+      "loss": 2.292,
+      "step": 6217
+    },
+    {
+      "epoch": 0.26305101954480076,
+      "grad_norm": 0.6104854941368103,
+      "learning_rate": 0.001,
+      "loss": 1.9547,
+      "step": 6218
+    },
+    {
+      "epoch": 0.2630933243083171,
+      "grad_norm": 1.6165142059326172,
+      "learning_rate": 0.001,
+      "loss": 2.2211,
+      "step": 6219
+    },
+    {
+      "epoch": 0.26313562907183347,
+      "grad_norm": 2.914968252182007,
+      "learning_rate": 0.001,
+      "loss": 2.4817,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2631779338353499,
+      "grad_norm": 0.24600031971931458,
+      "learning_rate": 0.001,
+      "loss": 2.0624,
+      "step": 6221
+    },
+    {
+      "epoch": 0.26322023859886623,
+      "grad_norm": 0.23349520564079285,
+      "learning_rate": 0.001,
+      "loss": 1.9119,
+      "step": 6222
+    },
+    {
+      "epoch": 0.2632625433623826,
+      "grad_norm": 0.2566237151622772,
+      "learning_rate": 0.001,
+      "loss": 3.1059,
+      "step": 6223
+    },
+    {
+      "epoch": 0.263304848125899,
+      "grad_norm": 0.27912285923957825,
+      "learning_rate": 0.001,
+      "loss": 2.2333,
+      "step": 6224
+    },
+    {
+      "epoch": 0.26334715288941535,
+      "grad_norm": 0.2630113661289215,
+      "learning_rate": 0.001,
+      "loss": 2.0277,
+      "step": 6225
+    },
+    {
+      "epoch": 0.2633894576529317,
+      "grad_norm": 0.34040457010269165,
+      "learning_rate": 0.001,
+      "loss": 2.6417,
+      "step": 6226
+    },
+    {
+      "epoch": 0.2634317624164481,
+      "grad_norm": 0.7498903870582581,
+      "learning_rate": 0.001,
+      "loss": 2.513,
+      "step": 6227
+    },
+    {
+      "epoch": 0.26347406717996447,
+      "grad_norm": 0.4051353633403778,
+      "learning_rate": 0.001,
+      "loss": 2.2107,
+      "step": 6228
+    },
+    {
+      "epoch": 0.2635163719434808,
+      "grad_norm": 0.288908451795578,
+      "learning_rate": 0.001,
+      "loss": 3.0754,
+      "step": 6229
+    },
+    {
+      "epoch": 0.26355867670699723,
+      "grad_norm": 0.40644702315330505,
+      "learning_rate": 0.001,
+      "loss": 3.2407,
+      "step": 6230
+    },
+    {
+      "epoch": 0.2636009814705136,
+      "grad_norm": 0.19740119576454163,
+      "learning_rate": 0.001,
+      "loss": 2.0704,
+      "step": 6231
+    },
+    {
+      "epoch": 0.26364328623402994,
+      "grad_norm": 0.22680628299713135,
+      "learning_rate": 0.001,
+      "loss": 2.6613,
+      "step": 6232
+    },
+    {
+      "epoch": 0.26368559099754635,
+      "grad_norm": 0.34792256355285645,
+      "learning_rate": 0.001,
+      "loss": 1.6583,
+      "step": 6233
+    },
+    {
+      "epoch": 0.2637278957610627,
+      "grad_norm": 0.57359379529953,
+      "learning_rate": 0.001,
+      "loss": 3.129,
+      "step": 6234
+    },
+    {
+      "epoch": 0.26377020052457906,
+      "grad_norm": 0.3252086937427521,
+      "learning_rate": 0.001,
+      "loss": 4.0645,
+      "step": 6235
+    },
+    {
+      "epoch": 0.26381250528809547,
+      "grad_norm": 0.7273129224777222,
+      "learning_rate": 0.001,
+      "loss": 2.7411,
+      "step": 6236
+    },
+    {
+      "epoch": 0.2638548100516118,
+      "grad_norm": 0.22891153395175934,
+      "learning_rate": 0.001,
+      "loss": 2.2799,
+      "step": 6237
+    },
+    {
+      "epoch": 0.2638971148151282,
+      "grad_norm": 0.4040292501449585,
+      "learning_rate": 0.001,
+      "loss": 2.375,
+      "step": 6238
+    },
+    {
+      "epoch": 0.26393941957864453,
+      "grad_norm": 20.426698684692383,
+      "learning_rate": 0.001,
+      "loss": 1.8637,
+      "step": 6239
+    },
+    {
+      "epoch": 0.26398172434216094,
+      "grad_norm": 0.26762643456459045,
+      "learning_rate": 0.001,
+      "loss": 2.1107,
+      "step": 6240
+    },
+    {
+      "epoch": 0.2640240291056773,
+      "grad_norm": 1.0472038984298706,
+      "learning_rate": 0.001,
+      "loss": 2.1314,
+      "step": 6241
+    },
+    {
+      "epoch": 0.26406633386919365,
+      "grad_norm": 0.7407483458518982,
+      "learning_rate": 0.001,
+      "loss": 2.1423,
+      "step": 6242
+    },
+    {
+      "epoch": 0.26410863863271006,
+      "grad_norm": 0.45358023047447205,
+      "learning_rate": 0.001,
+      "loss": 2.017,
+      "step": 6243
+    },
+    {
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.40108269453048706,
+      "learning_rate": 0.001,
+      "loss": 2.1408,
+      "step": 6244
+    },
+    {
+      "epoch": 0.26419324815974277,
+      "grad_norm": 0.20317232608795166,
+      "learning_rate": 0.001,
+      "loss": 2.2763,
+      "step": 6245
+    },
+    {
+      "epoch": 0.2642355529232592,
+      "grad_norm": 0.21886859834194183,
+      "learning_rate": 0.001,
+      "loss": 2.9445,
+      "step": 6246
+    },
+    {
+      "epoch": 0.26427785768677553,
+      "grad_norm": 0.2402205467224121,
+      "learning_rate": 0.001,
+      "loss": 3.2153,
+      "step": 6247
+    },
+    {
+      "epoch": 0.2643201624502919,
+      "grad_norm": 0.4614834487438202,
+      "learning_rate": 0.001,
+      "loss": 2.5112,
+      "step": 6248
+    },
+    {
+      "epoch": 0.2643624672138083,
+      "grad_norm": 0.4552881717681885,
+      "learning_rate": 0.001,
+      "loss": 2.1616,
+      "step": 6249
+    },
+    {
+      "epoch": 0.26440477197732465,
+      "grad_norm": 0.20050367712974548,
+      "learning_rate": 0.001,
+      "loss": 2.7497,
+      "step": 6250
+    },
+    {
+      "epoch": 0.264447076740841,
+      "grad_norm": 0.21264247596263885,
+      "learning_rate": 0.001,
+      "loss": 2.3534,
+      "step": 6251
+    },
+    {
+      "epoch": 0.2644893815043574,
+      "grad_norm": 0.9750521183013916,
+      "learning_rate": 0.001,
+      "loss": 1.9891,
+      "step": 6252
+    },
+    {
+      "epoch": 0.26453168626787377,
+      "grad_norm": 0.7031003832817078,
+      "learning_rate": 0.001,
+      "loss": 3.7157,
+      "step": 6253
+    },
+    {
+      "epoch": 0.2645739910313901,
+      "grad_norm": 1.4832327365875244,
+      "learning_rate": 0.001,
+      "loss": 2.1368,
+      "step": 6254
+    },
+    {
+      "epoch": 0.26461629579490653,
+      "grad_norm": 0.6920870542526245,
+      "learning_rate": 0.001,
+      "loss": 2.8061,
+      "step": 6255
+    },
+    {
+      "epoch": 0.2646586005584229,
+      "grad_norm": 0.5719873309135437,
+      "learning_rate": 0.001,
+      "loss": 2.5427,
+      "step": 6256
+    },
+    {
+      "epoch": 0.26470090532193924,
+      "grad_norm": 0.23134736716747284,
+      "learning_rate": 0.001,
+      "loss": 2.2727,
+      "step": 6257
+    },
+    {
+      "epoch": 0.26474321008545565,
+      "grad_norm": 0.1953703761100769,
+      "learning_rate": 0.001,
+      "loss": 1.9514,
+      "step": 6258
+    },
+    {
+      "epoch": 0.264785514848972,
+      "grad_norm": 0.27613356709480286,
+      "learning_rate": 0.001,
+      "loss": 2.8069,
+      "step": 6259
+    },
+    {
+      "epoch": 0.26482781961248836,
+      "grad_norm": 33.50883102416992,
+      "learning_rate": 0.001,
+      "loss": 2.0961,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2648701243760047,
+      "grad_norm": 0.2973005771636963,
+      "learning_rate": 0.001,
+      "loss": 3.3931,
+      "step": 6261
+    },
+    {
+      "epoch": 0.2649124291395211,
+      "grad_norm": 1.8105061054229736,
+      "learning_rate": 0.001,
+      "loss": 2.3645,
+      "step": 6262
+    },
+    {
+      "epoch": 0.2649547339030375,
+      "grad_norm": 0.9858757853507996,
+      "learning_rate": 0.001,
+      "loss": 2.1567,
+      "step": 6263
+    },
+    {
+      "epoch": 0.26499703866655383,
+      "grad_norm": 0.3192642629146576,
+      "learning_rate": 0.001,
+      "loss": 2.5659,
+      "step": 6264
+    },
+    {
+      "epoch": 0.26503934343007024,
+      "grad_norm": 0.21430915594100952,
+      "learning_rate": 0.001,
+      "loss": 2.5984,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2650816481935866,
+      "grad_norm": 0.45341306924819946,
+      "learning_rate": 0.001,
+      "loss": 2.8938,
+      "step": 6266
+    },
+    {
+      "epoch": 0.26512395295710295,
+      "grad_norm": 0.2323470115661621,
+      "learning_rate": 0.001,
+      "loss": 1.904,
+      "step": 6267
+    },
+    {
+      "epoch": 0.26516625772061936,
+      "grad_norm": 0.1812925934791565,
+      "learning_rate": 0.001,
+      "loss": 1.8184,
+      "step": 6268
+    },
+    {
+      "epoch": 0.2652085624841357,
+      "grad_norm": 0.23729638755321503,
+      "learning_rate": 0.001,
+      "loss": 2.2054,
+      "step": 6269
+    },
+    {
+      "epoch": 0.26525086724765207,
+      "grad_norm": 0.24241091310977936,
+      "learning_rate": 0.001,
+      "loss": 2.2426,
+      "step": 6270
+    },
+    {
+      "epoch": 0.2652931720111685,
+      "grad_norm": 0.22026705741882324,
+      "learning_rate": 0.001,
+      "loss": 1.9184,
+      "step": 6271
+    },
+    {
+      "epoch": 0.26533547677468483,
+      "grad_norm": 0.3513158857822418,
+      "learning_rate": 0.001,
+      "loss": 2.5154,
+      "step": 6272
+    },
+    {
+      "epoch": 0.2653777815382012,
+      "grad_norm": 1.525829792022705,
+      "learning_rate": 0.001,
+      "loss": 1.7338,
+      "step": 6273
+    },
+    {
+      "epoch": 0.2654200863017176,
+      "grad_norm": 2.88657808303833,
+      "learning_rate": 0.001,
+      "loss": 2.0355,
+      "step": 6274
+    },
+    {
+      "epoch": 0.26546239106523395,
+      "grad_norm": 1.638237476348877,
+      "learning_rate": 0.001,
+      "loss": 2.125,
+      "step": 6275
+    },
+    {
+      "epoch": 0.2655046958287503,
+      "grad_norm": 0.27576902508735657,
+      "learning_rate": 0.001,
+      "loss": 2.2267,
+      "step": 6276
+    },
+    {
+      "epoch": 0.2655470005922667,
+      "grad_norm": 1.6162000894546509,
+      "learning_rate": 0.001,
+      "loss": 2.4581,
+      "step": 6277
+    },
+    {
+      "epoch": 0.26558930535578307,
+      "grad_norm": 0.23247942328453064,
+      "learning_rate": 0.001,
+      "loss": 2.4351,
+      "step": 6278
+    },
+    {
+      "epoch": 0.2656316101192994,
+      "grad_norm": 0.28002870082855225,
+      "learning_rate": 0.001,
+      "loss": 2.1524,
+      "step": 6279
+    },
+    {
+      "epoch": 0.26567391488281583,
+      "grad_norm": 0.23175565898418427,
+      "learning_rate": 0.001,
+      "loss": 2.1924,
+      "step": 6280
+    },
+    {
+      "epoch": 0.2657162196463322,
+      "grad_norm": 0.7777692675590515,
+      "learning_rate": 0.001,
+      "loss": 2.6283,
+      "step": 6281
+    },
+    {
+      "epoch": 0.26575852440984854,
+      "grad_norm": 1.0904306173324585,
+      "learning_rate": 0.001,
+      "loss": 2.1067,
+      "step": 6282
+    },
+    {
+      "epoch": 0.2658008291733649,
+      "grad_norm": 0.6720946431159973,
+      "learning_rate": 0.001,
+      "loss": 1.938,
+      "step": 6283
+    },
+    {
+      "epoch": 0.2658431339368813,
+      "grad_norm": 7.82070255279541,
+      "learning_rate": 0.001,
+      "loss": 2.4127,
+      "step": 6284
+    },
+    {
+      "epoch": 0.26588543870039766,
+      "grad_norm": 0.4037260115146637,
+      "learning_rate": 0.001,
+      "loss": 3.3877,
+      "step": 6285
+    },
+    {
+      "epoch": 0.265927743463914,
+      "grad_norm": 1.355433702468872,
+      "learning_rate": 0.001,
+      "loss": 1.7886,
+      "step": 6286
+    },
+    {
+      "epoch": 0.2659700482274304,
+      "grad_norm": 0.42245426774024963,
+      "learning_rate": 0.001,
+      "loss": 3.5764,
+      "step": 6287
+    },
+    {
+      "epoch": 0.2660123529909468,
+      "grad_norm": 0.24568170309066772,
+      "learning_rate": 0.001,
+      "loss": 1.8148,
+      "step": 6288
+    },
+    {
+      "epoch": 0.26605465775446313,
+      "grad_norm": 6.263239860534668,
+      "learning_rate": 0.001,
+      "loss": 1.8413,
+      "step": 6289
+    },
+    {
+      "epoch": 0.26609696251797954,
+      "grad_norm": 0.2733914256095886,
+      "learning_rate": 0.001,
+      "loss": 2.3739,
+      "step": 6290
+    },
+    {
+      "epoch": 0.2661392672814959,
+      "grad_norm": 0.30786386132240295,
+      "learning_rate": 0.001,
+      "loss": 2.0434,
+      "step": 6291
+    },
+    {
+      "epoch": 0.26618157204501225,
+      "grad_norm": 0.29940205812454224,
+      "learning_rate": 0.001,
+      "loss": 2.5419,
+      "step": 6292
+    },
+    {
+      "epoch": 0.26622387680852866,
+      "grad_norm": 0.43194735050201416,
+      "learning_rate": 0.001,
+      "loss": 2.0004,
+      "step": 6293
+    },
+    {
+      "epoch": 0.266266181572045,
+      "grad_norm": 0.3845593333244324,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 6294
+    },
+    {
+      "epoch": 0.26630848633556137,
+      "grad_norm": 4.627384185791016,
+      "learning_rate": 0.001,
+      "loss": 3.1407,
+      "step": 6295
+    },
+    {
+      "epoch": 0.2663507910990778,
+      "grad_norm": 0.3965384066104889,
+      "learning_rate": 0.001,
+      "loss": 2.1406,
+      "step": 6296
+    },
+    {
+      "epoch": 0.26639309586259413,
+      "grad_norm": 2.6226112842559814,
+      "learning_rate": 0.001,
+      "loss": 1.7126,
+      "step": 6297
+    },
+    {
+      "epoch": 0.2664354006261105,
+      "grad_norm": 1.5446035861968994,
+      "learning_rate": 0.001,
+      "loss": 1.9526,
+      "step": 6298
+    },
+    {
+      "epoch": 0.2664777053896269,
+      "grad_norm": 2.2716145515441895,
+      "learning_rate": 0.001,
+      "loss": 2.4368,
+      "step": 6299
+    },
+    {
+      "epoch": 0.26652001015314325,
+      "grad_norm": 0.33402782678604126,
+      "learning_rate": 0.001,
+      "loss": 1.8137,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2665623149166596,
+      "grad_norm": 0.9563067555427551,
+      "learning_rate": 0.001,
+      "loss": 2.6016,
+      "step": 6301
+    },
+    {
+      "epoch": 0.266604619680176,
+      "grad_norm": 0.3132352828979492,
+      "learning_rate": 0.001,
+      "loss": 2.9526,
+      "step": 6302
+    },
+    {
+      "epoch": 0.26664692444369237,
+      "grad_norm": 0.6881158351898193,
+      "learning_rate": 0.001,
+      "loss": 2.4045,
+      "step": 6303
+    },
+    {
+      "epoch": 0.2666892292072087,
+      "grad_norm": 1.5459450483322144,
+      "learning_rate": 0.001,
+      "loss": 2.9012,
+      "step": 6304
+    },
+    {
+      "epoch": 0.2667315339707251,
+      "grad_norm": 0.40609946846961975,
+      "learning_rate": 0.001,
+      "loss": 1.9667,
+      "step": 6305
+    },
+    {
+      "epoch": 0.2667738387342415,
+      "grad_norm": 0.35249534249305725,
+      "learning_rate": 0.001,
+      "loss": 3.4398,
+      "step": 6306
+    },
+    {
+      "epoch": 0.26681614349775784,
+      "grad_norm": 0.981610894203186,
+      "learning_rate": 0.001,
+      "loss": 3.0861,
+      "step": 6307
+    },
+    {
+      "epoch": 0.2668584482612742,
+      "grad_norm": 0.530106246471405,
+      "learning_rate": 0.001,
+      "loss": 2.5613,
+      "step": 6308
+    },
+    {
+      "epoch": 0.2669007530247906,
+      "grad_norm": 2.370054244995117,
+      "learning_rate": 0.001,
+      "loss": 2.989,
+      "step": 6309
+    },
+    {
+      "epoch": 0.26694305778830696,
+      "grad_norm": 0.7922685742378235,
+      "learning_rate": 0.001,
+      "loss": 2.8242,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2669853625518233,
+      "grad_norm": 0.472288578748703,
+      "learning_rate": 0.001,
+      "loss": 2.9797,
+      "step": 6311
+    },
+    {
+      "epoch": 0.2670276673153397,
+      "grad_norm": 7.536009788513184,
+      "learning_rate": 0.001,
+      "loss": 3.7025,
+      "step": 6312
+    },
+    {
+      "epoch": 0.2670699720788561,
+      "grad_norm": 0.35234659910202026,
+      "learning_rate": 0.001,
+      "loss": 2.187,
+      "step": 6313
+    },
+    {
+      "epoch": 0.26711227684237243,
+      "grad_norm": 0.45496541261672974,
+      "learning_rate": 0.001,
+      "loss": 2.8294,
+      "step": 6314
+    },
+    {
+      "epoch": 0.26715458160588884,
+      "grad_norm": 0.4241243302822113,
+      "learning_rate": 0.001,
+      "loss": 2.8042,
+      "step": 6315
+    },
+    {
+      "epoch": 0.2671968863694052,
+      "grad_norm": 0.49575382471084595,
+      "learning_rate": 0.001,
+      "loss": 2.7069,
+      "step": 6316
+    },
+    {
+      "epoch": 0.26723919113292155,
+      "grad_norm": 0.32094845175743103,
+      "learning_rate": 0.001,
+      "loss": 2.2842,
+      "step": 6317
+    },
+    {
+      "epoch": 0.26728149589643796,
+      "grad_norm": 2.19527268409729,
+      "learning_rate": 0.001,
+      "loss": 3.9412,
+      "step": 6318
+    },
+    {
+      "epoch": 0.2673238006599543,
+      "grad_norm": 0.24995988607406616,
+      "learning_rate": 0.001,
+      "loss": 2.4964,
+      "step": 6319
+    },
+    {
+      "epoch": 0.26736610542347067,
+      "grad_norm": 4.221808433532715,
+      "learning_rate": 0.001,
+      "loss": 2.6365,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2674084101869871,
+      "grad_norm": 0.434870183467865,
+      "learning_rate": 0.001,
+      "loss": 2.3938,
+      "step": 6321
+    },
+    {
+      "epoch": 0.26745071495050343,
+      "grad_norm": 0.32494959235191345,
+      "learning_rate": 0.001,
+      "loss": 3.9073,
+      "step": 6322
+    },
+    {
+      "epoch": 0.2674930197140198,
+      "grad_norm": 0.314816951751709,
+      "learning_rate": 0.001,
+      "loss": 3.404,
+      "step": 6323
+    },
+    {
+      "epoch": 0.2675353244775362,
+      "grad_norm": 0.3015631139278412,
+      "learning_rate": 0.001,
+      "loss": 2.1681,
+      "step": 6324
+    },
+    {
+      "epoch": 0.26757762924105255,
+      "grad_norm": 0.23250603675842285,
+      "learning_rate": 0.001,
+      "loss": 2.4507,
+      "step": 6325
+    },
+    {
+      "epoch": 0.2676199340045689,
+      "grad_norm": 0.77018141746521,
+      "learning_rate": 0.001,
+      "loss": 3.565,
+      "step": 6326
+    },
+    {
+      "epoch": 0.26766223876808526,
+      "grad_norm": 2.2710511684417725,
+      "learning_rate": 0.001,
+      "loss": 2.027,
+      "step": 6327
+    },
+    {
+      "epoch": 0.26770454353160167,
+      "grad_norm": 0.3492465317249298,
+      "learning_rate": 0.001,
+      "loss": 3.2861,
+      "step": 6328
+    },
+    {
+      "epoch": 0.267746848295118,
+      "grad_norm": 2.129352331161499,
+      "learning_rate": 0.001,
+      "loss": 1.7412,
+      "step": 6329
+    },
+    {
+      "epoch": 0.2677891530586344,
+      "grad_norm": 0.8317650556564331,
+      "learning_rate": 0.001,
+      "loss": 3.3198,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2678314578221508,
+      "grad_norm": 0.3606610894203186,
+      "learning_rate": 0.001,
+      "loss": 3.1525,
+      "step": 6331
+    },
+    {
+      "epoch": 0.26787376258566714,
+      "grad_norm": 0.2575969696044922,
+      "learning_rate": 0.001,
+      "loss": 1.9891,
+      "step": 6332
+    },
+    {
+      "epoch": 0.2679160673491835,
+      "grad_norm": 0.26098647713661194,
+      "learning_rate": 0.001,
+      "loss": 2.6293,
+      "step": 6333
+    },
+    {
+      "epoch": 0.2679583721126999,
+      "grad_norm": 0.30769988894462585,
+      "learning_rate": 0.001,
+      "loss": 2.5751,
+      "step": 6334
+    },
+    {
+      "epoch": 0.26800067687621626,
+      "grad_norm": 0.9987971186637878,
+      "learning_rate": 0.001,
+      "loss": 2.2541,
+      "step": 6335
+    },
+    {
+      "epoch": 0.2680429816397326,
+      "grad_norm": 0.20395010709762573,
+      "learning_rate": 0.001,
+      "loss": 1.7506,
+      "step": 6336
+    },
+    {
+      "epoch": 0.268085286403249,
+      "grad_norm": 0.23569338023662567,
+      "learning_rate": 0.001,
+      "loss": 1.9746,
+      "step": 6337
+    },
+    {
+      "epoch": 0.2681275911667654,
+      "grad_norm": 0.3824636936187744,
+      "learning_rate": 0.001,
+      "loss": 3.2225,
+      "step": 6338
+    },
+    {
+      "epoch": 0.26816989593028173,
+      "grad_norm": 0.46429404616355896,
+      "learning_rate": 0.001,
+      "loss": 3.0844,
+      "step": 6339
+    },
+    {
+      "epoch": 0.26821220069379814,
+      "grad_norm": 0.6086142063140869,
+      "learning_rate": 0.001,
+      "loss": 3.0908,
+      "step": 6340
+    },
+    {
+      "epoch": 0.2682545054573145,
+      "grad_norm": 0.24936416745185852,
+      "learning_rate": 0.001,
+      "loss": 2.1276,
+      "step": 6341
+    },
+    {
+      "epoch": 0.26829681022083085,
+      "grad_norm": 0.41220951080322266,
+      "learning_rate": 0.001,
+      "loss": 2.7437,
+      "step": 6342
+    },
+    {
+      "epoch": 0.26833911498434726,
+      "grad_norm": 0.2830599546432495,
+      "learning_rate": 0.001,
+      "loss": 2.2482,
+      "step": 6343
+    },
+    {
+      "epoch": 0.2683814197478636,
+      "grad_norm": 0.2458266317844391,
+      "learning_rate": 0.001,
+      "loss": 2.1775,
+      "step": 6344
+    },
+    {
+      "epoch": 0.26842372451137997,
+      "grad_norm": 0.33611828088760376,
+      "learning_rate": 0.001,
+      "loss": 2.4973,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2684660292748964,
+      "grad_norm": 0.7442197799682617,
+      "learning_rate": 0.001,
+      "loss": 3.3098,
+      "step": 6346
+    },
+    {
+      "epoch": 0.26850833403841273,
+      "grad_norm": 0.23901230096817017,
+      "learning_rate": 0.001,
+      "loss": 2.303,
+      "step": 6347
+    },
+    {
+      "epoch": 0.2685506388019291,
+      "grad_norm": 0.22979353368282318,
+      "learning_rate": 0.001,
+      "loss": 1.7865,
+      "step": 6348
+    },
+    {
+      "epoch": 0.2685929435654455,
+      "grad_norm": 0.3396184742450714,
+      "learning_rate": 0.001,
+      "loss": 3.8322,
+      "step": 6349
+    },
+    {
+      "epoch": 0.26863524832896185,
+      "grad_norm": 0.2517753541469574,
+      "learning_rate": 0.001,
+      "loss": 2.2335,
+      "step": 6350
+    },
+    {
+      "epoch": 0.2686775530924782,
+      "grad_norm": 0.4144297242164612,
+      "learning_rate": 0.001,
+      "loss": 3.4627,
+      "step": 6351
+    },
+    {
+      "epoch": 0.26871985785599456,
+      "grad_norm": 0.891348123550415,
+      "learning_rate": 0.001,
+      "loss": 2.9963,
+      "step": 6352
+    },
+    {
+      "epoch": 0.26876216261951097,
+      "grad_norm": 0.18582716584205627,
+      "learning_rate": 0.001,
+      "loss": 2.2934,
+      "step": 6353
+    },
+    {
+      "epoch": 0.2688044673830273,
+      "grad_norm": 0.27463603019714355,
+      "learning_rate": 0.001,
+      "loss": 2.0024,
+      "step": 6354
+    },
+    {
+      "epoch": 0.2688467721465437,
+      "grad_norm": 7.595641613006592,
+      "learning_rate": 0.001,
+      "loss": 2.3449,
+      "step": 6355
+    },
+    {
+      "epoch": 0.2688890769100601,
+      "grad_norm": 1.0186480283737183,
+      "learning_rate": 0.001,
+      "loss": 2.4801,
+      "step": 6356
+    },
+    {
+      "epoch": 0.26893138167357644,
+      "grad_norm": 0.25770995020866394,
+      "learning_rate": 0.001,
+      "loss": 3.5221,
+      "step": 6357
+    },
+    {
+      "epoch": 0.2689736864370928,
+      "grad_norm": 0.5055752396583557,
+      "learning_rate": 0.001,
+      "loss": 3.0373,
+      "step": 6358
+    },
+    {
+      "epoch": 0.2690159912006092,
+      "grad_norm": 0.25883978605270386,
+      "learning_rate": 0.001,
+      "loss": 2.6142,
+      "step": 6359
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.2824064791202545,
+      "learning_rate": 0.001,
+      "loss": 2.4801,
+      "step": 6360
+    },
+    {
+      "epoch": 0.2691006007276419,
+      "grad_norm": 0.3055369555950165,
+      "learning_rate": 0.001,
+      "loss": 2.5225,
+      "step": 6361
+    },
+    {
+      "epoch": 0.2691429054911583,
+      "grad_norm": 0.27755987644195557,
+      "learning_rate": 0.001,
+      "loss": 2.5036,
+      "step": 6362
+    },
+    {
+      "epoch": 0.2691852102546747,
+      "grad_norm": 0.24744181334972382,
+      "learning_rate": 0.001,
+      "loss": 1.9392,
+      "step": 6363
+    },
+    {
+      "epoch": 0.26922751501819103,
+      "grad_norm": 0.23263895511627197,
+      "learning_rate": 0.001,
+      "loss": 1.7806,
+      "step": 6364
+    },
+    {
+      "epoch": 0.26926981978170744,
+      "grad_norm": 0.30707335472106934,
+      "learning_rate": 0.001,
+      "loss": 2.3249,
+      "step": 6365
+    },
+    {
+      "epoch": 0.2693121245452238,
+      "grad_norm": 0.17739908397197723,
+      "learning_rate": 0.001,
+      "loss": 1.8609,
+      "step": 6366
+    },
+    {
+      "epoch": 0.26935442930874015,
+      "grad_norm": 0.299203097820282,
+      "learning_rate": 0.001,
+      "loss": 2.6187,
+      "step": 6367
+    },
+    {
+      "epoch": 0.26939673407225656,
+      "grad_norm": 0.27614837884902954,
+      "learning_rate": 0.001,
+      "loss": 2.5299,
+      "step": 6368
+    },
+    {
+      "epoch": 0.2694390388357729,
+      "grad_norm": 1.2516486644744873,
+      "learning_rate": 0.001,
+      "loss": 2.1565,
+      "step": 6369
+    },
+    {
+      "epoch": 0.26948134359928927,
+      "grad_norm": 0.43937593698501587,
+      "learning_rate": 0.001,
+      "loss": 3.1544,
+      "step": 6370
+    },
+    {
+      "epoch": 0.2695236483628057,
+      "grad_norm": 0.272873193025589,
+      "learning_rate": 0.001,
+      "loss": 2.6493,
+      "step": 6371
+    },
+    {
+      "epoch": 0.26956595312632203,
+      "grad_norm": 0.3647904694080353,
+      "learning_rate": 0.001,
+      "loss": 2.4063,
+      "step": 6372
+    },
+    {
+      "epoch": 0.2696082578898384,
+      "grad_norm": 1.3242371082305908,
+      "learning_rate": 0.001,
+      "loss": 3.4809,
+      "step": 6373
+    },
+    {
+      "epoch": 0.26965056265335474,
+      "grad_norm": 0.23816105723381042,
+      "learning_rate": 0.001,
+      "loss": 1.8286,
+      "step": 6374
+    },
+    {
+      "epoch": 0.26969286741687115,
+      "grad_norm": 0.22475877404212952,
+      "learning_rate": 0.001,
+      "loss": 2.1492,
+      "step": 6375
+    },
+    {
+      "epoch": 0.2697351721803875,
+      "grad_norm": 0.33347398042678833,
+      "learning_rate": 0.001,
+      "loss": 2.8582,
+      "step": 6376
+    },
+    {
+      "epoch": 0.26977747694390386,
+      "grad_norm": 0.22079631686210632,
+      "learning_rate": 0.001,
+      "loss": 2.6284,
+      "step": 6377
+    },
+    {
+      "epoch": 0.26981978170742027,
+      "grad_norm": 0.24591165781021118,
+      "learning_rate": 0.001,
+      "loss": 2.1861,
+      "step": 6378
+    },
+    {
+      "epoch": 0.2698620864709366,
+      "grad_norm": 0.25623226165771484,
+      "learning_rate": 0.001,
+      "loss": 2.7529,
+      "step": 6379
+    },
+    {
+      "epoch": 0.269904391234453,
+      "grad_norm": 0.3840174078941345,
+      "learning_rate": 0.001,
+      "loss": 2.0479,
+      "step": 6380
+    },
+    {
+      "epoch": 0.2699466959979694,
+      "grad_norm": 0.29531073570251465,
+      "learning_rate": 0.001,
+      "loss": 2.1461,
+      "step": 6381
+    },
+    {
+      "epoch": 0.26998900076148574,
+      "grad_norm": 0.22676776349544525,
+      "learning_rate": 0.001,
+      "loss": 2.2194,
+      "step": 6382
+    },
+    {
+      "epoch": 0.2700313055250021,
+      "grad_norm": 0.233690544962883,
+      "learning_rate": 0.001,
+      "loss": 1.6935,
+      "step": 6383
+    },
+    {
+      "epoch": 0.2700736102885185,
+      "grad_norm": 0.5040211081504822,
+      "learning_rate": 0.001,
+      "loss": 3.0911,
+      "step": 6384
+    },
+    {
+      "epoch": 0.27011591505203486,
+      "grad_norm": 0.4640330374240875,
+      "learning_rate": 0.001,
+      "loss": 1.8026,
+      "step": 6385
+    },
+    {
+      "epoch": 0.2701582198155512,
+      "grad_norm": 0.21546930074691772,
+      "learning_rate": 0.001,
+      "loss": 3.1936,
+      "step": 6386
+    },
+    {
+      "epoch": 0.2702005245790676,
+      "grad_norm": 0.22518888115882874,
+      "learning_rate": 0.001,
+      "loss": 2.1559,
+      "step": 6387
+    },
+    {
+      "epoch": 0.270242829342584,
+      "grad_norm": 0.3616427183151245,
+      "learning_rate": 0.001,
+      "loss": 2.9413,
+      "step": 6388
+    },
+    {
+      "epoch": 0.27028513410610033,
+      "grad_norm": 0.22333014011383057,
+      "learning_rate": 0.001,
+      "loss": 2.1103,
+      "step": 6389
+    },
+    {
+      "epoch": 0.27032743886961674,
+      "grad_norm": 3.191633462905884,
+      "learning_rate": 0.001,
+      "loss": 2.508,
+      "step": 6390
+    },
+    {
+      "epoch": 0.2703697436331331,
+      "grad_norm": 0.24529823660850525,
+      "learning_rate": 0.001,
+      "loss": 2.4031,
+      "step": 6391
+    },
+    {
+      "epoch": 0.27041204839664945,
+      "grad_norm": 0.21744905412197113,
+      "learning_rate": 0.001,
+      "loss": 3.51,
+      "step": 6392
+    },
+    {
+      "epoch": 0.27045435316016586,
+      "grad_norm": 0.422944575548172,
+      "learning_rate": 0.001,
+      "loss": 3.0605,
+      "step": 6393
+    },
+    {
+      "epoch": 0.2704966579236822,
+      "grad_norm": 0.29786035418510437,
+      "learning_rate": 0.001,
+      "loss": 2.2543,
+      "step": 6394
+    },
+    {
+      "epoch": 0.27053896268719857,
+      "grad_norm": 0.5101473927497864,
+      "learning_rate": 0.001,
+      "loss": 3.697,
+      "step": 6395
+    },
+    {
+      "epoch": 0.2705812674507149,
+      "grad_norm": 0.2863130569458008,
+      "learning_rate": 0.001,
+      "loss": 2.2927,
+      "step": 6396
+    },
+    {
+      "epoch": 0.27062357221423133,
+      "grad_norm": 0.21719864010810852,
+      "learning_rate": 0.001,
+      "loss": 1.9034,
+      "step": 6397
+    },
+    {
+      "epoch": 0.2706658769777477,
+      "grad_norm": 0.5843220353126526,
+      "learning_rate": 0.001,
+      "loss": 2.0929,
+      "step": 6398
+    },
+    {
+      "epoch": 0.27070818174126404,
+      "grad_norm": 0.2397383600473404,
+      "learning_rate": 0.001,
+      "loss": 1.5715,
+      "step": 6399
+    },
+    {
+      "epoch": 0.27075048650478045,
+      "grad_norm": 0.29469746351242065,
+      "learning_rate": 0.001,
+      "loss": 3.1697,
+      "step": 6400
+    },
+    {
+      "epoch": 0.2707927912682968,
+      "grad_norm": 0.4289894998073578,
+      "learning_rate": 0.001,
+      "loss": 2.8917,
+      "step": 6401
+    },
+    {
+      "epoch": 0.27083509603181316,
+      "grad_norm": 0.40637311339378357,
+      "learning_rate": 0.001,
+      "loss": 2.5034,
+      "step": 6402
+    },
+    {
+      "epoch": 0.27087740079532957,
+      "grad_norm": 1.2302122116088867,
+      "learning_rate": 0.001,
+      "loss": 2.5266,
+      "step": 6403
+    },
+    {
+      "epoch": 0.2709197055588459,
+      "grad_norm": 0.7081871628761292,
+      "learning_rate": 0.001,
+      "loss": 2.5219,
+      "step": 6404
+    },
+    {
+      "epoch": 0.2709620103223623,
+      "grad_norm": 0.1945783644914627,
+      "learning_rate": 0.001,
+      "loss": 2.5327,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2710043150858787,
+      "grad_norm": 2.336282968521118,
+      "learning_rate": 0.001,
+      "loss": 2.8066,
+      "step": 6406
+    },
+    {
+      "epoch": 0.27104661984939504,
+      "grad_norm": 0.2778642177581787,
+      "learning_rate": 0.001,
+      "loss": 2.8345,
+      "step": 6407
+    },
+    {
+      "epoch": 0.2710889246129114,
+      "grad_norm": 0.3541242778301239,
+      "learning_rate": 0.001,
+      "loss": 2.5382,
+      "step": 6408
+    },
+    {
+      "epoch": 0.2711312293764278,
+      "grad_norm": 3.476750373840332,
+      "learning_rate": 0.001,
+      "loss": 2.3818,
+      "step": 6409
+    },
+    {
+      "epoch": 0.27117353413994416,
+      "grad_norm": 0.2823469936847687,
+      "learning_rate": 0.001,
+      "loss": 2.2147,
+      "step": 6410
+    },
+    {
+      "epoch": 0.2712158389034605,
+      "grad_norm": 0.2130715698003769,
+      "learning_rate": 0.001,
+      "loss": 2.3666,
+      "step": 6411
+    },
+    {
+      "epoch": 0.2712581436669769,
+      "grad_norm": 0.210503488779068,
+      "learning_rate": 0.001,
+      "loss": 1.7467,
+      "step": 6412
+    },
+    {
+      "epoch": 0.2713004484304933,
+      "grad_norm": 0.22151534259319305,
+      "learning_rate": 0.001,
+      "loss": 2.9495,
+      "step": 6413
+    },
+    {
+      "epoch": 0.27134275319400963,
+      "grad_norm": 0.24885092675685883,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 6414
+    },
+    {
+      "epoch": 0.27138505795752604,
+      "grad_norm": 0.23562543094158173,
+      "learning_rate": 0.001,
+      "loss": 2.2867,
+      "step": 6415
+    },
+    {
+      "epoch": 0.2714273627210424,
+      "grad_norm": 0.20105452835559845,
+      "learning_rate": 0.001,
+      "loss": 2.7994,
+      "step": 6416
+    },
+    {
+      "epoch": 0.27146966748455875,
+      "grad_norm": 0.2746374011039734,
+      "learning_rate": 0.001,
+      "loss": 1.9023,
+      "step": 6417
+    },
+    {
+      "epoch": 0.2715119722480751,
+      "grad_norm": 1.5132687091827393,
+      "learning_rate": 0.001,
+      "loss": 1.8792,
+      "step": 6418
+    },
+    {
+      "epoch": 0.2715542770115915,
+      "grad_norm": 0.24032434821128845,
+      "learning_rate": 0.001,
+      "loss": 1.9926,
+      "step": 6419
+    },
+    {
+      "epoch": 0.27159658177510787,
+      "grad_norm": 0.37214019894599915,
+      "learning_rate": 0.001,
+      "loss": 2.6278,
+      "step": 6420
+    },
+    {
+      "epoch": 0.2716388865386242,
+      "grad_norm": 0.2632519006729126,
+      "learning_rate": 0.001,
+      "loss": 2.4603,
+      "step": 6421
+    },
+    {
+      "epoch": 0.27168119130214063,
+      "grad_norm": 1.0849531888961792,
+      "learning_rate": 0.001,
+      "loss": 2.8918,
+      "step": 6422
+    },
+    {
+      "epoch": 0.271723496065657,
+      "grad_norm": 3.291983127593994,
+      "learning_rate": 0.001,
+      "loss": 2.1872,
+      "step": 6423
+    },
+    {
+      "epoch": 0.27176580082917334,
+      "grad_norm": 0.26407113671302795,
+      "learning_rate": 0.001,
+      "loss": 1.9061,
+      "step": 6424
+    },
+    {
+      "epoch": 0.27180810559268975,
+      "grad_norm": 0.2467721402645111,
+      "learning_rate": 0.001,
+      "loss": 2.1413,
+      "step": 6425
+    },
+    {
+      "epoch": 0.2718504103562061,
+      "grad_norm": 2.2249884605407715,
+      "learning_rate": 0.001,
+      "loss": 2.9308,
+      "step": 6426
+    },
+    {
+      "epoch": 0.27189271511972246,
+      "grad_norm": 0.866671621799469,
+      "learning_rate": 0.001,
+      "loss": 2.4567,
+      "step": 6427
+    },
+    {
+      "epoch": 0.27193501988323887,
+      "grad_norm": 0.32258787751197815,
+      "learning_rate": 0.001,
+      "loss": 3.198,
+      "step": 6428
+    },
+    {
+      "epoch": 0.2719773246467552,
+      "grad_norm": 2.03613018989563,
+      "learning_rate": 0.001,
+      "loss": 3.3261,
+      "step": 6429
+    },
+    {
+      "epoch": 0.2720196294102716,
+      "grad_norm": 0.9766329526901245,
+      "learning_rate": 0.001,
+      "loss": 2.2824,
+      "step": 6430
+    },
+    {
+      "epoch": 0.272061934173788,
+      "grad_norm": 0.31005382537841797,
+      "learning_rate": 0.001,
+      "loss": 3.2165,
+      "step": 6431
+    },
+    {
+      "epoch": 0.27210423893730434,
+      "grad_norm": 0.32901322841644287,
+      "learning_rate": 0.001,
+      "loss": 1.9921,
+      "step": 6432
+    },
+    {
+      "epoch": 0.2721465437008207,
+      "grad_norm": 0.24402910470962524,
+      "learning_rate": 0.001,
+      "loss": 3.7147,
+      "step": 6433
+    },
+    {
+      "epoch": 0.2721888484643371,
+      "grad_norm": 0.24346059560775757,
+      "learning_rate": 0.001,
+      "loss": 2.2568,
+      "step": 6434
+    },
+    {
+      "epoch": 0.27223115322785346,
+      "grad_norm": 0.27896568179130554,
+      "learning_rate": 0.001,
+      "loss": 2.3955,
+      "step": 6435
+    },
+    {
+      "epoch": 0.2722734579913698,
+      "grad_norm": 0.7607430219650269,
+      "learning_rate": 0.001,
+      "loss": 2.6034,
+      "step": 6436
+    },
+    {
+      "epoch": 0.2723157627548862,
+      "grad_norm": 0.5905636548995972,
+      "learning_rate": 0.001,
+      "loss": 2.1869,
+      "step": 6437
+    },
+    {
+      "epoch": 0.2723580675184026,
+      "grad_norm": 0.24976111948490143,
+      "learning_rate": 0.001,
+      "loss": 2.6681,
+      "step": 6438
+    },
+    {
+      "epoch": 0.27240037228191893,
+      "grad_norm": 0.4961627423763275,
+      "learning_rate": 0.001,
+      "loss": 3.1772,
+      "step": 6439
+    },
+    {
+      "epoch": 0.2724426770454353,
+      "grad_norm": 0.23998108506202698,
+      "learning_rate": 0.001,
+      "loss": 2.0783,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2724849818089517,
+      "grad_norm": 1.2053577899932861,
+      "learning_rate": 0.001,
+      "loss": 3.2813,
+      "step": 6441
+    },
+    {
+      "epoch": 0.27252728657246805,
+      "grad_norm": 0.2748907804489136,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 6442
+    },
+    {
+      "epoch": 0.2725695913359844,
+      "grad_norm": 0.31637445092201233,
+      "learning_rate": 0.001,
+      "loss": 2.2979,
+      "step": 6443
+    },
+    {
+      "epoch": 0.2726118960995008,
+      "grad_norm": 1.9223753213882446,
+      "learning_rate": 0.001,
+      "loss": 2.6186,
+      "step": 6444
+    },
+    {
+      "epoch": 0.27265420086301717,
+      "grad_norm": 0.6588495969772339,
+      "learning_rate": 0.001,
+      "loss": 2.2728,
+      "step": 6445
+    },
+    {
+      "epoch": 0.2726965056265335,
+      "grad_norm": 4.742616176605225,
+      "learning_rate": 0.001,
+      "loss": 2.0405,
+      "step": 6446
+    },
+    {
+      "epoch": 0.27273881039004993,
+      "grad_norm": 0.25673648715019226,
+      "learning_rate": 0.001,
+      "loss": 2.5662,
+      "step": 6447
+    },
+    {
+      "epoch": 0.2727811151535663,
+      "grad_norm": 0.22692641615867615,
+      "learning_rate": 0.001,
+      "loss": 1.8061,
+      "step": 6448
+    },
+    {
+      "epoch": 0.27282341991708264,
+      "grad_norm": 0.33545932173728943,
+      "learning_rate": 0.001,
+      "loss": 1.6433,
+      "step": 6449
+    },
+    {
+      "epoch": 0.27286572468059905,
+      "grad_norm": 0.2357272207736969,
+      "learning_rate": 0.001,
+      "loss": 2.672,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2729080294441154,
+      "grad_norm": 0.28300443291664124,
+      "learning_rate": 0.001,
+      "loss": 2.1577,
+      "step": 6451
+    },
+    {
+      "epoch": 0.27295033420763176,
+      "grad_norm": 0.2952788472175598,
+      "learning_rate": 0.001,
+      "loss": 2.1633,
+      "step": 6452
+    },
+    {
+      "epoch": 0.27299263897114817,
+      "grad_norm": 0.2804337441921234,
+      "learning_rate": 0.001,
+      "loss": 1.9271,
+      "step": 6453
+    },
+    {
+      "epoch": 0.2730349437346645,
+      "grad_norm": 0.24086658656597137,
+      "learning_rate": 0.001,
+      "loss": 2.162,
+      "step": 6454
+    },
+    {
+      "epoch": 0.2730772484981809,
+      "grad_norm": 0.26707324385643005,
+      "learning_rate": 0.001,
+      "loss": 2.1178,
+      "step": 6455
+    },
+    {
+      "epoch": 0.2731195532616973,
+      "grad_norm": 0.26330190896987915,
+      "learning_rate": 0.001,
+      "loss": 2.5571,
+      "step": 6456
+    },
+    {
+      "epoch": 0.27316185802521364,
+      "grad_norm": 1.0483336448669434,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 6457
+    },
+    {
+      "epoch": 0.27320416278873,
+      "grad_norm": 0.28875404596328735,
+      "learning_rate": 0.001,
+      "loss": 2.8507,
+      "step": 6458
+    },
+    {
+      "epoch": 0.2732464675522464,
+      "grad_norm": 0.3092382550239563,
+      "learning_rate": 0.001,
+      "loss": 2.0755,
+      "step": 6459
+    },
+    {
+      "epoch": 0.27328877231576276,
+      "grad_norm": 0.24076971411705017,
+      "learning_rate": 0.001,
+      "loss": 2.2558,
+      "step": 6460
+    },
+    {
+      "epoch": 0.2733310770792791,
+      "grad_norm": 0.26437196135520935,
+      "learning_rate": 0.001,
+      "loss": 3.098,
+      "step": 6461
+    },
+    {
+      "epoch": 0.2733733818427955,
+      "grad_norm": 0.2675298750400543,
+      "learning_rate": 0.001,
+      "loss": 3.5501,
+      "step": 6462
+    },
+    {
+      "epoch": 0.2734156866063119,
+      "grad_norm": 0.3043130338191986,
+      "learning_rate": 0.001,
+      "loss": 2.3021,
+      "step": 6463
+    },
+    {
+      "epoch": 0.27345799136982823,
+      "grad_norm": 1.052797555923462,
+      "learning_rate": 0.001,
+      "loss": 2.6146,
+      "step": 6464
+    },
+    {
+      "epoch": 0.2735002961333446,
+      "grad_norm": 0.23038017749786377,
+      "learning_rate": 0.001,
+      "loss": 1.6309,
+      "step": 6465
+    },
+    {
+      "epoch": 0.273542600896861,
+      "grad_norm": 0.3373502194881439,
+      "learning_rate": 0.001,
+      "loss": 2.1514,
+      "step": 6466
+    },
+    {
+      "epoch": 0.27358490566037735,
+      "grad_norm": 0.23279638588428497,
+      "learning_rate": 0.001,
+      "loss": 2.5891,
+      "step": 6467
+    },
+    {
+      "epoch": 0.2736272104238937,
+      "grad_norm": 0.507875919342041,
+      "learning_rate": 0.001,
+      "loss": 2.5316,
+      "step": 6468
+    },
+    {
+      "epoch": 0.2736695151874101,
+      "grad_norm": 0.21502485871315002,
+      "learning_rate": 0.001,
+      "loss": 2.2808,
+      "step": 6469
+    },
+    {
+      "epoch": 0.27371181995092647,
+      "grad_norm": 0.19854310154914856,
+      "learning_rate": 0.001,
+      "loss": 3.3566,
+      "step": 6470
+    },
+    {
+      "epoch": 0.2737541247144428,
+      "grad_norm": 0.2249358594417572,
+      "learning_rate": 0.001,
+      "loss": 1.8404,
+      "step": 6471
+    },
+    {
+      "epoch": 0.27379642947795924,
+      "grad_norm": 0.2288307249546051,
+      "learning_rate": 0.001,
+      "loss": 1.924,
+      "step": 6472
+    },
+    {
+      "epoch": 0.2738387342414756,
+      "grad_norm": 0.24865660071372986,
+      "learning_rate": 0.001,
+      "loss": 2.3673,
+      "step": 6473
+    },
+    {
+      "epoch": 0.27388103900499194,
+      "grad_norm": 0.24267597496509552,
+      "learning_rate": 0.001,
+      "loss": 1.9281,
+      "step": 6474
+    },
+    {
+      "epoch": 0.27392334376850835,
+      "grad_norm": 0.2626087963581085,
+      "learning_rate": 0.001,
+      "loss": 2.4203,
+      "step": 6475
+    },
+    {
+      "epoch": 0.2739656485320247,
+      "grad_norm": 0.8168473839759827,
+      "learning_rate": 0.001,
+      "loss": 2.6164,
+      "step": 6476
+    },
+    {
+      "epoch": 0.27400795329554106,
+      "grad_norm": 0.18421576917171478,
+      "learning_rate": 0.001,
+      "loss": 1.9259,
+      "step": 6477
+    },
+    {
+      "epoch": 0.27405025805905747,
+      "grad_norm": 0.2525613307952881,
+      "learning_rate": 0.001,
+      "loss": 2.3877,
+      "step": 6478
+    },
+    {
+      "epoch": 0.2740925628225738,
+      "grad_norm": 0.20929360389709473,
+      "learning_rate": 0.001,
+      "loss": 1.6114,
+      "step": 6479
+    },
+    {
+      "epoch": 0.2741348675860902,
+      "grad_norm": 0.4461418688297272,
+      "learning_rate": 0.001,
+      "loss": 2.563,
+      "step": 6480
+    },
+    {
+      "epoch": 0.2741771723496066,
+      "grad_norm": 0.2533968389034271,
+      "learning_rate": 0.001,
+      "loss": 2.7533,
+      "step": 6481
+    },
+    {
+      "epoch": 0.27421947711312294,
+      "grad_norm": 1.817432165145874,
+      "learning_rate": 0.001,
+      "loss": 2.2708,
+      "step": 6482
+    },
+    {
+      "epoch": 0.2742617818766393,
+      "grad_norm": 1.1035431623458862,
+      "learning_rate": 0.001,
+      "loss": 2.2485,
+      "step": 6483
+    },
+    {
+      "epoch": 0.2743040866401557,
+      "grad_norm": 0.4007866680622101,
+      "learning_rate": 0.001,
+      "loss": 2.566,
+      "step": 6484
+    },
+    {
+      "epoch": 0.27434639140367206,
+      "grad_norm": 0.569590151309967,
+      "learning_rate": 0.001,
+      "loss": 1.9645,
+      "step": 6485
+    },
+    {
+      "epoch": 0.2743886961671884,
+      "grad_norm": 0.2620263695716858,
+      "learning_rate": 0.001,
+      "loss": 2.1197,
+      "step": 6486
+    },
+    {
+      "epoch": 0.27443100093070477,
+      "grad_norm": 0.6177977323532104,
+      "learning_rate": 0.001,
+      "loss": 2.6985,
+      "step": 6487
+    },
+    {
+      "epoch": 0.2744733056942212,
+      "grad_norm": 0.22808204591274261,
+      "learning_rate": 0.001,
+      "loss": 3.8149,
+      "step": 6488
+    },
+    {
+      "epoch": 0.27451561045773754,
+      "grad_norm": 0.2721755802631378,
+      "learning_rate": 0.001,
+      "loss": 2.389,
+      "step": 6489
+    },
+    {
+      "epoch": 0.2745579152212539,
+      "grad_norm": 0.35955166816711426,
+      "learning_rate": 0.001,
+      "loss": 2.1531,
+      "step": 6490
+    },
+    {
+      "epoch": 0.2746002199847703,
+      "grad_norm": 0.2631028890609741,
+      "learning_rate": 0.001,
+      "loss": 1.944,
+      "step": 6491
+    },
+    {
+      "epoch": 0.27464252474828665,
+      "grad_norm": 0.40836191177368164,
+      "learning_rate": 0.001,
+      "loss": 2.219,
+      "step": 6492
+    },
+    {
+      "epoch": 0.274684829511803,
+      "grad_norm": 0.29734811186790466,
+      "learning_rate": 0.001,
+      "loss": 2.4754,
+      "step": 6493
+    },
+    {
+      "epoch": 0.2747271342753194,
+      "grad_norm": 0.3713732063770294,
+      "learning_rate": 0.001,
+      "loss": 3.1679,
+      "step": 6494
+    },
+    {
+      "epoch": 0.27476943903883577,
+      "grad_norm": 0.23647406697273254,
+      "learning_rate": 0.001,
+      "loss": 2.3628,
+      "step": 6495
+    },
+    {
+      "epoch": 0.2748117438023521,
+      "grad_norm": 0.19687795639038086,
+      "learning_rate": 0.001,
+      "loss": 2.8735,
+      "step": 6496
+    },
+    {
+      "epoch": 0.27485404856586854,
+      "grad_norm": 2.3540210723876953,
+      "learning_rate": 0.001,
+      "loss": 2.1741,
+      "step": 6497
+    },
+    {
+      "epoch": 0.2748963533293849,
+      "grad_norm": 0.5432139039039612,
+      "learning_rate": 0.001,
+      "loss": 2.0557,
+      "step": 6498
+    },
+    {
+      "epoch": 0.27493865809290124,
+      "grad_norm": 0.4527101516723633,
+      "learning_rate": 0.001,
+      "loss": 3.2012,
+      "step": 6499
+    },
+    {
+      "epoch": 0.27498096285641765,
+      "grad_norm": 0.22198386490345,
+      "learning_rate": 0.001,
+      "loss": 2.1226,
+      "step": 6500
+    },
+    {
+      "epoch": 0.275023267619934,
+      "grad_norm": 0.2468535304069519,
+      "learning_rate": 0.001,
+      "loss": 2.4435,
+      "step": 6501
+    },
+    {
+      "epoch": 0.27506557238345036,
+      "grad_norm": 0.22474850714206696,
+      "learning_rate": 0.001,
+      "loss": 2.0643,
+      "step": 6502
+    },
+    {
+      "epoch": 0.2751078771469668,
+      "grad_norm": 0.38860535621643066,
+      "learning_rate": 0.001,
+      "loss": 2.0566,
+      "step": 6503
+    },
+    {
+      "epoch": 0.2751501819104831,
+      "grad_norm": 0.4149872064590454,
+      "learning_rate": 0.001,
+      "loss": 2.1151,
+      "step": 6504
+    },
+    {
+      "epoch": 0.2751924866739995,
+      "grad_norm": 0.8003644943237305,
+      "learning_rate": 0.001,
+      "loss": 3.1268,
+      "step": 6505
+    },
+    {
+      "epoch": 0.2752347914375159,
+      "grad_norm": 0.2186277061700821,
+      "learning_rate": 0.001,
+      "loss": 1.8286,
+      "step": 6506
+    },
+    {
+      "epoch": 0.27527709620103225,
+      "grad_norm": 0.2277793288230896,
+      "learning_rate": 0.001,
+      "loss": 2.1968,
+      "step": 6507
+    },
+    {
+      "epoch": 0.2753194009645486,
+      "grad_norm": 0.2010040283203125,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 6508
+    },
+    {
+      "epoch": 0.27536170572806495,
+      "grad_norm": 0.9116567373275757,
+      "learning_rate": 0.001,
+      "loss": 2.6635,
+      "step": 6509
+    },
+    {
+      "epoch": 0.27540401049158136,
+      "grad_norm": 0.20740988850593567,
+      "learning_rate": 0.001,
+      "loss": 2.2698,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2754463152550977,
+      "grad_norm": 0.22273097932338715,
+      "learning_rate": 0.001,
+      "loss": 2.3287,
+      "step": 6511
+    },
+    {
+      "epoch": 0.27548862001861407,
+      "grad_norm": 0.28017768263816833,
+      "learning_rate": 0.001,
+      "loss": 2.3398,
+      "step": 6512
+    },
+    {
+      "epoch": 0.2755309247821305,
+      "grad_norm": 0.19713011384010315,
+      "learning_rate": 0.001,
+      "loss": 1.8994,
+      "step": 6513
+    },
+    {
+      "epoch": 0.27557322954564684,
+      "grad_norm": 0.19678883254528046,
+      "learning_rate": 0.001,
+      "loss": 2.0113,
+      "step": 6514
+    },
+    {
+      "epoch": 0.2756155343091632,
+      "grad_norm": 25.500926971435547,
+      "learning_rate": 0.001,
+      "loss": 1.8373,
+      "step": 6515
+    },
+    {
+      "epoch": 0.2756578390726796,
+      "grad_norm": 2.405385732650757,
+      "learning_rate": 0.001,
+      "loss": 2.077,
+      "step": 6516
+    },
+    {
+      "epoch": 0.27570014383619595,
+      "grad_norm": 0.9735079407691956,
+      "learning_rate": 0.001,
+      "loss": 2.5276,
+      "step": 6517
+    },
+    {
+      "epoch": 0.2757424485997123,
+      "grad_norm": 0.3777754604816437,
+      "learning_rate": 0.001,
+      "loss": 2.6313,
+      "step": 6518
+    },
+    {
+      "epoch": 0.2757847533632287,
+      "grad_norm": 0.25702303647994995,
+      "learning_rate": 0.001,
+      "loss": 2.9321,
+      "step": 6519
+    },
+    {
+      "epoch": 0.2758270581267451,
+      "grad_norm": 1.9004207849502563,
+      "learning_rate": 0.001,
+      "loss": 1.8612,
+      "step": 6520
+    },
+    {
+      "epoch": 0.2758693628902614,
+      "grad_norm": 0.2935592532157898,
+      "learning_rate": 0.001,
+      "loss": 3.1381,
+      "step": 6521
+    },
+    {
+      "epoch": 0.27591166765377784,
+      "grad_norm": 0.7922021150588989,
+      "learning_rate": 0.001,
+      "loss": 2.1313,
+      "step": 6522
+    },
+    {
+      "epoch": 0.2759539724172942,
+      "grad_norm": 0.2765513062477112,
+      "learning_rate": 0.001,
+      "loss": 2.7907,
+      "step": 6523
+    },
+    {
+      "epoch": 0.27599627718081055,
+      "grad_norm": 0.24716876447200775,
+      "learning_rate": 0.001,
+      "loss": 2.1889,
+      "step": 6524
+    },
+    {
+      "epoch": 0.27603858194432696,
+      "grad_norm": 0.24158848822116852,
+      "learning_rate": 0.001,
+      "loss": 1.8922,
+      "step": 6525
+    },
+    {
+      "epoch": 0.2760808867078433,
+      "grad_norm": 0.20398221909999847,
+      "learning_rate": 0.001,
+      "loss": 2.2493,
+      "step": 6526
+    },
+    {
+      "epoch": 0.27612319147135966,
+      "grad_norm": 0.22118747234344482,
+      "learning_rate": 0.001,
+      "loss": 2.6995,
+      "step": 6527
+    },
+    {
+      "epoch": 0.2761654962348761,
+      "grad_norm": 0.21023960411548615,
+      "learning_rate": 0.001,
+      "loss": 2.1655,
+      "step": 6528
+    },
+    {
+      "epoch": 0.2762078009983924,
+      "grad_norm": 0.2214033603668213,
+      "learning_rate": 0.001,
+      "loss": 1.7361,
+      "step": 6529
+    },
+    {
+      "epoch": 0.2762501057619088,
+      "grad_norm": 0.4276762306690216,
+      "learning_rate": 0.001,
+      "loss": 2.3999,
+      "step": 6530
+    },
+    {
+      "epoch": 0.27629241052542514,
+      "grad_norm": 0.2657619118690491,
+      "learning_rate": 0.001,
+      "loss": 3.2196,
+      "step": 6531
+    },
+    {
+      "epoch": 0.27633471528894155,
+      "grad_norm": 0.47785693407058716,
+      "learning_rate": 0.001,
+      "loss": 1.933,
+      "step": 6532
+    },
+    {
+      "epoch": 0.2763770200524579,
+      "grad_norm": 0.6065228581428528,
+      "learning_rate": 0.001,
+      "loss": 2.8416,
+      "step": 6533
+    },
+    {
+      "epoch": 0.27641932481597425,
+      "grad_norm": 0.18479658663272858,
+      "learning_rate": 0.001,
+      "loss": 1.9656,
+      "step": 6534
+    },
+    {
+      "epoch": 0.27646162957949066,
+      "grad_norm": 0.2248462438583374,
+      "learning_rate": 0.001,
+      "loss": 2.1931,
+      "step": 6535
+    },
+    {
+      "epoch": 0.276503934343007,
+      "grad_norm": 0.23013022541999817,
+      "learning_rate": 0.001,
+      "loss": 3.1781,
+      "step": 6536
+    },
+    {
+      "epoch": 0.2765462391065234,
+      "grad_norm": 0.23272770643234253,
+      "learning_rate": 0.001,
+      "loss": 2.345,
+      "step": 6537
+    },
+    {
+      "epoch": 0.2765885438700398,
+      "grad_norm": 6.299062252044678,
+      "learning_rate": 0.001,
+      "loss": 2.0466,
+      "step": 6538
+    },
+    {
+      "epoch": 0.27663084863355614,
+      "grad_norm": 0.2376430779695511,
+      "learning_rate": 0.001,
+      "loss": 2.2214,
+      "step": 6539
+    },
+    {
+      "epoch": 0.2766731533970725,
+      "grad_norm": 4.812066555023193,
+      "learning_rate": 0.001,
+      "loss": 2.6048,
+      "step": 6540
+    },
+    {
+      "epoch": 0.2767154581605889,
+      "grad_norm": 2.5265612602233887,
+      "learning_rate": 0.001,
+      "loss": 2.3992,
+      "step": 6541
+    },
+    {
+      "epoch": 0.27675776292410526,
+      "grad_norm": 0.3985212445259094,
+      "learning_rate": 0.001,
+      "loss": 3.2,
+      "step": 6542
+    },
+    {
+      "epoch": 0.2768000676876216,
+      "grad_norm": 4.070377349853516,
+      "learning_rate": 0.001,
+      "loss": 2.1483,
+      "step": 6543
+    },
+    {
+      "epoch": 0.276842372451138,
+      "grad_norm": 0.7375186681747437,
+      "learning_rate": 0.001,
+      "loss": 1.9092,
+      "step": 6544
+    },
+    {
+      "epoch": 0.2768846772146544,
+      "grad_norm": 0.2642837464809418,
+      "learning_rate": 0.001,
+      "loss": 2.5145,
+      "step": 6545
+    },
+    {
+      "epoch": 0.2769269819781707,
+      "grad_norm": 0.23197656869888306,
+      "learning_rate": 0.001,
+      "loss": 2.6426,
+      "step": 6546
+    },
+    {
+      "epoch": 0.27696928674168714,
+      "grad_norm": 0.4172825813293457,
+      "learning_rate": 0.001,
+      "loss": 1.7941,
+      "step": 6547
+    },
+    {
+      "epoch": 0.2770115915052035,
+      "grad_norm": 0.1864205151796341,
+      "learning_rate": 0.001,
+      "loss": 1.6709,
+      "step": 6548
+    },
+    {
+      "epoch": 0.27705389626871985,
+      "grad_norm": 1.1688998937606812,
+      "learning_rate": 0.001,
+      "loss": 2.374,
+      "step": 6549
+    },
+    {
+      "epoch": 0.27709620103223626,
+      "grad_norm": 0.26885920763015747,
+      "learning_rate": 0.001,
+      "loss": 2.6946,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2771385057957526,
+      "grad_norm": 0.6076298952102661,
+      "learning_rate": 0.001,
+      "loss": 2.1671,
+      "step": 6551
+    },
+    {
+      "epoch": 0.27718081055926896,
+      "grad_norm": 333.4671325683594,
+      "learning_rate": 0.001,
+      "loss": 2.3423,
+      "step": 6552
+    },
+    {
+      "epoch": 0.2772231153227853,
+      "grad_norm": 0.7827109098434448,
+      "learning_rate": 0.001,
+      "loss": 2.1145,
+      "step": 6553
+    },
+    {
+      "epoch": 0.27726542008630173,
+      "grad_norm": 2.1822445392608643,
+      "learning_rate": 0.001,
+      "loss": 1.8533,
+      "step": 6554
+    },
+    {
+      "epoch": 0.2773077248498181,
+      "grad_norm": 81.6081314086914,
+      "learning_rate": 0.001,
+      "loss": 2.4029,
+      "step": 6555
+    },
+    {
+      "epoch": 0.27735002961333444,
+      "grad_norm": 1.5109821557998657,
+      "learning_rate": 0.001,
+      "loss": 2.849,
+      "step": 6556
+    },
+    {
+      "epoch": 0.27739233437685085,
+      "grad_norm": 3.2216641902923584,
+      "learning_rate": 0.001,
+      "loss": 2.4335,
+      "step": 6557
+    },
+    {
+      "epoch": 0.2774346391403672,
+      "grad_norm": 0.28260374069213867,
+      "learning_rate": 0.001,
+      "loss": 2.8565,
+      "step": 6558
+    },
+    {
+      "epoch": 0.27747694390388356,
+      "grad_norm": 0.2404414415359497,
+      "learning_rate": 0.001,
+      "loss": 2.3991,
+      "step": 6559
+    },
+    {
+      "epoch": 0.27751924866739996,
+      "grad_norm": 0.2975144386291504,
+      "learning_rate": 0.001,
+      "loss": 2.8319,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2775615534309163,
+      "grad_norm": 8.33987045288086,
+      "learning_rate": 0.001,
+      "loss": 2.9683,
+      "step": 6561
+    },
+    {
+      "epoch": 0.2776038581944327,
+      "grad_norm": 1.107771396636963,
+      "learning_rate": 0.001,
+      "loss": 2.3978,
+      "step": 6562
+    },
+    {
+      "epoch": 0.2776461629579491,
+      "grad_norm": 0.24816367030143738,
+      "learning_rate": 0.001,
+      "loss": 2.367,
+      "step": 6563
+    },
+    {
+      "epoch": 0.27768846772146544,
+      "grad_norm": 1.1018351316452026,
+      "learning_rate": 0.001,
+      "loss": 2.62,
+      "step": 6564
+    },
+    {
+      "epoch": 0.2777307724849818,
+      "grad_norm": 1.640609860420227,
+      "learning_rate": 0.001,
+      "loss": 1.867,
+      "step": 6565
+    },
+    {
+      "epoch": 0.2777730772484982,
+      "grad_norm": 0.2782337963581085,
+      "learning_rate": 0.001,
+      "loss": 2.3716,
+      "step": 6566
+    },
+    {
+      "epoch": 0.27781538201201456,
+      "grad_norm": 1.2613625526428223,
+      "learning_rate": 0.001,
+      "loss": 2.6911,
+      "step": 6567
+    },
+    {
+      "epoch": 0.2778576867755309,
+      "grad_norm": 3.5542197227478027,
+      "learning_rate": 0.001,
+      "loss": 2.6192,
+      "step": 6568
+    },
+    {
+      "epoch": 0.2778999915390473,
+      "grad_norm": 0.22724595665931702,
+      "learning_rate": 0.001,
+      "loss": 1.9024,
+      "step": 6569
+    },
+    {
+      "epoch": 0.2779422963025637,
+      "grad_norm": 0.21525609493255615,
+      "learning_rate": 0.001,
+      "loss": 2.0792,
+      "step": 6570
+    },
+    {
+      "epoch": 0.27798460106608003,
+      "grad_norm": 1.333457589149475,
+      "learning_rate": 0.001,
+      "loss": 2.5817,
+      "step": 6571
+    },
+    {
+      "epoch": 0.27802690582959644,
+      "grad_norm": 0.2638329863548279,
+      "learning_rate": 0.001,
+      "loss": 1.8625,
+      "step": 6572
+    },
+    {
+      "epoch": 0.2780692105931128,
+      "grad_norm": 0.5335271954536438,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 6573
+    },
+    {
+      "epoch": 0.27811151535662915,
+      "grad_norm": 0.8093989491462708,
+      "learning_rate": 0.001,
+      "loss": 2.6137,
+      "step": 6574
+    },
+    {
+      "epoch": 0.2781538201201455,
+      "grad_norm": 0.2589428126811981,
+      "learning_rate": 0.001,
+      "loss": 2.4491,
+      "step": 6575
+    },
+    {
+      "epoch": 0.2781961248836619,
+      "grad_norm": 1.0892338752746582,
+      "learning_rate": 0.001,
+      "loss": 2.1675,
+      "step": 6576
+    },
+    {
+      "epoch": 0.27823842964717826,
+      "grad_norm": 0.3066917657852173,
+      "learning_rate": 0.001,
+      "loss": 2.3011,
+      "step": 6577
+    },
+    {
+      "epoch": 0.2782807344106946,
+      "grad_norm": 0.6517425179481506,
+      "learning_rate": 0.001,
+      "loss": 2.241,
+      "step": 6578
+    },
+    {
+      "epoch": 0.27832303917421103,
+      "grad_norm": 0.7544508576393127,
+      "learning_rate": 0.001,
+      "loss": 2.6692,
+      "step": 6579
+    },
+    {
+      "epoch": 0.2783653439377274,
+      "grad_norm": 0.2718650698661804,
+      "learning_rate": 0.001,
+      "loss": 2.6371,
+      "step": 6580
+    },
+    {
+      "epoch": 0.27840764870124374,
+      "grad_norm": 1.2337608337402344,
+      "learning_rate": 0.001,
+      "loss": 2.6096,
+      "step": 6581
+    },
+    {
+      "epoch": 0.27844995346476015,
+      "grad_norm": 0.36760374903678894,
+      "learning_rate": 0.001,
+      "loss": 2.2351,
+      "step": 6582
+    },
+    {
+      "epoch": 0.2784922582282765,
+      "grad_norm": 0.25954657793045044,
+      "learning_rate": 0.001,
+      "loss": 2.1766,
+      "step": 6583
+    },
+    {
+      "epoch": 0.27853456299179286,
+      "grad_norm": 0.606953501701355,
+      "learning_rate": 0.001,
+      "loss": 2.1137,
+      "step": 6584
+    },
+    {
+      "epoch": 0.27857686775530927,
+      "grad_norm": 0.2433394342660904,
+      "learning_rate": 0.001,
+      "loss": 1.8739,
+      "step": 6585
+    },
+    {
+      "epoch": 0.2786191725188256,
+      "grad_norm": 0.2681048810482025,
+      "learning_rate": 0.001,
+      "loss": 2.2242,
+      "step": 6586
+    },
+    {
+      "epoch": 0.278661477282342,
+      "grad_norm": 0.37554433941841125,
+      "learning_rate": 0.001,
+      "loss": 2.5414,
+      "step": 6587
+    },
+    {
+      "epoch": 0.2787037820458584,
+      "grad_norm": 0.37288689613342285,
+      "learning_rate": 0.001,
+      "loss": 2.7379,
+      "step": 6588
+    },
+    {
+      "epoch": 0.27874608680937474,
+      "grad_norm": 1.6531394720077515,
+      "learning_rate": 0.001,
+      "loss": 2.3285,
+      "step": 6589
+    },
+    {
+      "epoch": 0.2787883915728911,
+      "grad_norm": 0.4495515823364258,
+      "learning_rate": 0.001,
+      "loss": 1.9095,
+      "step": 6590
+    },
+    {
+      "epoch": 0.2788306963364075,
+      "grad_norm": 0.22591541707515717,
+      "learning_rate": 0.001,
+      "loss": 1.9806,
+      "step": 6591
+    },
+    {
+      "epoch": 0.27887300109992386,
+      "grad_norm": 0.4093262255191803,
+      "learning_rate": 0.001,
+      "loss": 2.9568,
+      "step": 6592
+    },
+    {
+      "epoch": 0.2789153058634402,
+      "grad_norm": 0.4655383825302124,
+      "learning_rate": 0.001,
+      "loss": 2.2092,
+      "step": 6593
+    },
+    {
+      "epoch": 0.2789576106269566,
+      "grad_norm": 0.1933087557554245,
+      "learning_rate": 0.001,
+      "loss": 2.6561,
+      "step": 6594
+    },
+    {
+      "epoch": 0.278999915390473,
+      "grad_norm": 0.7289586663246155,
+      "learning_rate": 0.001,
+      "loss": 2.2029,
+      "step": 6595
+    },
+    {
+      "epoch": 0.27904222015398933,
+      "grad_norm": 0.33073756098747253,
+      "learning_rate": 0.001,
+      "loss": 2.1711,
+      "step": 6596
+    },
+    {
+      "epoch": 0.27908452491750574,
+      "grad_norm": 0.1785137802362442,
+      "learning_rate": 0.001,
+      "loss": 1.7439,
+      "step": 6597
+    },
+    {
+      "epoch": 0.2791268296810221,
+      "grad_norm": 0.23087824881076813,
+      "learning_rate": 0.001,
+      "loss": 2.3974,
+      "step": 6598
+    },
+    {
+      "epoch": 0.27916913444453845,
+      "grad_norm": 0.2256430834531784,
+      "learning_rate": 0.001,
+      "loss": 2.225,
+      "step": 6599
+    },
+    {
+      "epoch": 0.2792114392080548,
+      "grad_norm": 0.27692413330078125,
+      "learning_rate": 0.001,
+      "loss": 1.9553,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2792537439715712,
+      "grad_norm": 0.28464028239250183,
+      "learning_rate": 0.001,
+      "loss": 3.7032,
+      "step": 6601
+    },
+    {
+      "epoch": 0.27929604873508757,
+      "grad_norm": 0.22122737765312195,
+      "learning_rate": 0.001,
+      "loss": 1.9469,
+      "step": 6602
+    },
+    {
+      "epoch": 0.2793383534986039,
+      "grad_norm": 0.2612067461013794,
+      "learning_rate": 0.001,
+      "loss": 1.7212,
+      "step": 6603
+    },
+    {
+      "epoch": 0.27938065826212033,
+      "grad_norm": 0.21582959592342377,
+      "learning_rate": 0.001,
+      "loss": 2.3386,
+      "step": 6604
+    },
+    {
+      "epoch": 0.2794229630256367,
+      "grad_norm": 0.258429616689682,
+      "learning_rate": 0.001,
+      "loss": 3.188,
+      "step": 6605
+    },
+    {
+      "epoch": 0.27946526778915304,
+      "grad_norm": 0.23203891515731812,
+      "learning_rate": 0.001,
+      "loss": 2.3323,
+      "step": 6606
+    },
+    {
+      "epoch": 0.27950757255266945,
+      "grad_norm": 0.37351715564727783,
+      "learning_rate": 0.001,
+      "loss": 3.2925,
+      "step": 6607
+    },
+    {
+      "epoch": 0.2795498773161858,
+      "grad_norm": 0.2870527505874634,
+      "learning_rate": 0.001,
+      "loss": 2.128,
+      "step": 6608
+    },
+    {
+      "epoch": 0.27959218207970216,
+      "grad_norm": 0.2394646555185318,
+      "learning_rate": 0.001,
+      "loss": 2.1182,
+      "step": 6609
+    },
+    {
+      "epoch": 0.27963448684321857,
+      "grad_norm": 1.8435173034667969,
+      "learning_rate": 0.001,
+      "loss": 2.1496,
+      "step": 6610
+    },
+    {
+      "epoch": 0.2796767916067349,
+      "grad_norm": 0.689439058303833,
+      "learning_rate": 0.001,
+      "loss": 1.6171,
+      "step": 6611
+    },
+    {
+      "epoch": 0.2797190963702513,
+      "grad_norm": 0.3655761182308197,
+      "learning_rate": 0.001,
+      "loss": 1.8437,
+      "step": 6612
+    },
+    {
+      "epoch": 0.2797614011337677,
+      "grad_norm": 0.21056896448135376,
+      "learning_rate": 0.001,
+      "loss": 3.201,
+      "step": 6613
+    },
+    {
+      "epoch": 0.27980370589728404,
+      "grad_norm": 0.20994161069393158,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 6614
+    },
+    {
+      "epoch": 0.2798460106608004,
+      "grad_norm": 0.20594006776809692,
+      "learning_rate": 0.001,
+      "loss": 2.0774,
+      "step": 6615
+    },
+    {
+      "epoch": 0.2798883154243168,
+      "grad_norm": 0.24466033279895782,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 6616
+    },
+    {
+      "epoch": 0.27993062018783316,
+      "grad_norm": 0.23356914520263672,
+      "learning_rate": 0.001,
+      "loss": 3.0406,
+      "step": 6617
+    },
+    {
+      "epoch": 0.2799729249513495,
+      "grad_norm": 0.3080406188964844,
+      "learning_rate": 0.001,
+      "loss": 1.7894,
+      "step": 6618
+    },
+    {
+      "epoch": 0.2800152297148659,
+      "grad_norm": 0.28324589133262634,
+      "learning_rate": 0.001,
+      "loss": 2.5605,
+      "step": 6619
+    },
+    {
+      "epoch": 0.2800575344783823,
+      "grad_norm": 0.4697299599647522,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 6620
+    },
+    {
+      "epoch": 0.28009983924189863,
+      "grad_norm": 0.7480176091194153,
+      "learning_rate": 0.001,
+      "loss": 3.3983,
+      "step": 6621
+    },
+    {
+      "epoch": 0.280142144005415,
+      "grad_norm": 1.2225714921951294,
+      "learning_rate": 0.001,
+      "loss": 2.6653,
+      "step": 6622
+    },
+    {
+      "epoch": 0.2801844487689314,
+      "grad_norm": 1.5693175792694092,
+      "learning_rate": 0.001,
+      "loss": 2.5905,
+      "step": 6623
+    },
+    {
+      "epoch": 0.28022675353244775,
+      "grad_norm": 1.6206374168395996,
+      "learning_rate": 0.001,
+      "loss": 2.0437,
+      "step": 6624
+    },
+    {
+      "epoch": 0.2802690582959641,
+      "grad_norm": 0.19221502542495728,
+      "learning_rate": 0.001,
+      "loss": 2.4707,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2803113630594805,
+      "grad_norm": 0.8016247153282166,
+      "learning_rate": 0.001,
+      "loss": 2.6747,
+      "step": 6626
+    },
+    {
+      "epoch": 0.28035366782299687,
+      "grad_norm": 0.21207623183727264,
+      "learning_rate": 0.001,
+      "loss": 2.1313,
+      "step": 6627
+    },
+    {
+      "epoch": 0.2803959725865132,
+      "grad_norm": 0.22111967206001282,
+      "learning_rate": 0.001,
+      "loss": 3.0506,
+      "step": 6628
+    },
+    {
+      "epoch": 0.28043827735002963,
+      "grad_norm": 0.21571913361549377,
+      "learning_rate": 0.001,
+      "loss": 1.8644,
+      "step": 6629
+    },
+    {
+      "epoch": 0.280480582113546,
+      "grad_norm": 0.23307397961616516,
+      "learning_rate": 0.001,
+      "loss": 2.8528,
+      "step": 6630
+    },
+    {
+      "epoch": 0.28052288687706234,
+      "grad_norm": 1.7991856336593628,
+      "learning_rate": 0.001,
+      "loss": 2.2838,
+      "step": 6631
+    },
+    {
+      "epoch": 0.28056519164057875,
+      "grad_norm": 0.23802727460861206,
+      "learning_rate": 0.001,
+      "loss": 2.0143,
+      "step": 6632
+    },
+    {
+      "epoch": 0.2806074964040951,
+      "grad_norm": 0.19570757448673248,
+      "learning_rate": 0.001,
+      "loss": 1.7036,
+      "step": 6633
+    },
+    {
+      "epoch": 0.28064980116761146,
+      "grad_norm": 0.27703866362571716,
+      "learning_rate": 0.001,
+      "loss": 4.0551,
+      "step": 6634
+    },
+    {
+      "epoch": 0.28069210593112787,
+      "grad_norm": 0.30882275104522705,
+      "learning_rate": 0.001,
+      "loss": 2.4837,
+      "step": 6635
+    },
+    {
+      "epoch": 0.2807344106946442,
+      "grad_norm": 0.2575119137763977,
+      "learning_rate": 0.001,
+      "loss": 2.4555,
+      "step": 6636
+    },
+    {
+      "epoch": 0.2807767154581606,
+      "grad_norm": 3.0365078449249268,
+      "learning_rate": 0.001,
+      "loss": 1.8518,
+      "step": 6637
+    },
+    {
+      "epoch": 0.280819020221677,
+      "grad_norm": 0.21521303057670593,
+      "learning_rate": 0.001,
+      "loss": 2.1552,
+      "step": 6638
+    },
+    {
+      "epoch": 0.28086132498519334,
+      "grad_norm": 0.2531343698501587,
+      "learning_rate": 0.001,
+      "loss": 2.9282,
+      "step": 6639
+    },
+    {
+      "epoch": 0.2809036297487097,
+      "grad_norm": 0.1944054663181305,
+      "learning_rate": 0.001,
+      "loss": 2.3841,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2809459345122261,
+      "grad_norm": 0.18977178633213043,
+      "learning_rate": 0.001,
+      "loss": 1.9523,
+      "step": 6641
+    },
+    {
+      "epoch": 0.28098823927574246,
+      "grad_norm": 0.20107170939445496,
+      "learning_rate": 0.001,
+      "loss": 1.9213,
+      "step": 6642
+    },
+    {
+      "epoch": 0.2810305440392588,
+      "grad_norm": 1.5466896295547485,
+      "learning_rate": 0.001,
+      "loss": 2.2963,
+      "step": 6643
+    },
+    {
+      "epoch": 0.28107284880277517,
+      "grad_norm": 0.20323142409324646,
+      "learning_rate": 0.001,
+      "loss": 2.3657,
+      "step": 6644
+    },
+    {
+      "epoch": 0.2811151535662916,
+      "grad_norm": 0.20698612928390503,
+      "learning_rate": 0.001,
+      "loss": 2.1832,
+      "step": 6645
+    },
+    {
+      "epoch": 0.28115745832980793,
+      "grad_norm": 0.265480101108551,
+      "learning_rate": 0.001,
+      "loss": 4.3758,
+      "step": 6646
+    },
+    {
+      "epoch": 0.2811997630933243,
+      "grad_norm": 0.21881890296936035,
+      "learning_rate": 0.001,
+      "loss": 1.9424,
+      "step": 6647
+    },
+    {
+      "epoch": 0.2812420678568407,
+      "grad_norm": 0.2621309459209442,
+      "learning_rate": 0.001,
+      "loss": 3.2508,
+      "step": 6648
+    },
+    {
+      "epoch": 0.28128437262035705,
+      "grad_norm": 0.2738354206085205,
+      "learning_rate": 0.001,
+      "loss": 2.7365,
+      "step": 6649
+    },
+    {
+      "epoch": 0.2813266773838734,
+      "grad_norm": 0.44590675830841064,
+      "learning_rate": 0.001,
+      "loss": 2.2821,
+      "step": 6650
+    },
+    {
+      "epoch": 0.2813689821473898,
+      "grad_norm": 0.25638794898986816,
+      "learning_rate": 0.001,
+      "loss": 2.7701,
+      "step": 6651
+    },
+    {
+      "epoch": 0.28141128691090617,
+      "grad_norm": 0.275796115398407,
+      "learning_rate": 0.001,
+      "loss": 2.6728,
+      "step": 6652
+    },
+    {
+      "epoch": 0.2814535916744225,
+      "grad_norm": 0.2530985176563263,
+      "learning_rate": 0.001,
+      "loss": 1.9915,
+      "step": 6653
+    },
+    {
+      "epoch": 0.28149589643793893,
+      "grad_norm": 0.1915893405675888,
+      "learning_rate": 0.001,
+      "loss": 1.9204,
+      "step": 6654
+    },
+    {
+      "epoch": 0.2815382012014553,
+      "grad_norm": 2.445066213607788,
+      "learning_rate": 0.001,
+      "loss": 2.5217,
+      "step": 6655
+    },
+    {
+      "epoch": 0.28158050596497164,
+      "grad_norm": 3.7785122394561768,
+      "learning_rate": 0.001,
+      "loss": 1.6211,
+      "step": 6656
+    },
+    {
+      "epoch": 0.28162281072848805,
+      "grad_norm": 0.21273264288902283,
+      "learning_rate": 0.001,
+      "loss": 1.958,
+      "step": 6657
+    },
+    {
+      "epoch": 0.2816651154920044,
+      "grad_norm": 0.29316845536231995,
+      "learning_rate": 0.001,
+      "loss": 2.8814,
+      "step": 6658
+    },
+    {
+      "epoch": 0.28170742025552076,
+      "grad_norm": 16.398019790649414,
+      "learning_rate": 0.001,
+      "loss": 2.0424,
+      "step": 6659
+    },
+    {
+      "epoch": 0.28174972501903717,
+      "grad_norm": 0.5399495363235474,
+      "learning_rate": 0.001,
+      "loss": 3.4691,
+      "step": 6660
+    },
+    {
+      "epoch": 0.2817920297825535,
+      "grad_norm": 0.29314282536506653,
+      "learning_rate": 0.001,
+      "loss": 1.9568,
+      "step": 6661
+    },
+    {
+      "epoch": 0.2818343345460699,
+      "grad_norm": 0.27456262707710266,
+      "learning_rate": 0.001,
+      "loss": 2.6013,
+      "step": 6662
+    },
+    {
+      "epoch": 0.2818766393095863,
+      "grad_norm": 0.5269389152526855,
+      "learning_rate": 0.001,
+      "loss": 2.4911,
+      "step": 6663
+    },
+    {
+      "epoch": 0.28191894407310264,
+      "grad_norm": 0.5170745849609375,
+      "learning_rate": 0.001,
+      "loss": 3.7857,
+      "step": 6664
+    },
+    {
+      "epoch": 0.281961248836619,
+      "grad_norm": 0.21741123497486115,
+      "learning_rate": 0.001,
+      "loss": 2.2259,
+      "step": 6665
+    },
+    {
+      "epoch": 0.28200355360013535,
+      "grad_norm": 0.19433678686618805,
+      "learning_rate": 0.001,
+      "loss": 2.3927,
+      "step": 6666
+    },
+    {
+      "epoch": 0.28204585836365176,
+      "grad_norm": 0.2336796671152115,
+      "learning_rate": 0.001,
+      "loss": 2.2329,
+      "step": 6667
+    },
+    {
+      "epoch": 0.2820881631271681,
+      "grad_norm": 0.3010079264640808,
+      "learning_rate": 0.001,
+      "loss": 2.8767,
+      "step": 6668
+    },
+    {
+      "epoch": 0.28213046789068447,
+      "grad_norm": 23.44135856628418,
+      "learning_rate": 0.001,
+      "loss": 2.6002,
+      "step": 6669
+    },
+    {
+      "epoch": 0.2821727726542009,
+      "grad_norm": 0.3416197597980499,
+      "learning_rate": 0.001,
+      "loss": 2.3501,
+      "step": 6670
+    },
+    {
+      "epoch": 0.28221507741771723,
+      "grad_norm": 0.20520153641700745,
+      "learning_rate": 0.001,
+      "loss": 1.875,
+      "step": 6671
+    },
+    {
+      "epoch": 0.2822573821812336,
+      "grad_norm": 0.2823079526424408,
+      "learning_rate": 0.001,
+      "loss": 2.699,
+      "step": 6672
+    },
+    {
+      "epoch": 0.28229968694475,
+      "grad_norm": 0.23430481553077698,
+      "learning_rate": 0.001,
+      "loss": 3.0939,
+      "step": 6673
+    },
+    {
+      "epoch": 0.28234199170826635,
+      "grad_norm": 0.2834431231021881,
+      "learning_rate": 0.001,
+      "loss": 2.2496,
+      "step": 6674
+    },
+    {
+      "epoch": 0.2823842964717827,
+      "grad_norm": 0.5793172717094421,
+      "learning_rate": 0.001,
+      "loss": 2.2915,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2824266012352991,
+      "grad_norm": 0.4027354121208191,
+      "learning_rate": 0.001,
+      "loss": 2.1231,
+      "step": 6676
+    },
+    {
+      "epoch": 0.28246890599881547,
+      "grad_norm": 0.24830132722854614,
+      "learning_rate": 0.001,
+      "loss": 1.8969,
+      "step": 6677
+    },
+    {
+      "epoch": 0.2825112107623318,
+      "grad_norm": 0.27415531873703003,
+      "learning_rate": 0.001,
+      "loss": 2.1313,
+      "step": 6678
+    },
+    {
+      "epoch": 0.28255351552584823,
+      "grad_norm": 0.21020923554897308,
+      "learning_rate": 0.001,
+      "loss": 1.6612,
+      "step": 6679
+    },
+    {
+      "epoch": 0.2825958202893646,
+      "grad_norm": 0.4596908986568451,
+      "learning_rate": 0.001,
+      "loss": 3.1641,
+      "step": 6680
+    },
+    {
+      "epoch": 0.28263812505288094,
+      "grad_norm": 0.2096545398235321,
+      "learning_rate": 0.001,
+      "loss": 2.3925,
+      "step": 6681
+    },
+    {
+      "epoch": 0.28268042981639735,
+      "grad_norm": 0.21853958070278168,
+      "learning_rate": 0.001,
+      "loss": 1.7605,
+      "step": 6682
+    },
+    {
+      "epoch": 0.2827227345799137,
+      "grad_norm": 0.2005593180656433,
+      "learning_rate": 0.001,
+      "loss": 2.0397,
+      "step": 6683
+    },
+    {
+      "epoch": 0.28276503934343006,
+      "grad_norm": 9.324366569519043,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 6684
+    },
+    {
+      "epoch": 0.28280734410694647,
+      "grad_norm": 0.23955844342708588,
+      "learning_rate": 0.001,
+      "loss": 2.4135,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2828496488704628,
+      "grad_norm": 0.240536168217659,
+      "learning_rate": 0.001,
+      "loss": 2.2006,
+      "step": 6686
+    },
+    {
+      "epoch": 0.2828919536339792,
+      "grad_norm": 0.2282586544752121,
+      "learning_rate": 0.001,
+      "loss": 2.4402,
+      "step": 6687
+    },
+    {
+      "epoch": 0.28293425839749553,
+      "grad_norm": 0.23527540266513824,
+      "learning_rate": 0.001,
+      "loss": 2.5282,
+      "step": 6688
+    },
+    {
+      "epoch": 0.28297656316101194,
+      "grad_norm": 0.2016860693693161,
+      "learning_rate": 0.001,
+      "loss": 2.9175,
+      "step": 6689
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.46961989998817444,
+      "learning_rate": 0.001,
+      "loss": 1.9682,
+      "step": 6690
+    },
+    {
+      "epoch": 0.28306117268804465,
+      "grad_norm": 3.572152614593506,
+      "learning_rate": 0.001,
+      "loss": 2.3511,
+      "step": 6691
+    },
+    {
+      "epoch": 0.28310347745156106,
+      "grad_norm": 0.18426565825939178,
+      "learning_rate": 0.001,
+      "loss": 2.563,
+      "step": 6692
+    },
+    {
+      "epoch": 0.2831457822150774,
+      "grad_norm": 0.20007674396038055,
+      "learning_rate": 0.001,
+      "loss": 2.182,
+      "step": 6693
+    },
+    {
+      "epoch": 0.28318808697859377,
+      "grad_norm": 0.2859644591808319,
+      "learning_rate": 0.001,
+      "loss": 2.4287,
+      "step": 6694
+    },
+    {
+      "epoch": 0.2832303917421102,
+      "grad_norm": 0.2305973470211029,
+      "learning_rate": 0.001,
+      "loss": 2.3868,
+      "step": 6695
+    },
+    {
+      "epoch": 0.28327269650562653,
+      "grad_norm": 0.21478715538978577,
+      "learning_rate": 0.001,
+      "loss": 2.2786,
+      "step": 6696
+    },
+    {
+      "epoch": 0.2833150012691429,
+      "grad_norm": 0.2607964277267456,
+      "learning_rate": 0.001,
+      "loss": 2.7508,
+      "step": 6697
+    },
+    {
+      "epoch": 0.2833573060326593,
+      "grad_norm": 0.2294250726699829,
+      "learning_rate": 0.001,
+      "loss": 1.9693,
+      "step": 6698
+    },
+    {
+      "epoch": 0.28339961079617565,
+      "grad_norm": 0.382492333650589,
+      "learning_rate": 0.001,
+      "loss": 3.9725,
+      "step": 6699
+    },
+    {
+      "epoch": 0.283441915559692,
+      "grad_norm": 0.2415558397769928,
+      "learning_rate": 0.001,
+      "loss": 2.6939,
+      "step": 6700
+    },
+    {
+      "epoch": 0.2834842203232084,
+      "grad_norm": 0.18394537270069122,
+      "learning_rate": 0.001,
+      "loss": 2.3632,
+      "step": 6701
+    },
+    {
+      "epoch": 0.28352652508672477,
+      "grad_norm": 0.2569441795349121,
+      "learning_rate": 0.001,
+      "loss": 2.3462,
+      "step": 6702
+    },
+    {
+      "epoch": 0.2835688298502411,
+      "grad_norm": 0.19975893199443817,
+      "learning_rate": 0.001,
+      "loss": 1.6878,
+      "step": 6703
+    },
+    {
+      "epoch": 0.28361113461375753,
+      "grad_norm": 2.8916850090026855,
+      "learning_rate": 0.001,
+      "loss": 1.7501,
+      "step": 6704
+    },
+    {
+      "epoch": 0.2836534393772739,
+      "grad_norm": 0.24953538179397583,
+      "learning_rate": 0.001,
+      "loss": 2.7791,
+      "step": 6705
+    },
+    {
+      "epoch": 0.28369574414079024,
+      "grad_norm": 0.6636295318603516,
+      "learning_rate": 0.001,
+      "loss": 2.2578,
+      "step": 6706
+    },
+    {
+      "epoch": 0.28373804890430665,
+      "grad_norm": 2.3027749061584473,
+      "learning_rate": 0.001,
+      "loss": 1.9202,
+      "step": 6707
+    },
+    {
+      "epoch": 0.283780353667823,
+      "grad_norm": 0.17140471935272217,
+      "learning_rate": 0.001,
+      "loss": 1.9173,
+      "step": 6708
+    },
+    {
+      "epoch": 0.28382265843133936,
+      "grad_norm": 0.1749950796365738,
+      "learning_rate": 0.001,
+      "loss": 1.5205,
+      "step": 6709
+    },
+    {
+      "epoch": 0.28386496319485577,
+      "grad_norm": 0.8592566251754761,
+      "learning_rate": 0.001,
+      "loss": 3.4644,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2839072679583721,
+      "grad_norm": 0.17556998133659363,
+      "learning_rate": 0.001,
+      "loss": 1.9657,
+      "step": 6711
+    },
+    {
+      "epoch": 0.2839495727218885,
+      "grad_norm": 0.2964968681335449,
+      "learning_rate": 0.001,
+      "loss": 2.8859,
+      "step": 6712
+    },
+    {
+      "epoch": 0.28399187748540483,
+      "grad_norm": 0.6098200678825378,
+      "learning_rate": 0.001,
+      "loss": 1.7954,
+      "step": 6713
+    },
+    {
+      "epoch": 0.28403418224892124,
+      "grad_norm": 0.43372607231140137,
+      "learning_rate": 0.001,
+      "loss": 3.3326,
+      "step": 6714
+    },
+    {
+      "epoch": 0.2840764870124376,
+      "grad_norm": 0.4292704463005066,
+      "learning_rate": 0.001,
+      "loss": 2.5526,
+      "step": 6715
+    },
+    {
+      "epoch": 0.28411879177595395,
+      "grad_norm": 0.21659818291664124,
+      "learning_rate": 0.001,
+      "loss": 1.9127,
+      "step": 6716
+    },
+    {
+      "epoch": 0.28416109653947036,
+      "grad_norm": 0.4214901030063629,
+      "learning_rate": 0.001,
+      "loss": 1.7826,
+      "step": 6717
+    },
+    {
+      "epoch": 0.2842034013029867,
+      "grad_norm": 0.3387224078178406,
+      "learning_rate": 0.001,
+      "loss": 2.7385,
+      "step": 6718
+    },
+    {
+      "epoch": 0.28424570606650307,
+      "grad_norm": 0.5583015084266663,
+      "learning_rate": 0.001,
+      "loss": 3.1142,
+      "step": 6719
+    },
+    {
+      "epoch": 0.2842880108300195,
+      "grad_norm": 0.19958928227424622,
+      "learning_rate": 0.001,
+      "loss": 3.1751,
+      "step": 6720
+    },
+    {
+      "epoch": 0.28433031559353583,
+      "grad_norm": 0.22173653542995453,
+      "learning_rate": 0.001,
+      "loss": 3.0606,
+      "step": 6721
+    },
+    {
+      "epoch": 0.2843726203570522,
+      "grad_norm": 0.24500377476215363,
+      "learning_rate": 0.001,
+      "loss": 2.0964,
+      "step": 6722
+    },
+    {
+      "epoch": 0.2844149251205686,
+      "grad_norm": 0.21380500495433807,
+      "learning_rate": 0.001,
+      "loss": 2.1937,
+      "step": 6723
+    },
+    {
+      "epoch": 0.28445722988408495,
+      "grad_norm": 0.21296393871307373,
+      "learning_rate": 0.001,
+      "loss": 2.1262,
+      "step": 6724
+    },
+    {
+      "epoch": 0.2844995346476013,
+      "grad_norm": 0.19265830516815186,
+      "learning_rate": 0.001,
+      "loss": 2.587,
+      "step": 6725
+    },
+    {
+      "epoch": 0.2845418394111177,
+      "grad_norm": 0.19577600061893463,
+      "learning_rate": 0.001,
+      "loss": 1.7467,
+      "step": 6726
+    },
+    {
+      "epoch": 0.28458414417463407,
+      "grad_norm": 0.25337207317352295,
+      "learning_rate": 0.001,
+      "loss": 2.1288,
+      "step": 6727
+    },
+    {
+      "epoch": 0.2846264489381504,
+      "grad_norm": 0.4300183653831482,
+      "learning_rate": 0.001,
+      "loss": 2.7892,
+      "step": 6728
+    },
+    {
+      "epoch": 0.28466875370166683,
+      "grad_norm": 0.8722933530807495,
+      "learning_rate": 0.001,
+      "loss": 2.3566,
+      "step": 6729
+    },
+    {
+      "epoch": 0.2847110584651832,
+      "grad_norm": 0.17936961352825165,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 6730
+    },
+    {
+      "epoch": 0.28475336322869954,
+      "grad_norm": 0.18349401652812958,
+      "learning_rate": 0.001,
+      "loss": 1.7784,
+      "step": 6731
+    },
+    {
+      "epoch": 0.28479566799221595,
+      "grad_norm": 1.0016671419143677,
+      "learning_rate": 0.001,
+      "loss": 1.9157,
+      "step": 6732
+    },
+    {
+      "epoch": 0.2848379727557323,
+      "grad_norm": 0.14717590808868408,
+      "learning_rate": 0.001,
+      "loss": 1.3068,
+      "step": 6733
+    },
+    {
+      "epoch": 0.28488027751924866,
+      "grad_norm": 4.4579243659973145,
+      "learning_rate": 0.001,
+      "loss": 2.4236,
+      "step": 6734
+    },
+    {
+      "epoch": 0.284922582282765,
+      "grad_norm": 0.2059265822172165,
+      "learning_rate": 0.001,
+      "loss": 1.9139,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2849648870462814,
+      "grad_norm": 0.2114630937576294,
+      "learning_rate": 0.001,
+      "loss": 3.1348,
+      "step": 6736
+    },
+    {
+      "epoch": 0.2850071918097978,
+      "grad_norm": 0.2629503011703491,
+      "learning_rate": 0.001,
+      "loss": 2.6543,
+      "step": 6737
+    },
+    {
+      "epoch": 0.28504949657331413,
+      "grad_norm": 2.027836799621582,
+      "learning_rate": 0.001,
+      "loss": 2.1327,
+      "step": 6738
+    },
+    {
+      "epoch": 0.28509180133683054,
+      "grad_norm": 0.9196799993515015,
+      "learning_rate": 0.001,
+      "loss": 2.5672,
+      "step": 6739
+    },
+    {
+      "epoch": 0.2851341061003469,
+      "grad_norm": 5.23358154296875,
+      "learning_rate": 0.001,
+      "loss": 2.0236,
+      "step": 6740
+    },
+    {
+      "epoch": 0.28517641086386325,
+      "grad_norm": 0.17935171723365784,
+      "learning_rate": 0.001,
+      "loss": 2.6343,
+      "step": 6741
+    },
+    {
+      "epoch": 0.28521871562737966,
+      "grad_norm": 1.6367027759552002,
+      "learning_rate": 0.001,
+      "loss": 1.7514,
+      "step": 6742
+    },
+    {
+      "epoch": 0.285261020390896,
+      "grad_norm": 0.20048488676548004,
+      "learning_rate": 0.001,
+      "loss": 2.6243,
+      "step": 6743
+    },
+    {
+      "epoch": 0.28530332515441237,
+      "grad_norm": 0.37594085931777954,
+      "learning_rate": 0.001,
+      "loss": 3.2875,
+      "step": 6744
+    },
+    {
+      "epoch": 0.2853456299179288,
+      "grad_norm": 0.18674291670322418,
+      "learning_rate": 0.001,
+      "loss": 1.8737,
+      "step": 6745
+    },
+    {
+      "epoch": 0.28538793468144513,
+      "grad_norm": 0.40229612588882446,
+      "learning_rate": 0.001,
+      "loss": 2.3867,
+      "step": 6746
+    },
+    {
+      "epoch": 0.2854302394449615,
+      "grad_norm": 2.3481996059417725,
+      "learning_rate": 0.001,
+      "loss": 3.3213,
+      "step": 6747
+    },
+    {
+      "epoch": 0.2854725442084779,
+      "grad_norm": 0.23974232375621796,
+      "learning_rate": 0.001,
+      "loss": 3.0777,
+      "step": 6748
+    },
+    {
+      "epoch": 0.28551484897199425,
+      "grad_norm": 0.22430221736431122,
+      "learning_rate": 0.001,
+      "loss": 2.5951,
+      "step": 6749
+    },
+    {
+      "epoch": 0.2855571537355106,
+      "grad_norm": 1.6026118993759155,
+      "learning_rate": 0.001,
+      "loss": 2.3724,
+      "step": 6750
+    },
+    {
+      "epoch": 0.285599458499027,
+      "grad_norm": 2.978184938430786,
+      "learning_rate": 0.001,
+      "loss": 2.1911,
+      "step": 6751
+    },
+    {
+      "epoch": 0.28564176326254337,
+      "grad_norm": 11.393836975097656,
+      "learning_rate": 0.001,
+      "loss": 3.553,
+      "step": 6752
+    },
+    {
+      "epoch": 0.2856840680260597,
+      "grad_norm": 0.254300057888031,
+      "learning_rate": 0.001,
+      "loss": 2.9947,
+      "step": 6753
+    },
+    {
+      "epoch": 0.28572637278957613,
+      "grad_norm": 0.5926849246025085,
+      "learning_rate": 0.001,
+      "loss": 2.2195,
+      "step": 6754
+    },
+    {
+      "epoch": 0.2857686775530925,
+      "grad_norm": 6.964447975158691,
+      "learning_rate": 0.001,
+      "loss": 2.2015,
+      "step": 6755
+    },
+    {
+      "epoch": 0.28581098231660884,
+      "grad_norm": 4.038476943969727,
+      "learning_rate": 0.001,
+      "loss": 2.186,
+      "step": 6756
+    },
+    {
+      "epoch": 0.2858532870801252,
+      "grad_norm": 3.579437494277954,
+      "learning_rate": 0.001,
+      "loss": 2.3797,
+      "step": 6757
+    },
+    {
+      "epoch": 0.2858955918436416,
+      "grad_norm": 4.091229438781738,
+      "learning_rate": 0.001,
+      "loss": 2.5496,
+      "step": 6758
+    },
+    {
+      "epoch": 0.28593789660715796,
+      "grad_norm": 3.153745412826538,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 6759
+    },
+    {
+      "epoch": 0.2859802013706743,
+      "grad_norm": 1.0637880563735962,
+      "learning_rate": 0.001,
+      "loss": 2.5076,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2860225061341907,
+      "grad_norm": 1.7310835123062134,
+      "learning_rate": 0.001,
+      "loss": 2.6743,
+      "step": 6761
+    },
+    {
+      "epoch": 0.2860648108977071,
+      "grad_norm": 0.3798615336418152,
+      "learning_rate": 0.001,
+      "loss": 2.1774,
+      "step": 6762
+    },
+    {
+      "epoch": 0.28610711566122343,
+      "grad_norm": 0.35704922676086426,
+      "learning_rate": 0.001,
+      "loss": 1.6369,
+      "step": 6763
+    },
+    {
+      "epoch": 0.28614942042473984,
+      "grad_norm": 0.22196084260940552,
+      "learning_rate": 0.001,
+      "loss": 2.2773,
+      "step": 6764
+    },
+    {
+      "epoch": 0.2861917251882562,
+      "grad_norm": 0.2564329504966736,
+      "learning_rate": 0.001,
+      "loss": 2.0616,
+      "step": 6765
+    },
+    {
+      "epoch": 0.28623402995177255,
+      "grad_norm": 0.662453830242157,
+      "learning_rate": 0.001,
+      "loss": 2.425,
+      "step": 6766
+    },
+    {
+      "epoch": 0.28627633471528896,
+      "grad_norm": 0.275669127702713,
+      "learning_rate": 0.001,
+      "loss": 2.1394,
+      "step": 6767
+    },
+    {
+      "epoch": 0.2863186394788053,
+      "grad_norm": 1.1164103746414185,
+      "learning_rate": 0.001,
+      "loss": 3.1801,
+      "step": 6768
+    },
+    {
+      "epoch": 0.28636094424232167,
+      "grad_norm": 0.28561368584632874,
+      "learning_rate": 0.001,
+      "loss": 3.4721,
+      "step": 6769
+    },
+    {
+      "epoch": 0.2864032490058381,
+      "grad_norm": 0.44134315848350525,
+      "learning_rate": 0.001,
+      "loss": 2.4825,
+      "step": 6770
+    },
+    {
+      "epoch": 0.28644555376935443,
+      "grad_norm": 0.6540394425392151,
+      "learning_rate": 0.001,
+      "loss": 2.4181,
+      "step": 6771
+    },
+    {
+      "epoch": 0.2864878585328708,
+      "grad_norm": 0.9629650712013245,
+      "learning_rate": 0.001,
+      "loss": 1.801,
+      "step": 6772
+    },
+    {
+      "epoch": 0.2865301632963872,
+      "grad_norm": 1.2992072105407715,
+      "learning_rate": 0.001,
+      "loss": 2.4131,
+      "step": 6773
+    },
+    {
+      "epoch": 0.28657246805990355,
+      "grad_norm": 0.8524438142776489,
+      "learning_rate": 0.001,
+      "loss": 2.1917,
+      "step": 6774
+    },
+    {
+      "epoch": 0.2866147728234199,
+      "grad_norm": 18.05820083618164,
+      "learning_rate": 0.001,
+      "loss": 3.5782,
+      "step": 6775
+    },
+    {
+      "epoch": 0.2866570775869363,
+      "grad_norm": 0.2133456915616989,
+      "learning_rate": 0.001,
+      "loss": 2.8237,
+      "step": 6776
+    },
+    {
+      "epoch": 0.28669938235045267,
+      "grad_norm": 29.985143661499023,
+      "learning_rate": 0.001,
+      "loss": 1.9714,
+      "step": 6777
+    },
+    {
+      "epoch": 0.286741687113969,
+      "grad_norm": 0.23235024511814117,
+      "learning_rate": 0.001,
+      "loss": 2.2575,
+      "step": 6778
+    },
+    {
+      "epoch": 0.2867839918774854,
+      "grad_norm": 0.2286916971206665,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 6779
+    },
+    {
+      "epoch": 0.2868262966410018,
+      "grad_norm": 0.20938080549240112,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 6780
+    },
+    {
+      "epoch": 0.28686860140451814,
+      "grad_norm": 0.1856824904680252,
+      "learning_rate": 0.001,
+      "loss": 2.2964,
+      "step": 6781
+    },
+    {
+      "epoch": 0.2869109061680345,
+      "grad_norm": 0.18935956060886383,
+      "learning_rate": 0.001,
+      "loss": 2.8474,
+      "step": 6782
+    },
+    {
+      "epoch": 0.2869532109315509,
+      "grad_norm": 0.24744349718093872,
+      "learning_rate": 0.001,
+      "loss": 2.2822,
+      "step": 6783
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.26295721530914307,
+      "learning_rate": 0.001,
+      "loss": 2.3905,
+      "step": 6784
+    },
+    {
+      "epoch": 0.2870378204585836,
+      "grad_norm": 0.23481005430221558,
+      "learning_rate": 0.001,
+      "loss": 2.0109,
+      "step": 6785
+    },
+    {
+      "epoch": 0.2870801252221,
+      "grad_norm": 0.17823268473148346,
+      "learning_rate": 0.001,
+      "loss": 2.6713,
+      "step": 6786
+    },
+    {
+      "epoch": 0.2871224299856164,
+      "grad_norm": 0.28128448128700256,
+      "learning_rate": 0.001,
+      "loss": 2.5553,
+      "step": 6787
+    },
+    {
+      "epoch": 0.28716473474913273,
+      "grad_norm": 0.23149077594280243,
+      "learning_rate": 0.001,
+      "loss": 1.8728,
+      "step": 6788
+    },
+    {
+      "epoch": 0.28720703951264914,
+      "grad_norm": 0.21473278105258942,
+      "learning_rate": 0.001,
+      "loss": 2.493,
+      "step": 6789
+    },
+    {
+      "epoch": 0.2872493442761655,
+      "grad_norm": 0.19875210523605347,
+      "learning_rate": 0.001,
+      "loss": 1.7056,
+      "step": 6790
+    },
+    {
+      "epoch": 0.28729164903968185,
+      "grad_norm": 0.18944133818149567,
+      "learning_rate": 0.001,
+      "loss": 2.243,
+      "step": 6791
+    },
+    {
+      "epoch": 0.28733395380319826,
+      "grad_norm": 0.17879356443881989,
+      "learning_rate": 0.001,
+      "loss": 1.9253,
+      "step": 6792
+    },
+    {
+      "epoch": 0.2873762585667146,
+      "grad_norm": 0.39051762223243713,
+      "learning_rate": 0.001,
+      "loss": 2.477,
+      "step": 6793
+    },
+    {
+      "epoch": 0.28741856333023097,
+      "grad_norm": 0.2168491631746292,
+      "learning_rate": 0.001,
+      "loss": 2.2378,
+      "step": 6794
+    },
+    {
+      "epoch": 0.2874608680937474,
+      "grad_norm": 0.20587210357189178,
+      "learning_rate": 0.001,
+      "loss": 2.9123,
+      "step": 6795
+    },
+    {
+      "epoch": 0.28750317285726373,
+      "grad_norm": 0.16420647501945496,
+      "learning_rate": 0.001,
+      "loss": 1.4775,
+      "step": 6796
+    },
+    {
+      "epoch": 0.2875454776207801,
+      "grad_norm": 0.18433904647827148,
+      "learning_rate": 0.001,
+      "loss": 2.6196,
+      "step": 6797
+    },
+    {
+      "epoch": 0.2875877823842965,
+      "grad_norm": 0.18781892955303192,
+      "learning_rate": 0.001,
+      "loss": 2.1033,
+      "step": 6798
+    },
+    {
+      "epoch": 0.28763008714781285,
+      "grad_norm": 0.17372409999370575,
+      "learning_rate": 0.001,
+      "loss": 2.12,
+      "step": 6799
+    },
+    {
+      "epoch": 0.2876723919113292,
+      "grad_norm": 0.18070591986179352,
+      "learning_rate": 0.001,
+      "loss": 2.02,
+      "step": 6800
+    },
+    {
+      "epoch": 0.28771469667484556,
+      "grad_norm": 0.1879878044128418,
+      "learning_rate": 0.001,
+      "loss": 1.6728,
+      "step": 6801
+    },
+    {
+      "epoch": 0.28775700143836197,
+      "grad_norm": 0.18465609848499298,
+      "learning_rate": 0.001,
+      "loss": 2.1977,
+      "step": 6802
+    },
+    {
+      "epoch": 0.2877993062018783,
+      "grad_norm": 0.19959770143032074,
+      "learning_rate": 0.001,
+      "loss": 2.1808,
+      "step": 6803
+    },
+    {
+      "epoch": 0.2878416109653947,
+      "grad_norm": 0.2111247032880783,
+      "learning_rate": 0.001,
+      "loss": 2.2085,
+      "step": 6804
+    },
+    {
+      "epoch": 0.2878839157289111,
+      "grad_norm": 0.25699108839035034,
+      "learning_rate": 0.001,
+      "loss": 2.784,
+      "step": 6805
+    },
+    {
+      "epoch": 0.28792622049242744,
+      "grad_norm": 0.17606748640537262,
+      "learning_rate": 0.001,
+      "loss": 1.9087,
+      "step": 6806
+    },
+    {
+      "epoch": 0.2879685252559438,
+      "grad_norm": 0.17687219381332397,
+      "learning_rate": 0.001,
+      "loss": 1.848,
+      "step": 6807
+    },
+    {
+      "epoch": 0.2880108300194602,
+      "grad_norm": 0.22173765301704407,
+      "learning_rate": 0.001,
+      "loss": 2.3972,
+      "step": 6808
+    },
+    {
+      "epoch": 0.28805313478297656,
+      "grad_norm": 0.18182872235774994,
+      "learning_rate": 0.001,
+      "loss": 1.7431,
+      "step": 6809
+    },
+    {
+      "epoch": 0.2880954395464929,
+      "grad_norm": 0.25314095616340637,
+      "learning_rate": 0.001,
+      "loss": 1.7966,
+      "step": 6810
+    },
+    {
+      "epoch": 0.2881377443100093,
+      "grad_norm": 0.20727166533470154,
+      "learning_rate": 0.001,
+      "loss": 2.3798,
+      "step": 6811
+    },
+    {
+      "epoch": 0.2881800490735257,
+      "grad_norm": 0.9613415002822876,
+      "learning_rate": 0.001,
+      "loss": 2.2462,
+      "step": 6812
+    },
+    {
+      "epoch": 0.28822235383704203,
+      "grad_norm": 1.1291375160217285,
+      "learning_rate": 0.001,
+      "loss": 2.3434,
+      "step": 6813
+    },
+    {
+      "epoch": 0.28826465860055844,
+      "grad_norm": 0.1664966493844986,
+      "learning_rate": 0.001,
+      "loss": 1.8012,
+      "step": 6814
+    },
+    {
+      "epoch": 0.2883069633640748,
+      "grad_norm": 0.1893632560968399,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 6815
+    },
+    {
+      "epoch": 0.28834926812759115,
+      "grad_norm": 0.2506442070007324,
+      "learning_rate": 0.001,
+      "loss": 2.2874,
+      "step": 6816
+    },
+    {
+      "epoch": 0.28839157289110756,
+      "grad_norm": 0.20881643891334534,
+      "learning_rate": 0.001,
+      "loss": 2.5897,
+      "step": 6817
+    },
+    {
+      "epoch": 0.2884338776546239,
+      "grad_norm": 3.2081472873687744,
+      "learning_rate": 0.001,
+      "loss": 3.7027,
+      "step": 6818
+    },
+    {
+      "epoch": 0.28847618241814027,
+      "grad_norm": 0.17766347527503967,
+      "learning_rate": 0.001,
+      "loss": 2.7331,
+      "step": 6819
+    },
+    {
+      "epoch": 0.2885184871816567,
+      "grad_norm": 0.1828794777393341,
+      "learning_rate": 0.001,
+      "loss": 1.878,
+      "step": 6820
+    },
+    {
+      "epoch": 0.28856079194517303,
+      "grad_norm": 0.2575604021549225,
+      "learning_rate": 0.001,
+      "loss": 2.9143,
+      "step": 6821
+    },
+    {
+      "epoch": 0.2886030967086894,
+      "grad_norm": 0.21907007694244385,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 6822
+    },
+    {
+      "epoch": 0.28864540147220574,
+      "grad_norm": 0.19140586256980896,
+      "learning_rate": 0.001,
+      "loss": 1.9638,
+      "step": 6823
+    },
+    {
+      "epoch": 0.28868770623572215,
+      "grad_norm": 0.829940915107727,
+      "learning_rate": 0.001,
+      "loss": 2.1286,
+      "step": 6824
+    },
+    {
+      "epoch": 0.2887300109992385,
+      "grad_norm": 0.1640363186597824,
+      "learning_rate": 0.001,
+      "loss": 1.6881,
+      "step": 6825
+    },
+    {
+      "epoch": 0.28877231576275486,
+      "grad_norm": 0.22927752137184143,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 6826
+    },
+    {
+      "epoch": 0.28881462052627127,
+      "grad_norm": 0.21113528311252594,
+      "learning_rate": 0.001,
+      "loss": 2.4409,
+      "step": 6827
+    },
+    {
+      "epoch": 0.2888569252897876,
+      "grad_norm": 0.1978234350681305,
+      "learning_rate": 0.001,
+      "loss": 2.0009,
+      "step": 6828
+    },
+    {
+      "epoch": 0.288899230053304,
+      "grad_norm": 0.43694573640823364,
+      "learning_rate": 0.001,
+      "loss": 2.7192,
+      "step": 6829
+    },
+    {
+      "epoch": 0.2889415348168204,
+      "grad_norm": 0.20024238526821136,
+      "learning_rate": 0.001,
+      "loss": 2.2269,
+      "step": 6830
+    },
+    {
+      "epoch": 0.28898383958033674,
+      "grad_norm": 0.17758913338184357,
+      "learning_rate": 0.001,
+      "loss": 1.9114,
+      "step": 6831
+    },
+    {
+      "epoch": 0.2890261443438531,
+      "grad_norm": 0.2091837078332901,
+      "learning_rate": 0.001,
+      "loss": 2.1373,
+      "step": 6832
+    },
+    {
+      "epoch": 0.2890684491073695,
+      "grad_norm": 0.28614622354507446,
+      "learning_rate": 0.001,
+      "loss": 2.3553,
+      "step": 6833
+    },
+    {
+      "epoch": 0.28911075387088586,
+      "grad_norm": 0.22399480640888214,
+      "learning_rate": 0.001,
+      "loss": 2.0642,
+      "step": 6834
+    },
+    {
+      "epoch": 0.2891530586344022,
+      "grad_norm": 0.24540039896965027,
+      "learning_rate": 0.001,
+      "loss": 1.8619,
+      "step": 6835
+    },
+    {
+      "epoch": 0.2891953633979186,
+      "grad_norm": 0.6220614314079285,
+      "learning_rate": 0.001,
+      "loss": 1.9019,
+      "step": 6836
+    },
+    {
+      "epoch": 0.289237668161435,
+      "grad_norm": 0.15628793835639954,
+      "learning_rate": 0.001,
+      "loss": 2.5514,
+      "step": 6837
+    },
+    {
+      "epoch": 0.28927997292495133,
+      "grad_norm": 2.0576629638671875,
+      "learning_rate": 0.001,
+      "loss": 2.5113,
+      "step": 6838
+    },
+    {
+      "epoch": 0.28932227768846774,
+      "grad_norm": 0.19702525436878204,
+      "learning_rate": 0.001,
+      "loss": 2.2104,
+      "step": 6839
+    },
+    {
+      "epoch": 0.2893645824519841,
+      "grad_norm": 0.20886345207691193,
+      "learning_rate": 0.001,
+      "loss": 3.0677,
+      "step": 6840
+    },
+    {
+      "epoch": 0.28940688721550045,
+      "grad_norm": 0.18030396103858948,
+      "learning_rate": 0.001,
+      "loss": 2.0693,
+      "step": 6841
+    },
+    {
+      "epoch": 0.28944919197901686,
+      "grad_norm": 0.16617818176746368,
+      "learning_rate": 0.001,
+      "loss": 1.7641,
+      "step": 6842
+    },
+    {
+      "epoch": 0.2894914967425332,
+      "grad_norm": 0.22310234606266022,
+      "learning_rate": 0.001,
+      "loss": 1.9379,
+      "step": 6843
+    },
+    {
+      "epoch": 0.28953380150604957,
+      "grad_norm": 0.19781459867954254,
+      "learning_rate": 0.001,
+      "loss": 1.9957,
+      "step": 6844
+    },
+    {
+      "epoch": 0.289576106269566,
+      "grad_norm": 0.9519227743148804,
+      "learning_rate": 0.001,
+      "loss": 2.2693,
+      "step": 6845
+    },
+    {
+      "epoch": 0.28961841103308233,
+      "grad_norm": 0.69413161277771,
+      "learning_rate": 0.001,
+      "loss": 2.4312,
+      "step": 6846
+    },
+    {
+      "epoch": 0.2896607157965987,
+      "grad_norm": 0.23835915327072144,
+      "learning_rate": 0.001,
+      "loss": 2.2305,
+      "step": 6847
+    },
+    {
+      "epoch": 0.28970302056011504,
+      "grad_norm": 0.16248169541358948,
+      "learning_rate": 0.001,
+      "loss": 1.9366,
+      "step": 6848
+    },
+    {
+      "epoch": 0.28974532532363145,
+      "grad_norm": 1.342553973197937,
+      "learning_rate": 0.001,
+      "loss": 2.4784,
+      "step": 6849
+    },
+    {
+      "epoch": 0.2897876300871478,
+      "grad_norm": 0.2539532482624054,
+      "learning_rate": 0.001,
+      "loss": 2.0127,
+      "step": 6850
+    },
+    {
+      "epoch": 0.28982993485066416,
+      "grad_norm": 0.3332529664039612,
+      "learning_rate": 0.001,
+      "loss": 2.0516,
+      "step": 6851
+    },
+    {
+      "epoch": 0.28987223961418057,
+      "grad_norm": 0.25316402316093445,
+      "learning_rate": 0.001,
+      "loss": 2.1151,
+      "step": 6852
+    },
+    {
+      "epoch": 0.2899145443776969,
+      "grad_norm": 0.2206842005252838,
+      "learning_rate": 0.001,
+      "loss": 2.1286,
+      "step": 6853
+    },
+    {
+      "epoch": 0.2899568491412133,
+      "grad_norm": 0.18992824852466583,
+      "learning_rate": 0.001,
+      "loss": 2.1739,
+      "step": 6854
+    },
+    {
+      "epoch": 0.2899991539047297,
+      "grad_norm": 0.22509132325649261,
+      "learning_rate": 0.001,
+      "loss": 1.8315,
+      "step": 6855
+    },
+    {
+      "epoch": 0.29004145866824604,
+      "grad_norm": 0.22096151113510132,
+      "learning_rate": 0.001,
+      "loss": 2.0061,
+      "step": 6856
+    },
+    {
+      "epoch": 0.2900837634317624,
+      "grad_norm": 0.27027440071105957,
+      "learning_rate": 0.001,
+      "loss": 2.488,
+      "step": 6857
+    },
+    {
+      "epoch": 0.2901260681952788,
+      "grad_norm": 0.2823043465614319,
+      "learning_rate": 0.001,
+      "loss": 2.5708,
+      "step": 6858
+    },
+    {
+      "epoch": 0.29016837295879516,
+      "grad_norm": 0.2073088437318802,
+      "learning_rate": 0.001,
+      "loss": 2.6952,
+      "step": 6859
+    },
+    {
+      "epoch": 0.2902106777223115,
+      "grad_norm": 0.18988396227359772,
+      "learning_rate": 0.001,
+      "loss": 2.1142,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2902529824858279,
+      "grad_norm": 0.1997257024049759,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 6861
+    },
+    {
+      "epoch": 0.2902952872493443,
+      "grad_norm": 0.17761564254760742,
+      "learning_rate": 0.001,
+      "loss": 1.7944,
+      "step": 6862
+    },
+    {
+      "epoch": 0.29033759201286063,
+      "grad_norm": 0.258060485124588,
+      "learning_rate": 0.001,
+      "loss": 3.3614,
+      "step": 6863
+    },
+    {
+      "epoch": 0.29037989677637704,
+      "grad_norm": 0.17017389833927155,
+      "learning_rate": 0.001,
+      "loss": 1.5792,
+      "step": 6864
+    },
+    {
+      "epoch": 0.2904222015398934,
+      "grad_norm": 0.4084915518760681,
+      "learning_rate": 0.001,
+      "loss": 1.8688,
+      "step": 6865
+    },
+    {
+      "epoch": 0.29046450630340975,
+      "grad_norm": 0.2108771950006485,
+      "learning_rate": 0.001,
+      "loss": 2.1209,
+      "step": 6866
+    },
+    {
+      "epoch": 0.29050681106692616,
+      "grad_norm": 0.42666342854499817,
+      "learning_rate": 0.001,
+      "loss": 2.7563,
+      "step": 6867
+    },
+    {
+      "epoch": 0.2905491158304425,
+      "grad_norm": 0.18676477670669556,
+      "learning_rate": 0.001,
+      "loss": 2.3573,
+      "step": 6868
+    },
+    {
+      "epoch": 0.29059142059395887,
+      "grad_norm": 0.586787760257721,
+      "learning_rate": 0.001,
+      "loss": 2.2616,
+      "step": 6869
+    },
+    {
+      "epoch": 0.2906337253574752,
+      "grad_norm": 2.2827725410461426,
+      "learning_rate": 0.001,
+      "loss": 1.948,
+      "step": 6870
+    },
+    {
+      "epoch": 0.29067603012099164,
+      "grad_norm": 0.1944393664598465,
+      "learning_rate": 0.001,
+      "loss": 2.7219,
+      "step": 6871
+    },
+    {
+      "epoch": 0.290718334884508,
+      "grad_norm": 2.946514844894409,
+      "learning_rate": 0.001,
+      "loss": 2.3146,
+      "step": 6872
+    },
+    {
+      "epoch": 0.29076063964802434,
+      "grad_norm": 0.410149484872818,
+      "learning_rate": 0.001,
+      "loss": 2.6918,
+      "step": 6873
+    },
+    {
+      "epoch": 0.29080294441154075,
+      "grad_norm": 0.18303263187408447,
+      "learning_rate": 0.001,
+      "loss": 1.4522,
+      "step": 6874
+    },
+    {
+      "epoch": 0.2908452491750571,
+      "grad_norm": 0.8593881726264954,
+      "learning_rate": 0.001,
+      "loss": 2.6251,
+      "step": 6875
+    },
+    {
+      "epoch": 0.29088755393857346,
+      "grad_norm": 2.545989751815796,
+      "learning_rate": 0.001,
+      "loss": 2.0312,
+      "step": 6876
+    },
+    {
+      "epoch": 0.29092985870208987,
+      "grad_norm": 0.30340850353240967,
+      "learning_rate": 0.001,
+      "loss": 2.2144,
+      "step": 6877
+    },
+    {
+      "epoch": 0.2909721634656062,
+      "grad_norm": 0.23748478293418884,
+      "learning_rate": 0.001,
+      "loss": 3.4675,
+      "step": 6878
+    },
+    {
+      "epoch": 0.2910144682291226,
+      "grad_norm": 18.90664291381836,
+      "learning_rate": 0.001,
+      "loss": 2.7135,
+      "step": 6879
+    },
+    {
+      "epoch": 0.291056772992639,
+      "grad_norm": 0.2496711015701294,
+      "learning_rate": 0.001,
+      "loss": 2.5996,
+      "step": 6880
+    },
+    {
+      "epoch": 0.29109907775615534,
+      "grad_norm": 1.5730305910110474,
+      "learning_rate": 0.001,
+      "loss": 3.1332,
+      "step": 6881
+    },
+    {
+      "epoch": 0.2911413825196717,
+      "grad_norm": 0.4019278585910797,
+      "learning_rate": 0.001,
+      "loss": 1.94,
+      "step": 6882
+    },
+    {
+      "epoch": 0.2911836872831881,
+      "grad_norm": 2.1472578048706055,
+      "learning_rate": 0.001,
+      "loss": 2.5725,
+      "step": 6883
+    },
+    {
+      "epoch": 0.29122599204670446,
+      "grad_norm": 0.42344650626182556,
+      "learning_rate": 0.001,
+      "loss": 2.5348,
+      "step": 6884
+    },
+    {
+      "epoch": 0.2912682968102208,
+      "grad_norm": 0.7056704759597778,
+      "learning_rate": 0.001,
+      "loss": 2.6225,
+      "step": 6885
+    },
+    {
+      "epoch": 0.2913106015737372,
+      "grad_norm": 0.6706409454345703,
+      "learning_rate": 0.001,
+      "loss": 2.3904,
+      "step": 6886
+    },
+    {
+      "epoch": 0.2913529063372536,
+      "grad_norm": 0.35459625720977783,
+      "learning_rate": 0.001,
+      "loss": 2.1607,
+      "step": 6887
+    },
+    {
+      "epoch": 0.29139521110076994,
+      "grad_norm": 0.3938932418823242,
+      "learning_rate": 0.001,
+      "loss": 2.4787,
+      "step": 6888
+    },
+    {
+      "epoch": 0.29143751586428635,
+      "grad_norm": 0.2852618992328644,
+      "learning_rate": 0.001,
+      "loss": 2.3739,
+      "step": 6889
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 1.7088735103607178,
+      "learning_rate": 0.001,
+      "loss": 2.6956,
+      "step": 6890
+    },
+    {
+      "epoch": 0.29152212539131905,
+      "grad_norm": 0.21869906783103943,
+      "learning_rate": 0.001,
+      "loss": 2.1744,
+      "step": 6891
+    },
+    {
+      "epoch": 0.2915644301548354,
+      "grad_norm": 0.40579238533973694,
+      "learning_rate": 0.001,
+      "loss": 2.3586,
+      "step": 6892
+    },
+    {
+      "epoch": 0.2916067349183518,
+      "grad_norm": 0.26919931173324585,
+      "learning_rate": 0.001,
+      "loss": 1.6718,
+      "step": 6893
+    },
+    {
+      "epoch": 0.29164903968186817,
+      "grad_norm": 0.33402884006500244,
+      "learning_rate": 0.001,
+      "loss": 3.2099,
+      "step": 6894
+    },
+    {
+      "epoch": 0.2916913444453845,
+      "grad_norm": 5.1144490242004395,
+      "learning_rate": 0.001,
+      "loss": 2.5433,
+      "step": 6895
+    },
+    {
+      "epoch": 0.29173364920890094,
+      "grad_norm": 27.068161010742188,
+      "learning_rate": 0.001,
+      "loss": 2.5207,
+      "step": 6896
+    },
+    {
+      "epoch": 0.2917759539724173,
+      "grad_norm": 1.7865883111953735,
+      "learning_rate": 0.001,
+      "loss": 3.3911,
+      "step": 6897
+    },
+    {
+      "epoch": 0.29181825873593364,
+      "grad_norm": 2.192460298538208,
+      "learning_rate": 0.001,
+      "loss": 2.4101,
+      "step": 6898
+    },
+    {
+      "epoch": 0.29186056349945005,
+      "grad_norm": 0.3550018072128296,
+      "learning_rate": 0.001,
+      "loss": 2.3135,
+      "step": 6899
+    },
+    {
+      "epoch": 0.2919028682629664,
+      "grad_norm": 0.24947704374790192,
+      "learning_rate": 0.001,
+      "loss": 3.6274,
+      "step": 6900
+    },
+    {
+      "epoch": 0.29194517302648276,
+      "grad_norm": 0.23705358803272247,
+      "learning_rate": 0.001,
+      "loss": 3.0381,
+      "step": 6901
+    },
+    {
+      "epoch": 0.2919874777899992,
+      "grad_norm": 1.2429789304733276,
+      "learning_rate": 0.001,
+      "loss": 2.1528,
+      "step": 6902
+    },
+    {
+      "epoch": 0.2920297825535155,
+      "grad_norm": 0.28556519746780396,
+      "learning_rate": 0.001,
+      "loss": 1.9076,
+      "step": 6903
+    },
+    {
+      "epoch": 0.2920720873170319,
+      "grad_norm": 3.0117995738983154,
+      "learning_rate": 0.001,
+      "loss": 2.4768,
+      "step": 6904
+    },
+    {
+      "epoch": 0.2921143920805483,
+      "grad_norm": 0.43369370698928833,
+      "learning_rate": 0.001,
+      "loss": 1.9007,
+      "step": 6905
+    },
+    {
+      "epoch": 0.29215669684406465,
+      "grad_norm": 0.2250383198261261,
+      "learning_rate": 0.001,
+      "loss": 2.0614,
+      "step": 6906
+    },
+    {
+      "epoch": 0.292199001607581,
+      "grad_norm": 0.26837220788002014,
+      "learning_rate": 0.001,
+      "loss": 2.5939,
+      "step": 6907
+    },
+    {
+      "epoch": 0.2922413063710974,
+      "grad_norm": 0.20127150416374207,
+      "learning_rate": 0.001,
+      "loss": 2.2609,
+      "step": 6908
+    },
+    {
+      "epoch": 0.29228361113461376,
+      "grad_norm": 0.9822574257850647,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 6909
+    },
+    {
+      "epoch": 0.2923259158981301,
+      "grad_norm": 0.2298155575990677,
+      "learning_rate": 0.001,
+      "loss": 1.9496,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2923682206616465,
+      "grad_norm": 1.8333818912506104,
+      "learning_rate": 0.001,
+      "loss": 2.3752,
+      "step": 6911
+    },
+    {
+      "epoch": 0.2924105254251629,
+      "grad_norm": 0.18054918944835663,
+      "learning_rate": 0.001,
+      "loss": 2.273,
+      "step": 6912
+    },
+    {
+      "epoch": 0.29245283018867924,
+      "grad_norm": 0.22334930300712585,
+      "learning_rate": 0.001,
+      "loss": 1.8855,
+      "step": 6913
+    },
+    {
+      "epoch": 0.2924951349521956,
+      "grad_norm": 0.22284521162509918,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 6914
+    },
+    {
+      "epoch": 0.292537439715712,
+      "grad_norm": 0.18738651275634766,
+      "learning_rate": 0.001,
+      "loss": 3.3555,
+      "step": 6915
+    },
+    {
+      "epoch": 0.29257974447922835,
+      "grad_norm": 0.23615574836730957,
+      "learning_rate": 0.001,
+      "loss": 2.1617,
+      "step": 6916
+    },
+    {
+      "epoch": 0.2926220492427447,
+      "grad_norm": 0.19093096256256104,
+      "learning_rate": 0.001,
+      "loss": 2.0053,
+      "step": 6917
+    },
+    {
+      "epoch": 0.2926643540062611,
+      "grad_norm": 0.22282744944095612,
+      "learning_rate": 0.001,
+      "loss": 2.3443,
+      "step": 6918
+    },
+    {
+      "epoch": 0.2927066587697775,
+      "grad_norm": 0.35610735416412354,
+      "learning_rate": 0.001,
+      "loss": 2.5339,
+      "step": 6919
+    },
+    {
+      "epoch": 0.2927489635332938,
+      "grad_norm": 0.2231018990278244,
+      "learning_rate": 0.001,
+      "loss": 2.6499,
+      "step": 6920
+    },
+    {
+      "epoch": 0.29279126829681024,
+      "grad_norm": 9.339323043823242,
+      "learning_rate": 0.001,
+      "loss": 2.2003,
+      "step": 6921
+    },
+    {
+      "epoch": 0.2928335730603266,
+      "grad_norm": 0.27400538325309753,
+      "learning_rate": 0.001,
+      "loss": 2.139,
+      "step": 6922
+    },
+    {
+      "epoch": 0.29287587782384294,
+      "grad_norm": 1.792051076889038,
+      "learning_rate": 0.001,
+      "loss": 3.2322,
+      "step": 6923
+    },
+    {
+      "epoch": 0.29291818258735935,
+      "grad_norm": 0.20132683217525482,
+      "learning_rate": 0.001,
+      "loss": 1.7977,
+      "step": 6924
+    },
+    {
+      "epoch": 0.2929604873508757,
+      "grad_norm": 0.22404009103775024,
+      "learning_rate": 0.001,
+      "loss": 2.1195,
+      "step": 6925
+    },
+    {
+      "epoch": 0.29300279211439206,
+      "grad_norm": 0.32715481519699097,
+      "learning_rate": 0.001,
+      "loss": 2.158,
+      "step": 6926
+    },
+    {
+      "epoch": 0.2930450968779085,
+      "grad_norm": 0.4097978472709656,
+      "learning_rate": 0.001,
+      "loss": 2.2345,
+      "step": 6927
+    },
+    {
+      "epoch": 0.2930874016414248,
+      "grad_norm": 15.934835433959961,
+      "learning_rate": 0.001,
+      "loss": 2.2525,
+      "step": 6928
+    },
+    {
+      "epoch": 0.2931297064049412,
+      "grad_norm": 0.20492605865001678,
+      "learning_rate": 0.001,
+      "loss": 2.2868,
+      "step": 6929
+    },
+    {
+      "epoch": 0.2931720111684576,
+      "grad_norm": 0.22894705832004547,
+      "learning_rate": 0.001,
+      "loss": 3.1435,
+      "step": 6930
+    },
+    {
+      "epoch": 0.29321431593197395,
+      "grad_norm": 0.19995182752609253,
+      "learning_rate": 0.001,
+      "loss": 2.3582,
+      "step": 6931
+    },
+    {
+      "epoch": 0.2932566206954903,
+      "grad_norm": 13.218109130859375,
+      "learning_rate": 0.001,
+      "loss": 4.1725,
+      "step": 6932
+    },
+    {
+      "epoch": 0.2932989254590067,
+      "grad_norm": 151.23451232910156,
+      "learning_rate": 0.001,
+      "loss": 1.6978,
+      "step": 6933
+    },
+    {
+      "epoch": 0.29334123022252306,
+      "grad_norm": 0.6989339590072632,
+      "learning_rate": 0.001,
+      "loss": 1.5756,
+      "step": 6934
+    },
+    {
+      "epoch": 0.2933835349860394,
+      "grad_norm": 11.406463623046875,
+      "learning_rate": 0.001,
+      "loss": 2.1872,
+      "step": 6935
+    },
+    {
+      "epoch": 0.2934258397495558,
+      "grad_norm": 0.19447572529315948,
+      "learning_rate": 0.001,
+      "loss": 1.9257,
+      "step": 6936
+    },
+    {
+      "epoch": 0.2934681445130722,
+      "grad_norm": 0.7553207874298096,
+      "learning_rate": 0.001,
+      "loss": 2.0954,
+      "step": 6937
+    },
+    {
+      "epoch": 0.29351044927658854,
+      "grad_norm": 0.6458339095115662,
+      "learning_rate": 0.001,
+      "loss": 1.9439,
+      "step": 6938
+    },
+    {
+      "epoch": 0.2935527540401049,
+      "grad_norm": 4.88887357711792,
+      "learning_rate": 0.001,
+      "loss": 2.7551,
+      "step": 6939
+    },
+    {
+      "epoch": 0.2935950588036213,
+      "grad_norm": 0.16294671595096588,
+      "learning_rate": 0.001,
+      "loss": 2.1921,
+      "step": 6940
+    },
+    {
+      "epoch": 0.29363736356713765,
+      "grad_norm": 0.1700448840856552,
+      "learning_rate": 0.001,
+      "loss": 2.3544,
+      "step": 6941
+    },
+    {
+      "epoch": 0.293679668330654,
+      "grad_norm": 0.9373911023139954,
+      "learning_rate": 0.001,
+      "loss": 3.519,
+      "step": 6942
+    },
+    {
+      "epoch": 0.2937219730941704,
+      "grad_norm": 0.16765141487121582,
+      "learning_rate": 0.001,
+      "loss": 1.753,
+      "step": 6943
+    },
+    {
+      "epoch": 0.2937642778576868,
+      "grad_norm": 0.785889744758606,
+      "learning_rate": 0.001,
+      "loss": 2.0261,
+      "step": 6944
+    },
+    {
+      "epoch": 0.2938065826212031,
+      "grad_norm": 0.36694565415382385,
+      "learning_rate": 0.001,
+      "loss": 2.8762,
+      "step": 6945
+    },
+    {
+      "epoch": 0.29384888738471954,
+      "grad_norm": 2.4932944774627686,
+      "learning_rate": 0.001,
+      "loss": 3.5164,
+      "step": 6946
+    },
+    {
+      "epoch": 0.2938911921482359,
+      "grad_norm": 0.3365854024887085,
+      "learning_rate": 0.001,
+      "loss": 2.5705,
+      "step": 6947
+    },
+    {
+      "epoch": 0.29393349691175225,
+      "grad_norm": 0.18186473846435547,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 6948
+    },
+    {
+      "epoch": 0.29397580167526866,
+      "grad_norm": 0.18802011013031006,
+      "learning_rate": 0.001,
+      "loss": 1.6162,
+      "step": 6949
+    },
+    {
+      "epoch": 0.294018106438785,
+      "grad_norm": 0.6282293796539307,
+      "learning_rate": 0.001,
+      "loss": 2.0641,
+      "step": 6950
+    },
+    {
+      "epoch": 0.29406041120230136,
+      "grad_norm": 0.18980659544467926,
+      "learning_rate": 0.001,
+      "loss": 2.1692,
+      "step": 6951
+    },
+    {
+      "epoch": 0.2941027159658178,
+      "grad_norm": 0.25274673104286194,
+      "learning_rate": 0.001,
+      "loss": 2.1386,
+      "step": 6952
+    },
+    {
+      "epoch": 0.29414502072933413,
+      "grad_norm": 1.6487905979156494,
+      "learning_rate": 0.001,
+      "loss": 2.0647,
+      "step": 6953
+    },
+    {
+      "epoch": 0.2941873254928505,
+      "grad_norm": 0.38848909735679626,
+      "learning_rate": 0.001,
+      "loss": 1.3862,
+      "step": 6954
+    },
+    {
+      "epoch": 0.2942296302563669,
+      "grad_norm": 0.5868200063705444,
+      "learning_rate": 0.001,
+      "loss": 2.8731,
+      "step": 6955
+    },
+    {
+      "epoch": 0.29427193501988325,
+      "grad_norm": 117.72666931152344,
+      "learning_rate": 0.001,
+      "loss": 3.0126,
+      "step": 6956
+    },
+    {
+      "epoch": 0.2943142397833996,
+      "grad_norm": 0.19856488704681396,
+      "learning_rate": 0.001,
+      "loss": 2.3552,
+      "step": 6957
+    },
+    {
+      "epoch": 0.294356544546916,
+      "grad_norm": 0.29041945934295654,
+      "learning_rate": 0.001,
+      "loss": 3.019,
+      "step": 6958
+    },
+    {
+      "epoch": 0.29439884931043236,
+      "grad_norm": 0.22051405906677246,
+      "learning_rate": 0.001,
+      "loss": 2.6603,
+      "step": 6959
+    },
+    {
+      "epoch": 0.2944411540739487,
+      "grad_norm": 0.1814699023962021,
+      "learning_rate": 0.001,
+      "loss": 1.5751,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2944834588374651,
+      "grad_norm": 8.469980239868164,
+      "learning_rate": 0.001,
+      "loss": 1.8405,
+      "step": 6961
+    },
+    {
+      "epoch": 0.2945257636009815,
+      "grad_norm": 0.3591289222240448,
+      "learning_rate": 0.001,
+      "loss": 1.9221,
+      "step": 6962
+    },
+    {
+      "epoch": 0.29456806836449784,
+      "grad_norm": 0.23890052735805511,
+      "learning_rate": 0.001,
+      "loss": 2.0497,
+      "step": 6963
+    },
+    {
+      "epoch": 0.2946103731280142,
+      "grad_norm": 0.29208847880363464,
+      "learning_rate": 0.001,
+      "loss": 2.6245,
+      "step": 6964
+    },
+    {
+      "epoch": 0.2946526778915306,
+      "grad_norm": 0.19358113408088684,
+      "learning_rate": 0.001,
+      "loss": 2.2433,
+      "step": 6965
+    },
+    {
+      "epoch": 0.29469498265504696,
+      "grad_norm": 0.24694213271141052,
+      "learning_rate": 0.001,
+      "loss": 1.7421,
+      "step": 6966
+    },
+    {
+      "epoch": 0.2947372874185633,
+      "grad_norm": 0.22472073137760162,
+      "learning_rate": 0.001,
+      "loss": 2.1198,
+      "step": 6967
+    },
+    {
+      "epoch": 0.2947795921820797,
+      "grad_norm": 0.3076910376548767,
+      "learning_rate": 0.001,
+      "loss": 1.9558,
+      "step": 6968
+    },
+    {
+      "epoch": 0.2948218969455961,
+      "grad_norm": 0.21519528329372406,
+      "learning_rate": 0.001,
+      "loss": 3.0599,
+      "step": 6969
+    },
+    {
+      "epoch": 0.29486420170911243,
+      "grad_norm": 0.2277740240097046,
+      "learning_rate": 0.001,
+      "loss": 2.8818,
+      "step": 6970
+    },
+    {
+      "epoch": 0.29490650647262884,
+      "grad_norm": 4.881925106048584,
+      "learning_rate": 0.001,
+      "loss": 1.9913,
+      "step": 6971
+    },
+    {
+      "epoch": 0.2949488112361452,
+      "grad_norm": 0.2809251844882965,
+      "learning_rate": 0.001,
+      "loss": 2.8118,
+      "step": 6972
+    },
+    {
+      "epoch": 0.29499111599966155,
+      "grad_norm": 0.322844535112381,
+      "learning_rate": 0.001,
+      "loss": 3.4413,
+      "step": 6973
+    },
+    {
+      "epoch": 0.29503342076317796,
+      "grad_norm": 0.2614366114139557,
+      "learning_rate": 0.001,
+      "loss": 2.5353,
+      "step": 6974
+    },
+    {
+      "epoch": 0.2950757255266943,
+      "grad_norm": 0.23935487866401672,
+      "learning_rate": 0.001,
+      "loss": 2.6036,
+      "step": 6975
+    },
+    {
+      "epoch": 0.29511803029021066,
+      "grad_norm": 0.2764032185077667,
+      "learning_rate": 0.001,
+      "loss": 2.577,
+      "step": 6976
+    },
+    {
+      "epoch": 0.2951603350537271,
+      "grad_norm": 0.2084626853466034,
+      "learning_rate": 0.001,
+      "loss": 2.0175,
+      "step": 6977
+    },
+    {
+      "epoch": 0.29520263981724343,
+      "grad_norm": 0.39110100269317627,
+      "learning_rate": 0.001,
+      "loss": 3.6172,
+      "step": 6978
+    },
+    {
+      "epoch": 0.2952449445807598,
+      "grad_norm": 0.5247201919555664,
+      "learning_rate": 0.001,
+      "loss": 3.2458,
+      "step": 6979
+    },
+    {
+      "epoch": 0.2952872493442762,
+      "grad_norm": 0.2991565465927124,
+      "learning_rate": 0.001,
+      "loss": 3.4346,
+      "step": 6980
+    },
+    {
+      "epoch": 0.29532955410779255,
+      "grad_norm": 0.218788743019104,
+      "learning_rate": 0.001,
+      "loss": 2.654,
+      "step": 6981
+    },
+    {
+      "epoch": 0.2953718588713089,
+      "grad_norm": 0.19704671204090118,
+      "learning_rate": 0.001,
+      "loss": 2.1183,
+      "step": 6982
+    },
+    {
+      "epoch": 0.29541416363482526,
+      "grad_norm": 0.1847701370716095,
+      "learning_rate": 0.001,
+      "loss": 2.2644,
+      "step": 6983
+    },
+    {
+      "epoch": 0.29545646839834167,
+      "grad_norm": 0.20297656953334808,
+      "learning_rate": 0.001,
+      "loss": 1.8673,
+      "step": 6984
+    },
+    {
+      "epoch": 0.295498773161858,
+      "grad_norm": 1.8230969905853271,
+      "learning_rate": 0.001,
+      "loss": 1.7627,
+      "step": 6985
+    },
+    {
+      "epoch": 0.2955410779253744,
+      "grad_norm": 0.22176900506019592,
+      "learning_rate": 0.001,
+      "loss": 2.2876,
+      "step": 6986
+    },
+    {
+      "epoch": 0.2955833826888908,
+      "grad_norm": 0.1842503547668457,
+      "learning_rate": 0.001,
+      "loss": 2.764,
+      "step": 6987
+    },
+    {
+      "epoch": 0.29562568745240714,
+      "grad_norm": 0.2775016725063324,
+      "learning_rate": 0.001,
+      "loss": 1.8702,
+      "step": 6988
+    },
+    {
+      "epoch": 0.2956679922159235,
+      "grad_norm": 0.21142591536045074,
+      "learning_rate": 0.001,
+      "loss": 3.3746,
+      "step": 6989
+    },
+    {
+      "epoch": 0.2957102969794399,
+      "grad_norm": 0.2473433017730713,
+      "learning_rate": 0.001,
+      "loss": 2.1965,
+      "step": 6990
+    },
+    {
+      "epoch": 0.29575260174295626,
+      "grad_norm": 0.2217579185962677,
+      "learning_rate": 0.001,
+      "loss": 3.064,
+      "step": 6991
+    },
+    {
+      "epoch": 0.2957949065064726,
+      "grad_norm": 0.2861359417438507,
+      "learning_rate": 0.001,
+      "loss": 2.5936,
+      "step": 6992
+    },
+    {
+      "epoch": 0.295837211269989,
+      "grad_norm": 0.6411170363426208,
+      "learning_rate": 0.001,
+      "loss": 2.0748,
+      "step": 6993
+    },
+    {
+      "epoch": 0.2958795160335054,
+      "grad_norm": 0.20746761560440063,
+      "learning_rate": 0.001,
+      "loss": 1.8649,
+      "step": 6994
+    },
+    {
+      "epoch": 0.29592182079702173,
+      "grad_norm": 0.19803252816200256,
+      "learning_rate": 0.001,
+      "loss": 2.6345,
+      "step": 6995
+    },
+    {
+      "epoch": 0.29596412556053814,
+      "grad_norm": 0.19498823583126068,
+      "learning_rate": 0.001,
+      "loss": 2.8652,
+      "step": 6996
+    },
+    {
+      "epoch": 0.2960064303240545,
+      "grad_norm": 0.18126071989536285,
+      "learning_rate": 0.001,
+      "loss": 1.7674,
+      "step": 6997
+    },
+    {
+      "epoch": 0.29604873508757085,
+      "grad_norm": 0.17771077156066895,
+      "learning_rate": 0.001,
+      "loss": 1.8077,
+      "step": 6998
+    },
+    {
+      "epoch": 0.29609103985108726,
+      "grad_norm": 0.2599290609359741,
+      "learning_rate": 0.001,
+      "loss": 2.336,
+      "step": 6999
+    },
+    {
+      "epoch": 0.2961333446146036,
+      "grad_norm": 0.3005380928516388,
+      "learning_rate": 0.001,
+      "loss": 2.4787,
+      "step": 7000
+    },
+    {
+      "epoch": 0.29617564937811997,
+      "grad_norm": 0.18659712374210358,
+      "learning_rate": 0.001,
+      "loss": 2.1932,
+      "step": 7001
+    },
+    {
+      "epoch": 0.2962179541416364,
+      "grad_norm": 0.19324898719787598,
+      "learning_rate": 0.001,
+      "loss": 2.0008,
+      "step": 7002
+    },
+    {
+      "epoch": 0.29626025890515273,
+      "grad_norm": 0.3487750291824341,
+      "learning_rate": 0.001,
+      "loss": 2.2476,
+      "step": 7003
+    },
+    {
+      "epoch": 0.2963025636686691,
+      "grad_norm": 0.20598524808883667,
+      "learning_rate": 0.001,
+      "loss": 2.0175,
+      "step": 7004
+    },
+    {
+      "epoch": 0.29634486843218544,
+      "grad_norm": 0.20581454038619995,
+      "learning_rate": 0.001,
+      "loss": 2.5366,
+      "step": 7005
+    },
+    {
+      "epoch": 0.29638717319570185,
+      "grad_norm": 0.19093303382396698,
+      "learning_rate": 0.001,
+      "loss": 2.4369,
+      "step": 7006
+    },
+    {
+      "epoch": 0.2964294779592182,
+      "grad_norm": 0.1642286777496338,
+      "learning_rate": 0.001,
+      "loss": 2.6144,
+      "step": 7007
+    },
+    {
+      "epoch": 0.29647178272273456,
+      "grad_norm": 0.5392095446586609,
+      "learning_rate": 0.001,
+      "loss": 2.2483,
+      "step": 7008
+    },
+    {
+      "epoch": 0.29651408748625097,
+      "grad_norm": 0.22686930000782013,
+      "learning_rate": 0.001,
+      "loss": 1.9133,
+      "step": 7009
+    },
+    {
+      "epoch": 0.2965563922497673,
+      "grad_norm": 0.22740709781646729,
+      "learning_rate": 0.001,
+      "loss": 1.4307,
+      "step": 7010
+    },
+    {
+      "epoch": 0.2965986970132837,
+      "grad_norm": 0.19837690889835358,
+      "learning_rate": 0.001,
+      "loss": 1.8655,
+      "step": 7011
+    },
+    {
+      "epoch": 0.2966410017768001,
+      "grad_norm": 0.32683438062667847,
+      "learning_rate": 0.001,
+      "loss": 2.1361,
+      "step": 7012
+    },
+    {
+      "epoch": 0.29668330654031644,
+      "grad_norm": 4.837608814239502,
+      "learning_rate": 0.001,
+      "loss": 3.105,
+      "step": 7013
+    },
+    {
+      "epoch": 0.2967256113038328,
+      "grad_norm": 0.1853974610567093,
+      "learning_rate": 0.001,
+      "loss": 1.8663,
+      "step": 7014
+    },
+    {
+      "epoch": 0.2967679160673492,
+      "grad_norm": 3.4521446228027344,
+      "learning_rate": 0.001,
+      "loss": 2.0969,
+      "step": 7015
+    },
+    {
+      "epoch": 0.29681022083086556,
+      "grad_norm": 11.476473808288574,
+      "learning_rate": 0.001,
+      "loss": 2.1661,
+      "step": 7016
+    },
+    {
+      "epoch": 0.2968525255943819,
+      "grad_norm": 0.2935603857040405,
+      "learning_rate": 0.001,
+      "loss": 2.8407,
+      "step": 7017
+    },
+    {
+      "epoch": 0.2968948303578983,
+      "grad_norm": 0.22967737913131714,
+      "learning_rate": 0.001,
+      "loss": 2.347,
+      "step": 7018
+    },
+    {
+      "epoch": 0.2969371351214147,
+      "grad_norm": 0.24542132019996643,
+      "learning_rate": 0.001,
+      "loss": 2.7476,
+      "step": 7019
+    },
+    {
+      "epoch": 0.29697943988493103,
+      "grad_norm": 0.3539610207080841,
+      "learning_rate": 0.001,
+      "loss": 3.4266,
+      "step": 7020
+    },
+    {
+      "epoch": 0.29702174464844744,
+      "grad_norm": 2.1936635971069336,
+      "learning_rate": 0.001,
+      "loss": 2.4714,
+      "step": 7021
+    },
+    {
+      "epoch": 0.2970640494119638,
+      "grad_norm": 36.311824798583984,
+      "learning_rate": 0.001,
+      "loss": 2.2848,
+      "step": 7022
+    },
+    {
+      "epoch": 0.29710635417548015,
+      "grad_norm": 0.7815760970115662,
+      "learning_rate": 0.001,
+      "loss": 3.0078,
+      "step": 7023
+    },
+    {
+      "epoch": 0.29714865893899656,
+      "grad_norm": 0.25058576464653015,
+      "learning_rate": 0.001,
+      "loss": 1.7609,
+      "step": 7024
+    },
+    {
+      "epoch": 0.2971909637025129,
+      "grad_norm": 0.32843679189682007,
+      "learning_rate": 0.001,
+      "loss": 2.2564,
+      "step": 7025
+    },
+    {
+      "epoch": 0.29723326846602927,
+      "grad_norm": 0.31752750277519226,
+      "learning_rate": 0.001,
+      "loss": 3.8093,
+      "step": 7026
+    },
+    {
+      "epoch": 0.2972755732295456,
+      "grad_norm": 0.3511388301849365,
+      "learning_rate": 0.001,
+      "loss": 2.9753,
+      "step": 7027
+    },
+    {
+      "epoch": 0.29731787799306203,
+      "grad_norm": 0.18950346112251282,
+      "learning_rate": 0.001,
+      "loss": 1.8184,
+      "step": 7028
+    },
+    {
+      "epoch": 0.2973601827565784,
+      "grad_norm": 1.3205153942108154,
+      "learning_rate": 0.001,
+      "loss": 2.4554,
+      "step": 7029
+    },
+    {
+      "epoch": 0.29740248752009474,
+      "grad_norm": 0.5435582995414734,
+      "learning_rate": 0.001,
+      "loss": 2.9064,
+      "step": 7030
+    },
+    {
+      "epoch": 0.29744479228361115,
+      "grad_norm": 0.6610992550849915,
+      "learning_rate": 0.001,
+      "loss": 2.4811,
+      "step": 7031
+    },
+    {
+      "epoch": 0.2974870970471275,
+      "grad_norm": 0.43292543292045593,
+      "learning_rate": 0.001,
+      "loss": 2.1861,
+      "step": 7032
+    },
+    {
+      "epoch": 0.29752940181064386,
+      "grad_norm": 0.29178565740585327,
+      "learning_rate": 0.001,
+      "loss": 2.0465,
+      "step": 7033
+    },
+    {
+      "epoch": 0.29757170657416027,
+      "grad_norm": 0.4427778720855713,
+      "learning_rate": 0.001,
+      "loss": 2.1454,
+      "step": 7034
+    },
+    {
+      "epoch": 0.2976140113376766,
+      "grad_norm": 0.36568501591682434,
+      "learning_rate": 0.001,
+      "loss": 2.513,
+      "step": 7035
+    },
+    {
+      "epoch": 0.297656316101193,
+      "grad_norm": 0.7165125608444214,
+      "learning_rate": 0.001,
+      "loss": 1.9268,
+      "step": 7036
+    },
+    {
+      "epoch": 0.2976986208647094,
+      "grad_norm": 0.23342393338680267,
+      "learning_rate": 0.001,
+      "loss": 2.8504,
+      "step": 7037
+    },
+    {
+      "epoch": 0.29774092562822574,
+      "grad_norm": 0.24309587478637695,
+      "learning_rate": 0.001,
+      "loss": 2.6698,
+      "step": 7038
+    },
+    {
+      "epoch": 0.2977832303917421,
+      "grad_norm": 1.0469741821289062,
+      "learning_rate": 0.001,
+      "loss": 4.0021,
+      "step": 7039
+    },
+    {
+      "epoch": 0.2978255351552585,
+      "grad_norm": 4.104727268218994,
+      "learning_rate": 0.001,
+      "loss": 2.4115,
+      "step": 7040
+    },
+    {
+      "epoch": 0.29786783991877486,
+      "grad_norm": 2.293821096420288,
+      "learning_rate": 0.001,
+      "loss": 2.3395,
+      "step": 7041
+    },
+    {
+      "epoch": 0.2979101446822912,
+      "grad_norm": 0.2298813909292221,
+      "learning_rate": 0.001,
+      "loss": 1.913,
+      "step": 7042
+    },
+    {
+      "epoch": 0.2979524494458076,
+      "grad_norm": 0.5691928267478943,
+      "learning_rate": 0.001,
+      "loss": 2.0014,
+      "step": 7043
+    },
+    {
+      "epoch": 0.297994754209324,
+      "grad_norm": 0.4775361120700836,
+      "learning_rate": 0.001,
+      "loss": 2.5362,
+      "step": 7044
+    },
+    {
+      "epoch": 0.29803705897284033,
+      "grad_norm": 0.364666223526001,
+      "learning_rate": 0.001,
+      "loss": 2.8652,
+      "step": 7045
+    },
+    {
+      "epoch": 0.29807936373635674,
+      "grad_norm": 0.9387134313583374,
+      "learning_rate": 0.001,
+      "loss": 2.1165,
+      "step": 7046
+    },
+    {
+      "epoch": 0.2981216684998731,
+      "grad_norm": 0.2410208135843277,
+      "learning_rate": 0.001,
+      "loss": 3.0745,
+      "step": 7047
+    },
+    {
+      "epoch": 0.29816397326338945,
+      "grad_norm": 0.2509887218475342,
+      "learning_rate": 0.001,
+      "loss": 1.7976,
+      "step": 7048
+    },
+    {
+      "epoch": 0.2982062780269058,
+      "grad_norm": 1.8324601650238037,
+      "learning_rate": 0.001,
+      "loss": 2.6896,
+      "step": 7049
+    },
+    {
+      "epoch": 0.2982485827904222,
+      "grad_norm": 0.342830091714859,
+      "learning_rate": 0.001,
+      "loss": 2.2083,
+      "step": 7050
+    },
+    {
+      "epoch": 0.29829088755393857,
+      "grad_norm": 0.263128399848938,
+      "learning_rate": 0.001,
+      "loss": 2.8128,
+      "step": 7051
+    },
+    {
+      "epoch": 0.2983331923174549,
+      "grad_norm": 0.2615799903869629,
+      "learning_rate": 0.001,
+      "loss": 1.9154,
+      "step": 7052
+    },
+    {
+      "epoch": 0.29837549708097133,
+      "grad_norm": 0.3741340935230255,
+      "learning_rate": 0.001,
+      "loss": 2.8024,
+      "step": 7053
+    },
+    {
+      "epoch": 0.2984178018444877,
+      "grad_norm": 0.21344539523124695,
+      "learning_rate": 0.001,
+      "loss": 2.0821,
+      "step": 7054
+    },
+    {
+      "epoch": 0.29846010660800404,
+      "grad_norm": 7.319907188415527,
+      "learning_rate": 0.001,
+      "loss": 2.5422,
+      "step": 7055
+    },
+    {
+      "epoch": 0.29850241137152045,
+      "grad_norm": 0.1762029081583023,
+      "learning_rate": 0.001,
+      "loss": 1.8821,
+      "step": 7056
+    },
+    {
+      "epoch": 0.2985447161350368,
+      "grad_norm": 0.21324655413627625,
+      "learning_rate": 0.001,
+      "loss": 1.4527,
+      "step": 7057
+    },
+    {
+      "epoch": 0.29858702089855316,
+      "grad_norm": 0.35666510462760925,
+      "learning_rate": 0.001,
+      "loss": 3.0456,
+      "step": 7058
+    },
+    {
+      "epoch": 0.29862932566206957,
+      "grad_norm": 0.21469701826572418,
+      "learning_rate": 0.001,
+      "loss": 2.0336,
+      "step": 7059
+    },
+    {
+      "epoch": 0.2986716304255859,
+      "grad_norm": 0.39980030059814453,
+      "learning_rate": 0.001,
+      "loss": 3.3548,
+      "step": 7060
+    },
+    {
+      "epoch": 0.2987139351891023,
+      "grad_norm": 0.23621757328510284,
+      "learning_rate": 0.001,
+      "loss": 2.3665,
+      "step": 7061
+    },
+    {
+      "epoch": 0.2987562399526187,
+      "grad_norm": 0.26678037643432617,
+      "learning_rate": 0.001,
+      "loss": 2.1806,
+      "step": 7062
+    },
+    {
+      "epoch": 0.29879854471613504,
+      "grad_norm": 0.5849965810775757,
+      "learning_rate": 0.001,
+      "loss": 1.8085,
+      "step": 7063
+    },
+    {
+      "epoch": 0.2988408494796514,
+      "grad_norm": 0.4421706199645996,
+      "learning_rate": 0.001,
+      "loss": 2.1855,
+      "step": 7064
+    },
+    {
+      "epoch": 0.2988831542431678,
+      "grad_norm": 0.2031574845314026,
+      "learning_rate": 0.001,
+      "loss": 1.6082,
+      "step": 7065
+    },
+    {
+      "epoch": 0.29892545900668416,
+      "grad_norm": 0.24453617632389069,
+      "learning_rate": 0.001,
+      "loss": 1.9515,
+      "step": 7066
+    },
+    {
+      "epoch": 0.2989677637702005,
+      "grad_norm": 0.28984296321868896,
+      "learning_rate": 0.001,
+      "loss": 2.3851,
+      "step": 7067
+    },
+    {
+      "epoch": 0.2990100685337169,
+      "grad_norm": 0.22082221508026123,
+      "learning_rate": 0.001,
+      "loss": 1.8154,
+      "step": 7068
+    },
+    {
+      "epoch": 0.2990523732972333,
+      "grad_norm": 0.20017731189727783,
+      "learning_rate": 0.001,
+      "loss": 1.9377,
+      "step": 7069
+    },
+    {
+      "epoch": 0.29909467806074963,
+      "grad_norm": 0.1992572695016861,
+      "learning_rate": 0.001,
+      "loss": 1.5338,
+      "step": 7070
+    },
+    {
+      "epoch": 0.29913698282426604,
+      "grad_norm": 3.3462400436401367,
+      "learning_rate": 0.001,
+      "loss": 2.8609,
+      "step": 7071
+    },
+    {
+      "epoch": 0.2991792875877824,
+      "grad_norm": 0.1810547411441803,
+      "learning_rate": 0.001,
+      "loss": 2.7506,
+      "step": 7072
+    },
+    {
+      "epoch": 0.29922159235129875,
+      "grad_norm": 0.351933091878891,
+      "learning_rate": 0.001,
+      "loss": 1.9723,
+      "step": 7073
+    },
+    {
+      "epoch": 0.2992638971148151,
+      "grad_norm": 0.2719806134700775,
+      "learning_rate": 0.001,
+      "loss": 2.5537,
+      "step": 7074
+    },
+    {
+      "epoch": 0.2993062018783315,
+      "grad_norm": 1.2375730276107788,
+      "learning_rate": 0.001,
+      "loss": 2.7143,
+      "step": 7075
+    },
+    {
+      "epoch": 0.29934850664184787,
+      "grad_norm": 0.27189958095550537,
+      "learning_rate": 0.001,
+      "loss": 2.8949,
+      "step": 7076
+    },
+    {
+      "epoch": 0.2993908114053642,
+      "grad_norm": 2.186654567718506,
+      "learning_rate": 0.001,
+      "loss": 3.3108,
+      "step": 7077
+    },
+    {
+      "epoch": 0.29943311616888063,
+      "grad_norm": 0.2271018624305725,
+      "learning_rate": 0.001,
+      "loss": 1.9664,
+      "step": 7078
+    },
+    {
+      "epoch": 0.299475420932397,
+      "grad_norm": 0.20025469362735748,
+      "learning_rate": 0.001,
+      "loss": 2.14,
+      "step": 7079
+    },
+    {
+      "epoch": 0.29951772569591334,
+      "grad_norm": 1.0243349075317383,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 7080
+    },
+    {
+      "epoch": 0.29956003045942975,
+      "grad_norm": 0.27229562401771545,
+      "learning_rate": 0.001,
+      "loss": 3.3795,
+      "step": 7081
+    },
+    {
+      "epoch": 0.2996023352229461,
+      "grad_norm": 0.20607994496822357,
+      "learning_rate": 0.001,
+      "loss": 2.6859,
+      "step": 7082
+    },
+    {
+      "epoch": 0.29964463998646246,
+      "grad_norm": 0.18946023285388947,
+      "learning_rate": 0.001,
+      "loss": 1.9361,
+      "step": 7083
+    },
+    {
+      "epoch": 0.29968694474997887,
+      "grad_norm": 0.22694577276706696,
+      "learning_rate": 0.001,
+      "loss": 2.872,
+      "step": 7084
+    },
+    {
+      "epoch": 0.2997292495134952,
+      "grad_norm": 0.19697017967700958,
+      "learning_rate": 0.001,
+      "loss": 1.8643,
+      "step": 7085
+    },
+    {
+      "epoch": 0.2997715542770116,
+      "grad_norm": 0.20350822806358337,
+      "learning_rate": 0.001,
+      "loss": 1.8414,
+      "step": 7086
+    },
+    {
+      "epoch": 0.299813859040528,
+      "grad_norm": 1.2988654375076294,
+      "learning_rate": 0.001,
+      "loss": 2.5616,
+      "step": 7087
+    },
+    {
+      "epoch": 0.29985616380404434,
+      "grad_norm": 7.280746936798096,
+      "learning_rate": 0.001,
+      "loss": 2.5211,
+      "step": 7088
+    },
+    {
+      "epoch": 0.2998984685675607,
+      "grad_norm": 0.9409200549125671,
+      "learning_rate": 0.001,
+      "loss": 1.8807,
+      "step": 7089
+    },
+    {
+      "epoch": 0.2999407733310771,
+      "grad_norm": 0.20737378299236298,
+      "learning_rate": 0.001,
+      "loss": 2.2614,
+      "step": 7090
+    },
+    {
+      "epoch": 0.29998307809459346,
+      "grad_norm": 0.19399072229862213,
+      "learning_rate": 0.001,
+      "loss": 1.5657,
+      "step": 7091
+    },
+    {
+      "epoch": 0.3000253828581098,
+      "grad_norm": 2.7120447158813477,
+      "learning_rate": 0.001,
+      "loss": 3.1643,
+      "step": 7092
+    },
+    {
+      "epoch": 0.3000676876216262,
+      "grad_norm": 0.23883989453315735,
+      "learning_rate": 0.001,
+      "loss": 1.4332,
+      "step": 7093
+    },
+    {
+      "epoch": 0.3001099923851426,
+      "grad_norm": 0.24009618163108826,
+      "learning_rate": 0.001,
+      "loss": 2.0072,
+      "step": 7094
+    },
+    {
+      "epoch": 0.30015229714865893,
+      "grad_norm": 0.27275872230529785,
+      "learning_rate": 0.001,
+      "loss": 1.8004,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3001946019121753,
+      "grad_norm": 0.3240390419960022,
+      "learning_rate": 0.001,
+      "loss": 2.4354,
+      "step": 7096
+    },
+    {
+      "epoch": 0.3002369066756917,
+      "grad_norm": 0.2155575454235077,
+      "learning_rate": 0.001,
+      "loss": 2.4924,
+      "step": 7097
+    },
+    {
+      "epoch": 0.30027921143920805,
+      "grad_norm": 0.26140278577804565,
+      "learning_rate": 0.001,
+      "loss": 3.0696,
+      "step": 7098
+    },
+    {
+      "epoch": 0.3003215162027244,
+      "grad_norm": 0.20411476492881775,
+      "learning_rate": 0.001,
+      "loss": 2.8706,
+      "step": 7099
+    },
+    {
+      "epoch": 0.3003638209662408,
+      "grad_norm": 0.26278483867645264,
+      "learning_rate": 0.001,
+      "loss": 2.6213,
+      "step": 7100
+    },
+    {
+      "epoch": 0.30040612572975717,
+      "grad_norm": 0.23542357981204987,
+      "learning_rate": 0.001,
+      "loss": 2.4739,
+      "step": 7101
+    },
+    {
+      "epoch": 0.3004484304932735,
+      "grad_norm": 1.478679895401001,
+      "learning_rate": 0.001,
+      "loss": 2.8704,
+      "step": 7102
+    },
+    {
+      "epoch": 0.30049073525678993,
+      "grad_norm": 0.22880683839321136,
+      "learning_rate": 0.001,
+      "loss": 2.2796,
+      "step": 7103
+    },
+    {
+      "epoch": 0.3005330400203063,
+      "grad_norm": 0.22094206511974335,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 7104
+    },
+    {
+      "epoch": 0.30057534478382264,
+      "grad_norm": 0.2112313210964203,
+      "learning_rate": 0.001,
+      "loss": 2.5717,
+      "step": 7105
+    },
+    {
+      "epoch": 0.30061764954733905,
+      "grad_norm": 0.19635865092277527,
+      "learning_rate": 0.001,
+      "loss": 1.7046,
+      "step": 7106
+    },
+    {
+      "epoch": 0.3006599543108554,
+      "grad_norm": 0.42038843035697937,
+      "learning_rate": 0.001,
+      "loss": 3.0465,
+      "step": 7107
+    },
+    {
+      "epoch": 0.30070225907437176,
+      "grad_norm": 0.2118614763021469,
+      "learning_rate": 0.001,
+      "loss": 2.0218,
+      "step": 7108
+    },
+    {
+      "epoch": 0.30074456383788817,
+      "grad_norm": 0.1884838491678238,
+      "learning_rate": 0.001,
+      "loss": 1.6187,
+      "step": 7109
+    },
+    {
+      "epoch": 0.3007868686014045,
+      "grad_norm": 1.542157530784607,
+      "learning_rate": 0.001,
+      "loss": 2.7272,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3008291733649209,
+      "grad_norm": 0.19439256191253662,
+      "learning_rate": 0.001,
+      "loss": 1.8408,
+      "step": 7111
+    },
+    {
+      "epoch": 0.3008714781284373,
+      "grad_norm": 0.21478168666362762,
+      "learning_rate": 0.001,
+      "loss": 2.5348,
+      "step": 7112
+    },
+    {
+      "epoch": 0.30091378289195364,
+      "grad_norm": 0.2651658058166504,
+      "learning_rate": 0.001,
+      "loss": 2.2955,
+      "step": 7113
+    },
+    {
+      "epoch": 0.30095608765547,
+      "grad_norm": 0.20375876128673553,
+      "learning_rate": 0.001,
+      "loss": 2.7385,
+      "step": 7114
+    },
+    {
+      "epoch": 0.3009983924189864,
+      "grad_norm": 1.0855168104171753,
+      "learning_rate": 0.001,
+      "loss": 1.7983,
+      "step": 7115
+    },
+    {
+      "epoch": 0.30104069718250276,
+      "grad_norm": 1.9149056673049927,
+      "learning_rate": 0.001,
+      "loss": 2.4025,
+      "step": 7116
+    },
+    {
+      "epoch": 0.3010830019460191,
+      "grad_norm": 0.2241235226392746,
+      "learning_rate": 0.001,
+      "loss": 2.6799,
+      "step": 7117
+    },
+    {
+      "epoch": 0.30112530670953547,
+      "grad_norm": 0.21922898292541504,
+      "learning_rate": 0.001,
+      "loss": 2.2658,
+      "step": 7118
+    },
+    {
+      "epoch": 0.3011676114730519,
+      "grad_norm": 0.2675611078739166,
+      "learning_rate": 0.001,
+      "loss": 1.4285,
+      "step": 7119
+    },
+    {
+      "epoch": 0.30120991623656823,
+      "grad_norm": 0.5068369507789612,
+      "learning_rate": 0.001,
+      "loss": 2.1726,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3012522210000846,
+      "grad_norm": 0.2993006110191345,
+      "learning_rate": 0.001,
+      "loss": 2.4697,
+      "step": 7121
+    },
+    {
+      "epoch": 0.301294525763601,
+      "grad_norm": 0.31761422753334045,
+      "learning_rate": 0.001,
+      "loss": 2.6937,
+      "step": 7122
+    },
+    {
+      "epoch": 0.30133683052711735,
+      "grad_norm": 0.2203313410282135,
+      "learning_rate": 0.001,
+      "loss": 2.4712,
+      "step": 7123
+    },
+    {
+      "epoch": 0.3013791352906337,
+      "grad_norm": 0.2073364555835724,
+      "learning_rate": 0.001,
+      "loss": 2.3447,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3014214400541501,
+      "grad_norm": 0.27055227756500244,
+      "learning_rate": 0.001,
+      "loss": 1.9593,
+      "step": 7125
+    },
+    {
+      "epoch": 0.30146374481766647,
+      "grad_norm": 0.19267384707927704,
+      "learning_rate": 0.001,
+      "loss": 2.5415,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3015060495811828,
+      "grad_norm": 0.21630926430225372,
+      "learning_rate": 0.001,
+      "loss": 3.0923,
+      "step": 7127
+    },
+    {
+      "epoch": 0.30154835434469923,
+      "grad_norm": 0.17075297236442566,
+      "learning_rate": 0.001,
+      "loss": 1.9153,
+      "step": 7128
+    },
+    {
+      "epoch": 0.3015906591082156,
+      "grad_norm": 2.4271461963653564,
+      "learning_rate": 0.001,
+      "loss": 2.0476,
+      "step": 7129
+    },
+    {
+      "epoch": 0.30163296387173194,
+      "grad_norm": 0.1639959067106247,
+      "learning_rate": 0.001,
+      "loss": 1.7935,
+      "step": 7130
+    },
+    {
+      "epoch": 0.30167526863524835,
+      "grad_norm": 0.17364169657230377,
+      "learning_rate": 0.001,
+      "loss": 2.065,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3017175733987647,
+      "grad_norm": 0.19003529846668243,
+      "learning_rate": 0.001,
+      "loss": 2.4537,
+      "step": 7132
+    },
+    {
+      "epoch": 0.30175987816228106,
+      "grad_norm": 6.679607391357422,
+      "learning_rate": 0.001,
+      "loss": 2.5314,
+      "step": 7133
+    },
+    {
+      "epoch": 0.30180218292579747,
+      "grad_norm": 0.20183472335338593,
+      "learning_rate": 0.001,
+      "loss": 2.6579,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3018444876893138,
+      "grad_norm": 2.5921237468719482,
+      "learning_rate": 0.001,
+      "loss": 2.0,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.23701635003089905,
+      "learning_rate": 0.001,
+      "loss": 2.559,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3019290972163466,
+      "grad_norm": 0.5344852805137634,
+      "learning_rate": 0.001,
+      "loss": 2.0646,
+      "step": 7137
+    },
+    {
+      "epoch": 0.30197140197986294,
+      "grad_norm": 0.20276984572410583,
+      "learning_rate": 0.001,
+      "loss": 2.3369,
+      "step": 7138
+    },
+    {
+      "epoch": 0.3020137067433793,
+      "grad_norm": 0.23550361394882202,
+      "learning_rate": 0.001,
+      "loss": 2.3764,
+      "step": 7139
+    },
+    {
+      "epoch": 0.30205601150689565,
+      "grad_norm": 0.1865498572587967,
+      "learning_rate": 0.001,
+      "loss": 1.8937,
+      "step": 7140
+    },
+    {
+      "epoch": 0.30209831627041206,
+      "grad_norm": 0.18712198734283447,
+      "learning_rate": 0.001,
+      "loss": 2.3899,
+      "step": 7141
+    },
+    {
+      "epoch": 0.3021406210339284,
+      "grad_norm": 1.8322774171829224,
+      "learning_rate": 0.001,
+      "loss": 1.7824,
+      "step": 7142
+    },
+    {
+      "epoch": 0.30218292579744477,
+      "grad_norm": 1.2885193824768066,
+      "learning_rate": 0.001,
+      "loss": 1.8025,
+      "step": 7143
+    },
+    {
+      "epoch": 0.3022252305609612,
+      "grad_norm": 1.9234724044799805,
+      "learning_rate": 0.001,
+      "loss": 3.0237,
+      "step": 7144
+    },
+    {
+      "epoch": 0.30226753532447753,
+      "grad_norm": 0.21429021656513214,
+      "learning_rate": 0.001,
+      "loss": 2.3753,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3023098400879939,
+      "grad_norm": 0.1841968446969986,
+      "learning_rate": 0.001,
+      "loss": 1.9292,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3023521448515103,
+      "grad_norm": 1.093137264251709,
+      "learning_rate": 0.001,
+      "loss": 1.7962,
+      "step": 7147
+    },
+    {
+      "epoch": 0.30239444961502665,
+      "grad_norm": 0.3102741539478302,
+      "learning_rate": 0.001,
+      "loss": 2.7395,
+      "step": 7148
+    },
+    {
+      "epoch": 0.302436754378543,
+      "grad_norm": 1.5967626571655273,
+      "learning_rate": 0.001,
+      "loss": 2.026,
+      "step": 7149
+    },
+    {
+      "epoch": 0.3024790591420594,
+      "grad_norm": 0.1948966681957245,
+      "learning_rate": 0.001,
+      "loss": 2.3974,
+      "step": 7150
+    },
+    {
+      "epoch": 0.30252136390557577,
+      "grad_norm": 5.692415714263916,
+      "learning_rate": 0.001,
+      "loss": 1.7977,
+      "step": 7151
+    },
+    {
+      "epoch": 0.3025636686690921,
+      "grad_norm": 0.3406045436859131,
+      "learning_rate": 0.001,
+      "loss": 3.7244,
+      "step": 7152
+    },
+    {
+      "epoch": 0.30260597343260853,
+      "grad_norm": 0.19864249229431152,
+      "learning_rate": 0.001,
+      "loss": 2.0721,
+      "step": 7153
+    },
+    {
+      "epoch": 0.3026482781961249,
+      "grad_norm": 0.22038595378398895,
+      "learning_rate": 0.001,
+      "loss": 1.618,
+      "step": 7154
+    },
+    {
+      "epoch": 0.30269058295964124,
+      "grad_norm": 0.17631886899471283,
+      "learning_rate": 0.001,
+      "loss": 2.1884,
+      "step": 7155
+    },
+    {
+      "epoch": 0.30273288772315765,
+      "grad_norm": 0.1953791379928589,
+      "learning_rate": 0.001,
+      "loss": 1.4864,
+      "step": 7156
+    },
+    {
+      "epoch": 0.302775192486674,
+      "grad_norm": 0.3843629062175751,
+      "learning_rate": 0.001,
+      "loss": 2.1935,
+      "step": 7157
+    },
+    {
+      "epoch": 0.30281749725019036,
+      "grad_norm": 0.562131941318512,
+      "learning_rate": 0.001,
+      "loss": 2.6914,
+      "step": 7158
+    },
+    {
+      "epoch": 0.30285980201370677,
+      "grad_norm": 0.30789318680763245,
+      "learning_rate": 0.001,
+      "loss": 3.5069,
+      "step": 7159
+    },
+    {
+      "epoch": 0.3029021067772231,
+      "grad_norm": 5.395096778869629,
+      "learning_rate": 0.001,
+      "loss": 2.2879,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3029444115407395,
+      "grad_norm": 0.22398367524147034,
+      "learning_rate": 0.001,
+      "loss": 2.5926,
+      "step": 7161
+    },
+    {
+      "epoch": 0.30298671630425583,
+      "grad_norm": 0.3233366012573242,
+      "learning_rate": 0.001,
+      "loss": 2.6733,
+      "step": 7162
+    },
+    {
+      "epoch": 0.30302902106777224,
+      "grad_norm": 0.20136740803718567,
+      "learning_rate": 0.001,
+      "loss": 1.6234,
+      "step": 7163
+    },
+    {
+      "epoch": 0.3030713258312886,
+      "grad_norm": 0.20825202763080597,
+      "learning_rate": 0.001,
+      "loss": 2.3597,
+      "step": 7164
+    },
+    {
+      "epoch": 0.30311363059480495,
+      "grad_norm": 1.0632126331329346,
+      "learning_rate": 0.001,
+      "loss": 1.722,
+      "step": 7165
+    },
+    {
+      "epoch": 0.30315593535832136,
+      "grad_norm": 0.22308948636054993,
+      "learning_rate": 0.001,
+      "loss": 2.4832,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3031982401218377,
+      "grad_norm": 0.22294947504997253,
+      "learning_rate": 0.001,
+      "loss": 2.4689,
+      "step": 7167
+    },
+    {
+      "epoch": 0.30324054488535407,
+      "grad_norm": 0.18253259360790253,
+      "learning_rate": 0.001,
+      "loss": 2.1326,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3032828496488705,
+      "grad_norm": 2.8974242210388184,
+      "learning_rate": 0.001,
+      "loss": 2.3822,
+      "step": 7169
+    },
+    {
+      "epoch": 0.30332515441238683,
+      "grad_norm": 0.19481533765792847,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3033674591759032,
+      "grad_norm": 0.42890578508377075,
+      "learning_rate": 0.001,
+      "loss": 2.517,
+      "step": 7171
+    },
+    {
+      "epoch": 0.3034097639394196,
+      "grad_norm": 0.24793508648872375,
+      "learning_rate": 0.001,
+      "loss": 1.7478,
+      "step": 7172
+    },
+    {
+      "epoch": 0.30345206870293595,
+      "grad_norm": 0.1969098150730133,
+      "learning_rate": 0.001,
+      "loss": 1.9163,
+      "step": 7173
+    },
+    {
+      "epoch": 0.3034943734664523,
+      "grad_norm": 0.37608563899993896,
+      "learning_rate": 0.001,
+      "loss": 2.0317,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3035366782299687,
+      "grad_norm": 0.30871760845184326,
+      "learning_rate": 0.001,
+      "loss": 1.9419,
+      "step": 7175
+    },
+    {
+      "epoch": 0.30357898299348507,
+      "grad_norm": 0.18646669387817383,
+      "learning_rate": 0.001,
+      "loss": 1.7309,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3036212877570014,
+      "grad_norm": 0.17259854078292847,
+      "learning_rate": 0.001,
+      "loss": 1.7845,
+      "step": 7177
+    },
+    {
+      "epoch": 0.30366359252051783,
+      "grad_norm": 0.17856760323047638,
+      "learning_rate": 0.001,
+      "loss": 2.4272,
+      "step": 7178
+    },
+    {
+      "epoch": 0.3037058972840342,
+      "grad_norm": 2.3033905029296875,
+      "learning_rate": 0.001,
+      "loss": 2.732,
+      "step": 7179
+    },
+    {
+      "epoch": 0.30374820204755054,
+      "grad_norm": 0.4790777266025543,
+      "learning_rate": 0.001,
+      "loss": 2.2848,
+      "step": 7180
+    },
+    {
+      "epoch": 0.30379050681106695,
+      "grad_norm": 0.2084704339504242,
+      "learning_rate": 0.001,
+      "loss": 3.398,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3038328115745833,
+      "grad_norm": 0.3605867028236389,
+      "learning_rate": 0.001,
+      "loss": 3.1206,
+      "step": 7182
+    },
+    {
+      "epoch": 0.30387511633809966,
+      "grad_norm": 0.17509806156158447,
+      "learning_rate": 0.001,
+      "loss": 1.7837,
+      "step": 7183
+    },
+    {
+      "epoch": 0.303917421101616,
+      "grad_norm": 0.22098810970783234,
+      "learning_rate": 0.001,
+      "loss": 2.0828,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3039597258651324,
+      "grad_norm": 3.9246859550476074,
+      "learning_rate": 0.001,
+      "loss": 3.2388,
+      "step": 7185
+    },
+    {
+      "epoch": 0.3040020306286488,
+      "grad_norm": 0.1589023768901825,
+      "learning_rate": 0.001,
+      "loss": 2.3912,
+      "step": 7186
+    },
+    {
+      "epoch": 0.30404433539216513,
+      "grad_norm": 0.20874576270580292,
+      "learning_rate": 0.001,
+      "loss": 2.8516,
+      "step": 7187
+    },
+    {
+      "epoch": 0.30408664015568154,
+      "grad_norm": 1.3450440168380737,
+      "learning_rate": 0.001,
+      "loss": 1.8097,
+      "step": 7188
+    },
+    {
+      "epoch": 0.3041289449191979,
+      "grad_norm": 1.1989301443099976,
+      "learning_rate": 0.001,
+      "loss": 1.8467,
+      "step": 7189
+    },
+    {
+      "epoch": 0.30417124968271425,
+      "grad_norm": 0.21062646806240082,
+      "learning_rate": 0.001,
+      "loss": 2.1306,
+      "step": 7190
+    },
+    {
+      "epoch": 0.30421355444623066,
+      "grad_norm": 4.8882012367248535,
+      "learning_rate": 0.001,
+      "loss": 2.5283,
+      "step": 7191
+    },
+    {
+      "epoch": 0.304255859209747,
+      "grad_norm": 0.22382013499736786,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 7192
+    },
+    {
+      "epoch": 0.30429816397326337,
+      "grad_norm": 0.6270066499710083,
+      "learning_rate": 0.001,
+      "loss": 1.8903,
+      "step": 7193
+    },
+    {
+      "epoch": 0.3043404687367798,
+      "grad_norm": 0.21127991378307343,
+      "learning_rate": 0.001,
+      "loss": 3.1819,
+      "step": 7194
+    },
+    {
+      "epoch": 0.30438277350029613,
+      "grad_norm": 0.20672987401485443,
+      "learning_rate": 0.001,
+      "loss": 3.1049,
+      "step": 7195
+    },
+    {
+      "epoch": 0.3044250782638125,
+      "grad_norm": 0.5075833201408386,
+      "learning_rate": 0.001,
+      "loss": 1.8776,
+      "step": 7196
+    },
+    {
+      "epoch": 0.3044673830273289,
+      "grad_norm": 0.2034783810377121,
+      "learning_rate": 0.001,
+      "loss": 1.9582,
+      "step": 7197
+    },
+    {
+      "epoch": 0.30450968779084525,
+      "grad_norm": 0.2227412313222885,
+      "learning_rate": 0.001,
+      "loss": 2.315,
+      "step": 7198
+    },
+    {
+      "epoch": 0.3045519925543616,
+      "grad_norm": 0.38511255383491516,
+      "learning_rate": 0.001,
+      "loss": 2.3305,
+      "step": 7199
+    },
+    {
+      "epoch": 0.304594297317878,
+      "grad_norm": 0.48111987113952637,
+      "learning_rate": 0.001,
+      "loss": 2.6835,
+      "step": 7200
+    },
+    {
+      "epoch": 0.30463660208139437,
+      "grad_norm": 0.39257773756980896,
+      "learning_rate": 0.001,
+      "loss": 2.4686,
+      "step": 7201
+    },
+    {
+      "epoch": 0.3046789068449107,
+      "grad_norm": 0.24036632478237152,
+      "learning_rate": 0.001,
+      "loss": 3.5692,
+      "step": 7202
+    },
+    {
+      "epoch": 0.30472121160842713,
+      "grad_norm": 1.5374091863632202,
+      "learning_rate": 0.001,
+      "loss": 2.1117,
+      "step": 7203
+    },
+    {
+      "epoch": 0.3047635163719435,
+      "grad_norm": 0.21697300672531128,
+      "learning_rate": 0.001,
+      "loss": 3.4893,
+      "step": 7204
+    },
+    {
+      "epoch": 0.30480582113545984,
+      "grad_norm": 0.2033626139163971,
+      "learning_rate": 0.001,
+      "loss": 2.6569,
+      "step": 7205
+    },
+    {
+      "epoch": 0.30484812589897625,
+      "grad_norm": 0.5761461853981018,
+      "learning_rate": 0.001,
+      "loss": 2.5943,
+      "step": 7206
+    },
+    {
+      "epoch": 0.3048904306624926,
+      "grad_norm": 0.17697875201702118,
+      "learning_rate": 0.001,
+      "loss": 1.7242,
+      "step": 7207
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.20617154240608215,
+      "learning_rate": 0.001,
+      "loss": 1.9209,
+      "step": 7208
+    },
+    {
+      "epoch": 0.3049750401895253,
+      "grad_norm": 1.6445116996765137,
+      "learning_rate": 0.001,
+      "loss": 2.3285,
+      "step": 7209
+    },
+    {
+      "epoch": 0.3050173449530417,
+      "grad_norm": 0.2126307338476181,
+      "learning_rate": 0.001,
+      "loss": 2.9291,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3050596497165581,
+      "grad_norm": 0.2812674939632416,
+      "learning_rate": 0.001,
+      "loss": 2.0021,
+      "step": 7211
+    },
+    {
+      "epoch": 0.30510195448007443,
+      "grad_norm": 0.19099974632263184,
+      "learning_rate": 0.001,
+      "loss": 1.9334,
+      "step": 7212
+    },
+    {
+      "epoch": 0.30514425924359084,
+      "grad_norm": 0.23043900728225708,
+      "learning_rate": 0.001,
+      "loss": 3.7938,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3051865640071072,
+      "grad_norm": 0.8108448386192322,
+      "learning_rate": 0.001,
+      "loss": 2.4988,
+      "step": 7214
+    },
+    {
+      "epoch": 0.30522886877062355,
+      "grad_norm": 0.22695882618427277,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 7215
+    },
+    {
+      "epoch": 0.30527117353413996,
+      "grad_norm": 0.2140110582113266,
+      "learning_rate": 0.001,
+      "loss": 2.7109,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3053134782976563,
+      "grad_norm": 0.18746429681777954,
+      "learning_rate": 0.001,
+      "loss": 1.5037,
+      "step": 7217
+    },
+    {
+      "epoch": 0.30535578306117267,
+      "grad_norm": 0.5185785293579102,
+      "learning_rate": 0.001,
+      "loss": 2.1417,
+      "step": 7218
+    },
+    {
+      "epoch": 0.3053980878246891,
+      "grad_norm": 0.22843578457832336,
+      "learning_rate": 0.001,
+      "loss": 2.9488,
+      "step": 7219
+    },
+    {
+      "epoch": 0.30544039258820543,
+      "grad_norm": 0.22175458073616028,
+      "learning_rate": 0.001,
+      "loss": 1.8912,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3054826973517218,
+      "grad_norm": 0.3052138388156891,
+      "learning_rate": 0.001,
+      "loss": 2.7883,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3055250021152382,
+      "grad_norm": 0.20926591753959656,
+      "learning_rate": 0.001,
+      "loss": 2.2513,
+      "step": 7222
+    },
+    {
+      "epoch": 0.30556730687875455,
+      "grad_norm": 0.5725785493850708,
+      "learning_rate": 0.001,
+      "loss": 2.0737,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3056096116422709,
+      "grad_norm": 0.2553465962409973,
+      "learning_rate": 0.001,
+      "loss": 2.8104,
+      "step": 7224
+    },
+    {
+      "epoch": 0.3056519164057873,
+      "grad_norm": 0.17359638214111328,
+      "learning_rate": 0.001,
+      "loss": 1.9327,
+      "step": 7225
+    },
+    {
+      "epoch": 0.30569422116930367,
+      "grad_norm": 0.38395628333091736,
+      "learning_rate": 0.001,
+      "loss": 3.3217,
+      "step": 7226
+    },
+    {
+      "epoch": 0.30573652593282,
+      "grad_norm": 0.4329422414302826,
+      "learning_rate": 0.001,
+      "loss": 3.0564,
+      "step": 7227
+    },
+    {
+      "epoch": 0.30577883069633643,
+      "grad_norm": 0.19598042964935303,
+      "learning_rate": 0.001,
+      "loss": 2.1527,
+      "step": 7228
+    },
+    {
+      "epoch": 0.3058211354598528,
+      "grad_norm": 4.16204309463501,
+      "learning_rate": 0.001,
+      "loss": 2.1645,
+      "step": 7229
+    },
+    {
+      "epoch": 0.30586344022336914,
+      "grad_norm": 0.9371064901351929,
+      "learning_rate": 0.001,
+      "loss": 2.5554,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3059057449868855,
+      "grad_norm": 89.23577117919922,
+      "learning_rate": 0.001,
+      "loss": 1.957,
+      "step": 7231
+    },
+    {
+      "epoch": 0.3059480497504019,
+      "grad_norm": 2.158701181411743,
+      "learning_rate": 0.001,
+      "loss": 2.8319,
+      "step": 7232
+    },
+    {
+      "epoch": 0.30599035451391826,
+      "grad_norm": 0.17827636003494263,
+      "learning_rate": 0.001,
+      "loss": 2.1969,
+      "step": 7233
+    },
+    {
+      "epoch": 0.3060326592774346,
+      "grad_norm": 0.21226316690444946,
+      "learning_rate": 0.001,
+      "loss": 1.4754,
+      "step": 7234
+    },
+    {
+      "epoch": 0.306074964040951,
+      "grad_norm": 0.6273579001426697,
+      "learning_rate": 0.001,
+      "loss": 1.9398,
+      "step": 7235
+    },
+    {
+      "epoch": 0.3061172688044674,
+      "grad_norm": 3.001936197280884,
+      "learning_rate": 0.001,
+      "loss": 2.9792,
+      "step": 7236
+    },
+    {
+      "epoch": 0.30615957356798373,
+      "grad_norm": 0.8433371186256409,
+      "learning_rate": 0.001,
+      "loss": 2.1398,
+      "step": 7237
+    },
+    {
+      "epoch": 0.30620187833150014,
+      "grad_norm": 0.518166720867157,
+      "learning_rate": 0.001,
+      "loss": 2.1598,
+      "step": 7238
+    },
+    {
+      "epoch": 0.3062441830950165,
+      "grad_norm": 0.25381919741630554,
+      "learning_rate": 0.001,
+      "loss": 2.3477,
+      "step": 7239
+    },
+    {
+      "epoch": 0.30628648785853285,
+      "grad_norm": 0.3235541880130768,
+      "learning_rate": 0.001,
+      "loss": 2.4599,
+      "step": 7240
+    },
+    {
+      "epoch": 0.30632879262204926,
+      "grad_norm": 0.46788913011550903,
+      "learning_rate": 0.001,
+      "loss": 2.0962,
+      "step": 7241
+    },
+    {
+      "epoch": 0.3063710973855656,
+      "grad_norm": 0.26095959544181824,
+      "learning_rate": 0.001,
+      "loss": 3.9055,
+      "step": 7242
+    },
+    {
+      "epoch": 0.30641340214908197,
+      "grad_norm": 0.44010379910469055,
+      "learning_rate": 0.001,
+      "loss": 2.2192,
+      "step": 7243
+    },
+    {
+      "epoch": 0.3064557069125984,
+      "grad_norm": 0.24772557616233826,
+      "learning_rate": 0.001,
+      "loss": 1.9465,
+      "step": 7244
+    },
+    {
+      "epoch": 0.30649801167611473,
+      "grad_norm": 0.22386042773723602,
+      "learning_rate": 0.001,
+      "loss": 2.0204,
+      "step": 7245
+    },
+    {
+      "epoch": 0.3065403164396311,
+      "grad_norm": 0.1983170062303543,
+      "learning_rate": 0.001,
+      "loss": 1.712,
+      "step": 7246
+    },
+    {
+      "epoch": 0.3065826212031475,
+      "grad_norm": 0.2919590175151825,
+      "learning_rate": 0.001,
+      "loss": 2.2712,
+      "step": 7247
+    },
+    {
+      "epoch": 0.30662492596666385,
+      "grad_norm": 1.4634429216384888,
+      "learning_rate": 0.001,
+      "loss": 1.5808,
+      "step": 7248
+    },
+    {
+      "epoch": 0.3066672307301802,
+      "grad_norm": 0.24671746790409088,
+      "learning_rate": 0.001,
+      "loss": 3.1394,
+      "step": 7249
+    },
+    {
+      "epoch": 0.3067095354936966,
+      "grad_norm": 4.342376708984375,
+      "learning_rate": 0.001,
+      "loss": 1.5847,
+      "step": 7250
+    },
+    {
+      "epoch": 0.30675184025721297,
+      "grad_norm": 0.6360762715339661,
+      "learning_rate": 0.001,
+      "loss": 2.1356,
+      "step": 7251
+    },
+    {
+      "epoch": 0.3067941450207293,
+      "grad_norm": 0.19539184868335724,
+      "learning_rate": 0.001,
+      "loss": 3.3058,
+      "step": 7252
+    },
+    {
+      "epoch": 0.3068364497842457,
+      "grad_norm": 0.22951632738113403,
+      "learning_rate": 0.001,
+      "loss": 2.0753,
+      "step": 7253
+    },
+    {
+      "epoch": 0.3068787545477621,
+      "grad_norm": 0.3070700764656067,
+      "learning_rate": 0.001,
+      "loss": 1.9365,
+      "step": 7254
+    },
+    {
+      "epoch": 0.30692105931127844,
+      "grad_norm": 0.36672693490982056,
+      "learning_rate": 0.001,
+      "loss": 3.0584,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3069633640747948,
+      "grad_norm": 0.8749989867210388,
+      "learning_rate": 0.001,
+      "loss": 1.9865,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3070056688383112,
+      "grad_norm": 0.4010057747364044,
+      "learning_rate": 0.001,
+      "loss": 2.2224,
+      "step": 7257
+    },
+    {
+      "epoch": 0.30704797360182756,
+      "grad_norm": 0.19571296870708466,
+      "learning_rate": 0.001,
+      "loss": 1.8307,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3070902783653439,
+      "grad_norm": 0.6532837748527527,
+      "learning_rate": 0.001,
+      "loss": 2.183,
+      "step": 7259
+    },
+    {
+      "epoch": 0.3071325831288603,
+      "grad_norm": 0.22160778939723969,
+      "learning_rate": 0.001,
+      "loss": 2.2073,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3071748878923767,
+      "grad_norm": 0.24020320177078247,
+      "learning_rate": 0.001,
+      "loss": 2.6333,
+      "step": 7261
+    },
+    {
+      "epoch": 0.30721719265589303,
+      "grad_norm": 0.1901521235704422,
+      "learning_rate": 0.001,
+      "loss": 2.0718,
+      "step": 7262
+    },
+    {
+      "epoch": 0.30725949741940944,
+      "grad_norm": 0.19890253245830536,
+      "learning_rate": 0.001,
+      "loss": 1.6972,
+      "step": 7263
+    },
+    {
+      "epoch": 0.3073018021829258,
+      "grad_norm": 0.17532768845558167,
+      "learning_rate": 0.001,
+      "loss": 2.0171,
+      "step": 7264
+    },
+    {
+      "epoch": 0.30734410694644215,
+      "grad_norm": 0.21217653155326843,
+      "learning_rate": 0.001,
+      "loss": 3.6026,
+      "step": 7265
+    },
+    {
+      "epoch": 0.30738641170995856,
+      "grad_norm": 0.23952527344226837,
+      "learning_rate": 0.001,
+      "loss": 1.9304,
+      "step": 7266
+    },
+    {
+      "epoch": 0.3074287164734749,
+      "grad_norm": 0.2269655019044876,
+      "learning_rate": 0.001,
+      "loss": 2.0697,
+      "step": 7267
+    },
+    {
+      "epoch": 0.30747102123699127,
+      "grad_norm": 0.21532459557056427,
+      "learning_rate": 0.001,
+      "loss": 2.6024,
+      "step": 7268
+    },
+    {
+      "epoch": 0.3075133260005077,
+      "grad_norm": 0.20854906737804413,
+      "learning_rate": 0.001,
+      "loss": 2.4159,
+      "step": 7269
+    },
+    {
+      "epoch": 0.30755563076402403,
+      "grad_norm": 0.2222970575094223,
+      "learning_rate": 0.001,
+      "loss": 3.1089,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3075979355275404,
+      "grad_norm": 0.8760474920272827,
+      "learning_rate": 0.001,
+      "loss": 1.7882,
+      "step": 7271
+    },
+    {
+      "epoch": 0.3076402402910568,
+      "grad_norm": 0.18725240230560303,
+      "learning_rate": 0.001,
+      "loss": 2.9421,
+      "step": 7272
+    },
+    {
+      "epoch": 0.30768254505457315,
+      "grad_norm": 0.9830304384231567,
+      "learning_rate": 0.001,
+      "loss": 1.7632,
+      "step": 7273
+    },
+    {
+      "epoch": 0.3077248498180895,
+      "grad_norm": 0.2174868881702423,
+      "learning_rate": 0.001,
+      "loss": 3.1616,
+      "step": 7274
+    },
+    {
+      "epoch": 0.30776715458160586,
+      "grad_norm": 0.2975056767463684,
+      "learning_rate": 0.001,
+      "loss": 2.1744,
+      "step": 7275
+    },
+    {
+      "epoch": 0.30780945934512227,
+      "grad_norm": 0.23097103834152222,
+      "learning_rate": 0.001,
+      "loss": 2.6325,
+      "step": 7276
+    },
+    {
+      "epoch": 0.3078517641086386,
+      "grad_norm": 0.20248940587043762,
+      "learning_rate": 0.001,
+      "loss": 2.3101,
+      "step": 7277
+    },
+    {
+      "epoch": 0.307894068872155,
+      "grad_norm": 0.21180301904678345,
+      "learning_rate": 0.001,
+      "loss": 2.1885,
+      "step": 7278
+    },
+    {
+      "epoch": 0.3079363736356714,
+      "grad_norm": 0.20690077543258667,
+      "learning_rate": 0.001,
+      "loss": 2.6467,
+      "step": 7279
+    },
+    {
+      "epoch": 0.30797867839918774,
+      "grad_norm": 0.19214750826358795,
+      "learning_rate": 0.001,
+      "loss": 2.5726,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3080209831627041,
+      "grad_norm": 0.6494995355606079,
+      "learning_rate": 0.001,
+      "loss": 2.6384,
+      "step": 7281
+    },
+    {
+      "epoch": 0.3080632879262205,
+      "grad_norm": 0.19168926775455475,
+      "learning_rate": 0.001,
+      "loss": 2.1541,
+      "step": 7282
+    },
+    {
+      "epoch": 0.30810559268973686,
+      "grad_norm": 0.2198658436536789,
+      "learning_rate": 0.001,
+      "loss": 2.046,
+      "step": 7283
+    },
+    {
+      "epoch": 0.3081478974532532,
+      "grad_norm": 0.23235653340816498,
+      "learning_rate": 0.001,
+      "loss": 2.2637,
+      "step": 7284
+    },
+    {
+      "epoch": 0.3081902022167696,
+      "grad_norm": 0.48129093647003174,
+      "learning_rate": 0.001,
+      "loss": 2.2735,
+      "step": 7285
+    },
+    {
+      "epoch": 0.308232506980286,
+      "grad_norm": 0.1928509920835495,
+      "learning_rate": 0.001,
+      "loss": 2.4181,
+      "step": 7286
+    },
+    {
+      "epoch": 0.30827481174380233,
+      "grad_norm": 0.21180304884910583,
+      "learning_rate": 0.001,
+      "loss": 1.9391,
+      "step": 7287
+    },
+    {
+      "epoch": 0.30831711650731874,
+      "grad_norm": 0.5490750074386597,
+      "learning_rate": 0.001,
+      "loss": 2.0321,
+      "step": 7288
+    },
+    {
+      "epoch": 0.3083594212708351,
+      "grad_norm": 0.19193410873413086,
+      "learning_rate": 0.001,
+      "loss": 2.4427,
+      "step": 7289
+    },
+    {
+      "epoch": 0.30840172603435145,
+      "grad_norm": 0.1852174997329712,
+      "learning_rate": 0.001,
+      "loss": 1.7946,
+      "step": 7290
+    },
+    {
+      "epoch": 0.30844403079786786,
+      "grad_norm": 0.1935269683599472,
+      "learning_rate": 0.001,
+      "loss": 2.5008,
+      "step": 7291
+    },
+    {
+      "epoch": 0.3084863355613842,
+      "grad_norm": 0.5893869400024414,
+      "learning_rate": 0.001,
+      "loss": 1.7927,
+      "step": 7292
+    },
+    {
+      "epoch": 0.30852864032490057,
+      "grad_norm": 0.4908425807952881,
+      "learning_rate": 0.001,
+      "loss": 2.8926,
+      "step": 7293
+    },
+    {
+      "epoch": 0.308570945088417,
+      "grad_norm": 0.20681232213974,
+      "learning_rate": 0.001,
+      "loss": 1.8814,
+      "step": 7294
+    },
+    {
+      "epoch": 0.30861324985193334,
+      "grad_norm": 1.3799690008163452,
+      "learning_rate": 0.001,
+      "loss": 2.032,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3086555546154497,
+      "grad_norm": 0.26583728194236755,
+      "learning_rate": 0.001,
+      "loss": 2.3084,
+      "step": 7296
+    },
+    {
+      "epoch": 0.30869785937896604,
+      "grad_norm": 0.21364301443099976,
+      "learning_rate": 0.001,
+      "loss": 2.5819,
+      "step": 7297
+    },
+    {
+      "epoch": 0.30874016414248245,
+      "grad_norm": 0.1767185777425766,
+      "learning_rate": 0.001,
+      "loss": 2.2752,
+      "step": 7298
+    },
+    {
+      "epoch": 0.3087824689059988,
+      "grad_norm": 0.19230739772319794,
+      "learning_rate": 0.001,
+      "loss": 2.4451,
+      "step": 7299
+    },
+    {
+      "epoch": 0.30882477366951516,
+      "grad_norm": 0.19486257433891296,
+      "learning_rate": 0.001,
+      "loss": 2.663,
+      "step": 7300
+    },
+    {
+      "epoch": 0.30886707843303157,
+      "grad_norm": 0.23837964236736298,
+      "learning_rate": 0.001,
+      "loss": 2.2232,
+      "step": 7301
+    },
+    {
+      "epoch": 0.3089093831965479,
+      "grad_norm": 0.2171543687582016,
+      "learning_rate": 0.001,
+      "loss": 2.9941,
+      "step": 7302
+    },
+    {
+      "epoch": 0.3089516879600643,
+      "grad_norm": 0.21793997287750244,
+      "learning_rate": 0.001,
+      "loss": 2.36,
+      "step": 7303
+    },
+    {
+      "epoch": 0.3089939927235807,
+      "grad_norm": 0.22793272137641907,
+      "learning_rate": 0.001,
+      "loss": 1.9991,
+      "step": 7304
+    },
+    {
+      "epoch": 0.30903629748709704,
+      "grad_norm": 0.23069630563259125,
+      "learning_rate": 0.001,
+      "loss": 1.94,
+      "step": 7305
+    },
+    {
+      "epoch": 0.3090786022506134,
+      "grad_norm": 0.25365108251571655,
+      "learning_rate": 0.001,
+      "loss": 2.1211,
+      "step": 7306
+    },
+    {
+      "epoch": 0.3091209070141298,
+      "grad_norm": 0.981309175491333,
+      "learning_rate": 0.001,
+      "loss": 2.3867,
+      "step": 7307
+    },
+    {
+      "epoch": 0.30916321177764616,
+      "grad_norm": 0.2560966908931732,
+      "learning_rate": 0.001,
+      "loss": 3.7126,
+      "step": 7308
+    },
+    {
+      "epoch": 0.3092055165411625,
+      "grad_norm": 0.18280810117721558,
+      "learning_rate": 0.001,
+      "loss": 1.8863,
+      "step": 7309
+    },
+    {
+      "epoch": 0.3092478213046789,
+      "grad_norm": 1.1143485307693481,
+      "learning_rate": 0.001,
+      "loss": 2.3522,
+      "step": 7310
+    },
+    {
+      "epoch": 0.3092901260681953,
+      "grad_norm": 0.42736926674842834,
+      "learning_rate": 0.001,
+      "loss": 2.6357,
+      "step": 7311
+    },
+    {
+      "epoch": 0.30933243083171164,
+      "grad_norm": 0.3606073260307312,
+      "learning_rate": 0.001,
+      "loss": 3.6859,
+      "step": 7312
+    },
+    {
+      "epoch": 0.30937473559522805,
+      "grad_norm": 0.5770615339279175,
+      "learning_rate": 0.001,
+      "loss": 2.6721,
+      "step": 7313
+    },
+    {
+      "epoch": 0.3094170403587444,
+      "grad_norm": 0.3912082314491272,
+      "learning_rate": 0.001,
+      "loss": 2.4154,
+      "step": 7314
+    },
+    {
+      "epoch": 0.30945934512226075,
+      "grad_norm": 0.21277593076229095,
+      "learning_rate": 0.001,
+      "loss": 1.8694,
+      "step": 7315
+    },
+    {
+      "epoch": 0.30950164988577716,
+      "grad_norm": 0.17572276294231415,
+      "learning_rate": 0.001,
+      "loss": 2.1676,
+      "step": 7316
+    },
+    {
+      "epoch": 0.3095439546492935,
+      "grad_norm": 0.20815758407115936,
+      "learning_rate": 0.001,
+      "loss": 3.3823,
+      "step": 7317
+    },
+    {
+      "epoch": 0.30958625941280987,
+      "grad_norm": 0.745972752571106,
+      "learning_rate": 0.001,
+      "loss": 2.1945,
+      "step": 7318
+    },
+    {
+      "epoch": 0.3096285641763263,
+      "grad_norm": 0.20606304705142975,
+      "learning_rate": 0.001,
+      "loss": 1.9767,
+      "step": 7319
+    },
+    {
+      "epoch": 0.30967086893984264,
+      "grad_norm": 0.19630566239356995,
+      "learning_rate": 0.001,
+      "loss": 1.8691,
+      "step": 7320
+    },
+    {
+      "epoch": 0.309713173703359,
+      "grad_norm": 0.48055750131607056,
+      "learning_rate": 0.001,
+      "loss": 1.9261,
+      "step": 7321
+    },
+    {
+      "epoch": 0.30975547846687534,
+      "grad_norm": 2.7202658653259277,
+      "learning_rate": 0.001,
+      "loss": 2.2182,
+      "step": 7322
+    },
+    {
+      "epoch": 0.30979778323039175,
+      "grad_norm": 0.5181911587715149,
+      "learning_rate": 0.001,
+      "loss": 4.2096,
+      "step": 7323
+    },
+    {
+      "epoch": 0.3098400879939081,
+      "grad_norm": 0.17982327938079834,
+      "learning_rate": 0.001,
+      "loss": 1.9361,
+      "step": 7324
+    },
+    {
+      "epoch": 0.30988239275742446,
+      "grad_norm": 0.19139717519283295,
+      "learning_rate": 0.001,
+      "loss": 2.3097,
+      "step": 7325
+    },
+    {
+      "epoch": 0.3099246975209409,
+      "grad_norm": 0.2525288462638855,
+      "learning_rate": 0.001,
+      "loss": 2.176,
+      "step": 7326
+    },
+    {
+      "epoch": 0.3099670022844572,
+      "grad_norm": 0.3544459640979767,
+      "learning_rate": 0.001,
+      "loss": 3.4766,
+      "step": 7327
+    },
+    {
+      "epoch": 0.3100093070479736,
+      "grad_norm": 0.20294804871082306,
+      "learning_rate": 0.001,
+      "loss": 1.8903,
+      "step": 7328
+    },
+    {
+      "epoch": 0.31005161181149,
+      "grad_norm": 2.3171162605285645,
+      "learning_rate": 0.001,
+      "loss": 2.4474,
+      "step": 7329
+    },
+    {
+      "epoch": 0.31009391657500635,
+      "grad_norm": 0.22611065208911896,
+      "learning_rate": 0.001,
+      "loss": 2.5749,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3101362213385227,
+      "grad_norm": 1.3491861820220947,
+      "learning_rate": 0.001,
+      "loss": 1.9441,
+      "step": 7331
+    },
+    {
+      "epoch": 0.3101785261020391,
+      "grad_norm": 0.22623848915100098,
+      "learning_rate": 0.001,
+      "loss": 2.951,
+      "step": 7332
+    },
+    {
+      "epoch": 0.31022083086555546,
+      "grad_norm": 1.876600742340088,
+      "learning_rate": 0.001,
+      "loss": 1.9068,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3102631356290718,
+      "grad_norm": 3.5861880779266357,
+      "learning_rate": 0.001,
+      "loss": 2.7255,
+      "step": 7334
+    },
+    {
+      "epoch": 0.3103054403925882,
+      "grad_norm": 0.2411908656358719,
+      "learning_rate": 0.001,
+      "loss": 1.5527,
+      "step": 7335
+    },
+    {
+      "epoch": 0.3103477451561046,
+      "grad_norm": 1.0028561353683472,
+      "learning_rate": 0.001,
+      "loss": 2.5271,
+      "step": 7336
+    },
+    {
+      "epoch": 0.31039004991962094,
+      "grad_norm": 0.23291008174419403,
+      "learning_rate": 0.001,
+      "loss": 2.0163,
+      "step": 7337
+    },
+    {
+      "epoch": 0.31043235468313735,
+      "grad_norm": 0.2906794846057892,
+      "learning_rate": 0.001,
+      "loss": 2.2704,
+      "step": 7338
+    },
+    {
+      "epoch": 0.3104746594466537,
+      "grad_norm": 0.7202196717262268,
+      "learning_rate": 0.001,
+      "loss": 2.9688,
+      "step": 7339
+    },
+    {
+      "epoch": 0.31051696421017005,
+      "grad_norm": 0.2299274504184723,
+      "learning_rate": 0.001,
+      "loss": 1.9663,
+      "step": 7340
+    },
+    {
+      "epoch": 0.31055926897368646,
+      "grad_norm": 0.25570148229599,
+      "learning_rate": 0.001,
+      "loss": 2.3401,
+      "step": 7341
+    },
+    {
+      "epoch": 0.3106015737372028,
+      "grad_norm": 0.18588940799236298,
+      "learning_rate": 0.001,
+      "loss": 2.0283,
+      "step": 7342
+    },
+    {
+      "epoch": 0.3106438785007192,
+      "grad_norm": 2.1442577838897705,
+      "learning_rate": 0.001,
+      "loss": 2.3488,
+      "step": 7343
+    },
+    {
+      "epoch": 0.3106861832642355,
+      "grad_norm": 2.738429069519043,
+      "learning_rate": 0.001,
+      "loss": 1.9623,
+      "step": 7344
+    },
+    {
+      "epoch": 0.31072848802775194,
+      "grad_norm": 0.20081764459609985,
+      "learning_rate": 0.001,
+      "loss": 2.3359,
+      "step": 7345
+    },
+    {
+      "epoch": 0.3107707927912683,
+      "grad_norm": 0.44594720005989075,
+      "learning_rate": 0.001,
+      "loss": 2.2799,
+      "step": 7346
+    },
+    {
+      "epoch": 0.31081309755478465,
+      "grad_norm": 0.26621145009994507,
+      "learning_rate": 0.001,
+      "loss": 3.0893,
+      "step": 7347
+    },
+    {
+      "epoch": 0.31085540231830106,
+      "grad_norm": 1.0780677795410156,
+      "learning_rate": 0.001,
+      "loss": 2.2325,
+      "step": 7348
+    },
+    {
+      "epoch": 0.3108977070818174,
+      "grad_norm": 0.33734557032585144,
+      "learning_rate": 0.001,
+      "loss": 1.9325,
+      "step": 7349
+    },
+    {
+      "epoch": 0.31094001184533376,
+      "grad_norm": 0.7444893717765808,
+      "learning_rate": 0.001,
+      "loss": 2.5934,
+      "step": 7350
+    },
+    {
+      "epoch": 0.3109823166088502,
+      "grad_norm": 0.2100495994091034,
+      "learning_rate": 0.001,
+      "loss": 2.7241,
+      "step": 7351
+    },
+    {
+      "epoch": 0.3110246213723665,
+      "grad_norm": 0.2134474217891693,
+      "learning_rate": 0.001,
+      "loss": 3.2037,
+      "step": 7352
+    },
+    {
+      "epoch": 0.3110669261358829,
+      "grad_norm": 0.21119053661823273,
+      "learning_rate": 0.001,
+      "loss": 2.3662,
+      "step": 7353
+    },
+    {
+      "epoch": 0.3111092308993993,
+      "grad_norm": 2.0805516242980957,
+      "learning_rate": 0.001,
+      "loss": 2.2039,
+      "step": 7354
+    },
+    {
+      "epoch": 0.31115153566291565,
+      "grad_norm": 0.245794877409935,
+      "learning_rate": 0.001,
+      "loss": 2.1602,
+      "step": 7355
+    },
+    {
+      "epoch": 0.311193840426432,
+      "grad_norm": 0.25053372979164124,
+      "learning_rate": 0.001,
+      "loss": 2.8372,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3112361451899484,
+      "grad_norm": 0.2969318628311157,
+      "learning_rate": 0.001,
+      "loss": 2.893,
+      "step": 7357
+    },
+    {
+      "epoch": 0.31127844995346476,
+      "grad_norm": 0.20529517531394958,
+      "learning_rate": 0.001,
+      "loss": 1.6362,
+      "step": 7358
+    },
+    {
+      "epoch": 0.3113207547169811,
+      "grad_norm": 0.43395882844924927,
+      "learning_rate": 0.001,
+      "loss": 3.1913,
+      "step": 7359
+    },
+    {
+      "epoch": 0.31136305948049753,
+      "grad_norm": 0.18605674803256989,
+      "learning_rate": 0.001,
+      "loss": 2.0649,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3114053642440139,
+      "grad_norm": 16.036762237548828,
+      "learning_rate": 0.001,
+      "loss": 1.9725,
+      "step": 7361
+    },
+    {
+      "epoch": 0.31144766900753024,
+      "grad_norm": 0.7822217345237732,
+      "learning_rate": 0.001,
+      "loss": 2.0036,
+      "step": 7362
+    },
+    {
+      "epoch": 0.31148997377104665,
+      "grad_norm": 2.8043062686920166,
+      "learning_rate": 0.001,
+      "loss": 2.5385,
+      "step": 7363
+    },
+    {
+      "epoch": 0.311532278534563,
+      "grad_norm": 0.18811993300914764,
+      "learning_rate": 0.001,
+      "loss": 1.9102,
+      "step": 7364
+    },
+    {
+      "epoch": 0.31157458329807936,
+      "grad_norm": 0.9021945595741272,
+      "learning_rate": 0.001,
+      "loss": 2.5355,
+      "step": 7365
+    },
+    {
+      "epoch": 0.3116168880615957,
+      "grad_norm": 0.247837632894516,
+      "learning_rate": 0.001,
+      "loss": 1.8979,
+      "step": 7366
+    },
+    {
+      "epoch": 0.3116591928251121,
+      "grad_norm": 0.4068886637687683,
+      "learning_rate": 0.001,
+      "loss": 2.5863,
+      "step": 7367
+    },
+    {
+      "epoch": 0.3117014975886285,
+      "grad_norm": 0.3906690776348114,
+      "learning_rate": 0.001,
+      "loss": 1.8295,
+      "step": 7368
+    },
+    {
+      "epoch": 0.3117438023521448,
+      "grad_norm": 0.18249556422233582,
+      "learning_rate": 0.001,
+      "loss": 2.2368,
+      "step": 7369
+    },
+    {
+      "epoch": 0.31178610711566124,
+      "grad_norm": 0.22424672544002533,
+      "learning_rate": 0.001,
+      "loss": 1.736,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3118284118791776,
+      "grad_norm": 1.013100504875183,
+      "learning_rate": 0.001,
+      "loss": 2.3112,
+      "step": 7371
+    },
+    {
+      "epoch": 0.31187071664269395,
+      "grad_norm": 0.5906060338020325,
+      "learning_rate": 0.001,
+      "loss": 2.3438,
+      "step": 7372
+    },
+    {
+      "epoch": 0.31191302140621036,
+      "grad_norm": 0.22655218839645386,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 7373
+    },
+    {
+      "epoch": 0.3119553261697267,
+      "grad_norm": 0.8316282629966736,
+      "learning_rate": 0.001,
+      "loss": 2.0015,
+      "step": 7374
+    },
+    {
+      "epoch": 0.31199763093324306,
+      "grad_norm": 0.3435095548629761,
+      "learning_rate": 0.001,
+      "loss": 2.3837,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3120399356967595,
+      "grad_norm": 1.1091632843017578,
+      "learning_rate": 0.001,
+      "loss": 2.3188,
+      "step": 7376
+    },
+    {
+      "epoch": 0.31208224046027583,
+      "grad_norm": 0.35905349254608154,
+      "learning_rate": 0.001,
+      "loss": 2.6176,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3121245452237922,
+      "grad_norm": 0.2939832806587219,
+      "learning_rate": 0.001,
+      "loss": 2.6581,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3121668499873086,
+      "grad_norm": 0.21034960448741913,
+      "learning_rate": 0.001,
+      "loss": 1.7465,
+      "step": 7379
+    },
+    {
+      "epoch": 0.31220915475082495,
+      "grad_norm": 0.23007825016975403,
+      "learning_rate": 0.001,
+      "loss": 2.6893,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3122514595143413,
+      "grad_norm": 0.1853979229927063,
+      "learning_rate": 0.001,
+      "loss": 1.961,
+      "step": 7381
+    },
+    {
+      "epoch": 0.3122937642778577,
+      "grad_norm": 4.700438499450684,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 7382
+    },
+    {
+      "epoch": 0.31233606904137406,
+      "grad_norm": 0.210820272564888,
+      "learning_rate": 0.001,
+      "loss": 2.2501,
+      "step": 7383
+    },
+    {
+      "epoch": 0.3123783738048904,
+      "grad_norm": 0.2408655881881714,
+      "learning_rate": 0.001,
+      "loss": 2.5988,
+      "step": 7384
+    },
+    {
+      "epoch": 0.31242067856840683,
+      "grad_norm": 7.633557319641113,
+      "learning_rate": 0.001,
+      "loss": 2.0167,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3124629833319232,
+      "grad_norm": 0.8851701617240906,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 7386
+    },
+    {
+      "epoch": 0.31250528809543954,
+      "grad_norm": 0.4367370009422302,
+      "learning_rate": 0.001,
+      "loss": 2.7794,
+      "step": 7387
+    },
+    {
+      "epoch": 0.3125475928589559,
+      "grad_norm": 0.24811141192913055,
+      "learning_rate": 0.001,
+      "loss": 1.8663,
+      "step": 7388
+    },
+    {
+      "epoch": 0.3125898976224723,
+      "grad_norm": 0.24149896204471588,
+      "learning_rate": 0.001,
+      "loss": 2.0688,
+      "step": 7389
+    },
+    {
+      "epoch": 0.31263220238598866,
+      "grad_norm": 0.17393232882022858,
+      "learning_rate": 0.001,
+      "loss": 2.2648,
+      "step": 7390
+    },
+    {
+      "epoch": 0.312674507149505,
+      "grad_norm": 0.6203790903091431,
+      "learning_rate": 0.001,
+      "loss": 2.4404,
+      "step": 7391
+    },
+    {
+      "epoch": 0.3127168119130214,
+      "grad_norm": 0.21590745449066162,
+      "learning_rate": 0.001,
+      "loss": 2.0125,
+      "step": 7392
+    },
+    {
+      "epoch": 0.3127591166765378,
+      "grad_norm": 0.2051093429327011,
+      "learning_rate": 0.001,
+      "loss": 1.9774,
+      "step": 7393
+    },
+    {
+      "epoch": 0.31280142144005413,
+      "grad_norm": 0.9935820698738098,
+      "learning_rate": 0.001,
+      "loss": 1.9072,
+      "step": 7394
+    },
+    {
+      "epoch": 0.31284372620357054,
+      "grad_norm": 0.18654446303844452,
+      "learning_rate": 0.001,
+      "loss": 2.4551,
+      "step": 7395
+    },
+    {
+      "epoch": 0.3128860309670869,
+      "grad_norm": 0.19215311110019684,
+      "learning_rate": 0.001,
+      "loss": 1.9348,
+      "step": 7396
+    },
+    {
+      "epoch": 0.31292833573060325,
+      "grad_norm": 0.18548458814620972,
+      "learning_rate": 0.001,
+      "loss": 1.9123,
+      "step": 7397
+    },
+    {
+      "epoch": 0.31297064049411966,
+      "grad_norm": 0.2123180329799652,
+      "learning_rate": 0.001,
+      "loss": 2.5744,
+      "step": 7398
+    },
+    {
+      "epoch": 0.313012945257636,
+      "grad_norm": 0.20290523767471313,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 7399
+    },
+    {
+      "epoch": 0.31305525002115236,
+      "grad_norm": 0.2106354683637619,
+      "learning_rate": 0.001,
+      "loss": 2.2838,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3130975547846688,
+      "grad_norm": 0.9356090426445007,
+      "learning_rate": 0.001,
+      "loss": 2.6152,
+      "step": 7401
+    },
+    {
+      "epoch": 0.31313985954818513,
+      "grad_norm": 0.190611332654953,
+      "learning_rate": 0.001,
+      "loss": 3.1069,
+      "step": 7402
+    },
+    {
+      "epoch": 0.3131821643117015,
+      "grad_norm": 0.20556482672691345,
+      "learning_rate": 0.001,
+      "loss": 1.7953,
+      "step": 7403
+    },
+    {
+      "epoch": 0.3132244690752179,
+      "grad_norm": 0.2476384937763214,
+      "learning_rate": 0.001,
+      "loss": 1.7635,
+      "step": 7404
+    },
+    {
+      "epoch": 0.31326677383873425,
+      "grad_norm": 0.2387237846851349,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3133090786022506,
+      "grad_norm": 1.1636114120483398,
+      "learning_rate": 0.001,
+      "loss": 1.8156,
+      "step": 7406
+    },
+    {
+      "epoch": 0.313351383365767,
+      "grad_norm": 0.36299633979797363,
+      "learning_rate": 0.001,
+      "loss": 1.984,
+      "step": 7407
+    },
+    {
+      "epoch": 0.31339368812928337,
+      "grad_norm": 0.20854444801807404,
+      "learning_rate": 0.001,
+      "loss": 2.8094,
+      "step": 7408
+    },
+    {
+      "epoch": 0.3134359928927997,
+      "grad_norm": 0.20307183265686035,
+      "learning_rate": 0.001,
+      "loss": 2.0501,
+      "step": 7409
+    },
+    {
+      "epoch": 0.3134782976563161,
+      "grad_norm": 0.37564030289649963,
+      "learning_rate": 0.001,
+      "loss": 1.8919,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3135206024198325,
+      "grad_norm": 0.4626857340335846,
+      "learning_rate": 0.001,
+      "loss": 3.2498,
+      "step": 7411
+    },
+    {
+      "epoch": 0.31356290718334884,
+      "grad_norm": 1.1579762697219849,
+      "learning_rate": 0.001,
+      "loss": 2.1231,
+      "step": 7412
+    },
+    {
+      "epoch": 0.3136052119468652,
+      "grad_norm": 0.17268231511116028,
+      "learning_rate": 0.001,
+      "loss": 2.8643,
+      "step": 7413
+    },
+    {
+      "epoch": 0.3136475167103816,
+      "grad_norm": 0.2048521339893341,
+      "learning_rate": 0.001,
+      "loss": 3.5636,
+      "step": 7414
+    },
+    {
+      "epoch": 0.31368982147389796,
+      "grad_norm": 0.3608675003051758,
+      "learning_rate": 0.001,
+      "loss": 3.2036,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3137321262374143,
+      "grad_norm": 0.24735359847545624,
+      "learning_rate": 0.001,
+      "loss": 2.1055,
+      "step": 7416
+    },
+    {
+      "epoch": 0.3137744310009307,
+      "grad_norm": 0.6630021333694458,
+      "learning_rate": 0.001,
+      "loss": 3.1071,
+      "step": 7417
+    },
+    {
+      "epoch": 0.3138167357644471,
+      "grad_norm": 0.19361117482185364,
+      "learning_rate": 0.001,
+      "loss": 2.294,
+      "step": 7418
+    },
+    {
+      "epoch": 0.31385904052796343,
+      "grad_norm": 0.26215800642967224,
+      "learning_rate": 0.001,
+      "loss": 2.9492,
+      "step": 7419
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 0.20346727967262268,
+      "learning_rate": 0.001,
+      "loss": 2.087,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3139436500549962,
+      "grad_norm": 0.3235960304737091,
+      "learning_rate": 0.001,
+      "loss": 2.0558,
+      "step": 7421
+    },
+    {
+      "epoch": 0.31398595481851255,
+      "grad_norm": 0.2161642611026764,
+      "learning_rate": 0.001,
+      "loss": 2.1842,
+      "step": 7422
+    },
+    {
+      "epoch": 0.31402825958202896,
+      "grad_norm": 0.19181302189826965,
+      "learning_rate": 0.001,
+      "loss": 1.9424,
+      "step": 7423
+    },
+    {
+      "epoch": 0.3140705643455453,
+      "grad_norm": 19.93857765197754,
+      "learning_rate": 0.001,
+      "loss": 2.7717,
+      "step": 7424
+    },
+    {
+      "epoch": 0.31411286910906167,
+      "grad_norm": 0.2410607635974884,
+      "learning_rate": 0.001,
+      "loss": 2.7044,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3141551738725781,
+      "grad_norm": 1.307002067565918,
+      "learning_rate": 0.001,
+      "loss": 2.1455,
+      "step": 7426
+    },
+    {
+      "epoch": 0.31419747863609443,
+      "grad_norm": 0.4844810962677002,
+      "learning_rate": 0.001,
+      "loss": 2.3331,
+      "step": 7427
+    },
+    {
+      "epoch": 0.3142397833996108,
+      "grad_norm": 0.28429657220840454,
+      "learning_rate": 0.001,
+      "loss": 2.89,
+      "step": 7428
+    },
+    {
+      "epoch": 0.3142820881631272,
+      "grad_norm": 0.8632631897926331,
+      "learning_rate": 0.001,
+      "loss": 1.8695,
+      "step": 7429
+    },
+    {
+      "epoch": 0.31432439292664355,
+      "grad_norm": 0.3193078637123108,
+      "learning_rate": 0.001,
+      "loss": 3.2732,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3143666976901599,
+      "grad_norm": 0.20497079193592072,
+      "learning_rate": 0.001,
+      "loss": 2.1384,
+      "step": 7431
+    },
+    {
+      "epoch": 0.31440900245367626,
+      "grad_norm": 0.2516421973705292,
+      "learning_rate": 0.001,
+      "loss": 2.2816,
+      "step": 7432
+    },
+    {
+      "epoch": 0.31445130721719267,
+      "grad_norm": 1.0141093730926514,
+      "learning_rate": 0.001,
+      "loss": 2.6545,
+      "step": 7433
+    },
+    {
+      "epoch": 0.314493611980709,
+      "grad_norm": 0.18999618291854858,
+      "learning_rate": 0.001,
+      "loss": 2.0399,
+      "step": 7434
+    },
+    {
+      "epoch": 0.3145359167442254,
+      "grad_norm": 0.456870436668396,
+      "learning_rate": 0.001,
+      "loss": 2.8804,
+      "step": 7435
+    },
+    {
+      "epoch": 0.3145782215077418,
+      "grad_norm": 0.46013543009757996,
+      "learning_rate": 0.001,
+      "loss": 2.5783,
+      "step": 7436
+    },
+    {
+      "epoch": 0.31462052627125814,
+      "grad_norm": 0.20753584802150726,
+      "learning_rate": 0.001,
+      "loss": 2.4055,
+      "step": 7437
+    },
+    {
+      "epoch": 0.3146628310347745,
+      "grad_norm": 0.23144669830799103,
+      "learning_rate": 0.001,
+      "loss": 2.9632,
+      "step": 7438
+    },
+    {
+      "epoch": 0.3147051357982909,
+      "grad_norm": 0.3479333817958832,
+      "learning_rate": 0.001,
+      "loss": 1.792,
+      "step": 7439
+    },
+    {
+      "epoch": 0.31474744056180726,
+      "grad_norm": 4.283021926879883,
+      "learning_rate": 0.001,
+      "loss": 2.1915,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3147897453253236,
+      "grad_norm": 7.6823649406433105,
+      "learning_rate": 0.001,
+      "loss": 2.5411,
+      "step": 7441
+    },
+    {
+      "epoch": 0.31483205008884,
+      "grad_norm": 0.6836051344871521,
+      "learning_rate": 0.001,
+      "loss": 2.3475,
+      "step": 7442
+    },
+    {
+      "epoch": 0.3148743548523564,
+      "grad_norm": 1.2467254400253296,
+      "learning_rate": 0.001,
+      "loss": 2.448,
+      "step": 7443
+    },
+    {
+      "epoch": 0.31491665961587273,
+      "grad_norm": 0.4439103603363037,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 7444
+    },
+    {
+      "epoch": 0.31495896437938914,
+      "grad_norm": 0.23518876731395721,
+      "learning_rate": 0.001,
+      "loss": 2.3276,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3150012691429055,
+      "grad_norm": 0.5438497066497803,
+      "learning_rate": 0.001,
+      "loss": 1.9945,
+      "step": 7446
+    },
+    {
+      "epoch": 0.31504357390642185,
+      "grad_norm": 1.6677125692367554,
+      "learning_rate": 0.001,
+      "loss": 1.5225,
+      "step": 7447
+    },
+    {
+      "epoch": 0.31508587866993826,
+      "grad_norm": 0.20245139300823212,
+      "learning_rate": 0.001,
+      "loss": 1.973,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3151281834334546,
+      "grad_norm": 0.24195541441440582,
+      "learning_rate": 0.001,
+      "loss": 1.9298,
+      "step": 7449
+    },
+    {
+      "epoch": 0.31517048819697097,
+      "grad_norm": 0.21246738731861115,
+      "learning_rate": 0.001,
+      "loss": 2.0905,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3152127929604874,
+      "grad_norm": 0.2219218611717224,
+      "learning_rate": 0.001,
+      "loss": 1.9244,
+      "step": 7451
+    },
+    {
+      "epoch": 0.31525509772400373,
+      "grad_norm": 0.20514735579490662,
+      "learning_rate": 0.001,
+      "loss": 2.4173,
+      "step": 7452
+    },
+    {
+      "epoch": 0.3152974024875201,
+      "grad_norm": 0.1975151151418686,
+      "learning_rate": 0.001,
+      "loss": 2.7215,
+      "step": 7453
+    },
+    {
+      "epoch": 0.3153397072510365,
+      "grad_norm": 0.2547701299190521,
+      "learning_rate": 0.001,
+      "loss": 2.3066,
+      "step": 7454
+    },
+    {
+      "epoch": 0.31538201201455285,
+      "grad_norm": 0.2303609549999237,
+      "learning_rate": 0.001,
+      "loss": 2.6705,
+      "step": 7455
+    },
+    {
+      "epoch": 0.3154243167780692,
+      "grad_norm": 0.2129909247159958,
+      "learning_rate": 0.001,
+      "loss": 1.9109,
+      "step": 7456
+    },
+    {
+      "epoch": 0.31546662154158556,
+      "grad_norm": 0.2366720288991928,
+      "learning_rate": 0.001,
+      "loss": 2.8221,
+      "step": 7457
+    },
+    {
+      "epoch": 0.31550892630510197,
+      "grad_norm": 0.23550964891910553,
+      "learning_rate": 0.001,
+      "loss": 3.5912,
+      "step": 7458
+    },
+    {
+      "epoch": 0.3155512310686183,
+      "grad_norm": 0.21231484413146973,
+      "learning_rate": 0.001,
+      "loss": 1.7912,
+      "step": 7459
+    },
+    {
+      "epoch": 0.3155935358321347,
+      "grad_norm": 0.22212854027748108,
+      "learning_rate": 0.001,
+      "loss": 2.5323,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3156358405956511,
+      "grad_norm": 1.3847850561141968,
+      "learning_rate": 0.001,
+      "loss": 2.2599,
+      "step": 7461
+    },
+    {
+      "epoch": 0.31567814535916744,
+      "grad_norm": 0.6198214292526245,
+      "learning_rate": 0.001,
+      "loss": 2.3198,
+      "step": 7462
+    },
+    {
+      "epoch": 0.3157204501226838,
+      "grad_norm": 0.20347964763641357,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 7463
+    },
+    {
+      "epoch": 0.3157627548862002,
+      "grad_norm": 0.3482247292995453,
+      "learning_rate": 0.001,
+      "loss": 1.8097,
+      "step": 7464
+    },
+    {
+      "epoch": 0.31580505964971656,
+      "grad_norm": 0.22890900075435638,
+      "learning_rate": 0.001,
+      "loss": 2.2547,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3158473644132329,
+      "grad_norm": 0.21213634312152863,
+      "learning_rate": 0.001,
+      "loss": 2.1473,
+      "step": 7466
+    },
+    {
+      "epoch": 0.3158896691767493,
+      "grad_norm": 0.2835163176059723,
+      "learning_rate": 0.001,
+      "loss": 2.5594,
+      "step": 7467
+    },
+    {
+      "epoch": 0.3159319739402657,
+      "grad_norm": 0.22125250101089478,
+      "learning_rate": 0.001,
+      "loss": 2.9808,
+      "step": 7468
+    },
+    {
+      "epoch": 0.31597427870378203,
+      "grad_norm": 2.557072401046753,
+      "learning_rate": 0.001,
+      "loss": 3.2332,
+      "step": 7469
+    },
+    {
+      "epoch": 0.31601658346729844,
+      "grad_norm": 4.832421779632568,
+      "learning_rate": 0.001,
+      "loss": 2.2446,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3160588882308148,
+      "grad_norm": 0.19227334856987,
+      "learning_rate": 0.001,
+      "loss": 2.5917,
+      "step": 7471
+    },
+    {
+      "epoch": 0.31610119299433115,
+      "grad_norm": 0.23523429036140442,
+      "learning_rate": 0.001,
+      "loss": 2.5136,
+      "step": 7472
+    },
+    {
+      "epoch": 0.31614349775784756,
+      "grad_norm": 0.22452674806118011,
+      "learning_rate": 0.001,
+      "loss": 2.0461,
+      "step": 7473
+    },
+    {
+      "epoch": 0.3161858025213639,
+      "grad_norm": 0.43354332447052,
+      "learning_rate": 0.001,
+      "loss": 3.0801,
+      "step": 7474
+    },
+    {
+      "epoch": 0.31622810728488027,
+      "grad_norm": 0.21420817077159882,
+      "learning_rate": 0.001,
+      "loss": 1.6362,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3162704120483967,
+      "grad_norm": 5.511373043060303,
+      "learning_rate": 0.001,
+      "loss": 2.1619,
+      "step": 7476
+    },
+    {
+      "epoch": 0.31631271681191303,
+      "grad_norm": 0.647841215133667,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 7477
+    },
+    {
+      "epoch": 0.3163550215754294,
+      "grad_norm": 0.4985596537590027,
+      "learning_rate": 0.001,
+      "loss": 2.6925,
+      "step": 7478
+    },
+    {
+      "epoch": 0.31639732633894574,
+      "grad_norm": 0.6579802632331848,
+      "learning_rate": 0.001,
+      "loss": 2.7019,
+      "step": 7479
+    },
+    {
+      "epoch": 0.31643963110246215,
+      "grad_norm": 2.1017627716064453,
+      "learning_rate": 0.001,
+      "loss": 2.4639,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3164819358659785,
+      "grad_norm": 0.5756150484085083,
+      "learning_rate": 0.001,
+      "loss": 1.7458,
+      "step": 7481
+    },
+    {
+      "epoch": 0.31652424062949486,
+      "grad_norm": 0.24982336163520813,
+      "learning_rate": 0.001,
+      "loss": 3.0768,
+      "step": 7482
+    },
+    {
+      "epoch": 0.31656654539301127,
+      "grad_norm": 0.5327016711235046,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 7483
+    },
+    {
+      "epoch": 0.3166088501565276,
+      "grad_norm": 7.926521301269531,
+      "learning_rate": 0.001,
+      "loss": 2.0401,
+      "step": 7484
+    },
+    {
+      "epoch": 0.316651154920044,
+      "grad_norm": 0.2336919903755188,
+      "learning_rate": 0.001,
+      "loss": 2.5418,
+      "step": 7485
+    },
+    {
+      "epoch": 0.3166934596835604,
+      "grad_norm": 1.4902650117874146,
+      "learning_rate": 0.001,
+      "loss": 2.053,
+      "step": 7486
+    },
+    {
+      "epoch": 0.31673576444707674,
+      "grad_norm": 0.3811624348163605,
+      "learning_rate": 0.001,
+      "loss": 1.8232,
+      "step": 7487
+    },
+    {
+      "epoch": 0.3167780692105931,
+      "grad_norm": 0.37704822421073914,
+      "learning_rate": 0.001,
+      "loss": 2.9608,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3168203739741095,
+      "grad_norm": 0.35828453302383423,
+      "learning_rate": 0.001,
+      "loss": 1.9027,
+      "step": 7489
+    },
+    {
+      "epoch": 0.31686267873762586,
+      "grad_norm": 0.22911661863327026,
+      "learning_rate": 0.001,
+      "loss": 2.7649,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3169049835011422,
+      "grad_norm": 0.2562866806983948,
+      "learning_rate": 0.001,
+      "loss": 2.4122,
+      "step": 7491
+    },
+    {
+      "epoch": 0.3169472882646586,
+      "grad_norm": 0.24646058678627014,
+      "learning_rate": 0.001,
+      "loss": 2.0809,
+      "step": 7492
+    },
+    {
+      "epoch": 0.316989593028175,
+      "grad_norm": 1.1058540344238281,
+      "learning_rate": 0.001,
+      "loss": 2.0834,
+      "step": 7493
+    },
+    {
+      "epoch": 0.31703189779169133,
+      "grad_norm": 0.20626600086688995,
+      "learning_rate": 0.001,
+      "loss": 2.2533,
+      "step": 7494
+    },
+    {
+      "epoch": 0.31707420255520774,
+      "grad_norm": 0.32516318559646606,
+      "learning_rate": 0.001,
+      "loss": 3.1424,
+      "step": 7495
+    },
+    {
+      "epoch": 0.3171165073187241,
+      "grad_norm": 0.2439742237329483,
+      "learning_rate": 0.001,
+      "loss": 2.2094,
+      "step": 7496
+    },
+    {
+      "epoch": 0.31715881208224045,
+      "grad_norm": 0.34792065620422363,
+      "learning_rate": 0.001,
+      "loss": 1.9122,
+      "step": 7497
+    },
+    {
+      "epoch": 0.31720111684575686,
+      "grad_norm": 0.23733051121234894,
+      "learning_rate": 0.001,
+      "loss": 1.936,
+      "step": 7498
+    },
+    {
+      "epoch": 0.3172434216092732,
+      "grad_norm": 0.24519434571266174,
+      "learning_rate": 0.001,
+      "loss": 2.036,
+      "step": 7499
+    },
+    {
+      "epoch": 0.31728572637278957,
+      "grad_norm": 0.4781532883644104,
+      "learning_rate": 0.001,
+      "loss": 3.1208,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3173280311363059,
+      "grad_norm": 0.19272254407405853,
+      "learning_rate": 0.001,
+      "loss": 2.552,
+      "step": 7501
+    },
+    {
+      "epoch": 0.31737033589982233,
+      "grad_norm": 0.32179877161979675,
+      "learning_rate": 0.001,
+      "loss": 2.8888,
+      "step": 7502
+    },
+    {
+      "epoch": 0.3174126406633387,
+      "grad_norm": 0.7427499890327454,
+      "learning_rate": 0.001,
+      "loss": 2.615,
+      "step": 7503
+    },
+    {
+      "epoch": 0.31745494542685504,
+      "grad_norm": 0.31697311997413635,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 7504
+    },
+    {
+      "epoch": 0.31749725019037145,
+      "grad_norm": 0.16490371525287628,
+      "learning_rate": 0.001,
+      "loss": 2.6595,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3175395549538878,
+      "grad_norm": 0.2627149224281311,
+      "learning_rate": 0.001,
+      "loss": 4.131,
+      "step": 7506
+    },
+    {
+      "epoch": 0.31758185971740416,
+      "grad_norm": 0.8162776827812195,
+      "learning_rate": 0.001,
+      "loss": 1.8456,
+      "step": 7507
+    },
+    {
+      "epoch": 0.31762416448092057,
+      "grad_norm": 0.3852151036262512,
+      "learning_rate": 0.001,
+      "loss": 3.4449,
+      "step": 7508
+    },
+    {
+      "epoch": 0.3176664692444369,
+      "grad_norm": 1.2743273973464966,
+      "learning_rate": 0.001,
+      "loss": 2.073,
+      "step": 7509
+    },
+    {
+      "epoch": 0.3177087740079533,
+      "grad_norm": 0.23382671177387238,
+      "learning_rate": 0.001,
+      "loss": 1.7473,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3177510787714697,
+      "grad_norm": 0.25078028440475464,
+      "learning_rate": 0.001,
+      "loss": 2.8671,
+      "step": 7511
+    },
+    {
+      "epoch": 0.31779338353498604,
+      "grad_norm": 0.18183058500289917,
+      "learning_rate": 0.001,
+      "loss": 2.3517,
+      "step": 7512
+    },
+    {
+      "epoch": 0.3178356882985024,
+      "grad_norm": 0.20492641627788544,
+      "learning_rate": 0.001,
+      "loss": 1.6892,
+      "step": 7513
+    },
+    {
+      "epoch": 0.3178779930620188,
+      "grad_norm": 0.19157011806964874,
+      "learning_rate": 0.001,
+      "loss": 1.9524,
+      "step": 7514
+    },
+    {
+      "epoch": 0.31792029782553516,
+      "grad_norm": 0.20687264204025269,
+      "learning_rate": 0.001,
+      "loss": 2.091,
+      "step": 7515
+    },
+    {
+      "epoch": 0.3179626025890515,
+      "grad_norm": 0.3631476163864136,
+      "learning_rate": 0.001,
+      "loss": 2.5191,
+      "step": 7516
+    },
+    {
+      "epoch": 0.3180049073525679,
+      "grad_norm": 2.8174641132354736,
+      "learning_rate": 0.001,
+      "loss": 2.2289,
+      "step": 7517
+    },
+    {
+      "epoch": 0.3180472121160843,
+      "grad_norm": 0.21862949430942535,
+      "learning_rate": 0.001,
+      "loss": 2.8965,
+      "step": 7518
+    },
+    {
+      "epoch": 0.31808951687960063,
+      "grad_norm": 1.085720419883728,
+      "learning_rate": 0.001,
+      "loss": 2.622,
+      "step": 7519
+    },
+    {
+      "epoch": 0.31813182164311704,
+      "grad_norm": 0.18160761892795563,
+      "learning_rate": 0.001,
+      "loss": 2.2491,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3181741264066334,
+      "grad_norm": 0.8975410461425781,
+      "learning_rate": 0.001,
+      "loss": 2.8174,
+      "step": 7521
+    },
+    {
+      "epoch": 0.31821643117014975,
+      "grad_norm": 1.3470760583877563,
+      "learning_rate": 0.001,
+      "loss": 1.7925,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3182587359336661,
+      "grad_norm": 1.8409106731414795,
+      "learning_rate": 0.001,
+      "loss": 2.8657,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3183010406971825,
+      "grad_norm": 0.17481787502765656,
+      "learning_rate": 0.001,
+      "loss": 2.4448,
+      "step": 7524
+    },
+    {
+      "epoch": 0.31834334546069887,
+      "grad_norm": 0.2951485514640808,
+      "learning_rate": 0.001,
+      "loss": 2.678,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3183856502242152,
+      "grad_norm": 0.22341516613960266,
+      "learning_rate": 0.001,
+      "loss": 2.1617,
+      "step": 7526
+    },
+    {
+      "epoch": 0.31842795498773163,
+      "grad_norm": 0.2601909935474396,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 7527
+    },
+    {
+      "epoch": 0.318470259751248,
+      "grad_norm": 0.2952847480773926,
+      "learning_rate": 0.001,
+      "loss": 2.1086,
+      "step": 7528
+    },
+    {
+      "epoch": 0.31851256451476434,
+      "grad_norm": 0.4443856477737427,
+      "learning_rate": 0.001,
+      "loss": 3.2807,
+      "step": 7529
+    },
+    {
+      "epoch": 0.31855486927828075,
+      "grad_norm": 0.235245943069458,
+      "learning_rate": 0.001,
+      "loss": 2.9082,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3185971740417971,
+      "grad_norm": 1.0847169160842896,
+      "learning_rate": 0.001,
+      "loss": 2.0038,
+      "step": 7531
+    },
+    {
+      "epoch": 0.31863947880531346,
+      "grad_norm": 0.245633065700531,
+      "learning_rate": 0.001,
+      "loss": 2.6333,
+      "step": 7532
+    },
+    {
+      "epoch": 0.31868178356882987,
+      "grad_norm": 0.2640596330165863,
+      "learning_rate": 0.001,
+      "loss": 3.2857,
+      "step": 7533
+    },
+    {
+      "epoch": 0.3187240883323462,
+      "grad_norm": 0.5016381740570068,
+      "learning_rate": 0.001,
+      "loss": 3.0914,
+      "step": 7534
+    },
+    {
+      "epoch": 0.3187663930958626,
+      "grad_norm": 0.19767212867736816,
+      "learning_rate": 0.001,
+      "loss": 1.7764,
+      "step": 7535
+    },
+    {
+      "epoch": 0.318808697859379,
+      "grad_norm": 0.2402283400297165,
+      "learning_rate": 0.001,
+      "loss": 2.7833,
+      "step": 7536
+    },
+    {
+      "epoch": 0.31885100262289534,
+      "grad_norm": 4.479025363922119,
+      "learning_rate": 0.001,
+      "loss": 2.2062,
+      "step": 7537
+    },
+    {
+      "epoch": 0.3188933073864117,
+      "grad_norm": 0.584766685962677,
+      "learning_rate": 0.001,
+      "loss": 1.7729,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3189356121499281,
+      "grad_norm": 0.5840973258018494,
+      "learning_rate": 0.001,
+      "loss": 1.8462,
+      "step": 7539
+    },
+    {
+      "epoch": 0.31897791691344446,
+      "grad_norm": 0.2561521828174591,
+      "learning_rate": 0.001,
+      "loss": 2.3039,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3190202216769608,
+      "grad_norm": 4.742803573608398,
+      "learning_rate": 0.001,
+      "loss": 2.0092,
+      "step": 7541
+    },
+    {
+      "epoch": 0.3190625264404772,
+      "grad_norm": 0.24484741687774658,
+      "learning_rate": 0.001,
+      "loss": 2.2363,
+      "step": 7542
+    },
+    {
+      "epoch": 0.3191048312039936,
+      "grad_norm": 0.70278000831604,
+      "learning_rate": 0.001,
+      "loss": 3.5423,
+      "step": 7543
+    },
+    {
+      "epoch": 0.31914713596750993,
+      "grad_norm": 0.22903594374656677,
+      "learning_rate": 0.001,
+      "loss": 1.8414,
+      "step": 7544
+    },
+    {
+      "epoch": 0.3191894407310263,
+      "grad_norm": 0.5062339901924133,
+      "learning_rate": 0.001,
+      "loss": 2.1526,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3192317454945427,
+      "grad_norm": 0.2158339023590088,
+      "learning_rate": 0.001,
+      "loss": 1.9886,
+      "step": 7546
+    },
+    {
+      "epoch": 0.31927405025805905,
+      "grad_norm": 1.0523641109466553,
+      "learning_rate": 0.001,
+      "loss": 2.6195,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3193163550215754,
+      "grad_norm": 0.2197253257036209,
+      "learning_rate": 0.001,
+      "loss": 1.5022,
+      "step": 7548
+    },
+    {
+      "epoch": 0.3193586597850918,
+      "grad_norm": 2.6306278705596924,
+      "learning_rate": 0.001,
+      "loss": 2.143,
+      "step": 7549
+    },
+    {
+      "epoch": 0.31940096454860817,
+      "grad_norm": 0.7273126840591431,
+      "learning_rate": 0.001,
+      "loss": 2.4476,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3194432693121245,
+      "grad_norm": 0.29238003492355347,
+      "learning_rate": 0.001,
+      "loss": 1.9445,
+      "step": 7551
+    },
+    {
+      "epoch": 0.31948557407564093,
+      "grad_norm": 0.6033421158790588,
+      "learning_rate": 0.001,
+      "loss": 2.5735,
+      "step": 7552
+    },
+    {
+      "epoch": 0.3195278788391573,
+      "grad_norm": 0.723403811454773,
+      "learning_rate": 0.001,
+      "loss": 2.9264,
+      "step": 7553
+    },
+    {
+      "epoch": 0.31957018360267364,
+      "grad_norm": 0.3378523886203766,
+      "learning_rate": 0.001,
+      "loss": 2.0737,
+      "step": 7554
+    },
+    {
+      "epoch": 0.31961248836619005,
+      "grad_norm": 0.2882983684539795,
+      "learning_rate": 0.001,
+      "loss": 3.3238,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3196547931297064,
+      "grad_norm": 1.7969506978988647,
+      "learning_rate": 0.001,
+      "loss": 2.2955,
+      "step": 7556
+    },
+    {
+      "epoch": 0.31969709789322276,
+      "grad_norm": 1.0519235134124756,
+      "learning_rate": 0.001,
+      "loss": 2.9223,
+      "step": 7557
+    },
+    {
+      "epoch": 0.31973940265673917,
+      "grad_norm": 0.26675179600715637,
+      "learning_rate": 0.001,
+      "loss": 3.1042,
+      "step": 7558
+    },
+    {
+      "epoch": 0.3197817074202555,
+      "grad_norm": 0.210018128156662,
+      "learning_rate": 0.001,
+      "loss": 2.4462,
+      "step": 7559
+    },
+    {
+      "epoch": 0.3198240121837719,
+      "grad_norm": 0.24346289038658142,
+      "learning_rate": 0.001,
+      "loss": 2.5506,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3198663169472883,
+      "grad_norm": 0.2894313633441925,
+      "learning_rate": 0.001,
+      "loss": 2.6206,
+      "step": 7561
+    },
+    {
+      "epoch": 0.31990862171080464,
+      "grad_norm": 0.23805563151836395,
+      "learning_rate": 0.001,
+      "loss": 2.1749,
+      "step": 7562
+    },
+    {
+      "epoch": 0.319950926474321,
+      "grad_norm": 0.2397831827402115,
+      "learning_rate": 0.001,
+      "loss": 2.7243,
+      "step": 7563
+    },
+    {
+      "epoch": 0.3199932312378374,
+      "grad_norm": 0.23759059607982635,
+      "learning_rate": 0.001,
+      "loss": 1.9441,
+      "step": 7564
+    },
+    {
+      "epoch": 0.32003553600135376,
+      "grad_norm": 1.2827054262161255,
+      "learning_rate": 0.001,
+      "loss": 2.5177,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3200778407648701,
+      "grad_norm": 0.3205539882183075,
+      "learning_rate": 0.001,
+      "loss": 3.2377,
+      "step": 7566
+    },
+    {
+      "epoch": 0.3201201455283865,
+      "grad_norm": 0.3002196252346039,
+      "learning_rate": 0.001,
+      "loss": 2.2226,
+      "step": 7567
+    },
+    {
+      "epoch": 0.3201624502919029,
+      "grad_norm": 0.35749757289886475,
+      "learning_rate": 0.001,
+      "loss": 2.0588,
+      "step": 7568
+    },
+    {
+      "epoch": 0.32020475505541923,
+      "grad_norm": 0.4479646384716034,
+      "learning_rate": 0.001,
+      "loss": 2.8293,
+      "step": 7569
+    },
+    {
+      "epoch": 0.3202470598189356,
+      "grad_norm": 1.4561418294906616,
+      "learning_rate": 0.001,
+      "loss": 1.874,
+      "step": 7570
+    },
+    {
+      "epoch": 0.320289364582452,
+      "grad_norm": 0.25509679317474365,
+      "learning_rate": 0.001,
+      "loss": 1.9602,
+      "step": 7571
+    },
+    {
+      "epoch": 0.32033166934596835,
+      "grad_norm": 0.33940833806991577,
+      "learning_rate": 0.001,
+      "loss": 3.8938,
+      "step": 7572
+    },
+    {
+      "epoch": 0.3203739741094847,
+      "grad_norm": 0.23563840985298157,
+      "learning_rate": 0.001,
+      "loss": 1.9939,
+      "step": 7573
+    },
+    {
+      "epoch": 0.3204162788730011,
+      "grad_norm": 0.5354148745536804,
+      "learning_rate": 0.001,
+      "loss": 2.3731,
+      "step": 7574
+    },
+    {
+      "epoch": 0.32045858363651747,
+      "grad_norm": 0.6357294917106628,
+      "learning_rate": 0.001,
+      "loss": 2.4846,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3205008884000338,
+      "grad_norm": 0.24302570521831512,
+      "learning_rate": 0.001,
+      "loss": 2.7975,
+      "step": 7576
+    },
+    {
+      "epoch": 0.32054319316355023,
+      "grad_norm": 0.46470755338668823,
+      "learning_rate": 0.001,
+      "loss": 2.1249,
+      "step": 7577
+    },
+    {
+      "epoch": 0.3205854979270666,
+      "grad_norm": 0.24483340978622437,
+      "learning_rate": 0.001,
+      "loss": 3.0459,
+      "step": 7578
+    },
+    {
+      "epoch": 0.32062780269058294,
+      "grad_norm": 0.22136029601097107,
+      "learning_rate": 0.001,
+      "loss": 1.8347,
+      "step": 7579
+    },
+    {
+      "epoch": 0.32067010745409935,
+      "grad_norm": 7.803838729858398,
+      "learning_rate": 0.001,
+      "loss": 1.7036,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3207124122176157,
+      "grad_norm": 0.2764716148376465,
+      "learning_rate": 0.001,
+      "loss": 2.6748,
+      "step": 7581
+    },
+    {
+      "epoch": 0.32075471698113206,
+      "grad_norm": 0.44170138239860535,
+      "learning_rate": 0.001,
+      "loss": 2.1851,
+      "step": 7582
+    },
+    {
+      "epoch": 0.32079702174464847,
+      "grad_norm": 0.6049759387969971,
+      "learning_rate": 0.001,
+      "loss": 2.012,
+      "step": 7583
+    },
+    {
+      "epoch": 0.3208393265081648,
+      "grad_norm": 0.18830320239067078,
+      "learning_rate": 0.001,
+      "loss": 2.1773,
+      "step": 7584
+    },
+    {
+      "epoch": 0.3208816312716812,
+      "grad_norm": 0.3015429377555847,
+      "learning_rate": 0.001,
+      "loss": 2.196,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3209239360351976,
+      "grad_norm": 0.20359709858894348,
+      "learning_rate": 0.001,
+      "loss": 2.4333,
+      "step": 7586
+    },
+    {
+      "epoch": 0.32096624079871394,
+      "grad_norm": 0.4667384922504425,
+      "learning_rate": 0.001,
+      "loss": 2.0174,
+      "step": 7587
+    },
+    {
+      "epoch": 0.3210085455622303,
+      "grad_norm": 1.0664136409759521,
+      "learning_rate": 0.001,
+      "loss": 2.0681,
+      "step": 7588
+    },
+    {
+      "epoch": 0.3210508503257467,
+      "grad_norm": 0.37264055013656616,
+      "learning_rate": 0.001,
+      "loss": 2.3549,
+      "step": 7589
+    },
+    {
+      "epoch": 0.32109315508926306,
+      "grad_norm": 0.2381688803434372,
+      "learning_rate": 0.001,
+      "loss": 2.0513,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3211354598527794,
+      "grad_norm": 1.4251641035079956,
+      "learning_rate": 0.001,
+      "loss": 2.5208,
+      "step": 7591
+    },
+    {
+      "epoch": 0.32117776461629577,
+      "grad_norm": 0.6149337291717529,
+      "learning_rate": 0.001,
+      "loss": 2.0792,
+      "step": 7592
+    },
+    {
+      "epoch": 0.3212200693798122,
+      "grad_norm": 0.18450385332107544,
+      "learning_rate": 0.001,
+      "loss": 3.026,
+      "step": 7593
+    },
+    {
+      "epoch": 0.32126237414332853,
+      "grad_norm": 0.26451027393341064,
+      "learning_rate": 0.001,
+      "loss": 2.5821,
+      "step": 7594
+    },
+    {
+      "epoch": 0.3213046789068449,
+      "grad_norm": 0.2782851755619049,
+      "learning_rate": 0.001,
+      "loss": 2.0855,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3213469836703613,
+      "grad_norm": 0.20386971533298492,
+      "learning_rate": 0.001,
+      "loss": 1.6014,
+      "step": 7596
+    },
+    {
+      "epoch": 0.32138928843387765,
+      "grad_norm": 1.3683080673217773,
+      "learning_rate": 0.001,
+      "loss": 3.028,
+      "step": 7597
+    },
+    {
+      "epoch": 0.321431593197394,
+      "grad_norm": 1.1451998949050903,
+      "learning_rate": 0.001,
+      "loss": 2.1981,
+      "step": 7598
+    },
+    {
+      "epoch": 0.3214738979609104,
+      "grad_norm": 0.43921521306037903,
+      "learning_rate": 0.001,
+      "loss": 2.2234,
+      "step": 7599
+    },
+    {
+      "epoch": 0.32151620272442677,
+      "grad_norm": 0.2476237416267395,
+      "learning_rate": 0.001,
+      "loss": 3.1997,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3215585074879431,
+      "grad_norm": 0.21488797664642334,
+      "learning_rate": 0.001,
+      "loss": 2.2913,
+      "step": 7601
+    },
+    {
+      "epoch": 0.32160081225145953,
+      "grad_norm": 0.2706213891506195,
+      "learning_rate": 0.001,
+      "loss": 2.4037,
+      "step": 7602
+    },
+    {
+      "epoch": 0.3216431170149759,
+      "grad_norm": 0.23860187828540802,
+      "learning_rate": 0.001,
+      "loss": 2.116,
+      "step": 7603
+    },
+    {
+      "epoch": 0.32168542177849224,
+      "grad_norm": 0.23283398151397705,
+      "learning_rate": 0.001,
+      "loss": 2.6682,
+      "step": 7604
+    },
+    {
+      "epoch": 0.32172772654200865,
+      "grad_norm": 0.4454127550125122,
+      "learning_rate": 0.001,
+      "loss": 2.2326,
+      "step": 7605
+    },
+    {
+      "epoch": 0.321770031305525,
+      "grad_norm": 0.2188045084476471,
+      "learning_rate": 0.001,
+      "loss": 2.3459,
+      "step": 7606
+    },
+    {
+      "epoch": 0.32181233606904136,
+      "grad_norm": 0.2107754945755005,
+      "learning_rate": 0.001,
+      "loss": 1.8558,
+      "step": 7607
+    },
+    {
+      "epoch": 0.32185464083255777,
+      "grad_norm": 0.24413500726222992,
+      "learning_rate": 0.001,
+      "loss": 2.6748,
+      "step": 7608
+    },
+    {
+      "epoch": 0.3218969455960741,
+      "grad_norm": 0.1848825365304947,
+      "learning_rate": 0.001,
+      "loss": 1.7373,
+      "step": 7609
+    },
+    {
+      "epoch": 0.3219392503595905,
+      "grad_norm": 0.4209091365337372,
+      "learning_rate": 0.001,
+      "loss": 2.9954,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3219815551231069,
+      "grad_norm": 1.6776453256607056,
+      "learning_rate": 0.001,
+      "loss": 3.9771,
+      "step": 7611
+    },
+    {
+      "epoch": 0.32202385988662324,
+      "grad_norm": 0.21313358843326569,
+      "learning_rate": 0.001,
+      "loss": 2.2837,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3220661646501396,
+      "grad_norm": 0.4032410979270935,
+      "learning_rate": 0.001,
+      "loss": 2.0786,
+      "step": 7613
+    },
+    {
+      "epoch": 0.32210846941365595,
+      "grad_norm": 0.18987895548343658,
+      "learning_rate": 0.001,
+      "loss": 2.3311,
+      "step": 7614
+    },
+    {
+      "epoch": 0.32215077417717236,
+      "grad_norm": 7.441311359405518,
+      "learning_rate": 0.001,
+      "loss": 2.2271,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3221930789406887,
+      "grad_norm": 0.20409175753593445,
+      "learning_rate": 0.001,
+      "loss": 2.9311,
+      "step": 7616
+    },
+    {
+      "epoch": 0.32223538370420507,
+      "grad_norm": 0.19200804829597473,
+      "learning_rate": 0.001,
+      "loss": 1.5997,
+      "step": 7617
+    },
+    {
+      "epoch": 0.3222776884677215,
+      "grad_norm": 0.2112921178340912,
+      "learning_rate": 0.001,
+      "loss": 1.853,
+      "step": 7618
+    },
+    {
+      "epoch": 0.32231999323123783,
+      "grad_norm": 0.5670382976531982,
+      "learning_rate": 0.001,
+      "loss": 2.1103,
+      "step": 7619
+    },
+    {
+      "epoch": 0.3223622979947542,
+      "grad_norm": 1.1795889139175415,
+      "learning_rate": 0.001,
+      "loss": 1.5708,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3224046027582706,
+      "grad_norm": 0.19757835566997528,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 7621
+    },
+    {
+      "epoch": 0.32244690752178695,
+      "grad_norm": 0.3224509060382843,
+      "learning_rate": 0.001,
+      "loss": 2.1601,
+      "step": 7622
+    },
+    {
+      "epoch": 0.3224892122853033,
+      "grad_norm": 3.6661248207092285,
+      "learning_rate": 0.001,
+      "loss": 2.2917,
+      "step": 7623
+    },
+    {
+      "epoch": 0.3225315170488197,
+      "grad_norm": 0.26291152834892273,
+      "learning_rate": 0.001,
+      "loss": 2.5686,
+      "step": 7624
+    },
+    {
+      "epoch": 0.32257382181233607,
+      "grad_norm": 0.2524521052837372,
+      "learning_rate": 0.001,
+      "loss": 2.3755,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3226161265758524,
+      "grad_norm": 0.22929400205612183,
+      "learning_rate": 0.001,
+      "loss": 2.6761,
+      "step": 7626
+    },
+    {
+      "epoch": 0.32265843133936883,
+      "grad_norm": 0.2434311956167221,
+      "learning_rate": 0.001,
+      "loss": 2.5041,
+      "step": 7627
+    },
+    {
+      "epoch": 0.3227007361028852,
+      "grad_norm": 0.3644380569458008,
+      "learning_rate": 0.001,
+      "loss": 2.0181,
+      "step": 7628
+    },
+    {
+      "epoch": 0.32274304086640154,
+      "grad_norm": 0.8881004452705383,
+      "learning_rate": 0.001,
+      "loss": 2.3105,
+      "step": 7629
+    },
+    {
+      "epoch": 0.32278534562991795,
+      "grad_norm": 1.9201459884643555,
+      "learning_rate": 0.001,
+      "loss": 2.3565,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3228276503934343,
+      "grad_norm": 0.22108227014541626,
+      "learning_rate": 0.001,
+      "loss": 1.9022,
+      "step": 7631
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.5766029953956604,
+      "learning_rate": 0.001,
+      "loss": 1.7875,
+      "step": 7632
+    },
+    {
+      "epoch": 0.32291225992046707,
+      "grad_norm": 0.23156306147575378,
+      "learning_rate": 0.001,
+      "loss": 2.9351,
+      "step": 7633
+    },
+    {
+      "epoch": 0.3229545646839834,
+      "grad_norm": 0.20917218923568726,
+      "learning_rate": 0.001,
+      "loss": 1.7399,
+      "step": 7634
+    },
+    {
+      "epoch": 0.3229968694474998,
+      "grad_norm": 0.2092711329460144,
+      "learning_rate": 0.001,
+      "loss": 2.2154,
+      "step": 7635
+    },
+    {
+      "epoch": 0.32303917421101613,
+      "grad_norm": 1.8786296844482422,
+      "learning_rate": 0.001,
+      "loss": 1.755,
+      "step": 7636
+    },
+    {
+      "epoch": 0.32308147897453254,
+      "grad_norm": 5.879979133605957,
+      "learning_rate": 0.001,
+      "loss": 2.5399,
+      "step": 7637
+    },
+    {
+      "epoch": 0.3231237837380489,
+      "grad_norm": 1.0984258651733398,
+      "learning_rate": 0.001,
+      "loss": 2.2887,
+      "step": 7638
+    },
+    {
+      "epoch": 0.32316608850156525,
+      "grad_norm": 0.2462794929742813,
+      "learning_rate": 0.001,
+      "loss": 1.8517,
+      "step": 7639
+    },
+    {
+      "epoch": 0.32320839326508166,
+      "grad_norm": 0.22659048438072205,
+      "learning_rate": 0.001,
+      "loss": 1.6135,
+      "step": 7640
+    },
+    {
+      "epoch": 0.323250698028598,
+      "grad_norm": 0.2993631064891815,
+      "learning_rate": 0.001,
+      "loss": 1.9505,
+      "step": 7641
+    },
+    {
+      "epoch": 0.32329300279211437,
+      "grad_norm": 0.43561097979545593,
+      "learning_rate": 0.001,
+      "loss": 2.3488,
+      "step": 7642
+    },
+    {
+      "epoch": 0.3233353075556308,
+      "grad_norm": 0.2653340995311737,
+      "learning_rate": 0.001,
+      "loss": 3.7175,
+      "step": 7643
+    },
+    {
+      "epoch": 0.32337761231914713,
+      "grad_norm": 0.2567252218723297,
+      "learning_rate": 0.001,
+      "loss": 2.252,
+      "step": 7644
+    },
+    {
+      "epoch": 0.3234199170826635,
+      "grad_norm": 0.2634236514568329,
+      "learning_rate": 0.001,
+      "loss": 2.5914,
+      "step": 7645
+    },
+    {
+      "epoch": 0.3234622218461799,
+      "grad_norm": 0.21972741186618805,
+      "learning_rate": 0.001,
+      "loss": 1.9794,
+      "step": 7646
+    },
+    {
+      "epoch": 0.32350452660969625,
+      "grad_norm": 0.37712669372558594,
+      "learning_rate": 0.001,
+      "loss": 2.7509,
+      "step": 7647
+    },
+    {
+      "epoch": 0.3235468313732126,
+      "grad_norm": 0.7490605711936951,
+      "learning_rate": 0.001,
+      "loss": 2.8407,
+      "step": 7648
+    },
+    {
+      "epoch": 0.323589136136729,
+      "grad_norm": 0.38894638419151306,
+      "learning_rate": 0.001,
+      "loss": 2.2041,
+      "step": 7649
+    },
+    {
+      "epoch": 0.32363144090024537,
+      "grad_norm": 0.4028621315956116,
+      "learning_rate": 0.001,
+      "loss": 2.0586,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3236737456637617,
+      "grad_norm": 2.477647542953491,
+      "learning_rate": 0.001,
+      "loss": 3.4487,
+      "step": 7651
+    },
+    {
+      "epoch": 0.32371605042727813,
+      "grad_norm": 0.26158443093299866,
+      "learning_rate": 0.001,
+      "loss": 2.305,
+      "step": 7652
+    },
+    {
+      "epoch": 0.3237583551907945,
+      "grad_norm": 0.2559559643268585,
+      "learning_rate": 0.001,
+      "loss": 2.3287,
+      "step": 7653
+    },
+    {
+      "epoch": 0.32380065995431084,
+      "grad_norm": 0.2276749461889267,
+      "learning_rate": 0.001,
+      "loss": 2.2801,
+      "step": 7654
+    },
+    {
+      "epoch": 0.32384296471782725,
+      "grad_norm": 0.2165883332490921,
+      "learning_rate": 0.001,
+      "loss": 2.4184,
+      "step": 7655
+    },
+    {
+      "epoch": 0.3238852694813436,
+      "grad_norm": 1.6801161766052246,
+      "learning_rate": 0.001,
+      "loss": 2.2247,
+      "step": 7656
+    },
+    {
+      "epoch": 0.32392757424485996,
+      "grad_norm": 0.23615668714046478,
+      "learning_rate": 0.001,
+      "loss": 3.3127,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3239698790083763,
+      "grad_norm": 0.20711363852024078,
+      "learning_rate": 0.001,
+      "loss": 2.502,
+      "step": 7658
+    },
+    {
+      "epoch": 0.3240121837718927,
+      "grad_norm": 0.20527245104312897,
+      "learning_rate": 0.001,
+      "loss": 1.7657,
+      "step": 7659
+    },
+    {
+      "epoch": 0.3240544885354091,
+      "grad_norm": 0.18716438114643097,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 7660
+    },
+    {
+      "epoch": 0.32409679329892543,
+      "grad_norm": 0.25125744938850403,
+      "learning_rate": 0.001,
+      "loss": 2.0741,
+      "step": 7661
+    },
+    {
+      "epoch": 0.32413909806244184,
+      "grad_norm": 0.1993718147277832,
+      "learning_rate": 0.001,
+      "loss": 3.3968,
+      "step": 7662
+    },
+    {
+      "epoch": 0.3241814028259582,
+      "grad_norm": 0.1872645914554596,
+      "learning_rate": 0.001,
+      "loss": 2.0158,
+      "step": 7663
+    },
+    {
+      "epoch": 0.32422370758947455,
+      "grad_norm": 0.19530023634433746,
+      "learning_rate": 0.001,
+      "loss": 1.9639,
+      "step": 7664
+    },
+    {
+      "epoch": 0.32426601235299096,
+      "grad_norm": 0.2583797872066498,
+      "learning_rate": 0.001,
+      "loss": 1.7254,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3243083171165073,
+      "grad_norm": 2.2203147411346436,
+      "learning_rate": 0.001,
+      "loss": 2.6183,
+      "step": 7666
+    },
+    {
+      "epoch": 0.32435062188002367,
+      "grad_norm": 0.19957537949085236,
+      "learning_rate": 0.001,
+      "loss": 2.202,
+      "step": 7667
+    },
+    {
+      "epoch": 0.3243929266435401,
+      "grad_norm": 0.2118930220603943,
+      "learning_rate": 0.001,
+      "loss": 3.7787,
+      "step": 7668
+    },
+    {
+      "epoch": 0.32443523140705643,
+      "grad_norm": 0.21599611639976501,
+      "learning_rate": 0.001,
+      "loss": 2.5277,
+      "step": 7669
+    },
+    {
+      "epoch": 0.3244775361705728,
+      "grad_norm": 3.4629387855529785,
+      "learning_rate": 0.001,
+      "loss": 1.6031,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3245198409340892,
+      "grad_norm": 0.29955390095710754,
+      "learning_rate": 0.001,
+      "loss": 1.6658,
+      "step": 7671
+    },
+    {
+      "epoch": 0.32456214569760555,
+      "grad_norm": 0.20181111991405487,
+      "learning_rate": 0.001,
+      "loss": 1.9581,
+      "step": 7672
+    },
+    {
+      "epoch": 0.3246044504611219,
+      "grad_norm": 0.8020063638687134,
+      "learning_rate": 0.001,
+      "loss": 2.6512,
+      "step": 7673
+    },
+    {
+      "epoch": 0.3246467552246383,
+      "grad_norm": 0.17198266088962555,
+      "learning_rate": 0.001,
+      "loss": 2.0142,
+      "step": 7674
+    },
+    {
+      "epoch": 0.32468905998815467,
+      "grad_norm": 0.25351399183273315,
+      "learning_rate": 0.001,
+      "loss": 2.301,
+      "step": 7675
+    },
+    {
+      "epoch": 0.324731364751671,
+      "grad_norm": 0.28780749440193176,
+      "learning_rate": 0.001,
+      "loss": 1.8023,
+      "step": 7676
+    },
+    {
+      "epoch": 0.32477366951518744,
+      "grad_norm": 8.679863929748535,
+      "learning_rate": 0.001,
+      "loss": 2.0141,
+      "step": 7677
+    },
+    {
+      "epoch": 0.3248159742787038,
+      "grad_norm": 1.8151581287384033,
+      "learning_rate": 0.001,
+      "loss": 1.7588,
+      "step": 7678
+    },
+    {
+      "epoch": 0.32485827904222014,
+      "grad_norm": 0.19262565672397614,
+      "learning_rate": 0.001,
+      "loss": 1.7654,
+      "step": 7679
+    },
+    {
+      "epoch": 0.32490058380573655,
+      "grad_norm": 0.2888406217098236,
+      "learning_rate": 0.001,
+      "loss": 2.3853,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3249428885692529,
+      "grad_norm": 0.5525591373443604,
+      "learning_rate": 0.001,
+      "loss": 2.9946,
+      "step": 7681
+    },
+    {
+      "epoch": 0.32498519333276926,
+      "grad_norm": 0.319832444190979,
+      "learning_rate": 0.001,
+      "loss": 2.0593,
+      "step": 7682
+    },
+    {
+      "epoch": 0.3250274980962856,
+      "grad_norm": 0.3268459737300873,
+      "learning_rate": 0.001,
+      "loss": 2.1757,
+      "step": 7683
+    },
+    {
+      "epoch": 0.325069802859802,
+      "grad_norm": 2.9600412845611572,
+      "learning_rate": 0.001,
+      "loss": 2.3893,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3251121076233184,
+      "grad_norm": 0.9362667202949524,
+      "learning_rate": 0.001,
+      "loss": 2.2399,
+      "step": 7685
+    },
+    {
+      "epoch": 0.32515441238683473,
+      "grad_norm": 0.509442150592804,
+      "learning_rate": 0.001,
+      "loss": 2.1812,
+      "step": 7686
+    },
+    {
+      "epoch": 0.32519671715035114,
+      "grad_norm": 0.6412566900253296,
+      "learning_rate": 0.001,
+      "loss": 2.117,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3252390219138675,
+      "grad_norm": 0.25562146306037903,
+      "learning_rate": 0.001,
+      "loss": 2.2778,
+      "step": 7688
+    },
+    {
+      "epoch": 0.32528132667738385,
+      "grad_norm": 0.20876967906951904,
+      "learning_rate": 0.001,
+      "loss": 1.7797,
+      "step": 7689
+    },
+    {
+      "epoch": 0.32532363144090026,
+      "grad_norm": 0.2410915642976761,
+      "learning_rate": 0.001,
+      "loss": 1.6797,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3253659362044166,
+      "grad_norm": 0.26328933238983154,
+      "learning_rate": 0.001,
+      "loss": 2.514,
+      "step": 7691
+    },
+    {
+      "epoch": 0.32540824096793297,
+      "grad_norm": 0.23529529571533203,
+      "learning_rate": 0.001,
+      "loss": 2.1006,
+      "step": 7692
+    },
+    {
+      "epoch": 0.3254505457314494,
+      "grad_norm": 0.31655412912368774,
+      "learning_rate": 0.001,
+      "loss": 2.7627,
+      "step": 7693
+    },
+    {
+      "epoch": 0.32549285049496574,
+      "grad_norm": 0.6058570742607117,
+      "learning_rate": 0.001,
+      "loss": 2.5461,
+      "step": 7694
+    },
+    {
+      "epoch": 0.3255351552584821,
+      "grad_norm": 0.21855738759040833,
+      "learning_rate": 0.001,
+      "loss": 1.8018,
+      "step": 7695
+    },
+    {
+      "epoch": 0.3255774600219985,
+      "grad_norm": 0.22689220309257507,
+      "learning_rate": 0.001,
+      "loss": 2.122,
+      "step": 7696
+    },
+    {
+      "epoch": 0.32561976478551485,
+      "grad_norm": 0.22445081174373627,
+      "learning_rate": 0.001,
+      "loss": 1.8022,
+      "step": 7697
+    },
+    {
+      "epoch": 0.3256620695490312,
+      "grad_norm": 0.2299547791481018,
+      "learning_rate": 0.001,
+      "loss": 2.2955,
+      "step": 7698
+    },
+    {
+      "epoch": 0.3257043743125476,
+      "grad_norm": 0.5000811219215393,
+      "learning_rate": 0.001,
+      "loss": 2.6577,
+      "step": 7699
+    },
+    {
+      "epoch": 0.32574667907606397,
+      "grad_norm": 2.1569273471832275,
+      "learning_rate": 0.001,
+      "loss": 2.2027,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3257889838395803,
+      "grad_norm": 0.25455546379089355,
+      "learning_rate": 0.001,
+      "loss": 2.4252,
+      "step": 7701
+    },
+    {
+      "epoch": 0.32583128860309674,
+      "grad_norm": 0.21060289442539215,
+      "learning_rate": 0.001,
+      "loss": 2.2641,
+      "step": 7702
+    },
+    {
+      "epoch": 0.3258735933666131,
+      "grad_norm": 0.1968967467546463,
+      "learning_rate": 0.001,
+      "loss": 2.5999,
+      "step": 7703
+    },
+    {
+      "epoch": 0.32591589813012944,
+      "grad_norm": 0.31110793352127075,
+      "learning_rate": 0.001,
+      "loss": 2.2596,
+      "step": 7704
+    },
+    {
+      "epoch": 0.3259582028936458,
+      "grad_norm": 2.084595203399658,
+      "learning_rate": 0.001,
+      "loss": 2.7048,
+      "step": 7705
+    },
+    {
+      "epoch": 0.3260005076571622,
+      "grad_norm": 0.22130149602890015,
+      "learning_rate": 0.001,
+      "loss": 1.5586,
+      "step": 7706
+    },
+    {
+      "epoch": 0.32604281242067856,
+      "grad_norm": 0.2199217677116394,
+      "learning_rate": 0.001,
+      "loss": 1.9871,
+      "step": 7707
+    },
+    {
+      "epoch": 0.3260851171841949,
+      "grad_norm": 0.19622403383255005,
+      "learning_rate": 0.001,
+      "loss": 1.8239,
+      "step": 7708
+    },
+    {
+      "epoch": 0.3261274219477113,
+      "grad_norm": 0.23500365018844604,
+      "learning_rate": 0.001,
+      "loss": 2.7668,
+      "step": 7709
+    },
+    {
+      "epoch": 0.3261697267112277,
+      "grad_norm": 0.4715394973754883,
+      "learning_rate": 0.001,
+      "loss": 3.1356,
+      "step": 7710
+    },
+    {
+      "epoch": 0.32621203147474404,
+      "grad_norm": 0.6608024835586548,
+      "learning_rate": 0.001,
+      "loss": 2.1273,
+      "step": 7711
+    },
+    {
+      "epoch": 0.32625433623826045,
+      "grad_norm": 0.2573659420013428,
+      "learning_rate": 0.001,
+      "loss": 2.0172,
+      "step": 7712
+    },
+    {
+      "epoch": 0.3262966410017768,
+      "grad_norm": 3.757869243621826,
+      "learning_rate": 0.001,
+      "loss": 2.2053,
+      "step": 7713
+    },
+    {
+      "epoch": 0.32633894576529315,
+      "grad_norm": 0.36734887957572937,
+      "learning_rate": 0.001,
+      "loss": 2.188,
+      "step": 7714
+    },
+    {
+      "epoch": 0.32638125052880956,
+      "grad_norm": 0.8497750163078308,
+      "learning_rate": 0.001,
+      "loss": 3.191,
+      "step": 7715
+    },
+    {
+      "epoch": 0.3264235552923259,
+      "grad_norm": 1.0639852285385132,
+      "learning_rate": 0.001,
+      "loss": 2.1099,
+      "step": 7716
+    },
+    {
+      "epoch": 0.32646586005584227,
+      "grad_norm": 0.19022592902183533,
+      "learning_rate": 0.001,
+      "loss": 1.6883,
+      "step": 7717
+    },
+    {
+      "epoch": 0.3265081648193587,
+      "grad_norm": 0.19042854011058807,
+      "learning_rate": 0.001,
+      "loss": 2.3408,
+      "step": 7718
+    },
+    {
+      "epoch": 0.32655046958287504,
+      "grad_norm": 0.2183213233947754,
+      "learning_rate": 0.001,
+      "loss": 2.0902,
+      "step": 7719
+    },
+    {
+      "epoch": 0.3265927743463914,
+      "grad_norm": 0.2381005436182022,
+      "learning_rate": 0.001,
+      "loss": 2.5141,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3266350791099078,
+      "grad_norm": 0.2116747945547104,
+      "learning_rate": 0.001,
+      "loss": 1.8768,
+      "step": 7721
+    },
+    {
+      "epoch": 0.32667738387342415,
+      "grad_norm": 0.43030673265457153,
+      "learning_rate": 0.001,
+      "loss": 3.133,
+      "step": 7722
+    },
+    {
+      "epoch": 0.3267196886369405,
+      "grad_norm": 3.445666551589966,
+      "learning_rate": 0.001,
+      "loss": 1.8229,
+      "step": 7723
+    },
+    {
+      "epoch": 0.3267619934004569,
+      "grad_norm": 1.1962391138076782,
+      "learning_rate": 0.001,
+      "loss": 2.3559,
+      "step": 7724
+    },
+    {
+      "epoch": 0.3268042981639733,
+      "grad_norm": 0.4430636763572693,
+      "learning_rate": 0.001,
+      "loss": 2.3892,
+      "step": 7725
+    },
+    {
+      "epoch": 0.3268466029274896,
+      "grad_norm": 0.21626804769039154,
+      "learning_rate": 0.001,
+      "loss": 2.156,
+      "step": 7726
+    },
+    {
+      "epoch": 0.326888907691006,
+      "grad_norm": 0.21447373926639557,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 7727
+    },
+    {
+      "epoch": 0.3269312124545224,
+      "grad_norm": 0.2664428949356079,
+      "learning_rate": 0.001,
+      "loss": 2.1011,
+      "step": 7728
+    },
+    {
+      "epoch": 0.32697351721803875,
+      "grad_norm": 0.19551439583301544,
+      "learning_rate": 0.001,
+      "loss": 2.5206,
+      "step": 7729
+    },
+    {
+      "epoch": 0.3270158219815551,
+      "grad_norm": 0.1879674643278122,
+      "learning_rate": 0.001,
+      "loss": 1.8628,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3270581267450715,
+      "grad_norm": 0.20081692934036255,
+      "learning_rate": 0.001,
+      "loss": 1.8979,
+      "step": 7731
+    },
+    {
+      "epoch": 0.32710043150858786,
+      "grad_norm": 0.178413525223732,
+      "learning_rate": 0.001,
+      "loss": 2.3565,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3271427362721042,
+      "grad_norm": 0.384456992149353,
+      "learning_rate": 0.001,
+      "loss": 2.2038,
+      "step": 7733
+    },
+    {
+      "epoch": 0.3271850410356206,
+      "grad_norm": 0.29158031940460205,
+      "learning_rate": 0.001,
+      "loss": 2.7088,
+      "step": 7734
+    },
+    {
+      "epoch": 0.327227345799137,
+      "grad_norm": 0.17012235522270203,
+      "learning_rate": 0.001,
+      "loss": 1.5478,
+      "step": 7735
+    },
+    {
+      "epoch": 0.32726965056265334,
+      "grad_norm": 0.39595428109169006,
+      "learning_rate": 0.001,
+      "loss": 2.7015,
+      "step": 7736
+    },
+    {
+      "epoch": 0.32731195532616975,
+      "grad_norm": 0.58181232213974,
+      "learning_rate": 0.001,
+      "loss": 1.7588,
+      "step": 7737
+    },
+    {
+      "epoch": 0.3273542600896861,
+      "grad_norm": 0.20064769685268402,
+      "learning_rate": 0.001,
+      "loss": 1.8287,
+      "step": 7738
+    },
+    {
+      "epoch": 0.32739656485320245,
+      "grad_norm": 0.2265099585056305,
+      "learning_rate": 0.001,
+      "loss": 2.289,
+      "step": 7739
+    },
+    {
+      "epoch": 0.32743886961671886,
+      "grad_norm": 0.2981756031513214,
+      "learning_rate": 0.001,
+      "loss": 2.1712,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3274811743802352,
+      "grad_norm": 1.7779628038406372,
+      "learning_rate": 0.001,
+      "loss": 3.4488,
+      "step": 7741
+    },
+    {
+      "epoch": 0.3275234791437516,
+      "grad_norm": 0.2585592269897461,
+      "learning_rate": 0.001,
+      "loss": 2.4738,
+      "step": 7742
+    },
+    {
+      "epoch": 0.327565783907268,
+      "grad_norm": 0.9747766256332397,
+      "learning_rate": 0.001,
+      "loss": 3.1085,
+      "step": 7743
+    },
+    {
+      "epoch": 0.32760808867078434,
+      "grad_norm": 6.341303825378418,
+      "learning_rate": 0.001,
+      "loss": 2.3515,
+      "step": 7744
+    },
+    {
+      "epoch": 0.3276503934343007,
+      "grad_norm": 0.28042072057724,
+      "learning_rate": 0.001,
+      "loss": 1.8256,
+      "step": 7745
+    },
+    {
+      "epoch": 0.3276926981978171,
+      "grad_norm": 0.21017926931381226,
+      "learning_rate": 0.001,
+      "loss": 2.2271,
+      "step": 7746
+    },
+    {
+      "epoch": 0.32773500296133345,
+      "grad_norm": 0.259328156709671,
+      "learning_rate": 0.001,
+      "loss": 2.6266,
+      "step": 7747
+    },
+    {
+      "epoch": 0.3277773077248498,
+      "grad_norm": 0.16318392753601074,
+      "learning_rate": 0.001,
+      "loss": 1.8191,
+      "step": 7748
+    },
+    {
+      "epoch": 0.32781961248836616,
+      "grad_norm": 0.27874764800071716,
+      "learning_rate": 0.001,
+      "loss": 1.9758,
+      "step": 7749
+    },
+    {
+      "epoch": 0.3278619172518826,
+      "grad_norm": 0.2953718304634094,
+      "learning_rate": 0.001,
+      "loss": 3.011,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3279042220153989,
+      "grad_norm": 0.3343517780303955,
+      "learning_rate": 0.001,
+      "loss": 2.1521,
+      "step": 7751
+    },
+    {
+      "epoch": 0.3279465267789153,
+      "grad_norm": 0.25767582654953003,
+      "learning_rate": 0.001,
+      "loss": 2.672,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3279888315424317,
+      "grad_norm": 0.21539786458015442,
+      "learning_rate": 0.001,
+      "loss": 2.3109,
+      "step": 7753
+    },
+    {
+      "epoch": 0.32803113630594805,
+      "grad_norm": 0.22561384737491608,
+      "learning_rate": 0.001,
+      "loss": 2.5035,
+      "step": 7754
+    },
+    {
+      "epoch": 0.3280734410694644,
+      "grad_norm": 0.2779701054096222,
+      "learning_rate": 0.001,
+      "loss": 2.5272,
+      "step": 7755
+    },
+    {
+      "epoch": 0.3281157458329808,
+      "grad_norm": 0.19811490178108215,
+      "learning_rate": 0.001,
+      "loss": 1.7952,
+      "step": 7756
+    },
+    {
+      "epoch": 0.32815805059649716,
+      "grad_norm": 0.20879384875297546,
+      "learning_rate": 0.001,
+      "loss": 1.2741,
+      "step": 7757
+    },
+    {
+      "epoch": 0.3282003553600135,
+      "grad_norm": 0.8248985409736633,
+      "learning_rate": 0.001,
+      "loss": 1.9017,
+      "step": 7758
+    },
+    {
+      "epoch": 0.32824266012352993,
+      "grad_norm": 0.20458753407001495,
+      "learning_rate": 0.001,
+      "loss": 1.9238,
+      "step": 7759
+    },
+    {
+      "epoch": 0.3282849648870463,
+      "grad_norm": 0.32681992650032043,
+      "learning_rate": 0.001,
+      "loss": 3.2441,
+      "step": 7760
+    },
+    {
+      "epoch": 0.32832726965056264,
+      "grad_norm": 6.789393424987793,
+      "learning_rate": 0.001,
+      "loss": 2.4838,
+      "step": 7761
+    },
+    {
+      "epoch": 0.32836957441407905,
+      "grad_norm": 0.49927785992622375,
+      "learning_rate": 0.001,
+      "loss": 2.888,
+      "step": 7762
+    },
+    {
+      "epoch": 0.3284118791775954,
+      "grad_norm": 2.3099589347839355,
+      "learning_rate": 0.001,
+      "loss": 2.6976,
+      "step": 7763
+    },
+    {
+      "epoch": 0.32845418394111175,
+      "grad_norm": 2.1838667392730713,
+      "learning_rate": 0.001,
+      "loss": 2.2512,
+      "step": 7764
+    },
+    {
+      "epoch": 0.32849648870462816,
+      "grad_norm": 0.39657530188560486,
+      "learning_rate": 0.001,
+      "loss": 2.7728,
+      "step": 7765
+    },
+    {
+      "epoch": 0.3285387934681445,
+      "grad_norm": 0.2155158668756485,
+      "learning_rate": 0.001,
+      "loss": 2.2733,
+      "step": 7766
+    },
+    {
+      "epoch": 0.3285810982316609,
+      "grad_norm": 0.46449387073516846,
+      "learning_rate": 0.001,
+      "loss": 3.3834,
+      "step": 7767
+    },
+    {
+      "epoch": 0.3286234029951773,
+      "grad_norm": 0.20302799344062805,
+      "learning_rate": 0.001,
+      "loss": 2.1242,
+      "step": 7768
+    },
+    {
+      "epoch": 0.32866570775869364,
+      "grad_norm": 0.1897967904806137,
+      "learning_rate": 0.001,
+      "loss": 2.2495,
+      "step": 7769
+    },
+    {
+      "epoch": 0.32870801252221,
+      "grad_norm": 0.17885787785053253,
+      "learning_rate": 0.001,
+      "loss": 1.9588,
+      "step": 7770
+    },
+    {
+      "epoch": 0.32875031728572635,
+      "grad_norm": 0.25244593620300293,
+      "learning_rate": 0.001,
+      "loss": 2.8224,
+      "step": 7771
+    },
+    {
+      "epoch": 0.32879262204924276,
+      "grad_norm": 0.17659065127372742,
+      "learning_rate": 0.001,
+      "loss": 1.5998,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3288349268127591,
+      "grad_norm": 0.19181221723556519,
+      "learning_rate": 0.001,
+      "loss": 1.8639,
+      "step": 7773
+    },
+    {
+      "epoch": 0.32887723157627546,
+      "grad_norm": 0.3274793028831482,
+      "learning_rate": 0.001,
+      "loss": 2.3154,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3289195363397919,
+      "grad_norm": 0.21050041913986206,
+      "learning_rate": 0.001,
+      "loss": 2.1191,
+      "step": 7775
+    },
+    {
+      "epoch": 0.32896184110330823,
+      "grad_norm": 0.17837364971637726,
+      "learning_rate": 0.001,
+      "loss": 1.7701,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3290041458668246,
+      "grad_norm": 0.6773764491081238,
+      "learning_rate": 0.001,
+      "loss": 2.6309,
+      "step": 7777
+    },
+    {
+      "epoch": 0.329046450630341,
+      "grad_norm": 2.140749216079712,
+      "learning_rate": 0.001,
+      "loss": 3.2878,
+      "step": 7778
+    },
+    {
+      "epoch": 0.32908875539385735,
+      "grad_norm": 4.552440166473389,
+      "learning_rate": 0.001,
+      "loss": 2.2279,
+      "step": 7779
+    },
+    {
+      "epoch": 0.3291310601573737,
+      "grad_norm": 0.37157878279685974,
+      "learning_rate": 0.001,
+      "loss": 2.3027,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3291733649208901,
+      "grad_norm": 0.25460559129714966,
+      "learning_rate": 0.001,
+      "loss": 2.0759,
+      "step": 7781
+    },
+    {
+      "epoch": 0.32921566968440646,
+      "grad_norm": 1.6842424869537354,
+      "learning_rate": 0.001,
+      "loss": 2.857,
+      "step": 7782
+    },
+    {
+      "epoch": 0.3292579744479228,
+      "grad_norm": 0.2335149198770523,
+      "learning_rate": 0.001,
+      "loss": 1.7248,
+      "step": 7783
+    },
+    {
+      "epoch": 0.32930027921143923,
+      "grad_norm": 0.1995423585176468,
+      "learning_rate": 0.001,
+      "loss": 1.7434,
+      "step": 7784
+    },
+    {
+      "epoch": 0.3293425839749556,
+      "grad_norm": 0.2805491089820862,
+      "learning_rate": 0.001,
+      "loss": 2.6196,
+      "step": 7785
+    },
+    {
+      "epoch": 0.32938488873847194,
+      "grad_norm": 0.7963484525680542,
+      "learning_rate": 0.001,
+      "loss": 2.7678,
+      "step": 7786
+    },
+    {
+      "epoch": 0.32942719350198835,
+      "grad_norm": 0.2746239900588989,
+      "learning_rate": 0.001,
+      "loss": 2.0567,
+      "step": 7787
+    },
+    {
+      "epoch": 0.3294694982655047,
+      "grad_norm": 0.6149387955665588,
+      "learning_rate": 0.001,
+      "loss": 2.4013,
+      "step": 7788
+    },
+    {
+      "epoch": 0.32951180302902106,
+      "grad_norm": 0.5365425944328308,
+      "learning_rate": 0.001,
+      "loss": 2.2887,
+      "step": 7789
+    },
+    {
+      "epoch": 0.32955410779253747,
+      "grad_norm": 0.22529155015945435,
+      "learning_rate": 0.001,
+      "loss": 2.5844,
+      "step": 7790
+    },
+    {
+      "epoch": 0.3295964125560538,
+      "grad_norm": 0.28875237703323364,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 7791
+    },
+    {
+      "epoch": 0.3296387173195702,
+      "grad_norm": 0.20266857743263245,
+      "learning_rate": 0.001,
+      "loss": 2.5674,
+      "step": 7792
+    },
+    {
+      "epoch": 0.32968102208308653,
+      "grad_norm": 0.16716927289962769,
+      "learning_rate": 0.001,
+      "loss": 2.2583,
+      "step": 7793
+    },
+    {
+      "epoch": 0.32972332684660294,
+      "grad_norm": 0.25711336731910706,
+      "learning_rate": 0.001,
+      "loss": 2.9724,
+      "step": 7794
+    },
+    {
+      "epoch": 0.3297656316101193,
+      "grad_norm": 4.552552700042725,
+      "learning_rate": 0.001,
+      "loss": 2.6996,
+      "step": 7795
+    },
+    {
+      "epoch": 0.32980793637363565,
+      "grad_norm": 0.5906350612640381,
+      "learning_rate": 0.001,
+      "loss": 2.1236,
+      "step": 7796
+    },
+    {
+      "epoch": 0.32985024113715206,
+      "grad_norm": 0.2682677209377289,
+      "learning_rate": 0.001,
+      "loss": 2.4385,
+      "step": 7797
+    },
+    {
+      "epoch": 0.3298925459006684,
+      "grad_norm": 0.25783124566078186,
+      "learning_rate": 0.001,
+      "loss": 2.3089,
+      "step": 7798
+    },
+    {
+      "epoch": 0.32993485066418476,
+      "grad_norm": 0.18664343655109406,
+      "learning_rate": 0.001,
+      "loss": 2.0034,
+      "step": 7799
+    },
+    {
+      "epoch": 0.3299771554277012,
+      "grad_norm": 0.1630748063325882,
+      "learning_rate": 0.001,
+      "loss": 1.6401,
+      "step": 7800
+    },
+    {
+      "epoch": 0.33001946019121753,
+      "grad_norm": 0.30394893884658813,
+      "learning_rate": 0.001,
+      "loss": 2.3112,
+      "step": 7801
+    },
+    {
+      "epoch": 0.3300617649547339,
+      "grad_norm": 0.19138090312480927,
+      "learning_rate": 0.001,
+      "loss": 2.5035,
+      "step": 7802
+    },
+    {
+      "epoch": 0.3301040697182503,
+      "grad_norm": 0.16754819452762604,
+      "learning_rate": 0.001,
+      "loss": 1.6879,
+      "step": 7803
+    },
+    {
+      "epoch": 0.33014637448176665,
+      "grad_norm": 0.8580173850059509,
+      "learning_rate": 0.001,
+      "loss": 2.4691,
+      "step": 7804
+    },
+    {
+      "epoch": 0.330188679245283,
+      "grad_norm": 0.176096573472023,
+      "learning_rate": 0.001,
+      "loss": 1.8854,
+      "step": 7805
+    },
+    {
+      "epoch": 0.3302309840087994,
+      "grad_norm": 0.20934566855430603,
+      "learning_rate": 0.001,
+      "loss": 2.0659,
+      "step": 7806
+    },
+    {
+      "epoch": 0.33027328877231577,
+      "grad_norm": 5.532082557678223,
+      "learning_rate": 0.001,
+      "loss": 1.759,
+      "step": 7807
+    },
+    {
+      "epoch": 0.3303155935358321,
+      "grad_norm": 0.19679847359657288,
+      "learning_rate": 0.001,
+      "loss": 2.3835,
+      "step": 7808
+    },
+    {
+      "epoch": 0.33035789829934853,
+      "grad_norm": 0.19240596890449524,
+      "learning_rate": 0.001,
+      "loss": 2.0052,
+      "step": 7809
+    },
+    {
+      "epoch": 0.3304002030628649,
+      "grad_norm": 0.19509436190128326,
+      "learning_rate": 0.001,
+      "loss": 2.426,
+      "step": 7810
+    },
+    {
+      "epoch": 0.33044250782638124,
+      "grad_norm": 0.19828227162361145,
+      "learning_rate": 0.001,
+      "loss": 2.1933,
+      "step": 7811
+    },
+    {
+      "epoch": 0.33048481258989765,
+      "grad_norm": 2.229938507080078,
+      "learning_rate": 0.001,
+      "loss": 1.7994,
+      "step": 7812
+    },
+    {
+      "epoch": 0.330527117353414,
+      "grad_norm": 0.2188158929347992,
+      "learning_rate": 0.001,
+      "loss": 1.8493,
+      "step": 7813
+    },
+    {
+      "epoch": 0.33056942211693036,
+      "grad_norm": 0.21723762154579163,
+      "learning_rate": 0.001,
+      "loss": 1.8718,
+      "step": 7814
+    },
+    {
+      "epoch": 0.33061172688044677,
+      "grad_norm": 0.5078144669532776,
+      "learning_rate": 0.001,
+      "loss": 2.1031,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3306540316439631,
+      "grad_norm": 0.24922381341457367,
+      "learning_rate": 0.001,
+      "loss": 2.1853,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3306963364074795,
+      "grad_norm": 0.19677498936653137,
+      "learning_rate": 0.001,
+      "loss": 2.4259,
+      "step": 7817
+    },
+    {
+      "epoch": 0.33073864117099583,
+      "grad_norm": 0.2011793553829193,
+      "learning_rate": 0.001,
+      "loss": 3.2597,
+      "step": 7818
+    },
+    {
+      "epoch": 0.33078094593451224,
+      "grad_norm": 0.3408401608467102,
+      "learning_rate": 0.001,
+      "loss": 2.3159,
+      "step": 7819
+    },
+    {
+      "epoch": 0.3308232506980286,
+      "grad_norm": 0.1871691644191742,
+      "learning_rate": 0.001,
+      "loss": 2.5773,
+      "step": 7820
+    },
+    {
+      "epoch": 0.33086555546154495,
+      "grad_norm": 0.19966411590576172,
+      "learning_rate": 0.001,
+      "loss": 1.4954,
+      "step": 7821
+    },
+    {
+      "epoch": 0.33090786022506136,
+      "grad_norm": 0.20927104353904724,
+      "learning_rate": 0.001,
+      "loss": 2.0749,
+      "step": 7822
+    },
+    {
+      "epoch": 0.3309501649885777,
+      "grad_norm": 0.2717510759830475,
+      "learning_rate": 0.001,
+      "loss": 2.9177,
+      "step": 7823
+    },
+    {
+      "epoch": 0.33099246975209407,
+      "grad_norm": 0.2813604176044464,
+      "learning_rate": 0.001,
+      "loss": 1.8495,
+      "step": 7824
+    },
+    {
+      "epoch": 0.3310347745156105,
+      "grad_norm": 0.18087762594223022,
+      "learning_rate": 0.001,
+      "loss": 2.747,
+      "step": 7825
+    },
+    {
+      "epoch": 0.33107707927912683,
+      "grad_norm": 29.72042465209961,
+      "learning_rate": 0.001,
+      "loss": 3.0675,
+      "step": 7826
+    },
+    {
+      "epoch": 0.3311193840426432,
+      "grad_norm": 0.2071724832057953,
+      "learning_rate": 0.001,
+      "loss": 2.6322,
+      "step": 7827
+    },
+    {
+      "epoch": 0.3311616888061596,
+      "grad_norm": 0.208586648106575,
+      "learning_rate": 0.001,
+      "loss": 2.4795,
+      "step": 7828
+    },
+    {
+      "epoch": 0.33120399356967595,
+      "grad_norm": 0.19940099120140076,
+      "learning_rate": 0.001,
+      "loss": 2.5773,
+      "step": 7829
+    },
+    {
+      "epoch": 0.3312462983331923,
+      "grad_norm": 2.280409097671509,
+      "learning_rate": 0.001,
+      "loss": 2.3525,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3312886030967087,
+      "grad_norm": 0.977155864238739,
+      "learning_rate": 0.001,
+      "loss": 1.8458,
+      "step": 7831
+    },
+    {
+      "epoch": 0.33133090786022507,
+      "grad_norm": 0.33999037742614746,
+      "learning_rate": 0.001,
+      "loss": 1.8913,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3313732126237414,
+      "grad_norm": 0.23237483203411102,
+      "learning_rate": 0.001,
+      "loss": 2.69,
+      "step": 7833
+    },
+    {
+      "epoch": 0.33141551738725783,
+      "grad_norm": 0.1733536273241043,
+      "learning_rate": 0.001,
+      "loss": 2.289,
+      "step": 7834
+    },
+    {
+      "epoch": 0.3314578221507742,
+      "grad_norm": 0.19590957462787628,
+      "learning_rate": 0.001,
+      "loss": 2.3374,
+      "step": 7835
+    },
+    {
+      "epoch": 0.33150012691429054,
+      "grad_norm": 0.22737683355808258,
+      "learning_rate": 0.001,
+      "loss": 2.8243,
+      "step": 7836
+    },
+    {
+      "epoch": 0.33154243167780695,
+      "grad_norm": 0.22133977711200714,
+      "learning_rate": 0.001,
+      "loss": 1.9613,
+      "step": 7837
+    },
+    {
+      "epoch": 0.3315847364413233,
+      "grad_norm": 0.24154724180698395,
+      "learning_rate": 0.001,
+      "loss": 2.0886,
+      "step": 7838
+    },
+    {
+      "epoch": 0.33162704120483966,
+      "grad_norm": 0.1789299100637436,
+      "learning_rate": 0.001,
+      "loss": 2.6418,
+      "step": 7839
+    },
+    {
+      "epoch": 0.331669345968356,
+      "grad_norm": 0.27288487553596497,
+      "learning_rate": 0.001,
+      "loss": 2.4128,
+      "step": 7840
+    },
+    {
+      "epoch": 0.3317116507318724,
+      "grad_norm": 0.7873362302780151,
+      "learning_rate": 0.001,
+      "loss": 1.8761,
+      "step": 7841
+    },
+    {
+      "epoch": 0.3317539554953888,
+      "grad_norm": 0.32761895656585693,
+      "learning_rate": 0.001,
+      "loss": 2.2242,
+      "step": 7842
+    },
+    {
+      "epoch": 0.33179626025890513,
+      "grad_norm": 2.1952366828918457,
+      "learning_rate": 0.001,
+      "loss": 2.0018,
+      "step": 7843
+    },
+    {
+      "epoch": 0.33183856502242154,
+      "grad_norm": 0.18359605967998505,
+      "learning_rate": 0.001,
+      "loss": 1.7927,
+      "step": 7844
+    },
+    {
+      "epoch": 0.3318808697859379,
+      "grad_norm": 0.9010672569274902,
+      "learning_rate": 0.001,
+      "loss": 2.9856,
+      "step": 7845
+    },
+    {
+      "epoch": 0.33192317454945425,
+      "grad_norm": 0.23249398171901703,
+      "learning_rate": 0.001,
+      "loss": 2.6271,
+      "step": 7846
+    },
+    {
+      "epoch": 0.33196547931297066,
+      "grad_norm": 0.5741890072822571,
+      "learning_rate": 0.001,
+      "loss": 3.903,
+      "step": 7847
+    },
+    {
+      "epoch": 0.332007784076487,
+      "grad_norm": 0.5828627347946167,
+      "learning_rate": 0.001,
+      "loss": 2.3399,
+      "step": 7848
+    },
+    {
+      "epoch": 0.33205008884000337,
+      "grad_norm": 0.2302628457546234,
+      "learning_rate": 0.001,
+      "loss": 2.6404,
+      "step": 7849
+    },
+    {
+      "epoch": 0.3320923936035198,
+      "grad_norm": 1.7967714071273804,
+      "learning_rate": 0.001,
+      "loss": 2.3163,
+      "step": 7850
+    },
+    {
+      "epoch": 0.33213469836703613,
+      "grad_norm": 36.19169616699219,
+      "learning_rate": 0.001,
+      "loss": 2.172,
+      "step": 7851
+    },
+    {
+      "epoch": 0.3321770031305525,
+      "grad_norm": 4.201509952545166,
+      "learning_rate": 0.001,
+      "loss": 2.2538,
+      "step": 7852
+    },
+    {
+      "epoch": 0.3322193078940689,
+      "grad_norm": 0.23143364489078522,
+      "learning_rate": 0.001,
+      "loss": 3.6801,
+      "step": 7853
+    },
+    {
+      "epoch": 0.33226161265758525,
+      "grad_norm": 0.4308498203754425,
+      "learning_rate": 0.001,
+      "loss": 2.6071,
+      "step": 7854
+    },
+    {
+      "epoch": 0.3323039174211016,
+      "grad_norm": 0.1866970956325531,
+      "learning_rate": 0.001,
+      "loss": 3.0409,
+      "step": 7855
+    },
+    {
+      "epoch": 0.332346222184618,
+      "grad_norm": 0.20423516631126404,
+      "learning_rate": 0.001,
+      "loss": 2.1872,
+      "step": 7856
+    },
+    {
+      "epoch": 0.33238852694813437,
+      "grad_norm": 0.23156633973121643,
+      "learning_rate": 0.001,
+      "loss": 1.8005,
+      "step": 7857
+    },
+    {
+      "epoch": 0.3324308317116507,
+      "grad_norm": 0.29893800616264343,
+      "learning_rate": 0.001,
+      "loss": 1.9907,
+      "step": 7858
+    },
+    {
+      "epoch": 0.33247313647516713,
+      "grad_norm": 0.23436959087848663,
+      "learning_rate": 0.001,
+      "loss": 2.8521,
+      "step": 7859
+    },
+    {
+      "epoch": 0.3325154412386835,
+      "grad_norm": 0.1933186650276184,
+      "learning_rate": 0.001,
+      "loss": 1.6054,
+      "step": 7860
+    },
+    {
+      "epoch": 0.33255774600219984,
+      "grad_norm": 0.2026066929101944,
+      "learning_rate": 0.001,
+      "loss": 2.026,
+      "step": 7861
+    },
+    {
+      "epoch": 0.3326000507657162,
+      "grad_norm": 0.21507464349269867,
+      "learning_rate": 0.001,
+      "loss": 2.9487,
+      "step": 7862
+    },
+    {
+      "epoch": 0.3326423555292326,
+      "grad_norm": 0.2192848175764084,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 7863
+    },
+    {
+      "epoch": 0.33268466029274896,
+      "grad_norm": 2.8954203128814697,
+      "learning_rate": 0.001,
+      "loss": 1.8237,
+      "step": 7864
+    },
+    {
+      "epoch": 0.3327269650562653,
+      "grad_norm": 0.30463677644729614,
+      "learning_rate": 0.001,
+      "loss": 2.1563,
+      "step": 7865
+    },
+    {
+      "epoch": 0.3327692698197817,
+      "grad_norm": 0.45512622594833374,
+      "learning_rate": 0.001,
+      "loss": 2.8004,
+      "step": 7866
+    },
+    {
+      "epoch": 0.3328115745832981,
+      "grad_norm": 0.32765403389930725,
+      "learning_rate": 0.001,
+      "loss": 2.4544,
+      "step": 7867
+    },
+    {
+      "epoch": 0.33285387934681443,
+      "grad_norm": 0.21778345108032227,
+      "learning_rate": 0.001,
+      "loss": 2.2046,
+      "step": 7868
+    },
+    {
+      "epoch": 0.33289618411033084,
+      "grad_norm": 0.2558639645576477,
+      "learning_rate": 0.001,
+      "loss": 3.3903,
+      "step": 7869
+    },
+    {
+      "epoch": 0.3329384888738472,
+      "grad_norm": 0.27127140760421753,
+      "learning_rate": 0.001,
+      "loss": 2.1592,
+      "step": 7870
+    },
+    {
+      "epoch": 0.33298079363736355,
+      "grad_norm": 0.2440665364265442,
+      "learning_rate": 0.001,
+      "loss": 3.0212,
+      "step": 7871
+    },
+    {
+      "epoch": 0.33302309840087996,
+      "grad_norm": 0.23537136614322662,
+      "learning_rate": 0.001,
+      "loss": 2.1113,
+      "step": 7872
+    },
+    {
+      "epoch": 0.3330654031643963,
+      "grad_norm": 0.5498445630073547,
+      "learning_rate": 0.001,
+      "loss": 2.9246,
+      "step": 7873
+    },
+    {
+      "epoch": 0.33310770792791267,
+      "grad_norm": 0.19727638363838196,
+      "learning_rate": 0.001,
+      "loss": 1.9724,
+      "step": 7874
+    },
+    {
+      "epoch": 0.3331500126914291,
+      "grad_norm": 0.19722647964954376,
+      "learning_rate": 0.001,
+      "loss": 1.92,
+      "step": 7875
+    },
+    {
+      "epoch": 0.33319231745494543,
+      "grad_norm": 1.0840023756027222,
+      "learning_rate": 0.001,
+      "loss": 2.243,
+      "step": 7876
+    },
+    {
+      "epoch": 0.3332346222184618,
+      "grad_norm": 1.0607739686965942,
+      "learning_rate": 0.001,
+      "loss": 2.2591,
+      "step": 7877
+    },
+    {
+      "epoch": 0.3332769269819782,
+      "grad_norm": 0.19976572692394257,
+      "learning_rate": 0.001,
+      "loss": 2.5023,
+      "step": 7878
+    },
+    {
+      "epoch": 0.33331923174549455,
+      "grad_norm": 0.20749805867671967,
+      "learning_rate": 0.001,
+      "loss": 2.4522,
+      "step": 7879
+    },
+    {
+      "epoch": 0.3333615365090109,
+      "grad_norm": 0.1775732934474945,
+      "learning_rate": 0.001,
+      "loss": 2.4282,
+      "step": 7880
+    },
+    {
+      "epoch": 0.3334038412725273,
+      "grad_norm": 0.6955072283744812,
+      "learning_rate": 0.001,
+      "loss": 1.5459,
+      "step": 7881
+    },
+    {
+      "epoch": 0.33344614603604367,
+      "grad_norm": 2.1425440311431885,
+      "learning_rate": 0.001,
+      "loss": 1.8967,
+      "step": 7882
+    },
+    {
+      "epoch": 0.33348845079956,
+      "grad_norm": 3.9106364250183105,
+      "learning_rate": 0.001,
+      "loss": 2.2198,
+      "step": 7883
+    },
+    {
+      "epoch": 0.3335307555630764,
+      "grad_norm": 6.036744594573975,
+      "learning_rate": 0.001,
+      "loss": 2.4918,
+      "step": 7884
+    },
+    {
+      "epoch": 0.3335730603265928,
+      "grad_norm": 0.19927839934825897,
+      "learning_rate": 0.001,
+      "loss": 2.5132,
+      "step": 7885
+    },
+    {
+      "epoch": 0.33361536509010914,
+      "grad_norm": 1.190531849861145,
+      "learning_rate": 0.001,
+      "loss": 1.8756,
+      "step": 7886
+    },
+    {
+      "epoch": 0.3336576698536255,
+      "grad_norm": 0.2305949628353119,
+      "learning_rate": 0.001,
+      "loss": 3.0,
+      "step": 7887
+    },
+    {
+      "epoch": 0.3336999746171419,
+      "grad_norm": 0.26107683777809143,
+      "learning_rate": 0.001,
+      "loss": 2.8662,
+      "step": 7888
+    },
+    {
+      "epoch": 0.33374227938065826,
+      "grad_norm": 0.3804430365562439,
+      "learning_rate": 0.001,
+      "loss": 3.1798,
+      "step": 7889
+    },
+    {
+      "epoch": 0.3337845841441746,
+      "grad_norm": 0.32322415709495544,
+      "learning_rate": 0.001,
+      "loss": 2.1622,
+      "step": 7890
+    },
+    {
+      "epoch": 0.333826888907691,
+      "grad_norm": 1.6569503545761108,
+      "learning_rate": 0.001,
+      "loss": 2.8064,
+      "step": 7891
+    },
+    {
+      "epoch": 0.3338691936712074,
+      "grad_norm": 0.975721001625061,
+      "learning_rate": 0.001,
+      "loss": 2.7777,
+      "step": 7892
+    },
+    {
+      "epoch": 0.33391149843472373,
+      "grad_norm": 1.4115006923675537,
+      "learning_rate": 0.001,
+      "loss": 2.3504,
+      "step": 7893
+    },
+    {
+      "epoch": 0.33395380319824014,
+      "grad_norm": 0.6836025714874268,
+      "learning_rate": 0.001,
+      "loss": 2.1942,
+      "step": 7894
+    },
+    {
+      "epoch": 0.3339961079617565,
+      "grad_norm": 0.35912907123565674,
+      "learning_rate": 0.001,
+      "loss": 3.1702,
+      "step": 7895
+    },
+    {
+      "epoch": 0.33403841272527285,
+      "grad_norm": 1.0951977968215942,
+      "learning_rate": 0.001,
+      "loss": 3.3936,
+      "step": 7896
+    },
+    {
+      "epoch": 0.33408071748878926,
+      "grad_norm": 0.24415375292301178,
+      "learning_rate": 0.001,
+      "loss": 2.4679,
+      "step": 7897
+    },
+    {
+      "epoch": 0.3341230222523056,
+      "grad_norm": 0.23192666471004486,
+      "learning_rate": 0.001,
+      "loss": 1.486,
+      "step": 7898
+    },
+    {
+      "epoch": 0.33416532701582197,
+      "grad_norm": 0.3416637182235718,
+      "learning_rate": 0.001,
+      "loss": 1.8632,
+      "step": 7899
+    },
+    {
+      "epoch": 0.3342076317793384,
+      "grad_norm": 0.23700256645679474,
+      "learning_rate": 0.001,
+      "loss": 2.0956,
+      "step": 7900
+    },
+    {
+      "epoch": 0.33424993654285473,
+      "grad_norm": 0.2494289129972458,
+      "learning_rate": 0.001,
+      "loss": 2.3394,
+      "step": 7901
+    },
+    {
+      "epoch": 0.3342922413063711,
+      "grad_norm": 0.2167077511548996,
+      "learning_rate": 0.001,
+      "loss": 2.3218,
+      "step": 7902
+    },
+    {
+      "epoch": 0.3343345460698875,
+      "grad_norm": 1.0088181495666504,
+      "learning_rate": 0.001,
+      "loss": 2.4546,
+      "step": 7903
+    },
+    {
+      "epoch": 0.33437685083340385,
+      "grad_norm": 0.45194211602211,
+      "learning_rate": 0.001,
+      "loss": 1.9015,
+      "step": 7904
+    },
+    {
+      "epoch": 0.3344191555969202,
+      "grad_norm": 0.20241999626159668,
+      "learning_rate": 0.001,
+      "loss": 1.8324,
+      "step": 7905
+    },
+    {
+      "epoch": 0.33446146036043656,
+      "grad_norm": 0.26756224036216736,
+      "learning_rate": 0.001,
+      "loss": 3.2489,
+      "step": 7906
+    },
+    {
+      "epoch": 0.33450376512395297,
+      "grad_norm": 0.49798786640167236,
+      "learning_rate": 0.001,
+      "loss": 3.2647,
+      "step": 7907
+    },
+    {
+      "epoch": 0.3345460698874693,
+      "grad_norm": 0.2664455473423004,
+      "learning_rate": 0.001,
+      "loss": 2.7186,
+      "step": 7908
+    },
+    {
+      "epoch": 0.3345883746509857,
+      "grad_norm": 0.3100428879261017,
+      "learning_rate": 0.001,
+      "loss": 2.275,
+      "step": 7909
+    },
+    {
+      "epoch": 0.3346306794145021,
+      "grad_norm": 0.19534626603126526,
+      "learning_rate": 0.001,
+      "loss": 2.7456,
+      "step": 7910
+    },
+    {
+      "epoch": 0.33467298417801844,
+      "grad_norm": 0.6475488543510437,
+      "learning_rate": 0.001,
+      "loss": 2.2076,
+      "step": 7911
+    },
+    {
+      "epoch": 0.3347152889415348,
+      "grad_norm": 0.3301447927951813,
+      "learning_rate": 0.001,
+      "loss": 2.3495,
+      "step": 7912
+    },
+    {
+      "epoch": 0.3347575937050512,
+      "grad_norm": 2.3587558269500732,
+      "learning_rate": 0.001,
+      "loss": 2.3851,
+      "step": 7913
+    },
+    {
+      "epoch": 0.33479989846856756,
+      "grad_norm": 0.2016088366508484,
+      "learning_rate": 0.001,
+      "loss": 2.1589,
+      "step": 7914
+    },
+    {
+      "epoch": 0.3348422032320839,
+      "grad_norm": 0.2044316530227661,
+      "learning_rate": 0.001,
+      "loss": 3.0746,
+      "step": 7915
+    },
+    {
+      "epoch": 0.3348845079956003,
+      "grad_norm": 0.19156186282634735,
+      "learning_rate": 0.001,
+      "loss": 2.1999,
+      "step": 7916
+    },
+    {
+      "epoch": 0.3349268127591167,
+      "grad_norm": 0.23087522387504578,
+      "learning_rate": 0.001,
+      "loss": 3.6112,
+      "step": 7917
+    },
+    {
+      "epoch": 0.33496911752263303,
+      "grad_norm": 0.1780395656824112,
+      "learning_rate": 0.001,
+      "loss": 2.6512,
+      "step": 7918
+    },
+    {
+      "epoch": 0.33501142228614944,
+      "grad_norm": 0.20307819545269012,
+      "learning_rate": 0.001,
+      "loss": 1.8293,
+      "step": 7919
+    },
+    {
+      "epoch": 0.3350537270496658,
+      "grad_norm": 0.735317587852478,
+      "learning_rate": 0.001,
+      "loss": 1.861,
+      "step": 7920
+    },
+    {
+      "epoch": 0.33509603181318215,
+      "grad_norm": 0.2308008074760437,
+      "learning_rate": 0.001,
+      "loss": 2.4375,
+      "step": 7921
+    },
+    {
+      "epoch": 0.33513833657669856,
+      "grad_norm": 0.2518015205860138,
+      "learning_rate": 0.001,
+      "loss": 2.1128,
+      "step": 7922
+    },
+    {
+      "epoch": 0.3351806413402149,
+      "grad_norm": 0.22979722917079926,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 7923
+    },
+    {
+      "epoch": 0.33522294610373127,
+      "grad_norm": 0.20202603936195374,
+      "learning_rate": 0.001,
+      "loss": 2.629,
+      "step": 7924
+    },
+    {
+      "epoch": 0.3352652508672477,
+      "grad_norm": 0.19163471460342407,
+      "learning_rate": 0.001,
+      "loss": 2.8881,
+      "step": 7925
+    },
+    {
+      "epoch": 0.33530755563076403,
+      "grad_norm": 0.17413891851902008,
+      "learning_rate": 0.001,
+      "loss": 1.7156,
+      "step": 7926
+    },
+    {
+      "epoch": 0.3353498603942804,
+      "grad_norm": 0.3942449390888214,
+      "learning_rate": 0.001,
+      "loss": 1.8654,
+      "step": 7927
+    },
+    {
+      "epoch": 0.3353921651577968,
+      "grad_norm": 20.90395164489746,
+      "learning_rate": 0.001,
+      "loss": 1.722,
+      "step": 7928
+    },
+    {
+      "epoch": 0.33543446992131315,
+      "grad_norm": 2.177119016647339,
+      "learning_rate": 0.001,
+      "loss": 1.8256,
+      "step": 7929
+    },
+    {
+      "epoch": 0.3354767746848295,
+      "grad_norm": 0.31585386395454407,
+      "learning_rate": 0.001,
+      "loss": 2.4039,
+      "step": 7930
+    },
+    {
+      "epoch": 0.33551907944834586,
+      "grad_norm": 0.3052481710910797,
+      "learning_rate": 0.001,
+      "loss": 2.1275,
+      "step": 7931
+    },
+    {
+      "epoch": 0.33556138421186227,
+      "grad_norm": 0.23809614777565002,
+      "learning_rate": 0.001,
+      "loss": 1.7413,
+      "step": 7932
+    },
+    {
+      "epoch": 0.3356036889753786,
+      "grad_norm": 2.0751454830169678,
+      "learning_rate": 0.001,
+      "loss": 2.132,
+      "step": 7933
+    },
+    {
+      "epoch": 0.335645993738895,
+      "grad_norm": 0.6498151421546936,
+      "learning_rate": 0.001,
+      "loss": 2.0706,
+      "step": 7934
+    },
+    {
+      "epoch": 0.3356882985024114,
+      "grad_norm": 0.26023566722869873,
+      "learning_rate": 0.001,
+      "loss": 2.4264,
+      "step": 7935
+    },
+    {
+      "epoch": 0.33573060326592774,
+      "grad_norm": 0.203081876039505,
+      "learning_rate": 0.001,
+      "loss": 2.4726,
+      "step": 7936
+    },
+    {
+      "epoch": 0.3357729080294441,
+      "grad_norm": 0.2829737961292267,
+      "learning_rate": 0.001,
+      "loss": 3.3616,
+      "step": 7937
+    },
+    {
+      "epoch": 0.3358152127929605,
+      "grad_norm": 1.6824712753295898,
+      "learning_rate": 0.001,
+      "loss": 1.6086,
+      "step": 7938
+    },
+    {
+      "epoch": 0.33585751755647686,
+      "grad_norm": 0.19478239119052887,
+      "learning_rate": 0.001,
+      "loss": 2.579,
+      "step": 7939
+    },
+    {
+      "epoch": 0.3358998223199932,
+      "grad_norm": 0.22429609298706055,
+      "learning_rate": 0.001,
+      "loss": 2.8092,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3359421270835096,
+      "grad_norm": 0.16416104137897491,
+      "learning_rate": 0.001,
+      "loss": 2.2778,
+      "step": 7941
+    },
+    {
+      "epoch": 0.335984431847026,
+      "grad_norm": 0.27525100111961365,
+      "learning_rate": 0.001,
+      "loss": 2.1074,
+      "step": 7942
+    },
+    {
+      "epoch": 0.33602673661054233,
+      "grad_norm": 0.219032421708107,
+      "learning_rate": 0.001,
+      "loss": 2.4779,
+      "step": 7943
+    },
+    {
+      "epoch": 0.33606904137405874,
+      "grad_norm": 0.7257388830184937,
+      "learning_rate": 0.001,
+      "loss": 1.9499,
+      "step": 7944
+    },
+    {
+      "epoch": 0.3361113461375751,
+      "grad_norm": 0.8192842602729797,
+      "learning_rate": 0.001,
+      "loss": 1.9096,
+      "step": 7945
+    },
+    {
+      "epoch": 0.33615365090109145,
+      "grad_norm": 0.1924901306629181,
+      "learning_rate": 0.001,
+      "loss": 1.9039,
+      "step": 7946
+    },
+    {
+      "epoch": 0.33619595566460786,
+      "grad_norm": 0.18553310632705688,
+      "learning_rate": 0.001,
+      "loss": 1.9586,
+      "step": 7947
+    },
+    {
+      "epoch": 0.3362382604281242,
+      "grad_norm": 0.41920140385627747,
+      "learning_rate": 0.001,
+      "loss": 2.1416,
+      "step": 7948
+    },
+    {
+      "epoch": 0.33628056519164057,
+      "grad_norm": 0.21451827883720398,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 7949
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 0.2012207955121994,
+      "learning_rate": 0.001,
+      "loss": 2.3316,
+      "step": 7950
+    },
+    {
+      "epoch": 0.33636517471867333,
+      "grad_norm": 0.1847204715013504,
+      "learning_rate": 0.001,
+      "loss": 2.8531,
+      "step": 7951
+    },
+    {
+      "epoch": 0.3364074794821897,
+      "grad_norm": 0.2382209748029709,
+      "learning_rate": 0.001,
+      "loss": 2.1404,
+      "step": 7952
+    },
+    {
+      "epoch": 0.33644978424570604,
+      "grad_norm": 0.19776912033557892,
+      "learning_rate": 0.001,
+      "loss": 2.2637,
+      "step": 7953
+    },
+    {
+      "epoch": 0.33649208900922245,
+      "grad_norm": 0.3410530388355255,
+      "learning_rate": 0.001,
+      "loss": 2.5895,
+      "step": 7954
+    },
+    {
+      "epoch": 0.3365343937727388,
+      "grad_norm": 0.4076724648475647,
+      "learning_rate": 0.001,
+      "loss": 2.3355,
+      "step": 7955
+    },
+    {
+      "epoch": 0.33657669853625516,
+      "grad_norm": 0.24683041870594025,
+      "learning_rate": 0.001,
+      "loss": 2.0245,
+      "step": 7956
+    },
+    {
+      "epoch": 0.33661900329977157,
+      "grad_norm": 0.1744687408208847,
+      "learning_rate": 0.001,
+      "loss": 2.8672,
+      "step": 7957
+    },
+    {
+      "epoch": 0.3366613080632879,
+      "grad_norm": 0.1615762561559677,
+      "learning_rate": 0.001,
+      "loss": 2.8006,
+      "step": 7958
+    },
+    {
+      "epoch": 0.3367036128268043,
+      "grad_norm": 6.103714942932129,
+      "learning_rate": 0.001,
+      "loss": 2.3797,
+      "step": 7959
+    },
+    {
+      "epoch": 0.3367459175903207,
+      "grad_norm": 0.23423656821250916,
+      "learning_rate": 0.001,
+      "loss": 2.1547,
+      "step": 7960
+    },
+    {
+      "epoch": 0.33678822235383704,
+      "grad_norm": 0.36211827397346497,
+      "learning_rate": 0.001,
+      "loss": 2.6871,
+      "step": 7961
+    },
+    {
+      "epoch": 0.3368305271173534,
+      "grad_norm": 0.2331177294254303,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 7962
+    },
+    {
+      "epoch": 0.3368728318808698,
+      "grad_norm": 0.18622010946273804,
+      "learning_rate": 0.001,
+      "loss": 2.4594,
+      "step": 7963
+    },
+    {
+      "epoch": 0.33691513664438616,
+      "grad_norm": 0.26407673954963684,
+      "learning_rate": 0.001,
+      "loss": 2.1465,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3369574414079025,
+      "grad_norm": 0.20621120929718018,
+      "learning_rate": 0.001,
+      "loss": 1.7513,
+      "step": 7965
+    },
+    {
+      "epoch": 0.3369997461714189,
+      "grad_norm": 0.2361491471529007,
+      "learning_rate": 0.001,
+      "loss": 1.8199,
+      "step": 7966
+    },
+    {
+      "epoch": 0.3370420509349353,
+      "grad_norm": 0.20442907512187958,
+      "learning_rate": 0.001,
+      "loss": 3.1949,
+      "step": 7967
+    },
+    {
+      "epoch": 0.33708435569845163,
+      "grad_norm": 0.23592479526996613,
+      "learning_rate": 0.001,
+      "loss": 1.7339,
+      "step": 7968
+    },
+    {
+      "epoch": 0.33712666046196804,
+      "grad_norm": 19.943355560302734,
+      "learning_rate": 0.001,
+      "loss": 1.6594,
+      "step": 7969
+    },
+    {
+      "epoch": 0.3371689652254844,
+      "grad_norm": 4.2921223640441895,
+      "learning_rate": 0.001,
+      "loss": 2.8318,
+      "step": 7970
+    },
+    {
+      "epoch": 0.33721126998900075,
+      "grad_norm": 0.3829931616783142,
+      "learning_rate": 0.001,
+      "loss": 2.4734,
+      "step": 7971
+    },
+    {
+      "epoch": 0.33725357475251716,
+      "grad_norm": 1.2138727903366089,
+      "learning_rate": 0.001,
+      "loss": 3.164,
+      "step": 7972
+    },
+    {
+      "epoch": 0.3372958795160335,
+      "grad_norm": 0.18562185764312744,
+      "learning_rate": 0.001,
+      "loss": 2.0124,
+      "step": 7973
+    },
+    {
+      "epoch": 0.33733818427954987,
+      "grad_norm": 1.363898754119873,
+      "learning_rate": 0.001,
+      "loss": 2.4757,
+      "step": 7974
+    },
+    {
+      "epoch": 0.3373804890430662,
+      "grad_norm": 0.16648072004318237,
+      "learning_rate": 0.001,
+      "loss": 2.2298,
+      "step": 7975
+    },
+    {
+      "epoch": 0.33742279380658263,
+      "grad_norm": 0.27364203333854675,
+      "learning_rate": 0.001,
+      "loss": 2.1235,
+      "step": 7976
+    },
+    {
+      "epoch": 0.337465098570099,
+      "grad_norm": 0.2676752209663391,
+      "learning_rate": 0.001,
+      "loss": 2.7993,
+      "step": 7977
+    },
+    {
+      "epoch": 0.33750740333361534,
+      "grad_norm": 0.21944323182106018,
+      "learning_rate": 0.001,
+      "loss": 2.7508,
+      "step": 7978
+    },
+    {
+      "epoch": 0.33754970809713175,
+      "grad_norm": 0.18799695372581482,
+      "learning_rate": 0.001,
+      "loss": 1.9509,
+      "step": 7979
+    },
+    {
+      "epoch": 0.3375920128606481,
+      "grad_norm": 0.17753717303276062,
+      "learning_rate": 0.001,
+      "loss": 2.3597,
+      "step": 7980
+    },
+    {
+      "epoch": 0.33763431762416446,
+      "grad_norm": 0.22551682591438293,
+      "learning_rate": 0.001,
+      "loss": 1.6755,
+      "step": 7981
+    },
+    {
+      "epoch": 0.33767662238768087,
+      "grad_norm": 0.21229778230190277,
+      "learning_rate": 0.001,
+      "loss": 2.2954,
+      "step": 7982
+    },
+    {
+      "epoch": 0.3377189271511972,
+      "grad_norm": 0.21653328835964203,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 7983
+    },
+    {
+      "epoch": 0.3377612319147136,
+      "grad_norm": 0.17863360047340393,
+      "learning_rate": 0.001,
+      "loss": 2.0211,
+      "step": 7984
+    },
+    {
+      "epoch": 0.33780353667823,
+      "grad_norm": 0.3650844693183899,
+      "learning_rate": 0.001,
+      "loss": 2.7883,
+      "step": 7985
+    },
+    {
+      "epoch": 0.33784584144174634,
+      "grad_norm": 0.31294921040534973,
+      "learning_rate": 0.001,
+      "loss": 2.2396,
+      "step": 7986
+    },
+    {
+      "epoch": 0.3378881462052627,
+      "grad_norm": 0.8418748378753662,
+      "learning_rate": 0.001,
+      "loss": 1.9066,
+      "step": 7987
+    },
+    {
+      "epoch": 0.3379304509687791,
+      "grad_norm": 1.4910551309585571,
+      "learning_rate": 0.001,
+      "loss": 1.5615,
+      "step": 7988
+    },
+    {
+      "epoch": 0.33797275573229546,
+      "grad_norm": 0.7362372875213623,
+      "learning_rate": 0.001,
+      "loss": 2.7537,
+      "step": 7989
+    },
+    {
+      "epoch": 0.3380150604958118,
+      "grad_norm": 0.27851244807243347,
+      "learning_rate": 0.001,
+      "loss": 2.1922,
+      "step": 7990
+    },
+    {
+      "epoch": 0.3380573652593282,
+      "grad_norm": 0.18572275340557098,
+      "learning_rate": 0.001,
+      "loss": 1.9722,
+      "step": 7991
+    },
+    {
+      "epoch": 0.3380996700228446,
+      "grad_norm": 0.32508203387260437,
+      "learning_rate": 0.001,
+      "loss": 2.7969,
+      "step": 7992
+    },
+    {
+      "epoch": 0.33814197478636093,
+      "grad_norm": 0.22849464416503906,
+      "learning_rate": 0.001,
+      "loss": 1.5679,
+      "step": 7993
+    },
+    {
+      "epoch": 0.33818427954987734,
+      "grad_norm": 0.18476389348506927,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 7994
+    },
+    {
+      "epoch": 0.3382265843133937,
+      "grad_norm": 0.32618439197540283,
+      "learning_rate": 0.001,
+      "loss": 1.9877,
+      "step": 7995
+    },
+    {
+      "epoch": 0.33826888907691005,
+      "grad_norm": 0.20857444405555725,
+      "learning_rate": 0.001,
+      "loss": 1.8781,
+      "step": 7996
+    },
+    {
+      "epoch": 0.3383111938404264,
+      "grad_norm": 0.1851375550031662,
+      "learning_rate": 0.001,
+      "loss": 1.5724,
+      "step": 7997
+    },
+    {
+      "epoch": 0.3383534986039428,
+      "grad_norm": 0.5002090930938721,
+      "learning_rate": 0.001,
+      "loss": 2.1482,
+      "step": 7998
+    },
+    {
+      "epoch": 0.33839580336745917,
+      "grad_norm": 0.18497441709041595,
+      "learning_rate": 0.001,
+      "loss": 2.2619,
+      "step": 7999
+    },
+    {
+      "epoch": 0.3384381081309755,
+      "grad_norm": 0.2192855030298233,
+      "learning_rate": 0.001,
+      "loss": 2.4961,
+      "step": 8000
+    },
+    {
+      "epoch": 0.33848041289449193,
+      "grad_norm": 3.4979212284088135,
+      "learning_rate": 0.001,
+      "loss": 2.0733,
+      "step": 8001
+    },
+    {
+      "epoch": 0.3385227176580083,
+      "grad_norm": 0.2794019877910614,
+      "learning_rate": 0.001,
+      "loss": 2.9049,
+      "step": 8002
+    },
+    {
+      "epoch": 0.33856502242152464,
+      "grad_norm": 0.26850298047065735,
+      "learning_rate": 0.001,
+      "loss": 2.1762,
+      "step": 8003
+    },
+    {
+      "epoch": 0.33860732718504105,
+      "grad_norm": 0.19548161327838898,
+      "learning_rate": 0.001,
+      "loss": 1.9791,
+      "step": 8004
+    },
+    {
+      "epoch": 0.3386496319485574,
+      "grad_norm": 0.20981860160827637,
+      "learning_rate": 0.001,
+      "loss": 2.0878,
+      "step": 8005
+    },
+    {
+      "epoch": 0.33869193671207376,
+      "grad_norm": 0.3214268088340759,
+      "learning_rate": 0.001,
+      "loss": 1.628,
+      "step": 8006
+    },
+    {
+      "epoch": 0.33873424147559017,
+      "grad_norm": 0.2680968940258026,
+      "learning_rate": 0.001,
+      "loss": 1.6191,
+      "step": 8007
+    },
+    {
+      "epoch": 0.3387765462391065,
+      "grad_norm": 5.1096930503845215,
+      "learning_rate": 0.001,
+      "loss": 2.2251,
+      "step": 8008
+    },
+    {
+      "epoch": 0.3388188510026229,
+      "grad_norm": 0.2013072371482849,
+      "learning_rate": 0.001,
+      "loss": 2.0241,
+      "step": 8009
+    },
+    {
+      "epoch": 0.3388611557661393,
+      "grad_norm": 0.2397328019142151,
+      "learning_rate": 0.001,
+      "loss": 1.7325,
+      "step": 8010
+    },
+    {
+      "epoch": 0.33890346052965564,
+      "grad_norm": 0.20630115270614624,
+      "learning_rate": 0.001,
+      "loss": 1.7987,
+      "step": 8011
+    },
+    {
+      "epoch": 0.338945765293172,
+      "grad_norm": 0.3187551200389862,
+      "learning_rate": 0.001,
+      "loss": 2.6951,
+      "step": 8012
+    },
+    {
+      "epoch": 0.3389880700566884,
+      "grad_norm": 0.2622925937175751,
+      "learning_rate": 0.001,
+      "loss": 1.9729,
+      "step": 8013
+    },
+    {
+      "epoch": 0.33903037482020476,
+      "grad_norm": 0.20265188813209534,
+      "learning_rate": 0.001,
+      "loss": 3.4582,
+      "step": 8014
+    },
+    {
+      "epoch": 0.3390726795837211,
+      "grad_norm": 0.44326838850975037,
+      "learning_rate": 0.001,
+      "loss": 2.0444,
+      "step": 8015
+    },
+    {
+      "epoch": 0.3391149843472375,
+      "grad_norm": 6.31563663482666,
+      "learning_rate": 0.001,
+      "loss": 1.9386,
+      "step": 8016
+    },
+    {
+      "epoch": 0.3391572891107539,
+      "grad_norm": 0.7522348761558533,
+      "learning_rate": 0.001,
+      "loss": 2.9402,
+      "step": 8017
+    },
+    {
+      "epoch": 0.33919959387427023,
+      "grad_norm": 0.27529436349868774,
+      "learning_rate": 0.001,
+      "loss": 1.899,
+      "step": 8018
+    },
+    {
+      "epoch": 0.3392418986377866,
+      "grad_norm": 0.304376482963562,
+      "learning_rate": 0.001,
+      "loss": 2.4614,
+      "step": 8019
+    },
+    {
+      "epoch": 0.339284203401303,
+      "grad_norm": 0.2474507987499237,
+      "learning_rate": 0.001,
+      "loss": 2.5785,
+      "step": 8020
+    },
+    {
+      "epoch": 0.33932650816481935,
+      "grad_norm": 0.27842018008232117,
+      "learning_rate": 0.001,
+      "loss": 2.2808,
+      "step": 8021
+    },
+    {
+      "epoch": 0.3393688129283357,
+      "grad_norm": 0.2625916302204132,
+      "learning_rate": 0.001,
+      "loss": 2.1915,
+      "step": 8022
+    },
+    {
+      "epoch": 0.3394111176918521,
+      "grad_norm": 0.2093726396560669,
+      "learning_rate": 0.001,
+      "loss": 2.1816,
+      "step": 8023
+    },
+    {
+      "epoch": 0.33945342245536847,
+      "grad_norm": 0.2079465538263321,
+      "learning_rate": 0.001,
+      "loss": 1.8849,
+      "step": 8024
+    },
+    {
+      "epoch": 0.3394957272188848,
+      "grad_norm": 0.7054566740989685,
+      "learning_rate": 0.001,
+      "loss": 1.8009,
+      "step": 8025
+    },
+    {
+      "epoch": 0.33953803198240123,
+      "grad_norm": 0.24773012101650238,
+      "learning_rate": 0.001,
+      "loss": 3.0748,
+      "step": 8026
+    },
+    {
+      "epoch": 0.3395803367459176,
+      "grad_norm": 0.23611260950565338,
+      "learning_rate": 0.001,
+      "loss": 2.0335,
+      "step": 8027
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 1.1430414915084839,
+      "learning_rate": 0.001,
+      "loss": 3.0462,
+      "step": 8028
+    },
+    {
+      "epoch": 0.33966494627295035,
+      "grad_norm": 0.22795584797859192,
+      "learning_rate": 0.001,
+      "loss": 2.0713,
+      "step": 8029
+    },
+    {
+      "epoch": 0.3397072510364667,
+      "grad_norm": 0.2508586347103119,
+      "learning_rate": 0.001,
+      "loss": 3.3238,
+      "step": 8030
+    },
+    {
+      "epoch": 0.33974955579998306,
+      "grad_norm": 0.24516475200653076,
+      "learning_rate": 0.001,
+      "loss": 1.9439,
+      "step": 8031
+    },
+    {
+      "epoch": 0.33979186056349947,
+      "grad_norm": 0.1642148196697235,
+      "learning_rate": 0.001,
+      "loss": 2.5945,
+      "step": 8032
+    },
+    {
+      "epoch": 0.3398341653270158,
+      "grad_norm": 2.4271435737609863,
+      "learning_rate": 0.001,
+      "loss": 2.6635,
+      "step": 8033
+    },
+    {
+      "epoch": 0.3398764700905322,
+      "grad_norm": 0.8231348991394043,
+      "learning_rate": 0.001,
+      "loss": 2.5651,
+      "step": 8034
+    },
+    {
+      "epoch": 0.3399187748540486,
+      "grad_norm": 0.23157507181167603,
+      "learning_rate": 0.001,
+      "loss": 2.0043,
+      "step": 8035
+    },
+    {
+      "epoch": 0.33996107961756494,
+      "grad_norm": 0.24356773495674133,
+      "learning_rate": 0.001,
+      "loss": 2.0861,
+      "step": 8036
+    },
+    {
+      "epoch": 0.3400033843810813,
+      "grad_norm": 0.23287416994571686,
+      "learning_rate": 0.001,
+      "loss": 2.1728,
+      "step": 8037
+    },
+    {
+      "epoch": 0.3400456891445977,
+      "grad_norm": 0.1699216365814209,
+      "learning_rate": 0.001,
+      "loss": 1.8013,
+      "step": 8038
+    },
+    {
+      "epoch": 0.34008799390811406,
+      "grad_norm": 0.19102227687835693,
+      "learning_rate": 0.001,
+      "loss": 2.2718,
+      "step": 8039
+    },
+    {
+      "epoch": 0.3401302986716304,
+      "grad_norm": 0.2611904442310333,
+      "learning_rate": 0.001,
+      "loss": 3.0925,
+      "step": 8040
+    },
+    {
+      "epoch": 0.34017260343514677,
+      "grad_norm": 0.8404667973518372,
+      "learning_rate": 0.001,
+      "loss": 1.8287,
+      "step": 8041
+    },
+    {
+      "epoch": 0.3402149081986632,
+      "grad_norm": 0.3977033793926239,
+      "learning_rate": 0.001,
+      "loss": 2.1804,
+      "step": 8042
+    },
+    {
+      "epoch": 0.34025721296217953,
+      "grad_norm": 0.21507540345191956,
+      "learning_rate": 0.001,
+      "loss": 2.6397,
+      "step": 8043
+    },
+    {
+      "epoch": 0.3402995177256959,
+      "grad_norm": 0.22415781021118164,
+      "learning_rate": 0.001,
+      "loss": 2.7939,
+      "step": 8044
+    },
+    {
+      "epoch": 0.3403418224892123,
+      "grad_norm": 2.2323434352874756,
+      "learning_rate": 0.001,
+      "loss": 2.3274,
+      "step": 8045
+    },
+    {
+      "epoch": 0.34038412725272865,
+      "grad_norm": 5.563083171844482,
+      "learning_rate": 0.001,
+      "loss": 2.1133,
+      "step": 8046
+    },
+    {
+      "epoch": 0.340426432016245,
+      "grad_norm": 0.27088257670402527,
+      "learning_rate": 0.001,
+      "loss": 2.0518,
+      "step": 8047
+    },
+    {
+      "epoch": 0.3404687367797614,
+      "grad_norm": 1.1499382257461548,
+      "learning_rate": 0.001,
+      "loss": 2.057,
+      "step": 8048
+    },
+    {
+      "epoch": 0.34051104154327777,
+      "grad_norm": 0.2051892727613449,
+      "learning_rate": 0.001,
+      "loss": 2.125,
+      "step": 8049
+    },
+    {
+      "epoch": 0.3405533463067941,
+      "grad_norm": 0.3838144540786743,
+      "learning_rate": 0.001,
+      "loss": 2.1897,
+      "step": 8050
+    },
+    {
+      "epoch": 0.34059565107031053,
+      "grad_norm": 0.28759241104125977,
+      "learning_rate": 0.001,
+      "loss": 1.9844,
+      "step": 8051
+    },
+    {
+      "epoch": 0.3406379558338269,
+      "grad_norm": 0.2020261138677597,
+      "learning_rate": 0.001,
+      "loss": 2.8377,
+      "step": 8052
+    },
+    {
+      "epoch": 0.34068026059734324,
+      "grad_norm": 0.24153375625610352,
+      "learning_rate": 0.001,
+      "loss": 2.1826,
+      "step": 8053
+    },
+    {
+      "epoch": 0.34072256536085965,
+      "grad_norm": 0.43648314476013184,
+      "learning_rate": 0.001,
+      "loss": 2.8549,
+      "step": 8054
+    },
+    {
+      "epoch": 0.340764870124376,
+      "grad_norm": 0.19116896390914917,
+      "learning_rate": 0.001,
+      "loss": 3.6501,
+      "step": 8055
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 7.185187816619873,
+      "learning_rate": 0.001,
+      "loss": 2.6377,
+      "step": 8056
+    },
+    {
+      "epoch": 0.34084947965140877,
+      "grad_norm": 0.18975602090358734,
+      "learning_rate": 0.001,
+      "loss": 2.1213,
+      "step": 8057
+    },
+    {
+      "epoch": 0.3408917844149251,
+      "grad_norm": 0.2663729190826416,
+      "learning_rate": 0.001,
+      "loss": 2.6091,
+      "step": 8058
+    },
+    {
+      "epoch": 0.3409340891784415,
+      "grad_norm": 0.20250354707241058,
+      "learning_rate": 0.001,
+      "loss": 2.103,
+      "step": 8059
+    },
+    {
+      "epoch": 0.3409763939419579,
+      "grad_norm": 0.20965036749839783,
+      "learning_rate": 0.001,
+      "loss": 3.2814,
+      "step": 8060
+    },
+    {
+      "epoch": 0.34101869870547424,
+      "grad_norm": 0.2456333339214325,
+      "learning_rate": 0.001,
+      "loss": 3.3578,
+      "step": 8061
+    },
+    {
+      "epoch": 0.3410610034689906,
+      "grad_norm": 0.18389590084552765,
+      "learning_rate": 0.001,
+      "loss": 2.5921,
+      "step": 8062
+    },
+    {
+      "epoch": 0.341103308232507,
+      "grad_norm": 0.23083733022212982,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 8063
+    },
+    {
+      "epoch": 0.34114561299602336,
+      "grad_norm": 0.6850119829177856,
+      "learning_rate": 0.001,
+      "loss": 1.5674,
+      "step": 8064
+    },
+    {
+      "epoch": 0.3411879177595397,
+      "grad_norm": 0.2017279416322708,
+      "learning_rate": 0.001,
+      "loss": 2.0718,
+      "step": 8065
+    },
+    {
+      "epoch": 0.34123022252305607,
+      "grad_norm": 0.18653610348701477,
+      "learning_rate": 0.001,
+      "loss": 2.4179,
+      "step": 8066
+    },
+    {
+      "epoch": 0.3412725272865725,
+      "grad_norm": 0.3980189263820648,
+      "learning_rate": 0.001,
+      "loss": 2.1241,
+      "step": 8067
+    },
+    {
+      "epoch": 0.34131483205008883,
+      "grad_norm": 0.7478682398796082,
+      "learning_rate": 0.001,
+      "loss": 1.8917,
+      "step": 8068
+    },
+    {
+      "epoch": 0.3413571368136052,
+      "grad_norm": 0.1957026720046997,
+      "learning_rate": 0.001,
+      "loss": 2.3097,
+      "step": 8069
+    },
+    {
+      "epoch": 0.3413994415771216,
+      "grad_norm": 0.1947687864303589,
+      "learning_rate": 0.001,
+      "loss": 1.8799,
+      "step": 8070
+    },
+    {
+      "epoch": 0.34144174634063795,
+      "grad_norm": 0.216701477766037,
+      "learning_rate": 0.001,
+      "loss": 3.1663,
+      "step": 8071
+    },
+    {
+      "epoch": 0.3414840511041543,
+      "grad_norm": 0.2218198925256729,
+      "learning_rate": 0.001,
+      "loss": 2.1391,
+      "step": 8072
+    },
+    {
+      "epoch": 0.3415263558676707,
+      "grad_norm": 0.27254411578178406,
+      "learning_rate": 0.001,
+      "loss": 2.6124,
+      "step": 8073
+    },
+    {
+      "epoch": 0.34156866063118707,
+      "grad_norm": 0.27166926860809326,
+      "learning_rate": 0.001,
+      "loss": 1.6906,
+      "step": 8074
+    },
+    {
+      "epoch": 0.3416109653947034,
+      "grad_norm": 0.2239091545343399,
+      "learning_rate": 0.001,
+      "loss": 1.8461,
+      "step": 8075
+    },
+    {
+      "epoch": 0.34165327015821984,
+      "grad_norm": 0.8380588889122009,
+      "learning_rate": 0.001,
+      "loss": 2.1581,
+      "step": 8076
+    },
+    {
+      "epoch": 0.3416955749217362,
+      "grad_norm": 0.39979785680770874,
+      "learning_rate": 0.001,
+      "loss": 2.3237,
+      "step": 8077
+    },
+    {
+      "epoch": 0.34173787968525254,
+      "grad_norm": 0.25030001997947693,
+      "learning_rate": 0.001,
+      "loss": 1.9951,
+      "step": 8078
+    },
+    {
+      "epoch": 0.34178018444876895,
+      "grad_norm": 0.22300481796264648,
+      "learning_rate": 0.001,
+      "loss": 2.2102,
+      "step": 8079
+    },
+    {
+      "epoch": 0.3418224892122853,
+      "grad_norm": 0.1912737488746643,
+      "learning_rate": 0.001,
+      "loss": 2.1506,
+      "step": 8080
+    },
+    {
+      "epoch": 0.34186479397580166,
+      "grad_norm": 2.661966562271118,
+      "learning_rate": 0.001,
+      "loss": 1.5773,
+      "step": 8081
+    },
+    {
+      "epoch": 0.34190709873931807,
+      "grad_norm": 0.2479306310415268,
+      "learning_rate": 0.001,
+      "loss": 2.8477,
+      "step": 8082
+    },
+    {
+      "epoch": 0.3419494035028344,
+      "grad_norm": 0.43052566051483154,
+      "learning_rate": 0.001,
+      "loss": 3.5951,
+      "step": 8083
+    },
+    {
+      "epoch": 0.3419917082663508,
+      "grad_norm": 0.2106366604566574,
+      "learning_rate": 0.001,
+      "loss": 2.7114,
+      "step": 8084
+    },
+    {
+      "epoch": 0.3420340130298672,
+      "grad_norm": 1.2527594566345215,
+      "learning_rate": 0.001,
+      "loss": 2.2984,
+      "step": 8085
+    },
+    {
+      "epoch": 0.34207631779338354,
+      "grad_norm": 4.545886993408203,
+      "learning_rate": 0.001,
+      "loss": 1.636,
+      "step": 8086
+    },
+    {
+      "epoch": 0.3421186225568999,
+      "grad_norm": 0.3907563090324402,
+      "learning_rate": 0.001,
+      "loss": 3.0252,
+      "step": 8087
+    },
+    {
+      "epoch": 0.34216092732041625,
+      "grad_norm": 0.24621452391147614,
+      "learning_rate": 0.001,
+      "loss": 1.6867,
+      "step": 8088
+    },
+    {
+      "epoch": 0.34220323208393266,
+      "grad_norm": 0.26306939125061035,
+      "learning_rate": 0.001,
+      "loss": 1.9305,
+      "step": 8089
+    },
+    {
+      "epoch": 0.342245536847449,
+      "grad_norm": 0.26752984523773193,
+      "learning_rate": 0.001,
+      "loss": 2.902,
+      "step": 8090
+    },
+    {
+      "epoch": 0.34228784161096537,
+      "grad_norm": 0.2583836019039154,
+      "learning_rate": 0.001,
+      "loss": 3.4137,
+      "step": 8091
+    },
+    {
+      "epoch": 0.3423301463744818,
+      "grad_norm": 1.7589455842971802,
+      "learning_rate": 0.001,
+      "loss": 2.0823,
+      "step": 8092
+    },
+    {
+      "epoch": 0.34237245113799814,
+      "grad_norm": 3.6824162006378174,
+      "learning_rate": 0.001,
+      "loss": 2.4873,
+      "step": 8093
+    },
+    {
+      "epoch": 0.3424147559015145,
+      "grad_norm": 0.45065632462501526,
+      "learning_rate": 0.001,
+      "loss": 2.184,
+      "step": 8094
+    },
+    {
+      "epoch": 0.3424570606650309,
+      "grad_norm": 0.22970075905323029,
+      "learning_rate": 0.001,
+      "loss": 1.886,
+      "step": 8095
+    },
+    {
+      "epoch": 0.34249936542854725,
+      "grad_norm": 0.2680737376213074,
+      "learning_rate": 0.001,
+      "loss": 1.9278,
+      "step": 8096
+    },
+    {
+      "epoch": 0.3425416701920636,
+      "grad_norm": 0.39559051394462585,
+      "learning_rate": 0.001,
+      "loss": 2.6823,
+      "step": 8097
+    },
+    {
+      "epoch": 0.34258397495558,
+      "grad_norm": 0.23845720291137695,
+      "learning_rate": 0.001,
+      "loss": 2.6721,
+      "step": 8098
+    },
+    {
+      "epoch": 0.34262627971909637,
+      "grad_norm": 0.6681344509124756,
+      "learning_rate": 0.001,
+      "loss": 3.0333,
+      "step": 8099
+    },
+    {
+      "epoch": 0.3426685844826127,
+      "grad_norm": 0.935767650604248,
+      "learning_rate": 0.001,
+      "loss": 3.4422,
+      "step": 8100
+    },
+    {
+      "epoch": 0.34271088924612914,
+      "grad_norm": 0.21342290937900543,
+      "learning_rate": 0.001,
+      "loss": 2.4125,
+      "step": 8101
+    },
+    {
+      "epoch": 0.3427531940096455,
+      "grad_norm": 0.7626002430915833,
+      "learning_rate": 0.001,
+      "loss": 3.0488,
+      "step": 8102
+    },
+    {
+      "epoch": 0.34279549877316184,
+      "grad_norm": 0.991487979888916,
+      "learning_rate": 0.001,
+      "loss": 2.2147,
+      "step": 8103
+    },
+    {
+      "epoch": 0.34283780353667825,
+      "grad_norm": 0.19054241478443146,
+      "learning_rate": 0.001,
+      "loss": 1.7796,
+      "step": 8104
+    },
+    {
+      "epoch": 0.3428801083001946,
+      "grad_norm": 2.831486701965332,
+      "learning_rate": 0.001,
+      "loss": 2.9814,
+      "step": 8105
+    },
+    {
+      "epoch": 0.34292241306371096,
+      "grad_norm": 0.27352792024612427,
+      "learning_rate": 0.001,
+      "loss": 1.866,
+      "step": 8106
+    },
+    {
+      "epoch": 0.34296471782722737,
+      "grad_norm": 0.24455174803733826,
+      "learning_rate": 0.001,
+      "loss": 2.2868,
+      "step": 8107
+    },
+    {
+      "epoch": 0.3430070225907437,
+      "grad_norm": 0.21650566160678864,
+      "learning_rate": 0.001,
+      "loss": 2.3255,
+      "step": 8108
+    },
+    {
+      "epoch": 0.3430493273542601,
+      "grad_norm": 0.18031521141529083,
+      "learning_rate": 0.001,
+      "loss": 2.1578,
+      "step": 8109
+    },
+    {
+      "epoch": 0.34309163211777643,
+      "grad_norm": 0.18996316194534302,
+      "learning_rate": 0.001,
+      "loss": 1.8919,
+      "step": 8110
+    },
+    {
+      "epoch": 0.34313393688129284,
+      "grad_norm": 0.18360304832458496,
+      "learning_rate": 0.001,
+      "loss": 1.61,
+      "step": 8111
+    },
+    {
+      "epoch": 0.3431762416448092,
+      "grad_norm": 0.44647422432899475,
+      "learning_rate": 0.001,
+      "loss": 2.3315,
+      "step": 8112
+    },
+    {
+      "epoch": 0.34321854640832555,
+      "grad_norm": 0.937113344669342,
+      "learning_rate": 0.001,
+      "loss": 3.0726,
+      "step": 8113
+    },
+    {
+      "epoch": 0.34326085117184196,
+      "grad_norm": 0.2121163010597229,
+      "learning_rate": 0.001,
+      "loss": 2.0622,
+      "step": 8114
+    },
+    {
+      "epoch": 0.3433031559353583,
+      "grad_norm": 0.21113400161266327,
+      "learning_rate": 0.001,
+      "loss": 2.1562,
+      "step": 8115
+    },
+    {
+      "epoch": 0.34334546069887467,
+      "grad_norm": 0.22805431485176086,
+      "learning_rate": 0.001,
+      "loss": 1.9647,
+      "step": 8116
+    },
+    {
+      "epoch": 0.3433877654623911,
+      "grad_norm": 0.18042448163032532,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 8117
+    },
+    {
+      "epoch": 0.34343007022590744,
+      "grad_norm": 0.33136382699012756,
+      "learning_rate": 0.001,
+      "loss": 3.6987,
+      "step": 8118
+    },
+    {
+      "epoch": 0.3434723749894238,
+      "grad_norm": 0.865990936756134,
+      "learning_rate": 0.001,
+      "loss": 2.1322,
+      "step": 8119
+    },
+    {
+      "epoch": 0.3435146797529402,
+      "grad_norm": 0.2049923986196518,
+      "learning_rate": 0.001,
+      "loss": 2.1151,
+      "step": 8120
+    },
+    {
+      "epoch": 0.34355698451645655,
+      "grad_norm": 0.41140925884246826,
+      "learning_rate": 0.001,
+      "loss": 2.1526,
+      "step": 8121
+    },
+    {
+      "epoch": 0.3435992892799729,
+      "grad_norm": 0.2289661467075348,
+      "learning_rate": 0.001,
+      "loss": 2.3287,
+      "step": 8122
+    },
+    {
+      "epoch": 0.3436415940434893,
+      "grad_norm": 0.8169209957122803,
+      "learning_rate": 0.001,
+      "loss": 3.6942,
+      "step": 8123
+    },
+    {
+      "epoch": 0.34368389880700567,
+      "grad_norm": 0.21063531935214996,
+      "learning_rate": 0.001,
+      "loss": 2.1003,
+      "step": 8124
+    },
+    {
+      "epoch": 0.343726203570522,
+      "grad_norm": 0.20649611949920654,
+      "learning_rate": 0.001,
+      "loss": 1.8405,
+      "step": 8125
+    },
+    {
+      "epoch": 0.34376850833403844,
+      "grad_norm": 0.20465020835399628,
+      "learning_rate": 0.001,
+      "loss": 2.3314,
+      "step": 8126
+    },
+    {
+      "epoch": 0.3438108130975548,
+      "grad_norm": 0.20881281793117523,
+      "learning_rate": 0.001,
+      "loss": 2.8126,
+      "step": 8127
+    },
+    {
+      "epoch": 0.34385311786107114,
+      "grad_norm": 0.45468512177467346,
+      "learning_rate": 0.001,
+      "loss": 2.2817,
+      "step": 8128
+    },
+    {
+      "epoch": 0.34389542262458755,
+      "grad_norm": 0.4071137011051178,
+      "learning_rate": 0.001,
+      "loss": 2.9757,
+      "step": 8129
+    },
+    {
+      "epoch": 0.3439377273881039,
+      "grad_norm": 1.8788225650787354,
+      "learning_rate": 0.001,
+      "loss": 3.8213,
+      "step": 8130
+    },
+    {
+      "epoch": 0.34398003215162026,
+      "grad_norm": 0.18839234113693237,
+      "learning_rate": 0.001,
+      "loss": 2.2812,
+      "step": 8131
+    },
+    {
+      "epoch": 0.3440223369151366,
+      "grad_norm": 2.463320016860962,
+      "learning_rate": 0.001,
+      "loss": 2.9136,
+      "step": 8132
+    },
+    {
+      "epoch": 0.344064641678653,
+      "grad_norm": 0.2248656004667282,
+      "learning_rate": 0.001,
+      "loss": 3.0937,
+      "step": 8133
+    },
+    {
+      "epoch": 0.3441069464421694,
+      "grad_norm": 1.793042540550232,
+      "learning_rate": 0.001,
+      "loss": 3.3005,
+      "step": 8134
+    },
+    {
+      "epoch": 0.34414925120568574,
+      "grad_norm": 0.17570704221725464,
+      "learning_rate": 0.001,
+      "loss": 3.2353,
+      "step": 8135
+    },
+    {
+      "epoch": 0.34419155596920215,
+      "grad_norm": 1.5945035219192505,
+      "learning_rate": 0.001,
+      "loss": 1.9676,
+      "step": 8136
+    },
+    {
+      "epoch": 0.3442338607327185,
+      "grad_norm": 0.18541212379932404,
+      "learning_rate": 0.001,
+      "loss": 2.5174,
+      "step": 8137
+    },
+    {
+      "epoch": 0.34427616549623485,
+      "grad_norm": 0.1724303811788559,
+      "learning_rate": 0.001,
+      "loss": 2.6184,
+      "step": 8138
+    },
+    {
+      "epoch": 0.34431847025975126,
+      "grad_norm": 0.9634104371070862,
+      "learning_rate": 0.001,
+      "loss": 2.0628,
+      "step": 8139
+    },
+    {
+      "epoch": 0.3443607750232676,
+      "grad_norm": 0.23390017449855804,
+      "learning_rate": 0.001,
+      "loss": 2.6646,
+      "step": 8140
+    },
+    {
+      "epoch": 0.34440307978678397,
+      "grad_norm": 0.19771607220172882,
+      "learning_rate": 0.001,
+      "loss": 2.321,
+      "step": 8141
+    },
+    {
+      "epoch": 0.3444453845503004,
+      "grad_norm": 0.21361044049263,
+      "learning_rate": 0.001,
+      "loss": 1.8798,
+      "step": 8142
+    },
+    {
+      "epoch": 0.34448768931381674,
+      "grad_norm": 0.20966002345085144,
+      "learning_rate": 0.001,
+      "loss": 1.9735,
+      "step": 8143
+    },
+    {
+      "epoch": 0.3445299940773331,
+      "grad_norm": 0.24883241951465607,
+      "learning_rate": 0.001,
+      "loss": 2.6098,
+      "step": 8144
+    },
+    {
+      "epoch": 0.3445722988408495,
+      "grad_norm": 1.5156461000442505,
+      "learning_rate": 0.001,
+      "loss": 1.8247,
+      "step": 8145
+    },
+    {
+      "epoch": 0.34461460360436585,
+      "grad_norm": 0.18546625971794128,
+      "learning_rate": 0.001,
+      "loss": 1.8604,
+      "step": 8146
+    },
+    {
+      "epoch": 0.3446569083678822,
+      "grad_norm": 0.18728595972061157,
+      "learning_rate": 0.001,
+      "loss": 2.6729,
+      "step": 8147
+    },
+    {
+      "epoch": 0.3446992131313986,
+      "grad_norm": 0.17871029675006866,
+      "learning_rate": 0.001,
+      "loss": 2.6175,
+      "step": 8148
+    },
+    {
+      "epoch": 0.344741517894915,
+      "grad_norm": 0.1993844360113144,
+      "learning_rate": 0.001,
+      "loss": 2.3318,
+      "step": 8149
+    },
+    {
+      "epoch": 0.3447838226584313,
+      "grad_norm": 0.3129980266094208,
+      "learning_rate": 0.001,
+      "loss": 2.1132,
+      "step": 8150
+    },
+    {
+      "epoch": 0.34482612742194774,
+      "grad_norm": 0.2431798130273819,
+      "learning_rate": 0.001,
+      "loss": 1.7761,
+      "step": 8151
+    },
+    {
+      "epoch": 0.3448684321854641,
+      "grad_norm": 0.33781230449676514,
+      "learning_rate": 0.001,
+      "loss": 4.0398,
+      "step": 8152
+    },
+    {
+      "epoch": 0.34491073694898045,
+      "grad_norm": 1.1244258880615234,
+      "learning_rate": 0.001,
+      "loss": 3.1731,
+      "step": 8153
+    },
+    {
+      "epoch": 0.3449530417124968,
+      "grad_norm": 0.1972734034061432,
+      "learning_rate": 0.001,
+      "loss": 3.1912,
+      "step": 8154
+    },
+    {
+      "epoch": 0.3449953464760132,
+      "grad_norm": 0.19920934736728668,
+      "learning_rate": 0.001,
+      "loss": 2.0368,
+      "step": 8155
+    },
+    {
+      "epoch": 0.34503765123952956,
+      "grad_norm": 0.2313353568315506,
+      "learning_rate": 0.001,
+      "loss": 1.7179,
+      "step": 8156
+    },
+    {
+      "epoch": 0.3450799560030459,
+      "grad_norm": 0.3713974356651306,
+      "learning_rate": 0.001,
+      "loss": 2.0866,
+      "step": 8157
+    },
+    {
+      "epoch": 0.3451222607665623,
+      "grad_norm": 0.21588067710399628,
+      "learning_rate": 0.001,
+      "loss": 1.7316,
+      "step": 8158
+    },
+    {
+      "epoch": 0.3451645655300787,
+      "grad_norm": 0.3257901668548584,
+      "learning_rate": 0.001,
+      "loss": 1.9901,
+      "step": 8159
+    },
+    {
+      "epoch": 0.34520687029359504,
+      "grad_norm": 0.21106848120689392,
+      "learning_rate": 0.001,
+      "loss": 3.0461,
+      "step": 8160
+    },
+    {
+      "epoch": 0.34524917505711145,
+      "grad_norm": 0.6497496962547302,
+      "learning_rate": 0.001,
+      "loss": 3.1958,
+      "step": 8161
+    },
+    {
+      "epoch": 0.3452914798206278,
+      "grad_norm": 0.17610898613929749,
+      "learning_rate": 0.001,
+      "loss": 2.2529,
+      "step": 8162
+    },
+    {
+      "epoch": 0.34533378458414415,
+      "grad_norm": 0.19626934826374054,
+      "learning_rate": 0.001,
+      "loss": 1.6347,
+      "step": 8163
+    },
+    {
+      "epoch": 0.34537608934766056,
+      "grad_norm": 0.21596650779247284,
+      "learning_rate": 0.001,
+      "loss": 3.0951,
+      "step": 8164
+    },
+    {
+      "epoch": 0.3454183941111769,
+      "grad_norm": 0.635787308216095,
+      "learning_rate": 0.001,
+      "loss": 2.1464,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3454606988746933,
+      "grad_norm": 0.17580640316009521,
+      "learning_rate": 0.001,
+      "loss": 1.8562,
+      "step": 8166
+    },
+    {
+      "epoch": 0.3455030036382097,
+      "grad_norm": 0.35619786381721497,
+      "learning_rate": 0.001,
+      "loss": 2.0581,
+      "step": 8167
+    },
+    {
+      "epoch": 0.34554530840172604,
+      "grad_norm": 0.19072014093399048,
+      "learning_rate": 0.001,
+      "loss": 2.9908,
+      "step": 8168
+    },
+    {
+      "epoch": 0.3455876131652424,
+      "grad_norm": 0.21359659731388092,
+      "learning_rate": 0.001,
+      "loss": 1.9315,
+      "step": 8169
+    },
+    {
+      "epoch": 0.3456299179287588,
+      "grad_norm": 0.197892963886261,
+      "learning_rate": 0.001,
+      "loss": 2.3538,
+      "step": 8170
+    },
+    {
+      "epoch": 0.34567222269227516,
+      "grad_norm": 0.18982310593128204,
+      "learning_rate": 0.001,
+      "loss": 2.0689,
+      "step": 8171
+    },
+    {
+      "epoch": 0.3457145274557915,
+      "grad_norm": 0.19009721279144287,
+      "learning_rate": 0.001,
+      "loss": 2.123,
+      "step": 8172
+    },
+    {
+      "epoch": 0.3457568322193079,
+      "grad_norm": 0.5255012512207031,
+      "learning_rate": 0.001,
+      "loss": 2.5506,
+      "step": 8173
+    },
+    {
+      "epoch": 0.3457991369828243,
+      "grad_norm": 1.7404595613479614,
+      "learning_rate": 0.001,
+      "loss": 2.4197,
+      "step": 8174
+    },
+    {
+      "epoch": 0.3458414417463406,
+      "grad_norm": 0.2141122967004776,
+      "learning_rate": 0.001,
+      "loss": 2.2982,
+      "step": 8175
+    },
+    {
+      "epoch": 0.34588374650985704,
+      "grad_norm": 0.21013054251670837,
+      "learning_rate": 0.001,
+      "loss": 3.148,
+      "step": 8176
+    },
+    {
+      "epoch": 0.3459260512733734,
+      "grad_norm": 0.2299475222826004,
+      "learning_rate": 0.001,
+      "loss": 2.7599,
+      "step": 8177
+    },
+    {
+      "epoch": 0.34596835603688975,
+      "grad_norm": 0.735609769821167,
+      "learning_rate": 0.001,
+      "loss": 1.6265,
+      "step": 8178
+    },
+    {
+      "epoch": 0.3460106608004061,
+      "grad_norm": 0.3593994677066803,
+      "learning_rate": 0.001,
+      "loss": 2.6944,
+      "step": 8179
+    },
+    {
+      "epoch": 0.3460529655639225,
+      "grad_norm": 0.22775153815746307,
+      "learning_rate": 0.001,
+      "loss": 2.2529,
+      "step": 8180
+    },
+    {
+      "epoch": 0.34609527032743886,
+      "grad_norm": 0.2560899257659912,
+      "learning_rate": 0.001,
+      "loss": 1.9796,
+      "step": 8181
+    },
+    {
+      "epoch": 0.3461375750909552,
+      "grad_norm": 0.20973028242588043,
+      "learning_rate": 0.001,
+      "loss": 2.9659,
+      "step": 8182
+    },
+    {
+      "epoch": 0.34617987985447163,
+      "grad_norm": 2.2904813289642334,
+      "learning_rate": 0.001,
+      "loss": 1.5693,
+      "step": 8183
+    },
+    {
+      "epoch": 0.346222184617988,
+      "grad_norm": 0.2068173885345459,
+      "learning_rate": 0.001,
+      "loss": 2.2342,
+      "step": 8184
+    },
+    {
+      "epoch": 0.34626448938150434,
+      "grad_norm": 0.6102269291877747,
+      "learning_rate": 0.001,
+      "loss": 2.1191,
+      "step": 8185
+    },
+    {
+      "epoch": 0.34630679414502075,
+      "grad_norm": 0.23791219294071198,
+      "learning_rate": 0.001,
+      "loss": 1.873,
+      "step": 8186
+    },
+    {
+      "epoch": 0.3463490989085371,
+      "grad_norm": 0.2418316751718521,
+      "learning_rate": 0.001,
+      "loss": 2.0798,
+      "step": 8187
+    },
+    {
+      "epoch": 0.34639140367205346,
+      "grad_norm": 0.944127082824707,
+      "learning_rate": 0.001,
+      "loss": 2.1407,
+      "step": 8188
+    },
+    {
+      "epoch": 0.34643370843556986,
+      "grad_norm": 0.20907090604305267,
+      "learning_rate": 0.001,
+      "loss": 2.5714,
+      "step": 8189
+    },
+    {
+      "epoch": 0.3464760131990862,
+      "grad_norm": 0.26245564222335815,
+      "learning_rate": 0.001,
+      "loss": 2.214,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3465183179626026,
+      "grad_norm": 0.2782506048679352,
+      "learning_rate": 0.001,
+      "loss": 1.9815,
+      "step": 8191
+    },
+    {
+      "epoch": 0.346560622726119,
+      "grad_norm": 0.2448754757642746,
+      "learning_rate": 0.001,
+      "loss": 2.5002,
+      "step": 8192
+    },
+    {
+      "epoch": 0.34660292748963534,
+      "grad_norm": 0.21422387659549713,
+      "learning_rate": 0.001,
+      "loss": 2.2284,
+      "step": 8193
+    },
+    {
+      "epoch": 0.3466452322531517,
+      "grad_norm": 0.19360801577568054,
+      "learning_rate": 0.001,
+      "loss": 1.8096,
+      "step": 8194
+    },
+    {
+      "epoch": 0.3466875370166681,
+      "grad_norm": 0.17713648080825806,
+      "learning_rate": 0.001,
+      "loss": 2.179,
+      "step": 8195
+    },
+    {
+      "epoch": 0.34672984178018446,
+      "grad_norm": 0.23182488977909088,
+      "learning_rate": 0.001,
+      "loss": 1.9,
+      "step": 8196
+    },
+    {
+      "epoch": 0.3467721465437008,
+      "grad_norm": 0.6592618227005005,
+      "learning_rate": 0.001,
+      "loss": 2.6549,
+      "step": 8197
+    },
+    {
+      "epoch": 0.3468144513072172,
+      "grad_norm": 0.20647495985031128,
+      "learning_rate": 0.001,
+      "loss": 2.7395,
+      "step": 8198
+    },
+    {
+      "epoch": 0.3468567560707336,
+      "grad_norm": 18.073009490966797,
+      "learning_rate": 0.001,
+      "loss": 2.7044,
+      "step": 8199
+    },
+    {
+      "epoch": 0.34689906083424993,
+      "grad_norm": 0.1915636509656906,
+      "learning_rate": 0.001,
+      "loss": 2.1236,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3469413655977663,
+      "grad_norm": 0.18017853796482086,
+      "learning_rate": 0.001,
+      "loss": 1.8282,
+      "step": 8201
+    },
+    {
+      "epoch": 0.3469836703612827,
+      "grad_norm": 1.731067419052124,
+      "learning_rate": 0.001,
+      "loss": 2.718,
+      "step": 8202
+    },
+    {
+      "epoch": 0.34702597512479905,
+      "grad_norm": 0.2168511152267456,
+      "learning_rate": 0.001,
+      "loss": 2.0306,
+      "step": 8203
+    },
+    {
+      "epoch": 0.3470682798883154,
+      "grad_norm": 0.23651044070720673,
+      "learning_rate": 0.001,
+      "loss": 2.4044,
+      "step": 8204
+    },
+    {
+      "epoch": 0.3471105846518318,
+      "grad_norm": 0.2720677852630615,
+      "learning_rate": 0.001,
+      "loss": 1.8696,
+      "step": 8205
+    },
+    {
+      "epoch": 0.34715288941534816,
+      "grad_norm": 0.24406278133392334,
+      "learning_rate": 0.001,
+      "loss": 1.9719,
+      "step": 8206
+    },
+    {
+      "epoch": 0.3471951941788645,
+      "grad_norm": 0.2830585241317749,
+      "learning_rate": 0.001,
+      "loss": 2.2669,
+      "step": 8207
+    },
+    {
+      "epoch": 0.34723749894238093,
+      "grad_norm": 0.9110187292098999,
+      "learning_rate": 0.001,
+      "loss": 2.46,
+      "step": 8208
+    },
+    {
+      "epoch": 0.3472798037058973,
+      "grad_norm": 0.23988114297389984,
+      "learning_rate": 0.001,
+      "loss": 2.1952,
+      "step": 8209
+    },
+    {
+      "epoch": 0.34732210846941364,
+      "grad_norm": 0.23413024842739105,
+      "learning_rate": 0.001,
+      "loss": 1.7244,
+      "step": 8210
+    },
+    {
+      "epoch": 0.34736441323293005,
+      "grad_norm": 0.17257529497146606,
+      "learning_rate": 0.001,
+      "loss": 1.3189,
+      "step": 8211
+    },
+    {
+      "epoch": 0.3474067179964464,
+      "grad_norm": 0.3271627128124237,
+      "learning_rate": 0.001,
+      "loss": 2.999,
+      "step": 8212
+    },
+    {
+      "epoch": 0.34744902275996276,
+      "grad_norm": 0.4803444445133209,
+      "learning_rate": 0.001,
+      "loss": 3.5374,
+      "step": 8213
+    },
+    {
+      "epoch": 0.34749132752347917,
+      "grad_norm": 0.20940083265304565,
+      "learning_rate": 0.001,
+      "loss": 3.9058,
+      "step": 8214
+    },
+    {
+      "epoch": 0.3475336322869955,
+      "grad_norm": 2.5996439456939697,
+      "learning_rate": 0.001,
+      "loss": 2.0939,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3475759370505119,
+      "grad_norm": 3.1301825046539307,
+      "learning_rate": 0.001,
+      "loss": 3.033,
+      "step": 8216
+    },
+    {
+      "epoch": 0.3476182418140283,
+      "grad_norm": 0.2768004834651947,
+      "learning_rate": 0.001,
+      "loss": 2.4367,
+      "step": 8217
+    },
+    {
+      "epoch": 0.34766054657754464,
+      "grad_norm": 0.2278338372707367,
+      "learning_rate": 0.001,
+      "loss": 2.7675,
+      "step": 8218
+    },
+    {
+      "epoch": 0.347702851341061,
+      "grad_norm": 59.52205276489258,
+      "learning_rate": 0.001,
+      "loss": 1.8877,
+      "step": 8219
+    },
+    {
+      "epoch": 0.3477451561045774,
+      "grad_norm": 0.4094898998737335,
+      "learning_rate": 0.001,
+      "loss": 3.3673,
+      "step": 8220
+    },
+    {
+      "epoch": 0.34778746086809376,
+      "grad_norm": 0.24076704680919647,
+      "learning_rate": 0.001,
+      "loss": 1.7683,
+      "step": 8221
+    },
+    {
+      "epoch": 0.3478297656316101,
+      "grad_norm": 0.2221417874097824,
+      "learning_rate": 0.001,
+      "loss": 2.8931,
+      "step": 8222
+    },
+    {
+      "epoch": 0.34787207039512646,
+      "grad_norm": 0.2613305151462555,
+      "learning_rate": 0.001,
+      "loss": 2.6144,
+      "step": 8223
+    },
+    {
+      "epoch": 0.3479143751586429,
+      "grad_norm": 0.22772428393363953,
+      "learning_rate": 0.001,
+      "loss": 2.4166,
+      "step": 8224
+    },
+    {
+      "epoch": 0.34795667992215923,
+      "grad_norm": 0.21340018510818481,
+      "learning_rate": 0.001,
+      "loss": 1.5321,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3479989846856756,
+      "grad_norm": 0.24845480918884277,
+      "learning_rate": 0.001,
+      "loss": 1.9605,
+      "step": 8226
+    },
+    {
+      "epoch": 0.348041289449192,
+      "grad_norm": 0.24793021380901337,
+      "learning_rate": 0.001,
+      "loss": 3.3195,
+      "step": 8227
+    },
+    {
+      "epoch": 0.34808359421270835,
+      "grad_norm": 3.903813123703003,
+      "learning_rate": 0.001,
+      "loss": 2.8211,
+      "step": 8228
+    },
+    {
+      "epoch": 0.3481258989762247,
+      "grad_norm": 0.21152375638484955,
+      "learning_rate": 0.001,
+      "loss": 2.4791,
+      "step": 8229
+    },
+    {
+      "epoch": 0.3481682037397411,
+      "grad_norm": 0.8900184631347656,
+      "learning_rate": 0.001,
+      "loss": 2.0313,
+      "step": 8230
+    },
+    {
+      "epoch": 0.34821050850325747,
+      "grad_norm": 0.2712019383907318,
+      "learning_rate": 0.001,
+      "loss": 2.3294,
+      "step": 8231
+    },
+    {
+      "epoch": 0.3482528132667738,
+      "grad_norm": 5.551436424255371,
+      "learning_rate": 0.001,
+      "loss": 2.4235,
+      "step": 8232
+    },
+    {
+      "epoch": 0.34829511803029023,
+      "grad_norm": 0.28290316462516785,
+      "learning_rate": 0.001,
+      "loss": 2.4363,
+      "step": 8233
+    },
+    {
+      "epoch": 0.3483374227938066,
+      "grad_norm": 0.7481885552406311,
+      "learning_rate": 0.001,
+      "loss": 1.7503,
+      "step": 8234
+    },
+    {
+      "epoch": 0.34837972755732294,
+      "grad_norm": 0.2814173400402069,
+      "learning_rate": 0.001,
+      "loss": 3.2661,
+      "step": 8235
+    },
+    {
+      "epoch": 0.34842203232083935,
+      "grad_norm": 1.8952068090438843,
+      "learning_rate": 0.001,
+      "loss": 3.6599,
+      "step": 8236
+    },
+    {
+      "epoch": 0.3484643370843557,
+      "grad_norm": 1.625000238418579,
+      "learning_rate": 0.001,
+      "loss": 3.2106,
+      "step": 8237
+    },
+    {
+      "epoch": 0.34850664184787206,
+      "grad_norm": 0.29325634241104126,
+      "learning_rate": 0.001,
+      "loss": 1.8171,
+      "step": 8238
+    },
+    {
+      "epoch": 0.34854894661138847,
+      "grad_norm": 0.23248402774333954,
+      "learning_rate": 0.001,
+      "loss": 1.9535,
+      "step": 8239
+    },
+    {
+      "epoch": 0.3485912513749048,
+      "grad_norm": 0.36317363381385803,
+      "learning_rate": 0.001,
+      "loss": 2.1191,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3486335561384212,
+      "grad_norm": 7.670083522796631,
+      "learning_rate": 0.001,
+      "loss": 2.581,
+      "step": 8241
+    },
+    {
+      "epoch": 0.3486758609019376,
+      "grad_norm": 0.2423803061246872,
+      "learning_rate": 0.001,
+      "loss": 2.7115,
+      "step": 8242
+    },
+    {
+      "epoch": 0.34871816566545394,
+      "grad_norm": 0.21461081504821777,
+      "learning_rate": 0.001,
+      "loss": 2.5597,
+      "step": 8243
+    },
+    {
+      "epoch": 0.3487604704289703,
+      "grad_norm": 0.28708726167678833,
+      "learning_rate": 0.001,
+      "loss": 3.2198,
+      "step": 8244
+    },
+    {
+      "epoch": 0.34880277519248665,
+      "grad_norm": 6.893510341644287,
+      "learning_rate": 0.001,
+      "loss": 2.521,
+      "step": 8245
+    },
+    {
+      "epoch": 0.34884507995600306,
+      "grad_norm": 0.19617153704166412,
+      "learning_rate": 0.001,
+      "loss": 1.9453,
+      "step": 8246
+    },
+    {
+      "epoch": 0.3488873847195194,
+      "grad_norm": 0.20679624378681183,
+      "learning_rate": 0.001,
+      "loss": 2.4901,
+      "step": 8247
+    },
+    {
+      "epoch": 0.34892968948303577,
+      "grad_norm": 0.28653019666671753,
+      "learning_rate": 0.001,
+      "loss": 2.3008,
+      "step": 8248
+    },
+    {
+      "epoch": 0.3489719942465522,
+      "grad_norm": 0.7451533675193787,
+      "learning_rate": 0.001,
+      "loss": 3.0218,
+      "step": 8249
+    },
+    {
+      "epoch": 0.34901429901006853,
+      "grad_norm": 0.3160172998905182,
+      "learning_rate": 0.001,
+      "loss": 2.5261,
+      "step": 8250
+    },
+    {
+      "epoch": 0.3490566037735849,
+      "grad_norm": 0.20035721361637115,
+      "learning_rate": 0.001,
+      "loss": 2.3146,
+      "step": 8251
+    },
+    {
+      "epoch": 0.3490989085371013,
+      "grad_norm": 0.19196628034114838,
+      "learning_rate": 0.001,
+      "loss": 3.0744,
+      "step": 8252
+    },
+    {
+      "epoch": 0.34914121330061765,
+      "grad_norm": 0.2677168846130371,
+      "learning_rate": 0.001,
+      "loss": 2.6725,
+      "step": 8253
+    },
+    {
+      "epoch": 0.349183518064134,
+      "grad_norm": 0.210031196475029,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 8254
+    },
+    {
+      "epoch": 0.3492258228276504,
+      "grad_norm": 0.2728026211261749,
+      "learning_rate": 0.001,
+      "loss": 1.6813,
+      "step": 8255
+    },
+    {
+      "epoch": 0.34926812759116677,
+      "grad_norm": 0.2557034492492676,
+      "learning_rate": 0.001,
+      "loss": 2.1864,
+      "step": 8256
+    },
+    {
+      "epoch": 0.3493104323546831,
+      "grad_norm": 0.7835190892219543,
+      "learning_rate": 0.001,
+      "loss": 3.1137,
+      "step": 8257
+    },
+    {
+      "epoch": 0.34935273711819953,
+      "grad_norm": 0.28121432662010193,
+      "learning_rate": 0.001,
+      "loss": 2.9322,
+      "step": 8258
+    },
+    {
+      "epoch": 0.3493950418817159,
+      "grad_norm": 0.207332044839859,
+      "learning_rate": 0.001,
+      "loss": 2.1749,
+      "step": 8259
+    },
+    {
+      "epoch": 0.34943734664523224,
+      "grad_norm": 0.5446605682373047,
+      "learning_rate": 0.001,
+      "loss": 2.0835,
+      "step": 8260
+    },
+    {
+      "epoch": 0.34947965140874865,
+      "grad_norm": 0.17694032192230225,
+      "learning_rate": 0.001,
+      "loss": 1.9013,
+      "step": 8261
+    },
+    {
+      "epoch": 0.349521956172265,
+      "grad_norm": 0.546602725982666,
+      "learning_rate": 0.001,
+      "loss": 2.1906,
+      "step": 8262
+    },
+    {
+      "epoch": 0.34956426093578136,
+      "grad_norm": 0.26002398133277893,
+      "learning_rate": 0.001,
+      "loss": 2.5833,
+      "step": 8263
+    },
+    {
+      "epoch": 0.34960656569929777,
+      "grad_norm": 0.21473874151706696,
+      "learning_rate": 0.001,
+      "loss": 1.9894,
+      "step": 8264
+    },
+    {
+      "epoch": 0.3496488704628141,
+      "grad_norm": 0.22110356390476227,
+      "learning_rate": 0.001,
+      "loss": 1.663,
+      "step": 8265
+    },
+    {
+      "epoch": 0.3496911752263305,
+      "grad_norm": 0.8616834878921509,
+      "learning_rate": 0.001,
+      "loss": 2.6967,
+      "step": 8266
+    },
+    {
+      "epoch": 0.34973347998984683,
+      "grad_norm": 0.2699930965900421,
+      "learning_rate": 0.001,
+      "loss": 3.416,
+      "step": 8267
+    },
+    {
+      "epoch": 0.34977578475336324,
+      "grad_norm": 0.27595359086990356,
+      "learning_rate": 0.001,
+      "loss": 1.9483,
+      "step": 8268
+    },
+    {
+      "epoch": 0.3498180895168796,
+      "grad_norm": 0.9030215740203857,
+      "learning_rate": 0.001,
+      "loss": 1.9612,
+      "step": 8269
+    },
+    {
+      "epoch": 0.34986039428039595,
+      "grad_norm": 0.21707811951637268,
+      "learning_rate": 0.001,
+      "loss": 1.8743,
+      "step": 8270
+    },
+    {
+      "epoch": 0.34990269904391236,
+      "grad_norm": 0.22298790514469147,
+      "learning_rate": 0.001,
+      "loss": 2.299,
+      "step": 8271
+    },
+    {
+      "epoch": 0.3499450038074287,
+      "grad_norm": 0.4610617756843567,
+      "learning_rate": 0.001,
+      "loss": 2.5806,
+      "step": 8272
+    },
+    {
+      "epoch": 0.34998730857094507,
+      "grad_norm": 0.16975651681423187,
+      "learning_rate": 0.001,
+      "loss": 1.8847,
+      "step": 8273
+    },
+    {
+      "epoch": 0.3500296133344615,
+      "grad_norm": 0.1778561770915985,
+      "learning_rate": 0.001,
+      "loss": 1.9668,
+      "step": 8274
+    },
+    {
+      "epoch": 0.35007191809797783,
+      "grad_norm": 0.1916327327489853,
+      "learning_rate": 0.001,
+      "loss": 2.3812,
+      "step": 8275
+    },
+    {
+      "epoch": 0.3501142228614942,
+      "grad_norm": 0.18574245274066925,
+      "learning_rate": 0.001,
+      "loss": 2.6713,
+      "step": 8276
+    },
+    {
+      "epoch": 0.3501565276250106,
+      "grad_norm": 0.17055678367614746,
+      "learning_rate": 0.001,
+      "loss": 2.7099,
+      "step": 8277
+    },
+    {
+      "epoch": 0.35019883238852695,
+      "grad_norm": 3.7420198917388916,
+      "learning_rate": 0.001,
+      "loss": 2.2635,
+      "step": 8278
+    },
+    {
+      "epoch": 0.3502411371520433,
+      "grad_norm": 0.19405874609947205,
+      "learning_rate": 0.001,
+      "loss": 2.8745,
+      "step": 8279
+    },
+    {
+      "epoch": 0.3502834419155597,
+      "grad_norm": 0.43192586302757263,
+      "learning_rate": 0.001,
+      "loss": 2.0631,
+      "step": 8280
+    },
+    {
+      "epoch": 0.35032574667907607,
+      "grad_norm": 0.19240808486938477,
+      "learning_rate": 0.001,
+      "loss": 2.2849,
+      "step": 8281
+    },
+    {
+      "epoch": 0.3503680514425924,
+      "grad_norm": 0.17617124319076538,
+      "learning_rate": 0.001,
+      "loss": 1.8595,
+      "step": 8282
+    },
+    {
+      "epoch": 0.35041035620610883,
+      "grad_norm": 4.70356559753418,
+      "learning_rate": 0.001,
+      "loss": 2.568,
+      "step": 8283
+    },
+    {
+      "epoch": 0.3504526609696252,
+      "grad_norm": 0.2642245292663574,
+      "learning_rate": 0.001,
+      "loss": 3.0254,
+      "step": 8284
+    },
+    {
+      "epoch": 0.35049496573314154,
+      "grad_norm": 0.1709575653076172,
+      "learning_rate": 0.001,
+      "loss": 2.6415,
+      "step": 8285
+    },
+    {
+      "epoch": 0.35053727049665795,
+      "grad_norm": 0.2094145566225052,
+      "learning_rate": 0.001,
+      "loss": 2.4383,
+      "step": 8286
+    },
+    {
+      "epoch": 0.3505795752601743,
+      "grad_norm": 0.18870730698108673,
+      "learning_rate": 0.001,
+      "loss": 1.6139,
+      "step": 8287
+    },
+    {
+      "epoch": 0.35062188002369066,
+      "grad_norm": 0.23713693022727966,
+      "learning_rate": 0.001,
+      "loss": 1.9205,
+      "step": 8288
+    },
+    {
+      "epoch": 0.35066418478720707,
+      "grad_norm": 0.1588599532842636,
+      "learning_rate": 0.001,
+      "loss": 1.5372,
+      "step": 8289
+    },
+    {
+      "epoch": 0.3507064895507234,
+      "grad_norm": 7.636360168457031,
+      "learning_rate": 0.001,
+      "loss": 2.0058,
+      "step": 8290
+    },
+    {
+      "epoch": 0.3507487943142398,
+      "grad_norm": 0.8418156504631042,
+      "learning_rate": 0.001,
+      "loss": 2.6797,
+      "step": 8291
+    },
+    {
+      "epoch": 0.35079109907775613,
+      "grad_norm": 0.2791912853717804,
+      "learning_rate": 0.001,
+      "loss": 3.1777,
+      "step": 8292
+    },
+    {
+      "epoch": 0.35083340384127254,
+      "grad_norm": 0.22210825979709625,
+      "learning_rate": 0.001,
+      "loss": 2.0927,
+      "step": 8293
+    },
+    {
+      "epoch": 0.3508757086047889,
+      "grad_norm": 2.1510884761810303,
+      "learning_rate": 0.001,
+      "loss": 2.077,
+      "step": 8294
+    },
+    {
+      "epoch": 0.35091801336830525,
+      "grad_norm": 0.21856360137462616,
+      "learning_rate": 0.001,
+      "loss": 2.7959,
+      "step": 8295
+    },
+    {
+      "epoch": 0.35096031813182166,
+      "grad_norm": 0.2667442560195923,
+      "learning_rate": 0.001,
+      "loss": 2.2033,
+      "step": 8296
+    },
+    {
+      "epoch": 0.351002622895338,
+      "grad_norm": 0.24852122366428375,
+      "learning_rate": 0.001,
+      "loss": 2.1141,
+      "step": 8297
+    },
+    {
+      "epoch": 0.35104492765885437,
+      "grad_norm": 1.397088885307312,
+      "learning_rate": 0.001,
+      "loss": 1.803,
+      "step": 8298
+    },
+    {
+      "epoch": 0.3510872324223708,
+      "grad_norm": 17.205541610717773,
+      "learning_rate": 0.001,
+      "loss": 2.371,
+      "step": 8299
+    },
+    {
+      "epoch": 0.35112953718588713,
+      "grad_norm": 0.2123159021139145,
+      "learning_rate": 0.001,
+      "loss": 3.3105,
+      "step": 8300
+    },
+    {
+      "epoch": 0.3511718419494035,
+      "grad_norm": 2.099641799926758,
+      "learning_rate": 0.001,
+      "loss": 2.1354,
+      "step": 8301
+    },
+    {
+      "epoch": 0.3512141467129199,
+      "grad_norm": 6.61326789855957,
+      "learning_rate": 0.001,
+      "loss": 2.2409,
+      "step": 8302
+    },
+    {
+      "epoch": 0.35125645147643625,
+      "grad_norm": 2.718135118484497,
+      "learning_rate": 0.001,
+      "loss": 2.2279,
+      "step": 8303
+    },
+    {
+      "epoch": 0.3512987562399526,
+      "grad_norm": 1.8631720542907715,
+      "learning_rate": 0.001,
+      "loss": 3.8234,
+      "step": 8304
+    },
+    {
+      "epoch": 0.351341061003469,
+      "grad_norm": 20.797183990478516,
+      "learning_rate": 0.001,
+      "loss": 3.4096,
+      "step": 8305
+    },
+    {
+      "epoch": 0.35138336576698537,
+      "grad_norm": 0.22008393704891205,
+      "learning_rate": 0.001,
+      "loss": 1.8032,
+      "step": 8306
+    },
+    {
+      "epoch": 0.3514256705305017,
+      "grad_norm": 0.26470401883125305,
+      "learning_rate": 0.001,
+      "loss": 2.6094,
+      "step": 8307
+    },
+    {
+      "epoch": 0.35146797529401813,
+      "grad_norm": 0.19725409150123596,
+      "learning_rate": 0.001,
+      "loss": 1.5124,
+      "step": 8308
+    },
+    {
+      "epoch": 0.3515102800575345,
+      "grad_norm": 0.3234120011329651,
+      "learning_rate": 0.001,
+      "loss": 2.1466,
+      "step": 8309
+    },
+    {
+      "epoch": 0.35155258482105084,
+      "grad_norm": 0.32287687063217163,
+      "learning_rate": 0.001,
+      "loss": 2.0659,
+      "step": 8310
+    },
+    {
+      "epoch": 0.35159488958456725,
+      "grad_norm": 0.2961645722389221,
+      "learning_rate": 0.001,
+      "loss": 2.6533,
+      "step": 8311
+    },
+    {
+      "epoch": 0.3516371943480836,
+      "grad_norm": 2.627986192703247,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 8312
+    },
+    {
+      "epoch": 0.35167949911159996,
+      "grad_norm": 0.374530166387558,
+      "learning_rate": 0.001,
+      "loss": 2.9414,
+      "step": 8313
+    },
+    {
+      "epoch": 0.3517218038751163,
+      "grad_norm": 0.19504281878471375,
+      "learning_rate": 0.001,
+      "loss": 1.7939,
+      "step": 8314
+    },
+    {
+      "epoch": 0.3517641086386327,
+      "grad_norm": 0.6185072064399719,
+      "learning_rate": 0.001,
+      "loss": 2.1054,
+      "step": 8315
+    },
+    {
+      "epoch": 0.3518064134021491,
+      "grad_norm": 0.4495874345302582,
+      "learning_rate": 0.001,
+      "loss": 1.7725,
+      "step": 8316
+    },
+    {
+      "epoch": 0.35184871816566543,
+      "grad_norm": 0.2591138780117035,
+      "learning_rate": 0.001,
+      "loss": 2.4438,
+      "step": 8317
+    },
+    {
+      "epoch": 0.35189102292918184,
+      "grad_norm": 0.2160710096359253,
+      "learning_rate": 0.001,
+      "loss": 1.9199,
+      "step": 8318
+    },
+    {
+      "epoch": 0.3519333276926982,
+      "grad_norm": 0.22822026908397675,
+      "learning_rate": 0.001,
+      "loss": 2.8699,
+      "step": 8319
+    },
+    {
+      "epoch": 0.35197563245621455,
+      "grad_norm": 0.23242418467998505,
+      "learning_rate": 0.001,
+      "loss": 2.9319,
+      "step": 8320
+    },
+    {
+      "epoch": 0.35201793721973096,
+      "grad_norm": 59.025177001953125,
+      "learning_rate": 0.001,
+      "loss": 2.2009,
+      "step": 8321
+    },
+    {
+      "epoch": 0.3520602419832473,
+      "grad_norm": 0.3726635277271271,
+      "learning_rate": 0.001,
+      "loss": 1.9791,
+      "step": 8322
+    },
+    {
+      "epoch": 0.35210254674676367,
+      "grad_norm": 0.2542991638183594,
+      "learning_rate": 0.001,
+      "loss": 3.2745,
+      "step": 8323
+    },
+    {
+      "epoch": 0.3521448515102801,
+      "grad_norm": 0.25025397539138794,
+      "learning_rate": 0.001,
+      "loss": 1.8405,
+      "step": 8324
+    },
+    {
+      "epoch": 0.35218715627379643,
+      "grad_norm": 0.2506003677845001,
+      "learning_rate": 0.001,
+      "loss": 3.0609,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3522294610373128,
+      "grad_norm": 0.26360347867012024,
+      "learning_rate": 0.001,
+      "loss": 2.1514,
+      "step": 8326
+    },
+    {
+      "epoch": 0.3522717658008292,
+      "grad_norm": 0.23163963854312897,
+      "learning_rate": 0.001,
+      "loss": 2.0993,
+      "step": 8327
+    },
+    {
+      "epoch": 0.35231407056434555,
+      "grad_norm": 18.21137046813965,
+      "learning_rate": 0.001,
+      "loss": 2.8873,
+      "step": 8328
+    },
+    {
+      "epoch": 0.3523563753278619,
+      "grad_norm": 0.21554243564605713,
+      "learning_rate": 0.001,
+      "loss": 2.7229,
+      "step": 8329
+    },
+    {
+      "epoch": 0.3523986800913783,
+      "grad_norm": 0.26712241768836975,
+      "learning_rate": 0.001,
+      "loss": 2.5294,
+      "step": 8330
+    },
+    {
+      "epoch": 0.35244098485489467,
+      "grad_norm": 0.2301950454711914,
+      "learning_rate": 0.001,
+      "loss": 2.3371,
+      "step": 8331
+    },
+    {
+      "epoch": 0.352483289618411,
+      "grad_norm": 0.19002433121204376,
+      "learning_rate": 0.001,
+      "loss": 2.637,
+      "step": 8332
+    },
+    {
+      "epoch": 0.35252559438192743,
+      "grad_norm": 0.25531867146492004,
+      "learning_rate": 0.001,
+      "loss": 2.001,
+      "step": 8333
+    },
+    {
+      "epoch": 0.3525678991454438,
+      "grad_norm": 0.9937705397605896,
+      "learning_rate": 0.001,
+      "loss": 1.9568,
+      "step": 8334
+    },
+    {
+      "epoch": 0.35261020390896014,
+      "grad_norm": 0.3505503833293915,
+      "learning_rate": 0.001,
+      "loss": 2.2035,
+      "step": 8335
+    },
+    {
+      "epoch": 0.3526525086724765,
+      "grad_norm": 0.2179388403892517,
+      "learning_rate": 0.001,
+      "loss": 2.0458,
+      "step": 8336
+    },
+    {
+      "epoch": 0.3526948134359929,
+      "grad_norm": 0.17621107399463654,
+      "learning_rate": 0.001,
+      "loss": 1.9847,
+      "step": 8337
+    },
+    {
+      "epoch": 0.35273711819950926,
+      "grad_norm": 0.34067684412002563,
+      "learning_rate": 0.001,
+      "loss": 2.064,
+      "step": 8338
+    },
+    {
+      "epoch": 0.3527794229630256,
+      "grad_norm": 0.24471262097358704,
+      "learning_rate": 0.001,
+      "loss": 3.5153,
+      "step": 8339
+    },
+    {
+      "epoch": 0.352821727726542,
+      "grad_norm": 0.37908583879470825,
+      "learning_rate": 0.001,
+      "loss": 2.4995,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3528640324900584,
+      "grad_norm": 0.453393816947937,
+      "learning_rate": 0.001,
+      "loss": 2.0838,
+      "step": 8341
+    },
+    {
+      "epoch": 0.35290633725357473,
+      "grad_norm": 0.18824003636837006,
+      "learning_rate": 0.001,
+      "loss": 2.6714,
+      "step": 8342
+    },
+    {
+      "epoch": 0.35294864201709114,
+      "grad_norm": 0.16498175263404846,
+      "learning_rate": 0.001,
+      "loss": 1.8878,
+      "step": 8343
+    },
+    {
+      "epoch": 0.3529909467806075,
+      "grad_norm": 0.18404828011989594,
+      "learning_rate": 0.001,
+      "loss": 2.5525,
+      "step": 8344
+    },
+    {
+      "epoch": 0.35303325154412385,
+      "grad_norm": 0.19402121007442474,
+      "learning_rate": 0.001,
+      "loss": 1.3525,
+      "step": 8345
+    },
+    {
+      "epoch": 0.35307555630764026,
+      "grad_norm": 0.3609767258167267,
+      "learning_rate": 0.001,
+      "loss": 2.2392,
+      "step": 8346
+    },
+    {
+      "epoch": 0.3531178610711566,
+      "grad_norm": 0.17331457138061523,
+      "learning_rate": 0.001,
+      "loss": 1.9563,
+      "step": 8347
+    },
+    {
+      "epoch": 0.35316016583467297,
+      "grad_norm": 0.1629885584115982,
+      "learning_rate": 0.001,
+      "loss": 2.2316,
+      "step": 8348
+    },
+    {
+      "epoch": 0.3532024705981894,
+      "grad_norm": 0.4724940061569214,
+      "learning_rate": 0.001,
+      "loss": 2.8375,
+      "step": 8349
+    },
+    {
+      "epoch": 0.35324477536170573,
+      "grad_norm": 0.19430530071258545,
+      "learning_rate": 0.001,
+      "loss": 2.5702,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3532870801252221,
+      "grad_norm": 0.2884927988052368,
+      "learning_rate": 0.001,
+      "loss": 1.8518,
+      "step": 8351
+    },
+    {
+      "epoch": 0.3533293848887385,
+      "grad_norm": 0.30345895886421204,
+      "learning_rate": 0.001,
+      "loss": 2.2031,
+      "step": 8352
+    },
+    {
+      "epoch": 0.35337168965225485,
+      "grad_norm": 0.1875777542591095,
+      "learning_rate": 0.001,
+      "loss": 1.9054,
+      "step": 8353
+    },
+    {
+      "epoch": 0.3534139944157712,
+      "grad_norm": 0.17914628982543945,
+      "learning_rate": 0.001,
+      "loss": 2.1172,
+      "step": 8354
+    },
+    {
+      "epoch": 0.3534562991792876,
+      "grad_norm": 0.223419189453125,
+      "learning_rate": 0.001,
+      "loss": 2.0011,
+      "step": 8355
+    },
+    {
+      "epoch": 0.35349860394280397,
+      "grad_norm": 1.009746789932251,
+      "learning_rate": 0.001,
+      "loss": 3.5386,
+      "step": 8356
+    },
+    {
+      "epoch": 0.3535409087063203,
+      "grad_norm": 0.17974236607551575,
+      "learning_rate": 0.001,
+      "loss": 2.8207,
+      "step": 8357
+    },
+    {
+      "epoch": 0.3535832134698367,
+      "grad_norm": 0.20830687880516052,
+      "learning_rate": 0.001,
+      "loss": 2.6878,
+      "step": 8358
+    },
+    {
+      "epoch": 0.3536255182333531,
+      "grad_norm": 0.19284899532794952,
+      "learning_rate": 0.001,
+      "loss": 1.9765,
+      "step": 8359
+    },
+    {
+      "epoch": 0.35366782299686944,
+      "grad_norm": 1.829329252243042,
+      "learning_rate": 0.001,
+      "loss": 1.6846,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3537101277603858,
+      "grad_norm": 0.21594662964344025,
+      "learning_rate": 0.001,
+      "loss": 1.8569,
+      "step": 8361
+    },
+    {
+      "epoch": 0.3537524325239022,
+      "grad_norm": 4.540489673614502,
+      "learning_rate": 0.001,
+      "loss": 2.6528,
+      "step": 8362
+    },
+    {
+      "epoch": 0.35379473728741856,
+      "grad_norm": 0.24919438362121582,
+      "learning_rate": 0.001,
+      "loss": 3.6536,
+      "step": 8363
+    },
+    {
+      "epoch": 0.3538370420509349,
+      "grad_norm": 0.21543803811073303,
+      "learning_rate": 0.001,
+      "loss": 2.3457,
+      "step": 8364
+    },
+    {
+      "epoch": 0.3538793468144513,
+      "grad_norm": 0.18875446915626526,
+      "learning_rate": 0.001,
+      "loss": 1.6829,
+      "step": 8365
+    },
+    {
+      "epoch": 0.3539216515779677,
+      "grad_norm": 0.18854385614395142,
+      "learning_rate": 0.001,
+      "loss": 2.7787,
+      "step": 8366
+    },
+    {
+      "epoch": 0.35396395634148403,
+      "grad_norm": 0.23087264597415924,
+      "learning_rate": 0.001,
+      "loss": 2.1272,
+      "step": 8367
+    },
+    {
+      "epoch": 0.35400626110500044,
+      "grad_norm": 0.16182763874530792,
+      "learning_rate": 0.001,
+      "loss": 1.4934,
+      "step": 8368
+    },
+    {
+      "epoch": 0.3540485658685168,
+      "grad_norm": 0.19627897441387177,
+      "learning_rate": 0.001,
+      "loss": 2.9024,
+      "step": 8369
+    },
+    {
+      "epoch": 0.35409087063203315,
+      "grad_norm": 0.18524384498596191,
+      "learning_rate": 0.001,
+      "loss": 2.1512,
+      "step": 8370
+    },
+    {
+      "epoch": 0.35413317539554956,
+      "grad_norm": 0.1795620322227478,
+      "learning_rate": 0.001,
+      "loss": 2.0742,
+      "step": 8371
+    },
+    {
+      "epoch": 0.3541754801590659,
+      "grad_norm": 0.20785008370876312,
+      "learning_rate": 0.001,
+      "loss": 1.6379,
+      "step": 8372
+    },
+    {
+      "epoch": 0.35421778492258227,
+      "grad_norm": 0.20255346596240997,
+      "learning_rate": 0.001,
+      "loss": 2.1128,
+      "step": 8373
+    },
+    {
+      "epoch": 0.3542600896860987,
+      "grad_norm": 0.22817464172840118,
+      "learning_rate": 0.001,
+      "loss": 2.0911,
+      "step": 8374
+    },
+    {
+      "epoch": 0.35430239444961503,
+      "grad_norm": 1.5379173755645752,
+      "learning_rate": 0.001,
+      "loss": 2.2441,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3543446992131314,
+      "grad_norm": 1.9879519939422607,
+      "learning_rate": 0.001,
+      "loss": 1.7536,
+      "step": 8376
+    },
+    {
+      "epoch": 0.3543870039766478,
+      "grad_norm": 0.8693563938140869,
+      "learning_rate": 0.001,
+      "loss": 2.0423,
+      "step": 8377
+    },
+    {
+      "epoch": 0.35442930874016415,
+      "grad_norm": 0.20398443937301636,
+      "learning_rate": 0.001,
+      "loss": 1.926,
+      "step": 8378
+    },
+    {
+      "epoch": 0.3544716135036805,
+      "grad_norm": 1.420688509941101,
+      "learning_rate": 0.001,
+      "loss": 2.1442,
+      "step": 8379
+    },
+    {
+      "epoch": 0.35451391826719686,
+      "grad_norm": 8.911663055419922,
+      "learning_rate": 0.001,
+      "loss": 1.6632,
+      "step": 8380
+    },
+    {
+      "epoch": 0.35455622303071327,
+      "grad_norm": 0.2737337052822113,
+      "learning_rate": 0.001,
+      "loss": 1.9327,
+      "step": 8381
+    },
+    {
+      "epoch": 0.3545985277942296,
+      "grad_norm": 0.20059816539287567,
+      "learning_rate": 0.001,
+      "loss": 2.094,
+      "step": 8382
+    },
+    {
+      "epoch": 0.354640832557746,
+      "grad_norm": 0.21529340744018555,
+      "learning_rate": 0.001,
+      "loss": 1.811,
+      "step": 8383
+    },
+    {
+      "epoch": 0.3546831373212624,
+      "grad_norm": 0.21227684617042542,
+      "learning_rate": 0.001,
+      "loss": 1.9585,
+      "step": 8384
+    },
+    {
+      "epoch": 0.35472544208477874,
+      "grad_norm": 0.18570013344287872,
+      "learning_rate": 0.001,
+      "loss": 2.5025,
+      "step": 8385
+    },
+    {
+      "epoch": 0.3547677468482951,
+      "grad_norm": 0.21961857378482819,
+      "learning_rate": 0.001,
+      "loss": 1.7425,
+      "step": 8386
+    },
+    {
+      "epoch": 0.3548100516118115,
+      "grad_norm": 8.055773735046387,
+      "learning_rate": 0.001,
+      "loss": 2.41,
+      "step": 8387
+    },
+    {
+      "epoch": 0.35485235637532786,
+      "grad_norm": 0.33439183235168457,
+      "learning_rate": 0.001,
+      "loss": 2.141,
+      "step": 8388
+    },
+    {
+      "epoch": 0.3548946611388442,
+      "grad_norm": 0.2350430190563202,
+      "learning_rate": 0.001,
+      "loss": 2.0489,
+      "step": 8389
+    },
+    {
+      "epoch": 0.3549369659023606,
+      "grad_norm": 0.2935107946395874,
+      "learning_rate": 0.001,
+      "loss": 1.439,
+      "step": 8390
+    },
+    {
+      "epoch": 0.354979270665877,
+      "grad_norm": 2.349884033203125,
+      "learning_rate": 0.001,
+      "loss": 2.6664,
+      "step": 8391
+    },
+    {
+      "epoch": 0.35502157542939333,
+      "grad_norm": 2.4287772178649902,
+      "learning_rate": 0.001,
+      "loss": 2.796,
+      "step": 8392
+    },
+    {
+      "epoch": 0.35506388019290974,
+      "grad_norm": 0.24385391175746918,
+      "learning_rate": 0.001,
+      "loss": 1.9699,
+      "step": 8393
+    },
+    {
+      "epoch": 0.3551061849564261,
+      "grad_norm": 0.21865463256835938,
+      "learning_rate": 0.001,
+      "loss": 1.4755,
+      "step": 8394
+    },
+    {
+      "epoch": 0.35514848971994245,
+      "grad_norm": 0.221740260720253,
+      "learning_rate": 0.001,
+      "loss": 2.6338,
+      "step": 8395
+    },
+    {
+      "epoch": 0.35519079448345886,
+      "grad_norm": 0.212009459733963,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 8396
+    },
+    {
+      "epoch": 0.3552330992469752,
+      "grad_norm": 0.42118605971336365,
+      "learning_rate": 0.001,
+      "loss": 1.8155,
+      "step": 8397
+    },
+    {
+      "epoch": 0.35527540401049157,
+      "grad_norm": 0.19747601449489594,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 8398
+    },
+    {
+      "epoch": 0.355317708774008,
+      "grad_norm": 0.21434566378593445,
+      "learning_rate": 0.001,
+      "loss": 3.2472,
+      "step": 8399
+    },
+    {
+      "epoch": 0.35536001353752433,
+      "grad_norm": 0.1856713443994522,
+      "learning_rate": 0.001,
+      "loss": 1.8753,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3554023183010407,
+      "grad_norm": 0.215153768658638,
+      "learning_rate": 0.001,
+      "loss": 1.7379,
+      "step": 8401
+    },
+    {
+      "epoch": 0.35544462306455704,
+      "grad_norm": 0.16858762502670288,
+      "learning_rate": 0.001,
+      "loss": 1.8505,
+      "step": 8402
+    },
+    {
+      "epoch": 0.35548692782807345,
+      "grad_norm": 0.23868021368980408,
+      "learning_rate": 0.001,
+      "loss": 3.6215,
+      "step": 8403
+    },
+    {
+      "epoch": 0.3555292325915898,
+      "grad_norm": 0.21436399221420288,
+      "learning_rate": 0.001,
+      "loss": 1.5321,
+      "step": 8404
+    },
+    {
+      "epoch": 0.35557153735510616,
+      "grad_norm": 0.3087001442909241,
+      "learning_rate": 0.001,
+      "loss": 2.6408,
+      "step": 8405
+    },
+    {
+      "epoch": 0.35561384211862257,
+      "grad_norm": 0.20392103493213654,
+      "learning_rate": 0.001,
+      "loss": 3.0366,
+      "step": 8406
+    },
+    {
+      "epoch": 0.3556561468821389,
+      "grad_norm": 1.7128150463104248,
+      "learning_rate": 0.001,
+      "loss": 3.0189,
+      "step": 8407
+    },
+    {
+      "epoch": 0.3556984516456553,
+      "grad_norm": 0.2017560452222824,
+      "learning_rate": 0.001,
+      "loss": 2.2836,
+      "step": 8408
+    },
+    {
+      "epoch": 0.3557407564091717,
+      "grad_norm": 0.2031020075082779,
+      "learning_rate": 0.001,
+      "loss": 2.3527,
+      "step": 8409
+    },
+    {
+      "epoch": 0.35578306117268804,
+      "grad_norm": 0.19620314240455627,
+      "learning_rate": 0.001,
+      "loss": 2.9028,
+      "step": 8410
+    },
+    {
+      "epoch": 0.3558253659362044,
+      "grad_norm": 0.23667864501476288,
+      "learning_rate": 0.001,
+      "loss": 1.8299,
+      "step": 8411
+    },
+    {
+      "epoch": 0.3558676706997208,
+      "grad_norm": 0.18068136274814606,
+      "learning_rate": 0.001,
+      "loss": 2.8935,
+      "step": 8412
+    },
+    {
+      "epoch": 0.35590997546323716,
+      "grad_norm": 0.9692279696464539,
+      "learning_rate": 0.001,
+      "loss": 2.2835,
+      "step": 8413
+    },
+    {
+      "epoch": 0.3559522802267535,
+      "grad_norm": 0.1930302083492279,
+      "learning_rate": 0.001,
+      "loss": 1.8272,
+      "step": 8414
+    },
+    {
+      "epoch": 0.3559945849902699,
+      "grad_norm": 0.49082469940185547,
+      "learning_rate": 0.001,
+      "loss": 3.0818,
+      "step": 8415
+    },
+    {
+      "epoch": 0.3560368897537863,
+      "grad_norm": 0.17451566457748413,
+      "learning_rate": 0.001,
+      "loss": 1.9013,
+      "step": 8416
+    },
+    {
+      "epoch": 0.35607919451730263,
+      "grad_norm": 0.3995453715324402,
+      "learning_rate": 0.001,
+      "loss": 2.0403,
+      "step": 8417
+    },
+    {
+      "epoch": 0.35612149928081904,
+      "grad_norm": 0.18853165209293365,
+      "learning_rate": 0.001,
+      "loss": 1.9157,
+      "step": 8418
+    },
+    {
+      "epoch": 0.3561638040443354,
+      "grad_norm": 0.16779901087284088,
+      "learning_rate": 0.001,
+      "loss": 2.425,
+      "step": 8419
+    },
+    {
+      "epoch": 0.35620610880785175,
+      "grad_norm": 0.20070673525333405,
+      "learning_rate": 0.001,
+      "loss": 2.0243,
+      "step": 8420
+    },
+    {
+      "epoch": 0.35624841357136816,
+      "grad_norm": 0.26167789101600647,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 8421
+    },
+    {
+      "epoch": 0.3562907183348845,
+      "grad_norm": 0.39837929606437683,
+      "learning_rate": 0.001,
+      "loss": 3.0087,
+      "step": 8422
+    },
+    {
+      "epoch": 0.35633302309840087,
+      "grad_norm": 0.1800961196422577,
+      "learning_rate": 0.001,
+      "loss": 2.4717,
+      "step": 8423
+    },
+    {
+      "epoch": 0.3563753278619173,
+      "grad_norm": 0.2057555615901947,
+      "learning_rate": 0.001,
+      "loss": 2.3249,
+      "step": 8424
+    },
+    {
+      "epoch": 0.35641763262543363,
+      "grad_norm": 0.17443668842315674,
+      "learning_rate": 0.001,
+      "loss": 1.6971,
+      "step": 8425
+    },
+    {
+      "epoch": 0.35645993738895,
+      "grad_norm": 0.21079231798648834,
+      "learning_rate": 0.001,
+      "loss": 2.3975,
+      "step": 8426
+    },
+    {
+      "epoch": 0.35650224215246634,
+      "grad_norm": 3.70257830619812,
+      "learning_rate": 0.001,
+      "loss": 2.5433,
+      "step": 8427
+    },
+    {
+      "epoch": 0.35654454691598275,
+      "grad_norm": 0.17566591501235962,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 8428
+    },
+    {
+      "epoch": 0.3565868516794991,
+      "grad_norm": 0.4361303746700287,
+      "learning_rate": 0.001,
+      "loss": 2.4334,
+      "step": 8429
+    },
+    {
+      "epoch": 0.35662915644301546,
+      "grad_norm": 0.27633586525917053,
+      "learning_rate": 0.001,
+      "loss": 3.0307,
+      "step": 8430
+    },
+    {
+      "epoch": 0.35667146120653187,
+      "grad_norm": 0.21895642578601837,
+      "learning_rate": 0.001,
+      "loss": 2.015,
+      "step": 8431
+    },
+    {
+      "epoch": 0.3567137659700482,
+      "grad_norm": 0.15661796927452087,
+      "learning_rate": 0.001,
+      "loss": 1.8659,
+      "step": 8432
+    },
+    {
+      "epoch": 0.3567560707335646,
+      "grad_norm": 0.19141338765621185,
+      "learning_rate": 0.001,
+      "loss": 2.7277,
+      "step": 8433
+    },
+    {
+      "epoch": 0.356798375497081,
+      "grad_norm": 0.2092185765504837,
+      "learning_rate": 0.001,
+      "loss": 2.2695,
+      "step": 8434
+    },
+    {
+      "epoch": 0.35684068026059734,
+      "grad_norm": 0.19185788929462433,
+      "learning_rate": 0.001,
+      "loss": 2.2118,
+      "step": 8435
+    },
+    {
+      "epoch": 0.3568829850241137,
+      "grad_norm": 0.20594419538974762,
+      "learning_rate": 0.001,
+      "loss": 2.4456,
+      "step": 8436
+    },
+    {
+      "epoch": 0.3569252897876301,
+      "grad_norm": 0.843341588973999,
+      "learning_rate": 0.001,
+      "loss": 2.6461,
+      "step": 8437
+    },
+    {
+      "epoch": 0.35696759455114646,
+      "grad_norm": 0.18174798786640167,
+      "learning_rate": 0.001,
+      "loss": 2.0485,
+      "step": 8438
+    },
+    {
+      "epoch": 0.3570098993146628,
+      "grad_norm": 0.6228585839271545,
+      "learning_rate": 0.001,
+      "loss": 2.8919,
+      "step": 8439
+    },
+    {
+      "epoch": 0.3570522040781792,
+      "grad_norm": 0.21685943007469177,
+      "learning_rate": 0.001,
+      "loss": 2.1506,
+      "step": 8440
+    },
+    {
+      "epoch": 0.3570945088416956,
+      "grad_norm": 0.16078568994998932,
+      "learning_rate": 0.001,
+      "loss": 2.6238,
+      "step": 8441
+    },
+    {
+      "epoch": 0.35713681360521193,
+      "grad_norm": 0.2162536233663559,
+      "learning_rate": 0.001,
+      "loss": 2.6207,
+      "step": 8442
+    },
+    {
+      "epoch": 0.35717911836872834,
+      "grad_norm": 0.20838353037834167,
+      "learning_rate": 0.001,
+      "loss": 2.4904,
+      "step": 8443
+    },
+    {
+      "epoch": 0.3572214231322447,
+      "grad_norm": 0.20165249705314636,
+      "learning_rate": 0.001,
+      "loss": 3.4246,
+      "step": 8444
+    },
+    {
+      "epoch": 0.35726372789576105,
+      "grad_norm": 0.15616677701473236,
+      "learning_rate": 0.001,
+      "loss": 2.2008,
+      "step": 8445
+    },
+    {
+      "epoch": 0.35730603265927746,
+      "grad_norm": 0.48721373081207275,
+      "learning_rate": 0.001,
+      "loss": 1.8402,
+      "step": 8446
+    },
+    {
+      "epoch": 0.3573483374227938,
+      "grad_norm": 0.23544834554195404,
+      "learning_rate": 0.001,
+      "loss": 2.1721,
+      "step": 8447
+    },
+    {
+      "epoch": 0.35739064218631017,
+      "grad_norm": 27.898847579956055,
+      "learning_rate": 0.001,
+      "loss": 2.5454,
+      "step": 8448
+    },
+    {
+      "epoch": 0.3574329469498265,
+      "grad_norm": 0.23525142669677734,
+      "learning_rate": 0.001,
+      "loss": 3.1336,
+      "step": 8449
+    },
+    {
+      "epoch": 0.35747525171334293,
+      "grad_norm": 0.16481630504131317,
+      "learning_rate": 0.001,
+      "loss": 1.871,
+      "step": 8450
+    },
+    {
+      "epoch": 0.3575175564768593,
+      "grad_norm": 0.20045652985572815,
+      "learning_rate": 0.001,
+      "loss": 2.5494,
+      "step": 8451
+    },
+    {
+      "epoch": 0.35755986124037564,
+      "grad_norm": 0.4051211476325989,
+      "learning_rate": 0.001,
+      "loss": 2.1058,
+      "step": 8452
+    },
+    {
+      "epoch": 0.35760216600389205,
+      "grad_norm": 1.801464557647705,
+      "learning_rate": 0.001,
+      "loss": 2.0417,
+      "step": 8453
+    },
+    {
+      "epoch": 0.3576444707674084,
+      "grad_norm": 0.22075815498828888,
+      "learning_rate": 0.001,
+      "loss": 2.2037,
+      "step": 8454
+    },
+    {
+      "epoch": 0.35768677553092476,
+      "grad_norm": 0.20504413545131683,
+      "learning_rate": 0.001,
+      "loss": 2.048,
+      "step": 8455
+    },
+    {
+      "epoch": 0.35772908029444117,
+      "grad_norm": 0.20746447145938873,
+      "learning_rate": 0.001,
+      "loss": 2.2803,
+      "step": 8456
+    },
+    {
+      "epoch": 0.3577713850579575,
+      "grad_norm": 0.47997429966926575,
+      "learning_rate": 0.001,
+      "loss": 2.247,
+      "step": 8457
+    },
+    {
+      "epoch": 0.3578136898214739,
+      "grad_norm": 0.24447347223758698,
+      "learning_rate": 0.001,
+      "loss": 1.9872,
+      "step": 8458
+    },
+    {
+      "epoch": 0.3578559945849903,
+      "grad_norm": 0.2988087832927704,
+      "learning_rate": 0.001,
+      "loss": 3.0362,
+      "step": 8459
+    },
+    {
+      "epoch": 0.35789829934850664,
+      "grad_norm": 1.136828064918518,
+      "learning_rate": 0.001,
+      "loss": 2.3112,
+      "step": 8460
+    },
+    {
+      "epoch": 0.357940604112023,
+      "grad_norm": 0.4372611343860626,
+      "learning_rate": 0.001,
+      "loss": 1.7607,
+      "step": 8461
+    },
+    {
+      "epoch": 0.3579829088755394,
+      "grad_norm": 0.18738792836666107,
+      "learning_rate": 0.001,
+      "loss": 1.9763,
+      "step": 8462
+    },
+    {
+      "epoch": 0.35802521363905576,
+      "grad_norm": 0.4911549985408783,
+      "learning_rate": 0.001,
+      "loss": 3.0154,
+      "step": 8463
+    },
+    {
+      "epoch": 0.3580675184025721,
+      "grad_norm": 0.21335723996162415,
+      "learning_rate": 0.001,
+      "loss": 3.5891,
+      "step": 8464
+    },
+    {
+      "epoch": 0.3581098231660885,
+      "grad_norm": 0.9935727715492249,
+      "learning_rate": 0.001,
+      "loss": 2.0077,
+      "step": 8465
+    },
+    {
+      "epoch": 0.3581521279296049,
+      "grad_norm": 0.16692140698432922,
+      "learning_rate": 0.001,
+      "loss": 1.725,
+      "step": 8466
+    },
+    {
+      "epoch": 0.35819443269312123,
+      "grad_norm": 3.184319496154785,
+      "learning_rate": 0.001,
+      "loss": 1.9069,
+      "step": 8467
+    },
+    {
+      "epoch": 0.35823673745663764,
+      "grad_norm": 0.2190055400133133,
+      "learning_rate": 0.001,
+      "loss": 2.3108,
+      "step": 8468
+    },
+    {
+      "epoch": 0.358279042220154,
+      "grad_norm": 2.8575377464294434,
+      "learning_rate": 0.001,
+      "loss": 1.8183,
+      "step": 8469
+    },
+    {
+      "epoch": 0.35832134698367035,
+      "grad_norm": 0.27034443616867065,
+      "learning_rate": 0.001,
+      "loss": 1.9055,
+      "step": 8470
+    },
+    {
+      "epoch": 0.3583636517471867,
+      "grad_norm": 0.2202761024236679,
+      "learning_rate": 0.001,
+      "loss": 2.4092,
+      "step": 8471
+    },
+    {
+      "epoch": 0.3584059565107031,
+      "grad_norm": 1.0408316850662231,
+      "learning_rate": 0.001,
+      "loss": 2.3927,
+      "step": 8472
+    },
+    {
+      "epoch": 0.35844826127421947,
+      "grad_norm": 0.26200470328330994,
+      "learning_rate": 0.001,
+      "loss": 2.256,
+      "step": 8473
+    },
+    {
+      "epoch": 0.3584905660377358,
+      "grad_norm": 0.23284506797790527,
+      "learning_rate": 0.001,
+      "loss": 2.4034,
+      "step": 8474
+    },
+    {
+      "epoch": 0.35853287080125223,
+      "grad_norm": 0.20646698772907257,
+      "learning_rate": 0.001,
+      "loss": 1.829,
+      "step": 8475
+    },
+    {
+      "epoch": 0.3585751755647686,
+      "grad_norm": 0.25378456711769104,
+      "learning_rate": 0.001,
+      "loss": 2.9985,
+      "step": 8476
+    },
+    {
+      "epoch": 0.35861748032828494,
+      "grad_norm": 0.31622862815856934,
+      "learning_rate": 0.001,
+      "loss": 2.0001,
+      "step": 8477
+    },
+    {
+      "epoch": 0.35865978509180135,
+      "grad_norm": 0.21043705940246582,
+      "learning_rate": 0.001,
+      "loss": 2.5147,
+      "step": 8478
+    },
+    {
+      "epoch": 0.3587020898553177,
+      "grad_norm": 0.3560662865638733,
+      "learning_rate": 0.001,
+      "loss": 2.2411,
+      "step": 8479
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 0.4322994649410248,
+      "learning_rate": 0.001,
+      "loss": 3.4217,
+      "step": 8480
+    },
+    {
+      "epoch": 0.35878669938235047,
+      "grad_norm": 0.7399798631668091,
+      "learning_rate": 0.001,
+      "loss": 2.408,
+      "step": 8481
+    },
+    {
+      "epoch": 0.3588290041458668,
+      "grad_norm": 0.22962254285812378,
+      "learning_rate": 0.001,
+      "loss": 2.7897,
+      "step": 8482
+    },
+    {
+      "epoch": 0.3588713089093832,
+      "grad_norm": 0.20676597952842712,
+      "learning_rate": 0.001,
+      "loss": 2.0482,
+      "step": 8483
+    },
+    {
+      "epoch": 0.3589136136728996,
+      "grad_norm": 0.1842261403799057,
+      "learning_rate": 0.001,
+      "loss": 2.6388,
+      "step": 8484
+    },
+    {
+      "epoch": 0.35895591843641594,
+      "grad_norm": 0.28361448645591736,
+      "learning_rate": 0.001,
+      "loss": 2.6798,
+      "step": 8485
+    },
+    {
+      "epoch": 0.3589982231999323,
+      "grad_norm": 0.910611093044281,
+      "learning_rate": 0.001,
+      "loss": 2.1683,
+      "step": 8486
+    },
+    {
+      "epoch": 0.3590405279634487,
+      "grad_norm": 0.7657113671302795,
+      "learning_rate": 0.001,
+      "loss": 2.3814,
+      "step": 8487
+    },
+    {
+      "epoch": 0.35908283272696506,
+      "grad_norm": 0.29642629623413086,
+      "learning_rate": 0.001,
+      "loss": 1.888,
+      "step": 8488
+    },
+    {
+      "epoch": 0.3591251374904814,
+      "grad_norm": 0.1811065673828125,
+      "learning_rate": 0.001,
+      "loss": 3.645,
+      "step": 8489
+    },
+    {
+      "epoch": 0.3591674422539978,
+      "grad_norm": 0.2280883491039276,
+      "learning_rate": 0.001,
+      "loss": 2.1726,
+      "step": 8490
+    },
+    {
+      "epoch": 0.3592097470175142,
+      "grad_norm": 0.19528071582317352,
+      "learning_rate": 0.001,
+      "loss": 2.1867,
+      "step": 8491
+    },
+    {
+      "epoch": 0.35925205178103053,
+      "grad_norm": 0.24748535454273224,
+      "learning_rate": 0.001,
+      "loss": 2.0032,
+      "step": 8492
+    },
+    {
+      "epoch": 0.3592943565445469,
+      "grad_norm": 0.21031174063682556,
+      "learning_rate": 0.001,
+      "loss": 2.0499,
+      "step": 8493
+    },
+    {
+      "epoch": 0.3593366613080633,
+      "grad_norm": 0.22891457378864288,
+      "learning_rate": 0.001,
+      "loss": 3.2536,
+      "step": 8494
+    },
+    {
+      "epoch": 0.35937896607157965,
+      "grad_norm": 0.18080003559589386,
+      "learning_rate": 0.001,
+      "loss": 2.1618,
+      "step": 8495
+    },
+    {
+      "epoch": 0.359421270835096,
+      "grad_norm": 0.3403954803943634,
+      "learning_rate": 0.001,
+      "loss": 1.625,
+      "step": 8496
+    },
+    {
+      "epoch": 0.3594635755986124,
+      "grad_norm": 0.1686491221189499,
+      "learning_rate": 0.001,
+      "loss": 3.029,
+      "step": 8497
+    },
+    {
+      "epoch": 0.35950588036212877,
+      "grad_norm": 0.3256143629550934,
+      "learning_rate": 0.001,
+      "loss": 2.1193,
+      "step": 8498
+    },
+    {
+      "epoch": 0.3595481851256451,
+      "grad_norm": 7.151230335235596,
+      "learning_rate": 0.001,
+      "loss": 3.4917,
+      "step": 8499
+    },
+    {
+      "epoch": 0.35959048988916154,
+      "grad_norm": 0.2594936788082123,
+      "learning_rate": 0.001,
+      "loss": 2.7958,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3596327946526779,
+      "grad_norm": 1.2051527500152588,
+      "learning_rate": 0.001,
+      "loss": 2.0397,
+      "step": 8501
+    },
+    {
+      "epoch": 0.35967509941619424,
+      "grad_norm": 4.79257345199585,
+      "learning_rate": 0.001,
+      "loss": 2.5444,
+      "step": 8502
+    },
+    {
+      "epoch": 0.35971740417971065,
+      "grad_norm": 0.1828813999891281,
+      "learning_rate": 0.001,
+      "loss": 2.5241,
+      "step": 8503
+    },
+    {
+      "epoch": 0.359759708943227,
+      "grad_norm": 0.22619538009166718,
+      "learning_rate": 0.001,
+      "loss": 2.5545,
+      "step": 8504
+    },
+    {
+      "epoch": 0.35980201370674336,
+      "grad_norm": 0.16107860207557678,
+      "learning_rate": 0.001,
+      "loss": 2.2485,
+      "step": 8505
+    },
+    {
+      "epoch": 0.35984431847025977,
+      "grad_norm": 0.22678741812705994,
+      "learning_rate": 0.001,
+      "loss": 2.5525,
+      "step": 8506
+    },
+    {
+      "epoch": 0.3598866232337761,
+      "grad_norm": 0.20380182564258575,
+      "learning_rate": 0.001,
+      "loss": 2.495,
+      "step": 8507
+    },
+    {
+      "epoch": 0.3599289279972925,
+      "grad_norm": 0.32149720191955566,
+      "learning_rate": 0.001,
+      "loss": 2.2425,
+      "step": 8508
+    },
+    {
+      "epoch": 0.3599712327608089,
+      "grad_norm": 0.17898282408714294,
+      "learning_rate": 0.001,
+      "loss": 2.3266,
+      "step": 8509
+    },
+    {
+      "epoch": 0.36001353752432524,
+      "grad_norm": 0.742031455039978,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 8510
+    },
+    {
+      "epoch": 0.3600558422878416,
+      "grad_norm": 1.0595115423202515,
+      "learning_rate": 0.001,
+      "loss": 2.9696,
+      "step": 8511
+    },
+    {
+      "epoch": 0.360098147051358,
+      "grad_norm": 0.22999782860279083,
+      "learning_rate": 0.001,
+      "loss": 3.4851,
+      "step": 8512
+    },
+    {
+      "epoch": 0.36014045181487436,
+      "grad_norm": 0.2157551646232605,
+      "learning_rate": 0.001,
+      "loss": 2.5969,
+      "step": 8513
+    },
+    {
+      "epoch": 0.3601827565783907,
+      "grad_norm": 0.8261212706565857,
+      "learning_rate": 0.001,
+      "loss": 2.3355,
+      "step": 8514
+    },
+    {
+      "epoch": 0.36022506134190707,
+      "grad_norm": 0.40030571818351746,
+      "learning_rate": 0.001,
+      "loss": 2.4083,
+      "step": 8515
+    },
+    {
+      "epoch": 0.3602673661054235,
+      "grad_norm": 0.23546592891216278,
+      "learning_rate": 0.001,
+      "loss": 2.1861,
+      "step": 8516
+    },
+    {
+      "epoch": 0.36030967086893984,
+      "grad_norm": 0.20725642144680023,
+      "learning_rate": 0.001,
+      "loss": 2.3907,
+      "step": 8517
+    },
+    {
+      "epoch": 0.3603519756324562,
+      "grad_norm": 0.28327709436416626,
+      "learning_rate": 0.001,
+      "loss": 3.0746,
+      "step": 8518
+    },
+    {
+      "epoch": 0.3603942803959726,
+      "grad_norm": 0.4538307189941406,
+      "learning_rate": 0.001,
+      "loss": 2.6733,
+      "step": 8519
+    },
+    {
+      "epoch": 0.36043658515948895,
+      "grad_norm": 0.20344911515712738,
+      "learning_rate": 0.001,
+      "loss": 2.7097,
+      "step": 8520
+    },
+    {
+      "epoch": 0.3604788899230053,
+      "grad_norm": 0.2238440215587616,
+      "learning_rate": 0.001,
+      "loss": 2.2248,
+      "step": 8521
+    },
+    {
+      "epoch": 0.3605211946865217,
+      "grad_norm": 0.2143414467573166,
+      "learning_rate": 0.001,
+      "loss": 2.6659,
+      "step": 8522
+    },
+    {
+      "epoch": 0.36056349945003807,
+      "grad_norm": 0.21883174777030945,
+      "learning_rate": 0.001,
+      "loss": 2.8158,
+      "step": 8523
+    },
+    {
+      "epoch": 0.3606058042135544,
+      "grad_norm": 1.42238450050354,
+      "learning_rate": 0.001,
+      "loss": 2.9731,
+      "step": 8524
+    },
+    {
+      "epoch": 0.36064810897707084,
+      "grad_norm": 4.848227500915527,
+      "learning_rate": 0.001,
+      "loss": 2.352,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3606904137405872,
+      "grad_norm": 0.7891883850097656,
+      "learning_rate": 0.001,
+      "loss": 2.0768,
+      "step": 8526
+    },
+    {
+      "epoch": 0.36073271850410354,
+      "grad_norm": 84.44062042236328,
+      "learning_rate": 0.001,
+      "loss": 3.0048,
+      "step": 8527
+    },
+    {
+      "epoch": 0.36077502326761995,
+      "grad_norm": 0.2623269855976105,
+      "learning_rate": 0.001,
+      "loss": 2.887,
+      "step": 8528
+    },
+    {
+      "epoch": 0.3608173280311363,
+      "grad_norm": 0.2041042447090149,
+      "learning_rate": 0.001,
+      "loss": 1.7496,
+      "step": 8529
+    },
+    {
+      "epoch": 0.36085963279465266,
+      "grad_norm": 0.25694215297698975,
+      "learning_rate": 0.001,
+      "loss": 2.0294,
+      "step": 8530
+    },
+    {
+      "epoch": 0.3609019375581691,
+      "grad_norm": 0.4767371416091919,
+      "learning_rate": 0.001,
+      "loss": 1.9393,
+      "step": 8531
+    },
+    {
+      "epoch": 0.3609442423216854,
+      "grad_norm": 0.5682966113090515,
+      "learning_rate": 0.001,
+      "loss": 2.401,
+      "step": 8532
+    },
+    {
+      "epoch": 0.3609865470852018,
+      "grad_norm": 0.754736065864563,
+      "learning_rate": 0.001,
+      "loss": 1.53,
+      "step": 8533
+    },
+    {
+      "epoch": 0.3610288518487182,
+      "grad_norm": 0.19843953847885132,
+      "learning_rate": 0.001,
+      "loss": 2.3422,
+      "step": 8534
+    },
+    {
+      "epoch": 0.36107115661223455,
+      "grad_norm": 0.6914002299308777,
+      "learning_rate": 0.001,
+      "loss": 2.5332,
+      "step": 8535
+    },
+    {
+      "epoch": 0.3611134613757509,
+      "grad_norm": 0.9868711829185486,
+      "learning_rate": 0.001,
+      "loss": 3.0186,
+      "step": 8536
+    },
+    {
+      "epoch": 0.3611557661392673,
+      "grad_norm": 0.9456235766410828,
+      "learning_rate": 0.001,
+      "loss": 1.7753,
+      "step": 8537
+    },
+    {
+      "epoch": 0.36119807090278366,
+      "grad_norm": 0.3088882565498352,
+      "learning_rate": 0.001,
+      "loss": 2.0363,
+      "step": 8538
+    },
+    {
+      "epoch": 0.3612403756663,
+      "grad_norm": 0.8840659856796265,
+      "learning_rate": 0.001,
+      "loss": 3.2643,
+      "step": 8539
+    },
+    {
+      "epoch": 0.36128268042981637,
+      "grad_norm": 1.7744220495224,
+      "learning_rate": 0.001,
+      "loss": 1.981,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3613249851933328,
+      "grad_norm": 0.8132971525192261,
+      "learning_rate": 0.001,
+      "loss": 3.234,
+      "step": 8541
+    },
+    {
+      "epoch": 0.36136728995684914,
+      "grad_norm": 0.26763689517974854,
+      "learning_rate": 0.001,
+      "loss": 2.3408,
+      "step": 8542
+    },
+    {
+      "epoch": 0.3614095947203655,
+      "grad_norm": 0.18884502351284027,
+      "learning_rate": 0.001,
+      "loss": 3.1137,
+      "step": 8543
+    },
+    {
+      "epoch": 0.3614518994838819,
+      "grad_norm": 4.647204399108887,
+      "learning_rate": 0.001,
+      "loss": 2.6148,
+      "step": 8544
+    },
+    {
+      "epoch": 0.36149420424739825,
+      "grad_norm": 3.6507856845855713,
+      "learning_rate": 0.001,
+      "loss": 3.341,
+      "step": 8545
+    },
+    {
+      "epoch": 0.3615365090109146,
+      "grad_norm": 0.21972277760505676,
+      "learning_rate": 0.001,
+      "loss": 1.8723,
+      "step": 8546
+    },
+    {
+      "epoch": 0.361578813774431,
+      "grad_norm": 0.3465214967727661,
+      "learning_rate": 0.001,
+      "loss": 2.1504,
+      "step": 8547
+    },
+    {
+      "epoch": 0.3616211185379474,
+      "grad_norm": 0.40754181146621704,
+      "learning_rate": 0.001,
+      "loss": 2.694,
+      "step": 8548
+    },
+    {
+      "epoch": 0.3616634233014637,
+      "grad_norm": 0.20011767745018005,
+      "learning_rate": 0.001,
+      "loss": 2.4816,
+      "step": 8549
+    },
+    {
+      "epoch": 0.36170572806498014,
+      "grad_norm": 0.1964236944913864,
+      "learning_rate": 0.001,
+      "loss": 1.6528,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3617480328284965,
+      "grad_norm": 0.17661263048648834,
+      "learning_rate": 0.001,
+      "loss": 1.9885,
+      "step": 8551
+    },
+    {
+      "epoch": 0.36179033759201285,
+      "grad_norm": 0.7868698835372925,
+      "learning_rate": 0.001,
+      "loss": 2.6136,
+      "step": 8552
+    },
+    {
+      "epoch": 0.36183264235552925,
+      "grad_norm": 0.23645655810832977,
+      "learning_rate": 0.001,
+      "loss": 2.408,
+      "step": 8553
+    },
+    {
+      "epoch": 0.3618749471190456,
+      "grad_norm": 0.18343326449394226,
+      "learning_rate": 0.001,
+      "loss": 1.5529,
+      "step": 8554
+    },
+    {
+      "epoch": 0.36191725188256196,
+      "grad_norm": 0.18939687311649323,
+      "learning_rate": 0.001,
+      "loss": 2.5676,
+      "step": 8555
+    },
+    {
+      "epoch": 0.3619595566460784,
+      "grad_norm": 0.7980564832687378,
+      "learning_rate": 0.001,
+      "loss": 2.4449,
+      "step": 8556
+    },
+    {
+      "epoch": 0.3620018614095947,
+      "grad_norm": 0.8566358089447021,
+      "learning_rate": 0.001,
+      "loss": 3.1181,
+      "step": 8557
+    },
+    {
+      "epoch": 0.3620441661731111,
+      "grad_norm": 0.24437959492206573,
+      "learning_rate": 0.001,
+      "loss": 2.1782,
+      "step": 8558
+    },
+    {
+      "epoch": 0.3620864709366275,
+      "grad_norm": 0.22951102256774902,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 8559
+    },
+    {
+      "epoch": 0.36212877570014385,
+      "grad_norm": 0.5679842233657837,
+      "learning_rate": 0.001,
+      "loss": 2.8188,
+      "step": 8560
+    },
+    {
+      "epoch": 0.3621710804636602,
+      "grad_norm": 0.2738345265388489,
+      "learning_rate": 0.001,
+      "loss": 2.9867,
+      "step": 8561
+    },
+    {
+      "epoch": 0.36221338522717655,
+      "grad_norm": 0.6427478790283203,
+      "learning_rate": 0.001,
+      "loss": 2.0955,
+      "step": 8562
+    },
+    {
+      "epoch": 0.36225568999069296,
+      "grad_norm": 0.17962263524532318,
+      "learning_rate": 0.001,
+      "loss": 1.8587,
+      "step": 8563
+    },
+    {
+      "epoch": 0.3622979947542093,
+      "grad_norm": 1.4514254331588745,
+      "learning_rate": 0.001,
+      "loss": 1.7745,
+      "step": 8564
+    },
+    {
+      "epoch": 0.3623402995177257,
+      "grad_norm": 0.2207036018371582,
+      "learning_rate": 0.001,
+      "loss": 1.6446,
+      "step": 8565
+    },
+    {
+      "epoch": 0.3623826042812421,
+      "grad_norm": 0.3202604055404663,
+      "learning_rate": 0.001,
+      "loss": 2.5328,
+      "step": 8566
+    },
+    {
+      "epoch": 0.36242490904475844,
+      "grad_norm": 0.6351190209388733,
+      "learning_rate": 0.001,
+      "loss": 2.7094,
+      "step": 8567
+    },
+    {
+      "epoch": 0.3624672138082748,
+      "grad_norm": 0.21355757117271423,
+      "learning_rate": 0.001,
+      "loss": 2.6323,
+      "step": 8568
+    },
+    {
+      "epoch": 0.3625095185717912,
+      "grad_norm": 0.20924630761146545,
+      "learning_rate": 0.001,
+      "loss": 1.6898,
+      "step": 8569
+    },
+    {
+      "epoch": 0.36255182333530755,
+      "grad_norm": 0.4818645417690277,
+      "learning_rate": 0.001,
+      "loss": 1.6882,
+      "step": 8570
+    },
+    {
+      "epoch": 0.3625941280988239,
+      "grad_norm": 0.2274334728717804,
+      "learning_rate": 0.001,
+      "loss": 3.8046,
+      "step": 8571
+    },
+    {
+      "epoch": 0.3626364328623403,
+      "grad_norm": 4.399014949798584,
+      "learning_rate": 0.001,
+      "loss": 2.6037,
+      "step": 8572
+    },
+    {
+      "epoch": 0.3626787376258567,
+      "grad_norm": 0.20539069175720215,
+      "learning_rate": 0.001,
+      "loss": 1.8258,
+      "step": 8573
+    },
+    {
+      "epoch": 0.362721042389373,
+      "grad_norm": 0.3310551345348358,
+      "learning_rate": 0.001,
+      "loss": 3.1566,
+      "step": 8574
+    },
+    {
+      "epoch": 0.36276334715288944,
+      "grad_norm": 0.2262580245733261,
+      "learning_rate": 0.001,
+      "loss": 2.6769,
+      "step": 8575
+    },
+    {
+      "epoch": 0.3628056519164058,
+      "grad_norm": 0.20004798471927643,
+      "learning_rate": 0.001,
+      "loss": 1.7463,
+      "step": 8576
+    },
+    {
+      "epoch": 0.36284795667992215,
+      "grad_norm": 0.17591601610183716,
+      "learning_rate": 0.001,
+      "loss": 1.5559,
+      "step": 8577
+    },
+    {
+      "epoch": 0.36289026144343856,
+      "grad_norm": 0.17528226971626282,
+      "learning_rate": 0.001,
+      "loss": 1.9886,
+      "step": 8578
+    },
+    {
+      "epoch": 0.3629325662069549,
+      "grad_norm": 0.3700055181980133,
+      "learning_rate": 0.001,
+      "loss": 2.4776,
+      "step": 8579
+    },
+    {
+      "epoch": 0.36297487097047126,
+      "grad_norm": 0.17485986649990082,
+      "learning_rate": 0.001,
+      "loss": 2.0908,
+      "step": 8580
+    },
+    {
+      "epoch": 0.3630171757339877,
+      "grad_norm": 9.299296379089355,
+      "learning_rate": 0.001,
+      "loss": 2.3967,
+      "step": 8581
+    },
+    {
+      "epoch": 0.36305948049750403,
+      "grad_norm": 0.20761297643184662,
+      "learning_rate": 0.001,
+      "loss": 2.0552,
+      "step": 8582
+    },
+    {
+      "epoch": 0.3631017852610204,
+      "grad_norm": 0.19602414965629578,
+      "learning_rate": 0.001,
+      "loss": 2.1057,
+      "step": 8583
+    },
+    {
+      "epoch": 0.36314409002453674,
+      "grad_norm": 1.7627781629562378,
+      "learning_rate": 0.001,
+      "loss": 2.5504,
+      "step": 8584
+    },
+    {
+      "epoch": 0.36318639478805315,
+      "grad_norm": 1.635079026222229,
+      "learning_rate": 0.001,
+      "loss": 1.8716,
+      "step": 8585
+    },
+    {
+      "epoch": 0.3632286995515695,
+      "grad_norm": 2.4154467582702637,
+      "learning_rate": 0.001,
+      "loss": 2.5212,
+      "step": 8586
+    },
+    {
+      "epoch": 0.36327100431508585,
+      "grad_norm": 0.1748884916305542,
+      "learning_rate": 0.001,
+      "loss": 2.1198,
+      "step": 8587
+    },
+    {
+      "epoch": 0.36331330907860226,
+      "grad_norm": 1.158073902130127,
+      "learning_rate": 0.001,
+      "loss": 2.5192,
+      "step": 8588
+    },
+    {
+      "epoch": 0.3633556138421186,
+      "grad_norm": 0.24196279048919678,
+      "learning_rate": 0.001,
+      "loss": 2.1299,
+      "step": 8589
+    },
+    {
+      "epoch": 0.363397918605635,
+      "grad_norm": 0.21735741198062897,
+      "learning_rate": 0.001,
+      "loss": 2.2981,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3634402233691514,
+      "grad_norm": 0.22882144153118134,
+      "learning_rate": 0.001,
+      "loss": 2.1787,
+      "step": 8591
+    },
+    {
+      "epoch": 0.36348252813266774,
+      "grad_norm": 0.22966794669628143,
+      "learning_rate": 0.001,
+      "loss": 2.0843,
+      "step": 8592
+    },
+    {
+      "epoch": 0.3635248328961841,
+      "grad_norm": 3.5307583808898926,
+      "learning_rate": 0.001,
+      "loss": 2.6233,
+      "step": 8593
+    },
+    {
+      "epoch": 0.3635671376597005,
+      "grad_norm": 0.1808539777994156,
+      "learning_rate": 0.001,
+      "loss": 2.447,
+      "step": 8594
+    },
+    {
+      "epoch": 0.36360944242321686,
+      "grad_norm": 0.191551074385643,
+      "learning_rate": 0.001,
+      "loss": 1.9306,
+      "step": 8595
+    },
+    {
+      "epoch": 0.3636517471867332,
+      "grad_norm": 0.2742213308811188,
+      "learning_rate": 0.001,
+      "loss": 2.5281,
+      "step": 8596
+    },
+    {
+      "epoch": 0.3636940519502496,
+      "grad_norm": 3.3009042739868164,
+      "learning_rate": 0.001,
+      "loss": 3.4417,
+      "step": 8597
+    },
+    {
+      "epoch": 0.363736356713766,
+      "grad_norm": 0.22634568810462952,
+      "learning_rate": 0.001,
+      "loss": 1.9962,
+      "step": 8598
+    },
+    {
+      "epoch": 0.36377866147728233,
+      "grad_norm": 0.3063672184944153,
+      "learning_rate": 0.001,
+      "loss": 2.1643,
+      "step": 8599
+    },
+    {
+      "epoch": 0.36382096624079874,
+      "grad_norm": 0.4341367781162262,
+      "learning_rate": 0.001,
+      "loss": 2.8908,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3638632710043151,
+      "grad_norm": 0.2267148345708847,
+      "learning_rate": 0.001,
+      "loss": 3.3518,
+      "step": 8601
+    },
+    {
+      "epoch": 0.36390557576783145,
+      "grad_norm": 0.235674649477005,
+      "learning_rate": 0.001,
+      "loss": 2.0174,
+      "step": 8602
+    },
+    {
+      "epoch": 0.36394788053134786,
+      "grad_norm": 0.1691252589225769,
+      "learning_rate": 0.001,
+      "loss": 1.7468,
+      "step": 8603
+    },
+    {
+      "epoch": 0.3639901852948642,
+      "grad_norm": 0.4646351933479309,
+      "learning_rate": 0.001,
+      "loss": 2.9241,
+      "step": 8604
+    },
+    {
+      "epoch": 0.36403249005838056,
+      "grad_norm": 0.18952451646327972,
+      "learning_rate": 0.001,
+      "loss": 1.9208,
+      "step": 8605
+    },
+    {
+      "epoch": 0.3640747948218969,
+      "grad_norm": 1.2215399742126465,
+      "learning_rate": 0.001,
+      "loss": 2.0738,
+      "step": 8606
+    },
+    {
+      "epoch": 0.36411709958541333,
+      "grad_norm": 0.1823042780160904,
+      "learning_rate": 0.001,
+      "loss": 2.0435,
+      "step": 8607
+    },
+    {
+      "epoch": 0.3641594043489297,
+      "grad_norm": 0.18429160118103027,
+      "learning_rate": 0.001,
+      "loss": 2.1176,
+      "step": 8608
+    },
+    {
+      "epoch": 0.36420170911244604,
+      "grad_norm": 0.30278459191322327,
+      "learning_rate": 0.001,
+      "loss": 2.0551,
+      "step": 8609
+    },
+    {
+      "epoch": 0.36424401387596245,
+      "grad_norm": 0.27430135011672974,
+      "learning_rate": 0.001,
+      "loss": 2.7956,
+      "step": 8610
+    },
+    {
+      "epoch": 0.3642863186394788,
+      "grad_norm": 0.7540306448936462,
+      "learning_rate": 0.001,
+      "loss": 2.312,
+      "step": 8611
+    },
+    {
+      "epoch": 0.36432862340299516,
+      "grad_norm": 0.24535398185253143,
+      "learning_rate": 0.001,
+      "loss": 2.2725,
+      "step": 8612
+    },
+    {
+      "epoch": 0.36437092816651157,
+      "grad_norm": 0.39230287075042725,
+      "learning_rate": 0.001,
+      "loss": 2.2182,
+      "step": 8613
+    },
+    {
+      "epoch": 0.3644132329300279,
+      "grad_norm": 0.19538423418998718,
+      "learning_rate": 0.001,
+      "loss": 2.1788,
+      "step": 8614
+    },
+    {
+      "epoch": 0.3644555376935443,
+      "grad_norm": 0.40475207567214966,
+      "learning_rate": 0.001,
+      "loss": 2.1257,
+      "step": 8615
+    },
+    {
+      "epoch": 0.3644978424570607,
+      "grad_norm": 0.27652233839035034,
+      "learning_rate": 0.001,
+      "loss": 3.0038,
+      "step": 8616
+    },
+    {
+      "epoch": 0.36454014722057704,
+      "grad_norm": 0.19666795432567596,
+      "learning_rate": 0.001,
+      "loss": 1.4796,
+      "step": 8617
+    },
+    {
+      "epoch": 0.3645824519840934,
+      "grad_norm": 0.8510257601737976,
+      "learning_rate": 0.001,
+      "loss": 2.5566,
+      "step": 8618
+    },
+    {
+      "epoch": 0.3646247567476098,
+      "grad_norm": 0.2172936201095581,
+      "learning_rate": 0.001,
+      "loss": 2.1356,
+      "step": 8619
+    },
+    {
+      "epoch": 0.36466706151112616,
+      "grad_norm": 0.20282861590385437,
+      "learning_rate": 0.001,
+      "loss": 1.5411,
+      "step": 8620
+    },
+    {
+      "epoch": 0.3647093662746425,
+      "grad_norm": 0.1585301011800766,
+      "learning_rate": 0.001,
+      "loss": 1.3001,
+      "step": 8621
+    },
+    {
+      "epoch": 0.3647516710381589,
+      "grad_norm": 0.20633943378925323,
+      "learning_rate": 0.001,
+      "loss": 1.869,
+      "step": 8622
+    },
+    {
+      "epoch": 0.3647939758016753,
+      "grad_norm": 0.24952706694602966,
+      "learning_rate": 0.001,
+      "loss": 2.2658,
+      "step": 8623
+    },
+    {
+      "epoch": 0.36483628056519163,
+      "grad_norm": 0.24051597714424133,
+      "learning_rate": 0.001,
+      "loss": 2.0523,
+      "step": 8624
+    },
+    {
+      "epoch": 0.36487858532870804,
+      "grad_norm": 1.4806382656097412,
+      "learning_rate": 0.001,
+      "loss": 3.0184,
+      "step": 8625
+    },
+    {
+      "epoch": 0.3649208900922244,
+      "grad_norm": 0.2159508466720581,
+      "learning_rate": 0.001,
+      "loss": 1.995,
+      "step": 8626
+    },
+    {
+      "epoch": 0.36496319485574075,
+      "grad_norm": 0.5792630314826965,
+      "learning_rate": 0.001,
+      "loss": 3.4955,
+      "step": 8627
+    },
+    {
+      "epoch": 0.3650054996192571,
+      "grad_norm": 0.18533584475517273,
+      "learning_rate": 0.001,
+      "loss": 1.815,
+      "step": 8628
+    },
+    {
+      "epoch": 0.3650478043827735,
+      "grad_norm": 0.2772461175918579,
+      "learning_rate": 0.001,
+      "loss": 2.1131,
+      "step": 8629
+    },
+    {
+      "epoch": 0.36509010914628987,
+      "grad_norm": 0.27100545167922974,
+      "learning_rate": 0.001,
+      "loss": 2.2957,
+      "step": 8630
+    },
+    {
+      "epoch": 0.3651324139098062,
+      "grad_norm": 0.2463526576757431,
+      "learning_rate": 0.001,
+      "loss": 3.0297,
+      "step": 8631
+    },
+    {
+      "epoch": 0.36517471867332263,
+      "grad_norm": 1.035469889640808,
+      "learning_rate": 0.001,
+      "loss": 1.6165,
+      "step": 8632
+    },
+    {
+      "epoch": 0.365217023436839,
+      "grad_norm": 0.23822209239006042,
+      "learning_rate": 0.001,
+      "loss": 2.0529,
+      "step": 8633
+    },
+    {
+      "epoch": 0.36525932820035534,
+      "grad_norm": 0.20233659446239471,
+      "learning_rate": 0.001,
+      "loss": 1.6715,
+      "step": 8634
+    },
+    {
+      "epoch": 0.36530163296387175,
+      "grad_norm": 0.48247143626213074,
+      "learning_rate": 0.001,
+      "loss": 2.6857,
+      "step": 8635
+    },
+    {
+      "epoch": 0.3653439377273881,
+      "grad_norm": 0.19231975078582764,
+      "learning_rate": 0.001,
+      "loss": 1.9427,
+      "step": 8636
+    },
+    {
+      "epoch": 0.36538624249090446,
+      "grad_norm": 0.7569523453712463,
+      "learning_rate": 0.001,
+      "loss": 2.3201,
+      "step": 8637
+    },
+    {
+      "epoch": 0.36542854725442087,
+      "grad_norm": 0.27033358812332153,
+      "learning_rate": 0.001,
+      "loss": 2.4324,
+      "step": 8638
+    },
+    {
+      "epoch": 0.3654708520179372,
+      "grad_norm": 0.18519775569438934,
+      "learning_rate": 0.001,
+      "loss": 1.9692,
+      "step": 8639
+    },
+    {
+      "epoch": 0.3655131567814536,
+      "grad_norm": 0.25175315141677856,
+      "learning_rate": 0.001,
+      "loss": 3.5475,
+      "step": 8640
+    },
+    {
+      "epoch": 0.36555546154497,
+      "grad_norm": 0.18655212223529816,
+      "learning_rate": 0.001,
+      "loss": 1.7191,
+      "step": 8641
+    },
+    {
+      "epoch": 0.36559776630848634,
+      "grad_norm": 0.21611282229423523,
+      "learning_rate": 0.001,
+      "loss": 2.2513,
+      "step": 8642
+    },
+    {
+      "epoch": 0.3656400710720027,
+      "grad_norm": 0.21771200001239777,
+      "learning_rate": 0.001,
+      "loss": 1.6885,
+      "step": 8643
+    },
+    {
+      "epoch": 0.3656823758355191,
+      "grad_norm": 0.19719097018241882,
+      "learning_rate": 0.001,
+      "loss": 2.629,
+      "step": 8644
+    },
+    {
+      "epoch": 0.36572468059903546,
+      "grad_norm": 11.389663696289062,
+      "learning_rate": 0.001,
+      "loss": 2.8977,
+      "step": 8645
+    },
+    {
+      "epoch": 0.3657669853625518,
+      "grad_norm": 0.15473631024360657,
+      "learning_rate": 0.001,
+      "loss": 1.5634,
+      "step": 8646
+    },
+    {
+      "epoch": 0.3658092901260682,
+      "grad_norm": 0.4173217713832855,
+      "learning_rate": 0.001,
+      "loss": 2.0252,
+      "step": 8647
+    },
+    {
+      "epoch": 0.3658515948895846,
+      "grad_norm": 1.9161365032196045,
+      "learning_rate": 0.001,
+      "loss": 3.082,
+      "step": 8648
+    },
+    {
+      "epoch": 0.36589389965310093,
+      "grad_norm": 0.1819329559803009,
+      "learning_rate": 0.001,
+      "loss": 2.5534,
+      "step": 8649
+    },
+    {
+      "epoch": 0.3659362044166173,
+      "grad_norm": 0.2493087202310562,
+      "learning_rate": 0.001,
+      "loss": 2.2305,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3659785091801337,
+      "grad_norm": 0.20480546355247498,
+      "learning_rate": 0.001,
+      "loss": 2.044,
+      "step": 8651
+    },
+    {
+      "epoch": 0.36602081394365005,
+      "grad_norm": 0.22163717448711395,
+      "learning_rate": 0.001,
+      "loss": 1.9366,
+      "step": 8652
+    },
+    {
+      "epoch": 0.3660631187071664,
+      "grad_norm": 0.187017560005188,
+      "learning_rate": 0.001,
+      "loss": 1.992,
+      "step": 8653
+    },
+    {
+      "epoch": 0.3661054234706828,
+      "grad_norm": 0.21020790934562683,
+      "learning_rate": 0.001,
+      "loss": 2.3122,
+      "step": 8654
+    },
+    {
+      "epoch": 0.36614772823419917,
+      "grad_norm": 0.19258715212345123,
+      "learning_rate": 0.001,
+      "loss": 3.0791,
+      "step": 8655
+    },
+    {
+      "epoch": 0.3661900329977155,
+      "grad_norm": 0.18852315843105316,
+      "learning_rate": 0.001,
+      "loss": 2.7449,
+      "step": 8656
+    },
+    {
+      "epoch": 0.36623233776123193,
+      "grad_norm": 0.16643263399600983,
+      "learning_rate": 0.001,
+      "loss": 1.8381,
+      "step": 8657
+    },
+    {
+      "epoch": 0.3662746425247483,
+      "grad_norm": 0.17222201824188232,
+      "learning_rate": 0.001,
+      "loss": 1.7572,
+      "step": 8658
+    },
+    {
+      "epoch": 0.36631694728826464,
+      "grad_norm": 0.1717625856399536,
+      "learning_rate": 0.001,
+      "loss": 2.1745,
+      "step": 8659
+    },
+    {
+      "epoch": 0.36635925205178105,
+      "grad_norm": 0.22495992481708527,
+      "learning_rate": 0.001,
+      "loss": 2.179,
+      "step": 8660
+    },
+    {
+      "epoch": 0.3664015568152974,
+      "grad_norm": 2.0256707668304443,
+      "learning_rate": 0.001,
+      "loss": 3.8565,
+      "step": 8661
+    },
+    {
+      "epoch": 0.36644386157881376,
+      "grad_norm": 0.20122158527374268,
+      "learning_rate": 0.001,
+      "loss": 2.0525,
+      "step": 8662
+    },
+    {
+      "epoch": 0.36648616634233017,
+      "grad_norm": 0.2111087143421173,
+      "learning_rate": 0.001,
+      "loss": 2.4875,
+      "step": 8663
+    },
+    {
+      "epoch": 0.3665284711058465,
+      "grad_norm": 0.6038394570350647,
+      "learning_rate": 0.001,
+      "loss": 2.5007,
+      "step": 8664
+    },
+    {
+      "epoch": 0.3665707758693629,
+      "grad_norm": 0.18492920696735382,
+      "learning_rate": 0.001,
+      "loss": 1.9992,
+      "step": 8665
+    },
+    {
+      "epoch": 0.3666130806328793,
+      "grad_norm": 0.39475682377815247,
+      "learning_rate": 0.001,
+      "loss": 3.2232,
+      "step": 8666
+    },
+    {
+      "epoch": 0.36665538539639564,
+      "grad_norm": 0.2003537118434906,
+      "learning_rate": 0.001,
+      "loss": 1.8457,
+      "step": 8667
+    },
+    {
+      "epoch": 0.366697690159912,
+      "grad_norm": 0.20579084753990173,
+      "learning_rate": 0.001,
+      "loss": 2.3453,
+      "step": 8668
+    },
+    {
+      "epoch": 0.3667399949234284,
+      "grad_norm": 0.3103180527687073,
+      "learning_rate": 0.001,
+      "loss": 2.771,
+      "step": 8669
+    },
+    {
+      "epoch": 0.36678229968694476,
+      "grad_norm": 0.18858268857002258,
+      "learning_rate": 0.001,
+      "loss": 1.9894,
+      "step": 8670
+    },
+    {
+      "epoch": 0.3668246044504611,
+      "grad_norm": 0.2262255698442459,
+      "learning_rate": 0.001,
+      "loss": 1.9428,
+      "step": 8671
+    },
+    {
+      "epoch": 0.3668669092139775,
+      "grad_norm": 0.4273422062397003,
+      "learning_rate": 0.001,
+      "loss": 3.7103,
+      "step": 8672
+    },
+    {
+      "epoch": 0.3669092139774939,
+      "grad_norm": 0.22894766926765442,
+      "learning_rate": 0.001,
+      "loss": 2.0784,
+      "step": 8673
+    },
+    {
+      "epoch": 0.36695151874101023,
+      "grad_norm": 0.248250812292099,
+      "learning_rate": 0.001,
+      "loss": 1.7906,
+      "step": 8674
+    },
+    {
+      "epoch": 0.3669938235045266,
+      "grad_norm": 0.307745099067688,
+      "learning_rate": 0.001,
+      "loss": 3.2343,
+      "step": 8675
+    },
+    {
+      "epoch": 0.367036128268043,
+      "grad_norm": 0.21414296329021454,
+      "learning_rate": 0.001,
+      "loss": 3.1725,
+      "step": 8676
+    },
+    {
+      "epoch": 0.36707843303155935,
+      "grad_norm": 1.1623830795288086,
+      "learning_rate": 0.001,
+      "loss": 1.9302,
+      "step": 8677
+    },
+    {
+      "epoch": 0.3671207377950757,
+      "grad_norm": 0.218984916806221,
+      "learning_rate": 0.001,
+      "loss": 2.4448,
+      "step": 8678
+    },
+    {
+      "epoch": 0.3671630425585921,
+      "grad_norm": 0.24998073279857635,
+      "learning_rate": 0.001,
+      "loss": 2.1159,
+      "step": 8679
+    },
+    {
+      "epoch": 0.36720534732210847,
+      "grad_norm": 0.19620297849178314,
+      "learning_rate": 0.001,
+      "loss": 2.264,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3672476520856248,
+      "grad_norm": 0.1893426477909088,
+      "learning_rate": 0.001,
+      "loss": 1.677,
+      "step": 8681
+    },
+    {
+      "epoch": 0.36728995684914123,
+      "grad_norm": 0.22591541707515717,
+      "learning_rate": 0.001,
+      "loss": 3.0901,
+      "step": 8682
+    },
+    {
+      "epoch": 0.3673322616126576,
+      "grad_norm": 0.3786305785179138,
+      "learning_rate": 0.001,
+      "loss": 3.0062,
+      "step": 8683
+    },
+    {
+      "epoch": 0.36737456637617394,
+      "grad_norm": 5.867971897125244,
+      "learning_rate": 0.001,
+      "loss": 2.209,
+      "step": 8684
+    },
+    {
+      "epoch": 0.36741687113969035,
+      "grad_norm": 44.805763244628906,
+      "learning_rate": 0.001,
+      "loss": 2.1827,
+      "step": 8685
+    },
+    {
+      "epoch": 0.3674591759032067,
+      "grad_norm": 0.18080832064151764,
+      "learning_rate": 0.001,
+      "loss": 1.9619,
+      "step": 8686
+    },
+    {
+      "epoch": 0.36750148066672306,
+      "grad_norm": 0.22980724275112152,
+      "learning_rate": 0.001,
+      "loss": 2.2155,
+      "step": 8687
+    },
+    {
+      "epoch": 0.36754378543023947,
+      "grad_norm": 0.2389446198940277,
+      "learning_rate": 0.001,
+      "loss": 2.045,
+      "step": 8688
+    },
+    {
+      "epoch": 0.3675860901937558,
+      "grad_norm": 0.22380129992961884,
+      "learning_rate": 0.001,
+      "loss": 2.716,
+      "step": 8689
+    },
+    {
+      "epoch": 0.3676283949572722,
+      "grad_norm": 0.3434034585952759,
+      "learning_rate": 0.001,
+      "loss": 3.1835,
+      "step": 8690
+    },
+    {
+      "epoch": 0.3676706997207886,
+      "grad_norm": 0.237090066075325,
+      "learning_rate": 0.001,
+      "loss": 1.9271,
+      "step": 8691
+    },
+    {
+      "epoch": 0.36771300448430494,
+      "grad_norm": 0.22618965804576874,
+      "learning_rate": 0.001,
+      "loss": 2.1982,
+      "step": 8692
+    },
+    {
+      "epoch": 0.3677553092478213,
+      "grad_norm": 0.1993749737739563,
+      "learning_rate": 0.001,
+      "loss": 4.0091,
+      "step": 8693
+    },
+    {
+      "epoch": 0.3677976140113377,
+      "grad_norm": 2.4615116119384766,
+      "learning_rate": 0.001,
+      "loss": 2.3653,
+      "step": 8694
+    },
+    {
+      "epoch": 0.36783991877485406,
+      "grad_norm": 0.5453082323074341,
+      "learning_rate": 0.001,
+      "loss": 3.2241,
+      "step": 8695
+    },
+    {
+      "epoch": 0.3678822235383704,
+      "grad_norm": 0.22832508385181427,
+      "learning_rate": 0.001,
+      "loss": 2.1284,
+      "step": 8696
+    },
+    {
+      "epoch": 0.36792452830188677,
+      "grad_norm": 0.22793933749198914,
+      "learning_rate": 0.001,
+      "loss": 2.7905,
+      "step": 8697
+    },
+    {
+      "epoch": 0.3679668330654032,
+      "grad_norm": 0.2116621732711792,
+      "learning_rate": 0.001,
+      "loss": 3.0399,
+      "step": 8698
+    },
+    {
+      "epoch": 0.36800913782891953,
+      "grad_norm": 0.26181337237358093,
+      "learning_rate": 0.001,
+      "loss": 1.625,
+      "step": 8699
+    },
+    {
+      "epoch": 0.3680514425924359,
+      "grad_norm": 0.228001207113266,
+      "learning_rate": 0.001,
+      "loss": 3.5122,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3680937473559523,
+      "grad_norm": 0.1740439236164093,
+      "learning_rate": 0.001,
+      "loss": 2.094,
+      "step": 8701
+    },
+    {
+      "epoch": 0.36813605211946865,
+      "grad_norm": 0.19469280540943146,
+      "learning_rate": 0.001,
+      "loss": 2.2975,
+      "step": 8702
+    },
+    {
+      "epoch": 0.368178356882985,
+      "grad_norm": 0.1703725904226303,
+      "learning_rate": 0.001,
+      "loss": 3.327,
+      "step": 8703
+    },
+    {
+      "epoch": 0.3682206616465014,
+      "grad_norm": 0.20927943289279938,
+      "learning_rate": 0.001,
+      "loss": 1.7857,
+      "step": 8704
+    },
+    {
+      "epoch": 0.36826296641001777,
+      "grad_norm": 0.15202440321445465,
+      "learning_rate": 0.001,
+      "loss": 2.1899,
+      "step": 8705
+    },
+    {
+      "epoch": 0.3683052711735341,
+      "grad_norm": 0.17597658932209015,
+      "learning_rate": 0.001,
+      "loss": 1.7741,
+      "step": 8706
+    },
+    {
+      "epoch": 0.36834757593705053,
+      "grad_norm": 0.16804081201553345,
+      "learning_rate": 0.001,
+      "loss": 2.0558,
+      "step": 8707
+    },
+    {
+      "epoch": 0.3683898807005669,
+      "grad_norm": 0.18276292085647583,
+      "learning_rate": 0.001,
+      "loss": 2.2216,
+      "step": 8708
+    },
+    {
+      "epoch": 0.36843218546408324,
+      "grad_norm": 0.2148251235485077,
+      "learning_rate": 0.001,
+      "loss": 1.9215,
+      "step": 8709
+    },
+    {
+      "epoch": 0.36847449022759965,
+      "grad_norm": 0.9029409289360046,
+      "learning_rate": 0.001,
+      "loss": 1.6041,
+      "step": 8710
+    },
+    {
+      "epoch": 0.368516794991116,
+      "grad_norm": 0.19321629405021667,
+      "learning_rate": 0.001,
+      "loss": 1.8104,
+      "step": 8711
+    },
+    {
+      "epoch": 0.36855909975463236,
+      "grad_norm": 0.177422434091568,
+      "learning_rate": 0.001,
+      "loss": 1.8768,
+      "step": 8712
+    },
+    {
+      "epoch": 0.36860140451814877,
+      "grad_norm": 0.2311866581439972,
+      "learning_rate": 0.001,
+      "loss": 2.567,
+      "step": 8713
+    },
+    {
+      "epoch": 0.3686437092816651,
+      "grad_norm": 0.2138138860464096,
+      "learning_rate": 0.001,
+      "loss": 2.0323,
+      "step": 8714
+    },
+    {
+      "epoch": 0.3686860140451815,
+      "grad_norm": 0.17242108285427094,
+      "learning_rate": 0.001,
+      "loss": 2.3389,
+      "step": 8715
+    },
+    {
+      "epoch": 0.3687283188086979,
+      "grad_norm": 3.463766098022461,
+      "learning_rate": 0.001,
+      "loss": 1.9939,
+      "step": 8716
+    },
+    {
+      "epoch": 0.36877062357221424,
+      "grad_norm": 0.29998794198036194,
+      "learning_rate": 0.001,
+      "loss": 2.0793,
+      "step": 8717
+    },
+    {
+      "epoch": 0.3688129283357306,
+      "grad_norm": 0.22262133657932281,
+      "learning_rate": 0.001,
+      "loss": 2.7188,
+      "step": 8718
+    },
+    {
+      "epoch": 0.36885523309924695,
+      "grad_norm": 0.28944888710975647,
+      "learning_rate": 0.001,
+      "loss": 1.7407,
+      "step": 8719
+    },
+    {
+      "epoch": 0.36889753786276336,
+      "grad_norm": 0.20619124174118042,
+      "learning_rate": 0.001,
+      "loss": 1.8345,
+      "step": 8720
+    },
+    {
+      "epoch": 0.3689398426262797,
+      "grad_norm": 0.1661635786294937,
+      "learning_rate": 0.001,
+      "loss": 1.586,
+      "step": 8721
+    },
+    {
+      "epoch": 0.36898214738979607,
+      "grad_norm": 0.222077414393425,
+      "learning_rate": 0.001,
+      "loss": 3.3912,
+      "step": 8722
+    },
+    {
+      "epoch": 0.3690244521533125,
+      "grad_norm": 0.19446569681167603,
+      "learning_rate": 0.001,
+      "loss": 2.499,
+      "step": 8723
+    },
+    {
+      "epoch": 0.36906675691682883,
+      "grad_norm": 1.546207070350647,
+      "learning_rate": 0.001,
+      "loss": 1.9862,
+      "step": 8724
+    },
+    {
+      "epoch": 0.3691090616803452,
+      "grad_norm": 0.22588209807872772,
+      "learning_rate": 0.001,
+      "loss": 3.0398,
+      "step": 8725
+    },
+    {
+      "epoch": 0.3691513664438616,
+      "grad_norm": 0.22056500613689423,
+      "learning_rate": 0.001,
+      "loss": 2.7724,
+      "step": 8726
+    },
+    {
+      "epoch": 0.36919367120737795,
+      "grad_norm": 10.859556198120117,
+      "learning_rate": 0.001,
+      "loss": 1.9629,
+      "step": 8727
+    },
+    {
+      "epoch": 0.3692359759708943,
+      "grad_norm": 0.22937330603599548,
+      "learning_rate": 0.001,
+      "loss": 2.0512,
+      "step": 8728
+    },
+    {
+      "epoch": 0.3692782807344107,
+      "grad_norm": 2.563246965408325,
+      "learning_rate": 0.001,
+      "loss": 2.3327,
+      "step": 8729
+    },
+    {
+      "epoch": 0.36932058549792707,
+      "grad_norm": 0.31085848808288574,
+      "learning_rate": 0.001,
+      "loss": 2.0146,
+      "step": 8730
+    },
+    {
+      "epoch": 0.3693628902614434,
+      "grad_norm": 0.22053919732570648,
+      "learning_rate": 0.001,
+      "loss": 1.9071,
+      "step": 8731
+    },
+    {
+      "epoch": 0.36940519502495983,
+      "grad_norm": 0.3156310021877289,
+      "learning_rate": 0.001,
+      "loss": 1.7969,
+      "step": 8732
+    },
+    {
+      "epoch": 0.3694474997884762,
+      "grad_norm": 0.24050398170948029,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 8733
+    },
+    {
+      "epoch": 0.36948980455199254,
+      "grad_norm": 0.18906262516975403,
+      "learning_rate": 0.001,
+      "loss": 1.7419,
+      "step": 8734
+    },
+    {
+      "epoch": 0.36953210931550895,
+      "grad_norm": 0.17383328080177307,
+      "learning_rate": 0.001,
+      "loss": 2.0027,
+      "step": 8735
+    },
+    {
+      "epoch": 0.3695744140790253,
+      "grad_norm": 0.19866390526294708,
+      "learning_rate": 0.001,
+      "loss": 2.1751,
+      "step": 8736
+    },
+    {
+      "epoch": 0.36961671884254166,
+      "grad_norm": 0.2835475504398346,
+      "learning_rate": 0.001,
+      "loss": 1.9432,
+      "step": 8737
+    },
+    {
+      "epoch": 0.36965902360605807,
+      "grad_norm": 0.9217908978462219,
+      "learning_rate": 0.001,
+      "loss": 1.6859,
+      "step": 8738
+    },
+    {
+      "epoch": 0.3697013283695744,
+      "grad_norm": 0.2083401083946228,
+      "learning_rate": 0.001,
+      "loss": 2.3959,
+      "step": 8739
+    },
+    {
+      "epoch": 0.3697436331330908,
+      "grad_norm": 0.19344541430473328,
+      "learning_rate": 0.001,
+      "loss": 1.929,
+      "step": 8740
+    },
+    {
+      "epoch": 0.36978593789660713,
+      "grad_norm": 0.2091798186302185,
+      "learning_rate": 0.001,
+      "loss": 1.8886,
+      "step": 8741
+    },
+    {
+      "epoch": 0.36982824266012354,
+      "grad_norm": 0.2184278666973114,
+      "learning_rate": 0.001,
+      "loss": 2.0926,
+      "step": 8742
+    },
+    {
+      "epoch": 0.3698705474236399,
+      "grad_norm": 0.2625698149204254,
+      "learning_rate": 0.001,
+      "loss": 2.3432,
+      "step": 8743
+    },
+    {
+      "epoch": 0.36991285218715625,
+      "grad_norm": 0.2139061689376831,
+      "learning_rate": 0.001,
+      "loss": 2.6096,
+      "step": 8744
+    },
+    {
+      "epoch": 0.36995515695067266,
+      "grad_norm": 0.18271711468696594,
+      "learning_rate": 0.001,
+      "loss": 2.6203,
+      "step": 8745
+    },
+    {
+      "epoch": 0.369997461714189,
+      "grad_norm": 0.16942210495471954,
+      "learning_rate": 0.001,
+      "loss": 2.8279,
+      "step": 8746
+    },
+    {
+      "epoch": 0.37003976647770537,
+      "grad_norm": 0.19270247220993042,
+      "learning_rate": 0.001,
+      "loss": 1.8951,
+      "step": 8747
+    },
+    {
+      "epoch": 0.3700820712412218,
+      "grad_norm": 0.19946153461933136,
+      "learning_rate": 0.001,
+      "loss": 2.6455,
+      "step": 8748
+    },
+    {
+      "epoch": 0.37012437600473813,
+      "grad_norm": 0.21664415299892426,
+      "learning_rate": 0.001,
+      "loss": 2.2819,
+      "step": 8749
+    },
+    {
+      "epoch": 0.3701666807682545,
+      "grad_norm": 0.19284196197986603,
+      "learning_rate": 0.001,
+      "loss": 2.4914,
+      "step": 8750
+    },
+    {
+      "epoch": 0.3702089855317709,
+      "grad_norm": 0.18668252229690552,
+      "learning_rate": 0.001,
+      "loss": 2.0608,
+      "step": 8751
+    },
+    {
+      "epoch": 0.37025129029528725,
+      "grad_norm": 1.1074740886688232,
+      "learning_rate": 0.001,
+      "loss": 2.6925,
+      "step": 8752
+    },
+    {
+      "epoch": 0.3702935950588036,
+      "grad_norm": 0.20908206701278687,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 8753
+    },
+    {
+      "epoch": 0.37033589982232,
+      "grad_norm": 0.19691090285778046,
+      "learning_rate": 0.001,
+      "loss": 2.0874,
+      "step": 8754
+    },
+    {
+      "epoch": 0.37037820458583637,
+      "grad_norm": 0.25301340222358704,
+      "learning_rate": 0.001,
+      "loss": 2.1522,
+      "step": 8755
+    },
+    {
+      "epoch": 0.3704205093493527,
+      "grad_norm": 0.1993921846151352,
+      "learning_rate": 0.001,
+      "loss": 2.3689,
+      "step": 8756
+    },
+    {
+      "epoch": 0.37046281411286913,
+      "grad_norm": 0.5521764755249023,
+      "learning_rate": 0.001,
+      "loss": 2.1797,
+      "step": 8757
+    },
+    {
+      "epoch": 0.3705051188763855,
+      "grad_norm": 0.19061268866062164,
+      "learning_rate": 0.001,
+      "loss": 2.8067,
+      "step": 8758
+    },
+    {
+      "epoch": 0.37054742363990184,
+      "grad_norm": 0.2150365263223648,
+      "learning_rate": 0.001,
+      "loss": 1.9172,
+      "step": 8759
+    },
+    {
+      "epoch": 0.37058972840341825,
+      "grad_norm": 0.18955853581428528,
+      "learning_rate": 0.001,
+      "loss": 2.3987,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3706320331669346,
+      "grad_norm": 0.18300773203372955,
+      "learning_rate": 0.001,
+      "loss": 2.123,
+      "step": 8761
+    },
+    {
+      "epoch": 0.37067433793045096,
+      "grad_norm": 0.1935848891735077,
+      "learning_rate": 0.001,
+      "loss": 2.1818,
+      "step": 8762
+    },
+    {
+      "epoch": 0.3707166426939673,
+      "grad_norm": 0.20884737372398376,
+      "learning_rate": 0.001,
+      "loss": 2.2408,
+      "step": 8763
+    },
+    {
+      "epoch": 0.3707589474574837,
+      "grad_norm": 0.915891170501709,
+      "learning_rate": 0.001,
+      "loss": 2.0213,
+      "step": 8764
+    },
+    {
+      "epoch": 0.3708012522210001,
+      "grad_norm": 18.33778190612793,
+      "learning_rate": 0.001,
+      "loss": 2.0037,
+      "step": 8765
+    },
+    {
+      "epoch": 0.37084355698451643,
+      "grad_norm": 1.2620832920074463,
+      "learning_rate": 0.001,
+      "loss": 2.4491,
+      "step": 8766
+    },
+    {
+      "epoch": 0.37088586174803284,
+      "grad_norm": 0.18721136450767517,
+      "learning_rate": 0.001,
+      "loss": 1.9352,
+      "step": 8767
+    },
+    {
+      "epoch": 0.3709281665115492,
+      "grad_norm": 0.1763981133699417,
+      "learning_rate": 0.001,
+      "loss": 2.1145,
+      "step": 8768
+    },
+    {
+      "epoch": 0.37097047127506555,
+      "grad_norm": 0.20833328366279602,
+      "learning_rate": 0.001,
+      "loss": 2.5553,
+      "step": 8769
+    },
+    {
+      "epoch": 0.37101277603858196,
+      "grad_norm": 0.37538349628448486,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 8770
+    },
+    {
+      "epoch": 0.3710550808020983,
+      "grad_norm": 0.1779465526342392,
+      "learning_rate": 0.001,
+      "loss": 1.8611,
+      "step": 8771
+    },
+    {
+      "epoch": 0.37109738556561467,
+      "grad_norm": 0.18236777186393738,
+      "learning_rate": 0.001,
+      "loss": 1.5463,
+      "step": 8772
+    },
+    {
+      "epoch": 0.3711396903291311,
+      "grad_norm": 0.3060573935508728,
+      "learning_rate": 0.001,
+      "loss": 2.9194,
+      "step": 8773
+    },
+    {
+      "epoch": 0.37118199509264743,
+      "grad_norm": 0.20852862298488617,
+      "learning_rate": 0.001,
+      "loss": 2.5415,
+      "step": 8774
+    },
+    {
+      "epoch": 0.3712242998561638,
+      "grad_norm": 0.48245781660079956,
+      "learning_rate": 0.001,
+      "loss": 1.5005,
+      "step": 8775
+    },
+    {
+      "epoch": 0.3712666046196802,
+      "grad_norm": 0.31515905261039734,
+      "learning_rate": 0.001,
+      "loss": 3.4687,
+      "step": 8776
+    },
+    {
+      "epoch": 0.37130890938319655,
+      "grad_norm": 0.2978445887565613,
+      "learning_rate": 0.001,
+      "loss": 2.7744,
+      "step": 8777
+    },
+    {
+      "epoch": 0.3713512141467129,
+      "grad_norm": 0.18172970414161682,
+      "learning_rate": 0.001,
+      "loss": 1.9334,
+      "step": 8778
+    },
+    {
+      "epoch": 0.3713935189102293,
+      "grad_norm": 0.18180030584335327,
+      "learning_rate": 0.001,
+      "loss": 1.6701,
+      "step": 8779
+    },
+    {
+      "epoch": 0.37143582367374567,
+      "grad_norm": 0.19042405486106873,
+      "learning_rate": 0.001,
+      "loss": 2.0225,
+      "step": 8780
+    },
+    {
+      "epoch": 0.371478128437262,
+      "grad_norm": 0.1896408349275589,
+      "learning_rate": 0.001,
+      "loss": 1.7211,
+      "step": 8781
+    },
+    {
+      "epoch": 0.37152043320077843,
+      "grad_norm": 0.21694940328598022,
+      "learning_rate": 0.001,
+      "loss": 2.1461,
+      "step": 8782
+    },
+    {
+      "epoch": 0.3715627379642948,
+      "grad_norm": 0.2233157753944397,
+      "learning_rate": 0.001,
+      "loss": 2.1678,
+      "step": 8783
+    },
+    {
+      "epoch": 0.37160504272781114,
+      "grad_norm": 0.2073362171649933,
+      "learning_rate": 0.001,
+      "loss": 1.9724,
+      "step": 8784
+    },
+    {
+      "epoch": 0.37164734749132755,
+      "grad_norm": 0.3852141201496124,
+      "learning_rate": 0.001,
+      "loss": 2.5631,
+      "step": 8785
+    },
+    {
+      "epoch": 0.3716896522548439,
+      "grad_norm": 0.18070602416992188,
+      "learning_rate": 0.001,
+      "loss": 1.6071,
+      "step": 8786
+    },
+    {
+      "epoch": 0.37173195701836026,
+      "grad_norm": 0.16367939114570618,
+      "learning_rate": 0.001,
+      "loss": 2.2925,
+      "step": 8787
+    },
+    {
+      "epoch": 0.3717742617818766,
+      "grad_norm": 0.1724989265203476,
+      "learning_rate": 0.001,
+      "loss": 3.1093,
+      "step": 8788
+    },
+    {
+      "epoch": 0.371816566545393,
+      "grad_norm": 0.16434898972511292,
+      "learning_rate": 0.001,
+      "loss": 2.4629,
+      "step": 8789
+    },
+    {
+      "epoch": 0.3718588713089094,
+      "grad_norm": 0.2104768455028534,
+      "learning_rate": 0.001,
+      "loss": 2.8752,
+      "step": 8790
+    },
+    {
+      "epoch": 0.37190117607242573,
+      "grad_norm": 0.20089322328567505,
+      "learning_rate": 0.001,
+      "loss": 2.4023,
+      "step": 8791
+    },
+    {
+      "epoch": 0.37194348083594214,
+      "grad_norm": 0.2389354258775711,
+      "learning_rate": 0.001,
+      "loss": 2.2687,
+      "step": 8792
+    },
+    {
+      "epoch": 0.3719857855994585,
+      "grad_norm": 0.47742322087287903,
+      "learning_rate": 0.001,
+      "loss": 1.9946,
+      "step": 8793
+    },
+    {
+      "epoch": 0.37202809036297485,
+      "grad_norm": 0.1540321409702301,
+      "learning_rate": 0.001,
+      "loss": 1.7913,
+      "step": 8794
+    },
+    {
+      "epoch": 0.37207039512649126,
+      "grad_norm": 0.17690719664096832,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 8795
+    },
+    {
+      "epoch": 0.3721126998900076,
+      "grad_norm": 0.23479756712913513,
+      "learning_rate": 0.001,
+      "loss": 3.3659,
+      "step": 8796
+    },
+    {
+      "epoch": 0.37215500465352397,
+      "grad_norm": 0.89059978723526,
+      "learning_rate": 0.001,
+      "loss": 3.5161,
+      "step": 8797
+    },
+    {
+      "epoch": 0.3721973094170404,
+      "grad_norm": 5.015807628631592,
+      "learning_rate": 0.001,
+      "loss": 2.5261,
+      "step": 8798
+    },
+    {
+      "epoch": 0.37223961418055673,
+      "grad_norm": 0.21967126429080963,
+      "learning_rate": 0.001,
+      "loss": 2.07,
+      "step": 8799
+    },
+    {
+      "epoch": 0.3722819189440731,
+      "grad_norm": 0.24731267988681793,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 8800
+    },
+    {
+      "epoch": 0.3723242237075895,
+      "grad_norm": 0.20447765290737152,
+      "learning_rate": 0.001,
+      "loss": 1.8957,
+      "step": 8801
+    },
+    {
+      "epoch": 0.37236652847110585,
+      "grad_norm": 0.2456209510564804,
+      "learning_rate": 0.001,
+      "loss": 1.9937,
+      "step": 8802
+    },
+    {
+      "epoch": 0.3724088332346222,
+      "grad_norm": 0.22434556484222412,
+      "learning_rate": 0.001,
+      "loss": 1.5603,
+      "step": 8803
+    },
+    {
+      "epoch": 0.3724511379981386,
+      "grad_norm": 0.45701247453689575,
+      "learning_rate": 0.001,
+      "loss": 2.2196,
+      "step": 8804
+    },
+    {
+      "epoch": 0.37249344276165497,
+      "grad_norm": 0.2302723377943039,
+      "learning_rate": 0.001,
+      "loss": 1.9974,
+      "step": 8805
+    },
+    {
+      "epoch": 0.3725357475251713,
+      "grad_norm": 0.28021976351737976,
+      "learning_rate": 0.001,
+      "loss": 3.7377,
+      "step": 8806
+    },
+    {
+      "epoch": 0.37257805228868773,
+      "grad_norm": 0.8244222402572632,
+      "learning_rate": 0.001,
+      "loss": 2.1969,
+      "step": 8807
+    },
+    {
+      "epoch": 0.3726203570522041,
+      "grad_norm": 0.20261800289154053,
+      "learning_rate": 0.001,
+      "loss": 3.6426,
+      "step": 8808
+    },
+    {
+      "epoch": 0.37266266181572044,
+      "grad_norm": 0.19825433194637299,
+      "learning_rate": 0.001,
+      "loss": 1.9425,
+      "step": 8809
+    },
+    {
+      "epoch": 0.3727049665792368,
+      "grad_norm": 1.1609710454940796,
+      "learning_rate": 0.001,
+      "loss": 2.1458,
+      "step": 8810
+    },
+    {
+      "epoch": 0.3727472713427532,
+      "grad_norm": 0.5567569136619568,
+      "learning_rate": 0.001,
+      "loss": 1.7775,
+      "step": 8811
+    },
+    {
+      "epoch": 0.37278957610626956,
+      "grad_norm": 0.2688750922679901,
+      "learning_rate": 0.001,
+      "loss": 2.358,
+      "step": 8812
+    },
+    {
+      "epoch": 0.3728318808697859,
+      "grad_norm": 0.33358681201934814,
+      "learning_rate": 0.001,
+      "loss": 2.8446,
+      "step": 8813
+    },
+    {
+      "epoch": 0.3728741856333023,
+      "grad_norm": 0.20433993637561798,
+      "learning_rate": 0.001,
+      "loss": 2.3362,
+      "step": 8814
+    },
+    {
+      "epoch": 0.3729164903968187,
+      "grad_norm": 0.19342532753944397,
+      "learning_rate": 0.001,
+      "loss": 2.8176,
+      "step": 8815
+    },
+    {
+      "epoch": 0.37295879516033503,
+      "grad_norm": 0.23268815875053406,
+      "learning_rate": 0.001,
+      "loss": 1.9077,
+      "step": 8816
+    },
+    {
+      "epoch": 0.37300109992385144,
+      "grad_norm": 0.23792216181755066,
+      "learning_rate": 0.001,
+      "loss": 2.4612,
+      "step": 8817
+    },
+    {
+      "epoch": 0.3730434046873678,
+      "grad_norm": 179.4039764404297,
+      "learning_rate": 0.001,
+      "loss": 3.1672,
+      "step": 8818
+    },
+    {
+      "epoch": 0.37308570945088415,
+      "grad_norm": 0.29416170716285706,
+      "learning_rate": 0.001,
+      "loss": 2.0936,
+      "step": 8819
+    },
+    {
+      "epoch": 0.37312801421440056,
+      "grad_norm": 0.2061353474855423,
+      "learning_rate": 0.001,
+      "loss": 3.3697,
+      "step": 8820
+    },
+    {
+      "epoch": 0.3731703189779169,
+      "grad_norm": 0.22392816841602325,
+      "learning_rate": 0.001,
+      "loss": 1.8806,
+      "step": 8821
+    },
+    {
+      "epoch": 0.37321262374143327,
+      "grad_norm": 0.5261951088905334,
+      "learning_rate": 0.001,
+      "loss": 2.6985,
+      "step": 8822
+    },
+    {
+      "epoch": 0.3732549285049497,
+      "grad_norm": 0.9621092081069946,
+      "learning_rate": 0.001,
+      "loss": 2.2439,
+      "step": 8823
+    },
+    {
+      "epoch": 0.37329723326846603,
+      "grad_norm": 0.20049139857292175,
+      "learning_rate": 0.001,
+      "loss": 2.2026,
+      "step": 8824
+    },
+    {
+      "epoch": 0.3733395380319824,
+      "grad_norm": 1.3945945501327515,
+      "learning_rate": 0.001,
+      "loss": 2.6688,
+      "step": 8825
+    },
+    {
+      "epoch": 0.3733818427954988,
+      "grad_norm": 0.28076982498168945,
+      "learning_rate": 0.001,
+      "loss": 2.9429,
+      "step": 8826
+    },
+    {
+      "epoch": 0.37342414755901515,
+      "grad_norm": 0.45473167300224304,
+      "learning_rate": 0.001,
+      "loss": 2.7468,
+      "step": 8827
+    },
+    {
+      "epoch": 0.3734664523225315,
+      "grad_norm": 1.4571759700775146,
+      "learning_rate": 0.001,
+      "loss": 1.6316,
+      "step": 8828
+    },
+    {
+      "epoch": 0.3735087570860479,
+      "grad_norm": 1.2502529621124268,
+      "learning_rate": 0.001,
+      "loss": 3.0181,
+      "step": 8829
+    },
+    {
+      "epoch": 0.37355106184956427,
+      "grad_norm": 0.26605066657066345,
+      "learning_rate": 0.001,
+      "loss": 2.5819,
+      "step": 8830
+    },
+    {
+      "epoch": 0.3735933666130806,
+      "grad_norm": 0.4681839644908905,
+      "learning_rate": 0.001,
+      "loss": 2.5359,
+      "step": 8831
+    },
+    {
+      "epoch": 0.373635671376597,
+      "grad_norm": 0.24733616411685944,
+      "learning_rate": 0.001,
+      "loss": 3.0378,
+      "step": 8832
+    },
+    {
+      "epoch": 0.3736779761401134,
+      "grad_norm": 0.20918242633342743,
+      "learning_rate": 0.001,
+      "loss": 2.6718,
+      "step": 8833
+    },
+    {
+      "epoch": 0.37372028090362974,
+      "grad_norm": 0.7425049543380737,
+      "learning_rate": 0.001,
+      "loss": 1.8763,
+      "step": 8834
+    },
+    {
+      "epoch": 0.3737625856671461,
+      "grad_norm": 0.1861983686685562,
+      "learning_rate": 0.001,
+      "loss": 2.6386,
+      "step": 8835
+    },
+    {
+      "epoch": 0.3738048904306625,
+      "grad_norm": 0.22254261374473572,
+      "learning_rate": 0.001,
+      "loss": 2.9136,
+      "step": 8836
+    },
+    {
+      "epoch": 0.37384719519417886,
+      "grad_norm": 0.22637084126472473,
+      "learning_rate": 0.001,
+      "loss": 2.5875,
+      "step": 8837
+    },
+    {
+      "epoch": 0.3738894999576952,
+      "grad_norm": 0.2065073400735855,
+      "learning_rate": 0.001,
+      "loss": 2.6558,
+      "step": 8838
+    },
+    {
+      "epoch": 0.3739318047212116,
+      "grad_norm": 0.18488679826259613,
+      "learning_rate": 0.001,
+      "loss": 2.429,
+      "step": 8839
+    },
+    {
+      "epoch": 0.373974109484728,
+      "grad_norm": 0.22156591713428497,
+      "learning_rate": 0.001,
+      "loss": 1.8497,
+      "step": 8840
+    },
+    {
+      "epoch": 0.37401641424824433,
+      "grad_norm": 0.1818188726902008,
+      "learning_rate": 0.001,
+      "loss": 2.1918,
+      "step": 8841
+    },
+    {
+      "epoch": 0.37405871901176074,
+      "grad_norm": 0.1845444142818451,
+      "learning_rate": 0.001,
+      "loss": 1.7402,
+      "step": 8842
+    },
+    {
+      "epoch": 0.3741010237752771,
+      "grad_norm": 0.17082853615283966,
+      "learning_rate": 0.001,
+      "loss": 1.9857,
+      "step": 8843
+    },
+    {
+      "epoch": 0.37414332853879345,
+      "grad_norm": 0.1782820224761963,
+      "learning_rate": 0.001,
+      "loss": 2.0385,
+      "step": 8844
+    },
+    {
+      "epoch": 0.37418563330230986,
+      "grad_norm": 1.1810798645019531,
+      "learning_rate": 0.001,
+      "loss": 2.1752,
+      "step": 8845
+    },
+    {
+      "epoch": 0.3742279380658262,
+      "grad_norm": 0.19883759319782257,
+      "learning_rate": 0.001,
+      "loss": 2.543,
+      "step": 8846
+    },
+    {
+      "epoch": 0.37427024282934257,
+      "grad_norm": 0.47685497999191284,
+      "learning_rate": 0.001,
+      "loss": 2.2209,
+      "step": 8847
+    },
+    {
+      "epoch": 0.374312547592859,
+      "grad_norm": 0.6518121957778931,
+      "learning_rate": 0.001,
+      "loss": 2.8904,
+      "step": 8848
+    },
+    {
+      "epoch": 0.37435485235637533,
+      "grad_norm": 0.1763034462928772,
+      "learning_rate": 0.001,
+      "loss": 2.5656,
+      "step": 8849
+    },
+    {
+      "epoch": 0.3743971571198917,
+      "grad_norm": 13.006291389465332,
+      "learning_rate": 0.001,
+      "loss": 1.7881,
+      "step": 8850
+    },
+    {
+      "epoch": 0.3744394618834081,
+      "grad_norm": 0.25351783633232117,
+      "learning_rate": 0.001,
+      "loss": 2.2294,
+      "step": 8851
+    },
+    {
+      "epoch": 0.37448176664692445,
+      "grad_norm": 10.508666038513184,
+      "learning_rate": 0.001,
+      "loss": 2.6075,
+      "step": 8852
+    },
+    {
+      "epoch": 0.3745240714104408,
+      "grad_norm": 0.23762452602386475,
+      "learning_rate": 0.001,
+      "loss": 2.0531,
+      "step": 8853
+    },
+    {
+      "epoch": 0.37456637617395716,
+      "grad_norm": 0.21458260715007782,
+      "learning_rate": 0.001,
+      "loss": 1.6651,
+      "step": 8854
+    },
+    {
+      "epoch": 0.37460868093747357,
+      "grad_norm": 0.19314922392368317,
+      "learning_rate": 0.001,
+      "loss": 2.2319,
+      "step": 8855
+    },
+    {
+      "epoch": 0.3746509857009899,
+      "grad_norm": 0.21474038064479828,
+      "learning_rate": 0.001,
+      "loss": 1.684,
+      "step": 8856
+    },
+    {
+      "epoch": 0.3746932904645063,
+      "grad_norm": 0.9535976648330688,
+      "learning_rate": 0.001,
+      "loss": 2.1219,
+      "step": 8857
+    },
+    {
+      "epoch": 0.3747355952280227,
+      "grad_norm": 0.3602615296840668,
+      "learning_rate": 0.001,
+      "loss": 2.2903,
+      "step": 8858
+    },
+    {
+      "epoch": 0.37477789999153904,
+      "grad_norm": 0.8784622550010681,
+      "learning_rate": 0.001,
+      "loss": 2.7722,
+      "step": 8859
+    },
+    {
+      "epoch": 0.3748202047550554,
+      "grad_norm": 0.21422924101352692,
+      "learning_rate": 0.001,
+      "loss": 2.2356,
+      "step": 8860
+    },
+    {
+      "epoch": 0.3748625095185718,
+      "grad_norm": 0.3022599220275879,
+      "learning_rate": 0.001,
+      "loss": 2.8531,
+      "step": 8861
+    },
+    {
+      "epoch": 0.37490481428208816,
+      "grad_norm": 0.2399420291185379,
+      "learning_rate": 0.001,
+      "loss": 2.1652,
+      "step": 8862
+    },
+    {
+      "epoch": 0.3749471190456045,
+      "grad_norm": 0.24375419318675995,
+      "learning_rate": 0.001,
+      "loss": 1.7758,
+      "step": 8863
+    },
+    {
+      "epoch": 0.3749894238091209,
+      "grad_norm": 0.2895803451538086,
+      "learning_rate": 0.001,
+      "loss": 1.9264,
+      "step": 8864
+    },
+    {
+      "epoch": 0.3750317285726373,
+      "grad_norm": 0.1909046471118927,
+      "learning_rate": 0.001,
+      "loss": 2.2194,
+      "step": 8865
+    },
+    {
+      "epoch": 0.37507403333615363,
+      "grad_norm": 0.2433093786239624,
+      "learning_rate": 0.001,
+      "loss": 2.556,
+      "step": 8866
+    },
+    {
+      "epoch": 0.37511633809967004,
+      "grad_norm": 0.20016717910766602,
+      "learning_rate": 0.001,
+      "loss": 1.7043,
+      "step": 8867
+    },
+    {
+      "epoch": 0.3751586428631864,
+      "grad_norm": 0.25574326515197754,
+      "learning_rate": 0.001,
+      "loss": 2.0388,
+      "step": 8868
+    },
+    {
+      "epoch": 0.37520094762670275,
+      "grad_norm": 0.2795771062374115,
+      "learning_rate": 0.001,
+      "loss": 1.9324,
+      "step": 8869
+    },
+    {
+      "epoch": 0.37524325239021916,
+      "grad_norm": 0.26534348726272583,
+      "learning_rate": 0.001,
+      "loss": 2.8957,
+      "step": 8870
+    },
+    {
+      "epoch": 0.3752855571537355,
+      "grad_norm": 0.2329631894826889,
+      "learning_rate": 0.001,
+      "loss": 2.7308,
+      "step": 8871
+    },
+    {
+      "epoch": 0.37532786191725187,
+      "grad_norm": 0.21924613416194916,
+      "learning_rate": 0.001,
+      "loss": 2.1649,
+      "step": 8872
+    },
+    {
+      "epoch": 0.3753701666807683,
+      "grad_norm": 1.5583733320236206,
+      "learning_rate": 0.001,
+      "loss": 2.8169,
+      "step": 8873
+    },
+    {
+      "epoch": 0.37541247144428463,
+      "grad_norm": 0.2518419027328491,
+      "learning_rate": 0.001,
+      "loss": 2.3901,
+      "step": 8874
+    },
+    {
+      "epoch": 0.375454776207801,
+      "grad_norm": 0.21858522295951843,
+      "learning_rate": 0.001,
+      "loss": 2.339,
+      "step": 8875
+    },
+    {
+      "epoch": 0.37549708097131734,
+      "grad_norm": 0.4082587957382202,
+      "learning_rate": 0.001,
+      "loss": 2.3889,
+      "step": 8876
+    },
+    {
+      "epoch": 0.37553938573483375,
+      "grad_norm": 0.20391131937503815,
+      "learning_rate": 0.001,
+      "loss": 3.0828,
+      "step": 8877
+    },
+    {
+      "epoch": 0.3755816904983501,
+      "grad_norm": 0.21462643146514893,
+      "learning_rate": 0.001,
+      "loss": 2.1877,
+      "step": 8878
+    },
+    {
+      "epoch": 0.37562399526186646,
+      "grad_norm": 0.2579302489757538,
+      "learning_rate": 0.001,
+      "loss": 2.0629,
+      "step": 8879
+    },
+    {
+      "epoch": 0.37566630002538287,
+      "grad_norm": 0.17081435024738312,
+      "learning_rate": 0.001,
+      "loss": 2.2482,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3757086047888992,
+      "grad_norm": 0.21092277765274048,
+      "learning_rate": 0.001,
+      "loss": 2.7822,
+      "step": 8881
+    },
+    {
+      "epoch": 0.3757509095524156,
+      "grad_norm": 0.5651504397392273,
+      "learning_rate": 0.001,
+      "loss": 2.5021,
+      "step": 8882
+    },
+    {
+      "epoch": 0.375793214315932,
+      "grad_norm": 0.23301394283771515,
+      "learning_rate": 0.001,
+      "loss": 2.0126,
+      "step": 8883
+    },
+    {
+      "epoch": 0.37583551907944834,
+      "grad_norm": 2.354231595993042,
+      "learning_rate": 0.001,
+      "loss": 2.4646,
+      "step": 8884
+    },
+    {
+      "epoch": 0.3758778238429647,
+      "grad_norm": 2.867743492126465,
+      "learning_rate": 0.001,
+      "loss": 2.1439,
+      "step": 8885
+    },
+    {
+      "epoch": 0.3759201286064811,
+      "grad_norm": 1.451896071434021,
+      "learning_rate": 0.001,
+      "loss": 2.6435,
+      "step": 8886
+    },
+    {
+      "epoch": 0.37596243336999746,
+      "grad_norm": 0.2039012610912323,
+      "learning_rate": 0.001,
+      "loss": 2.3304,
+      "step": 8887
+    },
+    {
+      "epoch": 0.3760047381335138,
+      "grad_norm": 0.4442058801651001,
+      "learning_rate": 0.001,
+      "loss": 2.9109,
+      "step": 8888
+    },
+    {
+      "epoch": 0.3760470428970302,
+      "grad_norm": 0.2005094289779663,
+      "learning_rate": 0.001,
+      "loss": 3.5619,
+      "step": 8889
+    },
+    {
+      "epoch": 0.3760893476605466,
+      "grad_norm": 3.216552495956421,
+      "learning_rate": 0.001,
+      "loss": 2.8794,
+      "step": 8890
+    },
+    {
+      "epoch": 0.37613165242406293,
+      "grad_norm": 0.23342199623584747,
+      "learning_rate": 0.001,
+      "loss": 2.6378,
+      "step": 8891
+    },
+    {
+      "epoch": 0.37617395718757934,
+      "grad_norm": 0.23157887160778046,
+      "learning_rate": 0.001,
+      "loss": 2.9445,
+      "step": 8892
+    },
+    {
+      "epoch": 0.3762162619510957,
+      "grad_norm": 0.2705920338630676,
+      "learning_rate": 0.001,
+      "loss": 2.2957,
+      "step": 8893
+    },
+    {
+      "epoch": 0.37625856671461205,
+      "grad_norm": 0.28149649500846863,
+      "learning_rate": 0.001,
+      "loss": 1.7105,
+      "step": 8894
+    },
+    {
+      "epoch": 0.37630087147812846,
+      "grad_norm": 1.1188766956329346,
+      "learning_rate": 0.001,
+      "loss": 2.2677,
+      "step": 8895
+    },
+    {
+      "epoch": 0.3763431762416448,
+      "grad_norm": 0.8402219414710999,
+      "learning_rate": 0.001,
+      "loss": 3.4824,
+      "step": 8896
+    },
+    {
+      "epoch": 0.37638548100516117,
+      "grad_norm": 0.2254517376422882,
+      "learning_rate": 0.001,
+      "loss": 2.4484,
+      "step": 8897
+    },
+    {
+      "epoch": 0.3764277857686776,
+      "grad_norm": 0.18616388738155365,
+      "learning_rate": 0.001,
+      "loss": 1.8755,
+      "step": 8898
+    },
+    {
+      "epoch": 0.37647009053219394,
+      "grad_norm": 0.16204172372817993,
+      "learning_rate": 0.001,
+      "loss": 2.3035,
+      "step": 8899
+    },
+    {
+      "epoch": 0.3765123952957103,
+      "grad_norm": 0.655873715877533,
+      "learning_rate": 0.001,
+      "loss": 2.7063,
+      "step": 8900
+    },
+    {
+      "epoch": 0.37655470005922664,
+      "grad_norm": 0.31336209177970886,
+      "learning_rate": 0.001,
+      "loss": 2.2015,
+      "step": 8901
+    },
+    {
+      "epoch": 0.37659700482274305,
+      "grad_norm": 0.21138517558574677,
+      "learning_rate": 0.001,
+      "loss": 2.0808,
+      "step": 8902
+    },
+    {
+      "epoch": 0.3766393095862594,
+      "grad_norm": 0.32678109407424927,
+      "learning_rate": 0.001,
+      "loss": 3.5257,
+      "step": 8903
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.17282311618328094,
+      "learning_rate": 0.001,
+      "loss": 1.6933,
+      "step": 8904
+    },
+    {
+      "epoch": 0.37672391911329217,
+      "grad_norm": 0.19820040464401245,
+      "learning_rate": 0.001,
+      "loss": 1.8809,
+      "step": 8905
+    },
+    {
+      "epoch": 0.3767662238768085,
+      "grad_norm": 0.23431384563446045,
+      "learning_rate": 0.001,
+      "loss": 1.9761,
+      "step": 8906
+    },
+    {
+      "epoch": 0.3768085286403249,
+      "grad_norm": 0.275423139333725,
+      "learning_rate": 0.001,
+      "loss": 2.9449,
+      "step": 8907
+    },
+    {
+      "epoch": 0.3768508334038413,
+      "grad_norm": 0.4699144661426544,
+      "learning_rate": 0.001,
+      "loss": 1.8988,
+      "step": 8908
+    },
+    {
+      "epoch": 0.37689313816735764,
+      "grad_norm": 0.17784275114536285,
+      "learning_rate": 0.001,
+      "loss": 2.1766,
+      "step": 8909
+    },
+    {
+      "epoch": 0.376935442930874,
+      "grad_norm": 0.25600793957710266,
+      "learning_rate": 0.001,
+      "loss": 2.2169,
+      "step": 8910
+    },
+    {
+      "epoch": 0.3769777476943904,
+      "grad_norm": 0.19872885942459106,
+      "learning_rate": 0.001,
+      "loss": 1.8248,
+      "step": 8911
+    },
+    {
+      "epoch": 0.37702005245790676,
+      "grad_norm": 0.6820826530456543,
+      "learning_rate": 0.001,
+      "loss": 2.1578,
+      "step": 8912
+    },
+    {
+      "epoch": 0.3770623572214231,
+      "grad_norm": 0.17017528414726257,
+      "learning_rate": 0.001,
+      "loss": 3.678,
+      "step": 8913
+    },
+    {
+      "epoch": 0.3771046619849395,
+      "grad_norm": 0.1723102331161499,
+      "learning_rate": 0.001,
+      "loss": 3.2953,
+      "step": 8914
+    },
+    {
+      "epoch": 0.3771469667484559,
+      "grad_norm": 0.17692267894744873,
+      "learning_rate": 0.001,
+      "loss": 2.2835,
+      "step": 8915
+    },
+    {
+      "epoch": 0.37718927151197224,
+      "grad_norm": 0.18386520445346832,
+      "learning_rate": 0.001,
+      "loss": 1.7949,
+      "step": 8916
+    },
+    {
+      "epoch": 0.37723157627548864,
+      "grad_norm": 0.17489704489707947,
+      "learning_rate": 0.001,
+      "loss": 2.8251,
+      "step": 8917
+    },
+    {
+      "epoch": 0.377273881039005,
+      "grad_norm": 0.18046604096889496,
+      "learning_rate": 0.001,
+      "loss": 1.9294,
+      "step": 8918
+    },
+    {
+      "epoch": 0.37731618580252135,
+      "grad_norm": 0.19288641214370728,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 8919
+    },
+    {
+      "epoch": 0.37735849056603776,
+      "grad_norm": 20.963592529296875,
+      "learning_rate": 0.001,
+      "loss": 1.7886,
+      "step": 8920
+    },
+    {
+      "epoch": 0.3774007953295541,
+      "grad_norm": 0.20476023852825165,
+      "learning_rate": 0.001,
+      "loss": 2.0257,
+      "step": 8921
+    },
+    {
+      "epoch": 0.37744310009307047,
+      "grad_norm": 0.17102450132369995,
+      "learning_rate": 0.001,
+      "loss": 3.2415,
+      "step": 8922
+    },
+    {
+      "epoch": 0.3774854048565868,
+      "grad_norm": 0.24836905300617218,
+      "learning_rate": 0.001,
+      "loss": 1.6505,
+      "step": 8923
+    },
+    {
+      "epoch": 0.37752770962010324,
+      "grad_norm": 0.2269943505525589,
+      "learning_rate": 0.001,
+      "loss": 1.9437,
+      "step": 8924
+    },
+    {
+      "epoch": 0.3775700143836196,
+      "grad_norm": 0.18366478383541107,
+      "learning_rate": 0.001,
+      "loss": 1.9654,
+      "step": 8925
+    },
+    {
+      "epoch": 0.37761231914713594,
+      "grad_norm": 0.19033034145832062,
+      "learning_rate": 0.001,
+      "loss": 1.7305,
+      "step": 8926
+    },
+    {
+      "epoch": 0.37765462391065235,
+      "grad_norm": 0.16039331257343292,
+      "learning_rate": 0.001,
+      "loss": 2.4644,
+      "step": 8927
+    },
+    {
+      "epoch": 0.3776969286741687,
+      "grad_norm": 0.15948708355426788,
+      "learning_rate": 0.001,
+      "loss": 3.5417,
+      "step": 8928
+    },
+    {
+      "epoch": 0.37773923343768506,
+      "grad_norm": 0.16082708537578583,
+      "learning_rate": 0.001,
+      "loss": 2.1491,
+      "step": 8929
+    },
+    {
+      "epoch": 0.37778153820120147,
+      "grad_norm": 0.21579056978225708,
+      "learning_rate": 0.001,
+      "loss": 2.8686,
+      "step": 8930
+    },
+    {
+      "epoch": 0.3778238429647178,
+      "grad_norm": 0.18891681730747223,
+      "learning_rate": 0.001,
+      "loss": 2.1063,
+      "step": 8931
+    },
+    {
+      "epoch": 0.3778661477282342,
+      "grad_norm": 0.2987871468067169,
+      "learning_rate": 0.001,
+      "loss": 2.1846,
+      "step": 8932
+    },
+    {
+      "epoch": 0.3779084524917506,
+      "grad_norm": 0.19955849647521973,
+      "learning_rate": 0.001,
+      "loss": 3.7915,
+      "step": 8933
+    },
+    {
+      "epoch": 0.37795075725526694,
+      "grad_norm": 0.4888962507247925,
+      "learning_rate": 0.001,
+      "loss": 1.5937,
+      "step": 8934
+    },
+    {
+      "epoch": 0.3779930620187833,
+      "grad_norm": 0.23087970912456512,
+      "learning_rate": 0.001,
+      "loss": 2.2461,
+      "step": 8935
+    },
+    {
+      "epoch": 0.3780353667822997,
+      "grad_norm": 0.48626619577407837,
+      "learning_rate": 0.001,
+      "loss": 2.6076,
+      "step": 8936
+    },
+    {
+      "epoch": 0.37807767154581606,
+      "grad_norm": 0.21073201298713684,
+      "learning_rate": 0.001,
+      "loss": 1.6726,
+      "step": 8937
+    },
+    {
+      "epoch": 0.3781199763093324,
+      "grad_norm": 0.166831374168396,
+      "learning_rate": 0.001,
+      "loss": 1.709,
+      "step": 8938
+    },
+    {
+      "epoch": 0.3781622810728488,
+      "grad_norm": 0.19128407537937164,
+      "learning_rate": 0.001,
+      "loss": 2.1098,
+      "step": 8939
+    },
+    {
+      "epoch": 0.3782045858363652,
+      "grad_norm": 0.19305849075317383,
+      "learning_rate": 0.001,
+      "loss": 1.8852,
+      "step": 8940
+    },
+    {
+      "epoch": 0.37824689059988154,
+      "grad_norm": 0.18565382063388824,
+      "learning_rate": 0.001,
+      "loss": 2.1476,
+      "step": 8941
+    },
+    {
+      "epoch": 0.37828919536339795,
+      "grad_norm": 0.25521063804626465,
+      "learning_rate": 0.001,
+      "loss": 2.7811,
+      "step": 8942
+    },
+    {
+      "epoch": 0.3783315001269143,
+      "grad_norm": 0.49443572759628296,
+      "learning_rate": 0.001,
+      "loss": 1.8314,
+      "step": 8943
+    },
+    {
+      "epoch": 0.37837380489043065,
+      "grad_norm": 0.24259395897388458,
+      "learning_rate": 0.001,
+      "loss": 1.8752,
+      "step": 8944
+    },
+    {
+      "epoch": 0.378416109653947,
+      "grad_norm": 0.21471655368804932,
+      "learning_rate": 0.001,
+      "loss": 1.8431,
+      "step": 8945
+    },
+    {
+      "epoch": 0.3784584144174634,
+      "grad_norm": 0.34845396876335144,
+      "learning_rate": 0.001,
+      "loss": 2.63,
+      "step": 8946
+    },
+    {
+      "epoch": 0.37850071918097977,
+      "grad_norm": 0.4037695825099945,
+      "learning_rate": 0.001,
+      "loss": 2.641,
+      "step": 8947
+    },
+    {
+      "epoch": 0.3785430239444961,
+      "grad_norm": 0.1841561496257782,
+      "learning_rate": 0.001,
+      "loss": 2.0631,
+      "step": 8948
+    },
+    {
+      "epoch": 0.37858532870801254,
+      "grad_norm": 0.5983163714408875,
+      "learning_rate": 0.001,
+      "loss": 2.3617,
+      "step": 8949
+    },
+    {
+      "epoch": 0.3786276334715289,
+      "grad_norm": 0.1743028163909912,
+      "learning_rate": 0.001,
+      "loss": 2.2375,
+      "step": 8950
+    },
+    {
+      "epoch": 0.37866993823504524,
+      "grad_norm": 0.24144525825977325,
+      "learning_rate": 0.001,
+      "loss": 1.9207,
+      "step": 8951
+    },
+    {
+      "epoch": 0.37871224299856165,
+      "grad_norm": 0.19788867235183716,
+      "learning_rate": 0.001,
+      "loss": 2.8912,
+      "step": 8952
+    },
+    {
+      "epoch": 0.378754547762078,
+      "grad_norm": 0.2542145550251007,
+      "learning_rate": 0.001,
+      "loss": 1.792,
+      "step": 8953
+    },
+    {
+      "epoch": 0.37879685252559436,
+      "grad_norm": 0.19751232862472534,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 8954
+    },
+    {
+      "epoch": 0.3788391572891108,
+      "grad_norm": 0.3706015944480896,
+      "learning_rate": 0.001,
+      "loss": 2.3522,
+      "step": 8955
+    },
+    {
+      "epoch": 0.3788814620526271,
+      "grad_norm": 0.20275190472602844,
+      "learning_rate": 0.001,
+      "loss": 1.8339,
+      "step": 8956
+    },
+    {
+      "epoch": 0.3789237668161435,
+      "grad_norm": 0.18066799640655518,
+      "learning_rate": 0.001,
+      "loss": 1.6604,
+      "step": 8957
+    },
+    {
+      "epoch": 0.3789660715796599,
+      "grad_norm": 0.20664532482624054,
+      "learning_rate": 0.001,
+      "loss": 2.1049,
+      "step": 8958
+    },
+    {
+      "epoch": 0.37900837634317625,
+      "grad_norm": 0.15460434556007385,
+      "learning_rate": 0.001,
+      "loss": 2.0694,
+      "step": 8959
+    },
+    {
+      "epoch": 0.3790506811066926,
+      "grad_norm": 0.177540585398674,
+      "learning_rate": 0.001,
+      "loss": 2.5301,
+      "step": 8960
+    },
+    {
+      "epoch": 0.379092985870209,
+      "grad_norm": 0.21939094364643097,
+      "learning_rate": 0.001,
+      "loss": 2.1656,
+      "step": 8961
+    },
+    {
+      "epoch": 0.37913529063372536,
+      "grad_norm": 0.17154373228549957,
+      "learning_rate": 0.001,
+      "loss": 3.5702,
+      "step": 8962
+    },
+    {
+      "epoch": 0.3791775953972417,
+      "grad_norm": 1.2005940675735474,
+      "learning_rate": 0.001,
+      "loss": 3.8547,
+      "step": 8963
+    },
+    {
+      "epoch": 0.3792199001607581,
+      "grad_norm": 0.20012064278125763,
+      "learning_rate": 0.001,
+      "loss": 2.1547,
+      "step": 8964
+    },
+    {
+      "epoch": 0.3792622049242745,
+      "grad_norm": 0.2005445957183838,
+      "learning_rate": 0.001,
+      "loss": 1.9539,
+      "step": 8965
+    },
+    {
+      "epoch": 0.37930450968779084,
+      "grad_norm": 0.17973119020462036,
+      "learning_rate": 0.001,
+      "loss": 2.0098,
+      "step": 8966
+    },
+    {
+      "epoch": 0.3793468144513072,
+      "grad_norm": 0.16720591485500336,
+      "learning_rate": 0.001,
+      "loss": 1.3429,
+      "step": 8967
+    },
+    {
+      "epoch": 0.3793891192148236,
+      "grad_norm": 1.0415222644805908,
+      "learning_rate": 0.001,
+      "loss": 2.3825,
+      "step": 8968
+    },
+    {
+      "epoch": 0.37943142397833995,
+      "grad_norm": 0.19271378219127655,
+      "learning_rate": 0.001,
+      "loss": 1.7486,
+      "step": 8969
+    },
+    {
+      "epoch": 0.3794737287418563,
+      "grad_norm": 0.26010555028915405,
+      "learning_rate": 0.001,
+      "loss": 2.988,
+      "step": 8970
+    },
+    {
+      "epoch": 0.3795160335053727,
+      "grad_norm": 1.41874361038208,
+      "learning_rate": 0.001,
+      "loss": 2.4324,
+      "step": 8971
+    },
+    {
+      "epoch": 0.3795583382688891,
+      "grad_norm": 0.18716134130954742,
+      "learning_rate": 0.001,
+      "loss": 2.2786,
+      "step": 8972
+    },
+    {
+      "epoch": 0.3796006430324054,
+      "grad_norm": 0.21621747314929962,
+      "learning_rate": 0.001,
+      "loss": 2.0158,
+      "step": 8973
+    },
+    {
+      "epoch": 0.37964294779592184,
+      "grad_norm": 0.21100404858589172,
+      "learning_rate": 0.001,
+      "loss": 2.6913,
+      "step": 8974
+    },
+    {
+      "epoch": 0.3796852525594382,
+      "grad_norm": 0.23763734102249146,
+      "learning_rate": 0.001,
+      "loss": 2.5987,
+      "step": 8975
+    },
+    {
+      "epoch": 0.37972755732295455,
+      "grad_norm": 0.24332080781459808,
+      "learning_rate": 0.001,
+      "loss": 2.6919,
+      "step": 8976
+    },
+    {
+      "epoch": 0.37976986208647096,
+      "grad_norm": 0.27417856454849243,
+      "learning_rate": 0.001,
+      "loss": 2.2423,
+      "step": 8977
+    },
+    {
+      "epoch": 0.3798121668499873,
+      "grad_norm": 0.19818872213363647,
+      "learning_rate": 0.001,
+      "loss": 2.0502,
+      "step": 8978
+    },
+    {
+      "epoch": 0.37985447161350366,
+      "grad_norm": 0.16945703327655792,
+      "learning_rate": 0.001,
+      "loss": 2.848,
+      "step": 8979
+    },
+    {
+      "epoch": 0.3798967763770201,
+      "grad_norm": 0.15949320793151855,
+      "learning_rate": 0.001,
+      "loss": 1.7612,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3799390811405364,
+      "grad_norm": 0.17047828435897827,
+      "learning_rate": 0.001,
+      "loss": 1.9685,
+      "step": 8981
+    },
+    {
+      "epoch": 0.3799813859040528,
+      "grad_norm": 0.19749274849891663,
+      "learning_rate": 0.001,
+      "loss": 2.5775,
+      "step": 8982
+    },
+    {
+      "epoch": 0.3800236906675692,
+      "grad_norm": 0.19193464517593384,
+      "learning_rate": 0.001,
+      "loss": 2.1396,
+      "step": 8983
+    },
+    {
+      "epoch": 0.38006599543108555,
+      "grad_norm": 2.228489398956299,
+      "learning_rate": 0.001,
+      "loss": 2.439,
+      "step": 8984
+    },
+    {
+      "epoch": 0.3801083001946019,
+      "grad_norm": 0.18555498123168945,
+      "learning_rate": 0.001,
+      "loss": 2.0835,
+      "step": 8985
+    },
+    {
+      "epoch": 0.3801506049581183,
+      "grad_norm": 0.4127380847930908,
+      "learning_rate": 0.001,
+      "loss": 1.7705,
+      "step": 8986
+    },
+    {
+      "epoch": 0.38019290972163466,
+      "grad_norm": 0.2189057320356369,
+      "learning_rate": 0.001,
+      "loss": 2.3658,
+      "step": 8987
+    },
+    {
+      "epoch": 0.380235214485151,
+      "grad_norm": 0.8270202279090881,
+      "learning_rate": 0.001,
+      "loss": 2.0476,
+      "step": 8988
+    },
+    {
+      "epoch": 0.3802775192486674,
+      "grad_norm": 0.2435636818408966,
+      "learning_rate": 0.001,
+      "loss": 2.236,
+      "step": 8989
+    },
+    {
+      "epoch": 0.3803198240121838,
+      "grad_norm": 0.40932199358940125,
+      "learning_rate": 0.001,
+      "loss": 2.5558,
+      "step": 8990
+    },
+    {
+      "epoch": 0.38036212877570014,
+      "grad_norm": 0.5411979556083679,
+      "learning_rate": 0.001,
+      "loss": 2.7722,
+      "step": 8991
+    },
+    {
+      "epoch": 0.3804044335392165,
+      "grad_norm": 0.28621330857276917,
+      "learning_rate": 0.001,
+      "loss": 1.9874,
+      "step": 8992
+    },
+    {
+      "epoch": 0.3804467383027329,
+      "grad_norm": 0.20250999927520752,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 8993
+    },
+    {
+      "epoch": 0.38048904306624926,
+      "grad_norm": 0.7544026970863342,
+      "learning_rate": 0.001,
+      "loss": 2.3332,
+      "step": 8994
+    },
+    {
+      "epoch": 0.3805313478297656,
+      "grad_norm": 0.20567587018013,
+      "learning_rate": 0.001,
+      "loss": 1.698,
+      "step": 8995
+    },
+    {
+      "epoch": 0.380573652593282,
+      "grad_norm": 0.20252707600593567,
+      "learning_rate": 0.001,
+      "loss": 1.5407,
+      "step": 8996
+    },
+    {
+      "epoch": 0.3806159573567984,
+      "grad_norm": 0.3609828054904938,
+      "learning_rate": 0.001,
+      "loss": 3.2147,
+      "step": 8997
+    },
+    {
+      "epoch": 0.3806582621203147,
+      "grad_norm": 0.18369510769844055,
+      "learning_rate": 0.001,
+      "loss": 1.6158,
+      "step": 8998
+    },
+    {
+      "epoch": 0.38070056688383114,
+      "grad_norm": 3.380438804626465,
+      "learning_rate": 0.001,
+      "loss": 2.0616,
+      "step": 8999
+    },
+    {
+      "epoch": 0.3807428716473475,
+      "grad_norm": 0.17703545093536377,
+      "learning_rate": 0.001,
+      "loss": 2.9286,
+      "step": 9000
+    },
+    {
+      "epoch": 0.38078517641086385,
+      "grad_norm": 0.601148247718811,
+      "learning_rate": 0.001,
+      "loss": 2.1041,
+      "step": 9001
+    },
+    {
+      "epoch": 0.38082748117438026,
+      "grad_norm": 0.1904536634683609,
+      "learning_rate": 0.001,
+      "loss": 2.2669,
+      "step": 9002
+    },
+    {
+      "epoch": 0.3808697859378966,
+      "grad_norm": 0.5615403056144714,
+      "learning_rate": 0.001,
+      "loss": 2.1864,
+      "step": 9003
+    },
+    {
+      "epoch": 0.38091209070141296,
+      "grad_norm": 0.18499788641929626,
+      "learning_rate": 0.001,
+      "loss": 1.82,
+      "step": 9004
+    },
+    {
+      "epoch": 0.3809543954649294,
+      "grad_norm": 0.22329087555408478,
+      "learning_rate": 0.001,
+      "loss": 1.9953,
+      "step": 9005
+    },
+    {
+      "epoch": 0.38099670022844573,
+      "grad_norm": 0.21240036189556122,
+      "learning_rate": 0.001,
+      "loss": 2.3098,
+      "step": 9006
+    },
+    {
+      "epoch": 0.3810390049919621,
+      "grad_norm": 0.171017587184906,
+      "learning_rate": 0.001,
+      "loss": 2.0382,
+      "step": 9007
+    },
+    {
+      "epoch": 0.3810813097554785,
+      "grad_norm": 0.22148646414279938,
+      "learning_rate": 0.001,
+      "loss": 2.1392,
+      "step": 9008
+    },
+    {
+      "epoch": 0.38112361451899485,
+      "grad_norm": 0.18616536259651184,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 9009
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 10.153366088867188,
+      "learning_rate": 0.001,
+      "loss": 2.767,
+      "step": 9010
+    },
+    {
+      "epoch": 0.38120822404602756,
+      "grad_norm": 0.19627103209495544,
+      "learning_rate": 0.001,
+      "loss": 1.7014,
+      "step": 9011
+    },
+    {
+      "epoch": 0.38125052880954396,
+      "grad_norm": 0.2045442759990692,
+      "learning_rate": 0.001,
+      "loss": 1.4853,
+      "step": 9012
+    },
+    {
+      "epoch": 0.3812928335730603,
+      "grad_norm": 0.8567882180213928,
+      "learning_rate": 0.001,
+      "loss": 2.0106,
+      "step": 9013
+    },
+    {
+      "epoch": 0.3813351383365767,
+      "grad_norm": 0.21957813203334808,
+      "learning_rate": 0.001,
+      "loss": 2.1993,
+      "step": 9014
+    },
+    {
+      "epoch": 0.3813774431000931,
+      "grad_norm": 0.20184127986431122,
+      "learning_rate": 0.001,
+      "loss": 2.0607,
+      "step": 9015
+    },
+    {
+      "epoch": 0.38141974786360944,
+      "grad_norm": 1.3871818780899048,
+      "learning_rate": 0.001,
+      "loss": 1.851,
+      "step": 9016
+    },
+    {
+      "epoch": 0.3814620526271258,
+      "grad_norm": 0.26643404364585876,
+      "learning_rate": 0.001,
+      "loss": 2.1513,
+      "step": 9017
+    },
+    {
+      "epoch": 0.3815043573906422,
+      "grad_norm": 0.3258124887943268,
+      "learning_rate": 0.001,
+      "loss": 2.6778,
+      "step": 9018
+    },
+    {
+      "epoch": 0.38154666215415856,
+      "grad_norm": 0.2190716564655304,
+      "learning_rate": 0.001,
+      "loss": 2.1382,
+      "step": 9019
+    },
+    {
+      "epoch": 0.3815889669176749,
+      "grad_norm": 0.21376149356365204,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 9020
+    },
+    {
+      "epoch": 0.3816312716811913,
+      "grad_norm": 0.22805331647396088,
+      "learning_rate": 0.001,
+      "loss": 2.3052,
+      "step": 9021
+    },
+    {
+      "epoch": 0.3816735764447077,
+      "grad_norm": 0.18218421936035156,
+      "learning_rate": 0.001,
+      "loss": 2.7028,
+      "step": 9022
+    },
+    {
+      "epoch": 0.38171588120822403,
+      "grad_norm": 0.19406476616859436,
+      "learning_rate": 0.001,
+      "loss": 2.589,
+      "step": 9023
+    },
+    {
+      "epoch": 0.38175818597174044,
+      "grad_norm": 5.224491119384766,
+      "learning_rate": 0.001,
+      "loss": 2.0287,
+      "step": 9024
+    },
+    {
+      "epoch": 0.3818004907352568,
+      "grad_norm": 0.2605280578136444,
+      "learning_rate": 0.001,
+      "loss": 2.3784,
+      "step": 9025
+    },
+    {
+      "epoch": 0.38184279549877315,
+      "grad_norm": 2.6531484127044678,
+      "learning_rate": 0.001,
+      "loss": 2.6587,
+      "step": 9026
+    },
+    {
+      "epoch": 0.38188510026228956,
+      "grad_norm": 0.19211426377296448,
+      "learning_rate": 0.001,
+      "loss": 2.5609,
+      "step": 9027
+    },
+    {
+      "epoch": 0.3819274050258059,
+      "grad_norm": 0.31936317682266235,
+      "learning_rate": 0.001,
+      "loss": 2.2436,
+      "step": 9028
+    },
+    {
+      "epoch": 0.38196970978932226,
+      "grad_norm": 0.27096110582351685,
+      "learning_rate": 0.001,
+      "loss": 2.763,
+      "step": 9029
+    },
+    {
+      "epoch": 0.3820120145528387,
+      "grad_norm": 0.27728742361068726,
+      "learning_rate": 0.001,
+      "loss": 2.2142,
+      "step": 9030
+    },
+    {
+      "epoch": 0.38205431931635503,
+      "grad_norm": 0.23881641030311584,
+      "learning_rate": 0.001,
+      "loss": 2.6463,
+      "step": 9031
+    },
+    {
+      "epoch": 0.3820966240798714,
+      "grad_norm": 1.6963469982147217,
+      "learning_rate": 0.001,
+      "loss": 1.7634,
+      "step": 9032
+    },
+    {
+      "epoch": 0.3821389288433878,
+      "grad_norm": 0.9256619215011597,
+      "learning_rate": 0.001,
+      "loss": 2.8096,
+      "step": 9033
+    },
+    {
+      "epoch": 0.38218123360690415,
+      "grad_norm": 0.23679040372371674,
+      "learning_rate": 0.001,
+      "loss": 2.5447,
+      "step": 9034
+    },
+    {
+      "epoch": 0.3822235383704205,
+      "grad_norm": 0.266876220703125,
+      "learning_rate": 0.001,
+      "loss": 2.1042,
+      "step": 9035
+    },
+    {
+      "epoch": 0.38226584313393686,
+      "grad_norm": 0.47037845849990845,
+      "learning_rate": 0.001,
+      "loss": 4.1207,
+      "step": 9036
+    },
+    {
+      "epoch": 0.38230814789745327,
+      "grad_norm": 0.3769329786300659,
+      "learning_rate": 0.001,
+      "loss": 3.3571,
+      "step": 9037
+    },
+    {
+      "epoch": 0.3823504526609696,
+      "grad_norm": 1.5549075603485107,
+      "learning_rate": 0.001,
+      "loss": 3.0967,
+      "step": 9038
+    },
+    {
+      "epoch": 0.382392757424486,
+      "grad_norm": 0.48421478271484375,
+      "learning_rate": 0.001,
+      "loss": 2.0338,
+      "step": 9039
+    },
+    {
+      "epoch": 0.3824350621880024,
+      "grad_norm": 0.22943010926246643,
+      "learning_rate": 0.001,
+      "loss": 2.5297,
+      "step": 9040
+    },
+    {
+      "epoch": 0.38247736695151874,
+      "grad_norm": 0.2692186236381531,
+      "learning_rate": 0.001,
+      "loss": 2.6332,
+      "step": 9041
+    },
+    {
+      "epoch": 0.3825196717150351,
+      "grad_norm": 0.31615620851516724,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 9042
+    },
+    {
+      "epoch": 0.3825619764785515,
+      "grad_norm": 0.2741483449935913,
+      "learning_rate": 0.001,
+      "loss": 1.9714,
+      "step": 9043
+    },
+    {
+      "epoch": 0.38260428124206786,
+      "grad_norm": 0.21158187091350555,
+      "learning_rate": 0.001,
+      "loss": 1.8752,
+      "step": 9044
+    },
+    {
+      "epoch": 0.3826465860055842,
+      "grad_norm": 9.120562553405762,
+      "learning_rate": 0.001,
+      "loss": 3.4247,
+      "step": 9045
+    },
+    {
+      "epoch": 0.3826888907691006,
+      "grad_norm": 0.3110532760620117,
+      "learning_rate": 0.001,
+      "loss": 2.6979,
+      "step": 9046
+    },
+    {
+      "epoch": 0.382731195532617,
+      "grad_norm": 0.4028329849243164,
+      "learning_rate": 0.001,
+      "loss": 3.1487,
+      "step": 9047
+    },
+    {
+      "epoch": 0.38277350029613333,
+      "grad_norm": 13.722456932067871,
+      "learning_rate": 0.001,
+      "loss": 3.2165,
+      "step": 9048
+    },
+    {
+      "epoch": 0.38281580505964974,
+      "grad_norm": 0.2979806661605835,
+      "learning_rate": 0.001,
+      "loss": 1.7009,
+      "step": 9049
+    },
+    {
+      "epoch": 0.3828581098231661,
+      "grad_norm": 0.2089688926935196,
+      "learning_rate": 0.001,
+      "loss": 1.9347,
+      "step": 9050
+    },
+    {
+      "epoch": 0.38290041458668245,
+      "grad_norm": 0.19975943863391876,
+      "learning_rate": 0.001,
+      "loss": 2.757,
+      "step": 9051
+    },
+    {
+      "epoch": 0.38294271935019886,
+      "grad_norm": 0.1705322414636612,
+      "learning_rate": 0.001,
+      "loss": 1.8747,
+      "step": 9052
+    },
+    {
+      "epoch": 0.3829850241137152,
+      "grad_norm": 0.2010183334350586,
+      "learning_rate": 0.001,
+      "loss": 1.6273,
+      "step": 9053
+    },
+    {
+      "epoch": 0.38302732887723157,
+      "grad_norm": 9.024285316467285,
+      "learning_rate": 0.001,
+      "loss": 1.99,
+      "step": 9054
+    },
+    {
+      "epoch": 0.383069633640748,
+      "grad_norm": 0.34857797622680664,
+      "learning_rate": 0.001,
+      "loss": 2.5817,
+      "step": 9055
+    },
+    {
+      "epoch": 0.38311193840426433,
+      "grad_norm": 0.5103548169136047,
+      "learning_rate": 0.001,
+      "loss": 2.021,
+      "step": 9056
+    },
+    {
+      "epoch": 0.3831542431677807,
+      "grad_norm": 2.387510299682617,
+      "learning_rate": 0.001,
+      "loss": 2.2303,
+      "step": 9057
+    },
+    {
+      "epoch": 0.38319654793129704,
+      "grad_norm": 0.2850176990032196,
+      "learning_rate": 0.001,
+      "loss": 2.4001,
+      "step": 9058
+    },
+    {
+      "epoch": 0.38323885269481345,
+      "grad_norm": 0.6285756826400757,
+      "learning_rate": 0.001,
+      "loss": 1.8074,
+      "step": 9059
+    },
+    {
+      "epoch": 0.3832811574583298,
+      "grad_norm": 0.5569390654563904,
+      "learning_rate": 0.001,
+      "loss": 2.4452,
+      "step": 9060
+    },
+    {
+      "epoch": 0.38332346222184616,
+      "grad_norm": 5.331487655639648,
+      "learning_rate": 0.001,
+      "loss": 2.7975,
+      "step": 9061
+    },
+    {
+      "epoch": 0.38336576698536257,
+      "grad_norm": 0.30871254205703735,
+      "learning_rate": 0.001,
+      "loss": 2.3949,
+      "step": 9062
+    },
+    {
+      "epoch": 0.3834080717488789,
+      "grad_norm": 0.3623650074005127,
+      "learning_rate": 0.001,
+      "loss": 2.5391,
+      "step": 9063
+    },
+    {
+      "epoch": 0.3834503765123953,
+      "grad_norm": 0.22876878082752228,
+      "learning_rate": 0.001,
+      "loss": 2.3655,
+      "step": 9064
+    },
+    {
+      "epoch": 0.3834926812759117,
+      "grad_norm": 0.1963501125574112,
+      "learning_rate": 0.001,
+      "loss": 2.2311,
+      "step": 9065
+    },
+    {
+      "epoch": 0.38353498603942804,
+      "grad_norm": 0.8837391138076782,
+      "learning_rate": 0.001,
+      "loss": 2.8312,
+      "step": 9066
+    },
+    {
+      "epoch": 0.3835772908029444,
+      "grad_norm": 0.24529825150966644,
+      "learning_rate": 0.001,
+      "loss": 2.4636,
+      "step": 9067
+    },
+    {
+      "epoch": 0.3836195955664608,
+      "grad_norm": 0.21412113308906555,
+      "learning_rate": 0.001,
+      "loss": 2.0917,
+      "step": 9068
+    },
+    {
+      "epoch": 0.38366190032997716,
+      "grad_norm": 0.23029856383800507,
+      "learning_rate": 0.001,
+      "loss": 1.8907,
+      "step": 9069
+    },
+    {
+      "epoch": 0.3837042050934935,
+      "grad_norm": 1.920670509338379,
+      "learning_rate": 0.001,
+      "loss": 2.3119,
+      "step": 9070
+    },
+    {
+      "epoch": 0.3837465098570099,
+      "grad_norm": 0.18355774879455566,
+      "learning_rate": 0.001,
+      "loss": 2.456,
+      "step": 9071
+    },
+    {
+      "epoch": 0.3837888146205263,
+      "grad_norm": 3.794325351715088,
+      "learning_rate": 0.001,
+      "loss": 2.9978,
+      "step": 9072
+    },
+    {
+      "epoch": 0.38383111938404263,
+      "grad_norm": 0.41664114594459534,
+      "learning_rate": 0.001,
+      "loss": 3.2798,
+      "step": 9073
+    },
+    {
+      "epoch": 0.38387342414755904,
+      "grad_norm": 5.11715030670166,
+      "learning_rate": 0.001,
+      "loss": 1.6814,
+      "step": 9074
+    },
+    {
+      "epoch": 0.3839157289110754,
+      "grad_norm": 0.2116970419883728,
+      "learning_rate": 0.001,
+      "loss": 2.3122,
+      "step": 9075
+    },
+    {
+      "epoch": 0.38395803367459175,
+      "grad_norm": 0.6905900239944458,
+      "learning_rate": 0.001,
+      "loss": 2.5548,
+      "step": 9076
+    },
+    {
+      "epoch": 0.38400033843810816,
+      "grad_norm": 0.18607500195503235,
+      "learning_rate": 0.001,
+      "loss": 2.4136,
+      "step": 9077
+    },
+    {
+      "epoch": 0.3840426432016245,
+      "grad_norm": 0.18247579038143158,
+      "learning_rate": 0.001,
+      "loss": 1.8635,
+      "step": 9078
+    },
+    {
+      "epoch": 0.38408494796514087,
+      "grad_norm": 0.17803490161895752,
+      "learning_rate": 0.001,
+      "loss": 1.9766,
+      "step": 9079
+    },
+    {
+      "epoch": 0.3841272527286572,
+      "grad_norm": 0.898413896560669,
+      "learning_rate": 0.001,
+      "loss": 2.7229,
+      "step": 9080
+    },
+    {
+      "epoch": 0.38416955749217363,
+      "grad_norm": 0.25650283694267273,
+      "learning_rate": 0.001,
+      "loss": 2.6571,
+      "step": 9081
+    },
+    {
+      "epoch": 0.38421186225569,
+      "grad_norm": 0.18835589289665222,
+      "learning_rate": 0.001,
+      "loss": 2.0706,
+      "step": 9082
+    },
+    {
+      "epoch": 0.38425416701920634,
+      "grad_norm": 0.2816076874732971,
+      "learning_rate": 0.001,
+      "loss": 3.3339,
+      "step": 9083
+    },
+    {
+      "epoch": 0.38429647178272275,
+      "grad_norm": 0.22398167848587036,
+      "learning_rate": 0.001,
+      "loss": 2.4135,
+      "step": 9084
+    },
+    {
+      "epoch": 0.3843387765462391,
+      "grad_norm": 0.906223714351654,
+      "learning_rate": 0.001,
+      "loss": 2.5593,
+      "step": 9085
+    },
+    {
+      "epoch": 0.38438108130975546,
+      "grad_norm": 0.3350967764854431,
+      "learning_rate": 0.001,
+      "loss": 2.6067,
+      "step": 9086
+    },
+    {
+      "epoch": 0.38442338607327187,
+      "grad_norm": 0.21072322130203247,
+      "learning_rate": 0.001,
+      "loss": 2.4582,
+      "step": 9087
+    },
+    {
+      "epoch": 0.3844656908367882,
+      "grad_norm": 0.26963114738464355,
+      "learning_rate": 0.001,
+      "loss": 2.1451,
+      "step": 9088
+    },
+    {
+      "epoch": 0.3845079956003046,
+      "grad_norm": 0.1972511261701584,
+      "learning_rate": 0.001,
+      "loss": 2.2775,
+      "step": 9089
+    },
+    {
+      "epoch": 0.384550300363821,
+      "grad_norm": 0.19861550629138947,
+      "learning_rate": 0.001,
+      "loss": 2.1255,
+      "step": 9090
+    },
+    {
+      "epoch": 0.38459260512733734,
+      "grad_norm": 0.22689582407474518,
+      "learning_rate": 0.001,
+      "loss": 1.9905,
+      "step": 9091
+    },
+    {
+      "epoch": 0.3846349098908537,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.001,
+      "loss": 1.9154,
+      "step": 9092
+    },
+    {
+      "epoch": 0.3846772146543701,
+      "grad_norm": 0.20935285091400146,
+      "learning_rate": 0.001,
+      "loss": 2.0418,
+      "step": 9093
+    },
+    {
+      "epoch": 0.38471951941788646,
+      "grad_norm": 0.20221330225467682,
+      "learning_rate": 0.001,
+      "loss": 1.9986,
+      "step": 9094
+    },
+    {
+      "epoch": 0.3847618241814028,
+      "grad_norm": 0.22481901943683624,
+      "learning_rate": 0.001,
+      "loss": 2.7436,
+      "step": 9095
+    },
+    {
+      "epoch": 0.3848041289449192,
+      "grad_norm": 0.20071500539779663,
+      "learning_rate": 0.001,
+      "loss": 2.0842,
+      "step": 9096
+    },
+    {
+      "epoch": 0.3848464337084356,
+      "grad_norm": 1.3057656288146973,
+      "learning_rate": 0.001,
+      "loss": 2.072,
+      "step": 9097
+    },
+    {
+      "epoch": 0.38488873847195193,
+      "grad_norm": 0.1865609586238861,
+      "learning_rate": 0.001,
+      "loss": 1.8298,
+      "step": 9098
+    },
+    {
+      "epoch": 0.38493104323546834,
+      "grad_norm": 0.3449401557445526,
+      "learning_rate": 0.001,
+      "loss": 3.089,
+      "step": 9099
+    },
+    {
+      "epoch": 0.3849733479989847,
+      "grad_norm": 2.8522613048553467,
+      "learning_rate": 0.001,
+      "loss": 2.6214,
+      "step": 9100
+    },
+    {
+      "epoch": 0.38501565276250105,
+      "grad_norm": 4.267495632171631,
+      "learning_rate": 0.001,
+      "loss": 2.3361,
+      "step": 9101
+    },
+    {
+      "epoch": 0.3850579575260174,
+      "grad_norm": 0.6872550249099731,
+      "learning_rate": 0.001,
+      "loss": 2.6345,
+      "step": 9102
+    },
+    {
+      "epoch": 0.3851002622895338,
+      "grad_norm": 0.21168558299541473,
+      "learning_rate": 0.001,
+      "loss": 2.5256,
+      "step": 9103
+    },
+    {
+      "epoch": 0.38514256705305017,
+      "grad_norm": 0.2972700297832489,
+      "learning_rate": 0.001,
+      "loss": 3.0594,
+      "step": 9104
+    },
+    {
+      "epoch": 0.3851848718165665,
+      "grad_norm": 0.42669767141342163,
+      "learning_rate": 0.001,
+      "loss": 2.4855,
+      "step": 9105
+    },
+    {
+      "epoch": 0.38522717658008293,
+      "grad_norm": 4.5662126541137695,
+      "learning_rate": 0.001,
+      "loss": 3.0319,
+      "step": 9106
+    },
+    {
+      "epoch": 0.3852694813435993,
+      "grad_norm": 3.4174654483795166,
+      "learning_rate": 0.001,
+      "loss": 1.9545,
+      "step": 9107
+    },
+    {
+      "epoch": 0.38531178610711564,
+      "grad_norm": 0.5192535519599915,
+      "learning_rate": 0.001,
+      "loss": 2.3336,
+      "step": 9108
+    },
+    {
+      "epoch": 0.38535409087063205,
+      "grad_norm": 1.6883811950683594,
+      "learning_rate": 0.001,
+      "loss": 2.6045,
+      "step": 9109
+    },
+    {
+      "epoch": 0.3853963956341484,
+      "grad_norm": 0.24284732341766357,
+      "learning_rate": 0.001,
+      "loss": 2.2217,
+      "step": 9110
+    },
+    {
+      "epoch": 0.38543870039766476,
+      "grad_norm": 0.2053961157798767,
+      "learning_rate": 0.001,
+      "loss": 2.6769,
+      "step": 9111
+    },
+    {
+      "epoch": 0.38548100516118117,
+      "grad_norm": 0.32398176193237305,
+      "learning_rate": 0.001,
+      "loss": 2.1218,
+      "step": 9112
+    },
+    {
+      "epoch": 0.3855233099246975,
+      "grad_norm": 0.1831725686788559,
+      "learning_rate": 0.001,
+      "loss": 1.6788,
+      "step": 9113
+    },
+    {
+      "epoch": 0.3855656146882139,
+      "grad_norm": 0.4239247739315033,
+      "learning_rate": 0.001,
+      "loss": 3.2968,
+      "step": 9114
+    },
+    {
+      "epoch": 0.3856079194517303,
+      "grad_norm": 0.20434336364269257,
+      "learning_rate": 0.001,
+      "loss": 1.5334,
+      "step": 9115
+    },
+    {
+      "epoch": 0.38565022421524664,
+      "grad_norm": 0.18700295686721802,
+      "learning_rate": 0.001,
+      "loss": 2.1835,
+      "step": 9116
+    },
+    {
+      "epoch": 0.385692528978763,
+      "grad_norm": 0.33566123247146606,
+      "learning_rate": 0.001,
+      "loss": 2.4534,
+      "step": 9117
+    },
+    {
+      "epoch": 0.3857348337422794,
+      "grad_norm": 0.4300113022327423,
+      "learning_rate": 0.001,
+      "loss": 2.7559,
+      "step": 9118
+    },
+    {
+      "epoch": 0.38577713850579576,
+      "grad_norm": 0.21322444081306458,
+      "learning_rate": 0.001,
+      "loss": 1.8912,
+      "step": 9119
+    },
+    {
+      "epoch": 0.3858194432693121,
+      "grad_norm": 0.25216782093048096,
+      "learning_rate": 0.001,
+      "loss": 2.3039,
+      "step": 9120
+    },
+    {
+      "epoch": 0.3858617480328285,
+      "grad_norm": 0.3035780191421509,
+      "learning_rate": 0.001,
+      "loss": 2.1976,
+      "step": 9121
+    },
+    {
+      "epoch": 0.3859040527963449,
+      "grad_norm": 0.23805847764015198,
+      "learning_rate": 0.001,
+      "loss": 2.2747,
+      "step": 9122
+    },
+    {
+      "epoch": 0.38594635755986123,
+      "grad_norm": 4.429523944854736,
+      "learning_rate": 0.001,
+      "loss": 3.2502,
+      "step": 9123
+    },
+    {
+      "epoch": 0.3859886623233776,
+      "grad_norm": 0.2643257677555084,
+      "learning_rate": 0.001,
+      "loss": 3.0433,
+      "step": 9124
+    },
+    {
+      "epoch": 0.386030967086894,
+      "grad_norm": 0.1765228807926178,
+      "learning_rate": 0.001,
+      "loss": 1.8067,
+      "step": 9125
+    },
+    {
+      "epoch": 0.38607327185041035,
+      "grad_norm": 2.101315975189209,
+      "learning_rate": 0.001,
+      "loss": 2.885,
+      "step": 9126
+    },
+    {
+      "epoch": 0.3861155766139267,
+      "grad_norm": 0.20153525471687317,
+      "learning_rate": 0.001,
+      "loss": 1.8812,
+      "step": 9127
+    },
+    {
+      "epoch": 0.3861578813774431,
+      "grad_norm": 0.2016594409942627,
+      "learning_rate": 0.001,
+      "loss": 1.6079,
+      "step": 9128
+    },
+    {
+      "epoch": 0.38620018614095947,
+      "grad_norm": 0.23051926493644714,
+      "learning_rate": 0.001,
+      "loss": 2.3102,
+      "step": 9129
+    },
+    {
+      "epoch": 0.3862424909044758,
+      "grad_norm": 0.20287853479385376,
+      "learning_rate": 0.001,
+      "loss": 2.5548,
+      "step": 9130
+    },
+    {
+      "epoch": 0.38628479566799223,
+      "grad_norm": 0.2215650975704193,
+      "learning_rate": 0.001,
+      "loss": 3.1022,
+      "step": 9131
+    },
+    {
+      "epoch": 0.3863271004315086,
+      "grad_norm": 14.681644439697266,
+      "learning_rate": 0.001,
+      "loss": 2.5631,
+      "step": 9132
+    },
+    {
+      "epoch": 0.38636940519502494,
+      "grad_norm": 0.21271252632141113,
+      "learning_rate": 0.001,
+      "loss": 1.8983,
+      "step": 9133
+    },
+    {
+      "epoch": 0.38641170995854135,
+      "grad_norm": 0.5988287329673767,
+      "learning_rate": 0.001,
+      "loss": 2.0374,
+      "step": 9134
+    },
+    {
+      "epoch": 0.3864540147220577,
+      "grad_norm": 0.5086121559143066,
+      "learning_rate": 0.001,
+      "loss": 2.1648,
+      "step": 9135
+    },
+    {
+      "epoch": 0.38649631948557406,
+      "grad_norm": 0.28043192625045776,
+      "learning_rate": 0.001,
+      "loss": 2.6912,
+      "step": 9136
+    },
+    {
+      "epoch": 0.38653862424909047,
+      "grad_norm": 1.1043363809585571,
+      "learning_rate": 0.001,
+      "loss": 2.9766,
+      "step": 9137
+    },
+    {
+      "epoch": 0.3865809290126068,
+      "grad_norm": 0.42361801862716675,
+      "learning_rate": 0.001,
+      "loss": 2.2055,
+      "step": 9138
+    },
+    {
+      "epoch": 0.3866232337761232,
+      "grad_norm": 0.7368301749229431,
+      "learning_rate": 0.001,
+      "loss": 3.197,
+      "step": 9139
+    },
+    {
+      "epoch": 0.3866655385396396,
+      "grad_norm": 0.4298476278781891,
+      "learning_rate": 0.001,
+      "loss": 2.3723,
+      "step": 9140
+    },
+    {
+      "epoch": 0.38670784330315594,
+      "grad_norm": 0.20373845100402832,
+      "learning_rate": 0.001,
+      "loss": 2.2679,
+      "step": 9141
+    },
+    {
+      "epoch": 0.3867501480666723,
+      "grad_norm": 0.18816867470741272,
+      "learning_rate": 0.001,
+      "loss": 2.195,
+      "step": 9142
+    },
+    {
+      "epoch": 0.3867924528301887,
+      "grad_norm": 0.24059052765369415,
+      "learning_rate": 0.001,
+      "loss": 2.0657,
+      "step": 9143
+    },
+    {
+      "epoch": 0.38683475759370506,
+      "grad_norm": 0.19671255350112915,
+      "learning_rate": 0.001,
+      "loss": 3.1763,
+      "step": 9144
+    },
+    {
+      "epoch": 0.3868770623572214,
+      "grad_norm": 0.5776639580726624,
+      "learning_rate": 0.001,
+      "loss": 1.9861,
+      "step": 9145
+    },
+    {
+      "epoch": 0.3869193671207378,
+      "grad_norm": 0.8698708415031433,
+      "learning_rate": 0.001,
+      "loss": 2.8534,
+      "step": 9146
+    },
+    {
+      "epoch": 0.3869616718842542,
+      "grad_norm": 4.098544597625732,
+      "learning_rate": 0.001,
+      "loss": 1.9432,
+      "step": 9147
+    },
+    {
+      "epoch": 0.38700397664777053,
+      "grad_norm": 0.20377150177955627,
+      "learning_rate": 0.001,
+      "loss": 2.981,
+      "step": 9148
+    },
+    {
+      "epoch": 0.3870462814112869,
+      "grad_norm": 0.17152920365333557,
+      "learning_rate": 0.001,
+      "loss": 1.976,
+      "step": 9149
+    },
+    {
+      "epoch": 0.3870885861748033,
+      "grad_norm": 0.19490063190460205,
+      "learning_rate": 0.001,
+      "loss": 2.7471,
+      "step": 9150
+    },
+    {
+      "epoch": 0.38713089093831965,
+      "grad_norm": 0.21407170593738556,
+      "learning_rate": 0.001,
+      "loss": 2.155,
+      "step": 9151
+    },
+    {
+      "epoch": 0.387173195701836,
+      "grad_norm": 2.212458610534668,
+      "learning_rate": 0.001,
+      "loss": 1.773,
+      "step": 9152
+    },
+    {
+      "epoch": 0.3872155004653524,
+      "grad_norm": 0.15391570329666138,
+      "learning_rate": 0.001,
+      "loss": 2.4875,
+      "step": 9153
+    },
+    {
+      "epoch": 0.38725780522886877,
+      "grad_norm": 0.15746274590492249,
+      "learning_rate": 0.001,
+      "loss": 2.0196,
+      "step": 9154
+    },
+    {
+      "epoch": 0.3873001099923851,
+      "grad_norm": 0.2026894986629486,
+      "learning_rate": 0.001,
+      "loss": 2.7989,
+      "step": 9155
+    },
+    {
+      "epoch": 0.38734241475590153,
+      "grad_norm": 0.1813724935054779,
+      "learning_rate": 0.001,
+      "loss": 2.718,
+      "step": 9156
+    },
+    {
+      "epoch": 0.3873847195194179,
+      "grad_norm": 0.17108535766601562,
+      "learning_rate": 0.001,
+      "loss": 1.7214,
+      "step": 9157
+    },
+    {
+      "epoch": 0.38742702428293424,
+      "grad_norm": 0.1782073825597763,
+      "learning_rate": 0.001,
+      "loss": 2.4506,
+      "step": 9158
+    },
+    {
+      "epoch": 0.38746932904645065,
+      "grad_norm": 0.46579709649086,
+      "learning_rate": 0.001,
+      "loss": 1.7336,
+      "step": 9159
+    },
+    {
+      "epoch": 0.387511633809967,
+      "grad_norm": 0.19454286992549896,
+      "learning_rate": 0.001,
+      "loss": 1.7805,
+      "step": 9160
+    },
+    {
+      "epoch": 0.38755393857348336,
+      "grad_norm": 0.18374571204185486,
+      "learning_rate": 0.001,
+      "loss": 1.6501,
+      "step": 9161
+    },
+    {
+      "epoch": 0.38759624333699977,
+      "grad_norm": 0.23796077072620392,
+      "learning_rate": 0.001,
+      "loss": 2.6949,
+      "step": 9162
+    },
+    {
+      "epoch": 0.3876385481005161,
+      "grad_norm": 0.3394874036312103,
+      "learning_rate": 0.001,
+      "loss": 3.0172,
+      "step": 9163
+    },
+    {
+      "epoch": 0.3876808528640325,
+      "grad_norm": 0.20546482503414154,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 9164
+    },
+    {
+      "epoch": 0.3877231576275489,
+      "grad_norm": 0.23391731083393097,
+      "learning_rate": 0.001,
+      "loss": 2.7176,
+      "step": 9165
+    },
+    {
+      "epoch": 0.38776546239106524,
+      "grad_norm": 0.2764662206172943,
+      "learning_rate": 0.001,
+      "loss": 2.2979,
+      "step": 9166
+    },
+    {
+      "epoch": 0.3878077671545816,
+      "grad_norm": 0.17653588950634003,
+      "learning_rate": 0.001,
+      "loss": 2.3402,
+      "step": 9167
+    },
+    {
+      "epoch": 0.387850071918098,
+      "grad_norm": 0.4383090138435364,
+      "learning_rate": 0.001,
+      "loss": 2.0904,
+      "step": 9168
+    },
+    {
+      "epoch": 0.38789237668161436,
+      "grad_norm": 0.20846538245677948,
+      "learning_rate": 0.001,
+      "loss": 3.2569,
+      "step": 9169
+    },
+    {
+      "epoch": 0.3879346814451307,
+      "grad_norm": 0.6603441834449768,
+      "learning_rate": 0.001,
+      "loss": 2.309,
+      "step": 9170
+    },
+    {
+      "epoch": 0.38797698620864707,
+      "grad_norm": 0.18814191222190857,
+      "learning_rate": 0.001,
+      "loss": 1.4408,
+      "step": 9171
+    },
+    {
+      "epoch": 0.3880192909721635,
+      "grad_norm": 3.1111462116241455,
+      "learning_rate": 0.001,
+      "loss": 2.6072,
+      "step": 9172
+    },
+    {
+      "epoch": 0.38806159573567983,
+      "grad_norm": 0.2014855295419693,
+      "learning_rate": 0.001,
+      "loss": 1.6812,
+      "step": 9173
+    },
+    {
+      "epoch": 0.3881039004991962,
+      "grad_norm": 0.24328027665615082,
+      "learning_rate": 0.001,
+      "loss": 2.088,
+      "step": 9174
+    },
+    {
+      "epoch": 0.3881462052627126,
+      "grad_norm": 0.2333800196647644,
+      "learning_rate": 0.001,
+      "loss": 2.3285,
+      "step": 9175
+    },
+    {
+      "epoch": 0.38818851002622895,
+      "grad_norm": 0.21624669432640076,
+      "learning_rate": 0.001,
+      "loss": 2.7369,
+      "step": 9176
+    },
+    {
+      "epoch": 0.3882308147897453,
+      "grad_norm": 0.21028754115104675,
+      "learning_rate": 0.001,
+      "loss": 2.5539,
+      "step": 9177
+    },
+    {
+      "epoch": 0.3882731195532617,
+      "grad_norm": 0.26059690117836,
+      "learning_rate": 0.001,
+      "loss": 2.5894,
+      "step": 9178
+    },
+    {
+      "epoch": 0.38831542431677807,
+      "grad_norm": 0.20622827112674713,
+      "learning_rate": 0.001,
+      "loss": 2.2577,
+      "step": 9179
+    },
+    {
+      "epoch": 0.3883577290802944,
+      "grad_norm": 0.2566535472869873,
+      "learning_rate": 0.001,
+      "loss": 2.6067,
+      "step": 9180
+    },
+    {
+      "epoch": 0.38840003384381083,
+      "grad_norm": 0.1649394929409027,
+      "learning_rate": 0.001,
+      "loss": 1.9004,
+      "step": 9181
+    },
+    {
+      "epoch": 0.3884423386073272,
+      "grad_norm": 0.211260125041008,
+      "learning_rate": 0.001,
+      "loss": 2.304,
+      "step": 9182
+    },
+    {
+      "epoch": 0.38848464337084354,
+      "grad_norm": 5.116180896759033,
+      "learning_rate": 0.001,
+      "loss": 2.774,
+      "step": 9183
+    },
+    {
+      "epoch": 0.38852694813435995,
+      "grad_norm": 0.3661218583583832,
+      "learning_rate": 0.001,
+      "loss": 1.7813,
+      "step": 9184
+    },
+    {
+      "epoch": 0.3885692528978763,
+      "grad_norm": 0.2053448110818863,
+      "learning_rate": 0.001,
+      "loss": 2.1452,
+      "step": 9185
+    },
+    {
+      "epoch": 0.38861155766139266,
+      "grad_norm": 0.16580262780189514,
+      "learning_rate": 0.001,
+      "loss": 1.8883,
+      "step": 9186
+    },
+    {
+      "epoch": 0.38865386242490907,
+      "grad_norm": 0.1831020563840866,
+      "learning_rate": 0.001,
+      "loss": 2.1391,
+      "step": 9187
+    },
+    {
+      "epoch": 0.3886961671884254,
+      "grad_norm": 1.015125036239624,
+      "learning_rate": 0.001,
+      "loss": 1.8072,
+      "step": 9188
+    },
+    {
+      "epoch": 0.3887384719519418,
+      "grad_norm": 0.22336459159851074,
+      "learning_rate": 0.001,
+      "loss": 2.1674,
+      "step": 9189
+    },
+    {
+      "epoch": 0.3887807767154582,
+      "grad_norm": 0.20443737506866455,
+      "learning_rate": 0.001,
+      "loss": 2.1295,
+      "step": 9190
+    },
+    {
+      "epoch": 0.38882308147897454,
+      "grad_norm": 0.2641700208187103,
+      "learning_rate": 0.001,
+      "loss": 1.5808,
+      "step": 9191
+    },
+    {
+      "epoch": 0.3888653862424909,
+      "grad_norm": 0.5966519117355347,
+      "learning_rate": 0.001,
+      "loss": 1.7853,
+      "step": 9192
+    },
+    {
+      "epoch": 0.38890769100600725,
+      "grad_norm": 0.6837995648384094,
+      "learning_rate": 0.001,
+      "loss": 2.0421,
+      "step": 9193
+    },
+    {
+      "epoch": 0.38894999576952366,
+      "grad_norm": 0.20334653556346893,
+      "learning_rate": 0.001,
+      "loss": 1.995,
+      "step": 9194
+    },
+    {
+      "epoch": 0.38899230053304,
+      "grad_norm": 1.7629520893096924,
+      "learning_rate": 0.001,
+      "loss": 1.6059,
+      "step": 9195
+    },
+    {
+      "epoch": 0.38903460529655637,
+      "grad_norm": 0.4265093505382538,
+      "learning_rate": 0.001,
+      "loss": 2.4376,
+      "step": 9196
+    },
+    {
+      "epoch": 0.3890769100600728,
+      "grad_norm": 0.47653642296791077,
+      "learning_rate": 0.001,
+      "loss": 2.4451,
+      "step": 9197
+    },
+    {
+      "epoch": 0.38911921482358913,
+      "grad_norm": 0.43369871377944946,
+      "learning_rate": 0.001,
+      "loss": 1.8053,
+      "step": 9198
+    },
+    {
+      "epoch": 0.3891615195871055,
+      "grad_norm": 0.19080866873264313,
+      "learning_rate": 0.001,
+      "loss": 1.9405,
+      "step": 9199
+    },
+    {
+      "epoch": 0.3892038243506219,
+      "grad_norm": 0.22995217144489288,
+      "learning_rate": 0.001,
+      "loss": 2.9559,
+      "step": 9200
+    },
+    {
+      "epoch": 0.38924612911413825,
+      "grad_norm": 0.21692658960819244,
+      "learning_rate": 0.001,
+      "loss": 3.0938,
+      "step": 9201
+    },
+    {
+      "epoch": 0.3892884338776546,
+      "grad_norm": 0.6913968920707703,
+      "learning_rate": 0.001,
+      "loss": 2.2379,
+      "step": 9202
+    },
+    {
+      "epoch": 0.389330738641171,
+      "grad_norm": 0.34403377771377563,
+      "learning_rate": 0.001,
+      "loss": 2.1844,
+      "step": 9203
+    },
+    {
+      "epoch": 0.38937304340468737,
+      "grad_norm": 1.5377286672592163,
+      "learning_rate": 0.001,
+      "loss": 3.3703,
+      "step": 9204
+    },
+    {
+      "epoch": 0.3894153481682037,
+      "grad_norm": 0.212476909160614,
+      "learning_rate": 0.001,
+      "loss": 2.0148,
+      "step": 9205
+    },
+    {
+      "epoch": 0.38945765293172013,
+      "grad_norm": 0.8234764933586121,
+      "learning_rate": 0.001,
+      "loss": 2.8544,
+      "step": 9206
+    },
+    {
+      "epoch": 0.3894999576952365,
+      "grad_norm": 0.1971426010131836,
+      "learning_rate": 0.001,
+      "loss": 2.0184,
+      "step": 9207
+    },
+    {
+      "epoch": 0.38954226245875284,
+      "grad_norm": 0.23782867193222046,
+      "learning_rate": 0.001,
+      "loss": 2.1397,
+      "step": 9208
+    },
+    {
+      "epoch": 0.38958456722226925,
+      "grad_norm": 0.7499661445617676,
+      "learning_rate": 0.001,
+      "loss": 1.8995,
+      "step": 9209
+    },
+    {
+      "epoch": 0.3896268719857856,
+      "grad_norm": 15.050933837890625,
+      "learning_rate": 0.001,
+      "loss": 1.6647,
+      "step": 9210
+    },
+    {
+      "epoch": 0.38966917674930196,
+      "grad_norm": 0.3999849259853363,
+      "learning_rate": 0.001,
+      "loss": 2.5121,
+      "step": 9211
+    },
+    {
+      "epoch": 0.38971148151281837,
+      "grad_norm": 0.35282769799232483,
+      "learning_rate": 0.001,
+      "loss": 1.9899,
+      "step": 9212
+    },
+    {
+      "epoch": 0.3897537862763347,
+      "grad_norm": 0.534419059753418,
+      "learning_rate": 0.001,
+      "loss": 2.1861,
+      "step": 9213
+    },
+    {
+      "epoch": 0.3897960910398511,
+      "grad_norm": 0.2646177411079407,
+      "learning_rate": 0.001,
+      "loss": 2.0317,
+      "step": 9214
+    },
+    {
+      "epoch": 0.38983839580336743,
+      "grad_norm": 0.20612642168998718,
+      "learning_rate": 0.001,
+      "loss": 1.835,
+      "step": 9215
+    },
+    {
+      "epoch": 0.38988070056688384,
+      "grad_norm": 0.25167348980903625,
+      "learning_rate": 0.001,
+      "loss": 2.4175,
+      "step": 9216
+    },
+    {
+      "epoch": 0.3899230053304002,
+      "grad_norm": 0.21403273940086365,
+      "learning_rate": 0.001,
+      "loss": 2.954,
+      "step": 9217
+    },
+    {
+      "epoch": 0.38996531009391655,
+      "grad_norm": 0.17985409498214722,
+      "learning_rate": 0.001,
+      "loss": 1.8319,
+      "step": 9218
+    },
+    {
+      "epoch": 0.39000761485743296,
+      "grad_norm": 2.4308388233184814,
+      "learning_rate": 0.001,
+      "loss": 2.0903,
+      "step": 9219
+    },
+    {
+      "epoch": 0.3900499196209493,
+      "grad_norm": 0.2107587307691574,
+      "learning_rate": 0.001,
+      "loss": 2.7956,
+      "step": 9220
+    },
+    {
+      "epoch": 0.39009222438446567,
+      "grad_norm": 1.1186227798461914,
+      "learning_rate": 0.001,
+      "loss": 2.1013,
+      "step": 9221
+    },
+    {
+      "epoch": 0.3901345291479821,
+      "grad_norm": 0.2170724719762802,
+      "learning_rate": 0.001,
+      "loss": 2.1852,
+      "step": 9222
+    },
+    {
+      "epoch": 0.39017683391149843,
+      "grad_norm": 0.1864577829837799,
+      "learning_rate": 0.001,
+      "loss": 1.4238,
+      "step": 9223
+    },
+    {
+      "epoch": 0.3902191386750148,
+      "grad_norm": 0.23575836420059204,
+      "learning_rate": 0.001,
+      "loss": 2.6753,
+      "step": 9224
+    },
+    {
+      "epoch": 0.3902614434385312,
+      "grad_norm": 0.21837228536605835,
+      "learning_rate": 0.001,
+      "loss": 2.4262,
+      "step": 9225
+    },
+    {
+      "epoch": 0.39030374820204755,
+      "grad_norm": 4.643385410308838,
+      "learning_rate": 0.001,
+      "loss": 2.1914,
+      "step": 9226
+    },
+    {
+      "epoch": 0.3903460529655639,
+      "grad_norm": 0.25850534439086914,
+      "learning_rate": 0.001,
+      "loss": 2.254,
+      "step": 9227
+    },
+    {
+      "epoch": 0.3903883577290803,
+      "grad_norm": 0.4413652718067169,
+      "learning_rate": 0.001,
+      "loss": 4.012,
+      "step": 9228
+    },
+    {
+      "epoch": 0.39043066249259667,
+      "grad_norm": 0.20437230169773102,
+      "learning_rate": 0.001,
+      "loss": 2.7935,
+      "step": 9229
+    },
+    {
+      "epoch": 0.390472967256113,
+      "grad_norm": 0.2475084662437439,
+      "learning_rate": 0.001,
+      "loss": 2.2118,
+      "step": 9230
+    },
+    {
+      "epoch": 0.39051527201962943,
+      "grad_norm": 0.7936202883720398,
+      "learning_rate": 0.001,
+      "loss": 2.0084,
+      "step": 9231
+    },
+    {
+      "epoch": 0.3905575767831458,
+      "grad_norm": 0.21864911913871765,
+      "learning_rate": 0.001,
+      "loss": 2.0875,
+      "step": 9232
+    },
+    {
+      "epoch": 0.39059988154666214,
+      "grad_norm": 0.25647038221359253,
+      "learning_rate": 0.001,
+      "loss": 2.1307,
+      "step": 9233
+    },
+    {
+      "epoch": 0.39064218631017855,
+      "grad_norm": 1.2627050876617432,
+      "learning_rate": 0.001,
+      "loss": 1.862,
+      "step": 9234
+    },
+    {
+      "epoch": 0.3906844910736949,
+      "grad_norm": 0.18225641548633575,
+      "learning_rate": 0.001,
+      "loss": 2.1382,
+      "step": 9235
+    },
+    {
+      "epoch": 0.39072679583721126,
+      "grad_norm": 0.1784055531024933,
+      "learning_rate": 0.001,
+      "loss": 2.4409,
+      "step": 9236
+    },
+    {
+      "epoch": 0.3907691006007276,
+      "grad_norm": 1.4898735284805298,
+      "learning_rate": 0.001,
+      "loss": 2.1492,
+      "step": 9237
+    },
+    {
+      "epoch": 0.390811405364244,
+      "grad_norm": 5.816760540008545,
+      "learning_rate": 0.001,
+      "loss": 2.0967,
+      "step": 9238
+    },
+    {
+      "epoch": 0.3908537101277604,
+      "grad_norm": 0.18301312625408173,
+      "learning_rate": 0.001,
+      "loss": 1.8928,
+      "step": 9239
+    },
+    {
+      "epoch": 0.39089601489127673,
+      "grad_norm": 0.21848469972610474,
+      "learning_rate": 0.001,
+      "loss": 2.2617,
+      "step": 9240
+    },
+    {
+      "epoch": 0.39093831965479314,
+      "grad_norm": 0.23523452877998352,
+      "learning_rate": 0.001,
+      "loss": 1.8523,
+      "step": 9241
+    },
+    {
+      "epoch": 0.3909806244183095,
+      "grad_norm": 0.28411930799484253,
+      "learning_rate": 0.001,
+      "loss": 2.5036,
+      "step": 9242
+    },
+    {
+      "epoch": 0.39102292918182585,
+      "grad_norm": 0.39887842535972595,
+      "learning_rate": 0.001,
+      "loss": 3.1872,
+      "step": 9243
+    },
+    {
+      "epoch": 0.39106523394534226,
+      "grad_norm": 0.26545318961143494,
+      "learning_rate": 0.001,
+      "loss": 2.1511,
+      "step": 9244
+    },
+    {
+      "epoch": 0.3911075387088586,
+      "grad_norm": 0.3136458992958069,
+      "learning_rate": 0.001,
+      "loss": 2.8136,
+      "step": 9245
+    },
+    {
+      "epoch": 0.39114984347237497,
+      "grad_norm": 0.2994268238544464,
+      "learning_rate": 0.001,
+      "loss": 2.0385,
+      "step": 9246
+    },
+    {
+      "epoch": 0.3911921482358914,
+      "grad_norm": 0.21520791947841644,
+      "learning_rate": 0.001,
+      "loss": 3.8089,
+      "step": 9247
+    },
+    {
+      "epoch": 0.39123445299940773,
+      "grad_norm": 0.18887737393379211,
+      "learning_rate": 0.001,
+      "loss": 1.6818,
+      "step": 9248
+    },
+    {
+      "epoch": 0.3912767577629241,
+      "grad_norm": 0.23042213916778564,
+      "learning_rate": 0.001,
+      "loss": 2.0132,
+      "step": 9249
+    },
+    {
+      "epoch": 0.3913190625264405,
+      "grad_norm": 8.303719520568848,
+      "learning_rate": 0.001,
+      "loss": 1.782,
+      "step": 9250
+    },
+    {
+      "epoch": 0.39136136728995685,
+      "grad_norm": 0.25407400727272034,
+      "learning_rate": 0.001,
+      "loss": 3.071,
+      "step": 9251
+    },
+    {
+      "epoch": 0.3914036720534732,
+      "grad_norm": 0.19759856164455414,
+      "learning_rate": 0.001,
+      "loss": 1.8951,
+      "step": 9252
+    },
+    {
+      "epoch": 0.3914459768169896,
+      "grad_norm": 0.3458765745162964,
+      "learning_rate": 0.001,
+      "loss": 3.0209,
+      "step": 9253
+    },
+    {
+      "epoch": 0.39148828158050597,
+      "grad_norm": 1.5416405200958252,
+      "learning_rate": 0.001,
+      "loss": 2.7139,
+      "step": 9254
+    },
+    {
+      "epoch": 0.3915305863440223,
+      "grad_norm": 0.21925224363803864,
+      "learning_rate": 0.001,
+      "loss": 2.1197,
+      "step": 9255
+    },
+    {
+      "epoch": 0.39157289110753873,
+      "grad_norm": 0.6360104084014893,
+      "learning_rate": 0.001,
+      "loss": 1.9278,
+      "step": 9256
+    },
+    {
+      "epoch": 0.3916151958710551,
+      "grad_norm": 0.17912402749061584,
+      "learning_rate": 0.001,
+      "loss": 2.8046,
+      "step": 9257
+    },
+    {
+      "epoch": 0.39165750063457144,
+      "grad_norm": 0.22755807638168335,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 9258
+    },
+    {
+      "epoch": 0.3916998053980878,
+      "grad_norm": 0.3106396794319153,
+      "learning_rate": 0.001,
+      "loss": 2.5425,
+      "step": 9259
+    },
+    {
+      "epoch": 0.3917421101616042,
+      "grad_norm": 1.9510068893432617,
+      "learning_rate": 0.001,
+      "loss": 1.7647,
+      "step": 9260
+    },
+    {
+      "epoch": 0.39178441492512056,
+      "grad_norm": 0.20490862429141998,
+      "learning_rate": 0.001,
+      "loss": 2.5582,
+      "step": 9261
+    },
+    {
+      "epoch": 0.3918267196886369,
+      "grad_norm": 0.18178889155387878,
+      "learning_rate": 0.001,
+      "loss": 2.8619,
+      "step": 9262
+    },
+    {
+      "epoch": 0.3918690244521533,
+      "grad_norm": 0.22644934058189392,
+      "learning_rate": 0.001,
+      "loss": 2.4013,
+      "step": 9263
+    },
+    {
+      "epoch": 0.3919113292156697,
+      "grad_norm": 0.2479361891746521,
+      "learning_rate": 0.001,
+      "loss": 2.4039,
+      "step": 9264
+    },
+    {
+      "epoch": 0.39195363397918603,
+      "grad_norm": 0.20470456779003143,
+      "learning_rate": 0.001,
+      "loss": 2.086,
+      "step": 9265
+    },
+    {
+      "epoch": 0.39199593874270244,
+      "grad_norm": 0.3624879717826843,
+      "learning_rate": 0.001,
+      "loss": 1.9208,
+      "step": 9266
+    },
+    {
+      "epoch": 0.3920382435062188,
+      "grad_norm": 0.3792955279350281,
+      "learning_rate": 0.001,
+      "loss": 2.912,
+      "step": 9267
+    },
+    {
+      "epoch": 0.39208054826973515,
+      "grad_norm": 0.3835373818874359,
+      "learning_rate": 0.001,
+      "loss": 3.1838,
+      "step": 9268
+    },
+    {
+      "epoch": 0.39212285303325156,
+      "grad_norm": 0.25581711530685425,
+      "learning_rate": 0.001,
+      "loss": 2.0758,
+      "step": 9269
+    },
+    {
+      "epoch": 0.3921651577967679,
+      "grad_norm": 0.20404990017414093,
+      "learning_rate": 0.001,
+      "loss": 1.8675,
+      "step": 9270
+    },
+    {
+      "epoch": 0.39220746256028427,
+      "grad_norm": 0.22453588247299194,
+      "learning_rate": 0.001,
+      "loss": 2.3676,
+      "step": 9271
+    },
+    {
+      "epoch": 0.3922497673238007,
+      "grad_norm": 0.2473563551902771,
+      "learning_rate": 0.001,
+      "loss": 2.179,
+      "step": 9272
+    },
+    {
+      "epoch": 0.39229207208731703,
+      "grad_norm": 0.24559181928634644,
+      "learning_rate": 0.001,
+      "loss": 3.4248,
+      "step": 9273
+    },
+    {
+      "epoch": 0.3923343768508334,
+      "grad_norm": 0.2277313470840454,
+      "learning_rate": 0.001,
+      "loss": 2.0933,
+      "step": 9274
+    },
+    {
+      "epoch": 0.3923766816143498,
+      "grad_norm": 0.19083429872989655,
+      "learning_rate": 0.001,
+      "loss": 2.4522,
+      "step": 9275
+    },
+    {
+      "epoch": 0.39241898637786615,
+      "grad_norm": 0.2212873101234436,
+      "learning_rate": 0.001,
+      "loss": 2.3474,
+      "step": 9276
+    },
+    {
+      "epoch": 0.3924612911413825,
+      "grad_norm": 0.1985638290643692,
+      "learning_rate": 0.001,
+      "loss": 2.283,
+      "step": 9277
+    },
+    {
+      "epoch": 0.3925035959048989,
+      "grad_norm": 1.7838554382324219,
+      "learning_rate": 0.001,
+      "loss": 2.1613,
+      "step": 9278
+    },
+    {
+      "epoch": 0.39254590066841527,
+      "grad_norm": 0.4087802469730377,
+      "learning_rate": 0.001,
+      "loss": 1.7939,
+      "step": 9279
+    },
+    {
+      "epoch": 0.3925882054319316,
+      "grad_norm": 0.17536307871341705,
+      "learning_rate": 0.001,
+      "loss": 1.9135,
+      "step": 9280
+    },
+    {
+      "epoch": 0.39263051019544803,
+      "grad_norm": 0.6189838647842407,
+      "learning_rate": 0.001,
+      "loss": 3.2527,
+      "step": 9281
+    },
+    {
+      "epoch": 0.3926728149589644,
+      "grad_norm": 0.2154485583305359,
+      "learning_rate": 0.001,
+      "loss": 1.8292,
+      "step": 9282
+    },
+    {
+      "epoch": 0.39271511972248074,
+      "grad_norm": 0.3119647204875946,
+      "learning_rate": 0.001,
+      "loss": 3.5083,
+      "step": 9283
+    },
+    {
+      "epoch": 0.3927574244859971,
+      "grad_norm": 0.28993380069732666,
+      "learning_rate": 0.001,
+      "loss": 3.3037,
+      "step": 9284
+    },
+    {
+      "epoch": 0.3927997292495135,
+      "grad_norm": 0.2555803954601288,
+      "learning_rate": 0.001,
+      "loss": 3.6532,
+      "step": 9285
+    },
+    {
+      "epoch": 0.39284203401302986,
+      "grad_norm": 0.22608381509780884,
+      "learning_rate": 0.001,
+      "loss": 2.7748,
+      "step": 9286
+    },
+    {
+      "epoch": 0.3928843387765462,
+      "grad_norm": 0.2289711982011795,
+      "learning_rate": 0.001,
+      "loss": 2.0334,
+      "step": 9287
+    },
+    {
+      "epoch": 0.3929266435400626,
+      "grad_norm": 0.7672654390335083,
+      "learning_rate": 0.001,
+      "loss": 1.7421,
+      "step": 9288
+    },
+    {
+      "epoch": 0.392968948303579,
+      "grad_norm": 0.210458442568779,
+      "learning_rate": 0.001,
+      "loss": 2.2449,
+      "step": 9289
+    },
+    {
+      "epoch": 0.39301125306709533,
+      "grad_norm": 0.4783534109592438,
+      "learning_rate": 0.001,
+      "loss": 1.7884,
+      "step": 9290
+    },
+    {
+      "epoch": 0.39305355783061174,
+      "grad_norm": 0.16665317118167877,
+      "learning_rate": 0.001,
+      "loss": 1.8335,
+      "step": 9291
+    },
+    {
+      "epoch": 0.3930958625941281,
+      "grad_norm": 0.6642040610313416,
+      "learning_rate": 0.001,
+      "loss": 2.3598,
+      "step": 9292
+    },
+    {
+      "epoch": 0.39313816735764445,
+      "grad_norm": 0.16919787228107452,
+      "learning_rate": 0.001,
+      "loss": 2.0728,
+      "step": 9293
+    },
+    {
+      "epoch": 0.39318047212116086,
+      "grad_norm": 0.17005786299705505,
+      "learning_rate": 0.001,
+      "loss": 1.8163,
+      "step": 9294
+    },
+    {
+      "epoch": 0.3932227768846772,
+      "grad_norm": 0.29996180534362793,
+      "learning_rate": 0.001,
+      "loss": 2.3236,
+      "step": 9295
+    },
+    {
+      "epoch": 0.39326508164819357,
+      "grad_norm": 0.19820648431777954,
+      "learning_rate": 0.001,
+      "loss": 2.0907,
+      "step": 9296
+    },
+    {
+      "epoch": 0.39330738641171,
+      "grad_norm": 1.3484530448913574,
+      "learning_rate": 0.001,
+      "loss": 1.6248,
+      "step": 9297
+    },
+    {
+      "epoch": 0.39334969117522633,
+      "grad_norm": 0.38515207171440125,
+      "learning_rate": 0.001,
+      "loss": 1.6151,
+      "step": 9298
+    },
+    {
+      "epoch": 0.3933919959387427,
+      "grad_norm": 0.2751673758029938,
+      "learning_rate": 0.001,
+      "loss": 1.8327,
+      "step": 9299
+    },
+    {
+      "epoch": 0.3934343007022591,
+      "grad_norm": 0.1867271363735199,
+      "learning_rate": 0.001,
+      "loss": 2.3518,
+      "step": 9300
+    },
+    {
+      "epoch": 0.39347660546577545,
+      "grad_norm": 1.881629467010498,
+      "learning_rate": 0.001,
+      "loss": 2.6431,
+      "step": 9301
+    },
+    {
+      "epoch": 0.3935189102292918,
+      "grad_norm": 0.1862645000219345,
+      "learning_rate": 0.001,
+      "loss": 2.3873,
+      "step": 9302
+    },
+    {
+      "epoch": 0.3935612149928082,
+      "grad_norm": 4.848597526550293,
+      "learning_rate": 0.001,
+      "loss": 2.0754,
+      "step": 9303
+    },
+    {
+      "epoch": 0.39360351975632457,
+      "grad_norm": 0.8090160489082336,
+      "learning_rate": 0.001,
+      "loss": 2.0008,
+      "step": 9304
+    },
+    {
+      "epoch": 0.3936458245198409,
+      "grad_norm": 60.91786575317383,
+      "learning_rate": 0.001,
+      "loss": 2.1857,
+      "step": 9305
+    },
+    {
+      "epoch": 0.3936881292833573,
+      "grad_norm": 0.5015795826911926,
+      "learning_rate": 0.001,
+      "loss": 2.5673,
+      "step": 9306
+    },
+    {
+      "epoch": 0.3937304340468737,
+      "grad_norm": 0.21446509659290314,
+      "learning_rate": 0.001,
+      "loss": 2.0362,
+      "step": 9307
+    },
+    {
+      "epoch": 0.39377273881039004,
+      "grad_norm": 1.1446090936660767,
+      "learning_rate": 0.001,
+      "loss": 1.639,
+      "step": 9308
+    },
+    {
+      "epoch": 0.3938150435739064,
+      "grad_norm": 0.2021520733833313,
+      "learning_rate": 0.001,
+      "loss": 3.2664,
+      "step": 9309
+    },
+    {
+      "epoch": 0.3938573483374228,
+      "grad_norm": 0.2120133638381958,
+      "learning_rate": 0.001,
+      "loss": 2.2566,
+      "step": 9310
+    },
+    {
+      "epoch": 0.39389965310093916,
+      "grad_norm": 0.226138174533844,
+      "learning_rate": 0.001,
+      "loss": 1.8226,
+      "step": 9311
+    },
+    {
+      "epoch": 0.3939419578644555,
+      "grad_norm": 4.0954694747924805,
+      "learning_rate": 0.001,
+      "loss": 2.0812,
+      "step": 9312
+    },
+    {
+      "epoch": 0.3939842626279719,
+      "grad_norm": 0.37310054898262024,
+      "learning_rate": 0.001,
+      "loss": 3.0503,
+      "step": 9313
+    },
+    {
+      "epoch": 0.3940265673914883,
+      "grad_norm": 0.6909289360046387,
+      "learning_rate": 0.001,
+      "loss": 2.0039,
+      "step": 9314
+    },
+    {
+      "epoch": 0.39406887215500463,
+      "grad_norm": 0.23392127454280853,
+      "learning_rate": 0.001,
+      "loss": 1.9528,
+      "step": 9315
+    },
+    {
+      "epoch": 0.39411117691852104,
+      "grad_norm": 1.0609430074691772,
+      "learning_rate": 0.001,
+      "loss": 2.6997,
+      "step": 9316
+    },
+    {
+      "epoch": 0.3941534816820374,
+      "grad_norm": 0.2690190374851227,
+      "learning_rate": 0.001,
+      "loss": 2.5265,
+      "step": 9317
+    },
+    {
+      "epoch": 0.39419578644555375,
+      "grad_norm": 0.19583603739738464,
+      "learning_rate": 0.001,
+      "loss": 2.9415,
+      "step": 9318
+    },
+    {
+      "epoch": 0.39423809120907016,
+      "grad_norm": 0.2094552218914032,
+      "learning_rate": 0.001,
+      "loss": 2.0034,
+      "step": 9319
+    },
+    {
+      "epoch": 0.3942803959725865,
+      "grad_norm": 0.33723655343055725,
+      "learning_rate": 0.001,
+      "loss": 1.8156,
+      "step": 9320
+    },
+    {
+      "epoch": 0.39432270073610287,
+      "grad_norm": 0.2463519275188446,
+      "learning_rate": 0.001,
+      "loss": 2.1634,
+      "step": 9321
+    },
+    {
+      "epoch": 0.3943650054996193,
+      "grad_norm": 2.194249153137207,
+      "learning_rate": 0.001,
+      "loss": 1.8768,
+      "step": 9322
+    },
+    {
+      "epoch": 0.39440731026313564,
+      "grad_norm": 0.2177824079990387,
+      "learning_rate": 0.001,
+      "loss": 1.7509,
+      "step": 9323
+    },
+    {
+      "epoch": 0.394449615026652,
+      "grad_norm": 0.2521611452102661,
+      "learning_rate": 0.001,
+      "loss": 2.2049,
+      "step": 9324
+    },
+    {
+      "epoch": 0.3944919197901684,
+      "grad_norm": 0.5440561175346375,
+      "learning_rate": 0.001,
+      "loss": 2.5381,
+      "step": 9325
+    },
+    {
+      "epoch": 0.39453422455368475,
+      "grad_norm": 0.22166214883327484,
+      "learning_rate": 0.001,
+      "loss": 1.8115,
+      "step": 9326
+    },
+    {
+      "epoch": 0.3945765293172011,
+      "grad_norm": 0.8539856672286987,
+      "learning_rate": 0.001,
+      "loss": 3.4899,
+      "step": 9327
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.34720534086227417,
+      "learning_rate": 0.001,
+      "loss": 1.8842,
+      "step": 9328
+    },
+    {
+      "epoch": 0.39466113884423387,
+      "grad_norm": 0.2507597804069519,
+      "learning_rate": 0.001,
+      "loss": 2.0993,
+      "step": 9329
+    },
+    {
+      "epoch": 0.3947034436077502,
+      "grad_norm": 17.17608642578125,
+      "learning_rate": 0.001,
+      "loss": 2.5217,
+      "step": 9330
+    },
+    {
+      "epoch": 0.3947457483712666,
+      "grad_norm": 0.21708784997463226,
+      "learning_rate": 0.001,
+      "loss": 3.1434,
+      "step": 9331
+    },
+    {
+      "epoch": 0.394788053134783,
+      "grad_norm": 8.15025520324707,
+      "learning_rate": 0.001,
+      "loss": 1.6658,
+      "step": 9332
+    },
+    {
+      "epoch": 0.39483035789829934,
+      "grad_norm": 0.1978793740272522,
+      "learning_rate": 0.001,
+      "loss": 1.6641,
+      "step": 9333
+    },
+    {
+      "epoch": 0.3948726626618157,
+      "grad_norm": 0.22561314702033997,
+      "learning_rate": 0.001,
+      "loss": 1.9712,
+      "step": 9334
+    },
+    {
+      "epoch": 0.3949149674253321,
+      "grad_norm": 8.708436012268066,
+      "learning_rate": 0.001,
+      "loss": 1.9502,
+      "step": 9335
+    },
+    {
+      "epoch": 0.39495727218884846,
+      "grad_norm": 0.2591575086116791,
+      "learning_rate": 0.001,
+      "loss": 2.4213,
+      "step": 9336
+    },
+    {
+      "epoch": 0.3949995769523648,
+      "grad_norm": 0.22696593403816223,
+      "learning_rate": 0.001,
+      "loss": 1.8231,
+      "step": 9337
+    },
+    {
+      "epoch": 0.3950418817158812,
+      "grad_norm": 0.4304518401622772,
+      "learning_rate": 0.001,
+      "loss": 1.6622,
+      "step": 9338
+    },
+    {
+      "epoch": 0.3950841864793976,
+      "grad_norm": 0.27557769417762756,
+      "learning_rate": 0.001,
+      "loss": 2.2719,
+      "step": 9339
+    },
+    {
+      "epoch": 0.39512649124291394,
+      "grad_norm": 0.2860799729824066,
+      "learning_rate": 0.001,
+      "loss": 1.9818,
+      "step": 9340
+    },
+    {
+      "epoch": 0.39516879600643035,
+      "grad_norm": 119.59538269042969,
+      "learning_rate": 0.001,
+      "loss": 1.7143,
+      "step": 9341
+    },
+    {
+      "epoch": 0.3952111007699467,
+      "grad_norm": 0.20414665341377258,
+      "learning_rate": 0.001,
+      "loss": 3.6217,
+      "step": 9342
+    },
+    {
+      "epoch": 0.39525340553346305,
+      "grad_norm": 0.2606525123119354,
+      "learning_rate": 0.001,
+      "loss": 2.0212,
+      "step": 9343
+    },
+    {
+      "epoch": 0.39529571029697946,
+      "grad_norm": 0.2075478881597519,
+      "learning_rate": 0.001,
+      "loss": 2.1644,
+      "step": 9344
+    },
+    {
+      "epoch": 0.3953380150604958,
+      "grad_norm": 0.24295449256896973,
+      "learning_rate": 0.001,
+      "loss": 2.2285,
+      "step": 9345
+    },
+    {
+      "epoch": 0.39538031982401217,
+      "grad_norm": 0.22193071246147156,
+      "learning_rate": 0.001,
+      "loss": 1.8047,
+      "step": 9346
+    },
+    {
+      "epoch": 0.3954226245875286,
+      "grad_norm": 0.21409432590007782,
+      "learning_rate": 0.001,
+      "loss": 2.1791,
+      "step": 9347
+    },
+    {
+      "epoch": 0.39546492935104494,
+      "grad_norm": 2.0014984607696533,
+      "learning_rate": 0.001,
+      "loss": 2.5374,
+      "step": 9348
+    },
+    {
+      "epoch": 0.3955072341145613,
+      "grad_norm": 5.059160232543945,
+      "learning_rate": 0.001,
+      "loss": 2.2516,
+      "step": 9349
+    },
+    {
+      "epoch": 0.39554953887807764,
+      "grad_norm": 0.1997808814048767,
+      "learning_rate": 0.001,
+      "loss": 2.2374,
+      "step": 9350
+    },
+    {
+      "epoch": 0.39559184364159405,
+      "grad_norm": 0.19376425445079803,
+      "learning_rate": 0.001,
+      "loss": 1.8633,
+      "step": 9351
+    },
+    {
+      "epoch": 0.3956341484051104,
+      "grad_norm": 0.8050264716148376,
+      "learning_rate": 0.001,
+      "loss": 2.5749,
+      "step": 9352
+    },
+    {
+      "epoch": 0.39567645316862676,
+      "grad_norm": 0.3786727786064148,
+      "learning_rate": 0.001,
+      "loss": 3.1153,
+      "step": 9353
+    },
+    {
+      "epoch": 0.3957187579321432,
+      "grad_norm": 0.2291092574596405,
+      "learning_rate": 0.001,
+      "loss": 1.8893,
+      "step": 9354
+    },
+    {
+      "epoch": 0.3957610626956595,
+      "grad_norm": 0.49110499024391174,
+      "learning_rate": 0.001,
+      "loss": 2.2142,
+      "step": 9355
+    },
+    {
+      "epoch": 0.3958033674591759,
+      "grad_norm": 0.18595391511917114,
+      "learning_rate": 0.001,
+      "loss": 1.9032,
+      "step": 9356
+    },
+    {
+      "epoch": 0.3958456722226923,
+      "grad_norm": 2.1453158855438232,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 9357
+    },
+    {
+      "epoch": 0.39588797698620865,
+      "grad_norm": 0.33591386675834656,
+      "learning_rate": 0.001,
+      "loss": 2.0517,
+      "step": 9358
+    },
+    {
+      "epoch": 0.395930281749725,
+      "grad_norm": 1.8636958599090576,
+      "learning_rate": 0.001,
+      "loss": 2.7847,
+      "step": 9359
+    },
+    {
+      "epoch": 0.3959725865132414,
+      "grad_norm": 0.20229916274547577,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 9360
+    },
+    {
+      "epoch": 0.39601489127675776,
+      "grad_norm": 0.2586257755756378,
+      "learning_rate": 0.001,
+      "loss": 1.8981,
+      "step": 9361
+    },
+    {
+      "epoch": 0.3960571960402741,
+      "grad_norm": 0.340988427400589,
+      "learning_rate": 0.001,
+      "loss": 2.7096,
+      "step": 9362
+    },
+    {
+      "epoch": 0.3960995008037905,
+      "grad_norm": 0.21785105764865875,
+      "learning_rate": 0.001,
+      "loss": 1.8415,
+      "step": 9363
+    },
+    {
+      "epoch": 0.3961418055673069,
+      "grad_norm": 0.7364193201065063,
+      "learning_rate": 0.001,
+      "loss": 2.9645,
+      "step": 9364
+    },
+    {
+      "epoch": 0.39618411033082324,
+      "grad_norm": 0.32688039541244507,
+      "learning_rate": 0.001,
+      "loss": 2.9784,
+      "step": 9365
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.2986684739589691,
+      "learning_rate": 0.001,
+      "loss": 3.2292,
+      "step": 9366
+    },
+    {
+      "epoch": 0.396268719857856,
+      "grad_norm": 2.457265853881836,
+      "learning_rate": 0.001,
+      "loss": 1.9641,
+      "step": 9367
+    },
+    {
+      "epoch": 0.39631102462137235,
+      "grad_norm": 0.28799575567245483,
+      "learning_rate": 0.001,
+      "loss": 2.2059,
+      "step": 9368
+    },
+    {
+      "epoch": 0.39635332938488876,
+      "grad_norm": 0.7766205668449402,
+      "learning_rate": 0.001,
+      "loss": 3.7114,
+      "step": 9369
+    },
+    {
+      "epoch": 0.3963956341484051,
+      "grad_norm": 0.7524096369743347,
+      "learning_rate": 0.001,
+      "loss": 2.6579,
+      "step": 9370
+    },
+    {
+      "epoch": 0.3964379389119215,
+      "grad_norm": 1.014512062072754,
+      "learning_rate": 0.001,
+      "loss": 2.8398,
+      "step": 9371
+    },
+    {
+      "epoch": 0.3964802436754378,
+      "grad_norm": 0.22580230236053467,
+      "learning_rate": 0.001,
+      "loss": 1.9793,
+      "step": 9372
+    },
+    {
+      "epoch": 0.39652254843895424,
+      "grad_norm": 0.24192282557487488,
+      "learning_rate": 0.001,
+      "loss": 2.111,
+      "step": 9373
+    },
+    {
+      "epoch": 0.3965648532024706,
+      "grad_norm": 1.9858895540237427,
+      "learning_rate": 0.001,
+      "loss": 2.8711,
+      "step": 9374
+    },
+    {
+      "epoch": 0.39660715796598695,
+      "grad_norm": 0.21744923293590546,
+      "learning_rate": 0.001,
+      "loss": 1.9478,
+      "step": 9375
+    },
+    {
+      "epoch": 0.39664946272950335,
+      "grad_norm": 1.180682897567749,
+      "learning_rate": 0.001,
+      "loss": 1.7087,
+      "step": 9376
+    },
+    {
+      "epoch": 0.3966917674930197,
+      "grad_norm": 0.7958380579948425,
+      "learning_rate": 0.001,
+      "loss": 3.7644,
+      "step": 9377
+    },
+    {
+      "epoch": 0.39673407225653606,
+      "grad_norm": 1.155543565750122,
+      "learning_rate": 0.001,
+      "loss": 2.2873,
+      "step": 9378
+    },
+    {
+      "epoch": 0.3967763770200525,
+      "grad_norm": 0.2220691293478012,
+      "learning_rate": 0.001,
+      "loss": 2.2881,
+      "step": 9379
+    },
+    {
+      "epoch": 0.3968186817835688,
+      "grad_norm": 2.1947951316833496,
+      "learning_rate": 0.001,
+      "loss": 2.9223,
+      "step": 9380
+    },
+    {
+      "epoch": 0.3968609865470852,
+      "grad_norm": 0.404043048620224,
+      "learning_rate": 0.001,
+      "loss": 2.0489,
+      "step": 9381
+    },
+    {
+      "epoch": 0.3969032913106016,
+      "grad_norm": 0.22369450330734253,
+      "learning_rate": 0.001,
+      "loss": 2.169,
+      "step": 9382
+    },
+    {
+      "epoch": 0.39694559607411795,
+      "grad_norm": 0.2130047082901001,
+      "learning_rate": 0.001,
+      "loss": 2.0125,
+      "step": 9383
+    },
+    {
+      "epoch": 0.3969879008376343,
+      "grad_norm": 0.42711809277534485,
+      "learning_rate": 0.001,
+      "loss": 2.6084,
+      "step": 9384
+    },
+    {
+      "epoch": 0.3970302056011507,
+      "grad_norm": 0.31725093722343445,
+      "learning_rate": 0.001,
+      "loss": 2.2298,
+      "step": 9385
+    },
+    {
+      "epoch": 0.39707251036466706,
+      "grad_norm": 0.2912001311779022,
+      "learning_rate": 0.001,
+      "loss": 2.0143,
+      "step": 9386
+    },
+    {
+      "epoch": 0.3971148151281834,
+      "grad_norm": 0.27698057889938354,
+      "learning_rate": 0.001,
+      "loss": 2.3999,
+      "step": 9387
+    },
+    {
+      "epoch": 0.39715711989169983,
+      "grad_norm": 0.7280207276344299,
+      "learning_rate": 0.001,
+      "loss": 2.1915,
+      "step": 9388
+    },
+    {
+      "epoch": 0.3971994246552162,
+      "grad_norm": 1.9393442869186401,
+      "learning_rate": 0.001,
+      "loss": 2.6316,
+      "step": 9389
+    },
+    {
+      "epoch": 0.39724172941873254,
+      "grad_norm": 0.20909947156906128,
+      "learning_rate": 0.001,
+      "loss": 2.5462,
+      "step": 9390
+    },
+    {
+      "epoch": 0.39728403418224895,
+      "grad_norm": 2.941432237625122,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 9391
+    },
+    {
+      "epoch": 0.3973263389457653,
+      "grad_norm": 0.18532417714595795,
+      "learning_rate": 0.001,
+      "loss": 2.9728,
+      "step": 9392
+    },
+    {
+      "epoch": 0.39736864370928165,
+      "grad_norm": 0.1898314207792282,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 9393
+    },
+    {
+      "epoch": 0.39741094847279806,
+      "grad_norm": 2.149635076522827,
+      "learning_rate": 0.001,
+      "loss": 2.0937,
+      "step": 9394
+    },
+    {
+      "epoch": 0.3974532532363144,
+      "grad_norm": 0.22059142589569092,
+      "learning_rate": 0.001,
+      "loss": 2.1177,
+      "step": 9395
+    },
+    {
+      "epoch": 0.3974955579998308,
+      "grad_norm": 0.38471415638923645,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 9396
+    },
+    {
+      "epoch": 0.3975378627633471,
+      "grad_norm": 0.6079134941101074,
+      "learning_rate": 0.001,
+      "loss": 2.0545,
+      "step": 9397
+    },
+    {
+      "epoch": 0.39758016752686354,
+      "grad_norm": 0.5451988577842712,
+      "learning_rate": 0.001,
+      "loss": 2.1774,
+      "step": 9398
+    },
+    {
+      "epoch": 0.3976224722903799,
+      "grad_norm": 0.20046019554138184,
+      "learning_rate": 0.001,
+      "loss": 1.909,
+      "step": 9399
+    },
+    {
+      "epoch": 0.39766477705389625,
+      "grad_norm": 0.7716697454452515,
+      "learning_rate": 0.001,
+      "loss": 2.4313,
+      "step": 9400
+    },
+    {
+      "epoch": 0.39770708181741266,
+      "grad_norm": 0.3297424018383026,
+      "learning_rate": 0.001,
+      "loss": 2.0079,
+      "step": 9401
+    },
+    {
+      "epoch": 0.397749386580929,
+      "grad_norm": 0.1952991932630539,
+      "learning_rate": 0.001,
+      "loss": 2.1102,
+      "step": 9402
+    },
+    {
+      "epoch": 0.39779169134444536,
+      "grad_norm": 0.19148588180541992,
+      "learning_rate": 0.001,
+      "loss": 1.9686,
+      "step": 9403
+    },
+    {
+      "epoch": 0.3978339961079618,
+      "grad_norm": 0.2454746812582016,
+      "learning_rate": 0.001,
+      "loss": 2.5814,
+      "step": 9404
+    },
+    {
+      "epoch": 0.39787630087147813,
+      "grad_norm": 1.7941993474960327,
+      "learning_rate": 0.001,
+      "loss": 2.0305,
+      "step": 9405
+    },
+    {
+      "epoch": 0.3979186056349945,
+      "grad_norm": 15.885468482971191,
+      "learning_rate": 0.001,
+      "loss": 2.1969,
+      "step": 9406
+    },
+    {
+      "epoch": 0.3979609103985109,
+      "grad_norm": 0.22626948356628418,
+      "learning_rate": 0.001,
+      "loss": 2.3471,
+      "step": 9407
+    },
+    {
+      "epoch": 0.39800321516202725,
+      "grad_norm": 0.3605894446372986,
+      "learning_rate": 0.001,
+      "loss": 2.5423,
+      "step": 9408
+    },
+    {
+      "epoch": 0.3980455199255436,
+      "grad_norm": 0.19169077277183533,
+      "learning_rate": 0.001,
+      "loss": 1.6805,
+      "step": 9409
+    },
+    {
+      "epoch": 0.39808782468906,
+      "grad_norm": 0.32516393065452576,
+      "learning_rate": 0.001,
+      "loss": 2.1722,
+      "step": 9410
+    },
+    {
+      "epoch": 0.39813012945257636,
+      "grad_norm": 0.7686219215393066,
+      "learning_rate": 0.001,
+      "loss": 2.8895,
+      "step": 9411
+    },
+    {
+      "epoch": 0.3981724342160927,
+      "grad_norm": 0.2653716802597046,
+      "learning_rate": 0.001,
+      "loss": 1.8654,
+      "step": 9412
+    },
+    {
+      "epoch": 0.39821473897960913,
+      "grad_norm": 3.2883455753326416,
+      "learning_rate": 0.001,
+      "loss": 2.0769,
+      "step": 9413
+    },
+    {
+      "epoch": 0.3982570437431255,
+      "grad_norm": 10.234563827514648,
+      "learning_rate": 0.001,
+      "loss": 2.4105,
+      "step": 9414
+    },
+    {
+      "epoch": 0.39829934850664184,
+      "grad_norm": 0.2494221329689026,
+      "learning_rate": 0.001,
+      "loss": 1.6076,
+      "step": 9415
+    },
+    {
+      "epoch": 0.39834165327015825,
+      "grad_norm": 0.28424543142318726,
+      "learning_rate": 0.001,
+      "loss": 2.3292,
+      "step": 9416
+    },
+    {
+      "epoch": 0.3983839580336746,
+      "grad_norm": 0.23830546438694,
+      "learning_rate": 0.001,
+      "loss": 2.6482,
+      "step": 9417
+    },
+    {
+      "epoch": 0.39842626279719096,
+      "grad_norm": 3.2902987003326416,
+      "learning_rate": 0.001,
+      "loss": 2.8442,
+      "step": 9418
+    },
+    {
+      "epoch": 0.3984685675607073,
+      "grad_norm": 1.0926035642623901,
+      "learning_rate": 0.001,
+      "loss": 1.8991,
+      "step": 9419
+    },
+    {
+      "epoch": 0.3985108723242237,
+      "grad_norm": 0.6159302592277527,
+      "learning_rate": 0.001,
+      "loss": 3.9763,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3985531770877401,
+      "grad_norm": 0.40966638922691345,
+      "learning_rate": 0.001,
+      "loss": 3.1978,
+      "step": 9421
+    },
+    {
+      "epoch": 0.39859548185125643,
+      "grad_norm": 1.2507386207580566,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 9422
+    },
+    {
+      "epoch": 0.39863778661477284,
+      "grad_norm": 0.210091695189476,
+      "learning_rate": 0.001,
+      "loss": 2.0291,
+      "step": 9423
+    },
+    {
+      "epoch": 0.3986800913782892,
+      "grad_norm": 0.6132796406745911,
+      "learning_rate": 0.001,
+      "loss": 1.9906,
+      "step": 9424
+    },
+    {
+      "epoch": 0.39872239614180555,
+      "grad_norm": 0.20599402487277985,
+      "learning_rate": 0.001,
+      "loss": 1.9661,
+      "step": 9425
+    },
+    {
+      "epoch": 0.39876470090532196,
+      "grad_norm": 75.12084197998047,
+      "learning_rate": 0.001,
+      "loss": 2.2098,
+      "step": 9426
+    },
+    {
+      "epoch": 0.3988070056688383,
+      "grad_norm": 0.19137638807296753,
+      "learning_rate": 0.001,
+      "loss": 2.4808,
+      "step": 9427
+    },
+    {
+      "epoch": 0.39884931043235466,
+      "grad_norm": 0.4978329539299011,
+      "learning_rate": 0.001,
+      "loss": 2.0099,
+      "step": 9428
+    },
+    {
+      "epoch": 0.3988916151958711,
+      "grad_norm": 1.2876667976379395,
+      "learning_rate": 0.001,
+      "loss": 2.6079,
+      "step": 9429
+    },
+    {
+      "epoch": 0.39893391995938743,
+      "grad_norm": 8.302372932434082,
+      "learning_rate": 0.001,
+      "loss": 2.3667,
+      "step": 9430
+    },
+    {
+      "epoch": 0.3989762247229038,
+      "grad_norm": 0.2350839525461197,
+      "learning_rate": 0.001,
+      "loss": 2.1824,
+      "step": 9431
+    },
+    {
+      "epoch": 0.3990185294864202,
+      "grad_norm": 7.219992160797119,
+      "learning_rate": 0.001,
+      "loss": 2.0594,
+      "step": 9432
+    },
+    {
+      "epoch": 0.39906083424993655,
+      "grad_norm": 0.8567195534706116,
+      "learning_rate": 0.001,
+      "loss": 2.2847,
+      "step": 9433
+    },
+    {
+      "epoch": 0.3991031390134529,
+      "grad_norm": 0.2659953236579895,
+      "learning_rate": 0.001,
+      "loss": 2.0949,
+      "step": 9434
+    },
+    {
+      "epoch": 0.3991454437769693,
+      "grad_norm": 0.27466630935668945,
+      "learning_rate": 0.001,
+      "loss": 2.9983,
+      "step": 9435
+    },
+    {
+      "epoch": 0.39918774854048567,
+      "grad_norm": 0.22787296772003174,
+      "learning_rate": 0.001,
+      "loss": 3.1284,
+      "step": 9436
+    },
+    {
+      "epoch": 0.399230053304002,
+      "grad_norm": 0.6802979111671448,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 9437
+    },
+    {
+      "epoch": 0.39927235806751843,
+      "grad_norm": 0.22264288365840912,
+      "learning_rate": 0.001,
+      "loss": 3.2098,
+      "step": 9438
+    },
+    {
+      "epoch": 0.3993146628310348,
+      "grad_norm": 0.20934641361236572,
+      "learning_rate": 0.001,
+      "loss": 2.5188,
+      "step": 9439
+    },
+    {
+      "epoch": 0.39935696759455114,
+      "grad_norm": 0.3321964144706726,
+      "learning_rate": 0.001,
+      "loss": 2.8836,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3993992723580675,
+      "grad_norm": 0.7503458857536316,
+      "learning_rate": 0.001,
+      "loss": 2.3137,
+      "step": 9441
+    },
+    {
+      "epoch": 0.3994415771215839,
+      "grad_norm": 0.20252926647663116,
+      "learning_rate": 0.001,
+      "loss": 3.1602,
+      "step": 9442
+    },
+    {
+      "epoch": 0.39948388188510026,
+      "grad_norm": 0.2206176519393921,
+      "learning_rate": 0.001,
+      "loss": 4.0381,
+      "step": 9443
+    },
+    {
+      "epoch": 0.3995261866486166,
+      "grad_norm": 0.16592377424240112,
+      "learning_rate": 0.001,
+      "loss": 2.0148,
+      "step": 9444
+    },
+    {
+      "epoch": 0.399568491412133,
+      "grad_norm": 0.19829799234867096,
+      "learning_rate": 0.001,
+      "loss": 1.8411,
+      "step": 9445
+    },
+    {
+      "epoch": 0.3996107961756494,
+      "grad_norm": 0.18881183862686157,
+      "learning_rate": 0.001,
+      "loss": 1.9147,
+      "step": 9446
+    },
+    {
+      "epoch": 0.39965310093916573,
+      "grad_norm": 0.24857249855995178,
+      "learning_rate": 0.001,
+      "loss": 2.3787,
+      "step": 9447
+    },
+    {
+      "epoch": 0.39969540570268214,
+      "grad_norm": 0.2752115726470947,
+      "learning_rate": 0.001,
+      "loss": 1.9453,
+      "step": 9448
+    },
+    {
+      "epoch": 0.3997377104661985,
+      "grad_norm": 0.1992688775062561,
+      "learning_rate": 0.001,
+      "loss": 2.4652,
+      "step": 9449
+    },
+    {
+      "epoch": 0.39978001522971485,
+      "grad_norm": 0.20093099772930145,
+      "learning_rate": 0.001,
+      "loss": 1.8111,
+      "step": 9450
+    },
+    {
+      "epoch": 0.39982231999323126,
+      "grad_norm": 0.1518988311290741,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 9451
+    },
+    {
+      "epoch": 0.3998646247567476,
+      "grad_norm": 0.20735451579093933,
+      "learning_rate": 0.001,
+      "loss": 2.1059,
+      "step": 9452
+    },
+    {
+      "epoch": 0.39990692952026397,
+      "grad_norm": 0.2454904466867447,
+      "learning_rate": 0.001,
+      "loss": 2.5582,
+      "step": 9453
+    },
+    {
+      "epoch": 0.3999492342837804,
+      "grad_norm": 0.8346079587936401,
+      "learning_rate": 0.001,
+      "loss": 3.135,
+      "step": 9454
+    },
+    {
+      "epoch": 0.39999153904729673,
+      "grad_norm": 0.1993485987186432,
+      "learning_rate": 0.001,
+      "loss": 2.1025,
+      "step": 9455
+    },
+    {
+      "epoch": 0.4000338438108131,
+      "grad_norm": 0.4179655909538269,
+      "learning_rate": 0.001,
+      "loss": 2.0312,
+      "step": 9456
+    },
+    {
+      "epoch": 0.4000761485743295,
+      "grad_norm": 0.17672595381736755,
+      "learning_rate": 0.001,
+      "loss": 2.5397,
+      "step": 9457
+    },
+    {
+      "epoch": 0.40011845333784585,
+      "grad_norm": 0.24849601089954376,
+      "learning_rate": 0.001,
+      "loss": 2.4992,
+      "step": 9458
+    },
+    {
+      "epoch": 0.4001607581013622,
+      "grad_norm": 0.2164924442768097,
+      "learning_rate": 0.001,
+      "loss": 2.0831,
+      "step": 9459
+    },
+    {
+      "epoch": 0.4002030628648786,
+      "grad_norm": 3.8473236560821533,
+      "learning_rate": 0.001,
+      "loss": 2.1948,
+      "step": 9460
+    },
+    {
+      "epoch": 0.40024536762839497,
+      "grad_norm": 0.24513396620750427,
+      "learning_rate": 0.001,
+      "loss": 1.9716,
+      "step": 9461
+    },
+    {
+      "epoch": 0.4002876723919113,
+      "grad_norm": 0.6947419047355652,
+      "learning_rate": 0.001,
+      "loss": 2.8307,
+      "step": 9462
+    },
+    {
+      "epoch": 0.4003299771554277,
+      "grad_norm": 0.20410794019699097,
+      "learning_rate": 0.001,
+      "loss": 3.0127,
+      "step": 9463
+    },
+    {
+      "epoch": 0.4003722819189441,
+      "grad_norm": 0.18092742562294006,
+      "learning_rate": 0.001,
+      "loss": 1.9074,
+      "step": 9464
+    },
+    {
+      "epoch": 0.40041458668246044,
+      "grad_norm": 0.17439773678779602,
+      "learning_rate": 0.001,
+      "loss": 1.8441,
+      "step": 9465
+    },
+    {
+      "epoch": 0.4004568914459768,
+      "grad_norm": 0.5184692144393921,
+      "learning_rate": 0.001,
+      "loss": 2.8331,
+      "step": 9466
+    },
+    {
+      "epoch": 0.4004991962094932,
+      "grad_norm": 0.19530414044857025,
+      "learning_rate": 0.001,
+      "loss": 2.3759,
+      "step": 9467
+    },
+    {
+      "epoch": 0.40054150097300956,
+      "grad_norm": 0.5446052551269531,
+      "learning_rate": 0.001,
+      "loss": 2.5317,
+      "step": 9468
+    },
+    {
+      "epoch": 0.4005838057365259,
+      "grad_norm": 1.0706979036331177,
+      "learning_rate": 0.001,
+      "loss": 3.0557,
+      "step": 9469
+    },
+    {
+      "epoch": 0.4006261105000423,
+      "grad_norm": 0.25208720564842224,
+      "learning_rate": 0.001,
+      "loss": 2.2757,
+      "step": 9470
+    },
+    {
+      "epoch": 0.4006684152635587,
+      "grad_norm": 0.1785879284143448,
+      "learning_rate": 0.001,
+      "loss": 1.8306,
+      "step": 9471
+    },
+    {
+      "epoch": 0.40071072002707503,
+      "grad_norm": 0.20661844313144684,
+      "learning_rate": 0.001,
+      "loss": 2.9331,
+      "step": 9472
+    },
+    {
+      "epoch": 0.40075302479059144,
+      "grad_norm": 0.24064034223556519,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 9473
+    },
+    {
+      "epoch": 0.4007953295541078,
+      "grad_norm": 28.610883712768555,
+      "learning_rate": 0.001,
+      "loss": 2.2863,
+      "step": 9474
+    },
+    {
+      "epoch": 0.40083763431762415,
+      "grad_norm": 0.6039111614227295,
+      "learning_rate": 0.001,
+      "loss": 2.4287,
+      "step": 9475
+    },
+    {
+      "epoch": 0.40087993908114056,
+      "grad_norm": 0.18513540923595428,
+      "learning_rate": 0.001,
+      "loss": 2.2021,
+      "step": 9476
+    },
+    {
+      "epoch": 0.4009222438446569,
+      "grad_norm": 1.6514339447021484,
+      "learning_rate": 0.001,
+      "loss": 1.9221,
+      "step": 9477
+    },
+    {
+      "epoch": 0.40096454860817327,
+      "grad_norm": 10.18622875213623,
+      "learning_rate": 0.001,
+      "loss": 2.3673,
+      "step": 9478
+    },
+    {
+      "epoch": 0.4010068533716897,
+      "grad_norm": 1.0164726972579956,
+      "learning_rate": 0.001,
+      "loss": 1.8288,
+      "step": 9479
+    },
+    {
+      "epoch": 0.40104915813520603,
+      "grad_norm": 0.5969464778900146,
+      "learning_rate": 0.001,
+      "loss": 2.2311,
+      "step": 9480
+    },
+    {
+      "epoch": 0.4010914628987224,
+      "grad_norm": 0.22844240069389343,
+      "learning_rate": 0.001,
+      "loss": 2.826,
+      "step": 9481
+    },
+    {
+      "epoch": 0.4011337676622388,
+      "grad_norm": 0.3800431191921234,
+      "learning_rate": 0.001,
+      "loss": 2.7672,
+      "step": 9482
+    },
+    {
+      "epoch": 0.40117607242575515,
+      "grad_norm": 0.4591025114059448,
+      "learning_rate": 0.001,
+      "loss": 2.5813,
+      "step": 9483
+    },
+    {
+      "epoch": 0.4012183771892715,
+      "grad_norm": 0.21790550649166107,
+      "learning_rate": 0.001,
+      "loss": 1.8184,
+      "step": 9484
+    },
+    {
+      "epoch": 0.40126068195278786,
+      "grad_norm": 0.7299678325653076,
+      "learning_rate": 0.001,
+      "loss": 1.9012,
+      "step": 9485
+    },
+    {
+      "epoch": 0.40130298671630427,
+      "grad_norm": 1.0232609510421753,
+      "learning_rate": 0.001,
+      "loss": 3.1905,
+      "step": 9486
+    },
+    {
+      "epoch": 0.4013452914798206,
+      "grad_norm": 0.1979575753211975,
+      "learning_rate": 0.001,
+      "loss": 1.8858,
+      "step": 9487
+    },
+    {
+      "epoch": 0.401387596243337,
+      "grad_norm": 1.3063571453094482,
+      "learning_rate": 0.001,
+      "loss": 1.5222,
+      "step": 9488
+    },
+    {
+      "epoch": 0.4014299010068534,
+      "grad_norm": 0.2541360855102539,
+      "learning_rate": 0.001,
+      "loss": 2.1482,
+      "step": 9489
+    },
+    {
+      "epoch": 0.40147220577036974,
+      "grad_norm": 2.719597816467285,
+      "learning_rate": 0.001,
+      "loss": 2.9177,
+      "step": 9490
+    },
+    {
+      "epoch": 0.4015145105338861,
+      "grad_norm": 0.18410396575927734,
+      "learning_rate": 0.001,
+      "loss": 1.3939,
+      "step": 9491
+    },
+    {
+      "epoch": 0.4015568152974025,
+      "grad_norm": 0.30777132511138916,
+      "learning_rate": 0.001,
+      "loss": 2.4803,
+      "step": 9492
+    },
+    {
+      "epoch": 0.40159912006091886,
+      "grad_norm": 0.5091328024864197,
+      "learning_rate": 0.001,
+      "loss": 2.1258,
+      "step": 9493
+    },
+    {
+      "epoch": 0.4016414248244352,
+      "grad_norm": 0.5316731929779053,
+      "learning_rate": 0.001,
+      "loss": 3.1333,
+      "step": 9494
+    },
+    {
+      "epoch": 0.4016837295879516,
+      "grad_norm": 0.5768591165542603,
+      "learning_rate": 0.001,
+      "loss": 2.1632,
+      "step": 9495
+    },
+    {
+      "epoch": 0.401726034351468,
+      "grad_norm": 0.2193114310503006,
+      "learning_rate": 0.001,
+      "loss": 2.4664,
+      "step": 9496
+    },
+    {
+      "epoch": 0.40176833911498433,
+      "grad_norm": 0.37915632128715515,
+      "learning_rate": 0.001,
+      "loss": 4.2133,
+      "step": 9497
+    },
+    {
+      "epoch": 0.40181064387850074,
+      "grad_norm": 0.18758665025234222,
+      "learning_rate": 0.001,
+      "loss": 1.8454,
+      "step": 9498
+    },
+    {
+      "epoch": 0.4018529486420171,
+      "grad_norm": 0.24519726634025574,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 9499
+    },
+    {
+      "epoch": 0.40189525340553345,
+      "grad_norm": 0.2464457005262375,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 9500
+    },
+    {
+      "epoch": 0.40193755816904986,
+      "grad_norm": 0.21936391294002533,
+      "learning_rate": 0.001,
+      "loss": 2.2305,
+      "step": 9501
+    },
+    {
+      "epoch": 0.4019798629325662,
+      "grad_norm": 0.248738095164299,
+      "learning_rate": 0.001,
+      "loss": 2.2684,
+      "step": 9502
+    },
+    {
+      "epoch": 0.40202216769608257,
+      "grad_norm": 0.25540632009506226,
+      "learning_rate": 0.001,
+      "loss": 3.4357,
+      "step": 9503
+    },
+    {
+      "epoch": 0.402064472459599,
+      "grad_norm": 1.870626449584961,
+      "learning_rate": 0.001,
+      "loss": 2.1011,
+      "step": 9504
+    },
+    {
+      "epoch": 0.40210677722311533,
+      "grad_norm": 0.23823359608650208,
+      "learning_rate": 0.001,
+      "loss": 1.7739,
+      "step": 9505
+    },
+    {
+      "epoch": 0.4021490819866317,
+      "grad_norm": 0.678466260433197,
+      "learning_rate": 0.001,
+      "loss": 1.9031,
+      "step": 9506
+    },
+    {
+      "epoch": 0.4021913867501481,
+      "grad_norm": 0.4503398537635803,
+      "learning_rate": 0.001,
+      "loss": 1.8608,
+      "step": 9507
+    },
+    {
+      "epoch": 0.40223369151366445,
+      "grad_norm": 0.528942346572876,
+      "learning_rate": 0.001,
+      "loss": 1.7426,
+      "step": 9508
+    },
+    {
+      "epoch": 0.4022759962771808,
+      "grad_norm": 0.2618485391139984,
+      "learning_rate": 0.001,
+      "loss": 2.144,
+      "step": 9509
+    },
+    {
+      "epoch": 0.40231830104069716,
+      "grad_norm": 4.79293155670166,
+      "learning_rate": 0.001,
+      "loss": 2.7917,
+      "step": 9510
+    },
+    {
+      "epoch": 0.40236060580421357,
+      "grad_norm": 0.18239739537239075,
+      "learning_rate": 0.001,
+      "loss": 2.5635,
+      "step": 9511
+    },
+    {
+      "epoch": 0.4024029105677299,
+      "grad_norm": 0.7269040942192078,
+      "learning_rate": 0.001,
+      "loss": 2.3232,
+      "step": 9512
+    },
+    {
+      "epoch": 0.4024452153312463,
+      "grad_norm": 0.18123169243335724,
+      "learning_rate": 0.001,
+      "loss": 1.7372,
+      "step": 9513
+    },
+    {
+      "epoch": 0.4024875200947627,
+      "grad_norm": 0.19387836754322052,
+      "learning_rate": 0.001,
+      "loss": 2.0364,
+      "step": 9514
+    },
+    {
+      "epoch": 0.40252982485827904,
+      "grad_norm": 0.18925026059150696,
+      "learning_rate": 0.001,
+      "loss": 1.7467,
+      "step": 9515
+    },
+    {
+      "epoch": 0.4025721296217954,
+      "grad_norm": 0.3890759348869324,
+      "learning_rate": 0.001,
+      "loss": 2.7014,
+      "step": 9516
+    },
+    {
+      "epoch": 0.4026144343853118,
+      "grad_norm": 0.3346877098083496,
+      "learning_rate": 0.001,
+      "loss": 3.0175,
+      "step": 9517
+    },
+    {
+      "epoch": 0.40265673914882816,
+      "grad_norm": 0.3895154893398285,
+      "learning_rate": 0.001,
+      "loss": 2.8014,
+      "step": 9518
+    },
+    {
+      "epoch": 0.4026990439123445,
+      "grad_norm": 3.064667224884033,
+      "learning_rate": 0.001,
+      "loss": 2.5548,
+      "step": 9519
+    },
+    {
+      "epoch": 0.4027413486758609,
+      "grad_norm": 0.25531595945358276,
+      "learning_rate": 0.001,
+      "loss": 1.9813,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4027836534393773,
+      "grad_norm": 0.23096615076065063,
+      "learning_rate": 0.001,
+      "loss": 2.8028,
+      "step": 9521
+    },
+    {
+      "epoch": 0.40282595820289363,
+      "grad_norm": 0.22149285674095154,
+      "learning_rate": 0.001,
+      "loss": 2.2017,
+      "step": 9522
+    },
+    {
+      "epoch": 0.40286826296641004,
+      "grad_norm": 0.32118022441864014,
+      "learning_rate": 0.001,
+      "loss": 3.6867,
+      "step": 9523
+    },
+    {
+      "epoch": 0.4029105677299264,
+      "grad_norm": 0.2108854353427887,
+      "learning_rate": 0.001,
+      "loss": 1.8694,
+      "step": 9524
+    },
+    {
+      "epoch": 0.40295287249344275,
+      "grad_norm": 0.9841258525848389,
+      "learning_rate": 0.001,
+      "loss": 3.0174,
+      "step": 9525
+    },
+    {
+      "epoch": 0.40299517725695916,
+      "grad_norm": 0.5012634992599487,
+      "learning_rate": 0.001,
+      "loss": 2.83,
+      "step": 9526
+    },
+    {
+      "epoch": 0.4030374820204755,
+      "grad_norm": 0.27034950256347656,
+      "learning_rate": 0.001,
+      "loss": 2.4939,
+      "step": 9527
+    },
+    {
+      "epoch": 0.40307978678399187,
+      "grad_norm": 0.18938782811164856,
+      "learning_rate": 0.001,
+      "loss": 2.5085,
+      "step": 9528
+    },
+    {
+      "epoch": 0.4031220915475083,
+      "grad_norm": 0.1947273164987564,
+      "learning_rate": 0.001,
+      "loss": 2.1869,
+      "step": 9529
+    },
+    {
+      "epoch": 0.40316439631102463,
+      "grad_norm": 0.26790064573287964,
+      "learning_rate": 0.001,
+      "loss": 2.4795,
+      "step": 9530
+    },
+    {
+      "epoch": 0.403206701074541,
+      "grad_norm": 0.19423073530197144,
+      "learning_rate": 0.001,
+      "loss": 2.0744,
+      "step": 9531
+    },
+    {
+      "epoch": 0.40324900583805734,
+      "grad_norm": 0.5601986646652222,
+      "learning_rate": 0.001,
+      "loss": 2.2896,
+      "step": 9532
+    },
+    {
+      "epoch": 0.40329131060157375,
+      "grad_norm": 1.604595422744751,
+      "learning_rate": 0.001,
+      "loss": 2.3243,
+      "step": 9533
+    },
+    {
+      "epoch": 0.4033336153650901,
+      "grad_norm": 0.31608372926712036,
+      "learning_rate": 0.001,
+      "loss": 2.3288,
+      "step": 9534
+    },
+    {
+      "epoch": 0.40337592012860646,
+      "grad_norm": 0.28224605321884155,
+      "learning_rate": 0.001,
+      "loss": 3.0434,
+      "step": 9535
+    },
+    {
+      "epoch": 0.40341822489212287,
+      "grad_norm": 0.21669940650463104,
+      "learning_rate": 0.001,
+      "loss": 2.5078,
+      "step": 9536
+    },
+    {
+      "epoch": 0.4034605296556392,
+      "grad_norm": 0.34247350692749023,
+      "learning_rate": 0.001,
+      "loss": 2.1365,
+      "step": 9537
+    },
+    {
+      "epoch": 0.4035028344191556,
+      "grad_norm": 0.3081550598144531,
+      "learning_rate": 0.001,
+      "loss": 1.9531,
+      "step": 9538
+    },
+    {
+      "epoch": 0.403545139182672,
+      "grad_norm": 1.4701812267303467,
+      "learning_rate": 0.001,
+      "loss": 2.1148,
+      "step": 9539
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 0.21030528843402863,
+      "learning_rate": 0.001,
+      "loss": 2.2116,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4036297487097047,
+      "grad_norm": 0.22468726336956024,
+      "learning_rate": 0.001,
+      "loss": 2.0276,
+      "step": 9541
+    },
+    {
+      "epoch": 0.4036720534732211,
+      "grad_norm": 0.2096341848373413,
+      "learning_rate": 0.001,
+      "loss": 2.1614,
+      "step": 9542
+    },
+    {
+      "epoch": 0.40371435823673746,
+      "grad_norm": 0.17588932812213898,
+      "learning_rate": 0.001,
+      "loss": 1.7986,
+      "step": 9543
+    },
+    {
+      "epoch": 0.4037566630002538,
+      "grad_norm": 0.21680901944637299,
+      "learning_rate": 0.001,
+      "loss": 2.815,
+      "step": 9544
+    },
+    {
+      "epoch": 0.4037989677637702,
+      "grad_norm": 0.16149887442588806,
+      "learning_rate": 0.001,
+      "loss": 1.4867,
+      "step": 9545
+    },
+    {
+      "epoch": 0.4038412725272866,
+      "grad_norm": 0.25393468141555786,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 9546
+    },
+    {
+      "epoch": 0.40388357729080293,
+      "grad_norm": 0.26233553886413574,
+      "learning_rate": 0.001,
+      "loss": 2.9283,
+      "step": 9547
+    },
+    {
+      "epoch": 0.40392588205431934,
+      "grad_norm": 0.21052797138690948,
+      "learning_rate": 0.001,
+      "loss": 2.1164,
+      "step": 9548
+    },
+    {
+      "epoch": 0.4039681868178357,
+      "grad_norm": 0.220393568277359,
+      "learning_rate": 0.001,
+      "loss": 2.5153,
+      "step": 9549
+    },
+    {
+      "epoch": 0.40401049158135205,
+      "grad_norm": 0.25661540031433105,
+      "learning_rate": 0.001,
+      "loss": 1.9917,
+      "step": 9550
+    },
+    {
+      "epoch": 0.40405279634486846,
+      "grad_norm": 2.5997157096862793,
+      "learning_rate": 0.001,
+      "loss": 2.2229,
+      "step": 9551
+    },
+    {
+      "epoch": 0.4040951011083848,
+      "grad_norm": 0.198783740401268,
+      "learning_rate": 0.001,
+      "loss": 2.0679,
+      "step": 9552
+    },
+    {
+      "epoch": 0.40413740587190117,
+      "grad_norm": 0.18976983428001404,
+      "learning_rate": 0.001,
+      "loss": 2.141,
+      "step": 9553
+    },
+    {
+      "epoch": 0.4041797106354175,
+      "grad_norm": 0.3730868101119995,
+      "learning_rate": 0.001,
+      "loss": 2.0335,
+      "step": 9554
+    },
+    {
+      "epoch": 0.40422201539893393,
+      "grad_norm": 0.1948307752609253,
+      "learning_rate": 0.001,
+      "loss": 1.7714,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4042643201624503,
+      "grad_norm": 0.2243911772966385,
+      "learning_rate": 0.001,
+      "loss": 2.2354,
+      "step": 9556
+    },
+    {
+      "epoch": 0.40430662492596664,
+      "grad_norm": 0.24743397533893585,
+      "learning_rate": 0.001,
+      "loss": 2.2232,
+      "step": 9557
+    },
+    {
+      "epoch": 0.40434892968948305,
+      "grad_norm": 0.3496690094470978,
+      "learning_rate": 0.001,
+      "loss": 1.7223,
+      "step": 9558
+    },
+    {
+      "epoch": 0.4043912344529994,
+      "grad_norm": 0.3667478561401367,
+      "learning_rate": 0.001,
+      "loss": 2.9957,
+      "step": 9559
+    },
+    {
+      "epoch": 0.40443353921651576,
+      "grad_norm": 0.32227084040641785,
+      "learning_rate": 0.001,
+      "loss": 2.0967,
+      "step": 9560
+    },
+    {
+      "epoch": 0.40447584398003217,
+      "grad_norm": 0.18110904097557068,
+      "learning_rate": 0.001,
+      "loss": 1.6972,
+      "step": 9561
+    },
+    {
+      "epoch": 0.4045181487435485,
+      "grad_norm": 0.19031678140163422,
+      "learning_rate": 0.001,
+      "loss": 2.3908,
+      "step": 9562
+    },
+    {
+      "epoch": 0.4045604535070649,
+      "grad_norm": 0.17075388133525848,
+      "learning_rate": 0.001,
+      "loss": 1.9596,
+      "step": 9563
+    },
+    {
+      "epoch": 0.4046027582705813,
+      "grad_norm": 0.20088161528110504,
+      "learning_rate": 0.001,
+      "loss": 2.8288,
+      "step": 9564
+    },
+    {
+      "epoch": 0.40464506303409764,
+      "grad_norm": 0.19151663780212402,
+      "learning_rate": 0.001,
+      "loss": 1.8171,
+      "step": 9565
+    },
+    {
+      "epoch": 0.404687367797614,
+      "grad_norm": 0.19099347293376923,
+      "learning_rate": 0.001,
+      "loss": 2.3548,
+      "step": 9566
+    },
+    {
+      "epoch": 0.4047296725611304,
+      "grad_norm": 0.2712372839450836,
+      "learning_rate": 0.001,
+      "loss": 2.8372,
+      "step": 9567
+    },
+    {
+      "epoch": 0.40477197732464676,
+      "grad_norm": 0.17886675894260406,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 9568
+    },
+    {
+      "epoch": 0.4048142820881631,
+      "grad_norm": 46.942535400390625,
+      "learning_rate": 0.001,
+      "loss": 2.8675,
+      "step": 9569
+    },
+    {
+      "epoch": 0.4048565868516795,
+      "grad_norm": 0.2914670407772064,
+      "learning_rate": 0.001,
+      "loss": 1.7127,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4048988916151959,
+      "grad_norm": 0.1782679408788681,
+      "learning_rate": 0.001,
+      "loss": 1.7707,
+      "step": 9571
+    },
+    {
+      "epoch": 0.40494119637871223,
+      "grad_norm": 0.446181982755661,
+      "learning_rate": 0.001,
+      "loss": 2.3157,
+      "step": 9572
+    },
+    {
+      "epoch": 0.40498350114222864,
+      "grad_norm": 1.071265697479248,
+      "learning_rate": 0.001,
+      "loss": 2.153,
+      "step": 9573
+    },
+    {
+      "epoch": 0.405025805905745,
+      "grad_norm": 0.2291143536567688,
+      "learning_rate": 0.001,
+      "loss": 1.8002,
+      "step": 9574
+    },
+    {
+      "epoch": 0.40506811066926135,
+      "grad_norm": 0.21641968190670013,
+      "learning_rate": 0.001,
+      "loss": 2.2813,
+      "step": 9575
+    },
+    {
+      "epoch": 0.4051104154327777,
+      "grad_norm": 3.91137957572937,
+      "learning_rate": 0.001,
+      "loss": 2.7253,
+      "step": 9576
+    },
+    {
+      "epoch": 0.4051527201962941,
+      "grad_norm": 0.15815329551696777,
+      "learning_rate": 0.001,
+      "loss": 1.5105,
+      "step": 9577
+    },
+    {
+      "epoch": 0.40519502495981047,
+      "grad_norm": 0.43796873092651367,
+      "learning_rate": 0.001,
+      "loss": 2.2207,
+      "step": 9578
+    },
+    {
+      "epoch": 0.4052373297233268,
+      "grad_norm": 0.17552462220191956,
+      "learning_rate": 0.001,
+      "loss": 2.1559,
+      "step": 9579
+    },
+    {
+      "epoch": 0.40527963448684323,
+      "grad_norm": 0.2455400824546814,
+      "learning_rate": 0.001,
+      "loss": 1.8954,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4053219392503596,
+      "grad_norm": 0.9045276045799255,
+      "learning_rate": 0.001,
+      "loss": 2.028,
+      "step": 9581
+    },
+    {
+      "epoch": 0.40536424401387594,
+      "grad_norm": 0.16627678275108337,
+      "learning_rate": 0.001,
+      "loss": 1.7033,
+      "step": 9582
+    },
+    {
+      "epoch": 0.40540654877739235,
+      "grad_norm": 0.23107825219631195,
+      "learning_rate": 0.001,
+      "loss": 3.9189,
+      "step": 9583
+    },
+    {
+      "epoch": 0.4054488535409087,
+      "grad_norm": 0.5262654423713684,
+      "learning_rate": 0.001,
+      "loss": 2.7847,
+      "step": 9584
+    },
+    {
+      "epoch": 0.40549115830442506,
+      "grad_norm": 0.3403107225894928,
+      "learning_rate": 0.001,
+      "loss": 2.8312,
+      "step": 9585
+    },
+    {
+      "epoch": 0.40553346306794147,
+      "grad_norm": 0.19533272087574005,
+      "learning_rate": 0.001,
+      "loss": 2.0224,
+      "step": 9586
+    },
+    {
+      "epoch": 0.4055757678314578,
+      "grad_norm": 0.2007928043603897,
+      "learning_rate": 0.001,
+      "loss": 1.8877,
+      "step": 9587
+    },
+    {
+      "epoch": 0.4056180725949742,
+      "grad_norm": 0.2459113597869873,
+      "learning_rate": 0.001,
+      "loss": 2.3136,
+      "step": 9588
+    },
+    {
+      "epoch": 0.4056603773584906,
+      "grad_norm": 0.17115989327430725,
+      "learning_rate": 0.001,
+      "loss": 2.79,
+      "step": 9589
+    },
+    {
+      "epoch": 0.40570268212200694,
+      "grad_norm": 1.789944052696228,
+      "learning_rate": 0.001,
+      "loss": 2.5636,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4057449868855233,
+      "grad_norm": 0.5894603133201599,
+      "learning_rate": 0.001,
+      "loss": 2.3501,
+      "step": 9591
+    },
+    {
+      "epoch": 0.4057872916490397,
+      "grad_norm": 0.24456271529197693,
+      "learning_rate": 0.001,
+      "loss": 2.8854,
+      "step": 9592
+    },
+    {
+      "epoch": 0.40582959641255606,
+      "grad_norm": 0.15908919274806976,
+      "learning_rate": 0.001,
+      "loss": 1.5415,
+      "step": 9593
+    },
+    {
+      "epoch": 0.4058719011760724,
+      "grad_norm": 0.17225715517997742,
+      "learning_rate": 0.001,
+      "loss": 2.4391,
+      "step": 9594
+    },
+    {
+      "epoch": 0.4059142059395888,
+      "grad_norm": 0.17913804948329926,
+      "learning_rate": 0.001,
+      "loss": 1.4686,
+      "step": 9595
+    },
+    {
+      "epoch": 0.4059565107031052,
+      "grad_norm": 0.4248640239238739,
+      "learning_rate": 0.001,
+      "loss": 2.7928,
+      "step": 9596
+    },
+    {
+      "epoch": 0.40599881546662153,
+      "grad_norm": 1.1703771352767944,
+      "learning_rate": 0.001,
+      "loss": 2.104,
+      "step": 9597
+    },
+    {
+      "epoch": 0.4060411202301379,
+      "grad_norm": 0.18261665105819702,
+      "learning_rate": 0.001,
+      "loss": 1.6789,
+      "step": 9598
+    },
+    {
+      "epoch": 0.4060834249936543,
+      "grad_norm": 1.6488425731658936,
+      "learning_rate": 0.001,
+      "loss": 2.1651,
+      "step": 9599
+    },
+    {
+      "epoch": 0.40612572975717065,
+      "grad_norm": 1.800002098083496,
+      "learning_rate": 0.001,
+      "loss": 2.8898,
+      "step": 9600
+    },
+    {
+      "epoch": 0.406168034520687,
+      "grad_norm": 2.6963353157043457,
+      "learning_rate": 0.001,
+      "loss": 2.6452,
+      "step": 9601
+    },
+    {
+      "epoch": 0.4062103392842034,
+      "grad_norm": 0.2065376192331314,
+      "learning_rate": 0.001,
+      "loss": 1.9318,
+      "step": 9602
+    },
+    {
+      "epoch": 0.40625264404771977,
+      "grad_norm": 0.9833797812461853,
+      "learning_rate": 0.001,
+      "loss": 2.7088,
+      "step": 9603
+    },
+    {
+      "epoch": 0.4062949488112361,
+      "grad_norm": 0.332534521818161,
+      "learning_rate": 0.001,
+      "loss": 1.9894,
+      "step": 9604
+    },
+    {
+      "epoch": 0.40633725357475253,
+      "grad_norm": 0.2684883177280426,
+      "learning_rate": 0.001,
+      "loss": 2.1797,
+      "step": 9605
+    },
+    {
+      "epoch": 0.4063795583382689,
+      "grad_norm": 1.313679814338684,
+      "learning_rate": 0.001,
+      "loss": 3.2259,
+      "step": 9606
+    },
+    {
+      "epoch": 0.40642186310178524,
+      "grad_norm": 0.4700590968132019,
+      "learning_rate": 0.001,
+      "loss": 1.8352,
+      "step": 9607
+    },
+    {
+      "epoch": 0.40646416786530165,
+      "grad_norm": 0.29726094007492065,
+      "learning_rate": 0.001,
+      "loss": 2.2151,
+      "step": 9608
+    },
+    {
+      "epoch": 0.406506472628818,
+      "grad_norm": 0.20610110461711884,
+      "learning_rate": 0.001,
+      "loss": 1.5279,
+      "step": 9609
+    },
+    {
+      "epoch": 0.40654877739233436,
+      "grad_norm": 0.19683901965618134,
+      "learning_rate": 0.001,
+      "loss": 2.1613,
+      "step": 9610
+    },
+    {
+      "epoch": 0.40659108215585077,
+      "grad_norm": 0.37825819849967957,
+      "learning_rate": 0.001,
+      "loss": 3.1917,
+      "step": 9611
+    },
+    {
+      "epoch": 0.4066333869193671,
+      "grad_norm": 0.18482539057731628,
+      "learning_rate": 0.001,
+      "loss": 2.0728,
+      "step": 9612
+    },
+    {
+      "epoch": 0.4066756916828835,
+      "grad_norm": 0.2475501000881195,
+      "learning_rate": 0.001,
+      "loss": 2.9863,
+      "step": 9613
+    },
+    {
+      "epoch": 0.4067179964463999,
+      "grad_norm": 0.14848104119300842,
+      "learning_rate": 0.001,
+      "loss": 2.2224,
+      "step": 9614
+    },
+    {
+      "epoch": 0.40676030120991624,
+      "grad_norm": 0.4939437806606293,
+      "learning_rate": 0.001,
+      "loss": 2.6319,
+      "step": 9615
+    },
+    {
+      "epoch": 0.4068026059734326,
+      "grad_norm": 1.2136601209640503,
+      "learning_rate": 0.001,
+      "loss": 2.591,
+      "step": 9616
+    },
+    {
+      "epoch": 0.406844910736949,
+      "grad_norm": 1.4348440170288086,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 9617
+    },
+    {
+      "epoch": 0.40688721550046536,
+      "grad_norm": 0.2134755551815033,
+      "learning_rate": 0.001,
+      "loss": 1.9658,
+      "step": 9618
+    },
+    {
+      "epoch": 0.4069295202639817,
+      "grad_norm": 1.9099276065826416,
+      "learning_rate": 0.001,
+      "loss": 2.0009,
+      "step": 9619
+    },
+    {
+      "epoch": 0.40697182502749807,
+      "grad_norm": 0.2304713875055313,
+      "learning_rate": 0.001,
+      "loss": 2.0757,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4070141297910145,
+      "grad_norm": 0.2379099726676941,
+      "learning_rate": 0.001,
+      "loss": 2.3855,
+      "step": 9621
+    },
+    {
+      "epoch": 0.40705643455453083,
+      "grad_norm": 0.2802619934082031,
+      "learning_rate": 0.001,
+      "loss": 2.562,
+      "step": 9622
+    },
+    {
+      "epoch": 0.4070987393180472,
+      "grad_norm": 0.2311367690563202,
+      "learning_rate": 0.001,
+      "loss": 3.7875,
+      "step": 9623
+    },
+    {
+      "epoch": 0.4071410440815636,
+      "grad_norm": 0.4283418655395508,
+      "learning_rate": 0.001,
+      "loss": 1.7183,
+      "step": 9624
+    },
+    {
+      "epoch": 0.40718334884507995,
+      "grad_norm": 0.31401526927948,
+      "learning_rate": 0.001,
+      "loss": 2.5655,
+      "step": 9625
+    },
+    {
+      "epoch": 0.4072256536085963,
+      "grad_norm": 0.22191263735294342,
+      "learning_rate": 0.001,
+      "loss": 2.251,
+      "step": 9626
+    },
+    {
+      "epoch": 0.4072679583721127,
+      "grad_norm": 0.23123349249362946,
+      "learning_rate": 0.001,
+      "loss": 2.5949,
+      "step": 9627
+    },
+    {
+      "epoch": 0.40731026313562907,
+      "grad_norm": 0.20630641281604767,
+      "learning_rate": 0.001,
+      "loss": 2.8717,
+      "step": 9628
+    },
+    {
+      "epoch": 0.4073525678991454,
+      "grad_norm": 1.072459101676941,
+      "learning_rate": 0.001,
+      "loss": 2.4395,
+      "step": 9629
+    },
+    {
+      "epoch": 0.40739487266266183,
+      "grad_norm": 0.22625629603862762,
+      "learning_rate": 0.001,
+      "loss": 2.2453,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4074371774261782,
+      "grad_norm": 1.8554446697235107,
+      "learning_rate": 0.001,
+      "loss": 2.1412,
+      "step": 9631
+    },
+    {
+      "epoch": 0.40747948218969454,
+      "grad_norm": 0.2199331372976303,
+      "learning_rate": 0.001,
+      "loss": 1.5662,
+      "step": 9632
+    },
+    {
+      "epoch": 0.40752178695321095,
+      "grad_norm": 1.69846773147583,
+      "learning_rate": 0.001,
+      "loss": 2.1245,
+      "step": 9633
+    },
+    {
+      "epoch": 0.4075640917167273,
+      "grad_norm": 0.22622716426849365,
+      "learning_rate": 0.001,
+      "loss": 2.1381,
+      "step": 9634
+    },
+    {
+      "epoch": 0.40760639648024366,
+      "grad_norm": 0.21170629560947418,
+      "learning_rate": 0.001,
+      "loss": 2.0326,
+      "step": 9635
+    },
+    {
+      "epoch": 0.40764870124376007,
+      "grad_norm": 0.26057133078575134,
+      "learning_rate": 0.001,
+      "loss": 1.9696,
+      "step": 9636
+    },
+    {
+      "epoch": 0.4076910060072764,
+      "grad_norm": 5.649323463439941,
+      "learning_rate": 0.001,
+      "loss": 2.8521,
+      "step": 9637
+    },
+    {
+      "epoch": 0.4077333107707928,
+      "grad_norm": 0.32760143280029297,
+      "learning_rate": 0.001,
+      "loss": 3.0351,
+      "step": 9638
+    },
+    {
+      "epoch": 0.4077756155343092,
+      "grad_norm": 1.6153209209442139,
+      "learning_rate": 0.001,
+      "loss": 1.9322,
+      "step": 9639
+    },
+    {
+      "epoch": 0.40781792029782554,
+      "grad_norm": 1.4140678644180298,
+      "learning_rate": 0.001,
+      "loss": 2.2613,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4078602250613419,
+      "grad_norm": 0.33353015780448914,
+      "learning_rate": 0.001,
+      "loss": 3.0011,
+      "step": 9641
+    },
+    {
+      "epoch": 0.4079025298248583,
+      "grad_norm": 27.971038818359375,
+      "learning_rate": 0.001,
+      "loss": 2.6894,
+      "step": 9642
+    },
+    {
+      "epoch": 0.40794483458837466,
+      "grad_norm": 0.9342984557151794,
+      "learning_rate": 0.001,
+      "loss": 2.7896,
+      "step": 9643
+    },
+    {
+      "epoch": 0.407987139351891,
+      "grad_norm": 2.035511016845703,
+      "learning_rate": 0.001,
+      "loss": 2.9743,
+      "step": 9644
+    },
+    {
+      "epoch": 0.40802944411540737,
+      "grad_norm": 0.682000994682312,
+      "learning_rate": 0.001,
+      "loss": 2.2909,
+      "step": 9645
+    },
+    {
+      "epoch": 0.4080717488789238,
+      "grad_norm": 6.955329418182373,
+      "learning_rate": 0.001,
+      "loss": 1.8326,
+      "step": 9646
+    },
+    {
+      "epoch": 0.40811405364244013,
+      "grad_norm": 0.23917068541049957,
+      "learning_rate": 0.001,
+      "loss": 2.0451,
+      "step": 9647
+    },
+    {
+      "epoch": 0.4081563584059565,
+      "grad_norm": 0.3635198473930359,
+      "learning_rate": 0.001,
+      "loss": 3.2204,
+      "step": 9648
+    },
+    {
+      "epoch": 0.4081986631694729,
+      "grad_norm": 0.5178560614585876,
+      "learning_rate": 0.001,
+      "loss": 3.7043,
+      "step": 9649
+    },
+    {
+      "epoch": 0.40824096793298925,
+      "grad_norm": 0.21071864664554596,
+      "learning_rate": 0.001,
+      "loss": 2.7752,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4082832726965056,
+      "grad_norm": 0.22183898091316223,
+      "learning_rate": 0.001,
+      "loss": 2.269,
+      "step": 9651
+    },
+    {
+      "epoch": 0.408325577460022,
+      "grad_norm": 0.19146625697612762,
+      "learning_rate": 0.001,
+      "loss": 2.5297,
+      "step": 9652
+    },
+    {
+      "epoch": 0.40836788222353837,
+      "grad_norm": 0.2734749913215637,
+      "learning_rate": 0.001,
+      "loss": 2.3894,
+      "step": 9653
+    },
+    {
+      "epoch": 0.4084101869870547,
+      "grad_norm": 0.24852819740772247,
+      "learning_rate": 0.001,
+      "loss": 2.2241,
+      "step": 9654
+    },
+    {
+      "epoch": 0.40845249175057113,
+      "grad_norm": 0.2272268384695053,
+      "learning_rate": 0.001,
+      "loss": 2.39,
+      "step": 9655
+    },
+    {
+      "epoch": 0.4084947965140875,
+      "grad_norm": 0.5135742425918579,
+      "learning_rate": 0.001,
+      "loss": 3.2057,
+      "step": 9656
+    },
+    {
+      "epoch": 0.40853710127760384,
+      "grad_norm": 0.4296305477619171,
+      "learning_rate": 0.001,
+      "loss": 2.3507,
+      "step": 9657
+    },
+    {
+      "epoch": 0.40857940604112025,
+      "grad_norm": 0.17050939798355103,
+      "learning_rate": 0.001,
+      "loss": 1.5197,
+      "step": 9658
+    },
+    {
+      "epoch": 0.4086217108046366,
+      "grad_norm": 4.443729877471924,
+      "learning_rate": 0.001,
+      "loss": 2.8856,
+      "step": 9659
+    },
+    {
+      "epoch": 0.40866401556815296,
+      "grad_norm": 0.18589919805526733,
+      "learning_rate": 0.001,
+      "loss": 1.711,
+      "step": 9660
+    },
+    {
+      "epoch": 0.40870632033166937,
+      "grad_norm": 0.9820543527603149,
+      "learning_rate": 0.001,
+      "loss": 2.7134,
+      "step": 9661
+    },
+    {
+      "epoch": 0.4087486250951857,
+      "grad_norm": 3.852842092514038,
+      "learning_rate": 0.001,
+      "loss": 1.894,
+      "step": 9662
+    },
+    {
+      "epoch": 0.4087909298587021,
+      "grad_norm": 0.6740947365760803,
+      "learning_rate": 0.001,
+      "loss": 1.9484,
+      "step": 9663
+    },
+    {
+      "epoch": 0.4088332346222185,
+      "grad_norm": 0.22364242374897003,
+      "learning_rate": 0.001,
+      "loss": 2.7015,
+      "step": 9664
+    },
+    {
+      "epoch": 0.40887553938573484,
+      "grad_norm": 0.283769428730011,
+      "learning_rate": 0.001,
+      "loss": 2.3642,
+      "step": 9665
+    },
+    {
+      "epoch": 0.4089178441492512,
+      "grad_norm": 0.9470567107200623,
+      "learning_rate": 0.001,
+      "loss": 2.4213,
+      "step": 9666
+    },
+    {
+      "epoch": 0.40896014891276755,
+      "grad_norm": 0.29617589712142944,
+      "learning_rate": 0.001,
+      "loss": 2.9135,
+      "step": 9667
+    },
+    {
+      "epoch": 0.40900245367628396,
+      "grad_norm": 0.20563291013240814,
+      "learning_rate": 0.001,
+      "loss": 2.1679,
+      "step": 9668
+    },
+    {
+      "epoch": 0.4090447584398003,
+      "grad_norm": 0.18204866349697113,
+      "learning_rate": 0.001,
+      "loss": 2.6882,
+      "step": 9669
+    },
+    {
+      "epoch": 0.40908706320331667,
+      "grad_norm": 0.2885321080684662,
+      "learning_rate": 0.001,
+      "loss": 2.095,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4091293679668331,
+      "grad_norm": 2.8684329986572266,
+      "learning_rate": 0.001,
+      "loss": 2.4832,
+      "step": 9671
+    },
+    {
+      "epoch": 0.40917167273034943,
+      "grad_norm": 0.32784703373908997,
+      "learning_rate": 0.001,
+      "loss": 2.5857,
+      "step": 9672
+    },
+    {
+      "epoch": 0.4092139774938658,
+      "grad_norm": 0.7078139185905457,
+      "learning_rate": 0.001,
+      "loss": 1.8139,
+      "step": 9673
+    },
+    {
+      "epoch": 0.4092562822573822,
+      "grad_norm": 0.25039342045783997,
+      "learning_rate": 0.001,
+      "loss": 3.3178,
+      "step": 9674
+    },
+    {
+      "epoch": 0.40929858702089855,
+      "grad_norm": 0.17222987115383148,
+      "learning_rate": 0.001,
+      "loss": 2.2217,
+      "step": 9675
+    },
+    {
+      "epoch": 0.4093408917844149,
+      "grad_norm": 0.15359899401664734,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 9676
+    },
+    {
+      "epoch": 0.4093831965479313,
+      "grad_norm": 0.2080516666173935,
+      "learning_rate": 0.001,
+      "loss": 2.7983,
+      "step": 9677
+    },
+    {
+      "epoch": 0.40942550131144767,
+      "grad_norm": 0.17450571060180664,
+      "learning_rate": 0.001,
+      "loss": 1.8784,
+      "step": 9678
+    },
+    {
+      "epoch": 0.409467806074964,
+      "grad_norm": 0.19042570888996124,
+      "learning_rate": 0.001,
+      "loss": 1.8635,
+      "step": 9679
+    },
+    {
+      "epoch": 0.40951011083848043,
+      "grad_norm": 0.17510387301445007,
+      "learning_rate": 0.001,
+      "loss": 2.8996,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4095524156019968,
+      "grad_norm": 0.7525598406791687,
+      "learning_rate": 0.001,
+      "loss": 2.7079,
+      "step": 9681
+    },
+    {
+      "epoch": 0.40959472036551314,
+      "grad_norm": 1.063751459121704,
+      "learning_rate": 0.001,
+      "loss": 2.0613,
+      "step": 9682
+    },
+    {
+      "epoch": 0.40963702512902955,
+      "grad_norm": 0.35080280900001526,
+      "learning_rate": 0.001,
+      "loss": 4.928,
+      "step": 9683
+    },
+    {
+      "epoch": 0.4096793298925459,
+      "grad_norm": 0.2547611594200134,
+      "learning_rate": 0.001,
+      "loss": 2.8568,
+      "step": 9684
+    },
+    {
+      "epoch": 0.40972163465606226,
+      "grad_norm": 0.23505434393882751,
+      "learning_rate": 0.001,
+      "loss": 2.4675,
+      "step": 9685
+    },
+    {
+      "epoch": 0.40976393941957867,
+      "grad_norm": 0.2235456109046936,
+      "learning_rate": 0.001,
+      "loss": 2.3584,
+      "step": 9686
+    },
+    {
+      "epoch": 0.409806244183095,
+      "grad_norm": 0.26593148708343506,
+      "learning_rate": 0.001,
+      "loss": 2.3399,
+      "step": 9687
+    },
+    {
+      "epoch": 0.4098485489466114,
+      "grad_norm": 0.38054850697517395,
+      "learning_rate": 0.001,
+      "loss": 1.9254,
+      "step": 9688
+    },
+    {
+      "epoch": 0.40989085371012773,
+      "grad_norm": 0.18785469233989716,
+      "learning_rate": 0.001,
+      "loss": 2.3267,
+      "step": 9689
+    },
+    {
+      "epoch": 0.40993315847364414,
+      "grad_norm": 0.19680415093898773,
+      "learning_rate": 0.001,
+      "loss": 2.1284,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4099754632371605,
+      "grad_norm": 0.8819651007652283,
+      "learning_rate": 0.001,
+      "loss": 2.2144,
+      "step": 9691
+    },
+    {
+      "epoch": 0.41001776800067685,
+      "grad_norm": 0.36599695682525635,
+      "learning_rate": 0.001,
+      "loss": 2.5738,
+      "step": 9692
+    },
+    {
+      "epoch": 0.41006007276419326,
+      "grad_norm": 1.350098967552185,
+      "learning_rate": 0.001,
+      "loss": 3.2411,
+      "step": 9693
+    },
+    {
+      "epoch": 0.4101023775277096,
+      "grad_norm": 0.252407968044281,
+      "learning_rate": 0.001,
+      "loss": 3.065,
+      "step": 9694
+    },
+    {
+      "epoch": 0.41014468229122597,
+      "grad_norm": 5.906357765197754,
+      "learning_rate": 0.001,
+      "loss": 1.996,
+      "step": 9695
+    },
+    {
+      "epoch": 0.4101869870547424,
+      "grad_norm": 0.30286088585853577,
+      "learning_rate": 0.001,
+      "loss": 2.4888,
+      "step": 9696
+    },
+    {
+      "epoch": 0.41022929181825873,
+      "grad_norm": 11.067898750305176,
+      "learning_rate": 0.001,
+      "loss": 2.2548,
+      "step": 9697
+    },
+    {
+      "epoch": 0.4102715965817751,
+      "grad_norm": 0.21052215993404388,
+      "learning_rate": 0.001,
+      "loss": 3.1103,
+      "step": 9698
+    },
+    {
+      "epoch": 0.4103139013452915,
+      "grad_norm": 0.24739009141921997,
+      "learning_rate": 0.001,
+      "loss": 3.126,
+      "step": 9699
+    },
+    {
+      "epoch": 0.41035620610880785,
+      "grad_norm": 0.22047722339630127,
+      "learning_rate": 0.001,
+      "loss": 2.3277,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4103985108723242,
+      "grad_norm": 0.7261704802513123,
+      "learning_rate": 0.001,
+      "loss": 2.8432,
+      "step": 9701
+    },
+    {
+      "epoch": 0.4104408156358406,
+      "grad_norm": 0.2856643795967102,
+      "learning_rate": 0.001,
+      "loss": 1.9512,
+      "step": 9702
+    },
+    {
+      "epoch": 0.41048312039935697,
+      "grad_norm": 0.49634721875190735,
+      "learning_rate": 0.001,
+      "loss": 2.1401,
+      "step": 9703
+    },
+    {
+      "epoch": 0.4105254251628733,
+      "grad_norm": 0.24824899435043335,
+      "learning_rate": 0.001,
+      "loss": 3.2889,
+      "step": 9704
+    },
+    {
+      "epoch": 0.41056772992638974,
+      "grad_norm": 0.20967476069927216,
+      "learning_rate": 0.001,
+      "loss": 2.091,
+      "step": 9705
+    },
+    {
+      "epoch": 0.4106100346899061,
+      "grad_norm": 0.1925218403339386,
+      "learning_rate": 0.001,
+      "loss": 1.8878,
+      "step": 9706
+    },
+    {
+      "epoch": 0.41065233945342244,
+      "grad_norm": 1.0082390308380127,
+      "learning_rate": 0.001,
+      "loss": 1.9122,
+      "step": 9707
+    },
+    {
+      "epoch": 0.41069464421693885,
+      "grad_norm": 6.59636926651001,
+      "learning_rate": 0.001,
+      "loss": 1.9186,
+      "step": 9708
+    },
+    {
+      "epoch": 0.4107369489804552,
+      "grad_norm": 0.2302708476781845,
+      "learning_rate": 0.001,
+      "loss": 2.3975,
+      "step": 9709
+    },
+    {
+      "epoch": 0.41077925374397156,
+      "grad_norm": 5.976405143737793,
+      "learning_rate": 0.001,
+      "loss": 2.8853,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4108215585074879,
+      "grad_norm": 5.888969421386719,
+      "learning_rate": 0.001,
+      "loss": 1.6746,
+      "step": 9711
+    },
+    {
+      "epoch": 0.4108638632710043,
+      "grad_norm": 2.457515001296997,
+      "learning_rate": 0.001,
+      "loss": 3.2117,
+      "step": 9712
+    },
+    {
+      "epoch": 0.4109061680345207,
+      "grad_norm": 0.266607403755188,
+      "learning_rate": 0.001,
+      "loss": 1.9498,
+      "step": 9713
+    },
+    {
+      "epoch": 0.41094847279803703,
+      "grad_norm": 1.2756965160369873,
+      "learning_rate": 0.001,
+      "loss": 2.7778,
+      "step": 9714
+    },
+    {
+      "epoch": 0.41099077756155344,
+      "grad_norm": 0.4476830065250397,
+      "learning_rate": 0.001,
+      "loss": 3.3292,
+      "step": 9715
+    },
+    {
+      "epoch": 0.4110330823250698,
+      "grad_norm": 0.2495070993900299,
+      "learning_rate": 0.001,
+      "loss": 2.3151,
+      "step": 9716
+    },
+    {
+      "epoch": 0.41107538708858615,
+      "grad_norm": 0.30039656162261963,
+      "learning_rate": 0.001,
+      "loss": 1.9914,
+      "step": 9717
+    },
+    {
+      "epoch": 0.41111769185210256,
+      "grad_norm": 0.47619497776031494,
+      "learning_rate": 0.001,
+      "loss": 3.4457,
+      "step": 9718
+    },
+    {
+      "epoch": 0.4111599966156189,
+      "grad_norm": 0.21117328107357025,
+      "learning_rate": 0.001,
+      "loss": 1.7526,
+      "step": 9719
+    },
+    {
+      "epoch": 0.41120230137913527,
+      "grad_norm": 0.4838405251502991,
+      "learning_rate": 0.001,
+      "loss": 2.325,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4112446061426517,
+      "grad_norm": 6.137927532196045,
+      "learning_rate": 0.001,
+      "loss": 3.0532,
+      "step": 9721
+    },
+    {
+      "epoch": 0.41128691090616804,
+      "grad_norm": 0.5445790886878967,
+      "learning_rate": 0.001,
+      "loss": 2.6303,
+      "step": 9722
+    },
+    {
+      "epoch": 0.4113292156696844,
+      "grad_norm": 0.24159128963947296,
+      "learning_rate": 0.001,
+      "loss": 2.4226,
+      "step": 9723
+    },
+    {
+      "epoch": 0.4113715204332008,
+      "grad_norm": 0.29332977533340454,
+      "learning_rate": 0.001,
+      "loss": 2.8964,
+      "step": 9724
+    },
+    {
+      "epoch": 0.41141382519671715,
+      "grad_norm": 0.6242585182189941,
+      "learning_rate": 0.001,
+      "loss": 2.5372,
+      "step": 9725
+    },
+    {
+      "epoch": 0.4114561299602335,
+      "grad_norm": 0.2993077039718628,
+      "learning_rate": 0.001,
+      "loss": 2.4038,
+      "step": 9726
+    },
+    {
+      "epoch": 0.4114984347237499,
+      "grad_norm": 0.4130201041698456,
+      "learning_rate": 0.001,
+      "loss": 2.5595,
+      "step": 9727
+    },
+    {
+      "epoch": 0.41154073948726627,
+      "grad_norm": 0.3834516406059265,
+      "learning_rate": 0.001,
+      "loss": 3.7084,
+      "step": 9728
+    },
+    {
+      "epoch": 0.4115830442507826,
+      "grad_norm": 0.29505565762519836,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 9729
+    },
+    {
+      "epoch": 0.41162534901429904,
+      "grad_norm": 0.4178231954574585,
+      "learning_rate": 0.001,
+      "loss": 2.9802,
+      "step": 9730
+    },
+    {
+      "epoch": 0.4116676537778154,
+      "grad_norm": 0.2813502252101898,
+      "learning_rate": 0.001,
+      "loss": 1.898,
+      "step": 9731
+    },
+    {
+      "epoch": 0.41170995854133174,
+      "grad_norm": 0.24282923340797424,
+      "learning_rate": 0.001,
+      "loss": 3.137,
+      "step": 9732
+    },
+    {
+      "epoch": 0.4117522633048481,
+      "grad_norm": 0.21911819279193878,
+      "learning_rate": 0.001,
+      "loss": 2.452,
+      "step": 9733
+    },
+    {
+      "epoch": 0.4117945680683645,
+      "grad_norm": 0.6716076731681824,
+      "learning_rate": 0.001,
+      "loss": 2.1814,
+      "step": 9734
+    },
+    {
+      "epoch": 0.41183687283188086,
+      "grad_norm": 0.9324064254760742,
+      "learning_rate": 0.001,
+      "loss": 3.5659,
+      "step": 9735
+    },
+    {
+      "epoch": 0.4118791775953972,
+      "grad_norm": 0.3203246593475342,
+      "learning_rate": 0.001,
+      "loss": 2.3815,
+      "step": 9736
+    },
+    {
+      "epoch": 0.4119214823589136,
+      "grad_norm": 1.0174567699432373,
+      "learning_rate": 0.001,
+      "loss": 2.8275,
+      "step": 9737
+    },
+    {
+      "epoch": 0.41196378712243,
+      "grad_norm": 5.53692626953125,
+      "learning_rate": 0.001,
+      "loss": 1.5753,
+      "step": 9738
+    },
+    {
+      "epoch": 0.41200609188594634,
+      "grad_norm": 0.3196747303009033,
+      "learning_rate": 0.001,
+      "loss": 2.2303,
+      "step": 9739
+    },
+    {
+      "epoch": 0.41204839664946274,
+      "grad_norm": 7.535372257232666,
+      "learning_rate": 0.001,
+      "loss": 2.7096,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4120907014129791,
+      "grad_norm": 0.5855947136878967,
+      "learning_rate": 0.001,
+      "loss": 2.4305,
+      "step": 9741
+    },
+    {
+      "epoch": 0.41213300617649545,
+      "grad_norm": 0.4567291736602783,
+      "learning_rate": 0.001,
+      "loss": 2.9843,
+      "step": 9742
+    },
+    {
+      "epoch": 0.41217531094001186,
+      "grad_norm": 0.276134729385376,
+      "learning_rate": 0.001,
+      "loss": 2.6929,
+      "step": 9743
+    },
+    {
+      "epoch": 0.4122176157035282,
+      "grad_norm": 2.817307472229004,
+      "learning_rate": 0.001,
+      "loss": 2.2315,
+      "step": 9744
+    },
+    {
+      "epoch": 0.41225992046704457,
+      "grad_norm": 1.8684515953063965,
+      "learning_rate": 0.001,
+      "loss": 1.7073,
+      "step": 9745
+    },
+    {
+      "epoch": 0.412302225230561,
+      "grad_norm": 0.2631063759326935,
+      "learning_rate": 0.001,
+      "loss": 2.4196,
+      "step": 9746
+    },
+    {
+      "epoch": 0.41234452999407734,
+      "grad_norm": 0.9968166947364807,
+      "learning_rate": 0.001,
+      "loss": 2.3738,
+      "step": 9747
+    },
+    {
+      "epoch": 0.4123868347575937,
+      "grad_norm": 0.2650030851364136,
+      "learning_rate": 0.001,
+      "loss": 2.0034,
+      "step": 9748
+    },
+    {
+      "epoch": 0.4124291395211101,
+      "grad_norm": 0.18941166996955872,
+      "learning_rate": 0.001,
+      "loss": 2.4076,
+      "step": 9749
+    },
+    {
+      "epoch": 0.41247144428462645,
+      "grad_norm": 16.938201904296875,
+      "learning_rate": 0.001,
+      "loss": 2.0145,
+      "step": 9750
+    },
+    {
+      "epoch": 0.4125137490481428,
+      "grad_norm": 2.8237738609313965,
+      "learning_rate": 0.001,
+      "loss": 2.1931,
+      "step": 9751
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.2568257749080658,
+      "learning_rate": 0.001,
+      "loss": 2.2738,
+      "step": 9752
+    },
+    {
+      "epoch": 0.41259835857517557,
+      "grad_norm": 0.2610342502593994,
+      "learning_rate": 0.001,
+      "loss": 2.0419,
+      "step": 9753
+    },
+    {
+      "epoch": 0.4126406633386919,
+      "grad_norm": 0.22543591260910034,
+      "learning_rate": 0.001,
+      "loss": 2.1172,
+      "step": 9754
+    },
+    {
+      "epoch": 0.41268296810220834,
+      "grad_norm": 1.7136520147323608,
+      "learning_rate": 0.001,
+      "loss": 2.7651,
+      "step": 9755
+    },
+    {
+      "epoch": 0.4127252728657247,
+      "grad_norm": 1.0864278078079224,
+      "learning_rate": 0.001,
+      "loss": 2.6813,
+      "step": 9756
+    },
+    {
+      "epoch": 0.41276757762924104,
+      "grad_norm": 0.16335022449493408,
+      "learning_rate": 0.001,
+      "loss": 2.7791,
+      "step": 9757
+    },
+    {
+      "epoch": 0.4128098823927574,
+      "grad_norm": 0.2040264904499054,
+      "learning_rate": 0.001,
+      "loss": 2.0147,
+      "step": 9758
+    },
+    {
+      "epoch": 0.4128521871562738,
+      "grad_norm": 0.23535273969173431,
+      "learning_rate": 0.001,
+      "loss": 2.2457,
+      "step": 9759
+    },
+    {
+      "epoch": 0.41289449191979016,
+      "grad_norm": 0.2834819257259369,
+      "learning_rate": 0.001,
+      "loss": 2.5383,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4129367966833065,
+      "grad_norm": 0.44039419293403625,
+      "learning_rate": 0.001,
+      "loss": 2.2677,
+      "step": 9761
+    },
+    {
+      "epoch": 0.4129791014468229,
+      "grad_norm": 0.19605425000190735,
+      "learning_rate": 0.001,
+      "loss": 2.0914,
+      "step": 9762
+    },
+    {
+      "epoch": 0.4130214062103393,
+      "grad_norm": 0.3128882050514221,
+      "learning_rate": 0.001,
+      "loss": 2.4405,
+      "step": 9763
+    },
+    {
+      "epoch": 0.41306371097385564,
+      "grad_norm": 0.19780918955802917,
+      "learning_rate": 0.001,
+      "loss": 2.9869,
+      "step": 9764
+    },
+    {
+      "epoch": 0.41310601573737205,
+      "grad_norm": 0.7669022679328918,
+      "learning_rate": 0.001,
+      "loss": 2.0915,
+      "step": 9765
+    },
+    {
+      "epoch": 0.4131483205008884,
+      "grad_norm": 1.3960890769958496,
+      "learning_rate": 0.001,
+      "loss": 2.018,
+      "step": 9766
+    },
+    {
+      "epoch": 0.41319062526440475,
+      "grad_norm": 0.4077872931957245,
+      "learning_rate": 0.001,
+      "loss": 2.4418,
+      "step": 9767
+    },
+    {
+      "epoch": 0.41323293002792116,
+      "grad_norm": 0.22895626723766327,
+      "learning_rate": 0.001,
+      "loss": 3.1219,
+      "step": 9768
+    },
+    {
+      "epoch": 0.4132752347914375,
+      "grad_norm": 0.3246360719203949,
+      "learning_rate": 0.001,
+      "loss": 2.5731,
+      "step": 9769
+    },
+    {
+      "epoch": 0.41331753955495387,
+      "grad_norm": 0.1943444311618805,
+      "learning_rate": 0.001,
+      "loss": 2.6833,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4133598443184703,
+      "grad_norm": 0.23944945633411407,
+      "learning_rate": 0.001,
+      "loss": 2.1968,
+      "step": 9771
+    },
+    {
+      "epoch": 0.41340214908198664,
+      "grad_norm": 0.2529198229312897,
+      "learning_rate": 0.001,
+      "loss": 2.5286,
+      "step": 9772
+    },
+    {
+      "epoch": 0.413444453845503,
+      "grad_norm": 1.6711100339889526,
+      "learning_rate": 0.001,
+      "loss": 2.4631,
+      "step": 9773
+    },
+    {
+      "epoch": 0.4134867586090194,
+      "grad_norm": 0.16065281629562378,
+      "learning_rate": 0.001,
+      "loss": 2.8067,
+      "step": 9774
+    },
+    {
+      "epoch": 0.41352906337253575,
+      "grad_norm": 0.1823013722896576,
+      "learning_rate": 0.001,
+      "loss": 2.4533,
+      "step": 9775
+    },
+    {
+      "epoch": 0.4135713681360521,
+      "grad_norm": 0.5769549012184143,
+      "learning_rate": 0.001,
+      "loss": 1.9118,
+      "step": 9776
+    },
+    {
+      "epoch": 0.4136136728995685,
+      "grad_norm": 0.3120274841785431,
+      "learning_rate": 0.001,
+      "loss": 2.6537,
+      "step": 9777
+    },
+    {
+      "epoch": 0.4136559776630849,
+      "grad_norm": 2.704646110534668,
+      "learning_rate": 0.001,
+      "loss": 2.6022,
+      "step": 9778
+    },
+    {
+      "epoch": 0.4136982824266012,
+      "grad_norm": 3.760073661804199,
+      "learning_rate": 0.001,
+      "loss": 2.4984,
+      "step": 9779
+    },
+    {
+      "epoch": 0.4137405871901176,
+      "grad_norm": 0.3028760254383087,
+      "learning_rate": 0.001,
+      "loss": 2.7438,
+      "step": 9780
+    },
+    {
+      "epoch": 0.413782891953634,
+      "grad_norm": 0.19027100503444672,
+      "learning_rate": 0.001,
+      "loss": 2.5773,
+      "step": 9781
+    },
+    {
+      "epoch": 0.41382519671715035,
+      "grad_norm": 0.703472375869751,
+      "learning_rate": 0.001,
+      "loss": 1.811,
+      "step": 9782
+    },
+    {
+      "epoch": 0.4138675014806667,
+      "grad_norm": 0.26506584882736206,
+      "learning_rate": 0.001,
+      "loss": 2.8884,
+      "step": 9783
+    },
+    {
+      "epoch": 0.4139098062441831,
+      "grad_norm": 0.2612300515174866,
+      "learning_rate": 0.001,
+      "loss": 2.7629,
+      "step": 9784
+    },
+    {
+      "epoch": 0.41395211100769946,
+      "grad_norm": 0.2701547145843506,
+      "learning_rate": 0.001,
+      "loss": 2.6545,
+      "step": 9785
+    },
+    {
+      "epoch": 0.4139944157712158,
+      "grad_norm": 0.5775189995765686,
+      "learning_rate": 0.001,
+      "loss": 2.2237,
+      "step": 9786
+    },
+    {
+      "epoch": 0.4140367205347322,
+      "grad_norm": 0.29414102435112,
+      "learning_rate": 0.001,
+      "loss": 2.5268,
+      "step": 9787
+    },
+    {
+      "epoch": 0.4140790252982486,
+      "grad_norm": 0.3888061046600342,
+      "learning_rate": 0.001,
+      "loss": 2.466,
+      "step": 9788
+    },
+    {
+      "epoch": 0.41412133006176494,
+      "grad_norm": 0.42892327904701233,
+      "learning_rate": 0.001,
+      "loss": 2.6945,
+      "step": 9789
+    },
+    {
+      "epoch": 0.41416363482528135,
+      "grad_norm": 0.3391939103603363,
+      "learning_rate": 0.001,
+      "loss": 2.2967,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4142059395887977,
+      "grad_norm": 0.1967991441488266,
+      "learning_rate": 0.001,
+      "loss": 2.2695,
+      "step": 9791
+    },
+    {
+      "epoch": 0.41424824435231405,
+      "grad_norm": 0.18619079887866974,
+      "learning_rate": 0.001,
+      "loss": 2.6969,
+      "step": 9792
+    },
+    {
+      "epoch": 0.41429054911583046,
+      "grad_norm": 0.21667589247226715,
+      "learning_rate": 0.001,
+      "loss": 2.346,
+      "step": 9793
+    },
+    {
+      "epoch": 0.4143328538793468,
+      "grad_norm": 0.1584491729736328,
+      "learning_rate": 0.001,
+      "loss": 1.6244,
+      "step": 9794
+    },
+    {
+      "epoch": 0.4143751586428632,
+      "grad_norm": 0.16956403851509094,
+      "learning_rate": 0.001,
+      "loss": 2.5346,
+      "step": 9795
+    },
+    {
+      "epoch": 0.4144174634063796,
+      "grad_norm": 0.2773437798023224,
+      "learning_rate": 0.001,
+      "loss": 3.2338,
+      "step": 9796
+    },
+    {
+      "epoch": 0.41445976816989594,
+      "grad_norm": 0.2252512127161026,
+      "learning_rate": 0.001,
+      "loss": 2.1501,
+      "step": 9797
+    },
+    {
+      "epoch": 0.4145020729334123,
+      "grad_norm": 0.9727465510368347,
+      "learning_rate": 0.001,
+      "loss": 2.7049,
+      "step": 9798
+    },
+    {
+      "epoch": 0.4145443776969287,
+      "grad_norm": 0.23791512846946716,
+      "learning_rate": 0.001,
+      "loss": 1.7446,
+      "step": 9799
+    },
+    {
+      "epoch": 0.41458668246044506,
+      "grad_norm": 14.202518463134766,
+      "learning_rate": 0.001,
+      "loss": 3.8168,
+      "step": 9800
+    },
+    {
+      "epoch": 0.4146289872239614,
+      "grad_norm": 0.44175243377685547,
+      "learning_rate": 0.001,
+      "loss": 2.1732,
+      "step": 9801
+    },
+    {
+      "epoch": 0.41467129198747776,
+      "grad_norm": 0.17590700089931488,
+      "learning_rate": 0.001,
+      "loss": 1.9734,
+      "step": 9802
+    },
+    {
+      "epoch": 0.4147135967509942,
+      "grad_norm": 0.18787901103496552,
+      "learning_rate": 0.001,
+      "loss": 2.4456,
+      "step": 9803
+    },
+    {
+      "epoch": 0.4147559015145105,
+      "grad_norm": 0.17629052698612213,
+      "learning_rate": 0.001,
+      "loss": 1.8945,
+      "step": 9804
+    },
+    {
+      "epoch": 0.4147982062780269,
+      "grad_norm": 0.17695337533950806,
+      "learning_rate": 0.001,
+      "loss": 2.3594,
+      "step": 9805
+    },
+    {
+      "epoch": 0.4148405110415433,
+      "grad_norm": 0.24011565744876862,
+      "learning_rate": 0.001,
+      "loss": 1.9115,
+      "step": 9806
+    },
+    {
+      "epoch": 0.41488281580505965,
+      "grad_norm": 3.5047898292541504,
+      "learning_rate": 0.001,
+      "loss": 2.0496,
+      "step": 9807
+    },
+    {
+      "epoch": 0.414925120568576,
+      "grad_norm": 3.9481778144836426,
+      "learning_rate": 0.001,
+      "loss": 3.1042,
+      "step": 9808
+    },
+    {
+      "epoch": 0.4149674253320924,
+      "grad_norm": 0.44774043560028076,
+      "learning_rate": 0.001,
+      "loss": 2.4119,
+      "step": 9809
+    },
+    {
+      "epoch": 0.41500973009560876,
+      "grad_norm": 0.9812766313552856,
+      "learning_rate": 0.001,
+      "loss": 2.6236,
+      "step": 9810
+    },
+    {
+      "epoch": 0.4150520348591251,
+      "grad_norm": 0.47764018177986145,
+      "learning_rate": 0.001,
+      "loss": 2.3742,
+      "step": 9811
+    },
+    {
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.18392595648765564,
+      "learning_rate": 0.001,
+      "loss": 2.4516,
+      "step": 9812
+    },
+    {
+      "epoch": 0.4151366443861579,
+      "grad_norm": 1.547833800315857,
+      "learning_rate": 0.001,
+      "loss": 1.7756,
+      "step": 9813
+    },
+    {
+      "epoch": 0.41517894914967424,
+      "grad_norm": 0.18931736052036285,
+      "learning_rate": 0.001,
+      "loss": 1.7584,
+      "step": 9814
+    },
+    {
+      "epoch": 0.41522125391319065,
+      "grad_norm": 0.1544259935617447,
+      "learning_rate": 0.001,
+      "loss": 1.4929,
+      "step": 9815
+    },
+    {
+      "epoch": 0.415263558676707,
+      "grad_norm": 0.2300853431224823,
+      "learning_rate": 0.001,
+      "loss": 2.7883,
+      "step": 9816
+    },
+    {
+      "epoch": 0.41530586344022336,
+      "grad_norm": 1.0780434608459473,
+      "learning_rate": 0.001,
+      "loss": 2.5039,
+      "step": 9817
+    },
+    {
+      "epoch": 0.41534816820373976,
+      "grad_norm": 0.5089141130447388,
+      "learning_rate": 0.001,
+      "loss": 1.7734,
+      "step": 9818
+    },
+    {
+      "epoch": 0.4153904729672561,
+      "grad_norm": 3.429790496826172,
+      "learning_rate": 0.001,
+      "loss": 2.3779,
+      "step": 9819
+    },
+    {
+      "epoch": 0.4154327777307725,
+      "grad_norm": 6.695871353149414,
+      "learning_rate": 0.001,
+      "loss": 2.3627,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4154750824942889,
+      "grad_norm": 0.1914941370487213,
+      "learning_rate": 0.001,
+      "loss": 2.655,
+      "step": 9821
+    },
+    {
+      "epoch": 0.41551738725780524,
+      "grad_norm": 0.6268720030784607,
+      "learning_rate": 0.001,
+      "loss": 2.0422,
+      "step": 9822
+    },
+    {
+      "epoch": 0.4155596920213216,
+      "grad_norm": 0.47744235396385193,
+      "learning_rate": 0.001,
+      "loss": 2.074,
+      "step": 9823
+    },
+    {
+      "epoch": 0.41560199678483795,
+      "grad_norm": 0.15372218191623688,
+      "learning_rate": 0.001,
+      "loss": 2.065,
+      "step": 9824
+    },
+    {
+      "epoch": 0.41564430154835436,
+      "grad_norm": 0.4260820746421814,
+      "learning_rate": 0.001,
+      "loss": 2.9168,
+      "step": 9825
+    },
+    {
+      "epoch": 0.4156866063118707,
+      "grad_norm": 0.19325514137744904,
+      "learning_rate": 0.001,
+      "loss": 1.8882,
+      "step": 9826
+    },
+    {
+      "epoch": 0.41572891107538706,
+      "grad_norm": 0.1592322140932083,
+      "learning_rate": 0.001,
+      "loss": 2.4256,
+      "step": 9827
+    },
+    {
+      "epoch": 0.4157712158389035,
+      "grad_norm": 0.17341256141662598,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 9828
+    },
+    {
+      "epoch": 0.41581352060241983,
+      "grad_norm": 0.20238442718982697,
+      "learning_rate": 0.001,
+      "loss": 2.2101,
+      "step": 9829
+    },
+    {
+      "epoch": 0.4158558253659362,
+      "grad_norm": 0.3207574784755707,
+      "learning_rate": 0.001,
+      "loss": 2.9256,
+      "step": 9830
+    },
+    {
+      "epoch": 0.4158981301294526,
+      "grad_norm": 0.1748972088098526,
+      "learning_rate": 0.001,
+      "loss": 1.8409,
+      "step": 9831
+    },
+    {
+      "epoch": 0.41594043489296895,
+      "grad_norm": 0.24550606310367584,
+      "learning_rate": 0.001,
+      "loss": 3.3066,
+      "step": 9832
+    },
+    {
+      "epoch": 0.4159827396564853,
+      "grad_norm": 0.9889479279518127,
+      "learning_rate": 0.001,
+      "loss": 2.1071,
+      "step": 9833
+    },
+    {
+      "epoch": 0.4160250444200017,
+      "grad_norm": 0.20585237443447113,
+      "learning_rate": 0.001,
+      "loss": 2.5545,
+      "step": 9834
+    },
+    {
+      "epoch": 0.41606734918351806,
+      "grad_norm": 0.13945859670639038,
+      "learning_rate": 0.001,
+      "loss": 1.4628,
+      "step": 9835
+    },
+    {
+      "epoch": 0.4161096539470344,
+      "grad_norm": 27.057586669921875,
+      "learning_rate": 0.001,
+      "loss": 2.8699,
+      "step": 9836
+    },
+    {
+      "epoch": 0.41615195871055083,
+      "grad_norm": 0.18365734815597534,
+      "learning_rate": 0.001,
+      "loss": 1.5084,
+      "step": 9837
+    },
+    {
+      "epoch": 0.4161942634740672,
+      "grad_norm": 0.43779316544532776,
+      "learning_rate": 0.001,
+      "loss": 2.1716,
+      "step": 9838
+    },
+    {
+      "epoch": 0.41623656823758354,
+      "grad_norm": 0.7131558060646057,
+      "learning_rate": 0.001,
+      "loss": 2.9858,
+      "step": 9839
+    },
+    {
+      "epoch": 0.41627887300109995,
+      "grad_norm": 3.8856091499328613,
+      "learning_rate": 0.001,
+      "loss": 3.9777,
+      "step": 9840
+    },
+    {
+      "epoch": 0.4163211777646163,
+      "grad_norm": 0.17805153131484985,
+      "learning_rate": 0.001,
+      "loss": 1.8416,
+      "step": 9841
+    },
+    {
+      "epoch": 0.41636348252813266,
+      "grad_norm": 0.5579624772071838,
+      "learning_rate": 0.001,
+      "loss": 3.2418,
+      "step": 9842
+    },
+    {
+      "epoch": 0.41640578729164907,
+      "grad_norm": 0.2010784149169922,
+      "learning_rate": 0.001,
+      "loss": 2.5104,
+      "step": 9843
+    },
+    {
+      "epoch": 0.4164480920551654,
+      "grad_norm": 0.1619592159986496,
+      "learning_rate": 0.001,
+      "loss": 1.9748,
+      "step": 9844
+    },
+    {
+      "epoch": 0.4164903968186818,
+      "grad_norm": 0.40515822172164917,
+      "learning_rate": 0.001,
+      "loss": 2.7918,
+      "step": 9845
+    },
+    {
+      "epoch": 0.41653270158219813,
+      "grad_norm": 0.22229692339897156,
+      "learning_rate": 0.001,
+      "loss": 2.0756,
+      "step": 9846
+    },
+    {
+      "epoch": 0.41657500634571454,
+      "grad_norm": 0.21411965787410736,
+      "learning_rate": 0.001,
+      "loss": 2.1632,
+      "step": 9847
+    },
+    {
+      "epoch": 0.4166173111092309,
+      "grad_norm": 0.21786244213581085,
+      "learning_rate": 0.001,
+      "loss": 3.3767,
+      "step": 9848
+    },
+    {
+      "epoch": 0.41665961587274725,
+      "grad_norm": 0.20544399321079254,
+      "learning_rate": 0.001,
+      "loss": 2.1629,
+      "step": 9849
+    },
+    {
+      "epoch": 0.41670192063626366,
+      "grad_norm": 0.17390744388103485,
+      "learning_rate": 0.001,
+      "loss": 2.2281,
+      "step": 9850
+    },
+    {
+      "epoch": 0.41674422539978,
+      "grad_norm": 0.5404853820800781,
+      "learning_rate": 0.001,
+      "loss": 2.8778,
+      "step": 9851
+    },
+    {
+      "epoch": 0.41678653016329636,
+      "grad_norm": 0.18228615820407867,
+      "learning_rate": 0.001,
+      "loss": 2.2249,
+      "step": 9852
+    },
+    {
+      "epoch": 0.4168288349268128,
+      "grad_norm": 1.0131566524505615,
+      "learning_rate": 0.001,
+      "loss": 2.8682,
+      "step": 9853
+    },
+    {
+      "epoch": 0.41687113969032913,
+      "grad_norm": 0.1611110270023346,
+      "learning_rate": 0.001,
+      "loss": 2.1683,
+      "step": 9854
+    },
+    {
+      "epoch": 0.4169134444538455,
+      "grad_norm": 0.7783699631690979,
+      "learning_rate": 0.001,
+      "loss": 2.6472,
+      "step": 9855
+    },
+    {
+      "epoch": 0.4169557492173619,
+      "grad_norm": 0.15802177786827087,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 9856
+    },
+    {
+      "epoch": 0.41699805398087825,
+      "grad_norm": 0.2226634919643402,
+      "learning_rate": 0.001,
+      "loss": 1.9478,
+      "step": 9857
+    },
+    {
+      "epoch": 0.4170403587443946,
+      "grad_norm": 0.4912348687648773,
+      "learning_rate": 0.001,
+      "loss": 1.7005,
+      "step": 9858
+    },
+    {
+      "epoch": 0.417082663507911,
+      "grad_norm": 0.1934366077184677,
+      "learning_rate": 0.001,
+      "loss": 2.5896,
+      "step": 9859
+    },
+    {
+      "epoch": 0.41712496827142737,
+      "grad_norm": 0.20569314062595367,
+      "learning_rate": 0.001,
+      "loss": 2.9554,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4171672730349437,
+      "grad_norm": 0.18698929250240326,
+      "learning_rate": 0.001,
+      "loss": 2.1069,
+      "step": 9861
+    },
+    {
+      "epoch": 0.41720957779846013,
+      "grad_norm": 0.2676628530025482,
+      "learning_rate": 0.001,
+      "loss": 2.5049,
+      "step": 9862
+    },
+    {
+      "epoch": 0.4172518825619765,
+      "grad_norm": 0.21008531749248505,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 9863
+    },
+    {
+      "epoch": 0.41729418732549284,
+      "grad_norm": 0.19284501671791077,
+      "learning_rate": 0.001,
+      "loss": 3.301,
+      "step": 9864
+    },
+    {
+      "epoch": 0.41733649208900925,
+      "grad_norm": 0.19138003885746002,
+      "learning_rate": 0.001,
+      "loss": 2.2343,
+      "step": 9865
+    },
+    {
+      "epoch": 0.4173787968525256,
+      "grad_norm": 3.3157238960266113,
+      "learning_rate": 0.001,
+      "loss": 1.9117,
+      "step": 9866
+    },
+    {
+      "epoch": 0.41742110161604196,
+      "grad_norm": 0.1964617371559143,
+      "learning_rate": 0.001,
+      "loss": 2.8301,
+      "step": 9867
+    },
+    {
+      "epoch": 0.4174634063795583,
+      "grad_norm": 0.27521830797195435,
+      "learning_rate": 0.001,
+      "loss": 1.853,
+      "step": 9868
+    },
+    {
+      "epoch": 0.4175057111430747,
+      "grad_norm": 0.26007047295570374,
+      "learning_rate": 0.001,
+      "loss": 2.6994,
+      "step": 9869
+    },
+    {
+      "epoch": 0.4175480159065911,
+      "grad_norm": 0.2345419079065323,
+      "learning_rate": 0.001,
+      "loss": 3.8451,
+      "step": 9870
+    },
+    {
+      "epoch": 0.41759032067010743,
+      "grad_norm": 0.5388814210891724,
+      "learning_rate": 0.001,
+      "loss": 2.7057,
+      "step": 9871
+    },
+    {
+      "epoch": 0.41763262543362384,
+      "grad_norm": 0.1674017757177353,
+      "learning_rate": 0.001,
+      "loss": 1.9514,
+      "step": 9872
+    },
+    {
+      "epoch": 0.4176749301971402,
+      "grad_norm": 0.2036234587430954,
+      "learning_rate": 0.001,
+      "loss": 2.4295,
+      "step": 9873
+    },
+    {
+      "epoch": 0.41771723496065655,
+      "grad_norm": 5.006783485412598,
+      "learning_rate": 0.001,
+      "loss": 2.2285,
+      "step": 9874
+    },
+    {
+      "epoch": 0.41775953972417296,
+      "grad_norm": 0.3043411374092102,
+      "learning_rate": 0.001,
+      "loss": 2.1786,
+      "step": 9875
+    },
+    {
+      "epoch": 0.4178018444876893,
+      "grad_norm": 0.18863193690776825,
+      "learning_rate": 0.001,
+      "loss": 2.0902,
+      "step": 9876
+    },
+    {
+      "epoch": 0.41784414925120567,
+      "grad_norm": 0.21312756836414337,
+      "learning_rate": 0.001,
+      "loss": 2.418,
+      "step": 9877
+    },
+    {
+      "epoch": 0.4178864540147221,
+      "grad_norm": 0.25817111134529114,
+      "learning_rate": 0.001,
+      "loss": 3.3019,
+      "step": 9878
+    },
+    {
+      "epoch": 0.41792875877823843,
+      "grad_norm": 0.17878010869026184,
+      "learning_rate": 0.001,
+      "loss": 2.1354,
+      "step": 9879
+    },
+    {
+      "epoch": 0.4179710635417548,
+      "grad_norm": 0.2418394237756729,
+      "learning_rate": 0.001,
+      "loss": 2.7489,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4180133683052712,
+      "grad_norm": 5.557830333709717,
+      "learning_rate": 0.001,
+      "loss": 1.8045,
+      "step": 9881
+    },
+    {
+      "epoch": 0.41805567306878755,
+      "grad_norm": 0.17561772465705872,
+      "learning_rate": 0.001,
+      "loss": 2.6087,
+      "step": 9882
+    },
+    {
+      "epoch": 0.4180979778323039,
+      "grad_norm": 0.1871245950460434,
+      "learning_rate": 0.001,
+      "loss": 2.7775,
+      "step": 9883
+    },
+    {
+      "epoch": 0.4181402825958203,
+      "grad_norm": 0.17754124104976654,
+      "learning_rate": 0.001,
+      "loss": 1.7533,
+      "step": 9884
+    },
+    {
+      "epoch": 0.41818258735933667,
+      "grad_norm": 0.406012624502182,
+      "learning_rate": 0.001,
+      "loss": 2.6556,
+      "step": 9885
+    },
+    {
+      "epoch": 0.418224892122853,
+      "grad_norm": 0.2209833264350891,
+      "learning_rate": 0.001,
+      "loss": 2.6001,
+      "step": 9886
+    },
+    {
+      "epoch": 0.41826719688636943,
+      "grad_norm": 5.273458480834961,
+      "learning_rate": 0.001,
+      "loss": 2.443,
+      "step": 9887
+    },
+    {
+      "epoch": 0.4183095016498858,
+      "grad_norm": 0.18473027646541595,
+      "learning_rate": 0.001,
+      "loss": 1.6151,
+      "step": 9888
+    },
+    {
+      "epoch": 0.41835180641340214,
+      "grad_norm": 3.541088342666626,
+      "learning_rate": 0.001,
+      "loss": 2.5512,
+      "step": 9889
+    },
+    {
+      "epoch": 0.41839411117691855,
+      "grad_norm": 0.17209482192993164,
+      "learning_rate": 0.001,
+      "loss": 2.3301,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4184364159404349,
+      "grad_norm": 0.2227262258529663,
+      "learning_rate": 0.001,
+      "loss": 2.3766,
+      "step": 9891
+    },
+    {
+      "epoch": 0.41847872070395126,
+      "grad_norm": 0.17431975901126862,
+      "learning_rate": 0.001,
+      "loss": 1.6998,
+      "step": 9892
+    },
+    {
+      "epoch": 0.4185210254674676,
+      "grad_norm": 0.3029531240463257,
+      "learning_rate": 0.001,
+      "loss": 2.4269,
+      "step": 9893
+    },
+    {
+      "epoch": 0.418563330230984,
+      "grad_norm": 0.22647684812545776,
+      "learning_rate": 0.001,
+      "loss": 1.8381,
+      "step": 9894
+    },
+    {
+      "epoch": 0.4186056349945004,
+      "grad_norm": 0.21035243570804596,
+      "learning_rate": 0.001,
+      "loss": 3.3303,
+      "step": 9895
+    },
+    {
+      "epoch": 0.41864793975801673,
+      "grad_norm": 1.0283381938934326,
+      "learning_rate": 0.001,
+      "loss": 2.258,
+      "step": 9896
+    },
+    {
+      "epoch": 0.41869024452153314,
+      "grad_norm": 1.2793562412261963,
+      "learning_rate": 0.001,
+      "loss": 3.1089,
+      "step": 9897
+    },
+    {
+      "epoch": 0.4187325492850495,
+      "grad_norm": 0.21786046028137207,
+      "learning_rate": 0.001,
+      "loss": 1.7205,
+      "step": 9898
+    },
+    {
+      "epoch": 0.41877485404856585,
+      "grad_norm": 0.16256938874721527,
+      "learning_rate": 0.001,
+      "loss": 2.8039,
+      "step": 9899
+    },
+    {
+      "epoch": 0.41881715881208226,
+      "grad_norm": 0.4168480634689331,
+      "learning_rate": 0.001,
+      "loss": 2.3668,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4188594635755986,
+      "grad_norm": 0.1868831068277359,
+      "learning_rate": 0.001,
+      "loss": 2.6872,
+      "step": 9901
+    },
+    {
+      "epoch": 0.41890176833911497,
+      "grad_norm": 2.076476573944092,
+      "learning_rate": 0.001,
+      "loss": 2.4575,
+      "step": 9902
+    },
+    {
+      "epoch": 0.4189440731026314,
+      "grad_norm": 0.22399549186229706,
+      "learning_rate": 0.001,
+      "loss": 2.7196,
+      "step": 9903
+    },
+    {
+      "epoch": 0.41898637786614773,
+      "grad_norm": 0.24914605915546417,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 9904
+    },
+    {
+      "epoch": 0.4190286826296641,
+      "grad_norm": 0.265962690114975,
+      "learning_rate": 0.001,
+      "loss": 1.5826,
+      "step": 9905
+    },
+    {
+      "epoch": 0.4190709873931805,
+      "grad_norm": 0.20687235891819,
+      "learning_rate": 0.001,
+      "loss": 2.7606,
+      "step": 9906
+    },
+    {
+      "epoch": 0.41911329215669685,
+      "grad_norm": 0.18030163645744324,
+      "learning_rate": 0.001,
+      "loss": 1.7922,
+      "step": 9907
+    },
+    {
+      "epoch": 0.4191555969202132,
+      "grad_norm": 0.17712494730949402,
+      "learning_rate": 0.001,
+      "loss": 1.9416,
+      "step": 9908
+    },
+    {
+      "epoch": 0.4191979016837296,
+      "grad_norm": 0.2647468149662018,
+      "learning_rate": 0.001,
+      "loss": 1.6112,
+      "step": 9909
+    },
+    {
+      "epoch": 0.41924020644724597,
+      "grad_norm": 10.648910522460938,
+      "learning_rate": 0.001,
+      "loss": 2.2037,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4192825112107623,
+      "grad_norm": 0.22634510695934296,
+      "learning_rate": 0.001,
+      "loss": 1.8658,
+      "step": 9911
+    },
+    {
+      "epoch": 0.41932481597427873,
+      "grad_norm": 0.19005584716796875,
+      "learning_rate": 0.001,
+      "loss": 1.784,
+      "step": 9912
+    },
+    {
+      "epoch": 0.4193671207377951,
+      "grad_norm": 11.0584077835083,
+      "learning_rate": 0.001,
+      "loss": 2.403,
+      "step": 9913
+    },
+    {
+      "epoch": 0.41940942550131144,
+      "grad_norm": 0.18908476829528809,
+      "learning_rate": 0.001,
+      "loss": 2.2306,
+      "step": 9914
+    },
+    {
+      "epoch": 0.4194517302648278,
+      "grad_norm": 0.196416437625885,
+      "learning_rate": 0.001,
+      "loss": 2.8057,
+      "step": 9915
+    },
+    {
+      "epoch": 0.4194940350283442,
+      "grad_norm": 0.37092915177345276,
+      "learning_rate": 0.001,
+      "loss": 3.7798,
+      "step": 9916
+    },
+    {
+      "epoch": 0.41953633979186056,
+      "grad_norm": 0.1975582391023636,
+      "learning_rate": 0.001,
+      "loss": 2.7689,
+      "step": 9917
+    },
+    {
+      "epoch": 0.4195786445553769,
+      "grad_norm": 16.982938766479492,
+      "learning_rate": 0.001,
+      "loss": 3.2579,
+      "step": 9918
+    },
+    {
+      "epoch": 0.4196209493188933,
+      "grad_norm": 0.22323065996170044,
+      "learning_rate": 0.001,
+      "loss": 2.1371,
+      "step": 9919
+    },
+    {
+      "epoch": 0.4196632540824097,
+      "grad_norm": 0.3615620732307434,
+      "learning_rate": 0.001,
+      "loss": 1.9409,
+      "step": 9920
+    },
+    {
+      "epoch": 0.41970555884592603,
+      "grad_norm": 0.7461709976196289,
+      "learning_rate": 0.001,
+      "loss": 2.4982,
+      "step": 9921
+    },
+    {
+      "epoch": 0.41974786360944244,
+      "grad_norm": 0.20795656740665436,
+      "learning_rate": 0.001,
+      "loss": 2.3513,
+      "step": 9922
+    },
+    {
+      "epoch": 0.4197901683729588,
+      "grad_norm": 0.19558289647102356,
+      "learning_rate": 0.001,
+      "loss": 2.7092,
+      "step": 9923
+    },
+    {
+      "epoch": 0.41983247313647515,
+      "grad_norm": 0.2396133542060852,
+      "learning_rate": 0.001,
+      "loss": 2.6774,
+      "step": 9924
+    },
+    {
+      "epoch": 0.41987477789999156,
+      "grad_norm": 0.2055635303258896,
+      "learning_rate": 0.001,
+      "loss": 3.1715,
+      "step": 9925
+    },
+    {
+      "epoch": 0.4199170826635079,
+      "grad_norm": 0.28740987181663513,
+      "learning_rate": 0.001,
+      "loss": 2.0135,
+      "step": 9926
+    },
+    {
+      "epoch": 0.41995938742702427,
+      "grad_norm": 3.5585386753082275,
+      "learning_rate": 0.001,
+      "loss": 2.0134,
+      "step": 9927
+    },
+    {
+      "epoch": 0.4200016921905407,
+      "grad_norm": 0.19014716148376465,
+      "learning_rate": 0.001,
+      "loss": 3.6427,
+      "step": 9928
+    },
+    {
+      "epoch": 0.42004399695405703,
+      "grad_norm": 0.32993394136428833,
+      "learning_rate": 0.001,
+      "loss": 1.8829,
+      "step": 9929
+    },
+    {
+      "epoch": 0.4200863017175734,
+      "grad_norm": 0.22422336041927338,
+      "learning_rate": 0.001,
+      "loss": 2.4582,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4201286064810898,
+      "grad_norm": 1.323824167251587,
+      "learning_rate": 0.001,
+      "loss": 3.1731,
+      "step": 9931
+    },
+    {
+      "epoch": 0.42017091124460615,
+      "grad_norm": 0.17324510216712952,
+      "learning_rate": 0.001,
+      "loss": 1.8609,
+      "step": 9932
+    },
+    {
+      "epoch": 0.4202132160081225,
+      "grad_norm": 1.34519624710083,
+      "learning_rate": 0.001,
+      "loss": 2.1091,
+      "step": 9933
+    },
+    {
+      "epoch": 0.4202555207716389,
+      "grad_norm": 2.9261980056762695,
+      "learning_rate": 0.001,
+      "loss": 2.0931,
+      "step": 9934
+    },
+    {
+      "epoch": 0.42029782553515527,
+      "grad_norm": 0.26872578263282776,
+      "learning_rate": 0.001,
+      "loss": 2.4059,
+      "step": 9935
+    },
+    {
+      "epoch": 0.4203401302986716,
+      "grad_norm": 10.175525665283203,
+      "learning_rate": 0.001,
+      "loss": 2.2073,
+      "step": 9936
+    },
+    {
+      "epoch": 0.420382435062188,
+      "grad_norm": 0.20974968373775482,
+      "learning_rate": 0.001,
+      "loss": 2.474,
+      "step": 9937
+    },
+    {
+      "epoch": 0.4204247398257044,
+      "grad_norm": 2.2708218097686768,
+      "learning_rate": 0.001,
+      "loss": 2.441,
+      "step": 9938
+    },
+    {
+      "epoch": 0.42046704458922074,
+      "grad_norm": 0.1942100077867508,
+      "learning_rate": 0.001,
+      "loss": 2.4447,
+      "step": 9939
+    },
+    {
+      "epoch": 0.4205093493527371,
+      "grad_norm": 0.24804472923278809,
+      "learning_rate": 0.001,
+      "loss": 2.365,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4205516541162535,
+      "grad_norm": 0.21028558909893036,
+      "learning_rate": 0.001,
+      "loss": 1.9267,
+      "step": 9941
+    },
+    {
+      "epoch": 0.42059395887976986,
+      "grad_norm": 0.25556495785713196,
+      "learning_rate": 0.001,
+      "loss": 2.5246,
+      "step": 9942
+    },
+    {
+      "epoch": 0.4206362636432862,
+      "grad_norm": 0.6381620764732361,
+      "learning_rate": 0.001,
+      "loss": 2.455,
+      "step": 9943
+    },
+    {
+      "epoch": 0.4206785684068026,
+      "grad_norm": 0.20870067179203033,
+      "learning_rate": 0.001,
+      "loss": 1.8444,
+      "step": 9944
+    },
+    {
+      "epoch": 0.420720873170319,
+      "grad_norm": 0.4335130751132965,
+      "learning_rate": 0.001,
+      "loss": 1.7678,
+      "step": 9945
+    },
+    {
+      "epoch": 0.42076317793383533,
+      "grad_norm": 0.20234449207782745,
+      "learning_rate": 0.001,
+      "loss": 1.6857,
+      "step": 9946
+    },
+    {
+      "epoch": 0.42080548269735174,
+      "grad_norm": 0.21633018553256989,
+      "learning_rate": 0.001,
+      "loss": 2.6976,
+      "step": 9947
+    },
+    {
+      "epoch": 0.4208477874608681,
+      "grad_norm": 0.20859219133853912,
+      "learning_rate": 0.001,
+      "loss": 1.9807,
+      "step": 9948
+    },
+    {
+      "epoch": 0.42089009222438445,
+      "grad_norm": 0.19663840532302856,
+      "learning_rate": 0.001,
+      "loss": 2.1092,
+      "step": 9949
+    },
+    {
+      "epoch": 0.42093239698790086,
+      "grad_norm": 0.24110959470272064,
+      "learning_rate": 0.001,
+      "loss": 2.0922,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4209747017514172,
+      "grad_norm": 0.18923258781433105,
+      "learning_rate": 0.001,
+      "loss": 1.8416,
+      "step": 9951
+    },
+    {
+      "epoch": 0.42101700651493357,
+      "grad_norm": 0.16037270426750183,
+      "learning_rate": 0.001,
+      "loss": 2.9326,
+      "step": 9952
+    },
+    {
+      "epoch": 0.42105931127845,
+      "grad_norm": 0.15264320373535156,
+      "learning_rate": 0.001,
+      "loss": 1.6577,
+      "step": 9953
+    },
+    {
+      "epoch": 0.42110161604196633,
+      "grad_norm": 4.530941486358643,
+      "learning_rate": 0.001,
+      "loss": 2.5381,
+      "step": 9954
+    },
+    {
+      "epoch": 0.4211439208054827,
+      "grad_norm": 0.17232206463813782,
+      "learning_rate": 0.001,
+      "loss": 3.0177,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4211862255689991,
+      "grad_norm": 0.18628808856010437,
+      "learning_rate": 0.001,
+      "loss": 2.0065,
+      "step": 9956
+    },
+    {
+      "epoch": 0.42122853033251545,
+      "grad_norm": 0.1941269338130951,
+      "learning_rate": 0.001,
+      "loss": 3.6867,
+      "step": 9957
+    },
+    {
+      "epoch": 0.4212708350960318,
+      "grad_norm": 0.18407343327999115,
+      "learning_rate": 0.001,
+      "loss": 2.8404,
+      "step": 9958
+    },
+    {
+      "epoch": 0.42131313985954816,
+      "grad_norm": 0.27422475814819336,
+      "learning_rate": 0.001,
+      "loss": 1.9193,
+      "step": 9959
+    },
+    {
+      "epoch": 0.42135544462306457,
+      "grad_norm": 0.3834380805492401,
+      "learning_rate": 0.001,
+      "loss": 1.587,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4213977493865809,
+      "grad_norm": 0.28580230474472046,
+      "learning_rate": 0.001,
+      "loss": 2.604,
+      "step": 9961
+    },
+    {
+      "epoch": 0.4214400541500973,
+      "grad_norm": 0.7395582795143127,
+      "learning_rate": 0.001,
+      "loss": 2.5038,
+      "step": 9962
+    },
+    {
+      "epoch": 0.4214823589136137,
+      "grad_norm": 0.19765476882457733,
+      "learning_rate": 0.001,
+      "loss": 2.0077,
+      "step": 9963
+    },
+    {
+      "epoch": 0.42152466367713004,
+      "grad_norm": 0.2048027217388153,
+      "learning_rate": 0.001,
+      "loss": 2.1928,
+      "step": 9964
+    },
+    {
+      "epoch": 0.4215669684406464,
+      "grad_norm": 0.1922098696231842,
+      "learning_rate": 0.001,
+      "loss": 1.5701,
+      "step": 9965
+    },
+    {
+      "epoch": 0.4216092732041628,
+      "grad_norm": 0.4606640040874481,
+      "learning_rate": 0.001,
+      "loss": 1.8854,
+      "step": 9966
+    },
+    {
+      "epoch": 0.42165157796767916,
+      "grad_norm": 0.17350491881370544,
+      "learning_rate": 0.001,
+      "loss": 3.4361,
+      "step": 9967
+    },
+    {
+      "epoch": 0.4216938827311955,
+      "grad_norm": 0.17392922937870026,
+      "learning_rate": 0.001,
+      "loss": 1.8771,
+      "step": 9968
+    },
+    {
+      "epoch": 0.4217361874947119,
+      "grad_norm": 0.25780975818634033,
+      "learning_rate": 0.001,
+      "loss": 2.3047,
+      "step": 9969
+    },
+    {
+      "epoch": 0.4217784922582283,
+      "grad_norm": 0.22652190923690796,
+      "learning_rate": 0.001,
+      "loss": 3.8245,
+      "step": 9970
+    },
+    {
+      "epoch": 0.42182079702174463,
+      "grad_norm": 0.1851012408733368,
+      "learning_rate": 0.001,
+      "loss": 2.5209,
+      "step": 9971
+    },
+    {
+      "epoch": 0.42186310178526104,
+      "grad_norm": 0.6701632738113403,
+      "learning_rate": 0.001,
+      "loss": 2.6308,
+      "step": 9972
+    },
+    {
+      "epoch": 0.4219054065487774,
+      "grad_norm": 0.19841192662715912,
+      "learning_rate": 0.001,
+      "loss": 3.4615,
+      "step": 9973
+    },
+    {
+      "epoch": 0.42194771131229375,
+      "grad_norm": 0.18321669101715088,
+      "learning_rate": 0.001,
+      "loss": 2.0243,
+      "step": 9974
+    },
+    {
+      "epoch": 0.42199001607581016,
+      "grad_norm": 0.1863401234149933,
+      "learning_rate": 0.001,
+      "loss": 2.2838,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4220323208393265,
+      "grad_norm": 0.5115291476249695,
+      "learning_rate": 0.001,
+      "loss": 2.6378,
+      "step": 9976
+    },
+    {
+      "epoch": 0.42207462560284287,
+      "grad_norm": 0.20143838226795197,
+      "learning_rate": 0.001,
+      "loss": 2.8185,
+      "step": 9977
+    },
+    {
+      "epoch": 0.4221169303663593,
+      "grad_norm": 0.7708411812782288,
+      "learning_rate": 0.001,
+      "loss": 2.9274,
+      "step": 9978
+    },
+    {
+      "epoch": 0.42215923512987563,
+      "grad_norm": 0.18255001306533813,
+      "learning_rate": 0.001,
+      "loss": 2.3446,
+      "step": 9979
+    },
+    {
+      "epoch": 0.422201539893392,
+      "grad_norm": 2.8448588848114014,
+      "learning_rate": 0.001,
+      "loss": 2.0094,
+      "step": 9980
+    },
+    {
+      "epoch": 0.42224384465690834,
+      "grad_norm": 0.2241508811712265,
+      "learning_rate": 0.001,
+      "loss": 2.4382,
+      "step": 9981
+    },
+    {
+      "epoch": 0.42228614942042475,
+      "grad_norm": 4.693768501281738,
+      "learning_rate": 0.001,
+      "loss": 2.416,
+      "step": 9982
+    },
+    {
+      "epoch": 0.4223284541839411,
+      "grad_norm": 0.18450792133808136,
+      "learning_rate": 0.001,
+      "loss": 1.5582,
+      "step": 9983
+    },
+    {
+      "epoch": 0.42237075894745746,
+      "grad_norm": 0.21260473132133484,
+      "learning_rate": 0.001,
+      "loss": 1.7709,
+      "step": 9984
+    },
+    {
+      "epoch": 0.42241306371097387,
+      "grad_norm": 0.23331928253173828,
+      "learning_rate": 0.001,
+      "loss": 2.4449,
+      "step": 9985
+    },
+    {
+      "epoch": 0.4224553684744902,
+      "grad_norm": 0.38122308254241943,
+      "learning_rate": 0.001,
+      "loss": 2.4113,
+      "step": 9986
+    },
+    {
+      "epoch": 0.4224976732380066,
+      "grad_norm": 0.16446325182914734,
+      "learning_rate": 0.001,
+      "loss": 2.0169,
+      "step": 9987
+    },
+    {
+      "epoch": 0.422539978001523,
+      "grad_norm": 0.1864539235830307,
+      "learning_rate": 0.001,
+      "loss": 2.2757,
+      "step": 9988
+    },
+    {
+      "epoch": 0.42258228276503934,
+      "grad_norm": 0.208041250705719,
+      "learning_rate": 0.001,
+      "loss": 1.9249,
+      "step": 9989
+    },
+    {
+      "epoch": 0.4226245875285557,
+      "grad_norm": 0.6653062105178833,
+      "learning_rate": 0.001,
+      "loss": 2.7106,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4226668922920721,
+      "grad_norm": 0.2309500128030777,
+      "learning_rate": 0.001,
+      "loss": 2.502,
+      "step": 9991
+    },
+    {
+      "epoch": 0.42270919705558846,
+      "grad_norm": 8.24671745300293,
+      "learning_rate": 0.001,
+      "loss": 1.6391,
+      "step": 9992
+    },
+    {
+      "epoch": 0.4227515018191048,
+      "grad_norm": 0.389308363199234,
+      "learning_rate": 0.001,
+      "loss": 1.9749,
+      "step": 9993
+    },
+    {
+      "epoch": 0.4227938065826212,
+      "grad_norm": 0.2275291532278061,
+      "learning_rate": 0.001,
+      "loss": 3.4856,
+      "step": 9994
+    },
+    {
+      "epoch": 0.4228361113461376,
+      "grad_norm": 0.8399461507797241,
+      "learning_rate": 0.001,
+      "loss": 3.256,
+      "step": 9995
+    },
+    {
+      "epoch": 0.42287841610965393,
+      "grad_norm": 0.9205453395843506,
+      "learning_rate": 0.001,
+      "loss": 2.8834,
+      "step": 9996
+    },
+    {
+      "epoch": 0.42292072087317034,
+      "grad_norm": 20.110586166381836,
+      "learning_rate": 0.001,
+      "loss": 3.0527,
+      "step": 9997
+    },
+    {
+      "epoch": 0.4229630256366867,
+      "grad_norm": 0.2495739907026291,
+      "learning_rate": 0.001,
+      "loss": 2.4493,
+      "step": 9998
+    },
+    {
+      "epoch": 0.42300533040020305,
+      "grad_norm": 16.760589599609375,
+      "learning_rate": 0.001,
+      "loss": 3.6418,
+      "step": 9999
+    },
+    {
+      "epoch": 0.42304763516371946,
+      "grad_norm": 0.20367243885993958,
+      "learning_rate": 0.001,
+      "loss": 2.063,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4230899399272358,
+      "grad_norm": 0.2168806493282318,
+      "learning_rate": 0.001,
+      "loss": 2.816,
+      "step": 10001
+    },
+    {
+      "epoch": 0.42313224469075217,
+      "grad_norm": 1.293717384338379,
+      "learning_rate": 0.001,
+      "loss": 2.3445,
+      "step": 10002
+    },
+    {
+      "epoch": 0.4231745494542686,
+      "grad_norm": 139.1713409423828,
+      "learning_rate": 0.001,
+      "loss": 3.2346,
+      "step": 10003
+    },
+    {
+      "epoch": 0.42321685421778493,
+      "grad_norm": 0.39124611020088196,
+      "learning_rate": 0.001,
+      "loss": 2.108,
+      "step": 10004
+    },
+    {
+      "epoch": 0.4232591589813013,
+      "grad_norm": 0.9725698828697205,
+      "learning_rate": 0.001,
+      "loss": 2.4943,
+      "step": 10005
+    },
+    {
+      "epoch": 0.42330146374481764,
+      "grad_norm": 2.8331100940704346,
+      "learning_rate": 0.001,
+      "loss": 2.6655,
+      "step": 10006
+    },
+    {
+      "epoch": 0.42334376850833405,
+      "grad_norm": 0.20734193921089172,
+      "learning_rate": 0.001,
+      "loss": 2.3108,
+      "step": 10007
+    },
+    {
+      "epoch": 0.4233860732718504,
+      "grad_norm": 1.9373948574066162,
+      "learning_rate": 0.001,
+      "loss": 1.867,
+      "step": 10008
+    },
+    {
+      "epoch": 0.42342837803536676,
+      "grad_norm": 1.3012683391571045,
+      "learning_rate": 0.001,
+      "loss": 2.6328,
+      "step": 10009
+    },
+    {
+      "epoch": 0.42347068279888317,
+      "grad_norm": 0.838748037815094,
+      "learning_rate": 0.001,
+      "loss": 2.1099,
+      "step": 10010
+    },
+    {
+      "epoch": 0.4235129875623995,
+      "grad_norm": 0.27665624022483826,
+      "learning_rate": 0.001,
+      "loss": 2.0056,
+      "step": 10011
+    },
+    {
+      "epoch": 0.4235552923259159,
+      "grad_norm": 0.2420514076948166,
+      "learning_rate": 0.001,
+      "loss": 1.8592,
+      "step": 10012
+    },
+    {
+      "epoch": 0.4235975970894323,
+      "grad_norm": 38.479976654052734,
+      "learning_rate": 0.001,
+      "loss": 1.7052,
+      "step": 10013
+    },
+    {
+      "epoch": 0.42363990185294864,
+      "grad_norm": 0.297911673784256,
+      "learning_rate": 0.001,
+      "loss": 2.694,
+      "step": 10014
+    },
+    {
+      "epoch": 0.423682206616465,
+      "grad_norm": 0.2878136932849884,
+      "learning_rate": 0.001,
+      "loss": 2.3265,
+      "step": 10015
+    },
+    {
+      "epoch": 0.4237245113799814,
+      "grad_norm": 0.3122667968273163,
+      "learning_rate": 0.001,
+      "loss": 3.1903,
+      "step": 10016
+    },
+    {
+      "epoch": 0.42376681614349776,
+      "grad_norm": 6.599530220031738,
+      "learning_rate": 0.001,
+      "loss": 3.0154,
+      "step": 10017
+    },
+    {
+      "epoch": 0.4238091209070141,
+      "grad_norm": 0.2963060438632965,
+      "learning_rate": 0.001,
+      "loss": 2.9263,
+      "step": 10018
+    },
+    {
+      "epoch": 0.4238514256705305,
+      "grad_norm": 0.3066433370113373,
+      "learning_rate": 0.001,
+      "loss": 2.8319,
+      "step": 10019
+    },
+    {
+      "epoch": 0.4238937304340469,
+      "grad_norm": 0.5115453004837036,
+      "learning_rate": 0.001,
+      "loss": 2.9406,
+      "step": 10020
+    },
+    {
+      "epoch": 0.42393603519756323,
+      "grad_norm": 0.24337489902973175,
+      "learning_rate": 0.001,
+      "loss": 2.2676,
+      "step": 10021
+    },
+    {
+      "epoch": 0.42397833996107964,
+      "grad_norm": 0.20824651420116425,
+      "learning_rate": 0.001,
+      "loss": 3.0662,
+      "step": 10022
+    },
+    {
+      "epoch": 0.424020644724596,
+      "grad_norm": 0.22388121485710144,
+      "learning_rate": 0.001,
+      "loss": 2.5818,
+      "step": 10023
+    },
+    {
+      "epoch": 0.42406294948811235,
+      "grad_norm": 0.29800713062286377,
+      "learning_rate": 0.001,
+      "loss": 3.4957,
+      "step": 10024
+    },
+    {
+      "epoch": 0.42410525425162876,
+      "grad_norm": 0.23895491659641266,
+      "learning_rate": 0.001,
+      "loss": 2.0719,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4241475590151451,
+      "grad_norm": 0.21432499587535858,
+      "learning_rate": 0.001,
+      "loss": 2.6525,
+      "step": 10026
+    },
+    {
+      "epoch": 0.42418986377866147,
+      "grad_norm": 0.22099368274211884,
+      "learning_rate": 0.001,
+      "loss": 2.1917,
+      "step": 10027
+    },
+    {
+      "epoch": 0.4242321685421778,
+      "grad_norm": 0.18679919838905334,
+      "learning_rate": 0.001,
+      "loss": 2.411,
+      "step": 10028
+    },
+    {
+      "epoch": 0.42427447330569423,
+      "grad_norm": 3.992774486541748,
+      "learning_rate": 0.001,
+      "loss": 2.1948,
+      "step": 10029
+    },
+    {
+      "epoch": 0.4243167780692106,
+      "grad_norm": 0.29614201188087463,
+      "learning_rate": 0.001,
+      "loss": 2.3378,
+      "step": 10030
+    },
+    {
+      "epoch": 0.42435908283272694,
+      "grad_norm": 0.2120267003774643,
+      "learning_rate": 0.001,
+      "loss": 1.9312,
+      "step": 10031
+    },
+    {
+      "epoch": 0.42440138759624335,
+      "grad_norm": 0.39450037479400635,
+      "learning_rate": 0.001,
+      "loss": 2.7093,
+      "step": 10032
+    },
+    {
+      "epoch": 0.4244436923597597,
+      "grad_norm": 0.28339096903800964,
+      "learning_rate": 0.001,
+      "loss": 2.5644,
+      "step": 10033
+    },
+    {
+      "epoch": 0.42448599712327606,
+      "grad_norm": 0.24640879034996033,
+      "learning_rate": 0.001,
+      "loss": 2.3985,
+      "step": 10034
+    },
+    {
+      "epoch": 0.42452830188679247,
+      "grad_norm": 4.143669128417969,
+      "learning_rate": 0.001,
+      "loss": 3.2367,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4245706066503088,
+      "grad_norm": 0.21093952655792236,
+      "learning_rate": 0.001,
+      "loss": 2.113,
+      "step": 10036
+    },
+    {
+      "epoch": 0.4246129114138252,
+      "grad_norm": 0.17772096395492554,
+      "learning_rate": 0.001,
+      "loss": 1.6009,
+      "step": 10037
+    },
+    {
+      "epoch": 0.4246552161773416,
+      "grad_norm": 2.9753193855285645,
+      "learning_rate": 0.001,
+      "loss": 2.4802,
+      "step": 10038
+    },
+    {
+      "epoch": 0.42469752094085794,
+      "grad_norm": 0.23098500072956085,
+      "learning_rate": 0.001,
+      "loss": 2.2026,
+      "step": 10039
+    },
+    {
+      "epoch": 0.4247398257043743,
+      "grad_norm": 4.376840114593506,
+      "learning_rate": 0.001,
+      "loss": 2.04,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4247821304678907,
+      "grad_norm": 1.3142906427383423,
+      "learning_rate": 0.001,
+      "loss": 2.0858,
+      "step": 10041
+    },
+    {
+      "epoch": 0.42482443523140706,
+      "grad_norm": 0.25867971777915955,
+      "learning_rate": 0.001,
+      "loss": 2.6186,
+      "step": 10042
+    },
+    {
+      "epoch": 0.4248667399949234,
+      "grad_norm": 0.8242712616920471,
+      "learning_rate": 0.001,
+      "loss": 2.2378,
+      "step": 10043
+    },
+    {
+      "epoch": 0.4249090447584398,
+      "grad_norm": 0.21803343296051025,
+      "learning_rate": 0.001,
+      "loss": 3.163,
+      "step": 10044
+    },
+    {
+      "epoch": 0.4249513495219562,
+      "grad_norm": 2.3811676502227783,
+      "learning_rate": 0.001,
+      "loss": 3.5696,
+      "step": 10045
+    },
+    {
+      "epoch": 0.42499365428547253,
+      "grad_norm": 0.5477252006530762,
+      "learning_rate": 0.001,
+      "loss": 2.5047,
+      "step": 10046
+    },
+    {
+      "epoch": 0.42503595904898894,
+      "grad_norm": 0.2212408185005188,
+      "learning_rate": 0.001,
+      "loss": 2.2622,
+      "step": 10047
+    },
+    {
+      "epoch": 0.4250782638125053,
+      "grad_norm": 0.40021783113479614,
+      "learning_rate": 0.001,
+      "loss": 2.4889,
+      "step": 10048
+    },
+    {
+      "epoch": 0.42512056857602165,
+      "grad_norm": 0.24213816225528717,
+      "learning_rate": 0.001,
+      "loss": 2.1517,
+      "step": 10049
+    },
+    {
+      "epoch": 0.425162873339538,
+      "grad_norm": 0.3857748210430145,
+      "learning_rate": 0.001,
+      "loss": 3.2273,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4252051781030544,
+      "grad_norm": 4.682729721069336,
+      "learning_rate": 0.001,
+      "loss": 2.5041,
+      "step": 10051
+    },
+    {
+      "epoch": 0.42524748286657077,
+      "grad_norm": 0.7291063070297241,
+      "learning_rate": 0.001,
+      "loss": 2.3562,
+      "step": 10052
+    },
+    {
+      "epoch": 0.4252897876300871,
+      "grad_norm": 1.8503894805908203,
+      "learning_rate": 0.001,
+      "loss": 2.678,
+      "step": 10053
+    },
+    {
+      "epoch": 0.42533209239360353,
+      "grad_norm": 0.2773977220058441,
+      "learning_rate": 0.001,
+      "loss": 2.6154,
+      "step": 10054
+    },
+    {
+      "epoch": 0.4253743971571199,
+      "grad_norm": 0.2241569608449936,
+      "learning_rate": 0.001,
+      "loss": 2.2345,
+      "step": 10055
+    },
+    {
+      "epoch": 0.42541670192063624,
+      "grad_norm": 0.5478442311286926,
+      "learning_rate": 0.001,
+      "loss": 3.3277,
+      "step": 10056
+    },
+    {
+      "epoch": 0.42545900668415265,
+      "grad_norm": 2.7203447818756104,
+      "learning_rate": 0.001,
+      "loss": 3.8307,
+      "step": 10057
+    },
+    {
+      "epoch": 0.425501311447669,
+      "grad_norm": 0.9797614216804504,
+      "learning_rate": 0.001,
+      "loss": 2.2508,
+      "step": 10058
+    },
+    {
+      "epoch": 0.42554361621118536,
+      "grad_norm": 0.4571799039840698,
+      "learning_rate": 0.001,
+      "loss": 2.9476,
+      "step": 10059
+    },
+    {
+      "epoch": 0.42558592097470177,
+      "grad_norm": 0.34742364287376404,
+      "learning_rate": 0.001,
+      "loss": 2.4439,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4256282257382181,
+      "grad_norm": 0.3077751100063324,
+      "learning_rate": 0.001,
+      "loss": 2.7386,
+      "step": 10061
+    },
+    {
+      "epoch": 0.4256705305017345,
+      "grad_norm": 0.34680673480033875,
+      "learning_rate": 0.001,
+      "loss": 2.7423,
+      "step": 10062
+    },
+    {
+      "epoch": 0.4257128352652509,
+      "grad_norm": 0.5962140560150146,
+      "learning_rate": 0.001,
+      "loss": 3.1924,
+      "step": 10063
+    },
+    {
+      "epoch": 0.42575514002876724,
+      "grad_norm": 0.4237496554851532,
+      "learning_rate": 0.001,
+      "loss": 2.7427,
+      "step": 10064
+    },
+    {
+      "epoch": 0.4257974447922836,
+      "grad_norm": 0.3010278046131134,
+      "learning_rate": 0.001,
+      "loss": 1.9587,
+      "step": 10065
+    },
+    {
+      "epoch": 0.4258397495558,
+      "grad_norm": 0.9290482997894287,
+      "learning_rate": 0.001,
+      "loss": 2.3115,
+      "step": 10066
+    },
+    {
+      "epoch": 0.42588205431931636,
+      "grad_norm": 0.18896250426769257,
+      "learning_rate": 0.001,
+      "loss": 1.8756,
+      "step": 10067
+    },
+    {
+      "epoch": 0.4259243590828327,
+      "grad_norm": 0.7285879254341125,
+      "learning_rate": 0.001,
+      "loss": 2.1316,
+      "step": 10068
+    },
+    {
+      "epoch": 0.4259666638463491,
+      "grad_norm": 0.1963866502046585,
+      "learning_rate": 0.001,
+      "loss": 2.351,
+      "step": 10069
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 5.0258283615112305,
+      "learning_rate": 0.001,
+      "loss": 2.4576,
+      "step": 10070
+    },
+    {
+      "epoch": 0.42605127337338183,
+      "grad_norm": 0.7055137157440186,
+      "learning_rate": 0.001,
+      "loss": 2.9171,
+      "step": 10071
+    },
+    {
+      "epoch": 0.4260935781368982,
+      "grad_norm": 0.2101747691631317,
+      "learning_rate": 0.001,
+      "loss": 1.9525,
+      "step": 10072
+    },
+    {
+      "epoch": 0.4261358829004146,
+      "grad_norm": 0.2830001711845398,
+      "learning_rate": 0.001,
+      "loss": 2.4924,
+      "step": 10073
+    },
+    {
+      "epoch": 0.42617818766393095,
+      "grad_norm": 0.2799762189388275,
+      "learning_rate": 0.001,
+      "loss": 3.067,
+      "step": 10074
+    },
+    {
+      "epoch": 0.4262204924274473,
+      "grad_norm": 0.2741358280181885,
+      "learning_rate": 0.001,
+      "loss": 2.3429,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4262627971909637,
+      "grad_norm": 0.5189446210861206,
+      "learning_rate": 0.001,
+      "loss": 3.3767,
+      "step": 10076
+    },
+    {
+      "epoch": 0.42630510195448007,
+      "grad_norm": 0.2622857391834259,
+      "learning_rate": 0.001,
+      "loss": 2.2139,
+      "step": 10077
+    },
+    {
+      "epoch": 0.4263474067179964,
+      "grad_norm": 0.25797808170318604,
+      "learning_rate": 0.001,
+      "loss": 2.4526,
+      "step": 10078
+    },
+    {
+      "epoch": 0.42638971148151283,
+      "grad_norm": 0.20280535519123077,
+      "learning_rate": 0.001,
+      "loss": 2.2921,
+      "step": 10079
+    },
+    {
+      "epoch": 0.4264320162450292,
+      "grad_norm": 0.18795685470104218,
+      "learning_rate": 0.001,
+      "loss": 2.6622,
+      "step": 10080
+    },
+    {
+      "epoch": 0.42647432100854554,
+      "grad_norm": 1.9836920499801636,
+      "learning_rate": 0.001,
+      "loss": 3.0886,
+      "step": 10081
+    },
+    {
+      "epoch": 0.42651662577206195,
+      "grad_norm": 0.41229698061943054,
+      "learning_rate": 0.001,
+      "loss": 2.5115,
+      "step": 10082
+    },
+    {
+      "epoch": 0.4265589305355783,
+      "grad_norm": 0.17537228763103485,
+      "learning_rate": 0.001,
+      "loss": 2.1755,
+      "step": 10083
+    },
+    {
+      "epoch": 0.42660123529909466,
+      "grad_norm": 0.2055114507675171,
+      "learning_rate": 0.001,
+      "loss": 2.0749,
+      "step": 10084
+    },
+    {
+      "epoch": 0.42664354006261107,
+      "grad_norm": 0.7603585720062256,
+      "learning_rate": 0.001,
+      "loss": 3.0483,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4266858448261274,
+      "grad_norm": 0.15949027240276337,
+      "learning_rate": 0.001,
+      "loss": 2.3814,
+      "step": 10086
+    },
+    {
+      "epoch": 0.4267281495896438,
+      "grad_norm": 0.18160079419612885,
+      "learning_rate": 0.001,
+      "loss": 2.764,
+      "step": 10087
+    },
+    {
+      "epoch": 0.4267704543531602,
+      "grad_norm": 1.1477245092391968,
+      "learning_rate": 0.001,
+      "loss": 2.4743,
+      "step": 10088
+    },
+    {
+      "epoch": 0.42681275911667654,
+      "grad_norm": 0.2228129804134369,
+      "learning_rate": 0.001,
+      "loss": 3.238,
+      "step": 10089
+    },
+    {
+      "epoch": 0.4268550638801929,
+      "grad_norm": 0.7572936415672302,
+      "learning_rate": 0.001,
+      "loss": 2.4797,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4268973686437093,
+      "grad_norm": 0.1875317394733429,
+      "learning_rate": 0.001,
+      "loss": 2.2772,
+      "step": 10091
+    },
+    {
+      "epoch": 0.42693967340722566,
+      "grad_norm": 0.16233602166175842,
+      "learning_rate": 0.001,
+      "loss": 1.489,
+      "step": 10092
+    },
+    {
+      "epoch": 0.426981978170742,
+      "grad_norm": 0.15060947835445404,
+      "learning_rate": 0.001,
+      "loss": 1.9052,
+      "step": 10093
+    },
+    {
+      "epoch": 0.42702428293425837,
+      "grad_norm": 0.28614911437034607,
+      "learning_rate": 0.001,
+      "loss": 2.401,
+      "step": 10094
+    },
+    {
+      "epoch": 0.4270665876977748,
+      "grad_norm": 0.1590694785118103,
+      "learning_rate": 0.001,
+      "loss": 2.4475,
+      "step": 10095
+    },
+    {
+      "epoch": 0.42710889246129113,
+      "grad_norm": 0.2464892566204071,
+      "learning_rate": 0.001,
+      "loss": 2.7287,
+      "step": 10096
+    },
+    {
+      "epoch": 0.4271511972248075,
+      "grad_norm": 0.25172311067581177,
+      "learning_rate": 0.001,
+      "loss": 2.6355,
+      "step": 10097
+    },
+    {
+      "epoch": 0.4271935019883239,
+      "grad_norm": 0.19006308913230896,
+      "learning_rate": 0.001,
+      "loss": 1.8972,
+      "step": 10098
+    },
+    {
+      "epoch": 0.42723580675184025,
+      "grad_norm": 0.17476560175418854,
+      "learning_rate": 0.001,
+      "loss": 2.0524,
+      "step": 10099
+    },
+    {
+      "epoch": 0.4272781115153566,
+      "grad_norm": 0.1643456220626831,
+      "learning_rate": 0.001,
+      "loss": 1.7454,
+      "step": 10100
+    },
+    {
+      "epoch": 0.427320416278873,
+      "grad_norm": 0.1704128235578537,
+      "learning_rate": 0.001,
+      "loss": 1.8189,
+      "step": 10101
+    },
+    {
+      "epoch": 0.42736272104238937,
+      "grad_norm": 0.16339196264743805,
+      "learning_rate": 0.001,
+      "loss": 2.1309,
+      "step": 10102
+    },
+    {
+      "epoch": 0.4274050258059057,
+      "grad_norm": 0.27944785356521606,
+      "learning_rate": 0.001,
+      "loss": 2.072,
+      "step": 10103
+    },
+    {
+      "epoch": 0.42744733056942213,
+      "grad_norm": 0.8265528082847595,
+      "learning_rate": 0.001,
+      "loss": 2.8271,
+      "step": 10104
+    },
+    {
+      "epoch": 0.4274896353329385,
+      "grad_norm": 0.16367796063423157,
+      "learning_rate": 0.001,
+      "loss": 2.0271,
+      "step": 10105
+    },
+    {
+      "epoch": 0.42753194009645484,
+      "grad_norm": 0.17860738933086395,
+      "learning_rate": 0.001,
+      "loss": 1.4789,
+      "step": 10106
+    },
+    {
+      "epoch": 0.42757424485997125,
+      "grad_norm": 0.2383193075656891,
+      "learning_rate": 0.001,
+      "loss": 2.6022,
+      "step": 10107
+    },
+    {
+      "epoch": 0.4276165496234876,
+      "grad_norm": 0.9572615027427673,
+      "learning_rate": 0.001,
+      "loss": 3.0684,
+      "step": 10108
+    },
+    {
+      "epoch": 0.42765885438700396,
+      "grad_norm": 0.31203147768974304,
+      "learning_rate": 0.001,
+      "loss": 2.1978,
+      "step": 10109
+    },
+    {
+      "epoch": 0.42770115915052037,
+      "grad_norm": 0.2197791486978531,
+      "learning_rate": 0.001,
+      "loss": 2.5708,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4277434639140367,
+      "grad_norm": 0.5677201151847839,
+      "learning_rate": 0.001,
+      "loss": 1.6285,
+      "step": 10111
+    },
+    {
+      "epoch": 0.4277857686775531,
+      "grad_norm": 0.18387927114963531,
+      "learning_rate": 0.001,
+      "loss": 1.9223,
+      "step": 10112
+    },
+    {
+      "epoch": 0.4278280734410695,
+      "grad_norm": 0.16864502429962158,
+      "learning_rate": 0.001,
+      "loss": 1.8276,
+      "step": 10113
+    },
+    {
+      "epoch": 0.42787037820458584,
+      "grad_norm": 0.23291771113872528,
+      "learning_rate": 0.001,
+      "loss": 3.0319,
+      "step": 10114
+    },
+    {
+      "epoch": 0.4279126829681022,
+      "grad_norm": 0.18459783494472504,
+      "learning_rate": 0.001,
+      "loss": 1.8791,
+      "step": 10115
+    },
+    {
+      "epoch": 0.4279549877316186,
+      "grad_norm": 1.1795158386230469,
+      "learning_rate": 0.001,
+      "loss": 2.6287,
+      "step": 10116
+    },
+    {
+      "epoch": 0.42799729249513496,
+      "grad_norm": 0.20951585471630096,
+      "learning_rate": 0.001,
+      "loss": 1.9405,
+      "step": 10117
+    },
+    {
+      "epoch": 0.4280395972586513,
+      "grad_norm": 0.31406983733177185,
+      "learning_rate": 0.001,
+      "loss": 3.4603,
+      "step": 10118
+    },
+    {
+      "epoch": 0.42808190202216767,
+      "grad_norm": 0.267755925655365,
+      "learning_rate": 0.001,
+      "loss": 3.2455,
+      "step": 10119
+    },
+    {
+      "epoch": 0.4281242067856841,
+      "grad_norm": 0.21892082691192627,
+      "learning_rate": 0.001,
+      "loss": 2.5571,
+      "step": 10120
+    },
+    {
+      "epoch": 0.42816651154920043,
+      "grad_norm": 0.17393141984939575,
+      "learning_rate": 0.001,
+      "loss": 1.6598,
+      "step": 10121
+    },
+    {
+      "epoch": 0.4282088163127168,
+      "grad_norm": 0.17320716381072998,
+      "learning_rate": 0.001,
+      "loss": 2.0091,
+      "step": 10122
+    },
+    {
+      "epoch": 0.4282511210762332,
+      "grad_norm": 0.1930369734764099,
+      "learning_rate": 0.001,
+      "loss": 1.7533,
+      "step": 10123
+    },
+    {
+      "epoch": 0.42829342583974955,
+      "grad_norm": 0.18624812364578247,
+      "learning_rate": 0.001,
+      "loss": 2.0146,
+      "step": 10124
+    },
+    {
+      "epoch": 0.4283357306032659,
+      "grad_norm": 0.7041241526603699,
+      "learning_rate": 0.001,
+      "loss": 2.8163,
+      "step": 10125
+    },
+    {
+      "epoch": 0.4283780353667823,
+      "grad_norm": 0.2105880081653595,
+      "learning_rate": 0.001,
+      "loss": 1.9341,
+      "step": 10126
+    },
+    {
+      "epoch": 0.42842034013029867,
+      "grad_norm": 0.1799003928899765,
+      "learning_rate": 0.001,
+      "loss": 2.2052,
+      "step": 10127
+    },
+    {
+      "epoch": 0.428462644893815,
+      "grad_norm": 1.5456318855285645,
+      "learning_rate": 0.001,
+      "loss": 1.8783,
+      "step": 10128
+    },
+    {
+      "epoch": 0.42850494965733144,
+      "grad_norm": 0.17844410240650177,
+      "learning_rate": 0.001,
+      "loss": 2.3216,
+      "step": 10129
+    },
+    {
+      "epoch": 0.4285472544208478,
+      "grad_norm": 0.16258619725704193,
+      "learning_rate": 0.001,
+      "loss": 2.56,
+      "step": 10130
+    },
+    {
+      "epoch": 0.42858955918436414,
+      "grad_norm": 0.18964774906635284,
+      "learning_rate": 0.001,
+      "loss": 1.9728,
+      "step": 10131
+    },
+    {
+      "epoch": 0.42863186394788055,
+      "grad_norm": 0.18737366795539856,
+      "learning_rate": 0.001,
+      "loss": 2.6866,
+      "step": 10132
+    },
+    {
+      "epoch": 0.4286741687113969,
+      "grad_norm": 0.19086334109306335,
+      "learning_rate": 0.001,
+      "loss": 2.0617,
+      "step": 10133
+    },
+    {
+      "epoch": 0.42871647347491326,
+      "grad_norm": 0.2464657872915268,
+      "learning_rate": 0.001,
+      "loss": 2.9955,
+      "step": 10134
+    },
+    {
+      "epoch": 0.42875877823842967,
+      "grad_norm": 0.3504067063331604,
+      "learning_rate": 0.001,
+      "loss": 2.5686,
+      "step": 10135
+    },
+    {
+      "epoch": 0.428801083001946,
+      "grad_norm": 0.17571739852428436,
+      "learning_rate": 0.001,
+      "loss": 2.1223,
+      "step": 10136
+    },
+    {
+      "epoch": 0.4288433877654624,
+      "grad_norm": 0.381346195936203,
+      "learning_rate": 0.001,
+      "loss": 1.5416,
+      "step": 10137
+    },
+    {
+      "epoch": 0.4288856925289788,
+      "grad_norm": 0.2737821936607361,
+      "learning_rate": 0.001,
+      "loss": 1.7345,
+      "step": 10138
+    },
+    {
+      "epoch": 0.42892799729249514,
+      "grad_norm": 0.17216061055660248,
+      "learning_rate": 0.001,
+      "loss": 1.6129,
+      "step": 10139
+    },
+    {
+      "epoch": 0.4289703020560115,
+      "grad_norm": 0.20674671232700348,
+      "learning_rate": 0.001,
+      "loss": 3.2899,
+      "step": 10140
+    },
+    {
+      "epoch": 0.42901260681952785,
+      "grad_norm": 1.2608942985534668,
+      "learning_rate": 0.001,
+      "loss": 1.8988,
+      "step": 10141
+    },
+    {
+      "epoch": 0.42905491158304426,
+      "grad_norm": 0.18615379929542542,
+      "learning_rate": 0.001,
+      "loss": 1.873,
+      "step": 10142
+    },
+    {
+      "epoch": 0.4290972163465606,
+      "grad_norm": 4.064700603485107,
+      "learning_rate": 0.001,
+      "loss": 2.7955,
+      "step": 10143
+    },
+    {
+      "epoch": 0.42913952111007697,
+      "grad_norm": 0.2223389595746994,
+      "learning_rate": 0.001,
+      "loss": 2.7764,
+      "step": 10144
+    },
+    {
+      "epoch": 0.4291818258735934,
+      "grad_norm": 0.4159873127937317,
+      "learning_rate": 0.001,
+      "loss": 3.8135,
+      "step": 10145
+    },
+    {
+      "epoch": 0.42922413063710974,
+      "grad_norm": 0.23633842170238495,
+      "learning_rate": 0.001,
+      "loss": 3.5594,
+      "step": 10146
+    },
+    {
+      "epoch": 0.4292664354006261,
+      "grad_norm": 1.0280625820159912,
+      "learning_rate": 0.001,
+      "loss": 1.7139,
+      "step": 10147
+    },
+    {
+      "epoch": 0.4293087401641425,
+      "grad_norm": 0.1797591596841812,
+      "learning_rate": 0.001,
+      "loss": 3.0655,
+      "step": 10148
+    },
+    {
+      "epoch": 0.42935104492765885,
+      "grad_norm": 0.4903676509857178,
+      "learning_rate": 0.001,
+      "loss": 1.8549,
+      "step": 10149
+    },
+    {
+      "epoch": 0.4293933496911752,
+      "grad_norm": 0.15862198173999786,
+      "learning_rate": 0.001,
+      "loss": 2.6685,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4294356544546916,
+      "grad_norm": 0.22118091583251953,
+      "learning_rate": 0.001,
+      "loss": 2.2547,
+      "step": 10151
+    },
+    {
+      "epoch": 0.42947795921820797,
+      "grad_norm": 0.18271955847740173,
+      "learning_rate": 0.001,
+      "loss": 2.2298,
+      "step": 10152
+    },
+    {
+      "epoch": 0.4295202639817243,
+      "grad_norm": 0.170980766415596,
+      "learning_rate": 0.001,
+      "loss": 2.8904,
+      "step": 10153
+    },
+    {
+      "epoch": 0.42956256874524074,
+      "grad_norm": 0.15399718284606934,
+      "learning_rate": 0.001,
+      "loss": 2.2953,
+      "step": 10154
+    },
+    {
+      "epoch": 0.4296048735087571,
+      "grad_norm": 0.17953842878341675,
+      "learning_rate": 0.001,
+      "loss": 1.9147,
+      "step": 10155
+    },
+    {
+      "epoch": 0.42964717827227344,
+      "grad_norm": 0.1702369898557663,
+      "learning_rate": 0.001,
+      "loss": 2.2872,
+      "step": 10156
+    },
+    {
+      "epoch": 0.42968948303578985,
+      "grad_norm": 0.19081392884254456,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 10157
+    },
+    {
+      "epoch": 0.4297317877993062,
+      "grad_norm": 0.1783342808485031,
+      "learning_rate": 0.001,
+      "loss": 1.9582,
+      "step": 10158
+    },
+    {
+      "epoch": 0.42977409256282256,
+      "grad_norm": 0.20043319463729858,
+      "learning_rate": 0.001,
+      "loss": 2.3066,
+      "step": 10159
+    },
+    {
+      "epoch": 0.429816397326339,
+      "grad_norm": 0.17219308018684387,
+      "learning_rate": 0.001,
+      "loss": 2.5435,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4298587020898553,
+      "grad_norm": 0.29747068881988525,
+      "learning_rate": 0.001,
+      "loss": 2.9951,
+      "step": 10161
+    },
+    {
+      "epoch": 0.4299010068533717,
+      "grad_norm": 4.550970077514648,
+      "learning_rate": 0.001,
+      "loss": 2.5154,
+      "step": 10162
+    },
+    {
+      "epoch": 0.42994331161688804,
+      "grad_norm": 0.18527233600616455,
+      "learning_rate": 0.001,
+      "loss": 2.0328,
+      "step": 10163
+    },
+    {
+      "epoch": 0.42998561638040445,
+      "grad_norm": 0.18644128739833832,
+      "learning_rate": 0.001,
+      "loss": 2.3224,
+      "step": 10164
+    },
+    {
+      "epoch": 0.4300279211439208,
+      "grad_norm": 0.2518557906150818,
+      "learning_rate": 0.001,
+      "loss": 2.1388,
+      "step": 10165
+    },
+    {
+      "epoch": 0.43007022590743715,
+      "grad_norm": 0.1625964343547821,
+      "learning_rate": 0.001,
+      "loss": 1.8495,
+      "step": 10166
+    },
+    {
+      "epoch": 0.43011253067095356,
+      "grad_norm": 0.2331293672323227,
+      "learning_rate": 0.001,
+      "loss": 2.216,
+      "step": 10167
+    },
+    {
+      "epoch": 0.4301548354344699,
+      "grad_norm": 0.2668088376522064,
+      "learning_rate": 0.001,
+      "loss": 2.3291,
+      "step": 10168
+    },
+    {
+      "epoch": 0.43019714019798627,
+      "grad_norm": 0.17521047592163086,
+      "learning_rate": 0.001,
+      "loss": 1.9921,
+      "step": 10169
+    },
+    {
+      "epoch": 0.4302394449615027,
+      "grad_norm": 0.19229163229465485,
+      "learning_rate": 0.001,
+      "loss": 2.0076,
+      "step": 10170
+    },
+    {
+      "epoch": 0.43028174972501904,
+      "grad_norm": 0.2229296714067459,
+      "learning_rate": 0.001,
+      "loss": 2.5076,
+      "step": 10171
+    },
+    {
+      "epoch": 0.4303240544885354,
+      "grad_norm": 0.47033241391181946,
+      "learning_rate": 0.001,
+      "loss": 2.2927,
+      "step": 10172
+    },
+    {
+      "epoch": 0.4303663592520518,
+      "grad_norm": 0.1660117357969284,
+      "learning_rate": 0.001,
+      "loss": 2.2447,
+      "step": 10173
+    },
+    {
+      "epoch": 0.43040866401556815,
+      "grad_norm": 2.12497615814209,
+      "learning_rate": 0.001,
+      "loss": 2.0389,
+      "step": 10174
+    },
+    {
+      "epoch": 0.4304509687790845,
+      "grad_norm": 0.18803949654102325,
+      "learning_rate": 0.001,
+      "loss": 3.1254,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.2218303233385086,
+      "learning_rate": 0.001,
+      "loss": 1.7379,
+      "step": 10176
+    },
+    {
+      "epoch": 0.4305355783061173,
+      "grad_norm": 0.14620518684387207,
+      "learning_rate": 0.001,
+      "loss": 1.6227,
+      "step": 10177
+    },
+    {
+      "epoch": 0.4305778830696336,
+      "grad_norm": 0.17080405354499817,
+      "learning_rate": 0.001,
+      "loss": 1.9532,
+      "step": 10178
+    },
+    {
+      "epoch": 0.43062018783315004,
+      "grad_norm": 2.5071756839752197,
+      "learning_rate": 0.001,
+      "loss": 3.2385,
+      "step": 10179
+    },
+    {
+      "epoch": 0.4306624925966664,
+      "grad_norm": 0.17002911865711212,
+      "learning_rate": 0.001,
+      "loss": 3.0736,
+      "step": 10180
+    },
+    {
+      "epoch": 0.43070479736018275,
+      "grad_norm": 0.1898619830608368,
+      "learning_rate": 0.001,
+      "loss": 1.6932,
+      "step": 10181
+    },
+    {
+      "epoch": 0.43074710212369915,
+      "grad_norm": 0.6082344651222229,
+      "learning_rate": 0.001,
+      "loss": 3.6207,
+      "step": 10182
+    },
+    {
+      "epoch": 0.4307894068872155,
+      "grad_norm": 0.6188779473304749,
+      "learning_rate": 0.001,
+      "loss": 1.6398,
+      "step": 10183
+    },
+    {
+      "epoch": 0.43083171165073186,
+      "grad_norm": 0.7821673154830933,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 10184
+    },
+    {
+      "epoch": 0.4308740164142482,
+      "grad_norm": 0.543019711971283,
+      "learning_rate": 0.001,
+      "loss": 1.6907,
+      "step": 10185
+    },
+    {
+      "epoch": 0.4309163211777646,
+      "grad_norm": 0.5596624612808228,
+      "learning_rate": 0.001,
+      "loss": 3.4859,
+      "step": 10186
+    },
+    {
+      "epoch": 0.430958625941281,
+      "grad_norm": 0.22357861697673798,
+      "learning_rate": 0.001,
+      "loss": 1.896,
+      "step": 10187
+    },
+    {
+      "epoch": 0.43100093070479734,
+      "grad_norm": 0.16170749068260193,
+      "learning_rate": 0.001,
+      "loss": 1.8071,
+      "step": 10188
+    },
+    {
+      "epoch": 0.43104323546831375,
+      "grad_norm": 0.15653282403945923,
+      "learning_rate": 0.001,
+      "loss": 1.6408,
+      "step": 10189
+    },
+    {
+      "epoch": 0.4310855402318301,
+      "grad_norm": 0.22891934216022491,
+      "learning_rate": 0.001,
+      "loss": 1.2615,
+      "step": 10190
+    },
+    {
+      "epoch": 0.43112784499534645,
+      "grad_norm": 0.9978185892105103,
+      "learning_rate": 0.001,
+      "loss": 2.0704,
+      "step": 10191
+    },
+    {
+      "epoch": 0.43117014975886286,
+      "grad_norm": 0.1806502640247345,
+      "learning_rate": 0.001,
+      "loss": 2.1097,
+      "step": 10192
+    },
+    {
+      "epoch": 0.4312124545223792,
+      "grad_norm": 0.16584782302379608,
+      "learning_rate": 0.001,
+      "loss": 2.528,
+      "step": 10193
+    },
+    {
+      "epoch": 0.4312547592858956,
+      "grad_norm": 0.39718106389045715,
+      "learning_rate": 0.001,
+      "loss": 2.2111,
+      "step": 10194
+    },
+    {
+      "epoch": 0.431297064049412,
+      "grad_norm": 0.37005481123924255,
+      "learning_rate": 0.001,
+      "loss": 2.8491,
+      "step": 10195
+    },
+    {
+      "epoch": 0.43133936881292834,
+      "grad_norm": 0.18187114596366882,
+      "learning_rate": 0.001,
+      "loss": 2.0542,
+      "step": 10196
+    },
+    {
+      "epoch": 0.4313816735764447,
+      "grad_norm": 0.21417534351348877,
+      "learning_rate": 0.001,
+      "loss": 1.842,
+      "step": 10197
+    },
+    {
+      "epoch": 0.4314239783399611,
+      "grad_norm": 0.2430461347103119,
+      "learning_rate": 0.001,
+      "loss": 3.2218,
+      "step": 10198
+    },
+    {
+      "epoch": 0.43146628310347745,
+      "grad_norm": 0.1812175065279007,
+      "learning_rate": 0.001,
+      "loss": 2.5477,
+      "step": 10199
+    },
+    {
+      "epoch": 0.4315085878669938,
+      "grad_norm": 0.5557557940483093,
+      "learning_rate": 0.001,
+      "loss": 2.0329,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4315508926305102,
+      "grad_norm": 0.16666504740715027,
+      "learning_rate": 0.001,
+      "loss": 1.8955,
+      "step": 10201
+    },
+    {
+      "epoch": 0.4315931973940266,
+      "grad_norm": 8.631728172302246,
+      "learning_rate": 0.001,
+      "loss": 2.2758,
+      "step": 10202
+    },
+    {
+      "epoch": 0.4316355021575429,
+      "grad_norm": 0.28974688053131104,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 10203
+    },
+    {
+      "epoch": 0.43167780692105934,
+      "grad_norm": 0.17004406452178955,
+      "learning_rate": 0.001,
+      "loss": 1.7401,
+      "step": 10204
+    },
+    {
+      "epoch": 0.4317201116845757,
+      "grad_norm": 0.3253629803657532,
+      "learning_rate": 0.001,
+      "loss": 2.3806,
+      "step": 10205
+    },
+    {
+      "epoch": 0.43176241644809205,
+      "grad_norm": 0.18079756200313568,
+      "learning_rate": 0.001,
+      "loss": 1.6866,
+      "step": 10206
+    },
+    {
+      "epoch": 0.4318047212116084,
+      "grad_norm": 0.16561074554920197,
+      "learning_rate": 0.001,
+      "loss": 2.1134,
+      "step": 10207
+    },
+    {
+      "epoch": 0.4318470259751248,
+      "grad_norm": 0.16736382246017456,
+      "learning_rate": 0.001,
+      "loss": 2.5646,
+      "step": 10208
+    },
+    {
+      "epoch": 0.43188933073864116,
+      "grad_norm": 0.16732652485370636,
+      "learning_rate": 0.001,
+      "loss": 1.7875,
+      "step": 10209
+    },
+    {
+      "epoch": 0.4319316355021575,
+      "grad_norm": 1.6806914806365967,
+      "learning_rate": 0.001,
+      "loss": 1.8155,
+      "step": 10210
+    },
+    {
+      "epoch": 0.43197394026567393,
+      "grad_norm": 0.902766764163971,
+      "learning_rate": 0.001,
+      "loss": 2.1906,
+      "step": 10211
+    },
+    {
+      "epoch": 0.4320162450291903,
+      "grad_norm": 0.22574925422668457,
+      "learning_rate": 0.001,
+      "loss": 2.4012,
+      "step": 10212
+    },
+    {
+      "epoch": 0.43205854979270664,
+      "grad_norm": 0.20272192358970642,
+      "learning_rate": 0.001,
+      "loss": 2.7397,
+      "step": 10213
+    },
+    {
+      "epoch": 0.43210085455622305,
+      "grad_norm": 0.18217982351779938,
+      "learning_rate": 0.001,
+      "loss": 1.8363,
+      "step": 10214
+    },
+    {
+      "epoch": 0.4321431593197394,
+      "grad_norm": 0.32555845379829407,
+      "learning_rate": 0.001,
+      "loss": 3.1817,
+      "step": 10215
+    },
+    {
+      "epoch": 0.43218546408325575,
+      "grad_norm": 4.589909076690674,
+      "learning_rate": 0.001,
+      "loss": 2.9862,
+      "step": 10216
+    },
+    {
+      "epoch": 0.43222776884677216,
+      "grad_norm": 0.24728332459926605,
+      "learning_rate": 0.001,
+      "loss": 1.834,
+      "step": 10217
+    },
+    {
+      "epoch": 0.4322700736102885,
+      "grad_norm": 0.331990510225296,
+      "learning_rate": 0.001,
+      "loss": 3.6178,
+      "step": 10218
+    },
+    {
+      "epoch": 0.4323123783738049,
+      "grad_norm": 0.31489211320877075,
+      "learning_rate": 0.001,
+      "loss": 1.8217,
+      "step": 10219
+    },
+    {
+      "epoch": 0.4323546831373213,
+      "grad_norm": 1.1140625476837158,
+      "learning_rate": 0.001,
+      "loss": 2.5692,
+      "step": 10220
+    },
+    {
+      "epoch": 0.43239698790083764,
+      "grad_norm": 0.1725742071866989,
+      "learning_rate": 0.001,
+      "loss": 1.7017,
+      "step": 10221
+    },
+    {
+      "epoch": 0.432439292664354,
+      "grad_norm": 0.19503003358840942,
+      "learning_rate": 0.001,
+      "loss": 2.5324,
+      "step": 10222
+    },
+    {
+      "epoch": 0.4324815974278704,
+      "grad_norm": 0.26874589920043945,
+      "learning_rate": 0.001,
+      "loss": 3.1633,
+      "step": 10223
+    },
+    {
+      "epoch": 0.43252390219138676,
+      "grad_norm": 0.24442531168460846,
+      "learning_rate": 0.001,
+      "loss": 2.4106,
+      "step": 10224
+    },
+    {
+      "epoch": 0.4325662069549031,
+      "grad_norm": 0.32704654335975647,
+      "learning_rate": 0.001,
+      "loss": 3.401,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4326085117184195,
+      "grad_norm": 0.2951880693435669,
+      "learning_rate": 0.001,
+      "loss": 2.9172,
+      "step": 10226
+    },
+    {
+      "epoch": 0.4326508164819359,
+      "grad_norm": 0.16954240202903748,
+      "learning_rate": 0.001,
+      "loss": 1.8404,
+      "step": 10227
+    },
+    {
+      "epoch": 0.43269312124545223,
+      "grad_norm": 0.1641477644443512,
+      "learning_rate": 0.001,
+      "loss": 3.1311,
+      "step": 10228
+    },
+    {
+      "epoch": 0.4327354260089686,
+      "grad_norm": 0.7109168171882629,
+      "learning_rate": 0.001,
+      "loss": 1.9146,
+      "step": 10229
+    },
+    {
+      "epoch": 0.432777730772485,
+      "grad_norm": 0.21353894472122192,
+      "learning_rate": 0.001,
+      "loss": 2.6354,
+      "step": 10230
+    },
+    {
+      "epoch": 0.43282003553600135,
+      "grad_norm": 0.32388636469841003,
+      "learning_rate": 0.001,
+      "loss": 1.8109,
+      "step": 10231
+    },
+    {
+      "epoch": 0.4328623402995177,
+      "grad_norm": 0.46089649200439453,
+      "learning_rate": 0.001,
+      "loss": 3.1302,
+      "step": 10232
+    },
+    {
+      "epoch": 0.4329046450630341,
+      "grad_norm": 0.28092247247695923,
+      "learning_rate": 0.001,
+      "loss": 2.9537,
+      "step": 10233
+    },
+    {
+      "epoch": 0.43294694982655046,
+      "grad_norm": 0.1795322597026825,
+      "learning_rate": 0.001,
+      "loss": 2.0469,
+      "step": 10234
+    },
+    {
+      "epoch": 0.4329892545900668,
+      "grad_norm": 1.1064038276672363,
+      "learning_rate": 0.001,
+      "loss": 2.0708,
+      "step": 10235
+    },
+    {
+      "epoch": 0.43303155935358323,
+      "grad_norm": 0.1590038537979126,
+      "learning_rate": 0.001,
+      "loss": 3.0817,
+      "step": 10236
+    },
+    {
+      "epoch": 0.4330738641170996,
+      "grad_norm": 0.1741926223039627,
+      "learning_rate": 0.001,
+      "loss": 2.1068,
+      "step": 10237
+    },
+    {
+      "epoch": 0.43311616888061594,
+      "grad_norm": 0.15864573419094086,
+      "learning_rate": 0.001,
+      "loss": 2.225,
+      "step": 10238
+    },
+    {
+      "epoch": 0.43315847364413235,
+      "grad_norm": 0.1976540982723236,
+      "learning_rate": 0.001,
+      "loss": 2.0396,
+      "step": 10239
+    },
+    {
+      "epoch": 0.4332007784076487,
+      "grad_norm": 0.1970503330230713,
+      "learning_rate": 0.001,
+      "loss": 2.3251,
+      "step": 10240
+    },
+    {
+      "epoch": 0.43324308317116506,
+      "grad_norm": 0.19585905969142914,
+      "learning_rate": 0.001,
+      "loss": 2.9537,
+      "step": 10241
+    },
+    {
+      "epoch": 0.43328538793468147,
+      "grad_norm": 0.5313712358474731,
+      "learning_rate": 0.001,
+      "loss": 2.066,
+      "step": 10242
+    },
+    {
+      "epoch": 0.4333276926981978,
+      "grad_norm": 0.3845883905887604,
+      "learning_rate": 0.001,
+      "loss": 3.0735,
+      "step": 10243
+    },
+    {
+      "epoch": 0.4333699974617142,
+      "grad_norm": 4.409493446350098,
+      "learning_rate": 0.001,
+      "loss": 2.7471,
+      "step": 10244
+    },
+    {
+      "epoch": 0.4334123022252306,
+      "grad_norm": 0.15902769565582275,
+      "learning_rate": 0.001,
+      "loss": 2.0389,
+      "step": 10245
+    },
+    {
+      "epoch": 0.43345460698874694,
+      "grad_norm": 0.19496986269950867,
+      "learning_rate": 0.001,
+      "loss": 2.2368,
+      "step": 10246
+    },
+    {
+      "epoch": 0.4334969117522633,
+      "grad_norm": 3.496062755584717,
+      "learning_rate": 0.001,
+      "loss": 2.2234,
+      "step": 10247
+    },
+    {
+      "epoch": 0.4335392165157797,
+      "grad_norm": 0.28134244680404663,
+      "learning_rate": 0.001,
+      "loss": 1.4976,
+      "step": 10248
+    },
+    {
+      "epoch": 0.43358152127929606,
+      "grad_norm": 0.2013697773218155,
+      "learning_rate": 0.001,
+      "loss": 1.9873,
+      "step": 10249
+    },
+    {
+      "epoch": 0.4336238260428124,
+      "grad_norm": 0.42853641510009766,
+      "learning_rate": 0.001,
+      "loss": 3.4437,
+      "step": 10250
+    },
+    {
+      "epoch": 0.4336661308063288,
+      "grad_norm": 2.750063896179199,
+      "learning_rate": 0.001,
+      "loss": 2.0177,
+      "step": 10251
+    },
+    {
+      "epoch": 0.4337084355698452,
+      "grad_norm": 0.21362249553203583,
+      "learning_rate": 0.001,
+      "loss": 2.6798,
+      "step": 10252
+    },
+    {
+      "epoch": 0.43375074033336153,
+      "grad_norm": 0.2875850200653076,
+      "learning_rate": 0.001,
+      "loss": 2.2333,
+      "step": 10253
+    },
+    {
+      "epoch": 0.4337930450968779,
+      "grad_norm": 0.2688775360584259,
+      "learning_rate": 0.001,
+      "loss": 2.2072,
+      "step": 10254
+    },
+    {
+      "epoch": 0.4338353498603943,
+      "grad_norm": 0.3297450840473175,
+      "learning_rate": 0.001,
+      "loss": 2.5938,
+      "step": 10255
+    },
+    {
+      "epoch": 0.43387765462391065,
+      "grad_norm": 0.4499950706958771,
+      "learning_rate": 0.001,
+      "loss": 2.221,
+      "step": 10256
+    },
+    {
+      "epoch": 0.433919959387427,
+      "grad_norm": 0.19243115186691284,
+      "learning_rate": 0.001,
+      "loss": 3.2843,
+      "step": 10257
+    },
+    {
+      "epoch": 0.4339622641509434,
+      "grad_norm": 0.18401968479156494,
+      "learning_rate": 0.001,
+      "loss": 1.8309,
+      "step": 10258
+    },
+    {
+      "epoch": 0.43400456891445977,
+      "grad_norm": 0.18404164910316467,
+      "learning_rate": 0.001,
+      "loss": 2.295,
+      "step": 10259
+    },
+    {
+      "epoch": 0.4340468736779761,
+      "grad_norm": 0.1946217566728592,
+      "learning_rate": 0.001,
+      "loss": 2.0939,
+      "step": 10260
+    },
+    {
+      "epoch": 0.43408917844149253,
+      "grad_norm": 0.1470000147819519,
+      "learning_rate": 0.001,
+      "loss": 1.7336,
+      "step": 10261
+    },
+    {
+      "epoch": 0.4341314832050089,
+      "grad_norm": 0.32551971077919006,
+      "learning_rate": 0.001,
+      "loss": 2.4746,
+      "step": 10262
+    },
+    {
+      "epoch": 0.43417378796852524,
+      "grad_norm": 2.652297019958496,
+      "learning_rate": 0.001,
+      "loss": 3.1451,
+      "step": 10263
+    },
+    {
+      "epoch": 0.43421609273204165,
+      "grad_norm": 0.1553354561328888,
+      "learning_rate": 0.001,
+      "loss": 1.976,
+      "step": 10264
+    },
+    {
+      "epoch": 0.434258397495558,
+      "grad_norm": 0.23101647198200226,
+      "learning_rate": 0.001,
+      "loss": 2.6715,
+      "step": 10265
+    },
+    {
+      "epoch": 0.43430070225907436,
+      "grad_norm": 0.20663489401340485,
+      "learning_rate": 0.001,
+      "loss": 1.561,
+      "step": 10266
+    },
+    {
+      "epoch": 0.43434300702259077,
+      "grad_norm": 0.19089651107788086,
+      "learning_rate": 0.001,
+      "loss": 2.8731,
+      "step": 10267
+    },
+    {
+      "epoch": 0.4343853117861071,
+      "grad_norm": 0.14528456330299377,
+      "learning_rate": 0.001,
+      "loss": 2.2012,
+      "step": 10268
+    },
+    {
+      "epoch": 0.4344276165496235,
+      "grad_norm": 0.1953110247850418,
+      "learning_rate": 0.001,
+      "loss": 2.5863,
+      "step": 10269
+    },
+    {
+      "epoch": 0.4344699213131399,
+      "grad_norm": 0.19322249293327332,
+      "learning_rate": 0.001,
+      "loss": 2.5648,
+      "step": 10270
+    },
+    {
+      "epoch": 0.43451222607665624,
+      "grad_norm": 0.15629512071609497,
+      "learning_rate": 0.001,
+      "loss": 2.6854,
+      "step": 10271
+    },
+    {
+      "epoch": 0.4345545308401726,
+      "grad_norm": 0.17981792986392975,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 10272
+    },
+    {
+      "epoch": 0.434596835603689,
+      "grad_norm": 0.18551938235759735,
+      "learning_rate": 0.001,
+      "loss": 2.6762,
+      "step": 10273
+    },
+    {
+      "epoch": 0.43463914036720536,
+      "grad_norm": 2.313119649887085,
+      "learning_rate": 0.001,
+      "loss": 3.5626,
+      "step": 10274
+    },
+    {
+      "epoch": 0.4346814451307217,
+      "grad_norm": 0.34999650716781616,
+      "learning_rate": 0.001,
+      "loss": 2.6788,
+      "step": 10275
+    },
+    {
+      "epoch": 0.43472374989423807,
+      "grad_norm": 0.36393770575523376,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 10276
+    },
+    {
+      "epoch": 0.4347660546577545,
+      "grad_norm": 0.4789801239967346,
+      "learning_rate": 0.001,
+      "loss": 1.7946,
+      "step": 10277
+    },
+    {
+      "epoch": 0.43480835942127083,
+      "grad_norm": 2.8181819915771484,
+      "learning_rate": 0.001,
+      "loss": 2.2287,
+      "step": 10278
+    },
+    {
+      "epoch": 0.4348506641847872,
+      "grad_norm": 0.5375362038612366,
+      "learning_rate": 0.001,
+      "loss": 2.2571,
+      "step": 10279
+    },
+    {
+      "epoch": 0.4348929689483036,
+      "grad_norm": 0.1631775200366974,
+      "learning_rate": 0.001,
+      "loss": 2.6874,
+      "step": 10280
+    },
+    {
+      "epoch": 0.43493527371181995,
+      "grad_norm": 0.1752665489912033,
+      "learning_rate": 0.001,
+      "loss": 2.6751,
+      "step": 10281
+    },
+    {
+      "epoch": 0.4349775784753363,
+      "grad_norm": 0.28829485177993774,
+      "learning_rate": 0.001,
+      "loss": 1.4542,
+      "step": 10282
+    },
+    {
+      "epoch": 0.4350198832388527,
+      "grad_norm": 0.6851754188537598,
+      "learning_rate": 0.001,
+      "loss": 2.094,
+      "step": 10283
+    },
+    {
+      "epoch": 0.43506218800236907,
+      "grad_norm": 0.1630009263753891,
+      "learning_rate": 0.001,
+      "loss": 2.3703,
+      "step": 10284
+    },
+    {
+      "epoch": 0.4351044927658854,
+      "grad_norm": 0.7183140516281128,
+      "learning_rate": 0.001,
+      "loss": 3.3134,
+      "step": 10285
+    },
+    {
+      "epoch": 0.43514679752940183,
+      "grad_norm": 0.17298482358455658,
+      "learning_rate": 0.001,
+      "loss": 2.7842,
+      "step": 10286
+    },
+    {
+      "epoch": 0.4351891022929182,
+      "grad_norm": 0.16329266130924225,
+      "learning_rate": 0.001,
+      "loss": 1.5573,
+      "step": 10287
+    },
+    {
+      "epoch": 0.43523140705643454,
+      "grad_norm": 0.15913952887058258,
+      "learning_rate": 0.001,
+      "loss": 2.8684,
+      "step": 10288
+    },
+    {
+      "epoch": 0.43527371181995095,
+      "grad_norm": 0.16433869302272797,
+      "learning_rate": 0.001,
+      "loss": 1.6984,
+      "step": 10289
+    },
+    {
+      "epoch": 0.4353160165834673,
+      "grad_norm": 0.1634620726108551,
+      "learning_rate": 0.001,
+      "loss": 2.2076,
+      "step": 10290
+    },
+    {
+      "epoch": 0.43535832134698366,
+      "grad_norm": 0.2353067547082901,
+      "learning_rate": 0.001,
+      "loss": 2.8474,
+      "step": 10291
+    },
+    {
+      "epoch": 0.43540062611050007,
+      "grad_norm": 0.1973012238740921,
+      "learning_rate": 0.001,
+      "loss": 2.5835,
+      "step": 10292
+    },
+    {
+      "epoch": 0.4354429308740164,
+      "grad_norm": 0.16465625166893005,
+      "learning_rate": 0.001,
+      "loss": 1.8952,
+      "step": 10293
+    },
+    {
+      "epoch": 0.4354852356375328,
+      "grad_norm": 1.9926971197128296,
+      "learning_rate": 0.001,
+      "loss": 1.7796,
+      "step": 10294
+    },
+    {
+      "epoch": 0.4355275404010492,
+      "grad_norm": 1.1302298307418823,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 10295
+    },
+    {
+      "epoch": 0.43556984516456554,
+      "grad_norm": 0.2590234577655792,
+      "learning_rate": 0.001,
+      "loss": 1.8019,
+      "step": 10296
+    },
+    {
+      "epoch": 0.4356121499280819,
+      "grad_norm": 0.14725469052791595,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 10297
+    },
+    {
+      "epoch": 0.43565445469159825,
+      "grad_norm": 3.516937255859375,
+      "learning_rate": 0.001,
+      "loss": 1.8395,
+      "step": 10298
+    },
+    {
+      "epoch": 0.43569675945511466,
+      "grad_norm": 0.1837252527475357,
+      "learning_rate": 0.001,
+      "loss": 2.3293,
+      "step": 10299
+    },
+    {
+      "epoch": 0.435739064218631,
+      "grad_norm": 0.15408559143543243,
+      "learning_rate": 0.001,
+      "loss": 2.7421,
+      "step": 10300
+    },
+    {
+      "epoch": 0.43578136898214737,
+      "grad_norm": 2.384270191192627,
+      "learning_rate": 0.001,
+      "loss": 2.6062,
+      "step": 10301
+    },
+    {
+      "epoch": 0.4358236737456638,
+      "grad_norm": 0.22565071284770966,
+      "learning_rate": 0.001,
+      "loss": 2.2583,
+      "step": 10302
+    },
+    {
+      "epoch": 0.43586597850918013,
+      "grad_norm": 0.19346702098846436,
+      "learning_rate": 0.001,
+      "loss": 2.36,
+      "step": 10303
+    },
+    {
+      "epoch": 0.4359082832726965,
+      "grad_norm": 0.15725278854370117,
+      "learning_rate": 0.001,
+      "loss": 1.6664,
+      "step": 10304
+    },
+    {
+      "epoch": 0.4359505880362129,
+      "grad_norm": 0.19711074233055115,
+      "learning_rate": 0.001,
+      "loss": 1.736,
+      "step": 10305
+    },
+    {
+      "epoch": 0.43599289279972925,
+      "grad_norm": 0.2101636379957199,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 10306
+    },
+    {
+      "epoch": 0.4360351975632456,
+      "grad_norm": 0.1750393956899643,
+      "learning_rate": 0.001,
+      "loss": 1.6158,
+      "step": 10307
+    },
+    {
+      "epoch": 0.436077502326762,
+      "grad_norm": 0.1906735897064209,
+      "learning_rate": 0.001,
+      "loss": 2.4902,
+      "step": 10308
+    },
+    {
+      "epoch": 0.43611980709027837,
+      "grad_norm": 1.6329114437103271,
+      "learning_rate": 0.001,
+      "loss": 2.1277,
+      "step": 10309
+    },
+    {
+      "epoch": 0.4361621118537947,
+      "grad_norm": 0.9235885739326477,
+      "learning_rate": 0.001,
+      "loss": 1.4367,
+      "step": 10310
+    },
+    {
+      "epoch": 0.43620441661731113,
+      "grad_norm": 0.14719220995903015,
+      "learning_rate": 0.001,
+      "loss": 2.4318,
+      "step": 10311
+    },
+    {
+      "epoch": 0.4362467213808275,
+      "grad_norm": 0.2758025825023651,
+      "learning_rate": 0.001,
+      "loss": 1.9603,
+      "step": 10312
+    },
+    {
+      "epoch": 0.43628902614434384,
+      "grad_norm": 0.25476765632629395,
+      "learning_rate": 0.001,
+      "loss": 4.3574,
+      "step": 10313
+    },
+    {
+      "epoch": 0.43633133090786025,
+      "grad_norm": 0.180466890335083,
+      "learning_rate": 0.001,
+      "loss": 2.5561,
+      "step": 10314
+    },
+    {
+      "epoch": 0.4363736356713766,
+      "grad_norm": 0.19408752024173737,
+      "learning_rate": 0.001,
+      "loss": 2.5026,
+      "step": 10315
+    },
+    {
+      "epoch": 0.43641594043489296,
+      "grad_norm": 0.25457096099853516,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 10316
+    },
+    {
+      "epoch": 0.43645824519840937,
+      "grad_norm": 0.3525673747062683,
+      "learning_rate": 0.001,
+      "loss": 2.2683,
+      "step": 10317
+    },
+    {
+      "epoch": 0.4365005499619257,
+      "grad_norm": 0.1872408390045166,
+      "learning_rate": 0.001,
+      "loss": 2.0554,
+      "step": 10318
+    },
+    {
+      "epoch": 0.4365428547254421,
+      "grad_norm": 0.31182733178138733,
+      "learning_rate": 0.001,
+      "loss": 2.6635,
+      "step": 10319
+    },
+    {
+      "epoch": 0.43658515948895843,
+      "grad_norm": 0.18361353874206543,
+      "learning_rate": 0.001,
+      "loss": 3.2744,
+      "step": 10320
+    },
+    {
+      "epoch": 0.43662746425247484,
+      "grad_norm": 0.20506177842617035,
+      "learning_rate": 0.001,
+      "loss": 2.086,
+      "step": 10321
+    },
+    {
+      "epoch": 0.4366697690159912,
+      "grad_norm": 0.1834511160850525,
+      "learning_rate": 0.001,
+      "loss": 2.1743,
+      "step": 10322
+    },
+    {
+      "epoch": 0.43671207377950755,
+      "grad_norm": 1.0289106369018555,
+      "learning_rate": 0.001,
+      "loss": 2.5491,
+      "step": 10323
+    },
+    {
+      "epoch": 0.43675437854302396,
+      "grad_norm": 1.9735060930252075,
+      "learning_rate": 0.001,
+      "loss": 2.4487,
+      "step": 10324
+    },
+    {
+      "epoch": 0.4367966833065403,
+      "grad_norm": 0.17474709451198578,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 10325
+    },
+    {
+      "epoch": 0.43683898807005667,
+      "grad_norm": 0.17368879914283752,
+      "learning_rate": 0.001,
+      "loss": 2.0555,
+      "step": 10326
+    },
+    {
+      "epoch": 0.4368812928335731,
+      "grad_norm": 0.2503129839897156,
+      "learning_rate": 0.001,
+      "loss": 2.6008,
+      "step": 10327
+    },
+    {
+      "epoch": 0.43692359759708943,
+      "grad_norm": 0.16766273975372314,
+      "learning_rate": 0.001,
+      "loss": 2.033,
+      "step": 10328
+    },
+    {
+      "epoch": 0.4369659023606058,
+      "grad_norm": 0.19881922006607056,
+      "learning_rate": 0.001,
+      "loss": 1.9835,
+      "step": 10329
+    },
+    {
+      "epoch": 0.4370082071241222,
+      "grad_norm": 2.48319149017334,
+      "learning_rate": 0.001,
+      "loss": 2.4763,
+      "step": 10330
+    },
+    {
+      "epoch": 0.43705051188763855,
+      "grad_norm": 0.19117386639118195,
+      "learning_rate": 0.001,
+      "loss": 1.7097,
+      "step": 10331
+    },
+    {
+      "epoch": 0.4370928166511549,
+      "grad_norm": 0.2317335158586502,
+      "learning_rate": 0.001,
+      "loss": 3.0248,
+      "step": 10332
+    },
+    {
+      "epoch": 0.4371351214146713,
+      "grad_norm": 3.407411813735962,
+      "learning_rate": 0.001,
+      "loss": 2.5108,
+      "step": 10333
+    },
+    {
+      "epoch": 0.43717742617818767,
+      "grad_norm": 0.32746651768684387,
+      "learning_rate": 0.001,
+      "loss": 2.6375,
+      "step": 10334
+    },
+    {
+      "epoch": 0.437219730941704,
+      "grad_norm": 0.4482516348361969,
+      "learning_rate": 0.001,
+      "loss": 1.3975,
+      "step": 10335
+    },
+    {
+      "epoch": 0.43726203570522043,
+      "grad_norm": 0.20397043228149414,
+      "learning_rate": 0.001,
+      "loss": 1.8973,
+      "step": 10336
+    },
+    {
+      "epoch": 0.4373043404687368,
+      "grad_norm": 0.20856736600399017,
+      "learning_rate": 0.001,
+      "loss": 2.7255,
+      "step": 10337
+    },
+    {
+      "epoch": 0.43734664523225314,
+      "grad_norm": 0.4428465664386749,
+      "learning_rate": 0.001,
+      "loss": 2.4357,
+      "step": 10338
+    },
+    {
+      "epoch": 0.43738894999576955,
+      "grad_norm": 0.2584536075592041,
+      "learning_rate": 0.001,
+      "loss": 3.332,
+      "step": 10339
+    },
+    {
+      "epoch": 0.4374312547592859,
+      "grad_norm": 0.2311519831418991,
+      "learning_rate": 0.001,
+      "loss": 3.5032,
+      "step": 10340
+    },
+    {
+      "epoch": 0.43747355952280226,
+      "grad_norm": 0.25852352380752563,
+      "learning_rate": 0.001,
+      "loss": 2.6795,
+      "step": 10341
+    },
+    {
+      "epoch": 0.4375158642863186,
+      "grad_norm": 0.1977977156639099,
+      "learning_rate": 0.001,
+      "loss": 2.0136,
+      "step": 10342
+    },
+    {
+      "epoch": 0.437558169049835,
+      "grad_norm": 0.20671911537647247,
+      "learning_rate": 0.001,
+      "loss": 2.0685,
+      "step": 10343
+    },
+    {
+      "epoch": 0.4376004738133514,
+      "grad_norm": 0.22240790724754333,
+      "learning_rate": 0.001,
+      "loss": 2.3391,
+      "step": 10344
+    },
+    {
+      "epoch": 0.43764277857686773,
+      "grad_norm": 0.9896284937858582,
+      "learning_rate": 0.001,
+      "loss": 2.4624,
+      "step": 10345
+    },
+    {
+      "epoch": 0.43768508334038414,
+      "grad_norm": 0.18324188888072968,
+      "learning_rate": 0.001,
+      "loss": 2.2438,
+      "step": 10346
+    },
+    {
+      "epoch": 0.4377273881039005,
+      "grad_norm": 0.20104867219924927,
+      "learning_rate": 0.001,
+      "loss": 2.8267,
+      "step": 10347
+    },
+    {
+      "epoch": 0.43776969286741685,
+      "grad_norm": 0.26405319571495056,
+      "learning_rate": 0.001,
+      "loss": 2.1973,
+      "step": 10348
+    },
+    {
+      "epoch": 0.43781199763093326,
+      "grad_norm": 0.5061947107315063,
+      "learning_rate": 0.001,
+      "loss": 2.5415,
+      "step": 10349
+    },
+    {
+      "epoch": 0.4378543023944496,
+      "grad_norm": 0.46758270263671875,
+      "learning_rate": 0.001,
+      "loss": 3.0251,
+      "step": 10350
+    },
+    {
+      "epoch": 0.43789660715796597,
+      "grad_norm": 1.8289541006088257,
+      "learning_rate": 0.001,
+      "loss": 2.4671,
+      "step": 10351
+    },
+    {
+      "epoch": 0.4379389119214824,
+      "grad_norm": 2.7915992736816406,
+      "learning_rate": 0.001,
+      "loss": 2.0793,
+      "step": 10352
+    },
+    {
+      "epoch": 0.43798121668499873,
+      "grad_norm": 0.1846030205488205,
+      "learning_rate": 0.001,
+      "loss": 1.7337,
+      "step": 10353
+    },
+    {
+      "epoch": 0.4380235214485151,
+      "grad_norm": 0.8970298171043396,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 10354
+    },
+    {
+      "epoch": 0.4380658262120315,
+      "grad_norm": 0.6228232979774475,
+      "learning_rate": 0.001,
+      "loss": 2.245,
+      "step": 10355
+    },
+    {
+      "epoch": 0.43810813097554785,
+      "grad_norm": 0.18022607266902924,
+      "learning_rate": 0.001,
+      "loss": 2.4163,
+      "step": 10356
+    },
+    {
+      "epoch": 0.4381504357390642,
+      "grad_norm": 0.3440026640892029,
+      "learning_rate": 0.001,
+      "loss": 2.0781,
+      "step": 10357
+    },
+    {
+      "epoch": 0.4381927405025806,
+      "grad_norm": 0.26871904730796814,
+      "learning_rate": 0.001,
+      "loss": 3.507,
+      "step": 10358
+    },
+    {
+      "epoch": 0.43823504526609697,
+      "grad_norm": 0.1928730607032776,
+      "learning_rate": 0.001,
+      "loss": 1.6934,
+      "step": 10359
+    },
+    {
+      "epoch": 0.4382773500296133,
+      "grad_norm": 0.23119626939296722,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 10360
+    },
+    {
+      "epoch": 0.43831965479312973,
+      "grad_norm": 0.20869280397891998,
+      "learning_rate": 0.001,
+      "loss": 2.3586,
+      "step": 10361
+    },
+    {
+      "epoch": 0.4383619595566461,
+      "grad_norm": 0.26448631286621094,
+      "learning_rate": 0.001,
+      "loss": 2.1414,
+      "step": 10362
+    },
+    {
+      "epoch": 0.43840426432016244,
+      "grad_norm": 0.2913392186164856,
+      "learning_rate": 0.001,
+      "loss": 2.2654,
+      "step": 10363
+    },
+    {
+      "epoch": 0.43844656908367885,
+      "grad_norm": 6.033486843109131,
+      "learning_rate": 0.001,
+      "loss": 2.6725,
+      "step": 10364
+    },
+    {
+      "epoch": 0.4384888738471952,
+      "grad_norm": 0.8384542465209961,
+      "learning_rate": 0.001,
+      "loss": 1.926,
+      "step": 10365
+    },
+    {
+      "epoch": 0.43853117861071156,
+      "grad_norm": 0.20631474256515503,
+      "learning_rate": 0.001,
+      "loss": 1.8508,
+      "step": 10366
+    },
+    {
+      "epoch": 0.4385734833742279,
+      "grad_norm": 0.3741599917411804,
+      "learning_rate": 0.001,
+      "loss": 2.8569,
+      "step": 10367
+    },
+    {
+      "epoch": 0.4386157881377443,
+      "grad_norm": 0.19838982820510864,
+      "learning_rate": 0.001,
+      "loss": 2.0593,
+      "step": 10368
+    },
+    {
+      "epoch": 0.4386580929012607,
+      "grad_norm": 0.16171355545520782,
+      "learning_rate": 0.001,
+      "loss": 1.7758,
+      "step": 10369
+    },
+    {
+      "epoch": 0.43870039766477703,
+      "grad_norm": 2.109577178955078,
+      "learning_rate": 0.001,
+      "loss": 2.4595,
+      "step": 10370
+    },
+    {
+      "epoch": 0.43874270242829344,
+      "grad_norm": 0.23121996223926544,
+      "learning_rate": 0.001,
+      "loss": 2.1405,
+      "step": 10371
+    },
+    {
+      "epoch": 0.4387850071918098,
+      "grad_norm": 0.19020533561706543,
+      "learning_rate": 0.001,
+      "loss": 2.0821,
+      "step": 10372
+    },
+    {
+      "epoch": 0.43882731195532615,
+      "grad_norm": 0.3466572165489197,
+      "learning_rate": 0.001,
+      "loss": 2.352,
+      "step": 10373
+    },
+    {
+      "epoch": 0.43886961671884256,
+      "grad_norm": 0.22979247570037842,
+      "learning_rate": 0.001,
+      "loss": 1.8425,
+      "step": 10374
+    },
+    {
+      "epoch": 0.4389119214823589,
+      "grad_norm": 0.2191963642835617,
+      "learning_rate": 0.001,
+      "loss": 1.9217,
+      "step": 10375
+    },
+    {
+      "epoch": 0.43895422624587527,
+      "grad_norm": 0.29709628224372864,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 10376
+    },
+    {
+      "epoch": 0.4389965310093917,
+      "grad_norm": 1.1428967714309692,
+      "learning_rate": 0.001,
+      "loss": 2.3209,
+      "step": 10377
+    },
+    {
+      "epoch": 0.43903883577290803,
+      "grad_norm": 0.21593493223190308,
+      "learning_rate": 0.001,
+      "loss": 2.2734,
+      "step": 10378
+    },
+    {
+      "epoch": 0.4390811405364244,
+      "grad_norm": 0.2102469801902771,
+      "learning_rate": 0.001,
+      "loss": 2.5221,
+      "step": 10379
+    },
+    {
+      "epoch": 0.4391234452999408,
+      "grad_norm": 0.19701038300991058,
+      "learning_rate": 0.001,
+      "loss": 2.7192,
+      "step": 10380
+    },
+    {
+      "epoch": 0.43916575006345715,
+      "grad_norm": 0.6864662170410156,
+      "learning_rate": 0.001,
+      "loss": 2.4796,
+      "step": 10381
+    },
+    {
+      "epoch": 0.4392080548269735,
+      "grad_norm": 0.5476522445678711,
+      "learning_rate": 0.001,
+      "loss": 1.725,
+      "step": 10382
+    },
+    {
+      "epoch": 0.4392503595904899,
+      "grad_norm": 0.19268248975276947,
+      "learning_rate": 0.001,
+      "loss": 1.8153,
+      "step": 10383
+    },
+    {
+      "epoch": 0.43929266435400627,
+      "grad_norm": 0.40893223881721497,
+      "learning_rate": 0.001,
+      "loss": 2.7319,
+      "step": 10384
+    },
+    {
+      "epoch": 0.4393349691175226,
+      "grad_norm": 2.664677143096924,
+      "learning_rate": 0.001,
+      "loss": 2.4025,
+      "step": 10385
+    },
+    {
+      "epoch": 0.43937727388103903,
+      "grad_norm": 0.18634748458862305,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 10386
+    },
+    {
+      "epoch": 0.4394195786445554,
+      "grad_norm": 0.1849360466003418,
+      "learning_rate": 0.001,
+      "loss": 2.2008,
+      "step": 10387
+    },
+    {
+      "epoch": 0.43946188340807174,
+      "grad_norm": 0.15529468655586243,
+      "learning_rate": 0.001,
+      "loss": 2.3172,
+      "step": 10388
+    },
+    {
+      "epoch": 0.4395041881715881,
+      "grad_norm": 0.15574440360069275,
+      "learning_rate": 0.001,
+      "loss": 1.7872,
+      "step": 10389
+    },
+    {
+      "epoch": 0.4395464929351045,
+      "grad_norm": 0.24909093976020813,
+      "learning_rate": 0.001,
+      "loss": 2.3089,
+      "step": 10390
+    },
+    {
+      "epoch": 0.43958879769862086,
+      "grad_norm": 0.27436402440071106,
+      "learning_rate": 0.001,
+      "loss": 2.7429,
+      "step": 10391
+    },
+    {
+      "epoch": 0.4396311024621372,
+      "grad_norm": 0.6394477486610413,
+      "learning_rate": 0.001,
+      "loss": 3.1377,
+      "step": 10392
+    },
+    {
+      "epoch": 0.4396734072256536,
+      "grad_norm": 0.1768225133419037,
+      "learning_rate": 0.001,
+      "loss": 1.7332,
+      "step": 10393
+    },
+    {
+      "epoch": 0.43971571198917,
+      "grad_norm": 0.18364182114601135,
+      "learning_rate": 0.001,
+      "loss": 3.3755,
+      "step": 10394
+    },
+    {
+      "epoch": 0.43975801675268633,
+      "grad_norm": 0.1930709332227707,
+      "learning_rate": 0.001,
+      "loss": 2.2478,
+      "step": 10395
+    },
+    {
+      "epoch": 0.43980032151620274,
+      "grad_norm": 4.500080585479736,
+      "learning_rate": 0.001,
+      "loss": 1.6433,
+      "step": 10396
+    },
+    {
+      "epoch": 0.4398426262797191,
+      "grad_norm": 0.20036698877811432,
+      "learning_rate": 0.001,
+      "loss": 2.2089,
+      "step": 10397
+    },
+    {
+      "epoch": 0.43988493104323545,
+      "grad_norm": 0.19765986502170563,
+      "learning_rate": 0.001,
+      "loss": 2.012,
+      "step": 10398
+    },
+    {
+      "epoch": 0.43992723580675186,
+      "grad_norm": 0.1463608741760254,
+      "learning_rate": 0.001,
+      "loss": 1.4149,
+      "step": 10399
+    },
+    {
+      "epoch": 0.4399695405702682,
+      "grad_norm": 6.610461235046387,
+      "learning_rate": 0.001,
+      "loss": 2.9959,
+      "step": 10400
+    },
+    {
+      "epoch": 0.44001184533378457,
+      "grad_norm": 0.23446570336818695,
+      "learning_rate": 0.001,
+      "loss": 2.249,
+      "step": 10401
+    },
+    {
+      "epoch": 0.440054150097301,
+      "grad_norm": 1.4213615655899048,
+      "learning_rate": 0.001,
+      "loss": 2.6993,
+      "step": 10402
+    },
+    {
+      "epoch": 0.44009645486081733,
+      "grad_norm": 0.6602863073348999,
+      "learning_rate": 0.001,
+      "loss": 1.8553,
+      "step": 10403
+    },
+    {
+      "epoch": 0.4401387596243337,
+      "grad_norm": 0.2995677590370178,
+      "learning_rate": 0.001,
+      "loss": 3.3502,
+      "step": 10404
+    },
+    {
+      "epoch": 0.4401810643878501,
+      "grad_norm": 0.1976313292980194,
+      "learning_rate": 0.001,
+      "loss": 2.0083,
+      "step": 10405
+    },
+    {
+      "epoch": 0.44022336915136645,
+      "grad_norm": 0.2297874391078949,
+      "learning_rate": 0.001,
+      "loss": 3.5225,
+      "step": 10406
+    },
+    {
+      "epoch": 0.4402656739148828,
+      "grad_norm": 0.24849973618984222,
+      "learning_rate": 0.001,
+      "loss": 2.3344,
+      "step": 10407
+    },
+    {
+      "epoch": 0.4403079786783992,
+      "grad_norm": 0.15684276819229126,
+      "learning_rate": 0.001,
+      "loss": 3.1674,
+      "step": 10408
+    },
+    {
+      "epoch": 0.44035028344191557,
+      "grad_norm": 0.24899160861968994,
+      "learning_rate": 0.001,
+      "loss": 1.7431,
+      "step": 10409
+    },
+    {
+      "epoch": 0.4403925882054319,
+      "grad_norm": 0.16840027272701263,
+      "learning_rate": 0.001,
+      "loss": 2.1913,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4404348929689483,
+      "grad_norm": 0.18736068904399872,
+      "learning_rate": 0.001,
+      "loss": 2.1678,
+      "step": 10411
+    },
+    {
+      "epoch": 0.4404771977324647,
+      "grad_norm": 0.19807054102420807,
+      "learning_rate": 0.001,
+      "loss": 2.3152,
+      "step": 10412
+    },
+    {
+      "epoch": 0.44051950249598104,
+      "grad_norm": 0.7695265412330627,
+      "learning_rate": 0.001,
+      "loss": 1.9602,
+      "step": 10413
+    },
+    {
+      "epoch": 0.4405618072594974,
+      "grad_norm": 0.19554787874221802,
+      "learning_rate": 0.001,
+      "loss": 1.9298,
+      "step": 10414
+    },
+    {
+      "epoch": 0.4406041120230138,
+      "grad_norm": 0.22768321633338928,
+      "learning_rate": 0.001,
+      "loss": 1.8096,
+      "step": 10415
+    },
+    {
+      "epoch": 0.44064641678653016,
+      "grad_norm": 0.16684657335281372,
+      "learning_rate": 0.001,
+      "loss": 2.6149,
+      "step": 10416
+    },
+    {
+      "epoch": 0.4406887215500465,
+      "grad_norm": 6.711855888366699,
+      "learning_rate": 0.001,
+      "loss": 2.506,
+      "step": 10417
+    },
+    {
+      "epoch": 0.4407310263135629,
+      "grad_norm": 0.199007049202919,
+      "learning_rate": 0.001,
+      "loss": 2.6235,
+      "step": 10418
+    },
+    {
+      "epoch": 0.4407733310770793,
+      "grad_norm": 4.696216106414795,
+      "learning_rate": 0.001,
+      "loss": 2.0736,
+      "step": 10419
+    },
+    {
+      "epoch": 0.44081563584059563,
+      "grad_norm": 0.18525268137454987,
+      "learning_rate": 0.001,
+      "loss": 1.6978,
+      "step": 10420
+    },
+    {
+      "epoch": 0.44085794060411204,
+      "grad_norm": 0.22784662246704102,
+      "learning_rate": 0.001,
+      "loss": 2.0242,
+      "step": 10421
+    },
+    {
+      "epoch": 0.4409002453676284,
+      "grad_norm": 0.49683111906051636,
+      "learning_rate": 0.001,
+      "loss": 3.07,
+      "step": 10422
+    },
+    {
+      "epoch": 0.44094255013114475,
+      "grad_norm": 0.29818785190582275,
+      "learning_rate": 0.001,
+      "loss": 2.0421,
+      "step": 10423
+    },
+    {
+      "epoch": 0.44098485489466116,
+      "grad_norm": 0.30528533458709717,
+      "learning_rate": 0.001,
+      "loss": 2.3055,
+      "step": 10424
+    },
+    {
+      "epoch": 0.4410271596581775,
+      "grad_norm": 0.17285211384296417,
+      "learning_rate": 0.001,
+      "loss": 2.8575,
+      "step": 10425
+    },
+    {
+      "epoch": 0.44106946442169387,
+      "grad_norm": 0.2074311375617981,
+      "learning_rate": 0.001,
+      "loss": 2.9005,
+      "step": 10426
+    },
+    {
+      "epoch": 0.4411117691852103,
+      "grad_norm": 13.308959007263184,
+      "learning_rate": 0.001,
+      "loss": 2.1316,
+      "step": 10427
+    },
+    {
+      "epoch": 0.44115407394872663,
+      "grad_norm": 8.290491104125977,
+      "learning_rate": 0.001,
+      "loss": 3.0784,
+      "step": 10428
+    },
+    {
+      "epoch": 0.441196378712243,
+      "grad_norm": 6.2142252922058105,
+      "learning_rate": 0.001,
+      "loss": 2.8017,
+      "step": 10429
+    },
+    {
+      "epoch": 0.4412386834757594,
+      "grad_norm": 0.41313034296035767,
+      "learning_rate": 0.001,
+      "loss": 2.1858,
+      "step": 10430
+    },
+    {
+      "epoch": 0.44128098823927575,
+      "grad_norm": 13.561888694763184,
+      "learning_rate": 0.001,
+      "loss": 2.6311,
+      "step": 10431
+    },
+    {
+      "epoch": 0.4413232930027921,
+      "grad_norm": 0.19502195715904236,
+      "learning_rate": 0.001,
+      "loss": 2.5145,
+      "step": 10432
+    },
+    {
+      "epoch": 0.44136559776630846,
+      "grad_norm": 0.2449101358652115,
+      "learning_rate": 0.001,
+      "loss": 2.9577,
+      "step": 10433
+    },
+    {
+      "epoch": 0.44140790252982487,
+      "grad_norm": 0.4394688606262207,
+      "learning_rate": 0.001,
+      "loss": 2.156,
+      "step": 10434
+    },
+    {
+      "epoch": 0.4414502072933412,
+      "grad_norm": 0.9801822900772095,
+      "learning_rate": 0.001,
+      "loss": 2.5387,
+      "step": 10435
+    },
+    {
+      "epoch": 0.4414925120568576,
+      "grad_norm": 0.2750179171562195,
+      "learning_rate": 0.001,
+      "loss": 2.1277,
+      "step": 10436
+    },
+    {
+      "epoch": 0.441534816820374,
+      "grad_norm": 0.17588399350643158,
+      "learning_rate": 0.001,
+      "loss": 3.0843,
+      "step": 10437
+    },
+    {
+      "epoch": 0.44157712158389034,
+      "grad_norm": 0.18611468374729156,
+      "learning_rate": 0.001,
+      "loss": 2.0977,
+      "step": 10438
+    },
+    {
+      "epoch": 0.4416194263474067,
+      "grad_norm": 0.4915761649608612,
+      "learning_rate": 0.001,
+      "loss": 2.4029,
+      "step": 10439
+    },
+    {
+      "epoch": 0.4416617311109231,
+      "grad_norm": 0.2895006835460663,
+      "learning_rate": 0.001,
+      "loss": 2.6397,
+      "step": 10440
+    },
+    {
+      "epoch": 0.44170403587443946,
+      "grad_norm": 0.19709734618663788,
+      "learning_rate": 0.001,
+      "loss": 2.1713,
+      "step": 10441
+    },
+    {
+      "epoch": 0.4417463406379558,
+      "grad_norm": 0.7510543465614319,
+      "learning_rate": 0.001,
+      "loss": 2.5878,
+      "step": 10442
+    },
+    {
+      "epoch": 0.4417886454014722,
+      "grad_norm": 0.18231795728206635,
+      "learning_rate": 0.001,
+      "loss": 1.8023,
+      "step": 10443
+    },
+    {
+      "epoch": 0.4418309501649886,
+      "grad_norm": 0.18479089438915253,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 10444
+    },
+    {
+      "epoch": 0.44187325492850493,
+      "grad_norm": 0.35781130194664,
+      "learning_rate": 0.001,
+      "loss": 2.4331,
+      "step": 10445
+    },
+    {
+      "epoch": 0.44191555969202134,
+      "grad_norm": 0.23527033627033234,
+      "learning_rate": 0.001,
+      "loss": 1.5157,
+      "step": 10446
+    },
+    {
+      "epoch": 0.4419578644555377,
+      "grad_norm": 0.21979889273643494,
+      "learning_rate": 0.001,
+      "loss": 2.0873,
+      "step": 10447
+    },
+    {
+      "epoch": 0.44200016921905405,
+      "grad_norm": 0.2087160348892212,
+      "learning_rate": 0.001,
+      "loss": 2.237,
+      "step": 10448
+    },
+    {
+      "epoch": 0.44204247398257046,
+      "grad_norm": 0.30770057439804077,
+      "learning_rate": 0.001,
+      "loss": 2.6998,
+      "step": 10449
+    },
+    {
+      "epoch": 0.4420847787460868,
+      "grad_norm": 0.27149367332458496,
+      "learning_rate": 0.001,
+      "loss": 2.438,
+      "step": 10450
+    },
+    {
+      "epoch": 0.44212708350960317,
+      "grad_norm": 0.18169645965099335,
+      "learning_rate": 0.001,
+      "loss": 1.4572,
+      "step": 10451
+    },
+    {
+      "epoch": 0.4421693882731196,
+      "grad_norm": 0.1930762231349945,
+      "learning_rate": 0.001,
+      "loss": 1.9931,
+      "step": 10452
+    },
+    {
+      "epoch": 0.44221169303663593,
+      "grad_norm": 1.637730598449707,
+      "learning_rate": 0.001,
+      "loss": 3.6527,
+      "step": 10453
+    },
+    {
+      "epoch": 0.4422539978001523,
+      "grad_norm": 0.19526685774326324,
+      "learning_rate": 0.001,
+      "loss": 2.715,
+      "step": 10454
+    },
+    {
+      "epoch": 0.44229630256366864,
+      "grad_norm": 0.25344404578208923,
+      "learning_rate": 0.001,
+      "loss": 2.2599,
+      "step": 10455
+    },
+    {
+      "epoch": 0.44233860732718505,
+      "grad_norm": 0.4373885989189148,
+      "learning_rate": 0.001,
+      "loss": 2.1229,
+      "step": 10456
+    },
+    {
+      "epoch": 0.4423809120907014,
+      "grad_norm": 0.49542349576950073,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 10457
+    },
+    {
+      "epoch": 0.44242321685421776,
+      "grad_norm": 0.21734249591827393,
+      "learning_rate": 0.001,
+      "loss": 2.5501,
+      "step": 10458
+    },
+    {
+      "epoch": 0.44246552161773417,
+      "grad_norm": 0.37125974893569946,
+      "learning_rate": 0.001,
+      "loss": 2.2032,
+      "step": 10459
+    },
+    {
+      "epoch": 0.4425078263812505,
+      "grad_norm": 1.202412724494934,
+      "learning_rate": 0.001,
+      "loss": 2.6156,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4425501311447669,
+      "grad_norm": 0.2056283950805664,
+      "learning_rate": 0.001,
+      "loss": 2.3052,
+      "step": 10461
+    },
+    {
+      "epoch": 0.4425924359082833,
+      "grad_norm": 0.24419522285461426,
+      "learning_rate": 0.001,
+      "loss": 1.9584,
+      "step": 10462
+    },
+    {
+      "epoch": 0.44263474067179964,
+      "grad_norm": 3.809471845626831,
+      "learning_rate": 0.001,
+      "loss": 2.0806,
+      "step": 10463
+    },
+    {
+      "epoch": 0.442677045435316,
+      "grad_norm": 0.1795853078365326,
+      "learning_rate": 0.001,
+      "loss": 1.8472,
+      "step": 10464
+    },
+    {
+      "epoch": 0.4427193501988324,
+      "grad_norm": 0.21403010189533234,
+      "learning_rate": 0.001,
+      "loss": 2.5621,
+      "step": 10465
+    },
+    {
+      "epoch": 0.44276165496234876,
+      "grad_norm": 0.23772773146629333,
+      "learning_rate": 0.001,
+      "loss": 4.0021,
+      "step": 10466
+    },
+    {
+      "epoch": 0.4428039597258651,
+      "grad_norm": 0.1654462069272995,
+      "learning_rate": 0.001,
+      "loss": 2.3355,
+      "step": 10467
+    },
+    {
+      "epoch": 0.4428462644893815,
+      "grad_norm": 0.16145402193069458,
+      "learning_rate": 0.001,
+      "loss": 3.1889,
+      "step": 10468
+    },
+    {
+      "epoch": 0.4428885692528979,
+      "grad_norm": 0.27223458886146545,
+      "learning_rate": 0.001,
+      "loss": 1.7634,
+      "step": 10469
+    },
+    {
+      "epoch": 0.44293087401641423,
+      "grad_norm": 0.1729728877544403,
+      "learning_rate": 0.001,
+      "loss": 1.2275,
+      "step": 10470
+    },
+    {
+      "epoch": 0.44297317877993064,
+      "grad_norm": 0.17519982159137726,
+      "learning_rate": 0.001,
+      "loss": 2.0601,
+      "step": 10471
+    },
+    {
+      "epoch": 0.443015483543447,
+      "grad_norm": 0.2915792167186737,
+      "learning_rate": 0.001,
+      "loss": 3.0095,
+      "step": 10472
+    },
+    {
+      "epoch": 0.44305778830696335,
+      "grad_norm": 0.212400883436203,
+      "learning_rate": 0.001,
+      "loss": 2.7761,
+      "step": 10473
+    },
+    {
+      "epoch": 0.44310009307047976,
+      "grad_norm": 1.2644349336624146,
+      "learning_rate": 0.001,
+      "loss": 2.631,
+      "step": 10474
+    },
+    {
+      "epoch": 0.4431423978339961,
+      "grad_norm": 0.2543438673019409,
+      "learning_rate": 0.001,
+      "loss": 2.9616,
+      "step": 10475
+    },
+    {
+      "epoch": 0.44318470259751247,
+      "grad_norm": 0.5095729827880859,
+      "learning_rate": 0.001,
+      "loss": 2.9212,
+      "step": 10476
+    },
+    {
+      "epoch": 0.4432270073610288,
+      "grad_norm": 0.22304564714431763,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 10477
+    },
+    {
+      "epoch": 0.44326931212454523,
+      "grad_norm": 0.2589484751224518,
+      "learning_rate": 0.001,
+      "loss": 1.8629,
+      "step": 10478
+    },
+    {
+      "epoch": 0.4433116168880616,
+      "grad_norm": 0.171627938747406,
+      "learning_rate": 0.001,
+      "loss": 2.1287,
+      "step": 10479
+    },
+    {
+      "epoch": 0.44335392165157794,
+      "grad_norm": 0.18898135423660278,
+      "learning_rate": 0.001,
+      "loss": 1.9837,
+      "step": 10480
+    },
+    {
+      "epoch": 0.44339622641509435,
+      "grad_norm": 0.17940890789031982,
+      "learning_rate": 0.001,
+      "loss": 1.8571,
+      "step": 10481
+    },
+    {
+      "epoch": 0.4434385311786107,
+      "grad_norm": 0.3940030634403229,
+      "learning_rate": 0.001,
+      "loss": 2.5347,
+      "step": 10482
+    },
+    {
+      "epoch": 0.44348083594212706,
+      "grad_norm": 0.16100460290908813,
+      "learning_rate": 0.001,
+      "loss": 3.2192,
+      "step": 10483
+    },
+    {
+      "epoch": 0.44352314070564347,
+      "grad_norm": 0.3118045926094055,
+      "learning_rate": 0.001,
+      "loss": 2.7277,
+      "step": 10484
+    },
+    {
+      "epoch": 0.4435654454691598,
+      "grad_norm": 0.38664117455482483,
+      "learning_rate": 0.001,
+      "loss": 3.5834,
+      "step": 10485
+    },
+    {
+      "epoch": 0.4436077502326762,
+      "grad_norm": 0.15630090236663818,
+      "learning_rate": 0.001,
+      "loss": 1.5872,
+      "step": 10486
+    },
+    {
+      "epoch": 0.4436500549961926,
+      "grad_norm": 0.1633961796760559,
+      "learning_rate": 0.001,
+      "loss": 2.2324,
+      "step": 10487
+    },
+    {
+      "epoch": 0.44369235975970894,
+      "grad_norm": 0.2306484431028366,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 10488
+    },
+    {
+      "epoch": 0.4437346645232253,
+      "grad_norm": 0.3159201741218567,
+      "learning_rate": 0.001,
+      "loss": 2.2158,
+      "step": 10489
+    },
+    {
+      "epoch": 0.4437769692867417,
+      "grad_norm": 0.1387176662683487,
+      "learning_rate": 0.001,
+      "loss": 1.596,
+      "step": 10490
+    },
+    {
+      "epoch": 0.44381927405025806,
+      "grad_norm": 0.17673452198505402,
+      "learning_rate": 0.001,
+      "loss": 2.1328,
+      "step": 10491
+    },
+    {
+      "epoch": 0.4438615788137744,
+      "grad_norm": 0.22988873720169067,
+      "learning_rate": 0.001,
+      "loss": 1.8124,
+      "step": 10492
+    },
+    {
+      "epoch": 0.4439038835772908,
+      "grad_norm": 0.1935664266347885,
+      "learning_rate": 0.001,
+      "loss": 1.9753,
+      "step": 10493
+    },
+    {
+      "epoch": 0.4439461883408072,
+      "grad_norm": 0.17069995403289795,
+      "learning_rate": 0.001,
+      "loss": 2.0083,
+      "step": 10494
+    },
+    {
+      "epoch": 0.44398849310432353,
+      "grad_norm": 1.7595449686050415,
+      "learning_rate": 0.001,
+      "loss": 2.3677,
+      "step": 10495
+    },
+    {
+      "epoch": 0.44403079786783994,
+      "grad_norm": 0.2332594096660614,
+      "learning_rate": 0.001,
+      "loss": 1.9425,
+      "step": 10496
+    },
+    {
+      "epoch": 0.4440731026313563,
+      "grad_norm": 3.590059757232666,
+      "learning_rate": 0.001,
+      "loss": 2.2229,
+      "step": 10497
+    },
+    {
+      "epoch": 0.44411540739487265,
+      "grad_norm": 0.1739518791437149,
+      "learning_rate": 0.001,
+      "loss": 1.8198,
+      "step": 10498
+    },
+    {
+      "epoch": 0.44415771215838906,
+      "grad_norm": 0.18792513012886047,
+      "learning_rate": 0.001,
+      "loss": 1.8256,
+      "step": 10499
+    },
+    {
+      "epoch": 0.4442000169219054,
+      "grad_norm": 0.16739851236343384,
+      "learning_rate": 0.001,
+      "loss": 1.96,
+      "step": 10500
+    },
+    {
+      "epoch": 0.44424232168542177,
+      "grad_norm": 0.17911168932914734,
+      "learning_rate": 0.001,
+      "loss": 2.0014,
+      "step": 10501
+    },
+    {
+      "epoch": 0.4442846264489381,
+      "grad_norm": 0.2152254283428192,
+      "learning_rate": 0.001,
+      "loss": 2.1837,
+      "step": 10502
+    },
+    {
+      "epoch": 0.44432693121245453,
+      "grad_norm": 0.16652332246303558,
+      "learning_rate": 0.001,
+      "loss": 1.8485,
+      "step": 10503
+    },
+    {
+      "epoch": 0.4443692359759709,
+      "grad_norm": 0.1486539989709854,
+      "learning_rate": 0.001,
+      "loss": 1.8012,
+      "step": 10504
+    },
+    {
+      "epoch": 0.44441154073948724,
+      "grad_norm": 0.4087856113910675,
+      "learning_rate": 0.001,
+      "loss": 1.6707,
+      "step": 10505
+    },
+    {
+      "epoch": 0.44445384550300365,
+      "grad_norm": 0.20888468623161316,
+      "learning_rate": 0.001,
+      "loss": 2.1961,
+      "step": 10506
+    },
+    {
+      "epoch": 0.44449615026652,
+      "grad_norm": 0.5997106432914734,
+      "learning_rate": 0.001,
+      "loss": 2.2202,
+      "step": 10507
+    },
+    {
+      "epoch": 0.44453845503003636,
+      "grad_norm": 3.278139591217041,
+      "learning_rate": 0.001,
+      "loss": 2.1766,
+      "step": 10508
+    },
+    {
+      "epoch": 0.44458075979355277,
+      "grad_norm": 0.16290752589702606,
+      "learning_rate": 0.001,
+      "loss": 3.9578,
+      "step": 10509
+    },
+    {
+      "epoch": 0.4446230645570691,
+      "grad_norm": 0.8776594996452332,
+      "learning_rate": 0.001,
+      "loss": 1.7571,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4446653693205855,
+      "grad_norm": 0.7716861367225647,
+      "learning_rate": 0.001,
+      "loss": 3.2192,
+      "step": 10511
+    },
+    {
+      "epoch": 0.4447076740841019,
+      "grad_norm": 1.821780800819397,
+      "learning_rate": 0.001,
+      "loss": 2.1203,
+      "step": 10512
+    },
+    {
+      "epoch": 0.44474997884761824,
+      "grad_norm": 4.94421911239624,
+      "learning_rate": 0.001,
+      "loss": 2.127,
+      "step": 10513
+    },
+    {
+      "epoch": 0.4447922836111346,
+      "grad_norm": 0.17742355167865753,
+      "learning_rate": 0.001,
+      "loss": 1.7438,
+      "step": 10514
+    },
+    {
+      "epoch": 0.444834588374651,
+      "grad_norm": 0.18554328382015228,
+      "learning_rate": 0.001,
+      "loss": 2.8967,
+      "step": 10515
+    },
+    {
+      "epoch": 0.44487689313816736,
+      "grad_norm": 0.1628502756357193,
+      "learning_rate": 0.001,
+      "loss": 1.8562,
+      "step": 10516
+    },
+    {
+      "epoch": 0.4449191979016837,
+      "grad_norm": 0.7208130955696106,
+      "learning_rate": 0.001,
+      "loss": 2.7975,
+      "step": 10517
+    },
+    {
+      "epoch": 0.4449615026652001,
+      "grad_norm": 0.6698981523513794,
+      "learning_rate": 0.001,
+      "loss": 2.7178,
+      "step": 10518
+    },
+    {
+      "epoch": 0.4450038074287165,
+      "grad_norm": 0.20229113101959229,
+      "learning_rate": 0.001,
+      "loss": 2.9919,
+      "step": 10519
+    },
+    {
+      "epoch": 0.44504611219223283,
+      "grad_norm": 0.1968151330947876,
+      "learning_rate": 0.001,
+      "loss": 3.2183,
+      "step": 10520
+    },
+    {
+      "epoch": 0.44508841695574924,
+      "grad_norm": 0.225248783826828,
+      "learning_rate": 0.001,
+      "loss": 3.2217,
+      "step": 10521
+    },
+    {
+      "epoch": 0.4451307217192656,
+      "grad_norm": 0.1952434927225113,
+      "learning_rate": 0.001,
+      "loss": 2.5245,
+      "step": 10522
+    },
+    {
+      "epoch": 0.44517302648278195,
+      "grad_norm": 1.0436367988586426,
+      "learning_rate": 0.001,
+      "loss": 3.3446,
+      "step": 10523
+    },
+    {
+      "epoch": 0.4452153312462983,
+      "grad_norm": 0.18723340332508087,
+      "learning_rate": 0.001,
+      "loss": 2.3264,
+      "step": 10524
+    },
+    {
+      "epoch": 0.4452576360098147,
+      "grad_norm": 0.2530616521835327,
+      "learning_rate": 0.001,
+      "loss": 2.052,
+      "step": 10525
+    },
+    {
+      "epoch": 0.44529994077333107,
+      "grad_norm": 0.22954061627388,
+      "learning_rate": 0.001,
+      "loss": 2.7963,
+      "step": 10526
+    },
+    {
+      "epoch": 0.4453422455368474,
+      "grad_norm": 0.18304547667503357,
+      "learning_rate": 0.001,
+      "loss": 2.3264,
+      "step": 10527
+    },
+    {
+      "epoch": 0.44538455030036384,
+      "grad_norm": 0.39500945806503296,
+      "learning_rate": 0.001,
+      "loss": 1.9825,
+      "step": 10528
+    },
+    {
+      "epoch": 0.4454268550638802,
+      "grad_norm": 2.5766665935516357,
+      "learning_rate": 0.001,
+      "loss": 2.5105,
+      "step": 10529
+    },
+    {
+      "epoch": 0.44546915982739654,
+      "grad_norm": 0.23480746150016785,
+      "learning_rate": 0.001,
+      "loss": 1.8403,
+      "step": 10530
+    },
+    {
+      "epoch": 0.44551146459091295,
+      "grad_norm": 0.18920212984085083,
+      "learning_rate": 0.001,
+      "loss": 2.8969,
+      "step": 10531
+    },
+    {
+      "epoch": 0.4455537693544293,
+      "grad_norm": 0.1778506338596344,
+      "learning_rate": 0.001,
+      "loss": 2.9153,
+      "step": 10532
+    },
+    {
+      "epoch": 0.44559607411794566,
+      "grad_norm": 0.28486740589141846,
+      "learning_rate": 0.001,
+      "loss": 3.6157,
+      "step": 10533
+    },
+    {
+      "epoch": 0.44563837888146207,
+      "grad_norm": 0.19953373074531555,
+      "learning_rate": 0.001,
+      "loss": 2.1183,
+      "step": 10534
+    },
+    {
+      "epoch": 0.4456806836449784,
+      "grad_norm": 0.17005318403244019,
+      "learning_rate": 0.001,
+      "loss": 3.2538,
+      "step": 10535
+    },
+    {
+      "epoch": 0.4457229884084948,
+      "grad_norm": 0.2750973105430603,
+      "learning_rate": 0.001,
+      "loss": 2.6134,
+      "step": 10536
+    },
+    {
+      "epoch": 0.4457652931720112,
+      "grad_norm": 0.18177230656147003,
+      "learning_rate": 0.001,
+      "loss": 2.8142,
+      "step": 10537
+    },
+    {
+      "epoch": 0.44580759793552754,
+      "grad_norm": 0.17410710453987122,
+      "learning_rate": 0.001,
+      "loss": 2.2946,
+      "step": 10538
+    },
+    {
+      "epoch": 0.4458499026990439,
+      "grad_norm": 0.42573174834251404,
+      "learning_rate": 0.001,
+      "loss": 2.5024,
+      "step": 10539
+    },
+    {
+      "epoch": 0.4458922074625603,
+      "grad_norm": 0.19186319410800934,
+      "learning_rate": 0.001,
+      "loss": 1.8359,
+      "step": 10540
+    },
+    {
+      "epoch": 0.44593451222607666,
+      "grad_norm": 0.3824551999568939,
+      "learning_rate": 0.001,
+      "loss": 2.5268,
+      "step": 10541
+    },
+    {
+      "epoch": 0.445976816989593,
+      "grad_norm": 0.23018528521060944,
+      "learning_rate": 0.001,
+      "loss": 2.834,
+      "step": 10542
+    },
+    {
+      "epoch": 0.4460191217531094,
+      "grad_norm": 0.1816767454147339,
+      "learning_rate": 0.001,
+      "loss": 2.0896,
+      "step": 10543
+    },
+    {
+      "epoch": 0.4460614265166258,
+      "grad_norm": 0.17159247398376465,
+      "learning_rate": 0.001,
+      "loss": 1.84,
+      "step": 10544
+    },
+    {
+      "epoch": 0.44610373128014214,
+      "grad_norm": 1.8396812677383423,
+      "learning_rate": 0.001,
+      "loss": 2.0751,
+      "step": 10545
+    },
+    {
+      "epoch": 0.4461460360436585,
+      "grad_norm": 0.3504880666732788,
+      "learning_rate": 0.001,
+      "loss": 2.3788,
+      "step": 10546
+    },
+    {
+      "epoch": 0.4461883408071749,
+      "grad_norm": 0.18512053787708282,
+      "learning_rate": 0.001,
+      "loss": 1.6916,
+      "step": 10547
+    },
+    {
+      "epoch": 0.44623064557069125,
+      "grad_norm": 0.18409253656864166,
+      "learning_rate": 0.001,
+      "loss": 1.5346,
+      "step": 10548
+    },
+    {
+      "epoch": 0.4462729503342076,
+      "grad_norm": 2.5886762142181396,
+      "learning_rate": 0.001,
+      "loss": 1.8567,
+      "step": 10549
+    },
+    {
+      "epoch": 0.446315255097724,
+      "grad_norm": 0.6384299993515015,
+      "learning_rate": 0.001,
+      "loss": 2.2381,
+      "step": 10550
+    },
+    {
+      "epoch": 0.44635755986124037,
+      "grad_norm": 2.2755072116851807,
+      "learning_rate": 0.001,
+      "loss": 2.2896,
+      "step": 10551
+    },
+    {
+      "epoch": 0.4463998646247567,
+      "grad_norm": 0.4430360198020935,
+      "learning_rate": 0.001,
+      "loss": 3.7153,
+      "step": 10552
+    },
+    {
+      "epoch": 0.44644216938827314,
+      "grad_norm": 1.0187450647354126,
+      "learning_rate": 0.001,
+      "loss": 3.2753,
+      "step": 10553
+    },
+    {
+      "epoch": 0.4464844741517895,
+      "grad_norm": 0.2957771122455597,
+      "learning_rate": 0.001,
+      "loss": 1.7454,
+      "step": 10554
+    },
+    {
+      "epoch": 0.44652677891530584,
+      "grad_norm": 0.2080879807472229,
+      "learning_rate": 0.001,
+      "loss": 2.0611,
+      "step": 10555
+    },
+    {
+      "epoch": 0.44656908367882225,
+      "grad_norm": 1.7243143320083618,
+      "learning_rate": 0.001,
+      "loss": 2.981,
+      "step": 10556
+    },
+    {
+      "epoch": 0.4466113884423386,
+      "grad_norm": 0.2195415198802948,
+      "learning_rate": 0.001,
+      "loss": 1.9373,
+      "step": 10557
+    },
+    {
+      "epoch": 0.44665369320585496,
+      "grad_norm": 0.19098582863807678,
+      "learning_rate": 0.001,
+      "loss": 1.9921,
+      "step": 10558
+    },
+    {
+      "epoch": 0.4466959979693714,
+      "grad_norm": 0.2138836830854416,
+      "learning_rate": 0.001,
+      "loss": 2.2429,
+      "step": 10559
+    },
+    {
+      "epoch": 0.4467383027328877,
+      "grad_norm": 0.18263570964336395,
+      "learning_rate": 0.001,
+      "loss": 2.5905,
+      "step": 10560
+    },
+    {
+      "epoch": 0.4467806074964041,
+      "grad_norm": 0.20292994379997253,
+      "learning_rate": 0.001,
+      "loss": 2.3356,
+      "step": 10561
+    },
+    {
+      "epoch": 0.4468229122599205,
+      "grad_norm": 0.2716735601425171,
+      "learning_rate": 0.001,
+      "loss": 1.7254,
+      "step": 10562
+    },
+    {
+      "epoch": 0.44686521702343684,
+      "grad_norm": 0.16340197622776031,
+      "learning_rate": 0.001,
+      "loss": 1.7591,
+      "step": 10563
+    },
+    {
+      "epoch": 0.4469075217869532,
+      "grad_norm": 0.2119547724723816,
+      "learning_rate": 0.001,
+      "loss": 2.3141,
+      "step": 10564
+    },
+    {
+      "epoch": 0.4469498265504696,
+      "grad_norm": 0.18071070313453674,
+      "learning_rate": 0.001,
+      "loss": 1.8618,
+      "step": 10565
+    },
+    {
+      "epoch": 0.44699213131398596,
+      "grad_norm": 0.3588818311691284,
+      "learning_rate": 0.001,
+      "loss": 3.3872,
+      "step": 10566
+    },
+    {
+      "epoch": 0.4470344360775023,
+      "grad_norm": 0.19287428259849548,
+      "learning_rate": 0.001,
+      "loss": 2.3166,
+      "step": 10567
+    },
+    {
+      "epoch": 0.44707674084101867,
+      "grad_norm": 0.43625664710998535,
+      "learning_rate": 0.001,
+      "loss": 2.8193,
+      "step": 10568
+    },
+    {
+      "epoch": 0.4471190456045351,
+      "grad_norm": 0.4928484559059143,
+      "learning_rate": 0.001,
+      "loss": 1.6347,
+      "step": 10569
+    },
+    {
+      "epoch": 0.44716135036805144,
+      "grad_norm": 0.685997724533081,
+      "learning_rate": 0.001,
+      "loss": 2.317,
+      "step": 10570
+    },
+    {
+      "epoch": 0.4472036551315678,
+      "grad_norm": 0.15223908424377441,
+      "learning_rate": 0.001,
+      "loss": 2.1719,
+      "step": 10571
+    },
+    {
+      "epoch": 0.4472459598950842,
+      "grad_norm": 0.18542735278606415,
+      "learning_rate": 0.001,
+      "loss": 1.5935,
+      "step": 10572
+    },
+    {
+      "epoch": 0.44728826465860055,
+      "grad_norm": 0.18469227850437164,
+      "learning_rate": 0.001,
+      "loss": 2.4737,
+      "step": 10573
+    },
+    {
+      "epoch": 0.4473305694221169,
+      "grad_norm": 0.24252432584762573,
+      "learning_rate": 0.001,
+      "loss": 1.9852,
+      "step": 10574
+    },
+    {
+      "epoch": 0.4473728741856333,
+      "grad_norm": 0.16159304976463318,
+      "learning_rate": 0.001,
+      "loss": 2.6181,
+      "step": 10575
+    },
+    {
+      "epoch": 0.44741517894914967,
+      "grad_norm": 3.23504900932312,
+      "learning_rate": 0.001,
+      "loss": 3.0528,
+      "step": 10576
+    },
+    {
+      "epoch": 0.447457483712666,
+      "grad_norm": 2.2757198810577393,
+      "learning_rate": 0.001,
+      "loss": 3.2706,
+      "step": 10577
+    },
+    {
+      "epoch": 0.44749978847618244,
+      "grad_norm": 0.22294031083583832,
+      "learning_rate": 0.001,
+      "loss": 2.1312,
+      "step": 10578
+    },
+    {
+      "epoch": 0.4475420932396988,
+      "grad_norm": 0.19114287197589874,
+      "learning_rate": 0.001,
+      "loss": 2.8998,
+      "step": 10579
+    },
+    {
+      "epoch": 0.44758439800321514,
+      "grad_norm": 0.6231216192245483,
+      "learning_rate": 0.001,
+      "loss": 3.3509,
+      "step": 10580
+    },
+    {
+      "epoch": 0.44762670276673155,
+      "grad_norm": 0.18667848408222198,
+      "learning_rate": 0.001,
+      "loss": 2.477,
+      "step": 10581
+    },
+    {
+      "epoch": 0.4476690075302479,
+      "grad_norm": 0.5230020880699158,
+      "learning_rate": 0.001,
+      "loss": 3.9596,
+      "step": 10582
+    },
+    {
+      "epoch": 0.44771131229376426,
+      "grad_norm": 0.2077375054359436,
+      "learning_rate": 0.001,
+      "loss": 2.1221,
+      "step": 10583
+    },
+    {
+      "epoch": 0.4477536170572807,
+      "grad_norm": 0.21029135584831238,
+      "learning_rate": 0.001,
+      "loss": 2.7445,
+      "step": 10584
+    },
+    {
+      "epoch": 0.447795921820797,
+      "grad_norm": 0.30405527353286743,
+      "learning_rate": 0.001,
+      "loss": 3.0613,
+      "step": 10585
+    },
+    {
+      "epoch": 0.4478382265843134,
+      "grad_norm": 0.16450397670269012,
+      "learning_rate": 0.001,
+      "loss": 2.518,
+      "step": 10586
+    },
+    {
+      "epoch": 0.4478805313478298,
+      "grad_norm": 1.3755722045898438,
+      "learning_rate": 0.001,
+      "loss": 2.3927,
+      "step": 10587
+    },
+    {
+      "epoch": 0.44792283611134615,
+      "grad_norm": 0.18358470499515533,
+      "learning_rate": 0.001,
+      "loss": 2.7458,
+      "step": 10588
+    },
+    {
+      "epoch": 0.4479651408748625,
+      "grad_norm": 5.526552200317383,
+      "learning_rate": 0.001,
+      "loss": 1.9233,
+      "step": 10589
+    },
+    {
+      "epoch": 0.44800744563837885,
+      "grad_norm": 0.2043481022119522,
+      "learning_rate": 0.001,
+      "loss": 2.0861,
+      "step": 10590
+    },
+    {
+      "epoch": 0.44804975040189526,
+      "grad_norm": 0.16888974606990814,
+      "learning_rate": 0.001,
+      "loss": 3.3986,
+      "step": 10591
+    },
+    {
+      "epoch": 0.4480920551654116,
+      "grad_norm": 0.208000048995018,
+      "learning_rate": 0.001,
+      "loss": 2.8552,
+      "step": 10592
+    },
+    {
+      "epoch": 0.44813435992892797,
+      "grad_norm": 2.04957914352417,
+      "learning_rate": 0.001,
+      "loss": 3.1259,
+      "step": 10593
+    },
+    {
+      "epoch": 0.4481766646924444,
+      "grad_norm": 0.4959050118923187,
+      "learning_rate": 0.001,
+      "loss": 2.161,
+      "step": 10594
+    },
+    {
+      "epoch": 0.44821896945596074,
+      "grad_norm": 0.20981314778327942,
+      "learning_rate": 0.001,
+      "loss": 2.1842,
+      "step": 10595
+    },
+    {
+      "epoch": 0.4482612742194771,
+      "grad_norm": 0.29932335019111633,
+      "learning_rate": 0.001,
+      "loss": 3.8183,
+      "step": 10596
+    },
+    {
+      "epoch": 0.4483035789829935,
+      "grad_norm": 0.22173571586608887,
+      "learning_rate": 0.001,
+      "loss": 2.4446,
+      "step": 10597
+    },
+    {
+      "epoch": 0.44834588374650985,
+      "grad_norm": 0.19373901188373566,
+      "learning_rate": 0.001,
+      "loss": 2.1564,
+      "step": 10598
+    },
+    {
+      "epoch": 0.4483881885100262,
+      "grad_norm": 0.802295446395874,
+      "learning_rate": 0.001,
+      "loss": 1.8698,
+      "step": 10599
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.1932401806116104,
+      "learning_rate": 0.001,
+      "loss": 2.4997,
+      "step": 10600
+    },
+    {
+      "epoch": 0.448472798037059,
+      "grad_norm": 0.19700828194618225,
+      "learning_rate": 0.001,
+      "loss": 2.4923,
+      "step": 10601
+    },
+    {
+      "epoch": 0.4485151028005753,
+      "grad_norm": 0.338742196559906,
+      "learning_rate": 0.001,
+      "loss": 1.6468,
+      "step": 10602
+    },
+    {
+      "epoch": 0.44855740756409174,
+      "grad_norm": 0.31011953949928284,
+      "learning_rate": 0.001,
+      "loss": 1.9476,
+      "step": 10603
+    },
+    {
+      "epoch": 0.4485997123276081,
+      "grad_norm": 0.1957470327615738,
+      "learning_rate": 0.001,
+      "loss": 1.7986,
+      "step": 10604
+    },
+    {
+      "epoch": 0.44864201709112445,
+      "grad_norm": 0.18266044557094574,
+      "learning_rate": 0.001,
+      "loss": 1.9285,
+      "step": 10605
+    },
+    {
+      "epoch": 0.44868432185464086,
+      "grad_norm": 0.17342445254325867,
+      "learning_rate": 0.001,
+      "loss": 2.4775,
+      "step": 10606
+    },
+    {
+      "epoch": 0.4487266266181572,
+      "grad_norm": 0.17653240263462067,
+      "learning_rate": 0.001,
+      "loss": 2.5431,
+      "step": 10607
+    },
+    {
+      "epoch": 0.44876893138167356,
+      "grad_norm": 0.1600387543439865,
+      "learning_rate": 0.001,
+      "loss": 3.0684,
+      "step": 10608
+    },
+    {
+      "epoch": 0.44881123614519,
+      "grad_norm": 0.20908235013484955,
+      "learning_rate": 0.001,
+      "loss": 2.4038,
+      "step": 10609
+    },
+    {
+      "epoch": 0.4488535409087063,
+      "grad_norm": 0.21198759973049164,
+      "learning_rate": 0.001,
+      "loss": 2.5214,
+      "step": 10610
+    },
+    {
+      "epoch": 0.4488958456722227,
+      "grad_norm": 0.16591830551624298,
+      "learning_rate": 0.001,
+      "loss": 2.178,
+      "step": 10611
+    },
+    {
+      "epoch": 0.4489381504357391,
+      "grad_norm": 3.6654651165008545,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 10612
+    },
+    {
+      "epoch": 0.44898045519925545,
+      "grad_norm": 0.17161937057971954,
+      "learning_rate": 0.001,
+      "loss": 2.2148,
+      "step": 10613
+    },
+    {
+      "epoch": 0.4490227599627718,
+      "grad_norm": 4.128261089324951,
+      "learning_rate": 0.001,
+      "loss": 2.2959,
+      "step": 10614
+    },
+    {
+      "epoch": 0.44906506472628815,
+      "grad_norm": 0.16964125633239746,
+      "learning_rate": 0.001,
+      "loss": 2.1389,
+      "step": 10615
+    },
+    {
+      "epoch": 0.44910736948980456,
+      "grad_norm": 1.7417608499526978,
+      "learning_rate": 0.001,
+      "loss": 1.705,
+      "step": 10616
+    },
+    {
+      "epoch": 0.4491496742533209,
+      "grad_norm": 0.9313798546791077,
+      "learning_rate": 0.001,
+      "loss": 2.6124,
+      "step": 10617
+    },
+    {
+      "epoch": 0.4491919790168373,
+      "grad_norm": 0.7544777393341064,
+      "learning_rate": 0.001,
+      "loss": 2.773,
+      "step": 10618
+    },
+    {
+      "epoch": 0.4492342837803537,
+      "grad_norm": 0.26840439438819885,
+      "learning_rate": 0.001,
+      "loss": 2.0986,
+      "step": 10619
+    },
+    {
+      "epoch": 0.44927658854387004,
+      "grad_norm": 0.18352295458316803,
+      "learning_rate": 0.001,
+      "loss": 2.9253,
+      "step": 10620
+    },
+    {
+      "epoch": 0.4493188933073864,
+      "grad_norm": 0.25015711784362793,
+      "learning_rate": 0.001,
+      "loss": 2.9165,
+      "step": 10621
+    },
+    {
+      "epoch": 0.4493611980709028,
+      "grad_norm": 0.15691161155700684,
+      "learning_rate": 0.001,
+      "loss": 1.6252,
+      "step": 10622
+    },
+    {
+      "epoch": 0.44940350283441916,
+      "grad_norm": 0.2128295749425888,
+      "learning_rate": 0.001,
+      "loss": 2.6712,
+      "step": 10623
+    },
+    {
+      "epoch": 0.4494458075979355,
+      "grad_norm": 2.622805118560791,
+      "learning_rate": 0.001,
+      "loss": 1.9613,
+      "step": 10624
+    },
+    {
+      "epoch": 0.4494881123614519,
+      "grad_norm": 1.922769546508789,
+      "learning_rate": 0.001,
+      "loss": 2.0337,
+      "step": 10625
+    },
+    {
+      "epoch": 0.4495304171249683,
+      "grad_norm": 0.1844593733549118,
+      "learning_rate": 0.001,
+      "loss": 2.4377,
+      "step": 10626
+    },
+    {
+      "epoch": 0.4495727218884846,
+      "grad_norm": 0.36522769927978516,
+      "learning_rate": 0.001,
+      "loss": 2.3181,
+      "step": 10627
+    },
+    {
+      "epoch": 0.44961502665200104,
+      "grad_norm": 0.796236515045166,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 10628
+    },
+    {
+      "epoch": 0.4496573314155174,
+      "grad_norm": 25.66647720336914,
+      "learning_rate": 0.001,
+      "loss": 1.9776,
+      "step": 10629
+    },
+    {
+      "epoch": 0.44969963617903375,
+      "grad_norm": 0.1876109093427658,
+      "learning_rate": 0.001,
+      "loss": 3.1408,
+      "step": 10630
+    },
+    {
+      "epoch": 0.44974194094255016,
+      "grad_norm": 0.2271474301815033,
+      "learning_rate": 0.001,
+      "loss": 2.7732,
+      "step": 10631
+    },
+    {
+      "epoch": 0.4497842457060665,
+      "grad_norm": 0.14895124733448029,
+      "learning_rate": 0.001,
+      "loss": 1.4203,
+      "step": 10632
+    },
+    {
+      "epoch": 0.44982655046958286,
+      "grad_norm": 0.5558005571365356,
+      "learning_rate": 0.001,
+      "loss": 2.295,
+      "step": 10633
+    },
+    {
+      "epoch": 0.4498688552330993,
+      "grad_norm": 0.2572641968727112,
+      "learning_rate": 0.001,
+      "loss": 2.5265,
+      "step": 10634
+    },
+    {
+      "epoch": 0.44991115999661563,
+      "grad_norm": 0.22297467291355133,
+      "learning_rate": 0.001,
+      "loss": 2.1774,
+      "step": 10635
+    },
+    {
+      "epoch": 0.449953464760132,
+      "grad_norm": 1.5451843738555908,
+      "learning_rate": 0.001,
+      "loss": 2.925,
+      "step": 10636
+    },
+    {
+      "epoch": 0.44999576952364834,
+      "grad_norm": 0.1914081871509552,
+      "learning_rate": 0.001,
+      "loss": 2.0183,
+      "step": 10637
+    },
+    {
+      "epoch": 0.45003807428716475,
+      "grad_norm": 0.192905992269516,
+      "learning_rate": 0.001,
+      "loss": 1.8431,
+      "step": 10638
+    },
+    {
+      "epoch": 0.4500803790506811,
+      "grad_norm": 0.1977374255657196,
+      "learning_rate": 0.001,
+      "loss": 2.5427,
+      "step": 10639
+    },
+    {
+      "epoch": 0.45012268381419746,
+      "grad_norm": 0.193924680352211,
+      "learning_rate": 0.001,
+      "loss": 2.4878,
+      "step": 10640
+    },
+    {
+      "epoch": 0.45016498857771386,
+      "grad_norm": 0.19579516351222992,
+      "learning_rate": 0.001,
+      "loss": 2.6807,
+      "step": 10641
+    },
+    {
+      "epoch": 0.4502072933412302,
+      "grad_norm": 0.17491202056407928,
+      "learning_rate": 0.001,
+      "loss": 1.4516,
+      "step": 10642
+    },
+    {
+      "epoch": 0.4502495981047466,
+      "grad_norm": 0.20388638973236084,
+      "learning_rate": 0.001,
+      "loss": 1.9766,
+      "step": 10643
+    },
+    {
+      "epoch": 0.450291902868263,
+      "grad_norm": 6.089061737060547,
+      "learning_rate": 0.001,
+      "loss": 1.7728,
+      "step": 10644
+    },
+    {
+      "epoch": 0.45033420763177934,
+      "grad_norm": 0.19175101816654205,
+      "learning_rate": 0.001,
+      "loss": 2.4468,
+      "step": 10645
+    },
+    {
+      "epoch": 0.4503765123952957,
+      "grad_norm": 0.18164491653442383,
+      "learning_rate": 0.001,
+      "loss": 1.6122,
+      "step": 10646
+    },
+    {
+      "epoch": 0.4504188171588121,
+      "grad_norm": 0.28620779514312744,
+      "learning_rate": 0.001,
+      "loss": 2.1983,
+      "step": 10647
+    },
+    {
+      "epoch": 0.45046112192232846,
+      "grad_norm": 0.17143702507019043,
+      "learning_rate": 0.001,
+      "loss": 3.1166,
+      "step": 10648
+    },
+    {
+      "epoch": 0.4505034266858448,
+      "grad_norm": 0.187702938914299,
+      "learning_rate": 0.001,
+      "loss": 2.7673,
+      "step": 10649
+    },
+    {
+      "epoch": 0.4505457314493612,
+      "grad_norm": 0.18621903657913208,
+      "learning_rate": 0.001,
+      "loss": 2.0963,
+      "step": 10650
+    },
+    {
+      "epoch": 0.4505880362128776,
+      "grad_norm": 0.18782265484333038,
+      "learning_rate": 0.001,
+      "loss": 2.6116,
+      "step": 10651
+    },
+    {
+      "epoch": 0.45063034097639393,
+      "grad_norm": 0.18947267532348633,
+      "learning_rate": 0.001,
+      "loss": 2.3372,
+      "step": 10652
+    },
+    {
+      "epoch": 0.45067264573991034,
+      "grad_norm": 0.6221209764480591,
+      "learning_rate": 0.001,
+      "loss": 1.4182,
+      "step": 10653
+    },
+    {
+      "epoch": 0.4507149505034267,
+      "grad_norm": 0.18391531705856323,
+      "learning_rate": 0.001,
+      "loss": 2.4517,
+      "step": 10654
+    },
+    {
+      "epoch": 0.45075725526694305,
+      "grad_norm": 0.31003570556640625,
+      "learning_rate": 0.001,
+      "loss": 2.6419,
+      "step": 10655
+    },
+    {
+      "epoch": 0.45079956003045946,
+      "grad_norm": 3.465576410293579,
+      "learning_rate": 0.001,
+      "loss": 2.7737,
+      "step": 10656
+    },
+    {
+      "epoch": 0.4508418647939758,
+      "grad_norm": 0.17568989098072052,
+      "learning_rate": 0.001,
+      "loss": 1.5883,
+      "step": 10657
+    },
+    {
+      "epoch": 0.45088416955749216,
+      "grad_norm": 0.18298137187957764,
+      "learning_rate": 0.001,
+      "loss": 2.9265,
+      "step": 10658
+    },
+    {
+      "epoch": 0.4509264743210085,
+      "grad_norm": 13.20496654510498,
+      "learning_rate": 0.001,
+      "loss": 2.632,
+      "step": 10659
+    },
+    {
+      "epoch": 0.45096877908452493,
+      "grad_norm": 0.2141726016998291,
+      "learning_rate": 0.001,
+      "loss": 1.5008,
+      "step": 10660
+    },
+    {
+      "epoch": 0.4510110838480413,
+      "grad_norm": 0.22003300487995148,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 10661
+    },
+    {
+      "epoch": 0.45105338861155764,
+      "grad_norm": 0.4876539409160614,
+      "learning_rate": 0.001,
+      "loss": 2.3984,
+      "step": 10662
+    },
+    {
+      "epoch": 0.45109569337507405,
+      "grad_norm": 0.2729082703590393,
+      "learning_rate": 0.001,
+      "loss": 1.7232,
+      "step": 10663
+    },
+    {
+      "epoch": 0.4511379981385904,
+      "grad_norm": 0.21046875417232513,
+      "learning_rate": 0.001,
+      "loss": 1.6638,
+      "step": 10664
+    },
+    {
+      "epoch": 0.45118030290210676,
+      "grad_norm": 6.246835708618164,
+      "learning_rate": 0.001,
+      "loss": 2.2441,
+      "step": 10665
+    },
+    {
+      "epoch": 0.45122260766562317,
+      "grad_norm": 0.24696436524391174,
+      "learning_rate": 0.001,
+      "loss": 2.4485,
+      "step": 10666
+    },
+    {
+      "epoch": 0.4512649124291395,
+      "grad_norm": 0.20760883390903473,
+      "learning_rate": 0.001,
+      "loss": 1.8772,
+      "step": 10667
+    },
+    {
+      "epoch": 0.4513072171926559,
+      "grad_norm": 0.2544911801815033,
+      "learning_rate": 0.001,
+      "loss": 3.1766,
+      "step": 10668
+    },
+    {
+      "epoch": 0.4513495219561723,
+      "grad_norm": 0.7529020309448242,
+      "learning_rate": 0.001,
+      "loss": 3.3731,
+      "step": 10669
+    },
+    {
+      "epoch": 0.45139182671968864,
+      "grad_norm": 0.23148514330387115,
+      "learning_rate": 0.001,
+      "loss": 2.8049,
+      "step": 10670
+    },
+    {
+      "epoch": 0.451434131483205,
+      "grad_norm": 0.28361719846725464,
+      "learning_rate": 0.001,
+      "loss": 2.4099,
+      "step": 10671
+    },
+    {
+      "epoch": 0.4514764362467214,
+      "grad_norm": 0.23970754444599152,
+      "learning_rate": 0.001,
+      "loss": 2.2309,
+      "step": 10672
+    },
+    {
+      "epoch": 0.45151874101023776,
+      "grad_norm": 0.20568141341209412,
+      "learning_rate": 0.001,
+      "loss": 2.9698,
+      "step": 10673
+    },
+    {
+      "epoch": 0.4515610457737541,
+      "grad_norm": 0.1887006312608719,
+      "learning_rate": 0.001,
+      "loss": 1.9361,
+      "step": 10674
+    },
+    {
+      "epoch": 0.4516033505372705,
+      "grad_norm": 0.31066638231277466,
+      "learning_rate": 0.001,
+      "loss": 2.2419,
+      "step": 10675
+    },
+    {
+      "epoch": 0.4516456553007869,
+      "grad_norm": 0.18941254913806915,
+      "learning_rate": 0.001,
+      "loss": 2.7735,
+      "step": 10676
+    },
+    {
+      "epoch": 0.45168796006430323,
+      "grad_norm": 0.16498833894729614,
+      "learning_rate": 0.001,
+      "loss": 3.2135,
+      "step": 10677
+    },
+    {
+      "epoch": 0.45173026482781964,
+      "grad_norm": 0.18185345828533173,
+      "learning_rate": 0.001,
+      "loss": 2.1368,
+      "step": 10678
+    },
+    {
+      "epoch": 0.451772569591336,
+      "grad_norm": 0.17669661343097687,
+      "learning_rate": 0.001,
+      "loss": 2.4485,
+      "step": 10679
+    },
+    {
+      "epoch": 0.45181487435485235,
+      "grad_norm": 0.22324764728546143,
+      "learning_rate": 0.001,
+      "loss": 2.542,
+      "step": 10680
+    },
+    {
+      "epoch": 0.4518571791183687,
+      "grad_norm": 0.17187261581420898,
+      "learning_rate": 0.001,
+      "loss": 2.604,
+      "step": 10681
+    },
+    {
+      "epoch": 0.4518994838818851,
+      "grad_norm": 0.184734508395195,
+      "learning_rate": 0.001,
+      "loss": 2.2802,
+      "step": 10682
+    },
+    {
+      "epoch": 0.45194178864540147,
+      "grad_norm": 0.14700573682785034,
+      "learning_rate": 0.001,
+      "loss": 3.5054,
+      "step": 10683
+    },
+    {
+      "epoch": 0.4519840934089178,
+      "grad_norm": 2.5030956268310547,
+      "learning_rate": 0.001,
+      "loss": 2.5592,
+      "step": 10684
+    },
+    {
+      "epoch": 0.45202639817243423,
+      "grad_norm": 0.20399565994739532,
+      "learning_rate": 0.001,
+      "loss": 3.7827,
+      "step": 10685
+    },
+    {
+      "epoch": 0.4520687029359506,
+      "grad_norm": 1.0105069875717163,
+      "learning_rate": 0.001,
+      "loss": 2.5169,
+      "step": 10686
+    },
+    {
+      "epoch": 0.45211100769946694,
+      "grad_norm": 0.39226314425468445,
+      "learning_rate": 0.001,
+      "loss": 1.9134,
+      "step": 10687
+    },
+    {
+      "epoch": 0.45215331246298335,
+      "grad_norm": 0.19782952964305878,
+      "learning_rate": 0.001,
+      "loss": 3.265,
+      "step": 10688
+    },
+    {
+      "epoch": 0.4521956172264997,
+      "grad_norm": 1.149984359741211,
+      "learning_rate": 0.001,
+      "loss": 2.6159,
+      "step": 10689
+    },
+    {
+      "epoch": 0.45223792199001606,
+      "grad_norm": 0.13431769609451294,
+      "learning_rate": 0.001,
+      "loss": 1.4885,
+      "step": 10690
+    },
+    {
+      "epoch": 0.45228022675353247,
+      "grad_norm": 0.18614532053470612,
+      "learning_rate": 0.001,
+      "loss": 1.9288,
+      "step": 10691
+    },
+    {
+      "epoch": 0.4523225315170488,
+      "grad_norm": 0.2908714711666107,
+      "learning_rate": 0.001,
+      "loss": 3.1537,
+      "step": 10692
+    },
+    {
+      "epoch": 0.4523648362805652,
+      "grad_norm": 0.21332935988903046,
+      "learning_rate": 0.001,
+      "loss": 2.2536,
+      "step": 10693
+    },
+    {
+      "epoch": 0.4524071410440816,
+      "grad_norm": 0.3098827302455902,
+      "learning_rate": 0.001,
+      "loss": 5.137,
+      "step": 10694
+    },
+    {
+      "epoch": 0.45244944580759794,
+      "grad_norm": 0.2915304899215698,
+      "learning_rate": 0.001,
+      "loss": 2.5853,
+      "step": 10695
+    },
+    {
+      "epoch": 0.4524917505711143,
+      "grad_norm": 1.5301575660705566,
+      "learning_rate": 0.001,
+      "loss": 2.4361,
+      "step": 10696
+    },
+    {
+      "epoch": 0.4525340553346307,
+      "grad_norm": 0.1704961210489273,
+      "learning_rate": 0.001,
+      "loss": 2.1001,
+      "step": 10697
+    },
+    {
+      "epoch": 0.45257636009814706,
+      "grad_norm": 0.3829928934574127,
+      "learning_rate": 0.001,
+      "loss": 1.9829,
+      "step": 10698
+    },
+    {
+      "epoch": 0.4526186648616634,
+      "grad_norm": 0.14519084990024567,
+      "learning_rate": 0.001,
+      "loss": 1.8952,
+      "step": 10699
+    },
+    {
+      "epoch": 0.4526609696251798,
+      "grad_norm": 5.465473651885986,
+      "learning_rate": 0.001,
+      "loss": 2.0073,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4527032743886962,
+      "grad_norm": 0.17210935056209564,
+      "learning_rate": 0.001,
+      "loss": 2.0146,
+      "step": 10701
+    },
+    {
+      "epoch": 0.45274557915221253,
+      "grad_norm": 2.133603811264038,
+      "learning_rate": 0.001,
+      "loss": 3.2162,
+      "step": 10702
+    },
+    {
+      "epoch": 0.4527878839157289,
+      "grad_norm": 0.194070965051651,
+      "learning_rate": 0.001,
+      "loss": 2.5802,
+      "step": 10703
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.21091453731060028,
+      "learning_rate": 0.001,
+      "loss": 2.4426,
+      "step": 10704
+    },
+    {
+      "epoch": 0.45287249344276165,
+      "grad_norm": 0.1854954957962036,
+      "learning_rate": 0.001,
+      "loss": 1.5763,
+      "step": 10705
+    },
+    {
+      "epoch": 0.452914798206278,
+      "grad_norm": 0.1935143917798996,
+      "learning_rate": 0.001,
+      "loss": 1.7313,
+      "step": 10706
+    },
+    {
+      "epoch": 0.4529571029697944,
+      "grad_norm": 0.32737401127815247,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 10707
+    },
+    {
+      "epoch": 0.45299940773331077,
+      "grad_norm": 3.649876832962036,
+      "learning_rate": 0.001,
+      "loss": 2.0724,
+      "step": 10708
+    },
+    {
+      "epoch": 0.4530417124968271,
+      "grad_norm": 1.8369063138961792,
+      "learning_rate": 0.001,
+      "loss": 1.6074,
+      "step": 10709
+    },
+    {
+      "epoch": 0.45308401726034353,
+      "grad_norm": 0.22719253599643707,
+      "learning_rate": 0.001,
+      "loss": 2.463,
+      "step": 10710
+    },
+    {
+      "epoch": 0.4531263220238599,
+      "grad_norm": 0.38920852541923523,
+      "learning_rate": 0.001,
+      "loss": 1.6571,
+      "step": 10711
+    },
+    {
+      "epoch": 0.45316862678737624,
+      "grad_norm": 0.18624505400657654,
+      "learning_rate": 0.001,
+      "loss": 2.507,
+      "step": 10712
+    },
+    {
+      "epoch": 0.45321093155089265,
+      "grad_norm": 0.2142438441514969,
+      "learning_rate": 0.001,
+      "loss": 2.3202,
+      "step": 10713
+    },
+    {
+      "epoch": 0.453253236314409,
+      "grad_norm": 1.3076987266540527,
+      "learning_rate": 0.001,
+      "loss": 1.6404,
+      "step": 10714
+    },
+    {
+      "epoch": 0.45329554107792536,
+      "grad_norm": 0.9134403467178345,
+      "learning_rate": 0.001,
+      "loss": 2.0657,
+      "step": 10715
+    },
+    {
+      "epoch": 0.45333784584144177,
+      "grad_norm": 0.3806881904602051,
+      "learning_rate": 0.001,
+      "loss": 2.0147,
+      "step": 10716
+    },
+    {
+      "epoch": 0.4533801506049581,
+      "grad_norm": 0.1724647432565689,
+      "learning_rate": 0.001,
+      "loss": 2.2622,
+      "step": 10717
+    },
+    {
+      "epoch": 0.4534224553684745,
+      "grad_norm": 0.5232478976249695,
+      "learning_rate": 0.001,
+      "loss": 3.3054,
+      "step": 10718
+    },
+    {
+      "epoch": 0.4534647601319909,
+      "grad_norm": 0.16044361889362335,
+      "learning_rate": 0.001,
+      "loss": 3.2017,
+      "step": 10719
+    },
+    {
+      "epoch": 0.45350706489550724,
+      "grad_norm": 0.17905324697494507,
+      "learning_rate": 0.001,
+      "loss": 1.8238,
+      "step": 10720
+    },
+    {
+      "epoch": 0.4535493696590236,
+      "grad_norm": 0.20264795422554016,
+      "learning_rate": 0.001,
+      "loss": 1.8288,
+      "step": 10721
+    },
+    {
+      "epoch": 0.45359167442254,
+      "grad_norm": 42.060760498046875,
+      "learning_rate": 0.001,
+      "loss": 3.1013,
+      "step": 10722
+    },
+    {
+      "epoch": 0.45363397918605636,
+      "grad_norm": 0.8245616555213928,
+      "learning_rate": 0.001,
+      "loss": 1.8905,
+      "step": 10723
+    },
+    {
+      "epoch": 0.4536762839495727,
+      "grad_norm": 934.2753295898438,
+      "learning_rate": 0.001,
+      "loss": 3.5141,
+      "step": 10724
+    },
+    {
+      "epoch": 0.4537185887130891,
+      "grad_norm": 0.4653104245662689,
+      "learning_rate": 0.001,
+      "loss": 2.0465,
+      "step": 10725
+    },
+    {
+      "epoch": 0.4537608934766055,
+      "grad_norm": 0.2089160531759262,
+      "learning_rate": 0.001,
+      "loss": 2.6989,
+      "step": 10726
+    },
+    {
+      "epoch": 0.45380319824012183,
+      "grad_norm": 1.048384428024292,
+      "learning_rate": 0.001,
+      "loss": 2.7521,
+      "step": 10727
+    },
+    {
+      "epoch": 0.4538455030036382,
+      "grad_norm": 0.18600456416606903,
+      "learning_rate": 0.001,
+      "loss": 2.0483,
+      "step": 10728
+    },
+    {
+      "epoch": 0.4538878077671546,
+      "grad_norm": 2.7133305072784424,
+      "learning_rate": 0.001,
+      "loss": 2.2104,
+      "step": 10729
+    },
+    {
+      "epoch": 0.45393011253067095,
+      "grad_norm": 0.19306029379367828,
+      "learning_rate": 0.001,
+      "loss": 2.466,
+      "step": 10730
+    },
+    {
+      "epoch": 0.4539724172941873,
+      "grad_norm": 0.1988273561000824,
+      "learning_rate": 0.001,
+      "loss": 2.7235,
+      "step": 10731
+    },
+    {
+      "epoch": 0.4540147220577037,
+      "grad_norm": 6.111566543579102,
+      "learning_rate": 0.001,
+      "loss": 2.7007,
+      "step": 10732
+    },
+    {
+      "epoch": 0.45405702682122007,
+      "grad_norm": 0.4075400233268738,
+      "learning_rate": 0.001,
+      "loss": 3.3743,
+      "step": 10733
+    },
+    {
+      "epoch": 0.4540993315847364,
+      "grad_norm": 0.1714090257883072,
+      "learning_rate": 0.001,
+      "loss": 2.4183,
+      "step": 10734
+    },
+    {
+      "epoch": 0.45414163634825283,
+      "grad_norm": 0.27098438143730164,
+      "learning_rate": 0.001,
+      "loss": 2.5263,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4541839411117692,
+      "grad_norm": 0.17201238870620728,
+      "learning_rate": 0.001,
+      "loss": 2.4768,
+      "step": 10736
+    },
+    {
+      "epoch": 0.45422624587528554,
+      "grad_norm": 0.23325662314891815,
+      "learning_rate": 0.001,
+      "loss": 3.6854,
+      "step": 10737
+    },
+    {
+      "epoch": 0.45426855063880195,
+      "grad_norm": 0.5161541104316711,
+      "learning_rate": 0.001,
+      "loss": 2.359,
+      "step": 10738
+    },
+    {
+      "epoch": 0.4543108554023183,
+      "grad_norm": 0.16155464947223663,
+      "learning_rate": 0.001,
+      "loss": 2.0626,
+      "step": 10739
+    },
+    {
+      "epoch": 0.45435316016583466,
+      "grad_norm": 0.43314120173454285,
+      "learning_rate": 0.001,
+      "loss": 2.4758,
+      "step": 10740
+    },
+    {
+      "epoch": 0.45439546492935107,
+      "grad_norm": 0.17236952483654022,
+      "learning_rate": 0.001,
+      "loss": 2.7448,
+      "step": 10741
+    },
+    {
+      "epoch": 0.4544377696928674,
+      "grad_norm": 1.6672812700271606,
+      "learning_rate": 0.001,
+      "loss": 1.8297,
+      "step": 10742
+    },
+    {
+      "epoch": 0.4544800744563838,
+      "grad_norm": 0.6431288123130798,
+      "learning_rate": 0.001,
+      "loss": 2.5279,
+      "step": 10743
+    },
+    {
+      "epoch": 0.4545223792199002,
+      "grad_norm": 0.1590161770582199,
+      "learning_rate": 0.001,
+      "loss": 1.7153,
+      "step": 10744
+    },
+    {
+      "epoch": 0.45456468398341654,
+      "grad_norm": 0.1368952840566635,
+      "learning_rate": 0.001,
+      "loss": 2.4392,
+      "step": 10745
+    },
+    {
+      "epoch": 0.4546069887469329,
+      "grad_norm": 0.9663378596305847,
+      "learning_rate": 0.001,
+      "loss": 2.5891,
+      "step": 10746
+    },
+    {
+      "epoch": 0.4546492935104493,
+      "grad_norm": 0.19427025318145752,
+      "learning_rate": 0.001,
+      "loss": 1.8954,
+      "step": 10747
+    },
+    {
+      "epoch": 0.45469159827396566,
+      "grad_norm": 2.1707139015197754,
+      "learning_rate": 0.001,
+      "loss": 3.155,
+      "step": 10748
+    },
+    {
+      "epoch": 0.454733903037482,
+      "grad_norm": 0.20545673370361328,
+      "learning_rate": 0.001,
+      "loss": 3.2715,
+      "step": 10749
+    },
+    {
+      "epoch": 0.45477620780099837,
+      "grad_norm": 0.1690516173839569,
+      "learning_rate": 0.001,
+      "loss": 2.7757,
+      "step": 10750
+    },
+    {
+      "epoch": 0.4548185125645148,
+      "grad_norm": 0.4233834743499756,
+      "learning_rate": 0.001,
+      "loss": 2.896,
+      "step": 10751
+    },
+    {
+      "epoch": 0.45486081732803113,
+      "grad_norm": 0.2022281438112259,
+      "learning_rate": 0.001,
+      "loss": 2.3454,
+      "step": 10752
+    },
+    {
+      "epoch": 0.4549031220915475,
+      "grad_norm": 0.17715947329998016,
+      "learning_rate": 0.001,
+      "loss": 2.0872,
+      "step": 10753
+    },
+    {
+      "epoch": 0.4549454268550639,
+      "grad_norm": 0.19292153418064117,
+      "learning_rate": 0.001,
+      "loss": 2.1691,
+      "step": 10754
+    },
+    {
+      "epoch": 0.45498773161858025,
+      "grad_norm": 1.835961103439331,
+      "learning_rate": 0.001,
+      "loss": 1.9661,
+      "step": 10755
+    },
+    {
+      "epoch": 0.4550300363820966,
+      "grad_norm": 0.16486118733882904,
+      "learning_rate": 0.001,
+      "loss": 2.5038,
+      "step": 10756
+    },
+    {
+      "epoch": 0.455072341145613,
+      "grad_norm": 0.7095552682876587,
+      "learning_rate": 0.001,
+      "loss": 2.1318,
+      "step": 10757
+    },
+    {
+      "epoch": 0.45511464590912937,
+      "grad_norm": 0.8125048875808716,
+      "learning_rate": 0.001,
+      "loss": 2.7182,
+      "step": 10758
+    },
+    {
+      "epoch": 0.4551569506726457,
+      "grad_norm": 6.972851753234863,
+      "learning_rate": 0.001,
+      "loss": 2.4598,
+      "step": 10759
+    },
+    {
+      "epoch": 0.45519925543616213,
+      "grad_norm": 1.0214871168136597,
+      "learning_rate": 0.001,
+      "loss": 2.0997,
+      "step": 10760
+    },
+    {
+      "epoch": 0.4552415601996785,
+      "grad_norm": 8.2625093460083,
+      "learning_rate": 0.001,
+      "loss": 2.4566,
+      "step": 10761
+    },
+    {
+      "epoch": 0.45528386496319484,
+      "grad_norm": 0.16502858698368073,
+      "learning_rate": 0.001,
+      "loss": 1.868,
+      "step": 10762
+    },
+    {
+      "epoch": 0.45532616972671125,
+      "grad_norm": 0.4280339777469635,
+      "learning_rate": 0.001,
+      "loss": 1.6946,
+      "step": 10763
+    },
+    {
+      "epoch": 0.4553684744902276,
+      "grad_norm": 0.23198087513446808,
+      "learning_rate": 0.001,
+      "loss": 3.1451,
+      "step": 10764
+    },
+    {
+      "epoch": 0.45541077925374396,
+      "grad_norm": 0.8714398145675659,
+      "learning_rate": 0.001,
+      "loss": 3.1186,
+      "step": 10765
+    },
+    {
+      "epoch": 0.45545308401726037,
+      "grad_norm": 0.2438887655735016,
+      "learning_rate": 0.001,
+      "loss": 2.7754,
+      "step": 10766
+    },
+    {
+      "epoch": 0.4554953887807767,
+      "grad_norm": 0.3168472349643707,
+      "learning_rate": 0.001,
+      "loss": 2.4378,
+      "step": 10767
+    },
+    {
+      "epoch": 0.4555376935442931,
+      "grad_norm": 0.214552104473114,
+      "learning_rate": 0.001,
+      "loss": 2.1332,
+      "step": 10768
+    },
+    {
+      "epoch": 0.4555799983078095,
+      "grad_norm": 0.18347600102424622,
+      "learning_rate": 0.001,
+      "loss": 1.9963,
+      "step": 10769
+    },
+    {
+      "epoch": 0.45562230307132584,
+      "grad_norm": 0.28068867325782776,
+      "learning_rate": 0.001,
+      "loss": 2.335,
+      "step": 10770
+    },
+    {
+      "epoch": 0.4556646078348422,
+      "grad_norm": 0.21685069799423218,
+      "learning_rate": 0.001,
+      "loss": 2.0136,
+      "step": 10771
+    },
+    {
+      "epoch": 0.45570691259835855,
+      "grad_norm": 0.18479874730110168,
+      "learning_rate": 0.001,
+      "loss": 2.1247,
+      "step": 10772
+    },
+    {
+      "epoch": 0.45574921736187496,
+      "grad_norm": 9.378438949584961,
+      "learning_rate": 0.001,
+      "loss": 1.8411,
+      "step": 10773
+    },
+    {
+      "epoch": 0.4557915221253913,
+      "grad_norm": 0.18619363009929657,
+      "learning_rate": 0.001,
+      "loss": 2.4149,
+      "step": 10774
+    },
+    {
+      "epoch": 0.45583382688890767,
+      "grad_norm": 0.48514658212661743,
+      "learning_rate": 0.001,
+      "loss": 1.8157,
+      "step": 10775
+    },
+    {
+      "epoch": 0.4558761316524241,
+      "grad_norm": 0.21699242293834686,
+      "learning_rate": 0.001,
+      "loss": 2.5075,
+      "step": 10776
+    },
+    {
+      "epoch": 0.45591843641594043,
+      "grad_norm": 0.5351647138595581,
+      "learning_rate": 0.001,
+      "loss": 2.0256,
+      "step": 10777
+    },
+    {
+      "epoch": 0.4559607411794568,
+      "grad_norm": 0.18083517253398895,
+      "learning_rate": 0.001,
+      "loss": 1.8436,
+      "step": 10778
+    },
+    {
+      "epoch": 0.4560030459429732,
+      "grad_norm": 0.16347528994083405,
+      "learning_rate": 0.001,
+      "loss": 1.6929,
+      "step": 10779
+    },
+    {
+      "epoch": 0.45604535070648955,
+      "grad_norm": 0.59482342004776,
+      "learning_rate": 0.001,
+      "loss": 2.8624,
+      "step": 10780
+    },
+    {
+      "epoch": 0.4560876554700059,
+      "grad_norm": 0.28643858432769775,
+      "learning_rate": 0.001,
+      "loss": 2.5811,
+      "step": 10781
+    },
+    {
+      "epoch": 0.4561299602335223,
+      "grad_norm": 0.22716449201107025,
+      "learning_rate": 0.001,
+      "loss": 2.7431,
+      "step": 10782
+    },
+    {
+      "epoch": 0.45617226499703867,
+      "grad_norm": 4.5622477531433105,
+      "learning_rate": 0.001,
+      "loss": 1.9823,
+      "step": 10783
+    },
+    {
+      "epoch": 0.456214569760555,
+      "grad_norm": 0.4299876391887665,
+      "learning_rate": 0.001,
+      "loss": 1.8034,
+      "step": 10784
+    },
+    {
+      "epoch": 0.45625687452407143,
+      "grad_norm": 0.2948051989078522,
+      "learning_rate": 0.001,
+      "loss": 2.2761,
+      "step": 10785
+    },
+    {
+      "epoch": 0.4562991792875878,
+      "grad_norm": 0.16705124080181122,
+      "learning_rate": 0.001,
+      "loss": 3.0812,
+      "step": 10786
+    },
+    {
+      "epoch": 0.45634148405110414,
+      "grad_norm": 0.33032622933387756,
+      "learning_rate": 0.001,
+      "loss": 2.9178,
+      "step": 10787
+    },
+    {
+      "epoch": 0.45638378881462055,
+      "grad_norm": 0.3047260642051697,
+      "learning_rate": 0.001,
+      "loss": 2.0885,
+      "step": 10788
+    },
+    {
+      "epoch": 0.4564260935781369,
+      "grad_norm": 0.2009318619966507,
+      "learning_rate": 0.001,
+      "loss": 2.5426,
+      "step": 10789
+    },
+    {
+      "epoch": 0.45646839834165326,
+      "grad_norm": 0.26665037870407104,
+      "learning_rate": 0.001,
+      "loss": 3.0866,
+      "step": 10790
+    },
+    {
+      "epoch": 0.45651070310516967,
+      "grad_norm": 0.15253359079360962,
+      "learning_rate": 0.001,
+      "loss": 1.9167,
+      "step": 10791
+    },
+    {
+      "epoch": 0.456553007868686,
+      "grad_norm": 0.17726293206214905,
+      "learning_rate": 0.001,
+      "loss": 1.9892,
+      "step": 10792
+    },
+    {
+      "epoch": 0.4565953126322024,
+      "grad_norm": 0.1824486255645752,
+      "learning_rate": 0.001,
+      "loss": 2.0296,
+      "step": 10793
+    },
+    {
+      "epoch": 0.45663761739571873,
+      "grad_norm": 0.7334975004196167,
+      "learning_rate": 0.001,
+      "loss": 2.5563,
+      "step": 10794
+    },
+    {
+      "epoch": 0.45667992215923514,
+      "grad_norm": 0.14659246802330017,
+      "learning_rate": 0.001,
+      "loss": 1.8249,
+      "step": 10795
+    },
+    {
+      "epoch": 0.4567222269227515,
+      "grad_norm": 1.9629788398742676,
+      "learning_rate": 0.001,
+      "loss": 3.2285,
+      "step": 10796
+    },
+    {
+      "epoch": 0.45676453168626785,
+      "grad_norm": 0.18042373657226562,
+      "learning_rate": 0.001,
+      "loss": 3.0951,
+      "step": 10797
+    },
+    {
+      "epoch": 0.45680683644978426,
+      "grad_norm": 0.2109190672636032,
+      "learning_rate": 0.001,
+      "loss": 2.0418,
+      "step": 10798
+    },
+    {
+      "epoch": 0.4568491412133006,
+      "grad_norm": 0.2082575410604477,
+      "learning_rate": 0.001,
+      "loss": 2.7988,
+      "step": 10799
+    },
+    {
+      "epoch": 0.45689144597681697,
+      "grad_norm": 0.14357468485832214,
+      "learning_rate": 0.001,
+      "loss": 1.6557,
+      "step": 10800
+    },
+    {
+      "epoch": 0.4569337507403334,
+      "grad_norm": 0.16632072627544403,
+      "learning_rate": 0.001,
+      "loss": 1.9405,
+      "step": 10801
+    },
+    {
+      "epoch": 0.45697605550384973,
+      "grad_norm": 0.4018622636795044,
+      "learning_rate": 0.001,
+      "loss": 2.1912,
+      "step": 10802
+    },
+    {
+      "epoch": 0.4570183602673661,
+      "grad_norm": 0.19781026244163513,
+      "learning_rate": 0.001,
+      "loss": 2.1403,
+      "step": 10803
+    },
+    {
+      "epoch": 0.4570606650308825,
+      "grad_norm": 0.16085539758205414,
+      "learning_rate": 0.001,
+      "loss": 1.8415,
+      "step": 10804
+    },
+    {
+      "epoch": 0.45710296979439885,
+      "grad_norm": 0.2825770080089569,
+      "learning_rate": 0.001,
+      "loss": 3.2822,
+      "step": 10805
+    },
+    {
+      "epoch": 0.4571452745579152,
+      "grad_norm": 0.15267670154571533,
+      "learning_rate": 0.001,
+      "loss": 2.0494,
+      "step": 10806
+    },
+    {
+      "epoch": 0.4571875793214316,
+      "grad_norm": 1.4387484788894653,
+      "learning_rate": 0.001,
+      "loss": 3.0406,
+      "step": 10807
+    },
+    {
+      "epoch": 0.45722988408494797,
+      "grad_norm": 0.14792972803115845,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 10808
+    },
+    {
+      "epoch": 0.4572721888484643,
+      "grad_norm": 0.17246343195438385,
+      "learning_rate": 0.001,
+      "loss": 2.1818,
+      "step": 10809
+    },
+    {
+      "epoch": 0.45731449361198073,
+      "grad_norm": 0.40068310499191284,
+      "learning_rate": 0.001,
+      "loss": 1.7939,
+      "step": 10810
+    },
+    {
+      "epoch": 0.4573567983754971,
+      "grad_norm": 0.19033183157444,
+      "learning_rate": 0.001,
+      "loss": 2.0297,
+      "step": 10811
+    },
+    {
+      "epoch": 0.45739910313901344,
+      "grad_norm": 0.16076108813285828,
+      "learning_rate": 0.001,
+      "loss": 1.9428,
+      "step": 10812
+    },
+    {
+      "epoch": 0.45744140790252985,
+      "grad_norm": 0.1628408133983612,
+      "learning_rate": 0.001,
+      "loss": 2.0893,
+      "step": 10813
+    },
+    {
+      "epoch": 0.4574837126660462,
+      "grad_norm": 0.29927513003349304,
+      "learning_rate": 0.001,
+      "loss": 2.7233,
+      "step": 10814
+    },
+    {
+      "epoch": 0.45752601742956256,
+      "grad_norm": 0.1823870986700058,
+      "learning_rate": 0.001,
+      "loss": 2.4618,
+      "step": 10815
+    },
+    {
+      "epoch": 0.4575683221930789,
+      "grad_norm": 0.4014187455177307,
+      "learning_rate": 0.001,
+      "loss": 2.2749,
+      "step": 10816
+    },
+    {
+      "epoch": 0.4576106269565953,
+      "grad_norm": 0.17997021973133087,
+      "learning_rate": 0.001,
+      "loss": 3.2688,
+      "step": 10817
+    },
+    {
+      "epoch": 0.4576529317201117,
+      "grad_norm": 0.15339773893356323,
+      "learning_rate": 0.001,
+      "loss": 1.9151,
+      "step": 10818
+    },
+    {
+      "epoch": 0.45769523648362803,
+      "grad_norm": 0.36581626534461975,
+      "learning_rate": 0.001,
+      "loss": 1.4648,
+      "step": 10819
+    },
+    {
+      "epoch": 0.45773754124714444,
+      "grad_norm": 0.17709127068519592,
+      "learning_rate": 0.001,
+      "loss": 2.4527,
+      "step": 10820
+    },
+    {
+      "epoch": 0.4577798460106608,
+      "grad_norm": 0.4405263066291809,
+      "learning_rate": 0.001,
+      "loss": 2.7126,
+      "step": 10821
+    },
+    {
+      "epoch": 0.45782215077417715,
+      "grad_norm": 0.6151844263076782,
+      "learning_rate": 0.001,
+      "loss": 2.4587,
+      "step": 10822
+    },
+    {
+      "epoch": 0.45786445553769356,
+      "grad_norm": 0.18310926854610443,
+      "learning_rate": 0.001,
+      "loss": 1.8446,
+      "step": 10823
+    },
+    {
+      "epoch": 0.4579067603012099,
+      "grad_norm": 0.18017029762268066,
+      "learning_rate": 0.001,
+      "loss": 2.2304,
+      "step": 10824
+    },
+    {
+      "epoch": 0.45794906506472627,
+      "grad_norm": 0.9790677428245544,
+      "learning_rate": 0.001,
+      "loss": 3.1814,
+      "step": 10825
+    },
+    {
+      "epoch": 0.4579913698282427,
+      "grad_norm": 0.20925338566303253,
+      "learning_rate": 0.001,
+      "loss": 3.3553,
+      "step": 10826
+    },
+    {
+      "epoch": 0.45803367459175903,
+      "grad_norm": 0.17660897970199585,
+      "learning_rate": 0.001,
+      "loss": 2.1409,
+      "step": 10827
+    },
+    {
+      "epoch": 0.4580759793552754,
+      "grad_norm": 0.48942068219184875,
+      "learning_rate": 0.001,
+      "loss": 2.4424,
+      "step": 10828
+    },
+    {
+      "epoch": 0.4581182841187918,
+      "grad_norm": 0.2551790475845337,
+      "learning_rate": 0.001,
+      "loss": 1.7616,
+      "step": 10829
+    },
+    {
+      "epoch": 0.45816058888230815,
+      "grad_norm": 0.20451773703098297,
+      "learning_rate": 0.001,
+      "loss": 2.0661,
+      "step": 10830
+    },
+    {
+      "epoch": 0.4582028936458245,
+      "grad_norm": 0.2622445225715637,
+      "learning_rate": 0.001,
+      "loss": 1.9234,
+      "step": 10831
+    },
+    {
+      "epoch": 0.4582451984093409,
+      "grad_norm": 0.16220338642597198,
+      "learning_rate": 0.001,
+      "loss": 1.6778,
+      "step": 10832
+    },
+    {
+      "epoch": 0.45828750317285727,
+      "grad_norm": 0.2827407121658325,
+      "learning_rate": 0.001,
+      "loss": 2.8145,
+      "step": 10833
+    },
+    {
+      "epoch": 0.4583298079363736,
+      "grad_norm": 0.21268029510974884,
+      "learning_rate": 0.001,
+      "loss": 2.3485,
+      "step": 10834
+    },
+    {
+      "epoch": 0.45837211269989003,
+      "grad_norm": 0.23748518526554108,
+      "learning_rate": 0.001,
+      "loss": 2.6988,
+      "step": 10835
+    },
+    {
+      "epoch": 0.4584144174634064,
+      "grad_norm": 0.4688171148300171,
+      "learning_rate": 0.001,
+      "loss": 2.2598,
+      "step": 10836
+    },
+    {
+      "epoch": 0.45845672222692274,
+      "grad_norm": 2.4489855766296387,
+      "learning_rate": 0.001,
+      "loss": 2.3562,
+      "step": 10837
+    },
+    {
+      "epoch": 0.4584990269904391,
+      "grad_norm": 0.1477189064025879,
+      "learning_rate": 0.001,
+      "loss": 2.064,
+      "step": 10838
+    },
+    {
+      "epoch": 0.4585413317539555,
+      "grad_norm": 0.20878483355045319,
+      "learning_rate": 0.001,
+      "loss": 2.477,
+      "step": 10839
+    },
+    {
+      "epoch": 0.45858363651747186,
+      "grad_norm": 0.47598791122436523,
+      "learning_rate": 0.001,
+      "loss": 2.7616,
+      "step": 10840
+    },
+    {
+      "epoch": 0.4586259412809882,
+      "grad_norm": 0.16954365372657776,
+      "learning_rate": 0.001,
+      "loss": 3.0694,
+      "step": 10841
+    },
+    {
+      "epoch": 0.4586682460445046,
+      "grad_norm": 0.25619927048683167,
+      "learning_rate": 0.001,
+      "loss": 1.9389,
+      "step": 10842
+    },
+    {
+      "epoch": 0.458710550808021,
+      "grad_norm": 1.3417062759399414,
+      "learning_rate": 0.001,
+      "loss": 2.3684,
+      "step": 10843
+    },
+    {
+      "epoch": 0.45875285557153733,
+      "grad_norm": 1.6794036626815796,
+      "learning_rate": 0.001,
+      "loss": 2.2654,
+      "step": 10844
+    },
+    {
+      "epoch": 0.45879516033505374,
+      "grad_norm": 0.185557559132576,
+      "learning_rate": 0.001,
+      "loss": 2.1782,
+      "step": 10845
+    },
+    {
+      "epoch": 0.4588374650985701,
+      "grad_norm": 0.1776900589466095,
+      "learning_rate": 0.001,
+      "loss": 2.9762,
+      "step": 10846
+    },
+    {
+      "epoch": 0.45887976986208645,
+      "grad_norm": 0.19009187817573547,
+      "learning_rate": 0.001,
+      "loss": 2.3947,
+      "step": 10847
+    },
+    {
+      "epoch": 0.45892207462560286,
+      "grad_norm": 0.17857715487480164,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 10848
+    },
+    {
+      "epoch": 0.4589643793891192,
+      "grad_norm": 0.19130466878414154,
+      "learning_rate": 0.001,
+      "loss": 3.6092,
+      "step": 10849
+    },
+    {
+      "epoch": 0.45900668415263557,
+      "grad_norm": 1.0953785181045532,
+      "learning_rate": 0.001,
+      "loss": 2.8111,
+      "step": 10850
+    },
+    {
+      "epoch": 0.459048988916152,
+      "grad_norm": 0.1537635773420334,
+      "learning_rate": 0.001,
+      "loss": 1.9802,
+      "step": 10851
+    },
+    {
+      "epoch": 0.45909129367966833,
+      "grad_norm": 0.16897273063659668,
+      "learning_rate": 0.001,
+      "loss": 2.0585,
+      "step": 10852
+    },
+    {
+      "epoch": 0.4591335984431847,
+      "grad_norm": 0.9091619253158569,
+      "learning_rate": 0.001,
+      "loss": 2.9428,
+      "step": 10853
+    },
+    {
+      "epoch": 0.4591759032067011,
+      "grad_norm": 0.16677002608776093,
+      "learning_rate": 0.001,
+      "loss": 1.8913,
+      "step": 10854
+    },
+    {
+      "epoch": 0.45921820797021745,
+      "grad_norm": 0.21832408010959625,
+      "learning_rate": 0.001,
+      "loss": 3.3002,
+      "step": 10855
+    },
+    {
+      "epoch": 0.4592605127337338,
+      "grad_norm": 0.1814076453447342,
+      "learning_rate": 0.001,
+      "loss": 1.5207,
+      "step": 10856
+    },
+    {
+      "epoch": 0.4593028174972502,
+      "grad_norm": 0.17598986625671387,
+      "learning_rate": 0.001,
+      "loss": 2.3064,
+      "step": 10857
+    },
+    {
+      "epoch": 0.45934512226076657,
+      "grad_norm": 0.24575437605381012,
+      "learning_rate": 0.001,
+      "loss": 2.6707,
+      "step": 10858
+    },
+    {
+      "epoch": 0.4593874270242829,
+      "grad_norm": 0.1765861064195633,
+      "learning_rate": 0.001,
+      "loss": 1.8149,
+      "step": 10859
+    },
+    {
+      "epoch": 0.45942973178779933,
+      "grad_norm": 0.7690317034721375,
+      "learning_rate": 0.001,
+      "loss": 2.0892,
+      "step": 10860
+    },
+    {
+      "epoch": 0.4594720365513157,
+      "grad_norm": 0.1727459579706192,
+      "learning_rate": 0.001,
+      "loss": 1.9199,
+      "step": 10861
+    },
+    {
+      "epoch": 0.45951434131483204,
+      "grad_norm": 0.1738310158252716,
+      "learning_rate": 0.001,
+      "loss": 2.377,
+      "step": 10862
+    },
+    {
+      "epoch": 0.4595566460783484,
+      "grad_norm": 12.975968360900879,
+      "learning_rate": 0.001,
+      "loss": 2.9698,
+      "step": 10863
+    },
+    {
+      "epoch": 0.4595989508418648,
+      "grad_norm": 0.23020650446414948,
+      "learning_rate": 0.001,
+      "loss": 2.3748,
+      "step": 10864
+    },
+    {
+      "epoch": 0.45964125560538116,
+      "grad_norm": 0.14543908834457397,
+      "learning_rate": 0.001,
+      "loss": 1.7494,
+      "step": 10865
+    },
+    {
+      "epoch": 0.4596835603688975,
+      "grad_norm": 1.8263896703720093,
+      "learning_rate": 0.001,
+      "loss": 2.5401,
+      "step": 10866
+    },
+    {
+      "epoch": 0.4597258651324139,
+      "grad_norm": 0.23865966498851776,
+      "learning_rate": 0.001,
+      "loss": 2.27,
+      "step": 10867
+    },
+    {
+      "epoch": 0.4597681698959303,
+      "grad_norm": 0.7951701283454895,
+      "learning_rate": 0.001,
+      "loss": 1.5266,
+      "step": 10868
+    },
+    {
+      "epoch": 0.45981047465944663,
+      "grad_norm": 0.18856553733348846,
+      "learning_rate": 0.001,
+      "loss": 3.1963,
+      "step": 10869
+    },
+    {
+      "epoch": 0.45985277942296304,
+      "grad_norm": 0.16450615227222443,
+      "learning_rate": 0.001,
+      "loss": 2.5576,
+      "step": 10870
+    },
+    {
+      "epoch": 0.4598950841864794,
+      "grad_norm": 0.18288759887218475,
+      "learning_rate": 0.001,
+      "loss": 1.5935,
+      "step": 10871
+    },
+    {
+      "epoch": 0.45993738894999575,
+      "grad_norm": 0.16834130883216858,
+      "learning_rate": 0.001,
+      "loss": 1.5081,
+      "step": 10872
+    },
+    {
+      "epoch": 0.45997969371351216,
+      "grad_norm": 0.1692802906036377,
+      "learning_rate": 0.001,
+      "loss": 1.8773,
+      "step": 10873
+    },
+    {
+      "epoch": 0.4600219984770285,
+      "grad_norm": 0.1614452451467514,
+      "learning_rate": 0.001,
+      "loss": 2.5756,
+      "step": 10874
+    },
+    {
+      "epoch": 0.46006430324054487,
+      "grad_norm": 0.18225987255573273,
+      "learning_rate": 0.001,
+      "loss": 1.9148,
+      "step": 10875
+    },
+    {
+      "epoch": 0.4601066080040613,
+      "grad_norm": 0.16945257782936096,
+      "learning_rate": 0.001,
+      "loss": 1.9408,
+      "step": 10876
+    },
+    {
+      "epoch": 0.46014891276757763,
+      "grad_norm": 0.18923348188400269,
+      "learning_rate": 0.001,
+      "loss": 3.1651,
+      "step": 10877
+    },
+    {
+      "epoch": 0.460191217531094,
+      "grad_norm": 0.7960704565048218,
+      "learning_rate": 0.001,
+      "loss": 2.0982,
+      "step": 10878
+    },
+    {
+      "epoch": 0.4602335222946104,
+      "grad_norm": 0.16281262040138245,
+      "learning_rate": 0.001,
+      "loss": 2.7911,
+      "step": 10879
+    },
+    {
+      "epoch": 0.46027582705812675,
+      "grad_norm": 0.15493243932724,
+      "learning_rate": 0.001,
+      "loss": 1.7558,
+      "step": 10880
+    },
+    {
+      "epoch": 0.4603181318216431,
+      "grad_norm": 1.4312360286712646,
+      "learning_rate": 0.001,
+      "loss": 2.6279,
+      "step": 10881
+    },
+    {
+      "epoch": 0.4603604365851595,
+      "grad_norm": 0.18182626366615295,
+      "learning_rate": 0.001,
+      "loss": 2.3233,
+      "step": 10882
+    },
+    {
+      "epoch": 0.46040274134867587,
+      "grad_norm": 0.17061913013458252,
+      "learning_rate": 0.001,
+      "loss": 2.0113,
+      "step": 10883
+    },
+    {
+      "epoch": 0.4604450461121922,
+      "grad_norm": 1.3097604513168335,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 10884
+    },
+    {
+      "epoch": 0.4604873508757086,
+      "grad_norm": 0.1752685308456421,
+      "learning_rate": 0.001,
+      "loss": 2.0929,
+      "step": 10885
+    },
+    {
+      "epoch": 0.460529655639225,
+      "grad_norm": 0.3267187178134918,
+      "learning_rate": 0.001,
+      "loss": 2.5284,
+      "step": 10886
+    },
+    {
+      "epoch": 0.46057196040274134,
+      "grad_norm": 0.17250603437423706,
+      "learning_rate": 0.001,
+      "loss": 2.8212,
+      "step": 10887
+    },
+    {
+      "epoch": 0.4606142651662577,
+      "grad_norm": 1.1300095319747925,
+      "learning_rate": 0.001,
+      "loss": 2.0045,
+      "step": 10888
+    },
+    {
+      "epoch": 0.4606565699297741,
+      "grad_norm": 0.18050777912139893,
+      "learning_rate": 0.001,
+      "loss": 2.8097,
+      "step": 10889
+    },
+    {
+      "epoch": 0.46069887469329046,
+      "grad_norm": 0.16347844898700714,
+      "learning_rate": 0.001,
+      "loss": 2.505,
+      "step": 10890
+    },
+    {
+      "epoch": 0.4607411794568068,
+      "grad_norm": 4.939105033874512,
+      "learning_rate": 0.001,
+      "loss": 2.3762,
+      "step": 10891
+    },
+    {
+      "epoch": 0.4607834842203232,
+      "grad_norm": 13.591337203979492,
+      "learning_rate": 0.001,
+      "loss": 2.09,
+      "step": 10892
+    },
+    {
+      "epoch": 0.4608257889838396,
+      "grad_norm": 0.9323223829269409,
+      "learning_rate": 0.001,
+      "loss": 1.6851,
+      "step": 10893
+    },
+    {
+      "epoch": 0.46086809374735593,
+      "grad_norm": 0.7729269862174988,
+      "learning_rate": 0.001,
+      "loss": 1.4762,
+      "step": 10894
+    },
+    {
+      "epoch": 0.46091039851087234,
+      "grad_norm": 8.235560417175293,
+      "learning_rate": 0.001,
+      "loss": 2.5788,
+      "step": 10895
+    },
+    {
+      "epoch": 0.4609527032743887,
+      "grad_norm": 0.15181410312652588,
+      "learning_rate": 0.001,
+      "loss": 1.448,
+      "step": 10896
+    },
+    {
+      "epoch": 0.46099500803790505,
+      "grad_norm": 0.18965478241443634,
+      "learning_rate": 0.001,
+      "loss": 1.7609,
+      "step": 10897
+    },
+    {
+      "epoch": 0.46103731280142146,
+      "grad_norm": 1.7368152141571045,
+      "learning_rate": 0.001,
+      "loss": 3.6529,
+      "step": 10898
+    },
+    {
+      "epoch": 0.4610796175649378,
+      "grad_norm": 0.22795704007148743,
+      "learning_rate": 0.001,
+      "loss": 2.778,
+      "step": 10899
+    },
+    {
+      "epoch": 0.46112192232845417,
+      "grad_norm": 0.15993206202983856,
+      "learning_rate": 0.001,
+      "loss": 1.6612,
+      "step": 10900
+    },
+    {
+      "epoch": 0.4611642270919706,
+      "grad_norm": 0.1846051812171936,
+      "learning_rate": 0.001,
+      "loss": 1.5179,
+      "step": 10901
+    },
+    {
+      "epoch": 0.46120653185548693,
+      "grad_norm": 0.22482596337795258,
+      "learning_rate": 0.001,
+      "loss": 2.6861,
+      "step": 10902
+    },
+    {
+      "epoch": 0.4612488366190033,
+      "grad_norm": 10.483909606933594,
+      "learning_rate": 0.001,
+      "loss": 2.1016,
+      "step": 10903
+    },
+    {
+      "epoch": 0.4612911413825197,
+      "grad_norm": 0.45602327585220337,
+      "learning_rate": 0.001,
+      "loss": 2.7724,
+      "step": 10904
+    },
+    {
+      "epoch": 0.46133344614603605,
+      "grad_norm": 0.19587081670761108,
+      "learning_rate": 0.001,
+      "loss": 1.8976,
+      "step": 10905
+    },
+    {
+      "epoch": 0.4613757509095524,
+      "grad_norm": 0.22737310826778412,
+      "learning_rate": 0.001,
+      "loss": 3.4405,
+      "step": 10906
+    },
+    {
+      "epoch": 0.46141805567306876,
+      "grad_norm": 0.17420297861099243,
+      "learning_rate": 0.001,
+      "loss": 2.4252,
+      "step": 10907
+    },
+    {
+      "epoch": 0.46146036043658517,
+      "grad_norm": 0.15364967286586761,
+      "learning_rate": 0.001,
+      "loss": 1.766,
+      "step": 10908
+    },
+    {
+      "epoch": 0.4615026652001015,
+      "grad_norm": 1.44530189037323,
+      "learning_rate": 0.001,
+      "loss": 2.0774,
+      "step": 10909
+    },
+    {
+      "epoch": 0.4615449699636179,
+      "grad_norm": 0.16581588983535767,
+      "learning_rate": 0.001,
+      "loss": 1.5537,
+      "step": 10910
+    },
+    {
+      "epoch": 0.4615872747271343,
+      "grad_norm": 0.18303801119327545,
+      "learning_rate": 0.001,
+      "loss": 2.2337,
+      "step": 10911
+    },
+    {
+      "epoch": 0.46162957949065064,
+      "grad_norm": 0.1899944394826889,
+      "learning_rate": 0.001,
+      "loss": 1.3449,
+      "step": 10912
+    },
+    {
+      "epoch": 0.461671884254167,
+      "grad_norm": 1.3561519384384155,
+      "learning_rate": 0.001,
+      "loss": 2.3431,
+      "step": 10913
+    },
+    {
+      "epoch": 0.4617141890176834,
+      "grad_norm": 0.17157615721225739,
+      "learning_rate": 0.001,
+      "loss": 2.5933,
+      "step": 10914
+    },
+    {
+      "epoch": 0.46175649378119976,
+      "grad_norm": 1.1707208156585693,
+      "learning_rate": 0.001,
+      "loss": 2.3276,
+      "step": 10915
+    },
+    {
+      "epoch": 0.4617987985447161,
+      "grad_norm": 0.3634940981864929,
+      "learning_rate": 0.001,
+      "loss": 2.8591,
+      "step": 10916
+    },
+    {
+      "epoch": 0.4618411033082325,
+      "grad_norm": 0.3375161588191986,
+      "learning_rate": 0.001,
+      "loss": 2.2901,
+      "step": 10917
+    },
+    {
+      "epoch": 0.4618834080717489,
+      "grad_norm": 31.233200073242188,
+      "learning_rate": 0.001,
+      "loss": 2.6661,
+      "step": 10918
+    },
+    {
+      "epoch": 0.46192571283526523,
+      "grad_norm": 0.4107927083969116,
+      "learning_rate": 0.001,
+      "loss": 2.019,
+      "step": 10919
+    },
+    {
+      "epoch": 0.46196801759878164,
+      "grad_norm": 0.7090280652046204,
+      "learning_rate": 0.001,
+      "loss": 2.5535,
+      "step": 10920
+    },
+    {
+      "epoch": 0.462010322362298,
+      "grad_norm": 2.35701060295105,
+      "learning_rate": 0.001,
+      "loss": 1.7864,
+      "step": 10921
+    },
+    {
+      "epoch": 0.46205262712581435,
+      "grad_norm": 0.16083607077598572,
+      "learning_rate": 0.001,
+      "loss": 1.9362,
+      "step": 10922
+    },
+    {
+      "epoch": 0.46209493188933076,
+      "grad_norm": 0.20161986351013184,
+      "learning_rate": 0.001,
+      "loss": 2.4446,
+      "step": 10923
+    },
+    {
+      "epoch": 0.4621372366528471,
+      "grad_norm": 7.899982929229736,
+      "learning_rate": 0.001,
+      "loss": 3.1712,
+      "step": 10924
+    },
+    {
+      "epoch": 0.46217954141636347,
+      "grad_norm": 0.15877409279346466,
+      "learning_rate": 0.001,
+      "loss": 2.2344,
+      "step": 10925
+    },
+    {
+      "epoch": 0.4622218461798799,
+      "grad_norm": 0.19421695172786713,
+      "learning_rate": 0.001,
+      "loss": 2.574,
+      "step": 10926
+    },
+    {
+      "epoch": 0.46226415094339623,
+      "grad_norm": 0.5524720549583435,
+      "learning_rate": 0.001,
+      "loss": 2.2769,
+      "step": 10927
+    },
+    {
+      "epoch": 0.4623064557069126,
+      "grad_norm": 0.30834636092185974,
+      "learning_rate": 0.001,
+      "loss": 1.5401,
+      "step": 10928
+    },
+    {
+      "epoch": 0.46234876047042894,
+      "grad_norm": 0.17443600296974182,
+      "learning_rate": 0.001,
+      "loss": 3.5821,
+      "step": 10929
+    },
+    {
+      "epoch": 0.46239106523394535,
+      "grad_norm": 0.17415951192378998,
+      "learning_rate": 0.001,
+      "loss": 2.5898,
+      "step": 10930
+    },
+    {
+      "epoch": 0.4624333699974617,
+      "grad_norm": 0.14136327803134918,
+      "learning_rate": 0.001,
+      "loss": 2.4831,
+      "step": 10931
+    },
+    {
+      "epoch": 0.46247567476097806,
+      "grad_norm": 13.823705673217773,
+      "learning_rate": 0.001,
+      "loss": 3.3177,
+      "step": 10932
+    },
+    {
+      "epoch": 0.46251797952449447,
+      "grad_norm": 0.24503834545612335,
+      "learning_rate": 0.001,
+      "loss": 2.119,
+      "step": 10933
+    },
+    {
+      "epoch": 0.4625602842880108,
+      "grad_norm": 0.24650239944458008,
+      "learning_rate": 0.001,
+      "loss": 2.9394,
+      "step": 10934
+    },
+    {
+      "epoch": 0.4626025890515272,
+      "grad_norm": 0.193034827709198,
+      "learning_rate": 0.001,
+      "loss": 2.3874,
+      "step": 10935
+    },
+    {
+      "epoch": 0.4626448938150436,
+      "grad_norm": 0.16548699140548706,
+      "learning_rate": 0.001,
+      "loss": 2.2558,
+      "step": 10936
+    },
+    {
+      "epoch": 0.46268719857855994,
+      "grad_norm": 0.2028469741344452,
+      "learning_rate": 0.001,
+      "loss": 3.2924,
+      "step": 10937
+    },
+    {
+      "epoch": 0.4627295033420763,
+      "grad_norm": 0.25958749651908875,
+      "learning_rate": 0.001,
+      "loss": 1.914,
+      "step": 10938
+    },
+    {
+      "epoch": 0.4627718081055927,
+      "grad_norm": 0.2172216922044754,
+      "learning_rate": 0.001,
+      "loss": 2.3915,
+      "step": 10939
+    },
+    {
+      "epoch": 0.46281411286910906,
+      "grad_norm": 0.22597426176071167,
+      "learning_rate": 0.001,
+      "loss": 2.447,
+      "step": 10940
+    },
+    {
+      "epoch": 0.4628564176326254,
+      "grad_norm": 0.6521003842353821,
+      "learning_rate": 0.001,
+      "loss": 3.1807,
+      "step": 10941
+    },
+    {
+      "epoch": 0.4628987223961418,
+      "grad_norm": 0.2350475788116455,
+      "learning_rate": 0.001,
+      "loss": 2.4876,
+      "step": 10942
+    },
+    {
+      "epoch": 0.4629410271596582,
+      "grad_norm": 0.15000639855861664,
+      "learning_rate": 0.001,
+      "loss": 2.5798,
+      "step": 10943
+    },
+    {
+      "epoch": 0.46298333192317453,
+      "grad_norm": 13.369725227355957,
+      "learning_rate": 0.001,
+      "loss": 2.4465,
+      "step": 10944
+    },
+    {
+      "epoch": 0.46302563668669094,
+      "grad_norm": 0.21107615530490875,
+      "learning_rate": 0.001,
+      "loss": 3.5766,
+      "step": 10945
+    },
+    {
+      "epoch": 0.4630679414502073,
+      "grad_norm": 0.45922183990478516,
+      "learning_rate": 0.001,
+      "loss": 3.318,
+      "step": 10946
+    },
+    {
+      "epoch": 0.46311024621372365,
+      "grad_norm": 0.15640473365783691,
+      "learning_rate": 0.001,
+      "loss": 1.9649,
+      "step": 10947
+    },
+    {
+      "epoch": 0.46315255097724006,
+      "grad_norm": 0.17740632593631744,
+      "learning_rate": 0.001,
+      "loss": 2.4352,
+      "step": 10948
+    },
+    {
+      "epoch": 0.4631948557407564,
+      "grad_norm": 0.21806718409061432,
+      "learning_rate": 0.001,
+      "loss": 3.8975,
+      "step": 10949
+    },
+    {
+      "epoch": 0.46323716050427277,
+      "grad_norm": 0.17442677915096283,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 10950
+    },
+    {
+      "epoch": 0.4632794652677891,
+      "grad_norm": 0.18725450336933136,
+      "learning_rate": 0.001,
+      "loss": 1.6942,
+      "step": 10951
+    },
+    {
+      "epoch": 0.46332177003130554,
+      "grad_norm": 0.2647855281829834,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 10952
+    },
+    {
+      "epoch": 0.4633640747948219,
+      "grad_norm": 0.22588206827640533,
+      "learning_rate": 0.001,
+      "loss": 2.6015,
+      "step": 10953
+    },
+    {
+      "epoch": 0.46340637955833824,
+      "grad_norm": 0.18170830607414246,
+      "learning_rate": 0.001,
+      "loss": 2.4247,
+      "step": 10954
+    },
+    {
+      "epoch": 0.46344868432185465,
+      "grad_norm": 0.19047954678535461,
+      "learning_rate": 0.001,
+      "loss": 2.5552,
+      "step": 10955
+    },
+    {
+      "epoch": 0.463490989085371,
+      "grad_norm": 0.18378345668315887,
+      "learning_rate": 0.001,
+      "loss": 3.1469,
+      "step": 10956
+    },
+    {
+      "epoch": 0.46353329384888736,
+      "grad_norm": 1.0354962348937988,
+      "learning_rate": 0.001,
+      "loss": 3.1662,
+      "step": 10957
+    },
+    {
+      "epoch": 0.46357559861240377,
+      "grad_norm": 0.16644562780857086,
+      "learning_rate": 0.001,
+      "loss": 1.9066,
+      "step": 10958
+    },
+    {
+      "epoch": 0.4636179033759201,
+      "grad_norm": 0.17657527327537537,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 10959
+    },
+    {
+      "epoch": 0.4636602081394365,
+      "grad_norm": 0.43034595251083374,
+      "learning_rate": 0.001,
+      "loss": 2.2832,
+      "step": 10960
+    },
+    {
+      "epoch": 0.4637025129029529,
+      "grad_norm": 1.4906646013259888,
+      "learning_rate": 0.001,
+      "loss": 2.355,
+      "step": 10961
+    },
+    {
+      "epoch": 0.46374481766646924,
+      "grad_norm": 0.324938029050827,
+      "learning_rate": 0.001,
+      "loss": 1.9415,
+      "step": 10962
+    },
+    {
+      "epoch": 0.4637871224299856,
+      "grad_norm": 1.2835636138916016,
+      "learning_rate": 0.001,
+      "loss": 1.9729,
+      "step": 10963
+    },
+    {
+      "epoch": 0.463829427193502,
+      "grad_norm": 0.1738107055425644,
+      "learning_rate": 0.001,
+      "loss": 2.4461,
+      "step": 10964
+    },
+    {
+      "epoch": 0.46387173195701836,
+      "grad_norm": 0.23237141966819763,
+      "learning_rate": 0.001,
+      "loss": 1.4232,
+      "step": 10965
+    },
+    {
+      "epoch": 0.4639140367205347,
+      "grad_norm": 0.30422407388687134,
+      "learning_rate": 0.001,
+      "loss": 2.2135,
+      "step": 10966
+    },
+    {
+      "epoch": 0.4639563414840511,
+      "grad_norm": 0.17090842127799988,
+      "learning_rate": 0.001,
+      "loss": 2.0125,
+      "step": 10967
+    },
+    {
+      "epoch": 0.4639986462475675,
+      "grad_norm": 0.1509457677602768,
+      "learning_rate": 0.001,
+      "loss": 2.2202,
+      "step": 10968
+    },
+    {
+      "epoch": 0.46404095101108384,
+      "grad_norm": 0.16378089785575867,
+      "learning_rate": 0.001,
+      "loss": 2.301,
+      "step": 10969
+    },
+    {
+      "epoch": 0.46408325577460025,
+      "grad_norm": 6.196132659912109,
+      "learning_rate": 0.001,
+      "loss": 2.2314,
+      "step": 10970
+    },
+    {
+      "epoch": 0.4641255605381166,
+      "grad_norm": 0.44474777579307556,
+      "learning_rate": 0.001,
+      "loss": 2.6964,
+      "step": 10971
+    },
+    {
+      "epoch": 0.46416786530163295,
+      "grad_norm": 0.21871718764305115,
+      "learning_rate": 0.001,
+      "loss": 2.0155,
+      "step": 10972
+    },
+    {
+      "epoch": 0.46421017006514936,
+      "grad_norm": 0.24714459478855133,
+      "learning_rate": 0.001,
+      "loss": 1.7315,
+      "step": 10973
+    },
+    {
+      "epoch": 0.4642524748286657,
+      "grad_norm": 0.2337181717157364,
+      "learning_rate": 0.001,
+      "loss": 3.461,
+      "step": 10974
+    },
+    {
+      "epoch": 0.46429477959218207,
+      "grad_norm": 0.19878096878528595,
+      "learning_rate": 0.001,
+      "loss": 2.4174,
+      "step": 10975
+    },
+    {
+      "epoch": 0.4643370843556984,
+      "grad_norm": 0.19646792113780975,
+      "learning_rate": 0.001,
+      "loss": 3.2401,
+      "step": 10976
+    },
+    {
+      "epoch": 0.46437938911921484,
+      "grad_norm": 0.5441222786903381,
+      "learning_rate": 0.001,
+      "loss": 2.0292,
+      "step": 10977
+    },
+    {
+      "epoch": 0.4644216938827312,
+      "grad_norm": 0.19950753450393677,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 10978
+    },
+    {
+      "epoch": 0.46446399864624754,
+      "grad_norm": 0.301084965467453,
+      "learning_rate": 0.001,
+      "loss": 1.6781,
+      "step": 10979
+    },
+    {
+      "epoch": 0.46450630340976395,
+      "grad_norm": 0.1772916615009308,
+      "learning_rate": 0.001,
+      "loss": 1.5019,
+      "step": 10980
+    },
+    {
+      "epoch": 0.4645486081732803,
+      "grad_norm": 0.170380100607872,
+      "learning_rate": 0.001,
+      "loss": 1.8409,
+      "step": 10981
+    },
+    {
+      "epoch": 0.46459091293679666,
+      "grad_norm": 0.1674899011850357,
+      "learning_rate": 0.001,
+      "loss": 1.9776,
+      "step": 10982
+    },
+    {
+      "epoch": 0.4646332177003131,
+      "grad_norm": 0.20366886258125305,
+      "learning_rate": 0.001,
+      "loss": 2.4532,
+      "step": 10983
+    },
+    {
+      "epoch": 0.4646755224638294,
+      "grad_norm": 0.18768104910850525,
+      "learning_rate": 0.001,
+      "loss": 1.868,
+      "step": 10984
+    },
+    {
+      "epoch": 0.4647178272273458,
+      "grad_norm": 0.19560806453227997,
+      "learning_rate": 0.001,
+      "loss": 2.4912,
+      "step": 10985
+    },
+    {
+      "epoch": 0.4647601319908622,
+      "grad_norm": 0.3951105773448944,
+      "learning_rate": 0.001,
+      "loss": 2.1868,
+      "step": 10986
+    },
+    {
+      "epoch": 0.46480243675437855,
+      "grad_norm": 0.9056179523468018,
+      "learning_rate": 0.001,
+      "loss": 2.7611,
+      "step": 10987
+    },
+    {
+      "epoch": 0.4648447415178949,
+      "grad_norm": 0.1860157996416092,
+      "learning_rate": 0.001,
+      "loss": 2.0899,
+      "step": 10988
+    },
+    {
+      "epoch": 0.4648870462814113,
+      "grad_norm": 0.7894105911254883,
+      "learning_rate": 0.001,
+      "loss": 2.2895,
+      "step": 10989
+    },
+    {
+      "epoch": 0.46492935104492766,
+      "grad_norm": 0.16466152667999268,
+      "learning_rate": 0.001,
+      "loss": 2.6064,
+      "step": 10990
+    },
+    {
+      "epoch": 0.464971655808444,
+      "grad_norm": 0.17930325865745544,
+      "learning_rate": 0.001,
+      "loss": 1.7395,
+      "step": 10991
+    },
+    {
+      "epoch": 0.4650139605719604,
+      "grad_norm": 0.1949310451745987,
+      "learning_rate": 0.001,
+      "loss": 2.3786,
+      "step": 10992
+    },
+    {
+      "epoch": 0.4650562653354768,
+      "grad_norm": 0.17911165952682495,
+      "learning_rate": 0.001,
+      "loss": 1.5416,
+      "step": 10993
+    },
+    {
+      "epoch": 0.46509857009899314,
+      "grad_norm": 0.207133486866951,
+      "learning_rate": 0.001,
+      "loss": 2.3997,
+      "step": 10994
+    },
+    {
+      "epoch": 0.46514087486250955,
+      "grad_norm": 0.1573377251625061,
+      "learning_rate": 0.001,
+      "loss": 1.966,
+      "step": 10995
+    },
+    {
+      "epoch": 0.4651831796260259,
+      "grad_norm": 0.7414722442626953,
+      "learning_rate": 0.001,
+      "loss": 3.0765,
+      "step": 10996
+    },
+    {
+      "epoch": 0.46522548438954225,
+      "grad_norm": 0.17348453402519226,
+      "learning_rate": 0.001,
+      "loss": 2.8971,
+      "step": 10997
+    },
+    {
+      "epoch": 0.4652677891530586,
+      "grad_norm": 1.594288945198059,
+      "learning_rate": 0.001,
+      "loss": 2.3042,
+      "step": 10998
+    },
+    {
+      "epoch": 0.465310093916575,
+      "grad_norm": 0.24077658355236053,
+      "learning_rate": 0.001,
+      "loss": 2.2233,
+      "step": 10999
+    },
+    {
+      "epoch": 0.4653523986800914,
+      "grad_norm": 0.2647017538547516,
+      "learning_rate": 0.001,
+      "loss": 1.2681,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4653947034436077,
+      "grad_norm": 0.15793390572071075,
+      "learning_rate": 0.001,
+      "loss": 1.6003,
+      "step": 11001
+    },
+    {
+      "epoch": 0.46543700820712414,
+      "grad_norm": 0.1997009813785553,
+      "learning_rate": 0.001,
+      "loss": 1.7316,
+      "step": 11002
+    },
+    {
+      "epoch": 0.4654793129706405,
+      "grad_norm": 0.18965008854866028,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 11003
+    },
+    {
+      "epoch": 0.46552161773415685,
+      "grad_norm": 0.24327543377876282,
+      "learning_rate": 0.001,
+      "loss": 2.5611,
+      "step": 11004
+    },
+    {
+      "epoch": 0.46556392249767325,
+      "grad_norm": 6.968198776245117,
+      "learning_rate": 0.001,
+      "loss": 2.3886,
+      "step": 11005
+    },
+    {
+      "epoch": 0.4656062272611896,
+      "grad_norm": 0.2511080503463745,
+      "learning_rate": 0.001,
+      "loss": 2.9208,
+      "step": 11006
+    },
+    {
+      "epoch": 0.46564853202470596,
+      "grad_norm": 0.1881697177886963,
+      "learning_rate": 0.001,
+      "loss": 1.5614,
+      "step": 11007
+    },
+    {
+      "epoch": 0.4656908367882224,
+      "grad_norm": 0.4397967755794525,
+      "learning_rate": 0.001,
+      "loss": 3.3356,
+      "step": 11008
+    },
+    {
+      "epoch": 0.4657331415517387,
+      "grad_norm": 0.2115083485841751,
+      "learning_rate": 0.001,
+      "loss": 2.1108,
+      "step": 11009
+    },
+    {
+      "epoch": 0.4657754463152551,
+      "grad_norm": 0.17709064483642578,
+      "learning_rate": 0.001,
+      "loss": 3.367,
+      "step": 11010
+    },
+    {
+      "epoch": 0.4658177510787715,
+      "grad_norm": 0.263375848531723,
+      "learning_rate": 0.001,
+      "loss": 2.0427,
+      "step": 11011
+    },
+    {
+      "epoch": 0.46586005584228785,
+      "grad_norm": 0.20178444683551788,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 11012
+    },
+    {
+      "epoch": 0.4659023606058042,
+      "grad_norm": 0.19885046780109406,
+      "learning_rate": 0.001,
+      "loss": 3.0593,
+      "step": 11013
+    },
+    {
+      "epoch": 0.4659446653693206,
+      "grad_norm": 0.27875369787216187,
+      "learning_rate": 0.001,
+      "loss": 3.4957,
+      "step": 11014
+    },
+    {
+      "epoch": 0.46598697013283696,
+      "grad_norm": 0.21964240074157715,
+      "learning_rate": 0.001,
+      "loss": 2.5046,
+      "step": 11015
+    },
+    {
+      "epoch": 0.4660292748963533,
+      "grad_norm": 0.16891005635261536,
+      "learning_rate": 0.001,
+      "loss": 2.9566,
+      "step": 11016
+    },
+    {
+      "epoch": 0.46607157965986973,
+      "grad_norm": 0.17071029543876648,
+      "learning_rate": 0.001,
+      "loss": 2.141,
+      "step": 11017
+    },
+    {
+      "epoch": 0.4661138844233861,
+      "grad_norm": 17.83469009399414,
+      "learning_rate": 0.001,
+      "loss": 2.0753,
+      "step": 11018
+    },
+    {
+      "epoch": 0.46615618918690244,
+      "grad_norm": 0.20334742963314056,
+      "learning_rate": 0.001,
+      "loss": 3.0571,
+      "step": 11019
+    },
+    {
+      "epoch": 0.4661984939504188,
+      "grad_norm": 0.1513107866048813,
+      "learning_rate": 0.001,
+      "loss": 2.1704,
+      "step": 11020
+    },
+    {
+      "epoch": 0.4662407987139352,
+      "grad_norm": 0.15699905157089233,
+      "learning_rate": 0.001,
+      "loss": 2.4362,
+      "step": 11021
+    },
+    {
+      "epoch": 0.46628310347745155,
+      "grad_norm": 0.25063201785087585,
+      "learning_rate": 0.001,
+      "loss": 2.6234,
+      "step": 11022
+    },
+    {
+      "epoch": 0.4663254082409679,
+      "grad_norm": 3.148904323577881,
+      "learning_rate": 0.001,
+      "loss": 3.4478,
+      "step": 11023
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.22315865755081177,
+      "learning_rate": 0.001,
+      "loss": 2.1858,
+      "step": 11024
+    },
+    {
+      "epoch": 0.4664100177680007,
+      "grad_norm": 0.24025224149227142,
+      "learning_rate": 0.001,
+      "loss": 2.9627,
+      "step": 11025
+    },
+    {
+      "epoch": 0.466452322531517,
+      "grad_norm": 0.2143600881099701,
+      "learning_rate": 0.001,
+      "loss": 1.9114,
+      "step": 11026
+    },
+    {
+      "epoch": 0.46649462729503344,
+      "grad_norm": 0.5619837641716003,
+      "learning_rate": 0.001,
+      "loss": 2.6092,
+      "step": 11027
+    },
+    {
+      "epoch": 0.4665369320585498,
+      "grad_norm": 0.2009797990322113,
+      "learning_rate": 0.001,
+      "loss": 1.9033,
+      "step": 11028
+    },
+    {
+      "epoch": 0.46657923682206615,
+      "grad_norm": 1.4359869956970215,
+      "learning_rate": 0.001,
+      "loss": 2.7386,
+      "step": 11029
+    },
+    {
+      "epoch": 0.46662154158558256,
+      "grad_norm": 0.15855303406715393,
+      "learning_rate": 0.001,
+      "loss": 3.3832,
+      "step": 11030
+    },
+    {
+      "epoch": 0.4666638463490989,
+      "grad_norm": 0.1745101362466812,
+      "learning_rate": 0.001,
+      "loss": 2.5427,
+      "step": 11031
+    },
+    {
+      "epoch": 0.46670615111261526,
+      "grad_norm": 0.20111802220344543,
+      "learning_rate": 0.001,
+      "loss": 2.204,
+      "step": 11032
+    },
+    {
+      "epoch": 0.4667484558761317,
+      "grad_norm": 0.22334951162338257,
+      "learning_rate": 0.001,
+      "loss": 2.4836,
+      "step": 11033
+    },
+    {
+      "epoch": 0.46679076063964803,
+      "grad_norm": 0.17813622951507568,
+      "learning_rate": 0.001,
+      "loss": 2.9059,
+      "step": 11034
+    },
+    {
+      "epoch": 0.4668330654031644,
+      "grad_norm": 0.19768601655960083,
+      "learning_rate": 0.001,
+      "loss": 1.7435,
+      "step": 11035
+    },
+    {
+      "epoch": 0.4668753701666808,
+      "grad_norm": 0.4055560529232025,
+      "learning_rate": 0.001,
+      "loss": 1.8149,
+      "step": 11036
+    },
+    {
+      "epoch": 0.46691767493019715,
+      "grad_norm": 0.14512808620929718,
+      "learning_rate": 0.001,
+      "loss": 2.3372,
+      "step": 11037
+    },
+    {
+      "epoch": 0.4669599796937135,
+      "grad_norm": 0.2370501309633255,
+      "learning_rate": 0.001,
+      "loss": 2.0583,
+      "step": 11038
+    },
+    {
+      "epoch": 0.4670022844572299,
+      "grad_norm": 0.16900447010993958,
+      "learning_rate": 0.001,
+      "loss": 1.8195,
+      "step": 11039
+    },
+    {
+      "epoch": 0.46704458922074626,
+      "grad_norm": 0.16092944145202637,
+      "learning_rate": 0.001,
+      "loss": 2.3941,
+      "step": 11040
+    },
+    {
+      "epoch": 0.4670868939842626,
+      "grad_norm": 0.1543259471654892,
+      "learning_rate": 0.001,
+      "loss": 1.7284,
+      "step": 11041
+    },
+    {
+      "epoch": 0.467129198747779,
+      "grad_norm": 0.16476312279701233,
+      "learning_rate": 0.001,
+      "loss": 2.297,
+      "step": 11042
+    },
+    {
+      "epoch": 0.4671715035112954,
+      "grad_norm": 0.1504514068365097,
+      "learning_rate": 0.001,
+      "loss": 2.2147,
+      "step": 11043
+    },
+    {
+      "epoch": 0.46721380827481174,
+      "grad_norm": 0.1745976358652115,
+      "learning_rate": 0.001,
+      "loss": 2.5716,
+      "step": 11044
+    },
+    {
+      "epoch": 0.4672561130383281,
+      "grad_norm": 0.17237746715545654,
+      "learning_rate": 0.001,
+      "loss": 2.1571,
+      "step": 11045
+    },
+    {
+      "epoch": 0.4672984178018445,
+      "grad_norm": 0.14179447293281555,
+      "learning_rate": 0.001,
+      "loss": 1.7693,
+      "step": 11046
+    },
+    {
+      "epoch": 0.46734072256536086,
+      "grad_norm": 0.2136073112487793,
+      "learning_rate": 0.001,
+      "loss": 1.9455,
+      "step": 11047
+    },
+    {
+      "epoch": 0.4673830273288772,
+      "grad_norm": 0.15894265472888947,
+      "learning_rate": 0.001,
+      "loss": 1.9635,
+      "step": 11048
+    },
+    {
+      "epoch": 0.4674253320923936,
+      "grad_norm": 1.5329841375350952,
+      "learning_rate": 0.001,
+      "loss": 2.4782,
+      "step": 11049
+    },
+    {
+      "epoch": 0.46746763685591,
+      "grad_norm": 0.17458941042423248,
+      "learning_rate": 0.001,
+      "loss": 2.1225,
+      "step": 11050
+    },
+    {
+      "epoch": 0.46750994161942633,
+      "grad_norm": 0.21708954870700836,
+      "learning_rate": 0.001,
+      "loss": 2.2734,
+      "step": 11051
+    },
+    {
+      "epoch": 0.46755224638294274,
+      "grad_norm": 0.7924477458000183,
+      "learning_rate": 0.001,
+      "loss": 1.9928,
+      "step": 11052
+    },
+    {
+      "epoch": 0.4675945511464591,
+      "grad_norm": 0.16856947541236877,
+      "learning_rate": 0.001,
+      "loss": 2.1516,
+      "step": 11053
+    },
+    {
+      "epoch": 0.46763685590997545,
+      "grad_norm": 0.15970207750797272,
+      "learning_rate": 0.001,
+      "loss": 2.7033,
+      "step": 11054
+    },
+    {
+      "epoch": 0.46767916067349186,
+      "grad_norm": 8.105451583862305,
+      "learning_rate": 0.001,
+      "loss": 2.0742,
+      "step": 11055
+    },
+    {
+      "epoch": 0.4677214654370082,
+      "grad_norm": 0.1840921938419342,
+      "learning_rate": 0.001,
+      "loss": 1.9796,
+      "step": 11056
+    },
+    {
+      "epoch": 0.46776377020052456,
+      "grad_norm": 0.15253804624080658,
+      "learning_rate": 0.001,
+      "loss": 2.0105,
+      "step": 11057
+    },
+    {
+      "epoch": 0.467806074964041,
+      "grad_norm": 5.88224458694458,
+      "learning_rate": 0.001,
+      "loss": 1.9661,
+      "step": 11058
+    },
+    {
+      "epoch": 0.46784837972755733,
+      "grad_norm": 0.16145947575569153,
+      "learning_rate": 0.001,
+      "loss": 2.1938,
+      "step": 11059
+    },
+    {
+      "epoch": 0.4678906844910737,
+      "grad_norm": 0.19684232771396637,
+      "learning_rate": 0.001,
+      "loss": 1.6197,
+      "step": 11060
+    },
+    {
+      "epoch": 0.4679329892545901,
+      "grad_norm": 45.53775405883789,
+      "learning_rate": 0.001,
+      "loss": 1.5319,
+      "step": 11061
+    },
+    {
+      "epoch": 0.46797529401810645,
+      "grad_norm": 2.51239275932312,
+      "learning_rate": 0.001,
+      "loss": 2.2339,
+      "step": 11062
+    },
+    {
+      "epoch": 0.4680175987816228,
+      "grad_norm": 0.19354106485843658,
+      "learning_rate": 0.001,
+      "loss": 2.2878,
+      "step": 11063
+    },
+    {
+      "epoch": 0.46805990354513916,
+      "grad_norm": 63.523895263671875,
+      "learning_rate": 0.001,
+      "loss": 2.1583,
+      "step": 11064
+    },
+    {
+      "epoch": 0.46810220830865557,
+      "grad_norm": 0.21123424172401428,
+      "learning_rate": 0.001,
+      "loss": 1.6651,
+      "step": 11065
+    },
+    {
+      "epoch": 0.4681445130721719,
+      "grad_norm": 0.5979301333427429,
+      "learning_rate": 0.001,
+      "loss": 1.8025,
+      "step": 11066
+    },
+    {
+      "epoch": 0.4681868178356883,
+      "grad_norm": 0.15575318038463593,
+      "learning_rate": 0.001,
+      "loss": 2.2109,
+      "step": 11067
+    },
+    {
+      "epoch": 0.4682291225992047,
+      "grad_norm": 0.2060658186674118,
+      "learning_rate": 0.001,
+      "loss": 1.9252,
+      "step": 11068
+    },
+    {
+      "epoch": 0.46827142736272104,
+      "grad_norm": 0.19070769846439362,
+      "learning_rate": 0.001,
+      "loss": 2.1284,
+      "step": 11069
+    },
+    {
+      "epoch": 0.4683137321262374,
+      "grad_norm": 0.14260248839855194,
+      "learning_rate": 0.001,
+      "loss": 1.8289,
+      "step": 11070
+    },
+    {
+      "epoch": 0.4683560368897538,
+      "grad_norm": 0.2190844863653183,
+      "learning_rate": 0.001,
+      "loss": 2.1024,
+      "step": 11071
+    },
+    {
+      "epoch": 0.46839834165327016,
+      "grad_norm": 0.1667780727148056,
+      "learning_rate": 0.001,
+      "loss": 2.5757,
+      "step": 11072
+    },
+    {
+      "epoch": 0.4684406464167865,
+      "grad_norm": 0.19789689779281616,
+      "learning_rate": 0.001,
+      "loss": 2.0722,
+      "step": 11073
+    },
+    {
+      "epoch": 0.4684829511803029,
+      "grad_norm": 0.16835686564445496,
+      "learning_rate": 0.001,
+      "loss": 1.8042,
+      "step": 11074
+    },
+    {
+      "epoch": 0.4685252559438193,
+      "grad_norm": 0.9263663291931152,
+      "learning_rate": 0.001,
+      "loss": 2.0782,
+      "step": 11075
+    },
+    {
+      "epoch": 0.46856756070733563,
+      "grad_norm": 0.21614162623882294,
+      "learning_rate": 0.001,
+      "loss": 2.924,
+      "step": 11076
+    },
+    {
+      "epoch": 0.46860986547085204,
+      "grad_norm": 3.207810163497925,
+      "learning_rate": 0.001,
+      "loss": 2.6613,
+      "step": 11077
+    },
+    {
+      "epoch": 0.4686521702343684,
+      "grad_norm": 0.5324118733406067,
+      "learning_rate": 0.001,
+      "loss": 2.3755,
+      "step": 11078
+    },
+    {
+      "epoch": 0.46869447499788475,
+      "grad_norm": 0.17789721488952637,
+      "learning_rate": 0.001,
+      "loss": 2.0421,
+      "step": 11079
+    },
+    {
+      "epoch": 0.46873677976140116,
+      "grad_norm": 0.17907847464084625,
+      "learning_rate": 0.001,
+      "loss": 2.6084,
+      "step": 11080
+    },
+    {
+      "epoch": 0.4687790845249175,
+      "grad_norm": 0.23135465383529663,
+      "learning_rate": 0.001,
+      "loss": 1.7174,
+      "step": 11081
+    },
+    {
+      "epoch": 0.46882138928843387,
+      "grad_norm": 0.17100471258163452,
+      "learning_rate": 0.001,
+      "loss": 2.5313,
+      "step": 11082
+    },
+    {
+      "epoch": 0.4688636940519503,
+      "grad_norm": 0.16610606014728546,
+      "learning_rate": 0.001,
+      "loss": 2.9865,
+      "step": 11083
+    },
+    {
+      "epoch": 0.46890599881546663,
+      "grad_norm": 0.15432819724082947,
+      "learning_rate": 0.001,
+      "loss": 3.2703,
+      "step": 11084
+    },
+    {
+      "epoch": 0.468948303578983,
+      "grad_norm": 0.18594351410865784,
+      "learning_rate": 0.001,
+      "loss": 3.4266,
+      "step": 11085
+    },
+    {
+      "epoch": 0.46899060834249934,
+      "grad_norm": 6.842674732208252,
+      "learning_rate": 0.001,
+      "loss": 2.4358,
+      "step": 11086
+    },
+    {
+      "epoch": 0.46903291310601575,
+      "grad_norm": 0.17697936296463013,
+      "learning_rate": 0.001,
+      "loss": 1.7158,
+      "step": 11087
+    },
+    {
+      "epoch": 0.4690752178695321,
+      "grad_norm": 0.17671042680740356,
+      "learning_rate": 0.001,
+      "loss": 2.4786,
+      "step": 11088
+    },
+    {
+      "epoch": 0.46911752263304846,
+      "grad_norm": 1.0439887046813965,
+      "learning_rate": 0.001,
+      "loss": 2.1289,
+      "step": 11089
+    },
+    {
+      "epoch": 0.46915982739656487,
+      "grad_norm": 0.21606509387493134,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4692021321600812,
+      "grad_norm": 0.24721235036849976,
+      "learning_rate": 0.001,
+      "loss": 2.8101,
+      "step": 11091
+    },
+    {
+      "epoch": 0.4692444369235976,
+      "grad_norm": 0.21566319465637207,
+      "learning_rate": 0.001,
+      "loss": 2.1605,
+      "step": 11092
+    },
+    {
+      "epoch": 0.469286741687114,
+      "grad_norm": 0.5193756818771362,
+      "learning_rate": 0.001,
+      "loss": 2.787,
+      "step": 11093
+    },
+    {
+      "epoch": 0.46932904645063034,
+      "grad_norm": 0.20842444896697998,
+      "learning_rate": 0.001,
+      "loss": 1.9445,
+      "step": 11094
+    },
+    {
+      "epoch": 0.4693713512141467,
+      "grad_norm": 0.1922614574432373,
+      "learning_rate": 0.001,
+      "loss": 2.0826,
+      "step": 11095
+    },
+    {
+      "epoch": 0.4694136559776631,
+      "grad_norm": 0.9516851305961609,
+      "learning_rate": 0.001,
+      "loss": 1.9346,
+      "step": 11096
+    },
+    {
+      "epoch": 0.46945596074117946,
+      "grad_norm": 0.18452809751033783,
+      "learning_rate": 0.001,
+      "loss": 2.7229,
+      "step": 11097
+    },
+    {
+      "epoch": 0.4694982655046958,
+      "grad_norm": 0.21180051565170288,
+      "learning_rate": 0.001,
+      "loss": 2.497,
+      "step": 11098
+    },
+    {
+      "epoch": 0.4695405702682122,
+      "grad_norm": 3.082894802093506,
+      "learning_rate": 0.001,
+      "loss": 1.9389,
+      "step": 11099
+    },
+    {
+      "epoch": 0.4695828750317286,
+      "grad_norm": 0.17565667629241943,
+      "learning_rate": 0.001,
+      "loss": 1.4547,
+      "step": 11100
+    },
+    {
+      "epoch": 0.46962517979524493,
+      "grad_norm": 0.2029939591884613,
+      "learning_rate": 0.001,
+      "loss": 2.4275,
+      "step": 11101
+    },
+    {
+      "epoch": 0.46966748455876134,
+      "grad_norm": 0.20764219760894775,
+      "learning_rate": 0.001,
+      "loss": 2.1751,
+      "step": 11102
+    },
+    {
+      "epoch": 0.4697097893222777,
+      "grad_norm": 0.19430102407932281,
+      "learning_rate": 0.001,
+      "loss": 1.736,
+      "step": 11103
+    },
+    {
+      "epoch": 0.46975209408579405,
+      "grad_norm": 0.180740088224411,
+      "learning_rate": 0.001,
+      "loss": 1.6614,
+      "step": 11104
+    },
+    {
+      "epoch": 0.46979439884931046,
+      "grad_norm": 8.314878463745117,
+      "learning_rate": 0.001,
+      "loss": 2.2919,
+      "step": 11105
+    },
+    {
+      "epoch": 0.4698367036128268,
+      "grad_norm": 0.21296004951000214,
+      "learning_rate": 0.001,
+      "loss": 2.1662,
+      "step": 11106
+    },
+    {
+      "epoch": 0.46987900837634317,
+      "grad_norm": 0.21107828617095947,
+      "learning_rate": 0.001,
+      "loss": 2.2709,
+      "step": 11107
+    },
+    {
+      "epoch": 0.4699213131398596,
+      "grad_norm": 0.19116267561912537,
+      "learning_rate": 0.001,
+      "loss": 1.8983,
+      "step": 11108
+    },
+    {
+      "epoch": 0.46996361790337593,
+      "grad_norm": 0.17168934643268585,
+      "learning_rate": 0.001,
+      "loss": 1.8434,
+      "step": 11109
+    },
+    {
+      "epoch": 0.4700059226668923,
+      "grad_norm": 0.18456289172172546,
+      "learning_rate": 0.001,
+      "loss": 2.4683,
+      "step": 11110
+    },
+    {
+      "epoch": 0.47004822743040864,
+      "grad_norm": 0.874439001083374,
+      "learning_rate": 0.001,
+      "loss": 3.1001,
+      "step": 11111
+    },
+    {
+      "epoch": 0.47009053219392505,
+      "grad_norm": 0.16725631058216095,
+      "learning_rate": 0.001,
+      "loss": 1.8565,
+      "step": 11112
+    },
+    {
+      "epoch": 0.4701328369574414,
+      "grad_norm": 0.1976451575756073,
+      "learning_rate": 0.001,
+      "loss": 2.5576,
+      "step": 11113
+    },
+    {
+      "epoch": 0.47017514172095776,
+      "grad_norm": 0.17343725264072418,
+      "learning_rate": 0.001,
+      "loss": 2.7862,
+      "step": 11114
+    },
+    {
+      "epoch": 0.47021744648447417,
+      "grad_norm": 0.22081388533115387,
+      "learning_rate": 0.001,
+      "loss": 2.4001,
+      "step": 11115
+    },
+    {
+      "epoch": 0.4702597512479905,
+      "grad_norm": 0.375754177570343,
+      "learning_rate": 0.001,
+      "loss": 2.0034,
+      "step": 11116
+    },
+    {
+      "epoch": 0.4703020560115069,
+      "grad_norm": 0.20346909761428833,
+      "learning_rate": 0.001,
+      "loss": 2.6006,
+      "step": 11117
+    },
+    {
+      "epoch": 0.4703443607750233,
+      "grad_norm": 0.21501335501670837,
+      "learning_rate": 0.001,
+      "loss": 1.8364,
+      "step": 11118
+    },
+    {
+      "epoch": 0.47038666553853964,
+      "grad_norm": 2.512514352798462,
+      "learning_rate": 0.001,
+      "loss": 2.0339,
+      "step": 11119
+    },
+    {
+      "epoch": 0.470428970302056,
+      "grad_norm": 0.18703396618366241,
+      "learning_rate": 0.001,
+      "loss": 2.2686,
+      "step": 11120
+    },
+    {
+      "epoch": 0.4704712750655724,
+      "grad_norm": 0.2161937803030014,
+      "learning_rate": 0.001,
+      "loss": 2.3499,
+      "step": 11121
+    },
+    {
+      "epoch": 0.47051357982908876,
+      "grad_norm": 0.22765906155109406,
+      "learning_rate": 0.001,
+      "loss": 2.2785,
+      "step": 11122
+    },
+    {
+      "epoch": 0.4705558845926051,
+      "grad_norm": 0.16340266168117523,
+      "learning_rate": 0.001,
+      "loss": 2.2066,
+      "step": 11123
+    },
+    {
+      "epoch": 0.4705981893561215,
+      "grad_norm": 0.17582248151302338,
+      "learning_rate": 0.001,
+      "loss": 3.2365,
+      "step": 11124
+    },
+    {
+      "epoch": 0.4706404941196379,
+      "grad_norm": 2.7968180179595947,
+      "learning_rate": 0.001,
+      "loss": 2.1311,
+      "step": 11125
+    },
+    {
+      "epoch": 0.47068279888315423,
+      "grad_norm": 0.19796128571033478,
+      "learning_rate": 0.001,
+      "loss": 2.6811,
+      "step": 11126
+    },
+    {
+      "epoch": 0.47072510364667064,
+      "grad_norm": 0.3247911334037781,
+      "learning_rate": 0.001,
+      "loss": 2.0883,
+      "step": 11127
+    },
+    {
+      "epoch": 0.470767408410187,
+      "grad_norm": 0.21730183064937592,
+      "learning_rate": 0.001,
+      "loss": 2.1616,
+      "step": 11128
+    },
+    {
+      "epoch": 0.47080971317370335,
+      "grad_norm": 0.226370707154274,
+      "learning_rate": 0.001,
+      "loss": 3.4228,
+      "step": 11129
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 2.1633243560791016,
+      "learning_rate": 0.001,
+      "loss": 2.2012,
+      "step": 11130
+    },
+    {
+      "epoch": 0.4708943227007361,
+      "grad_norm": 0.2938755750656128,
+      "learning_rate": 0.001,
+      "loss": 3.1938,
+      "step": 11131
+    },
+    {
+      "epoch": 0.47093662746425247,
+      "grad_norm": 0.15652383863925934,
+      "learning_rate": 0.001,
+      "loss": 1.8906,
+      "step": 11132
+    },
+    {
+      "epoch": 0.4709789322277688,
+      "grad_norm": 0.17891067266464233,
+      "learning_rate": 0.001,
+      "loss": 3.3451,
+      "step": 11133
+    },
+    {
+      "epoch": 0.47102123699128523,
+      "grad_norm": 0.47045284509658813,
+      "learning_rate": 0.001,
+      "loss": 2.3286,
+      "step": 11134
+    },
+    {
+      "epoch": 0.4710635417548016,
+      "grad_norm": 0.3203991949558258,
+      "learning_rate": 0.001,
+      "loss": 2.0339,
+      "step": 11135
+    },
+    {
+      "epoch": 0.47110584651831794,
+      "grad_norm": 0.1709599643945694,
+      "learning_rate": 0.001,
+      "loss": 1.9023,
+      "step": 11136
+    },
+    {
+      "epoch": 0.47114815128183435,
+      "grad_norm": 0.2922183871269226,
+      "learning_rate": 0.001,
+      "loss": 2.1608,
+      "step": 11137
+    },
+    {
+      "epoch": 0.4711904560453507,
+      "grad_norm": 8.45193862915039,
+      "learning_rate": 0.001,
+      "loss": 1.8432,
+      "step": 11138
+    },
+    {
+      "epoch": 0.47123276080886706,
+      "grad_norm": 0.18489998579025269,
+      "learning_rate": 0.001,
+      "loss": 2.5255,
+      "step": 11139
+    },
+    {
+      "epoch": 0.47127506557238347,
+      "grad_norm": 0.5812740325927734,
+      "learning_rate": 0.001,
+      "loss": 2.3951,
+      "step": 11140
+    },
+    {
+      "epoch": 0.4713173703358998,
+      "grad_norm": 0.18123511970043182,
+      "learning_rate": 0.001,
+      "loss": 1.8583,
+      "step": 11141
+    },
+    {
+      "epoch": 0.4713596750994162,
+      "grad_norm": 0.22358356416225433,
+      "learning_rate": 0.001,
+      "loss": 2.1568,
+      "step": 11142
+    },
+    {
+      "epoch": 0.4714019798629326,
+      "grad_norm": 0.363231360912323,
+      "learning_rate": 0.001,
+      "loss": 1.5782,
+      "step": 11143
+    },
+    {
+      "epoch": 0.47144428462644894,
+      "grad_norm": 0.2926989793777466,
+      "learning_rate": 0.001,
+      "loss": 2.6199,
+      "step": 11144
+    },
+    {
+      "epoch": 0.4714865893899653,
+      "grad_norm": 0.17961940169334412,
+      "learning_rate": 0.001,
+      "loss": 1.9892,
+      "step": 11145
+    },
+    {
+      "epoch": 0.4715288941534817,
+      "grad_norm": 0.17067760229110718,
+      "learning_rate": 0.001,
+      "loss": 2.2848,
+      "step": 11146
+    },
+    {
+      "epoch": 0.47157119891699806,
+      "grad_norm": 0.16751837730407715,
+      "learning_rate": 0.001,
+      "loss": 1.7813,
+      "step": 11147
+    },
+    {
+      "epoch": 0.4716135036805144,
+      "grad_norm": 0.20085522532463074,
+      "learning_rate": 0.001,
+      "loss": 2.9713,
+      "step": 11148
+    },
+    {
+      "epoch": 0.4716558084440308,
+      "grad_norm": 0.20381022989749908,
+      "learning_rate": 0.001,
+      "loss": 3.0651,
+      "step": 11149
+    },
+    {
+      "epoch": 0.4716981132075472,
+      "grad_norm": 0.2398488074541092,
+      "learning_rate": 0.001,
+      "loss": 2.5814,
+      "step": 11150
+    },
+    {
+      "epoch": 0.47174041797106353,
+      "grad_norm": 1.2999179363250732,
+      "learning_rate": 0.001,
+      "loss": 1.6057,
+      "step": 11151
+    },
+    {
+      "epoch": 0.47178272273457994,
+      "grad_norm": 0.7444708347320557,
+      "learning_rate": 0.001,
+      "loss": 2.5191,
+      "step": 11152
+    },
+    {
+      "epoch": 0.4718250274980963,
+      "grad_norm": 0.16335336863994598,
+      "learning_rate": 0.001,
+      "loss": 2.7914,
+      "step": 11153
+    },
+    {
+      "epoch": 0.47186733226161265,
+      "grad_norm": 0.5560107827186584,
+      "learning_rate": 0.001,
+      "loss": 2.2272,
+      "step": 11154
+    },
+    {
+      "epoch": 0.471909637025129,
+      "grad_norm": 0.16256298124790192,
+      "learning_rate": 0.001,
+      "loss": 1.6319,
+      "step": 11155
+    },
+    {
+      "epoch": 0.4719519417886454,
+      "grad_norm": 0.23074427247047424,
+      "learning_rate": 0.001,
+      "loss": 2.5899,
+      "step": 11156
+    },
+    {
+      "epoch": 0.47199424655216177,
+      "grad_norm": 0.17271144688129425,
+      "learning_rate": 0.001,
+      "loss": 2.4049,
+      "step": 11157
+    },
+    {
+      "epoch": 0.4720365513156781,
+      "grad_norm": 0.21863944828510284,
+      "learning_rate": 0.001,
+      "loss": 2.4901,
+      "step": 11158
+    },
+    {
+      "epoch": 0.47207885607919453,
+      "grad_norm": 0.15925271809101105,
+      "learning_rate": 0.001,
+      "loss": 2.034,
+      "step": 11159
+    },
+    {
+      "epoch": 0.4721211608427109,
+      "grad_norm": 0.1758195459842682,
+      "learning_rate": 0.001,
+      "loss": 2.6075,
+      "step": 11160
+    },
+    {
+      "epoch": 0.47216346560622724,
+      "grad_norm": 0.2449083775281906,
+      "learning_rate": 0.001,
+      "loss": 2.4067,
+      "step": 11161
+    },
+    {
+      "epoch": 0.47220577036974365,
+      "grad_norm": 0.14552642405033112,
+      "learning_rate": 0.001,
+      "loss": 1.9266,
+      "step": 11162
+    },
+    {
+      "epoch": 0.47224807513326,
+      "grad_norm": 0.3583971858024597,
+      "learning_rate": 0.001,
+      "loss": 1.9896,
+      "step": 11163
+    },
+    {
+      "epoch": 0.47229037989677636,
+      "grad_norm": 0.22816874086856842,
+      "learning_rate": 0.001,
+      "loss": 2.1876,
+      "step": 11164
+    },
+    {
+      "epoch": 0.47233268466029277,
+      "grad_norm": 0.18025673925876617,
+      "learning_rate": 0.001,
+      "loss": 1.8776,
+      "step": 11165
+    },
+    {
+      "epoch": 0.4723749894238091,
+      "grad_norm": 0.17543093860149384,
+      "learning_rate": 0.001,
+      "loss": 2.6487,
+      "step": 11166
+    },
+    {
+      "epoch": 0.4724172941873255,
+      "grad_norm": 0.1770016998052597,
+      "learning_rate": 0.001,
+      "loss": 3.3751,
+      "step": 11167
+    },
+    {
+      "epoch": 0.4724595989508419,
+      "grad_norm": 0.1971733570098877,
+      "learning_rate": 0.001,
+      "loss": 2.0229,
+      "step": 11168
+    },
+    {
+      "epoch": 0.47250190371435824,
+      "grad_norm": 0.22783638536930084,
+      "learning_rate": 0.001,
+      "loss": 1.7606,
+      "step": 11169
+    },
+    {
+      "epoch": 0.4725442084778746,
+      "grad_norm": 0.17248298227787018,
+      "learning_rate": 0.001,
+      "loss": 1.8782,
+      "step": 11170
+    },
+    {
+      "epoch": 0.472586513241391,
+      "grad_norm": 0.1579669564962387,
+      "learning_rate": 0.001,
+      "loss": 1.4778,
+      "step": 11171
+    },
+    {
+      "epoch": 0.47262881800490736,
+      "grad_norm": 0.5877176523208618,
+      "learning_rate": 0.001,
+      "loss": 1.8328,
+      "step": 11172
+    },
+    {
+      "epoch": 0.4726711227684237,
+      "grad_norm": 0.2934803366661072,
+      "learning_rate": 0.001,
+      "loss": 2.7896,
+      "step": 11173
+    },
+    {
+      "epoch": 0.4727134275319401,
+      "grad_norm": 0.28287965059280396,
+      "learning_rate": 0.001,
+      "loss": 3.311,
+      "step": 11174
+    },
+    {
+      "epoch": 0.4727557322954565,
+      "grad_norm": 0.1633968949317932,
+      "learning_rate": 0.001,
+      "loss": 1.8999,
+      "step": 11175
+    },
+    {
+      "epoch": 0.47279803705897283,
+      "grad_norm": 0.2607437074184418,
+      "learning_rate": 0.001,
+      "loss": 1.9724,
+      "step": 11176
+    },
+    {
+      "epoch": 0.4728403418224892,
+      "grad_norm": 0.6347095370292664,
+      "learning_rate": 0.001,
+      "loss": 2.791,
+      "step": 11177
+    },
+    {
+      "epoch": 0.4728826465860056,
+      "grad_norm": 0.20074185729026794,
+      "learning_rate": 0.001,
+      "loss": 3.0955,
+      "step": 11178
+    },
+    {
+      "epoch": 0.47292495134952195,
+      "grad_norm": 0.19658905267715454,
+      "learning_rate": 0.001,
+      "loss": 2.3828,
+      "step": 11179
+    },
+    {
+      "epoch": 0.4729672561130383,
+      "grad_norm": 0.14689096808433533,
+      "learning_rate": 0.001,
+      "loss": 2.1146,
+      "step": 11180
+    },
+    {
+      "epoch": 0.4730095608765547,
+      "grad_norm": 0.21568824350833893,
+      "learning_rate": 0.001,
+      "loss": 3.265,
+      "step": 11181
+    },
+    {
+      "epoch": 0.47305186564007107,
+      "grad_norm": 0.20874953269958496,
+      "learning_rate": 0.001,
+      "loss": 2.4134,
+      "step": 11182
+    },
+    {
+      "epoch": 0.4730941704035874,
+      "grad_norm": 5.027991771697998,
+      "learning_rate": 0.001,
+      "loss": 3.1703,
+      "step": 11183
+    },
+    {
+      "epoch": 0.47313647516710383,
+      "grad_norm": 0.1729307919740677,
+      "learning_rate": 0.001,
+      "loss": 1.7616,
+      "step": 11184
+    },
+    {
+      "epoch": 0.4731787799306202,
+      "grad_norm": 0.1840585619211197,
+      "learning_rate": 0.001,
+      "loss": 2.9509,
+      "step": 11185
+    },
+    {
+      "epoch": 0.47322108469413654,
+      "grad_norm": 0.19434517621994019,
+      "learning_rate": 0.001,
+      "loss": 2.1666,
+      "step": 11186
+    },
+    {
+      "epoch": 0.47326338945765295,
+      "grad_norm": 0.1548105627298355,
+      "learning_rate": 0.001,
+      "loss": 2.6967,
+      "step": 11187
+    },
+    {
+      "epoch": 0.4733056942211693,
+      "grad_norm": 0.15446540713310242,
+      "learning_rate": 0.001,
+      "loss": 1.9184,
+      "step": 11188
+    },
+    {
+      "epoch": 0.47334799898468566,
+      "grad_norm": 0.15472517907619476,
+      "learning_rate": 0.001,
+      "loss": 1.9038,
+      "step": 11189
+    },
+    {
+      "epoch": 0.47339030374820207,
+      "grad_norm": 0.29968515038490295,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 11190
+    },
+    {
+      "epoch": 0.4734326085117184,
+      "grad_norm": 3.557898759841919,
+      "learning_rate": 0.001,
+      "loss": 3.0507,
+      "step": 11191
+    },
+    {
+      "epoch": 0.4734749132752348,
+      "grad_norm": 1.0494880676269531,
+      "learning_rate": 0.001,
+      "loss": 1.6712,
+      "step": 11192
+    },
+    {
+      "epoch": 0.4735172180387512,
+      "grad_norm": 0.18932278454303741,
+      "learning_rate": 0.001,
+      "loss": 2.3845,
+      "step": 11193
+    },
+    {
+      "epoch": 0.47355952280226754,
+      "grad_norm": 0.7504813075065613,
+      "learning_rate": 0.001,
+      "loss": 2.8903,
+      "step": 11194
+    },
+    {
+      "epoch": 0.4736018275657839,
+      "grad_norm": 0.2353019416332245,
+      "learning_rate": 0.001,
+      "loss": 2.2606,
+      "step": 11195
+    },
+    {
+      "epoch": 0.4736441323293003,
+      "grad_norm": 0.22455672919750214,
+      "learning_rate": 0.001,
+      "loss": 2.4879,
+      "step": 11196
+    },
+    {
+      "epoch": 0.47368643709281666,
+      "grad_norm": 0.28473368287086487,
+      "learning_rate": 0.001,
+      "loss": 2.6755,
+      "step": 11197
+    },
+    {
+      "epoch": 0.473728741856333,
+      "grad_norm": 0.19479401409626007,
+      "learning_rate": 0.001,
+      "loss": 2.784,
+      "step": 11198
+    },
+    {
+      "epoch": 0.47377104661984937,
+      "grad_norm": 0.17094643414020538,
+      "learning_rate": 0.001,
+      "loss": 2.3451,
+      "step": 11199
+    },
+    {
+      "epoch": 0.4738133513833658,
+      "grad_norm": 0.17563167214393616,
+      "learning_rate": 0.001,
+      "loss": 2.3602,
+      "step": 11200
+    },
+    {
+      "epoch": 0.47385565614688213,
+      "grad_norm": 0.19318290054798126,
+      "learning_rate": 0.001,
+      "loss": 1.8098,
+      "step": 11201
+    },
+    {
+      "epoch": 0.4738979609103985,
+      "grad_norm": 0.1785099357366562,
+      "learning_rate": 0.001,
+      "loss": 2.6017,
+      "step": 11202
+    },
+    {
+      "epoch": 0.4739402656739149,
+      "grad_norm": 0.23971965909004211,
+      "learning_rate": 0.001,
+      "loss": 3.1176,
+      "step": 11203
+    },
+    {
+      "epoch": 0.47398257043743125,
+      "grad_norm": 0.17552334070205688,
+      "learning_rate": 0.001,
+      "loss": 2.6471,
+      "step": 11204
+    },
+    {
+      "epoch": 0.4740248752009476,
+      "grad_norm": 0.309535950422287,
+      "learning_rate": 0.001,
+      "loss": 1.5446,
+      "step": 11205
+    },
+    {
+      "epoch": 0.474067179964464,
+      "grad_norm": 0.23233279585838318,
+      "learning_rate": 0.001,
+      "loss": 2.4261,
+      "step": 11206
+    },
+    {
+      "epoch": 0.47410948472798037,
+      "grad_norm": 0.17341110110282898,
+      "learning_rate": 0.001,
+      "loss": 1.5443,
+      "step": 11207
+    },
+    {
+      "epoch": 0.4741517894914967,
+      "grad_norm": 0.9063680768013,
+      "learning_rate": 0.001,
+      "loss": 1.5549,
+      "step": 11208
+    },
+    {
+      "epoch": 0.47419409425501313,
+      "grad_norm": 0.20930500328540802,
+      "learning_rate": 0.001,
+      "loss": 2.9428,
+      "step": 11209
+    },
+    {
+      "epoch": 0.4742363990185295,
+      "grad_norm": 0.1631225347518921,
+      "learning_rate": 0.001,
+      "loss": 2.0326,
+      "step": 11210
+    },
+    {
+      "epoch": 0.47427870378204584,
+      "grad_norm": 0.3474618196487427,
+      "learning_rate": 0.001,
+      "loss": 2.0844,
+      "step": 11211
+    },
+    {
+      "epoch": 0.47432100854556225,
+      "grad_norm": 0.16792355477809906,
+      "learning_rate": 0.001,
+      "loss": 2.1918,
+      "step": 11212
+    },
+    {
+      "epoch": 0.4743633133090786,
+      "grad_norm": 0.2315523475408554,
+      "learning_rate": 0.001,
+      "loss": 2.7786,
+      "step": 11213
+    },
+    {
+      "epoch": 0.47440561807259496,
+      "grad_norm": 0.21026229858398438,
+      "learning_rate": 0.001,
+      "loss": 2.1381,
+      "step": 11214
+    },
+    {
+      "epoch": 0.47444792283611137,
+      "grad_norm": 0.1558665633201599,
+      "learning_rate": 0.001,
+      "loss": 1.828,
+      "step": 11215
+    },
+    {
+      "epoch": 0.4744902275996277,
+      "grad_norm": 0.19669584929943085,
+      "learning_rate": 0.001,
+      "loss": 2.3019,
+      "step": 11216
+    },
+    {
+      "epoch": 0.4745325323631441,
+      "grad_norm": 0.16526706516742706,
+      "learning_rate": 0.001,
+      "loss": 2.1734,
+      "step": 11217
+    },
+    {
+      "epoch": 0.4745748371266605,
+      "grad_norm": 0.1388012170791626,
+      "learning_rate": 0.001,
+      "loss": 2.5851,
+      "step": 11218
+    },
+    {
+      "epoch": 0.47461714189017684,
+      "grad_norm": 0.20129254460334778,
+      "learning_rate": 0.001,
+      "loss": 2.3251,
+      "step": 11219
+    },
+    {
+      "epoch": 0.4746594466536932,
+      "grad_norm": 0.1600094735622406,
+      "learning_rate": 0.001,
+      "loss": 1.8666,
+      "step": 11220
+    },
+    {
+      "epoch": 0.4747017514172096,
+      "grad_norm": 0.1788060963153839,
+      "learning_rate": 0.001,
+      "loss": 3.1572,
+      "step": 11221
+    },
+    {
+      "epoch": 0.47474405618072596,
+      "grad_norm": 0.34340453147888184,
+      "learning_rate": 0.001,
+      "loss": 1.7197,
+      "step": 11222
+    },
+    {
+      "epoch": 0.4747863609442423,
+      "grad_norm": 0.18429534137248993,
+      "learning_rate": 0.001,
+      "loss": 2.1796,
+      "step": 11223
+    },
+    {
+      "epoch": 0.47482866570775867,
+      "grad_norm": 0.19663597643375397,
+      "learning_rate": 0.001,
+      "loss": 2.397,
+      "step": 11224
+    },
+    {
+      "epoch": 0.4748709704712751,
+      "grad_norm": 0.16480304300785065,
+      "learning_rate": 0.001,
+      "loss": 1.901,
+      "step": 11225
+    },
+    {
+      "epoch": 0.47491327523479143,
+      "grad_norm": 0.6021777391433716,
+      "learning_rate": 0.001,
+      "loss": 2.8861,
+      "step": 11226
+    },
+    {
+      "epoch": 0.4749555799983078,
+      "grad_norm": 0.14651986956596375,
+      "learning_rate": 0.001,
+      "loss": 2.2698,
+      "step": 11227
+    },
+    {
+      "epoch": 0.4749978847618242,
+      "grad_norm": 0.521893322467804,
+      "learning_rate": 0.001,
+      "loss": 2.3953,
+      "step": 11228
+    },
+    {
+      "epoch": 0.47504018952534055,
+      "grad_norm": 0.20987311005592346,
+      "learning_rate": 0.001,
+      "loss": 2.5025,
+      "step": 11229
+    },
+    {
+      "epoch": 0.4750824942888569,
+      "grad_norm": 0.14994093775749207,
+      "learning_rate": 0.001,
+      "loss": 1.6545,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4751247990523733,
+      "grad_norm": 0.47113507986068726,
+      "learning_rate": 0.001,
+      "loss": 2.1639,
+      "step": 11231
+    },
+    {
+      "epoch": 0.47516710381588967,
+      "grad_norm": 0.14544245600700378,
+      "learning_rate": 0.001,
+      "loss": 2.5572,
+      "step": 11232
+    },
+    {
+      "epoch": 0.475209408579406,
+      "grad_norm": 2.366591215133667,
+      "learning_rate": 0.001,
+      "loss": 2.1375,
+      "step": 11233
+    },
+    {
+      "epoch": 0.47525171334292243,
+      "grad_norm": 0.14105114340782166,
+      "learning_rate": 0.001,
+      "loss": 2.686,
+      "step": 11234
+    },
+    {
+      "epoch": 0.4752940181064388,
+      "grad_norm": 0.49145644903182983,
+      "learning_rate": 0.001,
+      "loss": 1.8348,
+      "step": 11235
+    },
+    {
+      "epoch": 0.47533632286995514,
+      "grad_norm": 0.18383541703224182,
+      "learning_rate": 0.001,
+      "loss": 1.5315,
+      "step": 11236
+    },
+    {
+      "epoch": 0.47537862763347155,
+      "grad_norm": 0.17911598086357117,
+      "learning_rate": 0.001,
+      "loss": 1.9033,
+      "step": 11237
+    },
+    {
+      "epoch": 0.4754209323969879,
+      "grad_norm": 3.2058968544006348,
+      "learning_rate": 0.001,
+      "loss": 3.8925,
+      "step": 11238
+    },
+    {
+      "epoch": 0.47546323716050426,
+      "grad_norm": 0.1748027503490448,
+      "learning_rate": 0.001,
+      "loss": 2.9772,
+      "step": 11239
+    },
+    {
+      "epoch": 0.47550554192402067,
+      "grad_norm": 0.16488003730773926,
+      "learning_rate": 0.001,
+      "loss": 1.9438,
+      "step": 11240
+    },
+    {
+      "epoch": 0.475547846687537,
+      "grad_norm": 0.2031308114528656,
+      "learning_rate": 0.001,
+      "loss": 1.5888,
+      "step": 11241
+    },
+    {
+      "epoch": 0.4755901514510534,
+      "grad_norm": 0.2005867063999176,
+      "learning_rate": 0.001,
+      "loss": 2.3343,
+      "step": 11242
+    },
+    {
+      "epoch": 0.4756324562145698,
+      "grad_norm": 5.769382953643799,
+      "learning_rate": 0.001,
+      "loss": 1.9595,
+      "step": 11243
+    },
+    {
+      "epoch": 0.47567476097808614,
+      "grad_norm": 0.16878317296504974,
+      "learning_rate": 0.001,
+      "loss": 1.9647,
+      "step": 11244
+    },
+    {
+      "epoch": 0.4757170657416025,
+      "grad_norm": 0.1860329806804657,
+      "learning_rate": 0.001,
+      "loss": 1.8979,
+      "step": 11245
+    },
+    {
+      "epoch": 0.47575937050511885,
+      "grad_norm": 0.25839003920555115,
+      "learning_rate": 0.001,
+      "loss": 2.7318,
+      "step": 11246
+    },
+    {
+      "epoch": 0.47580167526863526,
+      "grad_norm": 3.520124912261963,
+      "learning_rate": 0.001,
+      "loss": 2.6687,
+      "step": 11247
+    },
+    {
+      "epoch": 0.4758439800321516,
+      "grad_norm": 0.2002669721841812,
+      "learning_rate": 0.001,
+      "loss": 3.1077,
+      "step": 11248
+    },
+    {
+      "epoch": 0.47588628479566797,
+      "grad_norm": 0.31239837408065796,
+      "learning_rate": 0.001,
+      "loss": 2.4069,
+      "step": 11249
+    },
+    {
+      "epoch": 0.4759285895591844,
+      "grad_norm": 0.2110671103000641,
+      "learning_rate": 0.001,
+      "loss": 2.3489,
+      "step": 11250
+    },
+    {
+      "epoch": 0.47597089432270073,
+      "grad_norm": 0.22103223204612732,
+      "learning_rate": 0.001,
+      "loss": 2.9316,
+      "step": 11251
+    },
+    {
+      "epoch": 0.4760131990862171,
+      "grad_norm": 0.7828096747398376,
+      "learning_rate": 0.001,
+      "loss": 2.5561,
+      "step": 11252
+    },
+    {
+      "epoch": 0.4760555038497335,
+      "grad_norm": 0.20706386864185333,
+      "learning_rate": 0.001,
+      "loss": 2.9993,
+      "step": 11253
+    },
+    {
+      "epoch": 0.47609780861324985,
+      "grad_norm": 0.19384732842445374,
+      "learning_rate": 0.001,
+      "loss": 1.7905,
+      "step": 11254
+    },
+    {
+      "epoch": 0.4761401133767662,
+      "grad_norm": 0.25102004408836365,
+      "learning_rate": 0.001,
+      "loss": 2.7746,
+      "step": 11255
+    },
+    {
+      "epoch": 0.4761824181402826,
+      "grad_norm": 0.1942043900489807,
+      "learning_rate": 0.001,
+      "loss": 2.215,
+      "step": 11256
+    },
+    {
+      "epoch": 0.47622472290379897,
+      "grad_norm": 0.2124514877796173,
+      "learning_rate": 0.001,
+      "loss": 2.1071,
+      "step": 11257
+    },
+    {
+      "epoch": 0.4762670276673153,
+      "grad_norm": 0.1634020060300827,
+      "learning_rate": 0.001,
+      "loss": 1.5839,
+      "step": 11258
+    },
+    {
+      "epoch": 0.47630933243083173,
+      "grad_norm": 0.19070233404636383,
+      "learning_rate": 0.001,
+      "loss": 3.202,
+      "step": 11259
+    },
+    {
+      "epoch": 0.4763516371943481,
+      "grad_norm": 0.18325842916965485,
+      "learning_rate": 0.001,
+      "loss": 3.1532,
+      "step": 11260
+    },
+    {
+      "epoch": 0.47639394195786444,
+      "grad_norm": 0.23473717272281647,
+      "learning_rate": 0.001,
+      "loss": 2.2308,
+      "step": 11261
+    },
+    {
+      "epoch": 0.47643624672138085,
+      "grad_norm": 0.17958499491214752,
+      "learning_rate": 0.001,
+      "loss": 2.0782,
+      "step": 11262
+    },
+    {
+      "epoch": 0.4764785514848972,
+      "grad_norm": 0.31303510069847107,
+      "learning_rate": 0.001,
+      "loss": 2.5105,
+      "step": 11263
+    },
+    {
+      "epoch": 0.47652085624841356,
+      "grad_norm": 0.18192079663276672,
+      "learning_rate": 0.001,
+      "loss": 1.7983,
+      "step": 11264
+    },
+    {
+      "epoch": 0.47656316101192997,
+      "grad_norm": 0.31585457921028137,
+      "learning_rate": 0.001,
+      "loss": 2.8891,
+      "step": 11265
+    },
+    {
+      "epoch": 0.4766054657754463,
+      "grad_norm": 0.19254378974437714,
+      "learning_rate": 0.001,
+      "loss": 1.9023,
+      "step": 11266
+    },
+    {
+      "epoch": 0.4766477705389627,
+      "grad_norm": 0.5017719268798828,
+      "learning_rate": 0.001,
+      "loss": 1.9545,
+      "step": 11267
+    },
+    {
+      "epoch": 0.47669007530247903,
+      "grad_norm": 0.1998423933982849,
+      "learning_rate": 0.001,
+      "loss": 2.9385,
+      "step": 11268
+    },
+    {
+      "epoch": 0.47673238006599544,
+      "grad_norm": 0.18288664519786835,
+      "learning_rate": 0.001,
+      "loss": 1.5928,
+      "step": 11269
+    },
+    {
+      "epoch": 0.4767746848295118,
+      "grad_norm": 0.1892680823802948,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 11270
+    },
+    {
+      "epoch": 0.47681698959302815,
+      "grad_norm": 0.7018859386444092,
+      "learning_rate": 0.001,
+      "loss": 2.2363,
+      "step": 11271
+    },
+    {
+      "epoch": 0.47685929435654456,
+      "grad_norm": 0.6510068774223328,
+      "learning_rate": 0.001,
+      "loss": 2.1761,
+      "step": 11272
+    },
+    {
+      "epoch": 0.4769015991200609,
+      "grad_norm": 1.0637061595916748,
+      "learning_rate": 0.001,
+      "loss": 1.9906,
+      "step": 11273
+    },
+    {
+      "epoch": 0.47694390388357727,
+      "grad_norm": 0.15826399624347687,
+      "learning_rate": 0.001,
+      "loss": 2.4301,
+      "step": 11274
+    },
+    {
+      "epoch": 0.4769862086470937,
+      "grad_norm": 0.17290960252285004,
+      "learning_rate": 0.001,
+      "loss": 3.1263,
+      "step": 11275
+    },
+    {
+      "epoch": 0.47702851341061003,
+      "grad_norm": 0.19678086042404175,
+      "learning_rate": 0.001,
+      "loss": 2.8927,
+      "step": 11276
+    },
+    {
+      "epoch": 0.4770708181741264,
+      "grad_norm": 0.19402678310871124,
+      "learning_rate": 0.001,
+      "loss": 2.1192,
+      "step": 11277
+    },
+    {
+      "epoch": 0.4771131229376428,
+      "grad_norm": 0.20166024565696716,
+      "learning_rate": 0.001,
+      "loss": 2.2653,
+      "step": 11278
+    },
+    {
+      "epoch": 0.47715542770115915,
+      "grad_norm": 0.15184548497200012,
+      "learning_rate": 0.001,
+      "loss": 3.5353,
+      "step": 11279
+    },
+    {
+      "epoch": 0.4771977324646755,
+      "grad_norm": 0.15294674038887024,
+      "learning_rate": 0.001,
+      "loss": 2.1343,
+      "step": 11280
+    },
+    {
+      "epoch": 0.4772400372281919,
+      "grad_norm": 0.23221707344055176,
+      "learning_rate": 0.001,
+      "loss": 1.9696,
+      "step": 11281
+    },
+    {
+      "epoch": 0.47728234199170827,
+      "grad_norm": 0.541606605052948,
+      "learning_rate": 0.001,
+      "loss": 1.8145,
+      "step": 11282
+    },
+    {
+      "epoch": 0.4773246467552246,
+      "grad_norm": 12.742988586425781,
+      "learning_rate": 0.001,
+      "loss": 2.5281,
+      "step": 11283
+    },
+    {
+      "epoch": 0.47736695151874103,
+      "grad_norm": 0.15748733282089233,
+      "learning_rate": 0.001,
+      "loss": 2.6597,
+      "step": 11284
+    },
+    {
+      "epoch": 0.4774092562822574,
+      "grad_norm": 0.1852605640888214,
+      "learning_rate": 0.001,
+      "loss": 2.1034,
+      "step": 11285
+    },
+    {
+      "epoch": 0.47745156104577374,
+      "grad_norm": 0.5693913102149963,
+      "learning_rate": 0.001,
+      "loss": 3.0643,
+      "step": 11286
+    },
+    {
+      "epoch": 0.47749386580929015,
+      "grad_norm": 0.570594072341919,
+      "learning_rate": 0.001,
+      "loss": 3.6691,
+      "step": 11287
+    },
+    {
+      "epoch": 0.4775361705728065,
+      "grad_norm": 0.20175811648368835,
+      "learning_rate": 0.001,
+      "loss": 2.628,
+      "step": 11288
+    },
+    {
+      "epoch": 0.47757847533632286,
+      "grad_norm": 0.2098332792520523,
+      "learning_rate": 0.001,
+      "loss": 2.3156,
+      "step": 11289
+    },
+    {
+      "epoch": 0.4776207800998392,
+      "grad_norm": 0.18259800970554352,
+      "learning_rate": 0.001,
+      "loss": 3.0061,
+      "step": 11290
+    },
+    {
+      "epoch": 0.4776630848633556,
+      "grad_norm": 0.17582936584949493,
+      "learning_rate": 0.001,
+      "loss": 2.0574,
+      "step": 11291
+    },
+    {
+      "epoch": 0.477705389626872,
+      "grad_norm": 3.1900362968444824,
+      "learning_rate": 0.001,
+      "loss": 2.023,
+      "step": 11292
+    },
+    {
+      "epoch": 0.47774769439038833,
+      "grad_norm": 0.18563833832740784,
+      "learning_rate": 0.001,
+      "loss": 2.3776,
+      "step": 11293
+    },
+    {
+      "epoch": 0.47778999915390474,
+      "grad_norm": 0.1735265702009201,
+      "learning_rate": 0.001,
+      "loss": 3.3993,
+      "step": 11294
+    },
+    {
+      "epoch": 0.4778323039174211,
+      "grad_norm": 0.17807775735855103,
+      "learning_rate": 0.001,
+      "loss": 1.9739,
+      "step": 11295
+    },
+    {
+      "epoch": 0.47787460868093745,
+      "grad_norm": 0.42924782633781433,
+      "learning_rate": 0.001,
+      "loss": 2.0357,
+      "step": 11296
+    },
+    {
+      "epoch": 0.47791691344445386,
+      "grad_norm": 0.2000165730714798,
+      "learning_rate": 0.001,
+      "loss": 2.0919,
+      "step": 11297
+    },
+    {
+      "epoch": 0.4779592182079702,
+      "grad_norm": 0.14181675016880035,
+      "learning_rate": 0.001,
+      "loss": 2.5228,
+      "step": 11298
+    },
+    {
+      "epoch": 0.47800152297148657,
+      "grad_norm": 5.60988187789917,
+      "learning_rate": 0.001,
+      "loss": 1.5685,
+      "step": 11299
+    },
+    {
+      "epoch": 0.478043827735003,
+      "grad_norm": 0.1896638423204422,
+      "learning_rate": 0.001,
+      "loss": 2.7071,
+      "step": 11300
+    },
+    {
+      "epoch": 0.47808613249851933,
+      "grad_norm": 46.45356750488281,
+      "learning_rate": 0.001,
+      "loss": 3.0575,
+      "step": 11301
+    },
+    {
+      "epoch": 0.4781284372620357,
+      "grad_norm": 0.5738785266876221,
+      "learning_rate": 0.001,
+      "loss": 2.3252,
+      "step": 11302
+    },
+    {
+      "epoch": 0.4781707420255521,
+      "grad_norm": 0.18505975604057312,
+      "learning_rate": 0.001,
+      "loss": 2.2331,
+      "step": 11303
+    },
+    {
+      "epoch": 0.47821304678906845,
+      "grad_norm": 0.18580776453018188,
+      "learning_rate": 0.001,
+      "loss": 2.7998,
+      "step": 11304
+    },
+    {
+      "epoch": 0.4782553515525848,
+      "grad_norm": 2.873171329498291,
+      "learning_rate": 0.001,
+      "loss": 3.3834,
+      "step": 11305
+    },
+    {
+      "epoch": 0.4782976563161012,
+      "grad_norm": 0.17090004682540894,
+      "learning_rate": 0.001,
+      "loss": 2.4539,
+      "step": 11306
+    },
+    {
+      "epoch": 0.47833996107961757,
+      "grad_norm": 0.5887268781661987,
+      "learning_rate": 0.001,
+      "loss": 1.9116,
+      "step": 11307
+    },
+    {
+      "epoch": 0.4783822658431339,
+      "grad_norm": 0.2121281623840332,
+      "learning_rate": 0.001,
+      "loss": 2.3149,
+      "step": 11308
+    },
+    {
+      "epoch": 0.47842457060665033,
+      "grad_norm": 0.31115978956222534,
+      "learning_rate": 0.001,
+      "loss": 2.1889,
+      "step": 11309
+    },
+    {
+      "epoch": 0.4784668753701667,
+      "grad_norm": 0.1875309944152832,
+      "learning_rate": 0.001,
+      "loss": 2.1295,
+      "step": 11310
+    },
+    {
+      "epoch": 0.47850918013368304,
+      "grad_norm": 0.1967727243900299,
+      "learning_rate": 0.001,
+      "loss": 1.7845,
+      "step": 11311
+    },
+    {
+      "epoch": 0.4785514848971994,
+      "grad_norm": 2.0214593410491943,
+      "learning_rate": 0.001,
+      "loss": 2.2464,
+      "step": 11312
+    },
+    {
+      "epoch": 0.4785937896607158,
+      "grad_norm": 0.18990519642829895,
+      "learning_rate": 0.001,
+      "loss": 2.0703,
+      "step": 11313
+    },
+    {
+      "epoch": 0.47863609442423216,
+      "grad_norm": 1.7356641292572021,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 11314
+    },
+    {
+      "epoch": 0.4786783991877485,
+      "grad_norm": 0.220098614692688,
+      "learning_rate": 0.001,
+      "loss": 2.4705,
+      "step": 11315
+    },
+    {
+      "epoch": 0.4787207039512649,
+      "grad_norm": 0.1770542711019516,
+      "learning_rate": 0.001,
+      "loss": 2.8367,
+      "step": 11316
+    },
+    {
+      "epoch": 0.4787630087147813,
+      "grad_norm": 0.22387665510177612,
+      "learning_rate": 0.001,
+      "loss": 2.1766,
+      "step": 11317
+    },
+    {
+      "epoch": 0.47880531347829763,
+      "grad_norm": 0.1825076937675476,
+      "learning_rate": 0.001,
+      "loss": 3.0979,
+      "step": 11318
+    },
+    {
+      "epoch": 0.47884761824181404,
+      "grad_norm": 0.2045036405324936,
+      "learning_rate": 0.001,
+      "loss": 2.9486,
+      "step": 11319
+    },
+    {
+      "epoch": 0.4788899230053304,
+      "grad_norm": 0.20427003502845764,
+      "learning_rate": 0.001,
+      "loss": 2.7808,
+      "step": 11320
+    },
+    {
+      "epoch": 0.47893222776884675,
+      "grad_norm": 0.21618923544883728,
+      "learning_rate": 0.001,
+      "loss": 2.3455,
+      "step": 11321
+    },
+    {
+      "epoch": 0.47897453253236316,
+      "grad_norm": 0.19083094596862793,
+      "learning_rate": 0.001,
+      "loss": 2.621,
+      "step": 11322
+    },
+    {
+      "epoch": 0.4790168372958795,
+      "grad_norm": 10.630045890808105,
+      "learning_rate": 0.001,
+      "loss": 2.8354,
+      "step": 11323
+    },
+    {
+      "epoch": 0.47905914205939587,
+      "grad_norm": 0.20413336157798767,
+      "learning_rate": 0.001,
+      "loss": 2.2648,
+      "step": 11324
+    },
+    {
+      "epoch": 0.4791014468229123,
+      "grad_norm": 0.23747549951076508,
+      "learning_rate": 0.001,
+      "loss": 3.3187,
+      "step": 11325
+    },
+    {
+      "epoch": 0.47914375158642863,
+      "grad_norm": 21.391681671142578,
+      "learning_rate": 0.001,
+      "loss": 1.6874,
+      "step": 11326
+    },
+    {
+      "epoch": 0.479186056349945,
+      "grad_norm": 0.2882399559020996,
+      "learning_rate": 0.001,
+      "loss": 2.4637,
+      "step": 11327
+    },
+    {
+      "epoch": 0.4792283611134614,
+      "grad_norm": 0.330314576625824,
+      "learning_rate": 0.001,
+      "loss": 3.253,
+      "step": 11328
+    },
+    {
+      "epoch": 0.47927066587697775,
+      "grad_norm": 1.4046648740768433,
+      "learning_rate": 0.001,
+      "loss": 2.8288,
+      "step": 11329
+    },
+    {
+      "epoch": 0.4793129706404941,
+      "grad_norm": 0.7811962962150574,
+      "learning_rate": 0.001,
+      "loss": 2.6857,
+      "step": 11330
+    },
+    {
+      "epoch": 0.4793552754040105,
+      "grad_norm": 0.2128724753856659,
+      "learning_rate": 0.001,
+      "loss": 2.478,
+      "step": 11331
+    },
+    {
+      "epoch": 0.47939758016752687,
+      "grad_norm": 0.2501393258571625,
+      "learning_rate": 0.001,
+      "loss": 2.4157,
+      "step": 11332
+    },
+    {
+      "epoch": 0.4794398849310432,
+      "grad_norm": 0.22114041447639465,
+      "learning_rate": 0.001,
+      "loss": 1.7828,
+      "step": 11333
+    },
+    {
+      "epoch": 0.47948218969455964,
+      "grad_norm": 0.1902758628129959,
+      "learning_rate": 0.001,
+      "loss": 2.0687,
+      "step": 11334
+    },
+    {
+      "epoch": 0.479524494458076,
+      "grad_norm": 0.548904538154602,
+      "learning_rate": 0.001,
+      "loss": 2.5692,
+      "step": 11335
+    },
+    {
+      "epoch": 0.47956679922159234,
+      "grad_norm": 0.18059691786766052,
+      "learning_rate": 0.001,
+      "loss": 2.0568,
+      "step": 11336
+    },
+    {
+      "epoch": 0.4796091039851087,
+      "grad_norm": 0.17880599200725555,
+      "learning_rate": 0.001,
+      "loss": 2.1907,
+      "step": 11337
+    },
+    {
+      "epoch": 0.4796514087486251,
+      "grad_norm": 0.3715362548828125,
+      "learning_rate": 0.001,
+      "loss": 2.5385,
+      "step": 11338
+    },
+    {
+      "epoch": 0.47969371351214146,
+      "grad_norm": 0.1764996498823166,
+      "learning_rate": 0.001,
+      "loss": 1.8525,
+      "step": 11339
+    },
+    {
+      "epoch": 0.4797360182756578,
+      "grad_norm": 0.8168947696685791,
+      "learning_rate": 0.001,
+      "loss": 2.5725,
+      "step": 11340
+    },
+    {
+      "epoch": 0.4797783230391742,
+      "grad_norm": 0.161908358335495,
+      "learning_rate": 0.001,
+      "loss": 2.0986,
+      "step": 11341
+    },
+    {
+      "epoch": 0.4798206278026906,
+      "grad_norm": 0.1834360510110855,
+      "learning_rate": 0.001,
+      "loss": 2.0397,
+      "step": 11342
+    },
+    {
+      "epoch": 0.47986293256620693,
+      "grad_norm": 0.17309635877609253,
+      "learning_rate": 0.001,
+      "loss": 2.5099,
+      "step": 11343
+    },
+    {
+      "epoch": 0.47990523732972334,
+      "grad_norm": 0.19647128880023956,
+      "learning_rate": 0.001,
+      "loss": 2.0393,
+      "step": 11344
+    },
+    {
+      "epoch": 0.4799475420932397,
+      "grad_norm": 0.670528769493103,
+      "learning_rate": 0.001,
+      "loss": 3.4804,
+      "step": 11345
+    },
+    {
+      "epoch": 0.47998984685675605,
+      "grad_norm": 0.1826171725988388,
+      "learning_rate": 0.001,
+      "loss": 2.0048,
+      "step": 11346
+    },
+    {
+      "epoch": 0.48003215162027246,
+      "grad_norm": 0.20341645181179047,
+      "learning_rate": 0.001,
+      "loss": 2.7362,
+      "step": 11347
+    },
+    {
+      "epoch": 0.4800744563837888,
+      "grad_norm": 0.2916952967643738,
+      "learning_rate": 0.001,
+      "loss": 1.6388,
+      "step": 11348
+    },
+    {
+      "epoch": 0.48011676114730517,
+      "grad_norm": 0.5783160328865051,
+      "learning_rate": 0.001,
+      "loss": 2.3783,
+      "step": 11349
+    },
+    {
+      "epoch": 0.4801590659108216,
+      "grad_norm": 0.20249702036380768,
+      "learning_rate": 0.001,
+      "loss": 2.3376,
+      "step": 11350
+    },
+    {
+      "epoch": 0.48020137067433794,
+      "grad_norm": 0.18847358226776123,
+      "learning_rate": 0.001,
+      "loss": 2.1271,
+      "step": 11351
+    },
+    {
+      "epoch": 0.4802436754378543,
+      "grad_norm": 2.4059906005859375,
+      "learning_rate": 0.001,
+      "loss": 1.7613,
+      "step": 11352
+    },
+    {
+      "epoch": 0.4802859802013707,
+      "grad_norm": 0.1835888773202896,
+      "learning_rate": 0.001,
+      "loss": 2.379,
+      "step": 11353
+    },
+    {
+      "epoch": 0.48032828496488705,
+      "grad_norm": 0.34716445207595825,
+      "learning_rate": 0.001,
+      "loss": 2.0613,
+      "step": 11354
+    },
+    {
+      "epoch": 0.4803705897284034,
+      "grad_norm": 0.154496431350708,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 11355
+    },
+    {
+      "epoch": 0.4804128944919198,
+      "grad_norm": 0.5748295783996582,
+      "learning_rate": 0.001,
+      "loss": 2.2405,
+      "step": 11356
+    },
+    {
+      "epoch": 0.48045519925543617,
+      "grad_norm": 0.3923312723636627,
+      "learning_rate": 0.001,
+      "loss": 2.7319,
+      "step": 11357
+    },
+    {
+      "epoch": 0.4804975040189525,
+      "grad_norm": 0.5746713280677795,
+      "learning_rate": 0.001,
+      "loss": 1.8951,
+      "step": 11358
+    },
+    {
+      "epoch": 0.4805398087824689,
+      "grad_norm": 1.0330185890197754,
+      "learning_rate": 0.001,
+      "loss": 3.1664,
+      "step": 11359
+    },
+    {
+      "epoch": 0.4805821135459853,
+      "grad_norm": 0.1738303303718567,
+      "learning_rate": 0.001,
+      "loss": 2.808,
+      "step": 11360
+    },
+    {
+      "epoch": 0.48062441830950164,
+      "grad_norm": 8.028282165527344,
+      "learning_rate": 0.001,
+      "loss": 1.9271,
+      "step": 11361
+    },
+    {
+      "epoch": 0.480666723073018,
+      "grad_norm": 0.18833428621292114,
+      "learning_rate": 0.001,
+      "loss": 2.1628,
+      "step": 11362
+    },
+    {
+      "epoch": 0.4807090278365344,
+      "grad_norm": 0.1686827838420868,
+      "learning_rate": 0.001,
+      "loss": 1.6737,
+      "step": 11363
+    },
+    {
+      "epoch": 0.48075133260005076,
+      "grad_norm": 0.18253150582313538,
+      "learning_rate": 0.001,
+      "loss": 2.337,
+      "step": 11364
+    },
+    {
+      "epoch": 0.4807936373635671,
+      "grad_norm": 0.23626413941383362,
+      "learning_rate": 0.001,
+      "loss": 3.2853,
+      "step": 11365
+    },
+    {
+      "epoch": 0.4808359421270835,
+      "grad_norm": 0.718061089515686,
+      "learning_rate": 0.001,
+      "loss": 2.8863,
+      "step": 11366
+    },
+    {
+      "epoch": 0.4808782468905999,
+      "grad_norm": 0.21760138869285583,
+      "learning_rate": 0.001,
+      "loss": 2.4297,
+      "step": 11367
+    },
+    {
+      "epoch": 0.48092055165411624,
+      "grad_norm": 0.19987843930721283,
+      "learning_rate": 0.001,
+      "loss": 3.1579,
+      "step": 11368
+    },
+    {
+      "epoch": 0.48096285641763264,
+      "grad_norm": 0.16361162066459656,
+      "learning_rate": 0.001,
+      "loss": 1.5716,
+      "step": 11369
+    },
+    {
+      "epoch": 0.481005161181149,
+      "grad_norm": 0.169572651386261,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 11370
+    },
+    {
+      "epoch": 0.48104746594466535,
+      "grad_norm": 1.216856598854065,
+      "learning_rate": 0.001,
+      "loss": 2.5188,
+      "step": 11371
+    },
+    {
+      "epoch": 0.48108977070818176,
+      "grad_norm": 0.20964156091213226,
+      "learning_rate": 0.001,
+      "loss": 2.9859,
+      "step": 11372
+    },
+    {
+      "epoch": 0.4811320754716981,
+      "grad_norm": 0.5650727152824402,
+      "learning_rate": 0.001,
+      "loss": 3.1159,
+      "step": 11373
+    },
+    {
+      "epoch": 0.48117438023521447,
+      "grad_norm": 0.2616911232471466,
+      "learning_rate": 0.001,
+      "loss": 2.1743,
+      "step": 11374
+    },
+    {
+      "epoch": 0.4812166849987309,
+      "grad_norm": 0.22309203445911407,
+      "learning_rate": 0.001,
+      "loss": 2.7001,
+      "step": 11375
+    },
+    {
+      "epoch": 0.48125898976224724,
+      "grad_norm": 0.18695230782032013,
+      "learning_rate": 0.001,
+      "loss": 2.203,
+      "step": 11376
+    },
+    {
+      "epoch": 0.4813012945257636,
+      "grad_norm": 0.1963418573141098,
+      "learning_rate": 0.001,
+      "loss": 3.2709,
+      "step": 11377
+    },
+    {
+      "epoch": 0.48134359928928,
+      "grad_norm": 0.16834786534309387,
+      "learning_rate": 0.001,
+      "loss": 2.769,
+      "step": 11378
+    },
+    {
+      "epoch": 0.48138590405279635,
+      "grad_norm": 0.1826411634683609,
+      "learning_rate": 0.001,
+      "loss": 1.7105,
+      "step": 11379
+    },
+    {
+      "epoch": 0.4814282088163127,
+      "grad_norm": 0.43672510981559753,
+      "learning_rate": 0.001,
+      "loss": 2.3934,
+      "step": 11380
+    },
+    {
+      "epoch": 0.48147051357982906,
+      "grad_norm": 0.16782554984092712,
+      "learning_rate": 0.001,
+      "loss": 2.589,
+      "step": 11381
+    },
+    {
+      "epoch": 0.4815128183433455,
+      "grad_norm": 0.16116951406002045,
+      "learning_rate": 0.001,
+      "loss": 2.2085,
+      "step": 11382
+    },
+    {
+      "epoch": 0.4815551231068618,
+      "grad_norm": 0.18043899536132812,
+      "learning_rate": 0.001,
+      "loss": 1.7793,
+      "step": 11383
+    },
+    {
+      "epoch": 0.4815974278703782,
+      "grad_norm": 0.5931171774864197,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 11384
+    },
+    {
+      "epoch": 0.4816397326338946,
+      "grad_norm": 0.4791402816772461,
+      "learning_rate": 0.001,
+      "loss": 3.4351,
+      "step": 11385
+    },
+    {
+      "epoch": 0.48168203739741094,
+      "grad_norm": 0.2714325487613678,
+      "learning_rate": 0.001,
+      "loss": 2.2362,
+      "step": 11386
+    },
+    {
+      "epoch": 0.4817243421609273,
+      "grad_norm": 0.19778069853782654,
+      "learning_rate": 0.001,
+      "loss": 2.5545,
+      "step": 11387
+    },
+    {
+      "epoch": 0.4817666469244437,
+      "grad_norm": 0.208540141582489,
+      "learning_rate": 0.001,
+      "loss": 2.9821,
+      "step": 11388
+    },
+    {
+      "epoch": 0.48180895168796006,
+      "grad_norm": 0.16977502405643463,
+      "learning_rate": 0.001,
+      "loss": 1.8149,
+      "step": 11389
+    },
+    {
+      "epoch": 0.4818512564514764,
+      "grad_norm": 0.22707661986351013,
+      "learning_rate": 0.001,
+      "loss": 2.7141,
+      "step": 11390
+    },
+    {
+      "epoch": 0.4818935612149928,
+      "grad_norm": 0.38424187898635864,
+      "learning_rate": 0.001,
+      "loss": 3.1576,
+      "step": 11391
+    },
+    {
+      "epoch": 0.4819358659785092,
+      "grad_norm": 1.370203971862793,
+      "learning_rate": 0.001,
+      "loss": 2.2048,
+      "step": 11392
+    },
+    {
+      "epoch": 0.48197817074202554,
+      "grad_norm": 0.18375247716903687,
+      "learning_rate": 0.001,
+      "loss": 2.3193,
+      "step": 11393
+    },
+    {
+      "epoch": 0.48202047550554195,
+      "grad_norm": 0.9614766836166382,
+      "learning_rate": 0.001,
+      "loss": 1.794,
+      "step": 11394
+    },
+    {
+      "epoch": 0.4820627802690583,
+      "grad_norm": 0.18902945518493652,
+      "learning_rate": 0.001,
+      "loss": 2.1396,
+      "step": 11395
+    },
+    {
+      "epoch": 0.48210508503257465,
+      "grad_norm": 0.16414183378219604,
+      "learning_rate": 0.001,
+      "loss": 1.6427,
+      "step": 11396
+    },
+    {
+      "epoch": 0.48214738979609106,
+      "grad_norm": 0.16200171411037445,
+      "learning_rate": 0.001,
+      "loss": 2.0166,
+      "step": 11397
+    },
+    {
+      "epoch": 0.4821896945596074,
+      "grad_norm": 3.9230146408081055,
+      "learning_rate": 0.001,
+      "loss": 2.5453,
+      "step": 11398
+    },
+    {
+      "epoch": 0.4822319993231238,
+      "grad_norm": 0.211333766579628,
+      "learning_rate": 0.001,
+      "loss": 2.9545,
+      "step": 11399
+    },
+    {
+      "epoch": 0.4822743040866402,
+      "grad_norm": 0.2078399807214737,
+      "learning_rate": 0.001,
+      "loss": 2.085,
+      "step": 11400
+    },
+    {
+      "epoch": 0.48231660885015654,
+      "grad_norm": 0.1923077553510666,
+      "learning_rate": 0.001,
+      "loss": 2.3923,
+      "step": 11401
+    },
+    {
+      "epoch": 0.4823589136136729,
+      "grad_norm": 1.6450941562652588,
+      "learning_rate": 0.001,
+      "loss": 2.4251,
+      "step": 11402
+    },
+    {
+      "epoch": 0.48240121837718924,
+      "grad_norm": 0.2333688586950302,
+      "learning_rate": 0.001,
+      "loss": 2.2939,
+      "step": 11403
+    },
+    {
+      "epoch": 0.48244352314070565,
+      "grad_norm": 6.017459392547607,
+      "learning_rate": 0.001,
+      "loss": 1.9482,
+      "step": 11404
+    },
+    {
+      "epoch": 0.482485827904222,
+      "grad_norm": 1.64377760887146,
+      "learning_rate": 0.001,
+      "loss": 1.9915,
+      "step": 11405
+    },
+    {
+      "epoch": 0.48252813266773836,
+      "grad_norm": 0.22818098962306976,
+      "learning_rate": 0.001,
+      "loss": 3.1697,
+      "step": 11406
+    },
+    {
+      "epoch": 0.4825704374312548,
+      "grad_norm": 0.20856161415576935,
+      "learning_rate": 0.001,
+      "loss": 1.7394,
+      "step": 11407
+    },
+    {
+      "epoch": 0.4826127421947711,
+      "grad_norm": 0.2079148292541504,
+      "learning_rate": 0.001,
+      "loss": 2.85,
+      "step": 11408
+    },
+    {
+      "epoch": 0.4826550469582875,
+      "grad_norm": 3.2609381675720215,
+      "learning_rate": 0.001,
+      "loss": 2.744,
+      "step": 11409
+    },
+    {
+      "epoch": 0.4826973517218039,
+      "grad_norm": 0.4268743395805359,
+      "learning_rate": 0.001,
+      "loss": 3.3981,
+      "step": 11410
+    },
+    {
+      "epoch": 0.48273965648532025,
+      "grad_norm": 0.260139137506485,
+      "learning_rate": 0.001,
+      "loss": 2.0163,
+      "step": 11411
+    },
+    {
+      "epoch": 0.4827819612488366,
+      "grad_norm": 0.21238379180431366,
+      "learning_rate": 0.001,
+      "loss": 2.566,
+      "step": 11412
+    },
+    {
+      "epoch": 0.482824266012353,
+      "grad_norm": 0.19492459297180176,
+      "learning_rate": 0.001,
+      "loss": 2.2858,
+      "step": 11413
+    },
+    {
+      "epoch": 0.48286657077586936,
+      "grad_norm": 2.3117878437042236,
+      "learning_rate": 0.001,
+      "loss": 2.0708,
+      "step": 11414
+    },
+    {
+      "epoch": 0.4829088755393857,
+      "grad_norm": 0.20078395307064056,
+      "learning_rate": 0.001,
+      "loss": 3.0703,
+      "step": 11415
+    },
+    {
+      "epoch": 0.4829511803029021,
+      "grad_norm": 0.18445464968681335,
+      "learning_rate": 0.001,
+      "loss": 2.7102,
+      "step": 11416
+    },
+    {
+      "epoch": 0.4829934850664185,
+      "grad_norm": 0.2198317050933838,
+      "learning_rate": 0.001,
+      "loss": 2.8357,
+      "step": 11417
+    },
+    {
+      "epoch": 0.48303578982993484,
+      "grad_norm": 2.0135302543640137,
+      "learning_rate": 0.001,
+      "loss": 1.8806,
+      "step": 11418
+    },
+    {
+      "epoch": 0.48307809459345125,
+      "grad_norm": 0.25936415791511536,
+      "learning_rate": 0.001,
+      "loss": 3.3516,
+      "step": 11419
+    },
+    {
+      "epoch": 0.4831203993569676,
+      "grad_norm": 0.2253940850496292,
+      "learning_rate": 0.001,
+      "loss": 2.3307,
+      "step": 11420
+    },
+    {
+      "epoch": 0.48316270412048395,
+      "grad_norm": 0.2977537214756012,
+      "learning_rate": 0.001,
+      "loss": 2.1588,
+      "step": 11421
+    },
+    {
+      "epoch": 0.48320500888400036,
+      "grad_norm": 1.3595587015151978,
+      "learning_rate": 0.001,
+      "loss": 2.6004,
+      "step": 11422
+    },
+    {
+      "epoch": 0.4832473136475167,
+      "grad_norm": 2.3829686641693115,
+      "learning_rate": 0.001,
+      "loss": 2.6128,
+      "step": 11423
+    },
+    {
+      "epoch": 0.4832896184110331,
+      "grad_norm": 0.2193344682455063,
+      "learning_rate": 0.001,
+      "loss": 2.9908,
+      "step": 11424
+    },
+    {
+      "epoch": 0.4833319231745494,
+      "grad_norm": 0.19066233932971954,
+      "learning_rate": 0.001,
+      "loss": 2.025,
+      "step": 11425
+    },
+    {
+      "epoch": 0.48337422793806584,
+      "grad_norm": 0.20256930589675903,
+      "learning_rate": 0.001,
+      "loss": 1.9787,
+      "step": 11426
+    },
+    {
+      "epoch": 0.4834165327015822,
+      "grad_norm": 0.22213421761989594,
+      "learning_rate": 0.001,
+      "loss": 1.9955,
+      "step": 11427
+    },
+    {
+      "epoch": 0.48345883746509855,
+      "grad_norm": 0.2642061412334442,
+      "learning_rate": 0.001,
+      "loss": 2.1242,
+      "step": 11428
+    },
+    {
+      "epoch": 0.48350114222861496,
+      "grad_norm": 0.23516133427619934,
+      "learning_rate": 0.001,
+      "loss": 2.2582,
+      "step": 11429
+    },
+    {
+      "epoch": 0.4835434469921313,
+      "grad_norm": 0.22922705113887787,
+      "learning_rate": 0.001,
+      "loss": 2.3993,
+      "step": 11430
+    },
+    {
+      "epoch": 0.48358575175564766,
+      "grad_norm": 0.2899877727031708,
+      "learning_rate": 0.001,
+      "loss": 3.879,
+      "step": 11431
+    },
+    {
+      "epoch": 0.4836280565191641,
+      "grad_norm": 0.17571720480918884,
+      "learning_rate": 0.001,
+      "loss": 2.1495,
+      "step": 11432
+    },
+    {
+      "epoch": 0.4836703612826804,
+      "grad_norm": 2.013770580291748,
+      "learning_rate": 0.001,
+      "loss": 2.7839,
+      "step": 11433
+    },
+    {
+      "epoch": 0.4837126660461968,
+      "grad_norm": 3.0192947387695312,
+      "learning_rate": 0.001,
+      "loss": 2.1995,
+      "step": 11434
+    },
+    {
+      "epoch": 0.4837549708097132,
+      "grad_norm": 0.20939669013023376,
+      "learning_rate": 0.001,
+      "loss": 2.1262,
+      "step": 11435
+    },
+    {
+      "epoch": 0.48379727557322955,
+      "grad_norm": 0.2533853352069855,
+      "learning_rate": 0.001,
+      "loss": 2.8203,
+      "step": 11436
+    },
+    {
+      "epoch": 0.4838395803367459,
+      "grad_norm": 0.2081190049648285,
+      "learning_rate": 0.001,
+      "loss": 2.3127,
+      "step": 11437
+    },
+    {
+      "epoch": 0.4838818851002623,
+      "grad_norm": 0.1944902241230011,
+      "learning_rate": 0.001,
+      "loss": 1.8991,
+      "step": 11438
+    },
+    {
+      "epoch": 0.48392418986377866,
+      "grad_norm": 0.19556209444999695,
+      "learning_rate": 0.001,
+      "loss": 2.3796,
+      "step": 11439
+    },
+    {
+      "epoch": 0.483966494627295,
+      "grad_norm": 0.27376073598861694,
+      "learning_rate": 0.001,
+      "loss": 2.5798,
+      "step": 11440
+    },
+    {
+      "epoch": 0.48400879939081143,
+      "grad_norm": 0.8901777863502502,
+      "learning_rate": 0.001,
+      "loss": 3.6993,
+      "step": 11441
+    },
+    {
+      "epoch": 0.4840511041543278,
+      "grad_norm": 0.5265764594078064,
+      "learning_rate": 0.001,
+      "loss": 1.7463,
+      "step": 11442
+    },
+    {
+      "epoch": 0.48409340891784414,
+      "grad_norm": 0.4732452630996704,
+      "learning_rate": 0.001,
+      "loss": 1.6442,
+      "step": 11443
+    },
+    {
+      "epoch": 0.48413571368136055,
+      "grad_norm": 0.23689015209674835,
+      "learning_rate": 0.001,
+      "loss": 2.367,
+      "step": 11444
+    },
+    {
+      "epoch": 0.4841780184448769,
+      "grad_norm": 0.1857021003961563,
+      "learning_rate": 0.001,
+      "loss": 2.0991,
+      "step": 11445
+    },
+    {
+      "epoch": 0.48422032320839326,
+      "grad_norm": 9.68715763092041,
+      "learning_rate": 0.001,
+      "loss": 1.781,
+      "step": 11446
+    },
+    {
+      "epoch": 0.4842626279719096,
+      "grad_norm": 0.18469716608524323,
+      "learning_rate": 0.001,
+      "loss": 2.3133,
+      "step": 11447
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.195147305727005,
+      "learning_rate": 0.001,
+      "loss": 3.6543,
+      "step": 11448
+    },
+    {
+      "epoch": 0.4843472374989424,
+      "grad_norm": 0.19625887274742126,
+      "learning_rate": 0.001,
+      "loss": 1.7704,
+      "step": 11449
+    },
+    {
+      "epoch": 0.4843895422624587,
+      "grad_norm": 1.0234781503677368,
+      "learning_rate": 0.001,
+      "loss": 2.4147,
+      "step": 11450
+    },
+    {
+      "epoch": 0.48443184702597514,
+      "grad_norm": 1.8966562747955322,
+      "learning_rate": 0.001,
+      "loss": 1.4547,
+      "step": 11451
+    },
+    {
+      "epoch": 0.4844741517894915,
+      "grad_norm": 0.29470208287239075,
+      "learning_rate": 0.001,
+      "loss": 2.3858,
+      "step": 11452
+    },
+    {
+      "epoch": 0.48451645655300785,
+      "grad_norm": 0.4111064672470093,
+      "learning_rate": 0.001,
+      "loss": 2.43,
+      "step": 11453
+    },
+    {
+      "epoch": 0.48455876131652426,
+      "grad_norm": 0.1965835988521576,
+      "learning_rate": 0.001,
+      "loss": 1.5909,
+      "step": 11454
+    },
+    {
+      "epoch": 0.4846010660800406,
+      "grad_norm": 0.19790798425674438,
+      "learning_rate": 0.001,
+      "loss": 2.1684,
+      "step": 11455
+    },
+    {
+      "epoch": 0.48464337084355696,
+      "grad_norm": 0.20888280868530273,
+      "learning_rate": 0.001,
+      "loss": 2.1558,
+      "step": 11456
+    },
+    {
+      "epoch": 0.4846856756070734,
+      "grad_norm": 0.16559502482414246,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 11457
+    },
+    {
+      "epoch": 0.48472798037058973,
+      "grad_norm": 0.1633719652891159,
+      "learning_rate": 0.001,
+      "loss": 2.0244,
+      "step": 11458
+    },
+    {
+      "epoch": 0.4847702851341061,
+      "grad_norm": 0.1771150678396225,
+      "learning_rate": 0.001,
+      "loss": 2.4581,
+      "step": 11459
+    },
+    {
+      "epoch": 0.4848125898976225,
+      "grad_norm": 0.2200511246919632,
+      "learning_rate": 0.001,
+      "loss": 2.4251,
+      "step": 11460
+    },
+    {
+      "epoch": 0.48485489466113885,
+      "grad_norm": 0.17225952446460724,
+      "learning_rate": 0.001,
+      "loss": 2.7225,
+      "step": 11461
+    },
+    {
+      "epoch": 0.4848971994246552,
+      "grad_norm": 0.555501401424408,
+      "learning_rate": 0.001,
+      "loss": 2.3918,
+      "step": 11462
+    },
+    {
+      "epoch": 0.4849395041881716,
+      "grad_norm": 0.16668234765529633,
+      "learning_rate": 0.001,
+      "loss": 2.0368,
+      "step": 11463
+    },
+    {
+      "epoch": 0.48498180895168796,
+      "grad_norm": 0.27358829975128174,
+      "learning_rate": 0.001,
+      "loss": 1.9445,
+      "step": 11464
+    },
+    {
+      "epoch": 0.4850241137152043,
+      "grad_norm": 0.18736232817173004,
+      "learning_rate": 0.001,
+      "loss": 2.7039,
+      "step": 11465
+    },
+    {
+      "epoch": 0.48506641847872073,
+      "grad_norm": 0.15612122416496277,
+      "learning_rate": 0.001,
+      "loss": 2.6935,
+      "step": 11466
+    },
+    {
+      "epoch": 0.4851087232422371,
+      "grad_norm": 2.6563851833343506,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 11467
+    },
+    {
+      "epoch": 0.48515102800575344,
+      "grad_norm": 0.1503165066242218,
+      "learning_rate": 0.001,
+      "loss": 2.1244,
+      "step": 11468
+    },
+    {
+      "epoch": 0.48519333276926985,
+      "grad_norm": 0.19594672322273254,
+      "learning_rate": 0.001,
+      "loss": 2.0981,
+      "step": 11469
+    },
+    {
+      "epoch": 0.4852356375327862,
+      "grad_norm": 0.1888262778520584,
+      "learning_rate": 0.001,
+      "loss": 2.8126,
+      "step": 11470
+    },
+    {
+      "epoch": 0.48527794229630256,
+      "grad_norm": 0.20222362875938416,
+      "learning_rate": 0.001,
+      "loss": 2.2918,
+      "step": 11471
+    },
+    {
+      "epoch": 0.4853202470598189,
+      "grad_norm": 1.1823304891586304,
+      "learning_rate": 0.001,
+      "loss": 1.7587,
+      "step": 11472
+    },
+    {
+      "epoch": 0.4853625518233353,
+      "grad_norm": 0.19512400031089783,
+      "learning_rate": 0.001,
+      "loss": 2.1734,
+      "step": 11473
+    },
+    {
+      "epoch": 0.4854048565868517,
+      "grad_norm": 1.402773380279541,
+      "learning_rate": 0.001,
+      "loss": 2.6338,
+      "step": 11474
+    },
+    {
+      "epoch": 0.48544716135036803,
+      "grad_norm": 0.22815269231796265,
+      "learning_rate": 0.001,
+      "loss": 2.9793,
+      "step": 11475
+    },
+    {
+      "epoch": 0.48548946611388444,
+      "grad_norm": 0.17088262736797333,
+      "learning_rate": 0.001,
+      "loss": 2.0868,
+      "step": 11476
+    },
+    {
+      "epoch": 0.4855317708774008,
+      "grad_norm": 0.16063320636749268,
+      "learning_rate": 0.001,
+      "loss": 1.6893,
+      "step": 11477
+    },
+    {
+      "epoch": 0.48557407564091715,
+      "grad_norm": 0.8555530309677124,
+      "learning_rate": 0.001,
+      "loss": 2.3043,
+      "step": 11478
+    },
+    {
+      "epoch": 0.48561638040443356,
+      "grad_norm": 0.27816876769065857,
+      "learning_rate": 0.001,
+      "loss": 3.0965,
+      "step": 11479
+    },
+    {
+      "epoch": 0.4856586851679499,
+      "grad_norm": 0.7349559664726257,
+      "learning_rate": 0.001,
+      "loss": 2.1048,
+      "step": 11480
+    },
+    {
+      "epoch": 0.48570098993146626,
+      "grad_norm": 0.16931132972240448,
+      "learning_rate": 0.001,
+      "loss": 2.0924,
+      "step": 11481
+    },
+    {
+      "epoch": 0.4857432946949827,
+      "grad_norm": 0.15484954416751862,
+      "learning_rate": 0.001,
+      "loss": 1.9129,
+      "step": 11482
+    },
+    {
+      "epoch": 0.48578559945849903,
+      "grad_norm": 0.3390078842639923,
+      "learning_rate": 0.001,
+      "loss": 2.2157,
+      "step": 11483
+    },
+    {
+      "epoch": 0.4858279042220154,
+      "grad_norm": 0.20084303617477417,
+      "learning_rate": 0.001,
+      "loss": 2.4884,
+      "step": 11484
+    },
+    {
+      "epoch": 0.4858702089855318,
+      "grad_norm": 0.24924889206886292,
+      "learning_rate": 0.001,
+      "loss": 2.4044,
+      "step": 11485
+    },
+    {
+      "epoch": 0.48591251374904815,
+      "grad_norm": 0.19741028547286987,
+      "learning_rate": 0.001,
+      "loss": 2.1854,
+      "step": 11486
+    },
+    {
+      "epoch": 0.4859548185125645,
+      "grad_norm": 0.20189212262630463,
+      "learning_rate": 0.001,
+      "loss": 2.4314,
+      "step": 11487
+    },
+    {
+      "epoch": 0.4859971232760809,
+      "grad_norm": 1.0460984706878662,
+      "learning_rate": 0.001,
+      "loss": 2.094,
+      "step": 11488
+    },
+    {
+      "epoch": 0.48603942803959727,
+      "grad_norm": 2.5869855880737305,
+      "learning_rate": 0.001,
+      "loss": 2.3084,
+      "step": 11489
+    },
+    {
+      "epoch": 0.4860817328031136,
+      "grad_norm": 0.25881245732307434,
+      "learning_rate": 0.001,
+      "loss": 3.5865,
+      "step": 11490
+    },
+    {
+      "epoch": 0.48612403756663003,
+      "grad_norm": 3.5187735557556152,
+      "learning_rate": 0.001,
+      "loss": 1.9536,
+      "step": 11491
+    },
+    {
+      "epoch": 0.4861663423301464,
+      "grad_norm": 0.179852694272995,
+      "learning_rate": 0.001,
+      "loss": 2.2011,
+      "step": 11492
+    },
+    {
+      "epoch": 0.48620864709366274,
+      "grad_norm": 0.37321174144744873,
+      "learning_rate": 0.001,
+      "loss": 2.0074,
+      "step": 11493
+    },
+    {
+      "epoch": 0.4862509518571791,
+      "grad_norm": 0.5720106959342957,
+      "learning_rate": 0.001,
+      "loss": 2.0811,
+      "step": 11494
+    },
+    {
+      "epoch": 0.4862932566206955,
+      "grad_norm": 0.31446918845176697,
+      "learning_rate": 0.001,
+      "loss": 2.0816,
+      "step": 11495
+    },
+    {
+      "epoch": 0.48633556138421186,
+      "grad_norm": 0.24225953221321106,
+      "learning_rate": 0.001,
+      "loss": 1.9998,
+      "step": 11496
+    },
+    {
+      "epoch": 0.4863778661477282,
+      "grad_norm": 0.20280331373214722,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 11497
+    },
+    {
+      "epoch": 0.4864201709112446,
+      "grad_norm": 0.1985984891653061,
+      "learning_rate": 0.001,
+      "loss": 2.5404,
+      "step": 11498
+    },
+    {
+      "epoch": 0.486462475674761,
+      "grad_norm": 0.25399044156074524,
+      "learning_rate": 0.001,
+      "loss": 2.2409,
+      "step": 11499
+    },
+    {
+      "epoch": 0.48650478043827733,
+      "grad_norm": 0.376498281955719,
+      "learning_rate": 0.001,
+      "loss": 2.584,
+      "step": 11500
+    },
+    {
+      "epoch": 0.48654708520179374,
+      "grad_norm": 0.15857632458209991,
+      "learning_rate": 0.001,
+      "loss": 2.2376,
+      "step": 11501
+    },
+    {
+      "epoch": 0.4865893899653101,
+      "grad_norm": 0.19442448019981384,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 11502
+    },
+    {
+      "epoch": 0.48663169472882645,
+      "grad_norm": 0.28401103615760803,
+      "learning_rate": 0.001,
+      "loss": 2.037,
+      "step": 11503
+    },
+    {
+      "epoch": 0.48667399949234286,
+      "grad_norm": 0.1915339082479477,
+      "learning_rate": 0.001,
+      "loss": 2.5886,
+      "step": 11504
+    },
+    {
+      "epoch": 0.4867163042558592,
+      "grad_norm": 0.13712449371814728,
+      "learning_rate": 0.001,
+      "loss": 1.6988,
+      "step": 11505
+    },
+    {
+      "epoch": 0.48675860901937557,
+      "grad_norm": 0.16903550922870636,
+      "learning_rate": 0.001,
+      "loss": 2.3553,
+      "step": 11506
+    },
+    {
+      "epoch": 0.486800913782892,
+      "grad_norm": 0.3302583396434784,
+      "learning_rate": 0.001,
+      "loss": 2.4041,
+      "step": 11507
+    },
+    {
+      "epoch": 0.48684321854640833,
+      "grad_norm": 0.16169388592243195,
+      "learning_rate": 0.001,
+      "loss": 1.5953,
+      "step": 11508
+    },
+    {
+      "epoch": 0.4868855233099247,
+      "grad_norm": 0.19090786576271057,
+      "learning_rate": 0.001,
+      "loss": 2.6609,
+      "step": 11509
+    },
+    {
+      "epoch": 0.4869278280734411,
+      "grad_norm": 0.2367997020483017,
+      "learning_rate": 0.001,
+      "loss": 2.7336,
+      "step": 11510
+    },
+    {
+      "epoch": 0.48697013283695745,
+      "grad_norm": 29.403594970703125,
+      "learning_rate": 0.001,
+      "loss": 2.4013,
+      "step": 11511
+    },
+    {
+      "epoch": 0.4870124376004738,
+      "grad_norm": 0.26779159903526306,
+      "learning_rate": 0.001,
+      "loss": 2.9784,
+      "step": 11512
+    },
+    {
+      "epoch": 0.4870547423639902,
+      "grad_norm": 0.4845396876335144,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 11513
+    },
+    {
+      "epoch": 0.48709704712750657,
+      "grad_norm": 0.5120859742164612,
+      "learning_rate": 0.001,
+      "loss": 1.9852,
+      "step": 11514
+    },
+    {
+      "epoch": 0.4871393518910229,
+      "grad_norm": 0.1959608644247055,
+      "learning_rate": 0.001,
+      "loss": 2.3048,
+      "step": 11515
+    },
+    {
+      "epoch": 0.4871816566545393,
+      "grad_norm": 5.542116641998291,
+      "learning_rate": 0.001,
+      "loss": 2.3137,
+      "step": 11516
+    },
+    {
+      "epoch": 0.4872239614180557,
+      "grad_norm": 0.26933497190475464,
+      "learning_rate": 0.001,
+      "loss": 2.1619,
+      "step": 11517
+    },
+    {
+      "epoch": 0.48726626618157204,
+      "grad_norm": 0.24525229632854462,
+      "learning_rate": 0.001,
+      "loss": 2.0777,
+      "step": 11518
+    },
+    {
+      "epoch": 0.4873085709450884,
+      "grad_norm": 0.20497070252895355,
+      "learning_rate": 0.001,
+      "loss": 2.317,
+      "step": 11519
+    },
+    {
+      "epoch": 0.4873508757086048,
+      "grad_norm": 0.31247276067733765,
+      "learning_rate": 0.001,
+      "loss": 2.4255,
+      "step": 11520
+    },
+    {
+      "epoch": 0.48739318047212116,
+      "grad_norm": 0.29333776235580444,
+      "learning_rate": 0.001,
+      "loss": 1.945,
+      "step": 11521
+    },
+    {
+      "epoch": 0.4874354852356375,
+      "grad_norm": 1.8194544315338135,
+      "learning_rate": 0.001,
+      "loss": 2.4755,
+      "step": 11522
+    },
+    {
+      "epoch": 0.4874777899991539,
+      "grad_norm": 0.182512104511261,
+      "learning_rate": 0.001,
+      "loss": 2.1394,
+      "step": 11523
+    },
+    {
+      "epoch": 0.4875200947626703,
+      "grad_norm": 0.19571569561958313,
+      "learning_rate": 0.001,
+      "loss": 2.2593,
+      "step": 11524
+    },
+    {
+      "epoch": 0.48756239952618663,
+      "grad_norm": 0.21386699378490448,
+      "learning_rate": 0.001,
+      "loss": 1.9803,
+      "step": 11525
+    },
+    {
+      "epoch": 0.48760470428970304,
+      "grad_norm": 1.0484811067581177,
+      "learning_rate": 0.001,
+      "loss": 1.9055,
+      "step": 11526
+    },
+    {
+      "epoch": 0.4876470090532194,
+      "grad_norm": 0.19381646811962128,
+      "learning_rate": 0.001,
+      "loss": 2.8984,
+      "step": 11527
+    },
+    {
+      "epoch": 0.48768931381673575,
+      "grad_norm": 0.18061847984790802,
+      "learning_rate": 0.001,
+      "loss": 2.9683,
+      "step": 11528
+    },
+    {
+      "epoch": 0.48773161858025216,
+      "grad_norm": 1.0904494524002075,
+      "learning_rate": 0.001,
+      "loss": 1.7487,
+      "step": 11529
+    },
+    {
+      "epoch": 0.4877739233437685,
+      "grad_norm": 0.5851990580558777,
+      "learning_rate": 0.001,
+      "loss": 4.4732,
+      "step": 11530
+    },
+    {
+      "epoch": 0.48781622810728487,
+      "grad_norm": 0.21629658341407776,
+      "learning_rate": 0.001,
+      "loss": 3.3176,
+      "step": 11531
+    },
+    {
+      "epoch": 0.4878585328708013,
+      "grad_norm": 0.1830849051475525,
+      "learning_rate": 0.001,
+      "loss": 1.9741,
+      "step": 11532
+    },
+    {
+      "epoch": 0.48790083763431763,
+      "grad_norm": 0.6447810530662537,
+      "learning_rate": 0.001,
+      "loss": 1.6206,
+      "step": 11533
+    },
+    {
+      "epoch": 0.487943142397834,
+      "grad_norm": 0.22213850915431976,
+      "learning_rate": 0.001,
+      "loss": 2.1542,
+      "step": 11534
+    },
+    {
+      "epoch": 0.4879854471613504,
+      "grad_norm": 0.6052437424659729,
+      "learning_rate": 0.001,
+      "loss": 2.0151,
+      "step": 11535
+    },
+    {
+      "epoch": 0.48802775192486675,
+      "grad_norm": 0.15839864313602448,
+      "learning_rate": 0.001,
+      "loss": 1.9729,
+      "step": 11536
+    },
+    {
+      "epoch": 0.4880700566883831,
+      "grad_norm": 1.3613576889038086,
+      "learning_rate": 0.001,
+      "loss": 2.4232,
+      "step": 11537
+    },
+    {
+      "epoch": 0.48811236145189946,
+      "grad_norm": 0.19417421519756317,
+      "learning_rate": 0.001,
+      "loss": 1.7537,
+      "step": 11538
+    },
+    {
+      "epoch": 0.48815466621541587,
+      "grad_norm": 0.3457423448562622,
+      "learning_rate": 0.001,
+      "loss": 2.1222,
+      "step": 11539
+    },
+    {
+      "epoch": 0.4881969709789322,
+      "grad_norm": 0.20159010589122772,
+      "learning_rate": 0.001,
+      "loss": 2.4426,
+      "step": 11540
+    },
+    {
+      "epoch": 0.4882392757424486,
+      "grad_norm": 0.188156396150589,
+      "learning_rate": 0.001,
+      "loss": 2.3599,
+      "step": 11541
+    },
+    {
+      "epoch": 0.488281580505965,
+      "grad_norm": 0.5900790095329285,
+      "learning_rate": 0.001,
+      "loss": 2.0175,
+      "step": 11542
+    },
+    {
+      "epoch": 0.48832388526948134,
+      "grad_norm": 0.2750971019268036,
+      "learning_rate": 0.001,
+      "loss": 2.0602,
+      "step": 11543
+    },
+    {
+      "epoch": 0.4883661900329977,
+      "grad_norm": 0.1912364661693573,
+      "learning_rate": 0.001,
+      "loss": 2.2475,
+      "step": 11544
+    },
+    {
+      "epoch": 0.4884084947965141,
+      "grad_norm": 0.20477798581123352,
+      "learning_rate": 0.001,
+      "loss": 2.7998,
+      "step": 11545
+    },
+    {
+      "epoch": 0.48845079956003046,
+      "grad_norm": 0.1736663430929184,
+      "learning_rate": 0.001,
+      "loss": 2.1871,
+      "step": 11546
+    },
+    {
+      "epoch": 0.4884931043235468,
+      "grad_norm": 0.1695953756570816,
+      "learning_rate": 0.001,
+      "loss": 2.4564,
+      "step": 11547
+    },
+    {
+      "epoch": 0.4885354090870632,
+      "grad_norm": 0.19503745436668396,
+      "learning_rate": 0.001,
+      "loss": 2.5058,
+      "step": 11548
+    },
+    {
+      "epoch": 0.4885777138505796,
+      "grad_norm": 0.1817982941865921,
+      "learning_rate": 0.001,
+      "loss": 2.7636,
+      "step": 11549
+    },
+    {
+      "epoch": 0.48862001861409593,
+      "grad_norm": 0.18275293707847595,
+      "learning_rate": 0.001,
+      "loss": 3.3698,
+      "step": 11550
+    },
+    {
+      "epoch": 0.48866232337761234,
+      "grad_norm": 5.7995758056640625,
+      "learning_rate": 0.001,
+      "loss": 3.3286,
+      "step": 11551
+    },
+    {
+      "epoch": 0.4887046281411287,
+      "grad_norm": 0.1749723106622696,
+      "learning_rate": 0.001,
+      "loss": 2.4678,
+      "step": 11552
+    },
+    {
+      "epoch": 0.48874693290464505,
+      "grad_norm": 0.22669129073619843,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 11553
+    },
+    {
+      "epoch": 0.48878923766816146,
+      "grad_norm": 0.4565039277076721,
+      "learning_rate": 0.001,
+      "loss": 2.2781,
+      "step": 11554
+    },
+    {
+      "epoch": 0.4888315424316778,
+      "grad_norm": 0.15998007357120514,
+      "learning_rate": 0.001,
+      "loss": 2.4679,
+      "step": 11555
+    },
+    {
+      "epoch": 0.48887384719519417,
+      "grad_norm": 0.17309017479419708,
+      "learning_rate": 0.001,
+      "loss": 2.0567,
+      "step": 11556
+    },
+    {
+      "epoch": 0.4889161519587106,
+      "grad_norm": 0.16389106214046478,
+      "learning_rate": 0.001,
+      "loss": 2.0645,
+      "step": 11557
+    },
+    {
+      "epoch": 0.48895845672222693,
+      "grad_norm": 0.1628233641386032,
+      "learning_rate": 0.001,
+      "loss": 2.142,
+      "step": 11558
+    },
+    {
+      "epoch": 0.4890007614857433,
+      "grad_norm": 0.17292705178260803,
+      "learning_rate": 0.001,
+      "loss": 2.4255,
+      "step": 11559
+    },
+    {
+      "epoch": 0.48904306624925964,
+      "grad_norm": 0.13833680748939514,
+      "learning_rate": 0.001,
+      "loss": 1.5506,
+      "step": 11560
+    },
+    {
+      "epoch": 0.48908537101277605,
+      "grad_norm": 0.1819964051246643,
+      "learning_rate": 0.001,
+      "loss": 2.4794,
+      "step": 11561
+    },
+    {
+      "epoch": 0.4891276757762924,
+      "grad_norm": 0.17787259817123413,
+      "learning_rate": 0.001,
+      "loss": 2.5472,
+      "step": 11562
+    },
+    {
+      "epoch": 0.48916998053980876,
+      "grad_norm": 0.1847681850194931,
+      "learning_rate": 0.001,
+      "loss": 1.929,
+      "step": 11563
+    },
+    {
+      "epoch": 0.48921228530332517,
+      "grad_norm": 0.16343270242214203,
+      "learning_rate": 0.001,
+      "loss": 1.9423,
+      "step": 11564
+    },
+    {
+      "epoch": 0.4892545900668415,
+      "grad_norm": 0.18636071681976318,
+      "learning_rate": 0.001,
+      "loss": 1.5396,
+      "step": 11565
+    },
+    {
+      "epoch": 0.4892968948303579,
+      "grad_norm": 0.3037877678871155,
+      "learning_rate": 0.001,
+      "loss": 1.9659,
+      "step": 11566
+    },
+    {
+      "epoch": 0.4893391995938743,
+      "grad_norm": 5.196141242980957,
+      "learning_rate": 0.001,
+      "loss": 2.4534,
+      "step": 11567
+    },
+    {
+      "epoch": 0.48938150435739064,
+      "grad_norm": 0.16297638416290283,
+      "learning_rate": 0.001,
+      "loss": 2.0671,
+      "step": 11568
+    },
+    {
+      "epoch": 0.489423809120907,
+      "grad_norm": 0.14544682204723358,
+      "learning_rate": 0.001,
+      "loss": 2.069,
+      "step": 11569
+    },
+    {
+      "epoch": 0.4894661138844234,
+      "grad_norm": 0.922217071056366,
+      "learning_rate": 0.001,
+      "loss": 2.0644,
+      "step": 11570
+    },
+    {
+      "epoch": 0.48950841864793976,
+      "grad_norm": 0.13835956156253815,
+      "learning_rate": 0.001,
+      "loss": 1.6519,
+      "step": 11571
+    },
+    {
+      "epoch": 0.4895507234114561,
+      "grad_norm": 0.3947632610797882,
+      "learning_rate": 0.001,
+      "loss": 2.2417,
+      "step": 11572
+    },
+    {
+      "epoch": 0.4895930281749725,
+      "grad_norm": 0.15445981919765472,
+      "learning_rate": 0.001,
+      "loss": 2.5499,
+      "step": 11573
+    },
+    {
+      "epoch": 0.4896353329384889,
+      "grad_norm": 0.17741024494171143,
+      "learning_rate": 0.001,
+      "loss": 1.8943,
+      "step": 11574
+    },
+    {
+      "epoch": 0.48967763770200523,
+      "grad_norm": 0.17265914380550385,
+      "learning_rate": 0.001,
+      "loss": 2.0673,
+      "step": 11575
+    },
+    {
+      "epoch": 0.48971994246552164,
+      "grad_norm": 0.640771210193634,
+      "learning_rate": 0.001,
+      "loss": 3.3352,
+      "step": 11576
+    },
+    {
+      "epoch": 0.489762247229038,
+      "grad_norm": 0.17222177982330322,
+      "learning_rate": 0.001,
+      "loss": 1.9422,
+      "step": 11577
+    },
+    {
+      "epoch": 0.48980455199255435,
+      "grad_norm": 0.18314416706562042,
+      "learning_rate": 0.001,
+      "loss": 1.9239,
+      "step": 11578
+    },
+    {
+      "epoch": 0.48984685675607076,
+      "grad_norm": 0.3542311489582062,
+      "learning_rate": 0.001,
+      "loss": 3.8196,
+      "step": 11579
+    },
+    {
+      "epoch": 0.4898891615195871,
+      "grad_norm": 0.14956578612327576,
+      "learning_rate": 0.001,
+      "loss": 1.9715,
+      "step": 11580
+    },
+    {
+      "epoch": 0.48993146628310347,
+      "grad_norm": 0.17212481796741486,
+      "learning_rate": 0.001,
+      "loss": 2.7058,
+      "step": 11581
+    },
+    {
+      "epoch": 0.4899737710466199,
+      "grad_norm": 0.2047460526227951,
+      "learning_rate": 0.001,
+      "loss": 2.4195,
+      "step": 11582
+    },
+    {
+      "epoch": 0.49001607581013623,
+      "grad_norm": 0.16925156116485596,
+      "learning_rate": 0.001,
+      "loss": 1.4925,
+      "step": 11583
+    },
+    {
+      "epoch": 0.4900583805736526,
+      "grad_norm": 0.17683045566082,
+      "learning_rate": 0.001,
+      "loss": 2.1021,
+      "step": 11584
+    },
+    {
+      "epoch": 0.49010068533716894,
+      "grad_norm": 0.19544382393360138,
+      "learning_rate": 0.001,
+      "loss": 2.8976,
+      "step": 11585
+    },
+    {
+      "epoch": 0.49014299010068535,
+      "grad_norm": 0.17539872229099274,
+      "learning_rate": 0.001,
+      "loss": 2.7534,
+      "step": 11586
+    },
+    {
+      "epoch": 0.4901852948642017,
+      "grad_norm": 0.17753051221370697,
+      "learning_rate": 0.001,
+      "loss": 2.2047,
+      "step": 11587
+    },
+    {
+      "epoch": 0.49022759962771806,
+      "grad_norm": 0.16815239191055298,
+      "learning_rate": 0.001,
+      "loss": 1.6325,
+      "step": 11588
+    },
+    {
+      "epoch": 0.49026990439123447,
+      "grad_norm": 0.2511793375015259,
+      "learning_rate": 0.001,
+      "loss": 2.8271,
+      "step": 11589
+    },
+    {
+      "epoch": 0.4903122091547508,
+      "grad_norm": 3.128819465637207,
+      "learning_rate": 0.001,
+      "loss": 1.7008,
+      "step": 11590
+    },
+    {
+      "epoch": 0.4903545139182672,
+      "grad_norm": 0.2257847785949707,
+      "learning_rate": 0.001,
+      "loss": 2.4141,
+      "step": 11591
+    },
+    {
+      "epoch": 0.4903968186817836,
+      "grad_norm": 1.591920256614685,
+      "learning_rate": 0.001,
+      "loss": 1.9112,
+      "step": 11592
+    },
+    {
+      "epoch": 0.49043912344529994,
+      "grad_norm": 0.2209502011537552,
+      "learning_rate": 0.001,
+      "loss": 2.9542,
+      "step": 11593
+    },
+    {
+      "epoch": 0.4904814282088163,
+      "grad_norm": 0.19603808224201202,
+      "learning_rate": 0.001,
+      "loss": 2.5474,
+      "step": 11594
+    },
+    {
+      "epoch": 0.4905237329723327,
+      "grad_norm": 0.6029075980186462,
+      "learning_rate": 0.001,
+      "loss": 1.8916,
+      "step": 11595
+    },
+    {
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.1723041832447052,
+      "learning_rate": 0.001,
+      "loss": 2.0029,
+      "step": 11596
+    },
+    {
+      "epoch": 0.4906083424993654,
+      "grad_norm": 0.156670480966568,
+      "learning_rate": 0.001,
+      "loss": 1.8345,
+      "step": 11597
+    },
+    {
+      "epoch": 0.4906506472628818,
+      "grad_norm": 0.2527156174182892,
+      "learning_rate": 0.001,
+      "loss": 3.4863,
+      "step": 11598
+    },
+    {
+      "epoch": 0.4906929520263982,
+      "grad_norm": 0.15961304306983948,
+      "learning_rate": 0.001,
+      "loss": 2.3809,
+      "step": 11599
+    },
+    {
+      "epoch": 0.49073525678991453,
+      "grad_norm": 0.16630440950393677,
+      "learning_rate": 0.001,
+      "loss": 2.33,
+      "step": 11600
+    },
+    {
+      "epoch": 0.49077756155343094,
+      "grad_norm": 0.1635364443063736,
+      "learning_rate": 0.001,
+      "loss": 2.1123,
+      "step": 11601
+    },
+    {
+      "epoch": 0.4908198663169473,
+      "grad_norm": 12.366936683654785,
+      "learning_rate": 0.001,
+      "loss": 2.755,
+      "step": 11602
+    },
+    {
+      "epoch": 0.49086217108046365,
+      "grad_norm": 0.16810069978237152,
+      "learning_rate": 0.001,
+      "loss": 2.0871,
+      "step": 11603
+    },
+    {
+      "epoch": 0.49090447584398006,
+      "grad_norm": 0.16595399379730225,
+      "learning_rate": 0.001,
+      "loss": 2.1468,
+      "step": 11604
+    },
+    {
+      "epoch": 0.4909467806074964,
+      "grad_norm": 0.5384510159492493,
+      "learning_rate": 0.001,
+      "loss": 2.1914,
+      "step": 11605
+    },
+    {
+      "epoch": 0.49098908537101277,
+      "grad_norm": 0.18624725937843323,
+      "learning_rate": 0.001,
+      "loss": 3.0921,
+      "step": 11606
+    },
+    {
+      "epoch": 0.4910313901345291,
+      "grad_norm": 0.18056440353393555,
+      "learning_rate": 0.001,
+      "loss": 1.936,
+      "step": 11607
+    },
+    {
+      "epoch": 0.49107369489804553,
+      "grad_norm": 1.4659069776535034,
+      "learning_rate": 0.001,
+      "loss": 2.5418,
+      "step": 11608
+    },
+    {
+      "epoch": 0.4911159996615619,
+      "grad_norm": 0.23289033770561218,
+      "learning_rate": 0.001,
+      "loss": 3.2947,
+      "step": 11609
+    },
+    {
+      "epoch": 0.49115830442507824,
+      "grad_norm": 0.18133248388767242,
+      "learning_rate": 0.001,
+      "loss": 1.8595,
+      "step": 11610
+    },
+    {
+      "epoch": 0.49120060918859465,
+      "grad_norm": 0.16130225360393524,
+      "learning_rate": 0.001,
+      "loss": 2.1366,
+      "step": 11611
+    },
+    {
+      "epoch": 0.491242913952111,
+      "grad_norm": 0.6706212759017944,
+      "learning_rate": 0.001,
+      "loss": 3.0816,
+      "step": 11612
+    },
+    {
+      "epoch": 0.49128521871562736,
+      "grad_norm": 0.2708061635494232,
+      "learning_rate": 0.001,
+      "loss": 2.2324,
+      "step": 11613
+    },
+    {
+      "epoch": 0.49132752347914377,
+      "grad_norm": 0.1868041753768921,
+      "learning_rate": 0.001,
+      "loss": 2.1089,
+      "step": 11614
+    },
+    {
+      "epoch": 0.4913698282426601,
+      "grad_norm": 0.21697242558002472,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 11615
+    },
+    {
+      "epoch": 0.4914121330061765,
+      "grad_norm": 0.1955055445432663,
+      "learning_rate": 0.001,
+      "loss": 2.3835,
+      "step": 11616
+    },
+    {
+      "epoch": 0.4914544377696929,
+      "grad_norm": 0.18908244371414185,
+      "learning_rate": 0.001,
+      "loss": 3.2207,
+      "step": 11617
+    },
+    {
+      "epoch": 0.49149674253320924,
+      "grad_norm": 0.17205405235290527,
+      "learning_rate": 0.001,
+      "loss": 2.0338,
+      "step": 11618
+    },
+    {
+      "epoch": 0.4915390472967256,
+      "grad_norm": 0.2222311645746231,
+      "learning_rate": 0.001,
+      "loss": 2.4758,
+      "step": 11619
+    },
+    {
+      "epoch": 0.491581352060242,
+      "grad_norm": 0.552980899810791,
+      "learning_rate": 0.001,
+      "loss": 1.7828,
+      "step": 11620
+    },
+    {
+      "epoch": 0.49162365682375836,
+      "grad_norm": 0.17426225543022156,
+      "learning_rate": 0.001,
+      "loss": 2.4719,
+      "step": 11621
+    },
+    {
+      "epoch": 0.4916659615872747,
+      "grad_norm": 0.17539294064044952,
+      "learning_rate": 0.001,
+      "loss": 2.2529,
+      "step": 11622
+    },
+    {
+      "epoch": 0.4917082663507911,
+      "grad_norm": 0.38558903336524963,
+      "learning_rate": 0.001,
+      "loss": 1.621,
+      "step": 11623
+    },
+    {
+      "epoch": 0.4917505711143075,
+      "grad_norm": 12.589545249938965,
+      "learning_rate": 0.001,
+      "loss": 3.9498,
+      "step": 11624
+    },
+    {
+      "epoch": 0.49179287587782383,
+      "grad_norm": 0.18124458193778992,
+      "learning_rate": 0.001,
+      "loss": 2.783,
+      "step": 11625
+    },
+    {
+      "epoch": 0.49183518064134024,
+      "grad_norm": 0.1963767558336258,
+      "learning_rate": 0.001,
+      "loss": 1.9605,
+      "step": 11626
+    },
+    {
+      "epoch": 0.4918774854048566,
+      "grad_norm": 0.6345813274383545,
+      "learning_rate": 0.001,
+      "loss": 2.0434,
+      "step": 11627
+    },
+    {
+      "epoch": 0.49191979016837295,
+      "grad_norm": 0.21551281213760376,
+      "learning_rate": 0.001,
+      "loss": 3.1873,
+      "step": 11628
+    },
+    {
+      "epoch": 0.4919620949318893,
+      "grad_norm": 0.17613928020000458,
+      "learning_rate": 0.001,
+      "loss": 2.0574,
+      "step": 11629
+    },
+    {
+      "epoch": 0.4920043996954057,
+      "grad_norm": 0.20698675513267517,
+      "learning_rate": 0.001,
+      "loss": 3.0403,
+      "step": 11630
+    },
+    {
+      "epoch": 0.49204670445892207,
+      "grad_norm": 0.2309427410364151,
+      "learning_rate": 0.001,
+      "loss": 2.8583,
+      "step": 11631
+    },
+    {
+      "epoch": 0.4920890092224384,
+      "grad_norm": 0.18390582501888275,
+      "learning_rate": 0.001,
+      "loss": 1.6705,
+      "step": 11632
+    },
+    {
+      "epoch": 0.49213131398595483,
+      "grad_norm": 0.2480992078781128,
+      "learning_rate": 0.001,
+      "loss": 2.2437,
+      "step": 11633
+    },
+    {
+      "epoch": 0.4921736187494712,
+      "grad_norm": 0.5748304724693298,
+      "learning_rate": 0.001,
+      "loss": 2.3204,
+      "step": 11634
+    },
+    {
+      "epoch": 0.49221592351298754,
+      "grad_norm": 0.151663139462471,
+      "learning_rate": 0.001,
+      "loss": 2.4468,
+      "step": 11635
+    },
+    {
+      "epoch": 0.49225822827650395,
+      "grad_norm": 0.7663713693618774,
+      "learning_rate": 0.001,
+      "loss": 2.0351,
+      "step": 11636
+    },
+    {
+      "epoch": 0.4923005330400203,
+      "grad_norm": 0.2127770632505417,
+      "learning_rate": 0.001,
+      "loss": 2.0751,
+      "step": 11637
+    },
+    {
+      "epoch": 0.49234283780353666,
+      "grad_norm": 0.6651235222816467,
+      "learning_rate": 0.001,
+      "loss": 2.5762,
+      "step": 11638
+    },
+    {
+      "epoch": 0.49238514256705307,
+      "grad_norm": 0.18715116381645203,
+      "learning_rate": 0.001,
+      "loss": 1.8795,
+      "step": 11639
+    },
+    {
+      "epoch": 0.4924274473305694,
+      "grad_norm": 0.45163238048553467,
+      "learning_rate": 0.001,
+      "loss": 2.6674,
+      "step": 11640
+    },
+    {
+      "epoch": 0.4924697520940858,
+      "grad_norm": 0.2655963599681854,
+      "learning_rate": 0.001,
+      "loss": 1.9527,
+      "step": 11641
+    },
+    {
+      "epoch": 0.4925120568576022,
+      "grad_norm": 0.2013363540172577,
+      "learning_rate": 0.001,
+      "loss": 3.415,
+      "step": 11642
+    },
+    {
+      "epoch": 0.49255436162111854,
+      "grad_norm": 0.20579899847507477,
+      "learning_rate": 0.001,
+      "loss": 2.5534,
+      "step": 11643
+    },
+    {
+      "epoch": 0.4925966663846349,
+      "grad_norm": 0.6283543705940247,
+      "learning_rate": 0.001,
+      "loss": 2.4812,
+      "step": 11644
+    },
+    {
+      "epoch": 0.4926389711481513,
+      "grad_norm": 0.19560718536376953,
+      "learning_rate": 0.001,
+      "loss": 1.6794,
+      "step": 11645
+    },
+    {
+      "epoch": 0.49268127591166766,
+      "grad_norm": 0.20248031616210938,
+      "learning_rate": 0.001,
+      "loss": 2.4745,
+      "step": 11646
+    },
+    {
+      "epoch": 0.492723580675184,
+      "grad_norm": 0.20276713371276855,
+      "learning_rate": 0.001,
+      "loss": 2.1166,
+      "step": 11647
+    },
+    {
+      "epoch": 0.4927658854387004,
+      "grad_norm": 0.21462062001228333,
+      "learning_rate": 0.001,
+      "loss": 2.0011,
+      "step": 11648
+    },
+    {
+      "epoch": 0.4928081902022168,
+      "grad_norm": 0.1634545773267746,
+      "learning_rate": 0.001,
+      "loss": 1.8337,
+      "step": 11649
+    },
+    {
+      "epoch": 0.49285049496573313,
+      "grad_norm": 0.18558084964752197,
+      "learning_rate": 0.001,
+      "loss": 1.4515,
+      "step": 11650
+    },
+    {
+      "epoch": 0.4928927997292495,
+      "grad_norm": 0.21327702701091766,
+      "learning_rate": 0.001,
+      "loss": 2.023,
+      "step": 11651
+    },
+    {
+      "epoch": 0.4929351044927659,
+      "grad_norm": 0.18165309727191925,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 11652
+    },
+    {
+      "epoch": 0.49297740925628225,
+      "grad_norm": 0.16911716759204865,
+      "learning_rate": 0.001,
+      "loss": 2.1337,
+      "step": 11653
+    },
+    {
+      "epoch": 0.4930197140197986,
+      "grad_norm": 0.7974897623062134,
+      "learning_rate": 0.001,
+      "loss": 3.1178,
+      "step": 11654
+    },
+    {
+      "epoch": 0.493062018783315,
+      "grad_norm": 0.2036375254392624,
+      "learning_rate": 0.001,
+      "loss": 2.142,
+      "step": 11655
+    },
+    {
+      "epoch": 0.49310432354683137,
+      "grad_norm": 0.21488189697265625,
+      "learning_rate": 0.001,
+      "loss": 2.9894,
+      "step": 11656
+    },
+    {
+      "epoch": 0.4931466283103477,
+      "grad_norm": 0.1535966992378235,
+      "learning_rate": 0.001,
+      "loss": 2.5775,
+      "step": 11657
+    },
+    {
+      "epoch": 0.49318893307386413,
+      "grad_norm": 0.19664905965328217,
+      "learning_rate": 0.001,
+      "loss": 2.1606,
+      "step": 11658
+    },
+    {
+      "epoch": 0.4932312378373805,
+      "grad_norm": 0.45215147733688354,
+      "learning_rate": 0.001,
+      "loss": 1.9469,
+      "step": 11659
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 0.17420543730258942,
+      "learning_rate": 0.001,
+      "loss": 1.9453,
+      "step": 11660
+    },
+    {
+      "epoch": 0.49331584736441325,
+      "grad_norm": 0.1706647127866745,
+      "learning_rate": 0.001,
+      "loss": 2.4864,
+      "step": 11661
+    },
+    {
+      "epoch": 0.4933581521279296,
+      "grad_norm": 0.17685231566429138,
+      "learning_rate": 0.001,
+      "loss": 2.1576,
+      "step": 11662
+    },
+    {
+      "epoch": 0.49340045689144596,
+      "grad_norm": 0.14751426875591278,
+      "learning_rate": 0.001,
+      "loss": 2.7795,
+      "step": 11663
+    },
+    {
+      "epoch": 0.49344276165496237,
+      "grad_norm": 0.33146944642066956,
+      "learning_rate": 0.001,
+      "loss": 2.1473,
+      "step": 11664
+    },
+    {
+      "epoch": 0.4934850664184787,
+      "grad_norm": 0.1822434514760971,
+      "learning_rate": 0.001,
+      "loss": 2.3431,
+      "step": 11665
+    },
+    {
+      "epoch": 0.4935273711819951,
+      "grad_norm": 0.17229153215885162,
+      "learning_rate": 0.001,
+      "loss": 1.9659,
+      "step": 11666
+    },
+    {
+      "epoch": 0.4935696759455115,
+      "grad_norm": 0.17021816968917847,
+      "learning_rate": 0.001,
+      "loss": 2.4419,
+      "step": 11667
+    },
+    {
+      "epoch": 0.49361198070902784,
+      "grad_norm": 0.15960778295993805,
+      "learning_rate": 0.001,
+      "loss": 1.9672,
+      "step": 11668
+    },
+    {
+      "epoch": 0.4936542854725442,
+      "grad_norm": 0.28872665762901306,
+      "learning_rate": 0.001,
+      "loss": 4.1413,
+      "step": 11669
+    },
+    {
+      "epoch": 0.4936965902360606,
+      "grad_norm": 0.16721804440021515,
+      "learning_rate": 0.001,
+      "loss": 1.9916,
+      "step": 11670
+    },
+    {
+      "epoch": 0.49373889499957696,
+      "grad_norm": 0.16407166421413422,
+      "learning_rate": 0.001,
+      "loss": 1.6337,
+      "step": 11671
+    },
+    {
+      "epoch": 0.4937811997630933,
+      "grad_norm": 0.16153931617736816,
+      "learning_rate": 0.001,
+      "loss": 1.9922,
+      "step": 11672
+    },
+    {
+      "epoch": 0.49382350452660967,
+      "grad_norm": 0.14842648804187775,
+      "learning_rate": 0.001,
+      "loss": 1.9083,
+      "step": 11673
+    },
+    {
+      "epoch": 0.4938658092901261,
+      "grad_norm": 0.301229327917099,
+      "learning_rate": 0.001,
+      "loss": 2.2091,
+      "step": 11674
+    },
+    {
+      "epoch": 0.49390811405364243,
+      "grad_norm": 0.1671200543642044,
+      "learning_rate": 0.001,
+      "loss": 2.6286,
+      "step": 11675
+    },
+    {
+      "epoch": 0.4939504188171588,
+      "grad_norm": 0.1930837333202362,
+      "learning_rate": 0.001,
+      "loss": 2.7532,
+      "step": 11676
+    },
+    {
+      "epoch": 0.4939927235806752,
+      "grad_norm": 0.1453826129436493,
+      "learning_rate": 0.001,
+      "loss": 2.1149,
+      "step": 11677
+    },
+    {
+      "epoch": 0.49403502834419155,
+      "grad_norm": 0.16736523807048798,
+      "learning_rate": 0.001,
+      "loss": 2.0743,
+      "step": 11678
+    },
+    {
+      "epoch": 0.4940773331077079,
+      "grad_norm": 0.14059995114803314,
+      "learning_rate": 0.001,
+      "loss": 1.8672,
+      "step": 11679
+    },
+    {
+      "epoch": 0.4941196378712243,
+      "grad_norm": 0.8272671699523926,
+      "learning_rate": 0.001,
+      "loss": 2.879,
+      "step": 11680
+    },
+    {
+      "epoch": 0.49416194263474067,
+      "grad_norm": 0.1517970710992813,
+      "learning_rate": 0.001,
+      "loss": 2.3603,
+      "step": 11681
+    },
+    {
+      "epoch": 0.494204247398257,
+      "grad_norm": 0.19491681456565857,
+      "learning_rate": 0.001,
+      "loss": 1.7711,
+      "step": 11682
+    },
+    {
+      "epoch": 0.49424655216177343,
+      "grad_norm": 0.16503435373306274,
+      "learning_rate": 0.001,
+      "loss": 2.2452,
+      "step": 11683
+    },
+    {
+      "epoch": 0.4942888569252898,
+      "grad_norm": 0.2520374059677124,
+      "learning_rate": 0.001,
+      "loss": 2.4305,
+      "step": 11684
+    },
+    {
+      "epoch": 0.49433116168880614,
+      "grad_norm": 0.325879842042923,
+      "learning_rate": 0.001,
+      "loss": 1.5981,
+      "step": 11685
+    },
+    {
+      "epoch": 0.49437346645232255,
+      "grad_norm": 0.17779172956943512,
+      "learning_rate": 0.001,
+      "loss": 2.5801,
+      "step": 11686
+    },
+    {
+      "epoch": 0.4944157712158389,
+      "grad_norm": 0.21204549074172974,
+      "learning_rate": 0.001,
+      "loss": 2.7278,
+      "step": 11687
+    },
+    {
+      "epoch": 0.49445807597935526,
+      "grad_norm": 0.21863149106502533,
+      "learning_rate": 0.001,
+      "loss": 2.3487,
+      "step": 11688
+    },
+    {
+      "epoch": 0.49450038074287167,
+      "grad_norm": 0.2039804309606552,
+      "learning_rate": 0.001,
+      "loss": 2.3773,
+      "step": 11689
+    },
+    {
+      "epoch": 0.494542685506388,
+      "grad_norm": 0.39557844400405884,
+      "learning_rate": 0.001,
+      "loss": 1.7114,
+      "step": 11690
+    },
+    {
+      "epoch": 0.4945849902699044,
+      "grad_norm": 0.28419822454452515,
+      "learning_rate": 0.001,
+      "loss": 1.9244,
+      "step": 11691
+    },
+    {
+      "epoch": 0.4946272950334208,
+      "grad_norm": 3.456125020980835,
+      "learning_rate": 0.001,
+      "loss": 2.3652,
+      "step": 11692
+    },
+    {
+      "epoch": 0.49466959979693714,
+      "grad_norm": 0.1735409051179886,
+      "learning_rate": 0.001,
+      "loss": 1.5861,
+      "step": 11693
+    },
+    {
+      "epoch": 0.4947119045604535,
+      "grad_norm": 0.2236192673444748,
+      "learning_rate": 0.001,
+      "loss": 2.8274,
+      "step": 11694
+    },
+    {
+      "epoch": 0.49475420932396985,
+      "grad_norm": 4.79661226272583,
+      "learning_rate": 0.001,
+      "loss": 2.3105,
+      "step": 11695
+    },
+    {
+      "epoch": 0.49479651408748626,
+      "grad_norm": 0.15469114482402802,
+      "learning_rate": 0.001,
+      "loss": 1.5461,
+      "step": 11696
+    },
+    {
+      "epoch": 0.4948388188510026,
+      "grad_norm": 0.383112370967865,
+      "learning_rate": 0.001,
+      "loss": 1.8817,
+      "step": 11697
+    },
+    {
+      "epoch": 0.49488112361451897,
+      "grad_norm": 0.17388460040092468,
+      "learning_rate": 0.001,
+      "loss": 2.0987,
+      "step": 11698
+    },
+    {
+      "epoch": 0.4949234283780354,
+      "grad_norm": 0.21153829991817474,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 11699
+    },
+    {
+      "epoch": 0.49496573314155173,
+      "grad_norm": 0.19076654314994812,
+      "learning_rate": 0.001,
+      "loss": 2.7168,
+      "step": 11700
+    },
+    {
+      "epoch": 0.4950080379050681,
+      "grad_norm": 0.17775197327136993,
+      "learning_rate": 0.001,
+      "loss": 2.3872,
+      "step": 11701
+    },
+    {
+      "epoch": 0.4950503426685845,
+      "grad_norm": 0.18127349019050598,
+      "learning_rate": 0.001,
+      "loss": 2.3911,
+      "step": 11702
+    },
+    {
+      "epoch": 0.49509264743210085,
+      "grad_norm": 0.20719602704048157,
+      "learning_rate": 0.001,
+      "loss": 2.4219,
+      "step": 11703
+    },
+    {
+      "epoch": 0.4951349521956172,
+      "grad_norm": 1.0996108055114746,
+      "learning_rate": 0.001,
+      "loss": 2.0785,
+      "step": 11704
+    },
+    {
+      "epoch": 0.4951772569591336,
+      "grad_norm": 0.20830421149730682,
+      "learning_rate": 0.001,
+      "loss": 2.3977,
+      "step": 11705
+    },
+    {
+      "epoch": 0.49521956172264997,
+      "grad_norm": 0.16215115785598755,
+      "learning_rate": 0.001,
+      "loss": 2.1351,
+      "step": 11706
+    },
+    {
+      "epoch": 0.4952618664861663,
+      "grad_norm": 0.18560096621513367,
+      "learning_rate": 0.001,
+      "loss": 2.045,
+      "step": 11707
+    },
+    {
+      "epoch": 0.49530417124968273,
+      "grad_norm": 1.037510871887207,
+      "learning_rate": 0.001,
+      "loss": 3.0518,
+      "step": 11708
+    },
+    {
+      "epoch": 0.4953464760131991,
+      "grad_norm": 0.197851300239563,
+      "learning_rate": 0.001,
+      "loss": 2.118,
+      "step": 11709
+    },
+    {
+      "epoch": 0.49538878077671544,
+      "grad_norm": 0.38033074140548706,
+      "learning_rate": 0.001,
+      "loss": 1.8406,
+      "step": 11710
+    },
+    {
+      "epoch": 0.49543108554023185,
+      "grad_norm": 0.17631156742572784,
+      "learning_rate": 0.001,
+      "loss": 1.9683,
+      "step": 11711
+    },
+    {
+      "epoch": 0.4954733903037482,
+      "grad_norm": 51.784454345703125,
+      "learning_rate": 0.001,
+      "loss": 2.3855,
+      "step": 11712
+    },
+    {
+      "epoch": 0.49551569506726456,
+      "grad_norm": 0.14018891751766205,
+      "learning_rate": 0.001,
+      "loss": 2.3842,
+      "step": 11713
+    },
+    {
+      "epoch": 0.49555799983078097,
+      "grad_norm": 0.2898732125759125,
+      "learning_rate": 0.001,
+      "loss": 2.951,
+      "step": 11714
+    },
+    {
+      "epoch": 0.4956003045942973,
+      "grad_norm": 0.1552821397781372,
+      "learning_rate": 0.001,
+      "loss": 1.5485,
+      "step": 11715
+    },
+    {
+      "epoch": 0.4956426093578137,
+      "grad_norm": 0.2228671759366989,
+      "learning_rate": 0.001,
+      "loss": 3.3397,
+      "step": 11716
+    },
+    {
+      "epoch": 0.4956849141213301,
+      "grad_norm": 0.17871029675006866,
+      "learning_rate": 0.001,
+      "loss": 2.4764,
+      "step": 11717
+    },
+    {
+      "epoch": 0.49572721888484644,
+      "grad_norm": 0.15905271470546722,
+      "learning_rate": 0.001,
+      "loss": 2.8845,
+      "step": 11718
+    },
+    {
+      "epoch": 0.4957695236483628,
+      "grad_norm": 0.19031955301761627,
+      "learning_rate": 0.001,
+      "loss": 2.0218,
+      "step": 11719
+    },
+    {
+      "epoch": 0.49581182841187915,
+      "grad_norm": 0.1882801651954651,
+      "learning_rate": 0.001,
+      "loss": 2.6585,
+      "step": 11720
+    },
+    {
+      "epoch": 0.49585413317539556,
+      "grad_norm": 0.19072453677654266,
+      "learning_rate": 0.001,
+      "loss": 1.9483,
+      "step": 11721
+    },
+    {
+      "epoch": 0.4958964379389119,
+      "grad_norm": 0.1833895891904831,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 11722
+    },
+    {
+      "epoch": 0.49593874270242827,
+      "grad_norm": 0.1797013133764267,
+      "learning_rate": 0.001,
+      "loss": 1.9646,
+      "step": 11723
+    },
+    {
+      "epoch": 0.4959810474659447,
+      "grad_norm": 0.15591877698898315,
+      "learning_rate": 0.001,
+      "loss": 1.7613,
+      "step": 11724
+    },
+    {
+      "epoch": 0.49602335222946103,
+      "grad_norm": 0.19704052805900574,
+      "learning_rate": 0.001,
+      "loss": 2.5827,
+      "step": 11725
+    },
+    {
+      "epoch": 0.4960656569929774,
+      "grad_norm": 0.14575530588626862,
+      "learning_rate": 0.001,
+      "loss": 1.9924,
+      "step": 11726
+    },
+    {
+      "epoch": 0.4961079617564938,
+      "grad_norm": 0.27594003081321716,
+      "learning_rate": 0.001,
+      "loss": 2.2579,
+      "step": 11727
+    },
+    {
+      "epoch": 0.49615026652001015,
+      "grad_norm": 0.4626797139644623,
+      "learning_rate": 0.001,
+      "loss": 1.6564,
+      "step": 11728
+    },
+    {
+      "epoch": 0.4961925712835265,
+      "grad_norm": 0.16353465616703033,
+      "learning_rate": 0.001,
+      "loss": 2.0006,
+      "step": 11729
+    },
+    {
+      "epoch": 0.4962348760470429,
+      "grad_norm": 0.12566418945789337,
+      "learning_rate": 0.001,
+      "loss": 1.9094,
+      "step": 11730
+    },
+    {
+      "epoch": 0.49627718081055927,
+      "grad_norm": 0.24055039882659912,
+      "learning_rate": 0.001,
+      "loss": 2.1373,
+      "step": 11731
+    },
+    {
+      "epoch": 0.4963194855740756,
+      "grad_norm": 0.5243616104125977,
+      "learning_rate": 0.001,
+      "loss": 1.875,
+      "step": 11732
+    },
+    {
+      "epoch": 0.49636179033759203,
+      "grad_norm": 0.2162822037935257,
+      "learning_rate": 0.001,
+      "loss": 1.5605,
+      "step": 11733
+    },
+    {
+      "epoch": 0.4964040951011084,
+      "grad_norm": 0.3283466398715973,
+      "learning_rate": 0.001,
+      "loss": 1.8846,
+      "step": 11734
+    },
+    {
+      "epoch": 0.49644639986462474,
+      "grad_norm": 0.20547711849212646,
+      "learning_rate": 0.001,
+      "loss": 2.197,
+      "step": 11735
+    },
+    {
+      "epoch": 0.49648870462814115,
+      "grad_norm": 0.26368358731269836,
+      "learning_rate": 0.001,
+      "loss": 3.4798,
+      "step": 11736
+    },
+    {
+      "epoch": 0.4965310093916575,
+      "grad_norm": 0.6923643350601196,
+      "learning_rate": 0.001,
+      "loss": 2.3685,
+      "step": 11737
+    },
+    {
+      "epoch": 0.49657331415517386,
+      "grad_norm": 0.21621890366077423,
+      "learning_rate": 0.001,
+      "loss": 2.4319,
+      "step": 11738
+    },
+    {
+      "epoch": 0.49661561891869027,
+      "grad_norm": 0.16339389979839325,
+      "learning_rate": 0.001,
+      "loss": 2.7695,
+      "step": 11739
+    },
+    {
+      "epoch": 0.4966579236822066,
+      "grad_norm": 0.16505463421344757,
+      "learning_rate": 0.001,
+      "loss": 2.034,
+      "step": 11740
+    },
+    {
+      "epoch": 0.496700228445723,
+      "grad_norm": 0.1787840873003006,
+      "learning_rate": 0.001,
+      "loss": 2.2629,
+      "step": 11741
+    },
+    {
+      "epoch": 0.49674253320923933,
+      "grad_norm": 0.34870707988739014,
+      "learning_rate": 0.001,
+      "loss": 2.0484,
+      "step": 11742
+    },
+    {
+      "epoch": 0.49678483797275574,
+      "grad_norm": 0.20475776493549347,
+      "learning_rate": 0.001,
+      "loss": 2.0176,
+      "step": 11743
+    },
+    {
+      "epoch": 0.4968271427362721,
+      "grad_norm": 3.398766040802002,
+      "learning_rate": 0.001,
+      "loss": 1.8543,
+      "step": 11744
+    },
+    {
+      "epoch": 0.49686944749978845,
+      "grad_norm": 0.15450245141983032,
+      "learning_rate": 0.001,
+      "loss": 2.6702,
+      "step": 11745
+    },
+    {
+      "epoch": 0.49691175226330486,
+      "grad_norm": 0.58338463306427,
+      "learning_rate": 0.001,
+      "loss": 3.001,
+      "step": 11746
+    },
+    {
+      "epoch": 0.4969540570268212,
+      "grad_norm": 0.1919245570898056,
+      "learning_rate": 0.001,
+      "loss": 2.2014,
+      "step": 11747
+    },
+    {
+      "epoch": 0.49699636179033757,
+      "grad_norm": 0.2006487101316452,
+      "learning_rate": 0.001,
+      "loss": 1.9795,
+      "step": 11748
+    },
+    {
+      "epoch": 0.497038666553854,
+      "grad_norm": 2.63030743598938,
+      "learning_rate": 0.001,
+      "loss": 2.3299,
+      "step": 11749
+    },
+    {
+      "epoch": 0.49708097131737033,
+      "grad_norm": 0.16151180863380432,
+      "learning_rate": 0.001,
+      "loss": 1.5329,
+      "step": 11750
+    },
+    {
+      "epoch": 0.4971232760808867,
+      "grad_norm": 0.2099246382713318,
+      "learning_rate": 0.001,
+      "loss": 1.9674,
+      "step": 11751
+    },
+    {
+      "epoch": 0.4971655808444031,
+      "grad_norm": 0.18836641311645508,
+      "learning_rate": 0.001,
+      "loss": 2.001,
+      "step": 11752
+    },
+    {
+      "epoch": 0.49720788560791945,
+      "grad_norm": 0.17593342065811157,
+      "learning_rate": 0.001,
+      "loss": 2.0003,
+      "step": 11753
+    },
+    {
+      "epoch": 0.4972501903714358,
+      "grad_norm": 0.21019577980041504,
+      "learning_rate": 0.001,
+      "loss": 3.3791,
+      "step": 11754
+    },
+    {
+      "epoch": 0.4972924951349522,
+      "grad_norm": 0.19833973050117493,
+      "learning_rate": 0.001,
+      "loss": 1.5016,
+      "step": 11755
+    },
+    {
+      "epoch": 0.49733479989846857,
+      "grad_norm": 0.1627396196126938,
+      "learning_rate": 0.001,
+      "loss": 1.9796,
+      "step": 11756
+    },
+    {
+      "epoch": 0.4973771046619849,
+      "grad_norm": 0.18876367807388306,
+      "learning_rate": 0.001,
+      "loss": 2.0286,
+      "step": 11757
+    },
+    {
+      "epoch": 0.49741940942550134,
+      "grad_norm": 0.2460365891456604,
+      "learning_rate": 0.001,
+      "loss": 2.237,
+      "step": 11758
+    },
+    {
+      "epoch": 0.4974617141890177,
+      "grad_norm": 0.34945163130760193,
+      "learning_rate": 0.001,
+      "loss": 1.7602,
+      "step": 11759
+    },
+    {
+      "epoch": 0.49750401895253404,
+      "grad_norm": 0.24172699451446533,
+      "learning_rate": 0.001,
+      "loss": 3.5869,
+      "step": 11760
+    },
+    {
+      "epoch": 0.49754632371605045,
+      "grad_norm": 0.19941669702529907,
+      "learning_rate": 0.001,
+      "loss": 1.7282,
+      "step": 11761
+    },
+    {
+      "epoch": 0.4975886284795668,
+      "grad_norm": 0.16542641818523407,
+      "learning_rate": 0.001,
+      "loss": 2.2762,
+      "step": 11762
+    },
+    {
+      "epoch": 0.49763093324308316,
+      "grad_norm": 0.18127164244651794,
+      "learning_rate": 0.001,
+      "loss": 1.7135,
+      "step": 11763
+    },
+    {
+      "epoch": 0.4976732380065995,
+      "grad_norm": 0.16587230563163757,
+      "learning_rate": 0.001,
+      "loss": 1.7269,
+      "step": 11764
+    },
+    {
+      "epoch": 0.4977155427701159,
+      "grad_norm": 0.15559205412864685,
+      "learning_rate": 0.001,
+      "loss": 2.0026,
+      "step": 11765
+    },
+    {
+      "epoch": 0.4977578475336323,
+      "grad_norm": 0.1646786332130432,
+      "learning_rate": 0.001,
+      "loss": 2.0569,
+      "step": 11766
+    },
+    {
+      "epoch": 0.49780015229714863,
+      "grad_norm": 0.1686537265777588,
+      "learning_rate": 0.001,
+      "loss": 2.0358,
+      "step": 11767
+    },
+    {
+      "epoch": 0.49784245706066504,
+      "grad_norm": 0.20397379994392395,
+      "learning_rate": 0.001,
+      "loss": 2.5713,
+      "step": 11768
+    },
+    {
+      "epoch": 0.4978847618241814,
+      "grad_norm": 0.2577773928642273,
+      "learning_rate": 0.001,
+      "loss": 1.9993,
+      "step": 11769
+    },
+    {
+      "epoch": 0.49792706658769775,
+      "grad_norm": 0.1761125773191452,
+      "learning_rate": 0.001,
+      "loss": 2.0988,
+      "step": 11770
+    },
+    {
+      "epoch": 0.49796937135121416,
+      "grad_norm": 0.3869670629501343,
+      "learning_rate": 0.001,
+      "loss": 2.7036,
+      "step": 11771
+    },
+    {
+      "epoch": 0.4980116761147305,
+      "grad_norm": 0.205720916390419,
+      "learning_rate": 0.001,
+      "loss": 2.7888,
+      "step": 11772
+    },
+    {
+      "epoch": 0.49805398087824687,
+      "grad_norm": 0.17144684493541718,
+      "learning_rate": 0.001,
+      "loss": 2.9138,
+      "step": 11773
+    },
+    {
+      "epoch": 0.4980962856417633,
+      "grad_norm": 0.4699496626853943,
+      "learning_rate": 0.001,
+      "loss": 2.1829,
+      "step": 11774
+    },
+    {
+      "epoch": 0.49813859040527964,
+      "grad_norm": 0.19264104962348938,
+      "learning_rate": 0.001,
+      "loss": 1.7184,
+      "step": 11775
+    },
+    {
+      "epoch": 0.498180895168796,
+      "grad_norm": 0.21069301664829254,
+      "learning_rate": 0.001,
+      "loss": 1.8648,
+      "step": 11776
+    },
+    {
+      "epoch": 0.4982231999323124,
+      "grad_norm": 0.17750690877437592,
+      "learning_rate": 0.001,
+      "loss": 2.6243,
+      "step": 11777
+    },
+    {
+      "epoch": 0.49826550469582875,
+      "grad_norm": 0.17451412975788116,
+      "learning_rate": 0.001,
+      "loss": 1.9341,
+      "step": 11778
+    },
+    {
+      "epoch": 0.4983078094593451,
+      "grad_norm": 0.18429610133171082,
+      "learning_rate": 0.001,
+      "loss": 1.8176,
+      "step": 11779
+    },
+    {
+      "epoch": 0.4983501142228615,
+      "grad_norm": 0.18034300208091736,
+      "learning_rate": 0.001,
+      "loss": 1.4398,
+      "step": 11780
+    },
+    {
+      "epoch": 0.49839241898637787,
+      "grad_norm": 0.2214241474866867,
+      "learning_rate": 0.001,
+      "loss": 2.8547,
+      "step": 11781
+    },
+    {
+      "epoch": 0.4984347237498942,
+      "grad_norm": 0.3767451047897339,
+      "learning_rate": 0.001,
+      "loss": 1.8602,
+      "step": 11782
+    },
+    {
+      "epoch": 0.49847702851341064,
+      "grad_norm": 0.16168661415576935,
+      "learning_rate": 0.001,
+      "loss": 1.378,
+      "step": 11783
+    },
+    {
+      "epoch": 0.498519333276927,
+      "grad_norm": 0.19141288101673126,
+      "learning_rate": 0.001,
+      "loss": 2.4181,
+      "step": 11784
+    },
+    {
+      "epoch": 0.49856163804044334,
+      "grad_norm": 0.17355014383792877,
+      "learning_rate": 0.001,
+      "loss": 1.8218,
+      "step": 11785
+    },
+    {
+      "epoch": 0.4986039428039597,
+      "grad_norm": 0.15057973563671112,
+      "learning_rate": 0.001,
+      "loss": 2.808,
+      "step": 11786
+    },
+    {
+      "epoch": 0.4986462475674761,
+      "grad_norm": 1.3171411752700806,
+      "learning_rate": 0.001,
+      "loss": 2.5091,
+      "step": 11787
+    },
+    {
+      "epoch": 0.49868855233099246,
+      "grad_norm": 0.20347653329372406,
+      "learning_rate": 0.001,
+      "loss": 2.2627,
+      "step": 11788
+    },
+    {
+      "epoch": 0.4987308570945088,
+      "grad_norm": 0.189688041806221,
+      "learning_rate": 0.001,
+      "loss": 2.2906,
+      "step": 11789
+    },
+    {
+      "epoch": 0.4987731618580252,
+      "grad_norm": 3.7641372680664062,
+      "learning_rate": 0.001,
+      "loss": 2.4856,
+      "step": 11790
+    },
+    {
+      "epoch": 0.4988154666215416,
+      "grad_norm": 0.2324923574924469,
+      "learning_rate": 0.001,
+      "loss": 3.4146,
+      "step": 11791
+    },
+    {
+      "epoch": 0.49885777138505794,
+      "grad_norm": 0.9900776147842407,
+      "learning_rate": 0.001,
+      "loss": 1.8225,
+      "step": 11792
+    },
+    {
+      "epoch": 0.49890007614857435,
+      "grad_norm": 0.4273008406162262,
+      "learning_rate": 0.001,
+      "loss": 2.9934,
+      "step": 11793
+    },
+    {
+      "epoch": 0.4989423809120907,
+      "grad_norm": 0.14736799895763397,
+      "learning_rate": 0.001,
+      "loss": 2.5134,
+      "step": 11794
+    },
+    {
+      "epoch": 0.49898468567560705,
+      "grad_norm": 2.7909343242645264,
+      "learning_rate": 0.001,
+      "loss": 1.9295,
+      "step": 11795
+    },
+    {
+      "epoch": 0.49902699043912346,
+      "grad_norm": 0.27986812591552734,
+      "learning_rate": 0.001,
+      "loss": 3.138,
+      "step": 11796
+    },
+    {
+      "epoch": 0.4990692952026398,
+      "grad_norm": 0.18478405475616455,
+      "learning_rate": 0.001,
+      "loss": 1.9531,
+      "step": 11797
+    },
+    {
+      "epoch": 0.49911159996615617,
+      "grad_norm": 0.18496938049793243,
+      "learning_rate": 0.001,
+      "loss": 2.3311,
+      "step": 11798
+    },
+    {
+      "epoch": 0.4991539047296726,
+      "grad_norm": 0.18009823560714722,
+      "learning_rate": 0.001,
+      "loss": 2.0114,
+      "step": 11799
+    },
+    {
+      "epoch": 0.49919620949318894,
+      "grad_norm": 0.23303596675395966,
+      "learning_rate": 0.001,
+      "loss": 1.9756,
+      "step": 11800
+    },
+    {
+      "epoch": 0.4992385142567053,
+      "grad_norm": 0.1994001865386963,
+      "learning_rate": 0.001,
+      "loss": 2.563,
+      "step": 11801
+    },
+    {
+      "epoch": 0.4992808190202217,
+      "grad_norm": 0.2099967747926712,
+      "learning_rate": 0.001,
+      "loss": 1.5605,
+      "step": 11802
+    },
+    {
+      "epoch": 0.49932312378373805,
+      "grad_norm": 0.34757867455482483,
+      "learning_rate": 0.001,
+      "loss": 3.3986,
+      "step": 11803
+    },
+    {
+      "epoch": 0.4993654285472544,
+      "grad_norm": 0.15620851516723633,
+      "learning_rate": 0.001,
+      "loss": 2.5944,
+      "step": 11804
+    },
+    {
+      "epoch": 0.4994077333107708,
+      "grad_norm": 0.2044965624809265,
+      "learning_rate": 0.001,
+      "loss": 3.5678,
+      "step": 11805
+    },
+    {
+      "epoch": 0.4994500380742872,
+      "grad_norm": 0.1616724729537964,
+      "learning_rate": 0.001,
+      "loss": 2.2451,
+      "step": 11806
+    },
+    {
+      "epoch": 0.4994923428378035,
+      "grad_norm": 0.21824952960014343,
+      "learning_rate": 0.001,
+      "loss": 3.1134,
+      "step": 11807
+    },
+    {
+      "epoch": 0.4995346476013199,
+      "grad_norm": 0.19981542229652405,
+      "learning_rate": 0.001,
+      "loss": 2.2409,
+      "step": 11808
+    },
+    {
+      "epoch": 0.4995769523648363,
+      "grad_norm": 0.538329541683197,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 11809
+    },
+    {
+      "epoch": 0.49961925712835265,
+      "grad_norm": 0.18705959618091583,
+      "learning_rate": 0.001,
+      "loss": 2.367,
+      "step": 11810
+    },
+    {
+      "epoch": 0.499661561891869,
+      "grad_norm": 1.9119819402694702,
+      "learning_rate": 0.001,
+      "loss": 1.9958,
+      "step": 11811
+    },
+    {
+      "epoch": 0.4997038666553854,
+      "grad_norm": 0.1916486769914627,
+      "learning_rate": 0.001,
+      "loss": 2.6751,
+      "step": 11812
+    },
+    {
+      "epoch": 0.49974617141890176,
+      "grad_norm": 0.17251789569854736,
+      "learning_rate": 0.001,
+      "loss": 1.6438,
+      "step": 11813
+    },
+    {
+      "epoch": 0.4997884761824181,
+      "grad_norm": 0.2549959421157837,
+      "learning_rate": 0.001,
+      "loss": 2.6528,
+      "step": 11814
+    },
+    {
+      "epoch": 0.4998307809459345,
+      "grad_norm": 0.16786809265613556,
+      "learning_rate": 0.001,
+      "loss": 1.7551,
+      "step": 11815
+    },
+    {
+      "epoch": 0.4998730857094509,
+      "grad_norm": 0.22102710604667664,
+      "learning_rate": 0.001,
+      "loss": 1.6668,
+      "step": 11816
+    },
+    {
+      "epoch": 0.49991539047296724,
+      "grad_norm": 0.2043444961309433,
+      "learning_rate": 0.001,
+      "loss": 2.0156,
+      "step": 11817
+    },
+    {
+      "epoch": 0.49995769523648365,
+      "grad_norm": 0.1693449169397354,
+      "learning_rate": 0.001,
+      "loss": 1.7542,
+      "step": 11818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18922848999500275,
+      "learning_rate": 0.001,
+      "loss": 3.11,
+      "step": 11819
+    },
+    {
+      "epoch": 0.5000423047635164,
+      "grad_norm": 0.19555483758449554,
+      "learning_rate": 0.001,
+      "loss": 3.5945,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5000846095270327,
+      "grad_norm": 0.22335830330848694,
+      "learning_rate": 0.001,
+      "loss": 3.515,
+      "step": 11821
+    },
+    {
+      "epoch": 0.5001269142905491,
+      "grad_norm": 0.17894285917282104,
+      "learning_rate": 0.001,
+      "loss": 1.6961,
+      "step": 11822
+    },
+    {
+      "epoch": 0.5001692190540655,
+      "grad_norm": 0.16279856860637665,
+      "learning_rate": 0.001,
+      "loss": 2.7065,
+      "step": 11823
+    },
+    {
+      "epoch": 0.5002115238175818,
+      "grad_norm": 0.2117103487253189,
+      "learning_rate": 0.001,
+      "loss": 2.481,
+      "step": 11824
+    },
+    {
+      "epoch": 0.5002538285810982,
+      "grad_norm": 0.1833728402853012,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 11825
+    },
+    {
+      "epoch": 0.5002961333446146,
+      "grad_norm": 0.1608177125453949,
+      "learning_rate": 0.001,
+      "loss": 1.8915,
+      "step": 11826
+    },
+    {
+      "epoch": 0.500338438108131,
+      "grad_norm": 0.15755756199359894,
+      "learning_rate": 0.001,
+      "loss": 2.0134,
+      "step": 11827
+    },
+    {
+      "epoch": 0.5003807428716474,
+      "grad_norm": 1.6536004543304443,
+      "learning_rate": 0.001,
+      "loss": 2.4632,
+      "step": 11828
+    },
+    {
+      "epoch": 0.5004230476351638,
+      "grad_norm": 0.17885924875736237,
+      "learning_rate": 0.001,
+      "loss": 2.8906,
+      "step": 11829
+    },
+    {
+      "epoch": 0.5004653523986801,
+      "grad_norm": 0.17914816737174988,
+      "learning_rate": 0.001,
+      "loss": 1.82,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5005076571621965,
+      "grad_norm": 0.4052477478981018,
+      "learning_rate": 0.001,
+      "loss": 1.7167,
+      "step": 11831
+    },
+    {
+      "epoch": 0.5005499619257129,
+      "grad_norm": 0.18463678658008575,
+      "learning_rate": 0.001,
+      "loss": 2.2511,
+      "step": 11832
+    },
+    {
+      "epoch": 0.5005922666892292,
+      "grad_norm": 0.194928377866745,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 11833
+    },
+    {
+      "epoch": 0.5006345714527456,
+      "grad_norm": 0.17245490849018097,
+      "learning_rate": 0.001,
+      "loss": 1.9821,
+      "step": 11834
+    },
+    {
+      "epoch": 0.500676876216262,
+      "grad_norm": 0.1762954294681549,
+      "learning_rate": 0.001,
+      "loss": 1.6249,
+      "step": 11835
+    },
+    {
+      "epoch": 0.5007191809797783,
+      "grad_norm": 0.16809524595737457,
+      "learning_rate": 0.001,
+      "loss": 1.6515,
+      "step": 11836
+    },
+    {
+      "epoch": 0.5007614857432947,
+      "grad_norm": 0.2633932828903198,
+      "learning_rate": 0.001,
+      "loss": 2.8323,
+      "step": 11837
+    },
+    {
+      "epoch": 0.5008037905068111,
+      "grad_norm": 1.2055695056915283,
+      "learning_rate": 0.001,
+      "loss": 2.7611,
+      "step": 11838
+    },
+    {
+      "epoch": 0.5008460952703274,
+      "grad_norm": 0.5635671019554138,
+      "learning_rate": 0.001,
+      "loss": 2.3381,
+      "step": 11839
+    },
+    {
+      "epoch": 0.5008884000338438,
+      "grad_norm": 0.22210033237934113,
+      "learning_rate": 0.001,
+      "loss": 2.4092,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5009307047973602,
+      "grad_norm": 0.3583562672138214,
+      "learning_rate": 0.001,
+      "loss": 2.3941,
+      "step": 11841
+    },
+    {
+      "epoch": 0.5009730095608765,
+      "grad_norm": 1.5744633674621582,
+      "learning_rate": 0.001,
+      "loss": 2.1961,
+      "step": 11842
+    },
+    {
+      "epoch": 0.501015314324393,
+      "grad_norm": 0.19284333288669586,
+      "learning_rate": 0.001,
+      "loss": 2.6481,
+      "step": 11843
+    },
+    {
+      "epoch": 0.5010576190879092,
+      "grad_norm": 0.16375435888767242,
+      "learning_rate": 0.001,
+      "loss": 3.0202,
+      "step": 11844
+    },
+    {
+      "epoch": 0.5010999238514257,
+      "grad_norm": 0.1810348629951477,
+      "learning_rate": 0.001,
+      "loss": 2.0206,
+      "step": 11845
+    },
+    {
+      "epoch": 0.5011422286149421,
+      "grad_norm": 0.15742111206054688,
+      "learning_rate": 0.001,
+      "loss": 2.0867,
+      "step": 11846
+    },
+    {
+      "epoch": 0.5011845333784584,
+      "grad_norm": 0.18012715876102448,
+      "learning_rate": 0.001,
+      "loss": 1.8844,
+      "step": 11847
+    },
+    {
+      "epoch": 0.5012268381419748,
+      "grad_norm": 0.21920199692249298,
+      "learning_rate": 0.001,
+      "loss": 2.0386,
+      "step": 11848
+    },
+    {
+      "epoch": 0.5012691429054912,
+      "grad_norm": 0.20530590415000916,
+      "learning_rate": 0.001,
+      "loss": 2.7012,
+      "step": 11849
+    },
+    {
+      "epoch": 0.5013114476690075,
+      "grad_norm": 0.1777116060256958,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5013537524325239,
+      "grad_norm": 0.9319034814834595,
+      "learning_rate": 0.001,
+      "loss": 2.1148,
+      "step": 11851
+    },
+    {
+      "epoch": 0.5013960571960403,
+      "grad_norm": 0.22064313292503357,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 11852
+    },
+    {
+      "epoch": 0.5014383619595566,
+      "grad_norm": 0.19049854576587677,
+      "learning_rate": 0.001,
+      "loss": 2.2153,
+      "step": 11853
+    },
+    {
+      "epoch": 0.501480666723073,
+      "grad_norm": 0.1864825040102005,
+      "learning_rate": 0.001,
+      "loss": 2.225,
+      "step": 11854
+    },
+    {
+      "epoch": 0.5015229714865894,
+      "grad_norm": 0.20713578164577484,
+      "learning_rate": 0.001,
+      "loss": 2.5584,
+      "step": 11855
+    },
+    {
+      "epoch": 0.5015652762501057,
+      "grad_norm": 0.2419637143611908,
+      "learning_rate": 0.001,
+      "loss": 1.9168,
+      "step": 11856
+    },
+    {
+      "epoch": 0.5016075810136221,
+      "grad_norm": 0.18344347178936005,
+      "learning_rate": 0.001,
+      "loss": 1.6852,
+      "step": 11857
+    },
+    {
+      "epoch": 0.5016498857771385,
+      "grad_norm": 3.606316328048706,
+      "learning_rate": 0.001,
+      "loss": 1.7524,
+      "step": 11858
+    },
+    {
+      "epoch": 0.5016921905406548,
+      "grad_norm": 0.1817534863948822,
+      "learning_rate": 0.001,
+      "loss": 1.6807,
+      "step": 11859
+    },
+    {
+      "epoch": 0.5017344953041712,
+      "grad_norm": 0.22895671427249908,
+      "learning_rate": 0.001,
+      "loss": 2.1781,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5017768000676877,
+      "grad_norm": 0.1826825588941574,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 11861
+    },
+    {
+      "epoch": 0.501819104831204,
+      "grad_norm": 0.1994072049856186,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 11862
+    },
+    {
+      "epoch": 0.5018614095947204,
+      "grad_norm": 0.1609574556350708,
+      "learning_rate": 0.001,
+      "loss": 2.5509,
+      "step": 11863
+    },
+    {
+      "epoch": 0.5019037143582368,
+      "grad_norm": 0.28358885645866394,
+      "learning_rate": 0.001,
+      "loss": 2.2666,
+      "step": 11864
+    },
+    {
+      "epoch": 0.5019460191217531,
+      "grad_norm": 0.19718047976493835,
+      "learning_rate": 0.001,
+      "loss": 2.8378,
+      "step": 11865
+    },
+    {
+      "epoch": 0.5019883238852695,
+      "grad_norm": 0.17806696891784668,
+      "learning_rate": 0.001,
+      "loss": 2.5277,
+      "step": 11866
+    },
+    {
+      "epoch": 0.5020306286487859,
+      "grad_norm": 0.22065989673137665,
+      "learning_rate": 0.001,
+      "loss": 3.4721,
+      "step": 11867
+    },
+    {
+      "epoch": 0.5020729334123022,
+      "grad_norm": 1.0695648193359375,
+      "learning_rate": 0.001,
+      "loss": 2.6934,
+      "step": 11868
+    },
+    {
+      "epoch": 0.5021152381758186,
+      "grad_norm": 0.34497925639152527,
+      "learning_rate": 0.001,
+      "loss": 2.5293,
+      "step": 11869
+    },
+    {
+      "epoch": 0.502157542939335,
+      "grad_norm": 0.2196062207221985,
+      "learning_rate": 0.001,
+      "loss": 1.7822,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5021998477028513,
+      "grad_norm": 1.1649377346038818,
+      "learning_rate": 0.001,
+      "loss": 2.6756,
+      "step": 11871
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.18098509311676025,
+      "learning_rate": 0.001,
+      "loss": 2.4486,
+      "step": 11872
+    },
+    {
+      "epoch": 0.5022844572298841,
+      "grad_norm": 0.21399042010307312,
+      "learning_rate": 0.001,
+      "loss": 2.5243,
+      "step": 11873
+    },
+    {
+      "epoch": 0.5023267619934004,
+      "grad_norm": 0.19336672127246857,
+      "learning_rate": 0.001,
+      "loss": 1.3858,
+      "step": 11874
+    },
+    {
+      "epoch": 0.5023690667569168,
+      "grad_norm": 1.1209372282028198,
+      "learning_rate": 0.001,
+      "loss": 2.7417,
+      "step": 11875
+    },
+    {
+      "epoch": 0.5024113715204332,
+      "grad_norm": 0.1669924110174179,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 11876
+    },
+    {
+      "epoch": 0.5024536762839495,
+      "grad_norm": 0.20557548105716705,
+      "learning_rate": 0.001,
+      "loss": 1.9532,
+      "step": 11877
+    },
+    {
+      "epoch": 0.502495981047466,
+      "grad_norm": 1.1194974184036255,
+      "learning_rate": 0.001,
+      "loss": 3.5188,
+      "step": 11878
+    },
+    {
+      "epoch": 0.5025382858109824,
+      "grad_norm": 1.5623433589935303,
+      "learning_rate": 0.001,
+      "loss": 2.5858,
+      "step": 11879
+    },
+    {
+      "epoch": 0.5025805905744987,
+      "grad_norm": 0.32952699065208435,
+      "learning_rate": 0.001,
+      "loss": 2.9636,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5026228953380151,
+      "grad_norm": 0.18571840226650238,
+      "learning_rate": 0.001,
+      "loss": 1.854,
+      "step": 11881
+    },
+    {
+      "epoch": 0.5026652001015315,
+      "grad_norm": 6.811452865600586,
+      "learning_rate": 0.001,
+      "loss": 1.903,
+      "step": 11882
+    },
+    {
+      "epoch": 0.5027075048650478,
+      "grad_norm": 0.1865660697221756,
+      "learning_rate": 0.001,
+      "loss": 2.0409,
+      "step": 11883
+    },
+    {
+      "epoch": 0.5027498096285642,
+      "grad_norm": 0.17085236310958862,
+      "learning_rate": 0.001,
+      "loss": 2.2718,
+      "step": 11884
+    },
+    {
+      "epoch": 0.5027921143920806,
+      "grad_norm": 0.19499720633029938,
+      "learning_rate": 0.001,
+      "loss": 2.1248,
+      "step": 11885
+    },
+    {
+      "epoch": 0.5028344191555969,
+      "grad_norm": 0.1698610633611679,
+      "learning_rate": 0.001,
+      "loss": 1.8143,
+      "step": 11886
+    },
+    {
+      "epoch": 0.5028767239191133,
+      "grad_norm": 0.18589156866073608,
+      "learning_rate": 0.001,
+      "loss": 2.1618,
+      "step": 11887
+    },
+    {
+      "epoch": 0.5029190286826296,
+      "grad_norm": 0.17679286003112793,
+      "learning_rate": 0.001,
+      "loss": 2.1766,
+      "step": 11888
+    },
+    {
+      "epoch": 0.502961333446146,
+      "grad_norm": 0.18507613241672516,
+      "learning_rate": 0.001,
+      "loss": 1.8563,
+      "step": 11889
+    },
+    {
+      "epoch": 0.5030036382096624,
+      "grad_norm": 0.1703113168478012,
+      "learning_rate": 0.001,
+      "loss": 1.8142,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5030459429731787,
+      "grad_norm": 0.1954764574766159,
+      "learning_rate": 0.001,
+      "loss": 2.0398,
+      "step": 11891
+    },
+    {
+      "epoch": 0.5030882477366951,
+      "grad_norm": 0.1964946985244751,
+      "learning_rate": 0.001,
+      "loss": 2.9447,
+      "step": 11892
+    },
+    {
+      "epoch": 0.5031305525002115,
+      "grad_norm": 0.23403804004192352,
+      "learning_rate": 0.001,
+      "loss": 2.4022,
+      "step": 11893
+    },
+    {
+      "epoch": 0.5031728572637278,
+      "grad_norm": 0.20903852581977844,
+      "learning_rate": 0.001,
+      "loss": 2.6347,
+      "step": 11894
+    },
+    {
+      "epoch": 0.5032151620272443,
+      "grad_norm": 0.14716748893260956,
+      "learning_rate": 0.001,
+      "loss": 2.6717,
+      "step": 11895
+    },
+    {
+      "epoch": 0.5032574667907607,
+      "grad_norm": 0.1724325567483902,
+      "learning_rate": 0.001,
+      "loss": 2.4629,
+      "step": 11896
+    },
+    {
+      "epoch": 0.503299771554277,
+      "grad_norm": 2.00115704536438,
+      "learning_rate": 0.001,
+      "loss": 2.8087,
+      "step": 11897
+    },
+    {
+      "epoch": 0.5033420763177934,
+      "grad_norm": 0.3015700876712799,
+      "learning_rate": 0.001,
+      "loss": 1.8208,
+      "step": 11898
+    },
+    {
+      "epoch": 0.5033843810813098,
+      "grad_norm": 0.18735451996326447,
+      "learning_rate": 0.001,
+      "loss": 1.9337,
+      "step": 11899
+    },
+    {
+      "epoch": 0.5034266858448261,
+      "grad_norm": 0.18726201355457306,
+      "learning_rate": 0.001,
+      "loss": 2.8956,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5034689906083425,
+      "grad_norm": 0.22728313505649567,
+      "learning_rate": 0.001,
+      "loss": 2.3571,
+      "step": 11901
+    },
+    {
+      "epoch": 0.5035112953718589,
+      "grad_norm": 0.3538459241390228,
+      "learning_rate": 0.001,
+      "loss": 2.0404,
+      "step": 11902
+    },
+    {
+      "epoch": 0.5035536001353752,
+      "grad_norm": 0.2191903442144394,
+      "learning_rate": 0.001,
+      "loss": 2.8035,
+      "step": 11903
+    },
+    {
+      "epoch": 0.5035959048988916,
+      "grad_norm": 1.2303228378295898,
+      "learning_rate": 0.001,
+      "loss": 2.9012,
+      "step": 11904
+    },
+    {
+      "epoch": 0.503638209662408,
+      "grad_norm": 0.6738932132720947,
+      "learning_rate": 0.001,
+      "loss": 2.0589,
+      "step": 11905
+    },
+    {
+      "epoch": 0.5036805144259243,
+      "grad_norm": 0.18187807500362396,
+      "learning_rate": 0.001,
+      "loss": 2.1458,
+      "step": 11906
+    },
+    {
+      "epoch": 0.5037228191894407,
+      "grad_norm": 0.20437604188919067,
+      "learning_rate": 0.001,
+      "loss": 2.5876,
+      "step": 11907
+    },
+    {
+      "epoch": 0.5037651239529571,
+      "grad_norm": 0.17892666161060333,
+      "learning_rate": 0.001,
+      "loss": 2.4489,
+      "step": 11908
+    },
+    {
+      "epoch": 0.5038074287164734,
+      "grad_norm": 0.1988033950328827,
+      "learning_rate": 0.001,
+      "loss": 2.935,
+      "step": 11909
+    },
+    {
+      "epoch": 0.5038497334799898,
+      "grad_norm": 0.16450054943561554,
+      "learning_rate": 0.001,
+      "loss": 1.6965,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5038920382435063,
+      "grad_norm": 0.1781654804944992,
+      "learning_rate": 0.001,
+      "loss": 2.5796,
+      "step": 11911
+    },
+    {
+      "epoch": 0.5039343430070226,
+      "grad_norm": 0.17082349956035614,
+      "learning_rate": 0.001,
+      "loss": 2.9056,
+      "step": 11912
+    },
+    {
+      "epoch": 0.503976647770539,
+      "grad_norm": 0.23962025344371796,
+      "learning_rate": 0.001,
+      "loss": 2.1747,
+      "step": 11913
+    },
+    {
+      "epoch": 0.5040189525340554,
+      "grad_norm": 0.18529817461967468,
+      "learning_rate": 0.001,
+      "loss": 2.4123,
+      "step": 11914
+    },
+    {
+      "epoch": 0.5040612572975717,
+      "grad_norm": 0.3135690689086914,
+      "learning_rate": 0.001,
+      "loss": 1.7927,
+      "step": 11915
+    },
+    {
+      "epoch": 0.5041035620610881,
+      "grad_norm": 0.7973864674568176,
+      "learning_rate": 0.001,
+      "loss": 1.7801,
+      "step": 11916
+    },
+    {
+      "epoch": 0.5041458668246045,
+      "grad_norm": 0.25282907485961914,
+      "learning_rate": 0.001,
+      "loss": 2.5156,
+      "step": 11917
+    },
+    {
+      "epoch": 0.5041881715881208,
+      "grad_norm": 0.8695905804634094,
+      "learning_rate": 0.001,
+      "loss": 2.0658,
+      "step": 11918
+    },
+    {
+      "epoch": 0.5042304763516372,
+      "grad_norm": 0.1739739030599594,
+      "learning_rate": 0.001,
+      "loss": 1.7482,
+      "step": 11919
+    },
+    {
+      "epoch": 0.5042727811151536,
+      "grad_norm": 0.639236330986023,
+      "learning_rate": 0.001,
+      "loss": 1.906,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5043150858786699,
+      "grad_norm": 0.2065761834383011,
+      "learning_rate": 0.001,
+      "loss": 1.6717,
+      "step": 11921
+    },
+    {
+      "epoch": 0.5043573906421863,
+      "grad_norm": 0.22129188477993011,
+      "learning_rate": 0.001,
+      "loss": 2.7184,
+      "step": 11922
+    },
+    {
+      "epoch": 0.5043996954057027,
+      "grad_norm": 0.16457290947437286,
+      "learning_rate": 0.001,
+      "loss": 1.9362,
+      "step": 11923
+    },
+    {
+      "epoch": 0.504442000169219,
+      "grad_norm": 0.7113840579986572,
+      "learning_rate": 0.001,
+      "loss": 1.3661,
+      "step": 11924
+    },
+    {
+      "epoch": 0.5044843049327354,
+      "grad_norm": 0.16901074349880219,
+      "learning_rate": 0.001,
+      "loss": 2.3126,
+      "step": 11925
+    },
+    {
+      "epoch": 0.5045266096962518,
+      "grad_norm": 0.2004851996898651,
+      "learning_rate": 0.001,
+      "loss": 2.6944,
+      "step": 11926
+    },
+    {
+      "epoch": 0.5045689144597681,
+      "grad_norm": 0.17636452615261078,
+      "learning_rate": 0.001,
+      "loss": 2.3345,
+      "step": 11927
+    },
+    {
+      "epoch": 0.5046112192232846,
+      "grad_norm": 0.16024649143218994,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 11928
+    },
+    {
+      "epoch": 0.504653523986801,
+      "grad_norm": 0.18146340548992157,
+      "learning_rate": 0.001,
+      "loss": 2.6534,
+      "step": 11929
+    },
+    {
+      "epoch": 0.5046958287503173,
+      "grad_norm": 0.19489388167858124,
+      "learning_rate": 0.001,
+      "loss": 2.5024,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5047381335138337,
+      "grad_norm": 0.5140373706817627,
+      "learning_rate": 0.001,
+      "loss": 3.1386,
+      "step": 11931
+    },
+    {
+      "epoch": 0.50478043827735,
+      "grad_norm": 1.8529261350631714,
+      "learning_rate": 0.001,
+      "loss": 2.1581,
+      "step": 11932
+    },
+    {
+      "epoch": 0.5048227430408664,
+      "grad_norm": 0.16722053289413452,
+      "learning_rate": 0.001,
+      "loss": 2.019,
+      "step": 11933
+    },
+    {
+      "epoch": 0.5048650478043828,
+      "grad_norm": 0.16939833760261536,
+      "learning_rate": 0.001,
+      "loss": 1.843,
+      "step": 11934
+    },
+    {
+      "epoch": 0.5049073525678991,
+      "grad_norm": 0.3722081780433655,
+      "learning_rate": 0.001,
+      "loss": 2.8935,
+      "step": 11935
+    },
+    {
+      "epoch": 0.5049496573314155,
+      "grad_norm": 0.15549683570861816,
+      "learning_rate": 0.001,
+      "loss": 1.5844,
+      "step": 11936
+    },
+    {
+      "epoch": 0.5049919620949319,
+      "grad_norm": 0.19663242995738983,
+      "learning_rate": 0.001,
+      "loss": 2.0955,
+      "step": 11937
+    },
+    {
+      "epoch": 0.5050342668584482,
+      "grad_norm": 0.557640016078949,
+      "learning_rate": 0.001,
+      "loss": 3.2234,
+      "step": 11938
+    },
+    {
+      "epoch": 0.5050765716219646,
+      "grad_norm": 0.24764712154865265,
+      "learning_rate": 0.001,
+      "loss": 2.313,
+      "step": 11939
+    },
+    {
+      "epoch": 0.505118876385481,
+      "grad_norm": 0.22433580458164215,
+      "learning_rate": 0.001,
+      "loss": 2.8558,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5051611811489973,
+      "grad_norm": 0.29622960090637207,
+      "learning_rate": 0.001,
+      "loss": 2.7506,
+      "step": 11941
+    },
+    {
+      "epoch": 0.5052034859125137,
+      "grad_norm": 0.23077332973480225,
+      "learning_rate": 0.001,
+      "loss": 1.8445,
+      "step": 11942
+    },
+    {
+      "epoch": 0.5052457906760301,
+      "grad_norm": 0.16419456899166107,
+      "learning_rate": 0.001,
+      "loss": 2.2429,
+      "step": 11943
+    },
+    {
+      "epoch": 0.5052880954395464,
+      "grad_norm": 1.3198111057281494,
+      "learning_rate": 0.001,
+      "loss": 2.769,
+      "step": 11944
+    },
+    {
+      "epoch": 0.5053304002030629,
+      "grad_norm": 0.32443439960479736,
+      "learning_rate": 0.001,
+      "loss": 2.6581,
+      "step": 11945
+    },
+    {
+      "epoch": 0.5053727049665793,
+      "grad_norm": 0.34906241297721863,
+      "learning_rate": 0.001,
+      "loss": 2.769,
+      "step": 11946
+    },
+    {
+      "epoch": 0.5054150097300956,
+      "grad_norm": 0.1365213692188263,
+      "learning_rate": 0.001,
+      "loss": 2.729,
+      "step": 11947
+    },
+    {
+      "epoch": 0.505457314493612,
+      "grad_norm": 0.3047373294830322,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 11948
+    },
+    {
+      "epoch": 0.5054996192571284,
+      "grad_norm": 0.4273305833339691,
+      "learning_rate": 0.001,
+      "loss": 2.3561,
+      "step": 11949
+    },
+    {
+      "epoch": 0.5055419240206447,
+      "grad_norm": 0.22199836373329163,
+      "learning_rate": 0.001,
+      "loss": 2.0293,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5055842287841611,
+      "grad_norm": 0.3782862424850464,
+      "learning_rate": 0.001,
+      "loss": 1.7579,
+      "step": 11951
+    },
+    {
+      "epoch": 0.5056265335476775,
+      "grad_norm": 0.19411024451255798,
+      "learning_rate": 0.001,
+      "loss": 1.4811,
+      "step": 11952
+    },
+    {
+      "epoch": 0.5056688383111938,
+      "grad_norm": 0.1856779307126999,
+      "learning_rate": 0.001,
+      "loss": 2.2312,
+      "step": 11953
+    },
+    {
+      "epoch": 0.5057111430747102,
+      "grad_norm": 0.17408660054206848,
+      "learning_rate": 0.001,
+      "loss": 2.3105,
+      "step": 11954
+    },
+    {
+      "epoch": 0.5057534478382266,
+      "grad_norm": 0.19032998383045197,
+      "learning_rate": 0.001,
+      "loss": 2.088,
+      "step": 11955
+    },
+    {
+      "epoch": 0.5057957526017429,
+      "grad_norm": 0.18069171905517578,
+      "learning_rate": 0.001,
+      "loss": 2.6122,
+      "step": 11956
+    },
+    {
+      "epoch": 0.5058380573652593,
+      "grad_norm": 0.16551326215267181,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 11957
+    },
+    {
+      "epoch": 0.5058803621287757,
+      "grad_norm": 2.8570401668548584,
+      "learning_rate": 0.001,
+      "loss": 1.9708,
+      "step": 11958
+    },
+    {
+      "epoch": 0.505922666892292,
+      "grad_norm": 0.20645208656787872,
+      "learning_rate": 0.001,
+      "loss": 2.3922,
+      "step": 11959
+    },
+    {
+      "epoch": 0.5059649716558084,
+      "grad_norm": 4.1343817710876465,
+      "learning_rate": 0.001,
+      "loss": 2.1069,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5060072764193249,
+      "grad_norm": 0.17171384394168854,
+      "learning_rate": 0.001,
+      "loss": 1.5699,
+      "step": 11961
+    },
+    {
+      "epoch": 0.5060495811828412,
+      "grad_norm": 0.4710404574871063,
+      "learning_rate": 0.001,
+      "loss": 1.9224,
+      "step": 11962
+    },
+    {
+      "epoch": 0.5060918859463576,
+      "grad_norm": 0.9586241245269775,
+      "learning_rate": 0.001,
+      "loss": 2.1856,
+      "step": 11963
+    },
+    {
+      "epoch": 0.506134190709874,
+      "grad_norm": 0.1798468381166458,
+      "learning_rate": 0.001,
+      "loss": 2.836,
+      "step": 11964
+    },
+    {
+      "epoch": 0.5061764954733903,
+      "grad_norm": 0.2874303162097931,
+      "learning_rate": 0.001,
+      "loss": 2.2059,
+      "step": 11965
+    },
+    {
+      "epoch": 0.5062188002369067,
+      "grad_norm": 0.1568385362625122,
+      "learning_rate": 0.001,
+      "loss": 1.7764,
+      "step": 11966
+    },
+    {
+      "epoch": 0.5062611050004231,
+      "grad_norm": 0.17159751057624817,
+      "learning_rate": 0.001,
+      "loss": 2.1703,
+      "step": 11967
+    },
+    {
+      "epoch": 0.5063034097639394,
+      "grad_norm": 0.20359356701374054,
+      "learning_rate": 0.001,
+      "loss": 2.4419,
+      "step": 11968
+    },
+    {
+      "epoch": 0.5063457145274558,
+      "grad_norm": 0.1408676654100418,
+      "learning_rate": 0.001,
+      "loss": 2.12,
+      "step": 11969
+    },
+    {
+      "epoch": 0.5063880192909722,
+      "grad_norm": 0.16593052446842194,
+      "learning_rate": 0.001,
+      "loss": 1.9553,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5064303240544885,
+      "grad_norm": 0.12770260870456696,
+      "learning_rate": 0.001,
+      "loss": 2.1708,
+      "step": 11971
+    },
+    {
+      "epoch": 0.5064726288180049,
+      "grad_norm": 0.15011616051197052,
+      "learning_rate": 0.001,
+      "loss": 2.1359,
+      "step": 11972
+    },
+    {
+      "epoch": 0.5065149335815213,
+      "grad_norm": 0.47157904505729675,
+      "learning_rate": 0.001,
+      "loss": 1.9827,
+      "step": 11973
+    },
+    {
+      "epoch": 0.5065572383450376,
+      "grad_norm": 0.15799900889396667,
+      "learning_rate": 0.001,
+      "loss": 2.2315,
+      "step": 11974
+    },
+    {
+      "epoch": 0.506599543108554,
+      "grad_norm": 0.20162101089954376,
+      "learning_rate": 0.001,
+      "loss": 2.2226,
+      "step": 11975
+    },
+    {
+      "epoch": 0.5066418478720704,
+      "grad_norm": 0.21195806562900543,
+      "learning_rate": 0.001,
+      "loss": 2.3964,
+      "step": 11976
+    },
+    {
+      "epoch": 0.5066841526355867,
+      "grad_norm": 0.18306732177734375,
+      "learning_rate": 0.001,
+      "loss": 3.3474,
+      "step": 11977
+    },
+    {
+      "epoch": 0.5067264573991032,
+      "grad_norm": 0.15694935619831085,
+      "learning_rate": 0.001,
+      "loss": 2.5094,
+      "step": 11978
+    },
+    {
+      "epoch": 0.5067687621626195,
+      "grad_norm": 0.22060388326644897,
+      "learning_rate": 0.001,
+      "loss": 2.6429,
+      "step": 11979
+    },
+    {
+      "epoch": 0.5068110669261359,
+      "grad_norm": 0.2225179672241211,
+      "learning_rate": 0.001,
+      "loss": 1.6857,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5068533716896523,
+      "grad_norm": 0.19372129440307617,
+      "learning_rate": 0.001,
+      "loss": 2.5101,
+      "step": 11981
+    },
+    {
+      "epoch": 0.5068956764531686,
+      "grad_norm": 0.13902443647384644,
+      "learning_rate": 0.001,
+      "loss": 2.4499,
+      "step": 11982
+    },
+    {
+      "epoch": 0.506937981216685,
+      "grad_norm": 0.15087106823921204,
+      "learning_rate": 0.001,
+      "loss": 2.1793,
+      "step": 11983
+    },
+    {
+      "epoch": 0.5069802859802014,
+      "grad_norm": 0.16260752081871033,
+      "learning_rate": 0.001,
+      "loss": 2.0164,
+      "step": 11984
+    },
+    {
+      "epoch": 0.5070225907437177,
+      "grad_norm": 0.159589946269989,
+      "learning_rate": 0.001,
+      "loss": 2.8375,
+      "step": 11985
+    },
+    {
+      "epoch": 0.5070648955072341,
+      "grad_norm": 0.21032920479774475,
+      "learning_rate": 0.001,
+      "loss": 4.2796,
+      "step": 11986
+    },
+    {
+      "epoch": 0.5071072002707505,
+      "grad_norm": 0.19876354932785034,
+      "learning_rate": 0.001,
+      "loss": 2.9237,
+      "step": 11987
+    },
+    {
+      "epoch": 0.5071495050342668,
+      "grad_norm": 0.35270553827285767,
+      "learning_rate": 0.001,
+      "loss": 2.9928,
+      "step": 11988
+    },
+    {
+      "epoch": 0.5071918097977832,
+      "grad_norm": 0.16005408763885498,
+      "learning_rate": 0.001,
+      "loss": 2.1871,
+      "step": 11989
+    },
+    {
+      "epoch": 0.5072341145612996,
+      "grad_norm": 0.13983991742134094,
+      "learning_rate": 0.001,
+      "loss": 2.6865,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5072764193248159,
+      "grad_norm": 0.17701087892055511,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 11991
+    },
+    {
+      "epoch": 0.5073187240883323,
+      "grad_norm": 0.15667004883289337,
+      "learning_rate": 0.001,
+      "loss": 1.466,
+      "step": 11992
+    },
+    {
+      "epoch": 0.5073610288518487,
+      "grad_norm": 0.22075319290161133,
+      "learning_rate": 0.001,
+      "loss": 2.0947,
+      "step": 11993
+    },
+    {
+      "epoch": 0.507403333615365,
+      "grad_norm": 0.14011307060718536,
+      "learning_rate": 0.001,
+      "loss": 2.5497,
+      "step": 11994
+    },
+    {
+      "epoch": 0.5074456383788815,
+      "grad_norm": 0.1682954877614975,
+      "learning_rate": 0.001,
+      "loss": 1.8733,
+      "step": 11995
+    },
+    {
+      "epoch": 0.5074879431423979,
+      "grad_norm": 30.817886352539062,
+      "learning_rate": 0.001,
+      "loss": 2.5975,
+      "step": 11996
+    },
+    {
+      "epoch": 0.5075302479059142,
+      "grad_norm": 0.17083485424518585,
+      "learning_rate": 0.001,
+      "loss": 2.1617,
+      "step": 11997
+    },
+    {
+      "epoch": 0.5075725526694306,
+      "grad_norm": 0.23582878708839417,
+      "learning_rate": 0.001,
+      "loss": 2.6255,
+      "step": 11998
+    },
+    {
+      "epoch": 0.507614857432947,
+      "grad_norm": 0.16995792090892792,
+      "learning_rate": 0.001,
+      "loss": 2.413,
+      "step": 11999
+    },
+    {
+      "epoch": 0.5076571621964633,
+      "grad_norm": 0.1696588695049286,
+      "learning_rate": 0.001,
+      "loss": 2.9008,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5076994669599797,
+      "grad_norm": 1.6405519247055054,
+      "learning_rate": 0.001,
+      "loss": 1.7866,
+      "step": 12001
+    },
+    {
+      "epoch": 0.5077417717234961,
+      "grad_norm": 0.8979272246360779,
+      "learning_rate": 0.001,
+      "loss": 2.6012,
+      "step": 12002
+    },
+    {
+      "epoch": 0.5077840764870124,
+      "grad_norm": 0.20839910209178925,
+      "learning_rate": 0.001,
+      "loss": 2.8517,
+      "step": 12003
+    },
+    {
+      "epoch": 0.5078263812505288,
+      "grad_norm": 0.2238219529390335,
+      "learning_rate": 0.001,
+      "loss": 2.2089,
+      "step": 12004
+    },
+    {
+      "epoch": 0.5078686860140452,
+      "grad_norm": 0.21982823312282562,
+      "learning_rate": 0.001,
+      "loss": 1.9233,
+      "step": 12005
+    },
+    {
+      "epoch": 0.5079109907775615,
+      "grad_norm": 0.20902688801288605,
+      "learning_rate": 0.001,
+      "loss": 2.416,
+      "step": 12006
+    },
+    {
+      "epoch": 0.5079532955410779,
+      "grad_norm": 0.4519989788532257,
+      "learning_rate": 0.001,
+      "loss": 1.7383,
+      "step": 12007
+    },
+    {
+      "epoch": 0.5079956003045943,
+      "grad_norm": 0.15626175701618195,
+      "learning_rate": 0.001,
+      "loss": 3.0287,
+      "step": 12008
+    },
+    {
+      "epoch": 0.5080379050681106,
+      "grad_norm": 0.1568712294101715,
+      "learning_rate": 0.001,
+      "loss": 2.3071,
+      "step": 12009
+    },
+    {
+      "epoch": 0.508080209831627,
+      "grad_norm": 0.17470332980155945,
+      "learning_rate": 0.001,
+      "loss": 1.9627,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5081225145951435,
+      "grad_norm": 0.29563674330711365,
+      "learning_rate": 0.001,
+      "loss": 2.2517,
+      "step": 12011
+    },
+    {
+      "epoch": 0.5081648193586598,
+      "grad_norm": 0.1537560075521469,
+      "learning_rate": 0.001,
+      "loss": 1.5154,
+      "step": 12012
+    },
+    {
+      "epoch": 0.5082071241221762,
+      "grad_norm": 0.24751636385917664,
+      "learning_rate": 0.001,
+      "loss": 2.1838,
+      "step": 12013
+    },
+    {
+      "epoch": 0.5082494288856926,
+      "grad_norm": 0.2212720662355423,
+      "learning_rate": 0.001,
+      "loss": 2.7298,
+      "step": 12014
+    },
+    {
+      "epoch": 0.5082917336492089,
+      "grad_norm": 0.1655520498752594,
+      "learning_rate": 0.001,
+      "loss": 2.1824,
+      "step": 12015
+    },
+    {
+      "epoch": 0.5083340384127253,
+      "grad_norm": 0.17084214091300964,
+      "learning_rate": 0.001,
+      "loss": 2.413,
+      "step": 12016
+    },
+    {
+      "epoch": 0.5083763431762417,
+      "grad_norm": 0.1613297313451767,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 12017
+    },
+    {
+      "epoch": 0.508418647939758,
+      "grad_norm": 0.28114500641822815,
+      "learning_rate": 0.001,
+      "loss": 1.6721,
+      "step": 12018
+    },
+    {
+      "epoch": 0.5084609527032744,
+      "grad_norm": 0.19233211874961853,
+      "learning_rate": 0.001,
+      "loss": 1.9725,
+      "step": 12019
+    },
+    {
+      "epoch": 0.5085032574667908,
+      "grad_norm": 0.17586541175842285,
+      "learning_rate": 0.001,
+      "loss": 1.7528,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5085455622303071,
+      "grad_norm": 0.1447594314813614,
+      "learning_rate": 0.001,
+      "loss": 2.6402,
+      "step": 12021
+    },
+    {
+      "epoch": 0.5085878669938235,
+      "grad_norm": 0.14185841381549835,
+      "learning_rate": 0.001,
+      "loss": 1.6817,
+      "step": 12022
+    },
+    {
+      "epoch": 0.5086301717573398,
+      "grad_norm": 0.171967551112175,
+      "learning_rate": 0.001,
+      "loss": 2.3004,
+      "step": 12023
+    },
+    {
+      "epoch": 0.5086724765208562,
+      "grad_norm": 0.18312491476535797,
+      "learning_rate": 0.001,
+      "loss": 2.5789,
+      "step": 12024
+    },
+    {
+      "epoch": 0.5087147812843726,
+      "grad_norm": 0.298218697309494,
+      "learning_rate": 0.001,
+      "loss": 2.98,
+      "step": 12025
+    },
+    {
+      "epoch": 0.5087570860478889,
+      "grad_norm": 0.1913619339466095,
+      "learning_rate": 0.001,
+      "loss": 3.5224,
+      "step": 12026
+    },
+    {
+      "epoch": 0.5087993908114053,
+      "grad_norm": 0.18118982017040253,
+      "learning_rate": 0.001,
+      "loss": 1.9072,
+      "step": 12027
+    },
+    {
+      "epoch": 0.5088416955749218,
+      "grad_norm": 0.6203705072402954,
+      "learning_rate": 0.001,
+      "loss": 2.6366,
+      "step": 12028
+    },
+    {
+      "epoch": 0.5088840003384381,
+      "grad_norm": 0.17132526636123657,
+      "learning_rate": 0.001,
+      "loss": 2.0598,
+      "step": 12029
+    },
+    {
+      "epoch": 0.5089263051019545,
+      "grad_norm": 0.18120494484901428,
+      "learning_rate": 0.001,
+      "loss": 2.783,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5089686098654709,
+      "grad_norm": 0.20556822419166565,
+      "learning_rate": 0.001,
+      "loss": 2.4337,
+      "step": 12031
+    },
+    {
+      "epoch": 0.5090109146289872,
+      "grad_norm": 0.16901428997516632,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5090532193925036,
+      "grad_norm": 0.16724789142608643,
+      "learning_rate": 0.001,
+      "loss": 2.1285,
+      "step": 12033
+    },
+    {
+      "epoch": 0.50909552415602,
+      "grad_norm": 0.15585707128047943,
+      "learning_rate": 0.001,
+      "loss": 1.7745,
+      "step": 12034
+    },
+    {
+      "epoch": 0.5091378289195363,
+      "grad_norm": 0.5937365889549255,
+      "learning_rate": 0.001,
+      "loss": 2.4572,
+      "step": 12035
+    },
+    {
+      "epoch": 0.5091801336830527,
+      "grad_norm": 0.17110083997249603,
+      "learning_rate": 0.001,
+      "loss": 2.4326,
+      "step": 12036
+    },
+    {
+      "epoch": 0.5092224384465691,
+      "grad_norm": 0.18722589313983917,
+      "learning_rate": 0.001,
+      "loss": 3.232,
+      "step": 12037
+    },
+    {
+      "epoch": 0.5092647432100854,
+      "grad_norm": 0.2509753704071045,
+      "learning_rate": 0.001,
+      "loss": 2.1435,
+      "step": 12038
+    },
+    {
+      "epoch": 0.5093070479736018,
+      "grad_norm": 0.1581326723098755,
+      "learning_rate": 0.001,
+      "loss": 2.3724,
+      "step": 12039
+    },
+    {
+      "epoch": 0.5093493527371182,
+      "grad_norm": 0.2504560053348541,
+      "learning_rate": 0.001,
+      "loss": 2.3174,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5093916575006345,
+      "grad_norm": 0.19486159086227417,
+      "learning_rate": 0.001,
+      "loss": 2.1135,
+      "step": 12041
+    },
+    {
+      "epoch": 0.5094339622641509,
+      "grad_norm": 0.16291144490242004,
+      "learning_rate": 0.001,
+      "loss": 1.9069,
+      "step": 12042
+    },
+    {
+      "epoch": 0.5094762670276674,
+      "grad_norm": 0.17306125164031982,
+      "learning_rate": 0.001,
+      "loss": 2.8266,
+      "step": 12043
+    },
+    {
+      "epoch": 0.5095185717911836,
+      "grad_norm": 0.17126202583312988,
+      "learning_rate": 0.001,
+      "loss": 3.3905,
+      "step": 12044
+    },
+    {
+      "epoch": 0.5095608765547001,
+      "grad_norm": 0.18580661714076996,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 12045
+    },
+    {
+      "epoch": 0.5096031813182165,
+      "grad_norm": 1.9326337575912476,
+      "learning_rate": 0.001,
+      "loss": 1.8201,
+      "step": 12046
+    },
+    {
+      "epoch": 0.5096454860817328,
+      "grad_norm": 0.24594885110855103,
+      "learning_rate": 0.001,
+      "loss": 2.2209,
+      "step": 12047
+    },
+    {
+      "epoch": 0.5096877908452492,
+      "grad_norm": 0.14384353160858154,
+      "learning_rate": 0.001,
+      "loss": 1.7018,
+      "step": 12048
+    },
+    {
+      "epoch": 0.5097300956087656,
+      "grad_norm": 0.1911388784646988,
+      "learning_rate": 0.001,
+      "loss": 2.9794,
+      "step": 12049
+    },
+    {
+      "epoch": 0.5097724003722819,
+      "grad_norm": 0.15660209953784943,
+      "learning_rate": 0.001,
+      "loss": 2.4576,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5098147051357983,
+      "grad_norm": 0.27428507804870605,
+      "learning_rate": 0.001,
+      "loss": 2.893,
+      "step": 12051
+    },
+    {
+      "epoch": 0.5098570098993147,
+      "grad_norm": 0.17285604774951935,
+      "learning_rate": 0.001,
+      "loss": 1.984,
+      "step": 12052
+    },
+    {
+      "epoch": 0.509899314662831,
+      "grad_norm": 0.17490775883197784,
+      "learning_rate": 0.001,
+      "loss": 2.0082,
+      "step": 12053
+    },
+    {
+      "epoch": 0.5099416194263474,
+      "grad_norm": 0.1558762937784195,
+      "learning_rate": 0.001,
+      "loss": 1.9196,
+      "step": 12054
+    },
+    {
+      "epoch": 0.5099839241898638,
+      "grad_norm": 0.42359229922294617,
+      "learning_rate": 0.001,
+      "loss": 2.7122,
+      "step": 12055
+    },
+    {
+      "epoch": 0.5100262289533801,
+      "grad_norm": 0.17612966895103455,
+      "learning_rate": 0.001,
+      "loss": 1.7214,
+      "step": 12056
+    },
+    {
+      "epoch": 0.5100685337168965,
+      "grad_norm": 0.3643682599067688,
+      "learning_rate": 0.001,
+      "loss": 1.7763,
+      "step": 12057
+    },
+    {
+      "epoch": 0.5101108384804129,
+      "grad_norm": 0.205791637301445,
+      "learning_rate": 0.001,
+      "loss": 2.0399,
+      "step": 12058
+    },
+    {
+      "epoch": 0.5101531432439292,
+      "grad_norm": 2.9326553344726562,
+      "learning_rate": 0.001,
+      "loss": 1.8045,
+      "step": 12059
+    },
+    {
+      "epoch": 0.5101954480074457,
+      "grad_norm": 0.22265328466892242,
+      "learning_rate": 0.001,
+      "loss": 2.1796,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5102377527709621,
+      "grad_norm": 0.17063187062740326,
+      "learning_rate": 0.001,
+      "loss": 2.4523,
+      "step": 12061
+    },
+    {
+      "epoch": 0.5102800575344784,
+      "grad_norm": 0.18064865469932556,
+      "learning_rate": 0.001,
+      "loss": 1.5731,
+      "step": 12062
+    },
+    {
+      "epoch": 0.5103223622979948,
+      "grad_norm": 2.2805631160736084,
+      "learning_rate": 0.001,
+      "loss": 3.7232,
+      "step": 12063
+    },
+    {
+      "epoch": 0.5103646670615112,
+      "grad_norm": 0.1840154379606247,
+      "learning_rate": 0.001,
+      "loss": 1.7874,
+      "step": 12064
+    },
+    {
+      "epoch": 0.5104069718250275,
+      "grad_norm": 0.16922450065612793,
+      "learning_rate": 0.001,
+      "loss": 1.9633,
+      "step": 12065
+    },
+    {
+      "epoch": 0.5104492765885439,
+      "grad_norm": 0.1468036025762558,
+      "learning_rate": 0.001,
+      "loss": 1.8785,
+      "step": 12066
+    },
+    {
+      "epoch": 0.5104915813520603,
+      "grad_norm": 0.15584182739257812,
+      "learning_rate": 0.001,
+      "loss": 2.47,
+      "step": 12067
+    },
+    {
+      "epoch": 0.5105338861155766,
+      "grad_norm": 0.19544000923633575,
+      "learning_rate": 0.001,
+      "loss": 2.2966,
+      "step": 12068
+    },
+    {
+      "epoch": 0.510576190879093,
+      "grad_norm": 0.15952937304973602,
+      "learning_rate": 0.001,
+      "loss": 1.9361,
+      "step": 12069
+    },
+    {
+      "epoch": 0.5106184956426093,
+      "grad_norm": 2.266671657562256,
+      "learning_rate": 0.001,
+      "loss": 3.9638,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5106608004061257,
+      "grad_norm": 0.20230835676193237,
+      "learning_rate": 0.001,
+      "loss": 2.0883,
+      "step": 12071
+    },
+    {
+      "epoch": 0.5107031051696421,
+      "grad_norm": 0.22591619193553925,
+      "learning_rate": 0.001,
+      "loss": 2.8491,
+      "step": 12072
+    },
+    {
+      "epoch": 0.5107454099331584,
+      "grad_norm": 0.18922626972198486,
+      "learning_rate": 0.001,
+      "loss": 1.9407,
+      "step": 12073
+    },
+    {
+      "epoch": 0.5107877146966748,
+      "grad_norm": 0.17691117525100708,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 12074
+    },
+    {
+      "epoch": 0.5108300194601912,
+      "grad_norm": 0.17793789505958557,
+      "learning_rate": 0.001,
+      "loss": 2.0651,
+      "step": 12075
+    },
+    {
+      "epoch": 0.5108723242237075,
+      "grad_norm": 0.2747514545917511,
+      "learning_rate": 0.001,
+      "loss": 1.9631,
+      "step": 12076
+    },
+    {
+      "epoch": 0.510914628987224,
+      "grad_norm": 0.16500215232372284,
+      "learning_rate": 0.001,
+      "loss": 1.5828,
+      "step": 12077
+    },
+    {
+      "epoch": 0.5109569337507404,
+      "grad_norm": 0.2846604883670807,
+      "learning_rate": 0.001,
+      "loss": 3.0458,
+      "step": 12078
+    },
+    {
+      "epoch": 0.5109992385142567,
+      "grad_norm": 0.18914783000946045,
+      "learning_rate": 0.001,
+      "loss": 2.501,
+      "step": 12079
+    },
+    {
+      "epoch": 0.5110415432777731,
+      "grad_norm": 3.4462015628814697,
+      "learning_rate": 0.001,
+      "loss": 1.8007,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5110838480412895,
+      "grad_norm": 0.22221924364566803,
+      "learning_rate": 0.001,
+      "loss": 2.4593,
+      "step": 12081
+    },
+    {
+      "epoch": 0.5111261528048058,
+      "grad_norm": 0.3206009268760681,
+      "learning_rate": 0.001,
+      "loss": 2.1852,
+      "step": 12082
+    },
+    {
+      "epoch": 0.5111684575683222,
+      "grad_norm": 0.19568482041358948,
+      "learning_rate": 0.001,
+      "loss": 2.0723,
+      "step": 12083
+    },
+    {
+      "epoch": 0.5112107623318386,
+      "grad_norm": 0.13651347160339355,
+      "learning_rate": 0.001,
+      "loss": 2.1792,
+      "step": 12084
+    },
+    {
+      "epoch": 0.5112530670953549,
+      "grad_norm": 0.9014352560043335,
+      "learning_rate": 0.001,
+      "loss": 2.1313,
+      "step": 12085
+    },
+    {
+      "epoch": 0.5112953718588713,
+      "grad_norm": 0.17774808406829834,
+      "learning_rate": 0.001,
+      "loss": 2.194,
+      "step": 12086
+    },
+    {
+      "epoch": 0.5113376766223877,
+      "grad_norm": 0.2228837013244629,
+      "learning_rate": 0.001,
+      "loss": 3.0247,
+      "step": 12087
+    },
+    {
+      "epoch": 0.511379981385904,
+      "grad_norm": 0.20427215099334717,
+      "learning_rate": 0.001,
+      "loss": 2.176,
+      "step": 12088
+    },
+    {
+      "epoch": 0.5114222861494204,
+      "grad_norm": 0.1658029705286026,
+      "learning_rate": 0.001,
+      "loss": 2.0688,
+      "step": 12089
+    },
+    {
+      "epoch": 0.5114645909129368,
+      "grad_norm": 0.19632932543754578,
+      "learning_rate": 0.001,
+      "loss": 3.3308,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5115068956764531,
+      "grad_norm": 0.17053432762622833,
+      "learning_rate": 0.001,
+      "loss": 2.1211,
+      "step": 12091
+    },
+    {
+      "epoch": 0.5115492004399695,
+      "grad_norm": 0.17195382714271545,
+      "learning_rate": 0.001,
+      "loss": 1.7132,
+      "step": 12092
+    },
+    {
+      "epoch": 0.511591505203486,
+      "grad_norm": 0.17372830212116241,
+      "learning_rate": 0.001,
+      "loss": 2.7497,
+      "step": 12093
+    },
+    {
+      "epoch": 0.5116338099670023,
+      "grad_norm": 0.18535274267196655,
+      "learning_rate": 0.001,
+      "loss": 2.2781,
+      "step": 12094
+    },
+    {
+      "epoch": 0.5116761147305187,
+      "grad_norm": 0.1690640151500702,
+      "learning_rate": 0.001,
+      "loss": 1.9845,
+      "step": 12095
+    },
+    {
+      "epoch": 0.5117184194940351,
+      "grad_norm": 0.18286339938640594,
+      "learning_rate": 0.001,
+      "loss": 3.2735,
+      "step": 12096
+    },
+    {
+      "epoch": 0.5117607242575514,
+      "grad_norm": 0.15377508103847504,
+      "learning_rate": 0.001,
+      "loss": 2.0673,
+      "step": 12097
+    },
+    {
+      "epoch": 0.5118030290210678,
+      "grad_norm": 1.7860561609268188,
+      "learning_rate": 0.001,
+      "loss": 2.9895,
+      "step": 12098
+    },
+    {
+      "epoch": 0.5118453337845842,
+      "grad_norm": 0.2976422607898712,
+      "learning_rate": 0.001,
+      "loss": 2.5911,
+      "step": 12099
+    },
+    {
+      "epoch": 0.5118876385481005,
+      "grad_norm": 0.6552213430404663,
+      "learning_rate": 0.001,
+      "loss": 3.01,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5119299433116169,
+      "grad_norm": 0.1674286127090454,
+      "learning_rate": 0.001,
+      "loss": 1.5554,
+      "step": 12101
+    },
+    {
+      "epoch": 0.5119722480751333,
+      "grad_norm": 0.17942678928375244,
+      "learning_rate": 0.001,
+      "loss": 2.5887,
+      "step": 12102
+    },
+    {
+      "epoch": 0.5120145528386496,
+      "grad_norm": 0.3610702455043793,
+      "learning_rate": 0.001,
+      "loss": 2.0485,
+      "step": 12103
+    },
+    {
+      "epoch": 0.512056857602166,
+      "grad_norm": 9.683119773864746,
+      "learning_rate": 0.001,
+      "loss": 1.5999,
+      "step": 12104
+    },
+    {
+      "epoch": 0.5120991623656824,
+      "grad_norm": 0.23564742505550385,
+      "learning_rate": 0.001,
+      "loss": 2.6848,
+      "step": 12105
+    },
+    {
+      "epoch": 0.5121414671291987,
+      "grad_norm": 0.20375941693782806,
+      "learning_rate": 0.001,
+      "loss": 2.348,
+      "step": 12106
+    },
+    {
+      "epoch": 0.5121837718927151,
+      "grad_norm": 0.17690829932689667,
+      "learning_rate": 0.001,
+      "loss": 1.8304,
+      "step": 12107
+    },
+    {
+      "epoch": 0.5122260766562315,
+      "grad_norm": 0.3078082501888275,
+      "learning_rate": 0.001,
+      "loss": 2.4127,
+      "step": 12108
+    },
+    {
+      "epoch": 0.5122683814197478,
+      "grad_norm": 0.17831426858901978,
+      "learning_rate": 0.001,
+      "loss": 2.4171,
+      "step": 12109
+    },
+    {
+      "epoch": 0.5123106861832643,
+      "grad_norm": 0.19849298894405365,
+      "learning_rate": 0.001,
+      "loss": 3.024,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5123529909467807,
+      "grad_norm": 0.25595778226852417,
+      "learning_rate": 0.001,
+      "loss": 1.8982,
+      "step": 12111
+    },
+    {
+      "epoch": 0.512395295710297,
+      "grad_norm": 0.19416412711143494,
+      "learning_rate": 0.001,
+      "loss": 2.3878,
+      "step": 12112
+    },
+    {
+      "epoch": 0.5124376004738134,
+      "grad_norm": 0.5815975666046143,
+      "learning_rate": 0.001,
+      "loss": 2.9711,
+      "step": 12113
+    },
+    {
+      "epoch": 0.5124799052373297,
+      "grad_norm": 0.34129148721694946,
+      "learning_rate": 0.001,
+      "loss": 2.6444,
+      "step": 12114
+    },
+    {
+      "epoch": 0.5125222100008461,
+      "grad_norm": 0.495937705039978,
+      "learning_rate": 0.001,
+      "loss": 2.1635,
+      "step": 12115
+    },
+    {
+      "epoch": 0.5125645147643625,
+      "grad_norm": 0.23061814904212952,
+      "learning_rate": 0.001,
+      "loss": 2.387,
+      "step": 12116
+    },
+    {
+      "epoch": 0.5126068195278788,
+      "grad_norm": 0.28698426485061646,
+      "learning_rate": 0.001,
+      "loss": 2.4555,
+      "step": 12117
+    },
+    {
+      "epoch": 0.5126491242913952,
+      "grad_norm": 0.3059033453464508,
+      "learning_rate": 0.001,
+      "loss": 2.024,
+      "step": 12118
+    },
+    {
+      "epoch": 0.5126914290549116,
+      "grad_norm": 0.24312277138233185,
+      "learning_rate": 0.001,
+      "loss": 2.1431,
+      "step": 12119
+    },
+    {
+      "epoch": 0.5127337338184279,
+      "grad_norm": 0.17924846708774567,
+      "learning_rate": 0.001,
+      "loss": 2.322,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5127760385819443,
+      "grad_norm": 0.15020184218883514,
+      "learning_rate": 0.001,
+      "loss": 1.4752,
+      "step": 12121
+    },
+    {
+      "epoch": 0.5128183433454607,
+      "grad_norm": 3.8008360862731934,
+      "learning_rate": 0.001,
+      "loss": 2.591,
+      "step": 12122
+    },
+    {
+      "epoch": 0.512860648108977,
+      "grad_norm": 0.20697496831417084,
+      "learning_rate": 0.001,
+      "loss": 2.5281,
+      "step": 12123
+    },
+    {
+      "epoch": 0.5129029528724934,
+      "grad_norm": 0.18046455085277557,
+      "learning_rate": 0.001,
+      "loss": 1.9783,
+      "step": 12124
+    },
+    {
+      "epoch": 0.5129452576360098,
+      "grad_norm": 4.528714656829834,
+      "learning_rate": 0.001,
+      "loss": 1.7568,
+      "step": 12125
+    },
+    {
+      "epoch": 0.5129875623995261,
+      "grad_norm": 0.520558774471283,
+      "learning_rate": 0.001,
+      "loss": 2.9131,
+      "step": 12126
+    },
+    {
+      "epoch": 0.5130298671630426,
+      "grad_norm": 0.529538631439209,
+      "learning_rate": 0.001,
+      "loss": 2.5839,
+      "step": 12127
+    },
+    {
+      "epoch": 0.513072171926559,
+      "grad_norm": 0.3380816578865051,
+      "learning_rate": 0.001,
+      "loss": 1.9767,
+      "step": 12128
+    },
+    {
+      "epoch": 0.5131144766900753,
+      "grad_norm": 0.5433550477027893,
+      "learning_rate": 0.001,
+      "loss": 2.5798,
+      "step": 12129
+    },
+    {
+      "epoch": 0.5131567814535917,
+      "grad_norm": 0.17710618674755096,
+      "learning_rate": 0.001,
+      "loss": 2.614,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5131990862171081,
+      "grad_norm": 0.23254965245723724,
+      "learning_rate": 0.001,
+      "loss": 2.8047,
+      "step": 12131
+    },
+    {
+      "epoch": 0.5132413909806244,
+      "grad_norm": 0.1852521151304245,
+      "learning_rate": 0.001,
+      "loss": 2.0367,
+      "step": 12132
+    },
+    {
+      "epoch": 0.5132836957441408,
+      "grad_norm": 0.19175705313682556,
+      "learning_rate": 0.001,
+      "loss": 2.0662,
+      "step": 12133
+    },
+    {
+      "epoch": 0.5133260005076572,
+      "grad_norm": 0.18979766964912415,
+      "learning_rate": 0.001,
+      "loss": 2.3063,
+      "step": 12134
+    },
+    {
+      "epoch": 0.5133683052711735,
+      "grad_norm": 0.1920190006494522,
+      "learning_rate": 0.001,
+      "loss": 2.7921,
+      "step": 12135
+    },
+    {
+      "epoch": 0.5134106100346899,
+      "grad_norm": 0.5775536298751831,
+      "learning_rate": 0.001,
+      "loss": 2.0244,
+      "step": 12136
+    },
+    {
+      "epoch": 0.5134529147982063,
+      "grad_norm": 0.1514301896095276,
+      "learning_rate": 0.001,
+      "loss": 1.7011,
+      "step": 12137
+    },
+    {
+      "epoch": 0.5134952195617226,
+      "grad_norm": 0.15868137776851654,
+      "learning_rate": 0.001,
+      "loss": 2.9314,
+      "step": 12138
+    },
+    {
+      "epoch": 0.513537524325239,
+      "grad_norm": 0.18454696238040924,
+      "learning_rate": 0.001,
+      "loss": 2.7521,
+      "step": 12139
+    },
+    {
+      "epoch": 0.5135798290887554,
+      "grad_norm": 0.21990300714969635,
+      "learning_rate": 0.001,
+      "loss": 1.5834,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5136221338522717,
+      "grad_norm": 0.21516460180282593,
+      "learning_rate": 0.001,
+      "loss": 2.0828,
+      "step": 12141
+    },
+    {
+      "epoch": 0.5136644386157881,
+      "grad_norm": 0.2636964023113251,
+      "learning_rate": 0.001,
+      "loss": 2.2502,
+      "step": 12142
+    },
+    {
+      "epoch": 0.5137067433793046,
+      "grad_norm": 0.21557240188121796,
+      "learning_rate": 0.001,
+      "loss": 2.4408,
+      "step": 12143
+    },
+    {
+      "epoch": 0.5137490481428209,
+      "grad_norm": 8.734152793884277,
+      "learning_rate": 0.001,
+      "loss": 2.9185,
+      "step": 12144
+    },
+    {
+      "epoch": 0.5137913529063373,
+      "grad_norm": 0.32210806012153625,
+      "learning_rate": 0.001,
+      "loss": 2.2702,
+      "step": 12145
+    },
+    {
+      "epoch": 0.5138336576698537,
+      "grad_norm": 0.3446987569332123,
+      "learning_rate": 0.001,
+      "loss": 2.9083,
+      "step": 12146
+    },
+    {
+      "epoch": 0.51387596243337,
+      "grad_norm": 3.7084527015686035,
+      "learning_rate": 0.001,
+      "loss": 1.4133,
+      "step": 12147
+    },
+    {
+      "epoch": 0.5139182671968864,
+      "grad_norm": 3.1599371433258057,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 12148
+    },
+    {
+      "epoch": 0.5139605719604028,
+      "grad_norm": 0.19752854108810425,
+      "learning_rate": 0.001,
+      "loss": 2.0721,
+      "step": 12149
+    },
+    {
+      "epoch": 0.5140028767239191,
+      "grad_norm": 0.49823862314224243,
+      "learning_rate": 0.001,
+      "loss": 1.9638,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5140451814874355,
+      "grad_norm": 0.1748315989971161,
+      "learning_rate": 0.001,
+      "loss": 3.1363,
+      "step": 12151
+    },
+    {
+      "epoch": 0.5140874862509519,
+      "grad_norm": 0.1747720092535019,
+      "learning_rate": 0.001,
+      "loss": 2.1459,
+      "step": 12152
+    },
+    {
+      "epoch": 0.5141297910144682,
+      "grad_norm": 2.5184714794158936,
+      "learning_rate": 0.001,
+      "loss": 1.7912,
+      "step": 12153
+    },
+    {
+      "epoch": 0.5141720957779846,
+      "grad_norm": 0.30171647667884827,
+      "learning_rate": 0.001,
+      "loss": 2.5741,
+      "step": 12154
+    },
+    {
+      "epoch": 0.514214400541501,
+      "grad_norm": 0.19412563741207123,
+      "learning_rate": 0.001,
+      "loss": 1.8221,
+      "step": 12155
+    },
+    {
+      "epoch": 0.5142567053050173,
+      "grad_norm": 0.44253069162368774,
+      "learning_rate": 0.001,
+      "loss": 2.7872,
+      "step": 12156
+    },
+    {
+      "epoch": 0.5142990100685337,
+      "grad_norm": 0.16020557284355164,
+      "learning_rate": 0.001,
+      "loss": 1.8939,
+      "step": 12157
+    },
+    {
+      "epoch": 0.51434131483205,
+      "grad_norm": 0.6392630934715271,
+      "learning_rate": 0.001,
+      "loss": 2.4378,
+      "step": 12158
+    },
+    {
+      "epoch": 0.5143836195955664,
+      "grad_norm": 0.15356384217739105,
+      "learning_rate": 0.001,
+      "loss": 2.2201,
+      "step": 12159
+    },
+    {
+      "epoch": 0.5144259243590829,
+      "grad_norm": 0.2277487814426422,
+      "learning_rate": 0.001,
+      "loss": 2.1108,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5144682291225992,
+      "grad_norm": 0.19493798911571503,
+      "learning_rate": 0.001,
+      "loss": 2.7649,
+      "step": 12161
+    },
+    {
+      "epoch": 0.5145105338861156,
+      "grad_norm": 0.8036705851554871,
+      "learning_rate": 0.001,
+      "loss": 2.6234,
+      "step": 12162
+    },
+    {
+      "epoch": 0.514552838649632,
+      "grad_norm": 0.23683510720729828,
+      "learning_rate": 0.001,
+      "loss": 3.3676,
+      "step": 12163
+    },
+    {
+      "epoch": 0.5145951434131483,
+      "grad_norm": 0.20313666760921478,
+      "learning_rate": 0.001,
+      "loss": 2.2018,
+      "step": 12164
+    },
+    {
+      "epoch": 0.5146374481766647,
+      "grad_norm": 0.16080090403556824,
+      "learning_rate": 0.001,
+      "loss": 1.4551,
+      "step": 12165
+    },
+    {
+      "epoch": 0.5146797529401811,
+      "grad_norm": 0.19314774870872498,
+      "learning_rate": 0.001,
+      "loss": 2.5364,
+      "step": 12166
+    },
+    {
+      "epoch": 0.5147220577036974,
+      "grad_norm": 0.2120058387517929,
+      "learning_rate": 0.001,
+      "loss": 2.1422,
+      "step": 12167
+    },
+    {
+      "epoch": 0.5147643624672138,
+      "grad_norm": 0.15870290994644165,
+      "learning_rate": 0.001,
+      "loss": 1.7913,
+      "step": 12168
+    },
+    {
+      "epoch": 0.5148066672307302,
+      "grad_norm": 1.8567591905593872,
+      "learning_rate": 0.001,
+      "loss": 1.8428,
+      "step": 12169
+    },
+    {
+      "epoch": 0.5148489719942465,
+      "grad_norm": 0.25935545563697815,
+      "learning_rate": 0.001,
+      "loss": 2.1154,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5148912767577629,
+      "grad_norm": 0.19411538541316986,
+      "learning_rate": 0.001,
+      "loss": 1.6982,
+      "step": 12171
+    },
+    {
+      "epoch": 0.5149335815212793,
+      "grad_norm": 0.1572023630142212,
+      "learning_rate": 0.001,
+      "loss": 1.9115,
+      "step": 12172
+    },
+    {
+      "epoch": 0.5149758862847956,
+      "grad_norm": 1.9859943389892578,
+      "learning_rate": 0.001,
+      "loss": 2.9479,
+      "step": 12173
+    },
+    {
+      "epoch": 0.515018191048312,
+      "grad_norm": 0.16157686710357666,
+      "learning_rate": 0.001,
+      "loss": 2.1507,
+      "step": 12174
+    },
+    {
+      "epoch": 0.5150604958118284,
+      "grad_norm": 1.8809139728546143,
+      "learning_rate": 0.001,
+      "loss": 2.054,
+      "step": 12175
+    },
+    {
+      "epoch": 0.5151028005753447,
+      "grad_norm": 0.17030459642410278,
+      "learning_rate": 0.001,
+      "loss": 1.8893,
+      "step": 12176
+    },
+    {
+      "epoch": 0.5151451053388612,
+      "grad_norm": 0.16130401194095612,
+      "learning_rate": 0.001,
+      "loss": 2.5668,
+      "step": 12177
+    },
+    {
+      "epoch": 0.5151874101023776,
+      "grad_norm": 0.19014528393745422,
+      "learning_rate": 0.001,
+      "loss": 2.4916,
+      "step": 12178
+    },
+    {
+      "epoch": 0.5152297148658939,
+      "grad_norm": 0.4937807321548462,
+      "learning_rate": 0.001,
+      "loss": 2.3189,
+      "step": 12179
+    },
+    {
+      "epoch": 0.5152720196294103,
+      "grad_norm": 0.24521322548389435,
+      "learning_rate": 0.001,
+      "loss": 1.4842,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5153143243929267,
+      "grad_norm": 0.17198513448238373,
+      "learning_rate": 0.001,
+      "loss": 2.2285,
+      "step": 12181
+    },
+    {
+      "epoch": 0.515356629156443,
+      "grad_norm": 0.2905202805995941,
+      "learning_rate": 0.001,
+      "loss": 3.2654,
+      "step": 12182
+    },
+    {
+      "epoch": 0.5153989339199594,
+      "grad_norm": 0.1737702488899231,
+      "learning_rate": 0.001,
+      "loss": 2.1419,
+      "step": 12183
+    },
+    {
+      "epoch": 0.5154412386834758,
+      "grad_norm": 0.23815499246120453,
+      "learning_rate": 0.001,
+      "loss": 1.6129,
+      "step": 12184
+    },
+    {
+      "epoch": 0.5154835434469921,
+      "grad_norm": 0.19331566989421844,
+      "learning_rate": 0.001,
+      "loss": 2.0372,
+      "step": 12185
+    },
+    {
+      "epoch": 0.5155258482105085,
+      "grad_norm": 0.20304280519485474,
+      "learning_rate": 0.001,
+      "loss": 2.8925,
+      "step": 12186
+    },
+    {
+      "epoch": 0.5155681529740249,
+      "grad_norm": 0.16830262541770935,
+      "learning_rate": 0.001,
+      "loss": 2.4063,
+      "step": 12187
+    },
+    {
+      "epoch": 0.5156104577375412,
+      "grad_norm": 0.2067635953426361,
+      "learning_rate": 0.001,
+      "loss": 2.4644,
+      "step": 12188
+    },
+    {
+      "epoch": 0.5156527625010576,
+      "grad_norm": 0.26112356781959534,
+      "learning_rate": 0.001,
+      "loss": 1.8161,
+      "step": 12189
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 0.317443311214447,
+      "learning_rate": 0.001,
+      "loss": 2.3216,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5157373720280903,
+      "grad_norm": 0.1734699159860611,
+      "learning_rate": 0.001,
+      "loss": 2.4626,
+      "step": 12191
+    },
+    {
+      "epoch": 0.5157796767916067,
+      "grad_norm": 0.21294571459293365,
+      "learning_rate": 0.001,
+      "loss": 3.3039,
+      "step": 12192
+    },
+    {
+      "epoch": 0.5158219815551232,
+      "grad_norm": 0.6978985071182251,
+      "learning_rate": 0.001,
+      "loss": 3.0031,
+      "step": 12193
+    },
+    {
+      "epoch": 0.5158642863186395,
+      "grad_norm": 0.38809531927108765,
+      "learning_rate": 0.001,
+      "loss": 1.5207,
+      "step": 12194
+    },
+    {
+      "epoch": 0.5159065910821559,
+      "grad_norm": 0.17903557419776917,
+      "learning_rate": 0.001,
+      "loss": 3.2565,
+      "step": 12195
+    },
+    {
+      "epoch": 0.5159488958456723,
+      "grad_norm": 3.361924648284912,
+      "learning_rate": 0.001,
+      "loss": 2.2003,
+      "step": 12196
+    },
+    {
+      "epoch": 0.5159912006091886,
+      "grad_norm": 0.16305619478225708,
+      "learning_rate": 0.001,
+      "loss": 1.7818,
+      "step": 12197
+    },
+    {
+      "epoch": 0.516033505372705,
+      "grad_norm": 2.384248733520508,
+      "learning_rate": 0.001,
+      "loss": 2.5049,
+      "step": 12198
+    },
+    {
+      "epoch": 0.5160758101362214,
+      "grad_norm": 0.1969674527645111,
+      "learning_rate": 0.001,
+      "loss": 2.1683,
+      "step": 12199
+    },
+    {
+      "epoch": 0.5161181148997377,
+      "grad_norm": 0.42641380429267883,
+      "learning_rate": 0.001,
+      "loss": 1.9276,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5161604196632541,
+      "grad_norm": 0.18019993603229523,
+      "learning_rate": 0.001,
+      "loss": 2.4077,
+      "step": 12201
+    },
+    {
+      "epoch": 0.5162027244267705,
+      "grad_norm": 0.2010054886341095,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 12202
+    },
+    {
+      "epoch": 0.5162450291902868,
+      "grad_norm": 0.31541866064071655,
+      "learning_rate": 0.001,
+      "loss": 2.5211,
+      "step": 12203
+    },
+    {
+      "epoch": 0.5162873339538032,
+      "grad_norm": 0.16396036744117737,
+      "learning_rate": 0.001,
+      "loss": 2.0067,
+      "step": 12204
+    },
+    {
+      "epoch": 0.5163296387173195,
+      "grad_norm": 0.1641075164079666,
+      "learning_rate": 0.001,
+      "loss": 1.8241,
+      "step": 12205
+    },
+    {
+      "epoch": 0.5163719434808359,
+      "grad_norm": 0.9481202960014343,
+      "learning_rate": 0.001,
+      "loss": 2.4869,
+      "step": 12206
+    },
+    {
+      "epoch": 0.5164142482443523,
+      "grad_norm": 0.17808358371257782,
+      "learning_rate": 0.001,
+      "loss": 1.9912,
+      "step": 12207
+    },
+    {
+      "epoch": 0.5164565530078686,
+      "grad_norm": 0.15392626821994781,
+      "learning_rate": 0.001,
+      "loss": 2.3439,
+      "step": 12208
+    },
+    {
+      "epoch": 0.516498857771385,
+      "grad_norm": 0.1524217277765274,
+      "learning_rate": 0.001,
+      "loss": 1.5912,
+      "step": 12209
+    },
+    {
+      "epoch": 0.5165411625349015,
+      "grad_norm": 0.21630914509296417,
+      "learning_rate": 0.001,
+      "loss": 1.8016,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5165834672984178,
+      "grad_norm": 0.1926201581954956,
+      "learning_rate": 0.001,
+      "loss": 2.1263,
+      "step": 12211
+    },
+    {
+      "epoch": 0.5166257720619342,
+      "grad_norm": 0.4192337691783905,
+      "learning_rate": 0.001,
+      "loss": 2.6866,
+      "step": 12212
+    },
+    {
+      "epoch": 0.5166680768254506,
+      "grad_norm": 2.8745205402374268,
+      "learning_rate": 0.001,
+      "loss": 1.7879,
+      "step": 12213
+    },
+    {
+      "epoch": 0.5167103815889669,
+      "grad_norm": 0.4913157820701599,
+      "learning_rate": 0.001,
+      "loss": 1.9727,
+      "step": 12214
+    },
+    {
+      "epoch": 0.5167526863524833,
+      "grad_norm": 0.2506055533885956,
+      "learning_rate": 0.001,
+      "loss": 2.5039,
+      "step": 12215
+    },
+    {
+      "epoch": 0.5167949911159997,
+      "grad_norm": 0.19540032744407654,
+      "learning_rate": 0.001,
+      "loss": 1.7671,
+      "step": 12216
+    },
+    {
+      "epoch": 0.516837295879516,
+      "grad_norm": 0.174033522605896,
+      "learning_rate": 0.001,
+      "loss": 2.5238,
+      "step": 12217
+    },
+    {
+      "epoch": 0.5168796006430324,
+      "grad_norm": 0.1758839190006256,
+      "learning_rate": 0.001,
+      "loss": 1.882,
+      "step": 12218
+    },
+    {
+      "epoch": 0.5169219054065488,
+      "grad_norm": 0.1755172461271286,
+      "learning_rate": 0.001,
+      "loss": 2.0126,
+      "step": 12219
+    },
+    {
+      "epoch": 0.5169642101700651,
+      "grad_norm": 0.2143266797065735,
+      "learning_rate": 0.001,
+      "loss": 2.2329,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5170065149335815,
+      "grad_norm": 0.20903651416301727,
+      "learning_rate": 0.001,
+      "loss": 2.5899,
+      "step": 12221
+    },
+    {
+      "epoch": 0.5170488196970979,
+      "grad_norm": 0.19313296675682068,
+      "learning_rate": 0.001,
+      "loss": 1.7994,
+      "step": 12222
+    },
+    {
+      "epoch": 0.5170911244606142,
+      "grad_norm": 0.1702556163072586,
+      "learning_rate": 0.001,
+      "loss": 1.8909,
+      "step": 12223
+    },
+    {
+      "epoch": 0.5171334292241306,
+      "grad_norm": 0.17211031913757324,
+      "learning_rate": 0.001,
+      "loss": 2.1869,
+      "step": 12224
+    },
+    {
+      "epoch": 0.517175733987647,
+      "grad_norm": 0.18440312147140503,
+      "learning_rate": 0.001,
+      "loss": 2.1057,
+      "step": 12225
+    },
+    {
+      "epoch": 0.5172180387511633,
+      "grad_norm": 0.18326327204704285,
+      "learning_rate": 0.001,
+      "loss": 2.9968,
+      "step": 12226
+    },
+    {
+      "epoch": 0.5172603435146798,
+      "grad_norm": 0.3492085933685303,
+      "learning_rate": 0.001,
+      "loss": 1.8782,
+      "step": 12227
+    },
+    {
+      "epoch": 0.5173026482781962,
+      "grad_norm": 0.20614147186279297,
+      "learning_rate": 0.001,
+      "loss": 1.6796,
+      "step": 12228
+    },
+    {
+      "epoch": 0.5173449530417125,
+      "grad_norm": 0.18441130220890045,
+      "learning_rate": 0.001,
+      "loss": 1.968,
+      "step": 12229
+    },
+    {
+      "epoch": 0.5173872578052289,
+      "grad_norm": 0.21896757185459137,
+      "learning_rate": 0.001,
+      "loss": 2.4805,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5174295625687453,
+      "grad_norm": 0.3186761736869812,
+      "learning_rate": 0.001,
+      "loss": 2.0246,
+      "step": 12231
+    },
+    {
+      "epoch": 0.5174718673322616,
+      "grad_norm": 0.24206700921058655,
+      "learning_rate": 0.001,
+      "loss": 2.3907,
+      "step": 12232
+    },
+    {
+      "epoch": 0.517514172095778,
+      "grad_norm": 0.19111651182174683,
+      "learning_rate": 0.001,
+      "loss": 1.9669,
+      "step": 12233
+    },
+    {
+      "epoch": 0.5175564768592944,
+      "grad_norm": 0.5492585897445679,
+      "learning_rate": 0.001,
+      "loss": 1.8345,
+      "step": 12234
+    },
+    {
+      "epoch": 0.5175987816228107,
+      "grad_norm": 0.4502498209476471,
+      "learning_rate": 0.001,
+      "loss": 2.326,
+      "step": 12235
+    },
+    {
+      "epoch": 0.5176410863863271,
+      "grad_norm": 0.22445209324359894,
+      "learning_rate": 0.001,
+      "loss": 1.5903,
+      "step": 12236
+    },
+    {
+      "epoch": 0.5176833911498435,
+      "grad_norm": 0.24608655273914337,
+      "learning_rate": 0.001,
+      "loss": 3.3088,
+      "step": 12237
+    },
+    {
+      "epoch": 0.5177256959133598,
+      "grad_norm": 0.25341930985450745,
+      "learning_rate": 0.001,
+      "loss": 2.5506,
+      "step": 12238
+    },
+    {
+      "epoch": 0.5177680006768762,
+      "grad_norm": 0.18000063300132751,
+      "learning_rate": 0.001,
+      "loss": 2.4819,
+      "step": 12239
+    },
+    {
+      "epoch": 0.5178103054403926,
+      "grad_norm": 0.21232323348522186,
+      "learning_rate": 0.001,
+      "loss": 2.9278,
+      "step": 12240
+    },
+    {
+      "epoch": 0.5178526102039089,
+      "grad_norm": 0.14666417241096497,
+      "learning_rate": 0.001,
+      "loss": 2.0134,
+      "step": 12241
+    },
+    {
+      "epoch": 0.5178949149674253,
+      "grad_norm": 0.15804670751094818,
+      "learning_rate": 0.001,
+      "loss": 1.7012,
+      "step": 12242
+    },
+    {
+      "epoch": 0.5179372197309418,
+      "grad_norm": 0.25225335359573364,
+      "learning_rate": 0.001,
+      "loss": 2.1588,
+      "step": 12243
+    },
+    {
+      "epoch": 0.517979524494458,
+      "grad_norm": 0.18607112765312195,
+      "learning_rate": 0.001,
+      "loss": 1.8788,
+      "step": 12244
+    },
+    {
+      "epoch": 0.5180218292579745,
+      "grad_norm": 0.20208168029785156,
+      "learning_rate": 0.001,
+      "loss": 1.9131,
+      "step": 12245
+    },
+    {
+      "epoch": 0.5180641340214909,
+      "grad_norm": 1.9069957733154297,
+      "learning_rate": 0.001,
+      "loss": 1.4981,
+      "step": 12246
+    },
+    {
+      "epoch": 0.5181064387850072,
+      "grad_norm": 1.0007692575454712,
+      "learning_rate": 0.001,
+      "loss": 1.9453,
+      "step": 12247
+    },
+    {
+      "epoch": 0.5181487435485236,
+      "grad_norm": 0.1494317203760147,
+      "learning_rate": 0.001,
+      "loss": 1.5441,
+      "step": 12248
+    },
+    {
+      "epoch": 0.5181910483120399,
+      "grad_norm": 0.19041891396045685,
+      "learning_rate": 0.001,
+      "loss": 1.7273,
+      "step": 12249
+    },
+    {
+      "epoch": 0.5182333530755563,
+      "grad_norm": 0.4171985983848572,
+      "learning_rate": 0.001,
+      "loss": 1.4179,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5182756578390727,
+      "grad_norm": 0.19176238775253296,
+      "learning_rate": 0.001,
+      "loss": 2.2115,
+      "step": 12251
+    },
+    {
+      "epoch": 0.518317962602589,
+      "grad_norm": 1.5357853174209595,
+      "learning_rate": 0.001,
+      "loss": 2.3413,
+      "step": 12252
+    },
+    {
+      "epoch": 0.5183602673661054,
+      "grad_norm": 0.19013294577598572,
+      "learning_rate": 0.001,
+      "loss": 2.2362,
+      "step": 12253
+    },
+    {
+      "epoch": 0.5184025721296218,
+      "grad_norm": 0.19564557075500488,
+      "learning_rate": 0.001,
+      "loss": 1.971,
+      "step": 12254
+    },
+    {
+      "epoch": 0.5184448768931381,
+      "grad_norm": 0.21926617622375488,
+      "learning_rate": 0.001,
+      "loss": 2.6159,
+      "step": 12255
+    },
+    {
+      "epoch": 0.5184871816566545,
+      "grad_norm": 0.17380297183990479,
+      "learning_rate": 0.001,
+      "loss": 2.8202,
+      "step": 12256
+    },
+    {
+      "epoch": 0.5185294864201709,
+      "grad_norm": 0.5753722190856934,
+      "learning_rate": 0.001,
+      "loss": 3.7983,
+      "step": 12257
+    },
+    {
+      "epoch": 0.5185717911836872,
+      "grad_norm": 0.25849559903144836,
+      "learning_rate": 0.001,
+      "loss": 2.8161,
+      "step": 12258
+    },
+    {
+      "epoch": 0.5186140959472036,
+      "grad_norm": 0.18519815802574158,
+      "learning_rate": 0.001,
+      "loss": 2.1676,
+      "step": 12259
+    },
+    {
+      "epoch": 0.5186564007107201,
+      "grad_norm": 0.19013738632202148,
+      "learning_rate": 0.001,
+      "loss": 1.5707,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5186987054742364,
+      "grad_norm": 0.17344141006469727,
+      "learning_rate": 0.001,
+      "loss": 2.1209,
+      "step": 12261
+    },
+    {
+      "epoch": 0.5187410102377528,
+      "grad_norm": 1.2750663757324219,
+      "learning_rate": 0.001,
+      "loss": 1.9932,
+      "step": 12262
+    },
+    {
+      "epoch": 0.5187833150012692,
+      "grad_norm": 0.16777728497982025,
+      "learning_rate": 0.001,
+      "loss": 1.5308,
+      "step": 12263
+    },
+    {
+      "epoch": 0.5188256197647855,
+      "grad_norm": 0.1650819033384323,
+      "learning_rate": 0.001,
+      "loss": 2.7328,
+      "step": 12264
+    },
+    {
+      "epoch": 0.5188679245283019,
+      "grad_norm": 0.14448492228984833,
+      "learning_rate": 0.001,
+      "loss": 2.4757,
+      "step": 12265
+    },
+    {
+      "epoch": 0.5189102292918183,
+      "grad_norm": 0.18940666317939758,
+      "learning_rate": 0.001,
+      "loss": 1.5308,
+      "step": 12266
+    },
+    {
+      "epoch": 0.5189525340553346,
+      "grad_norm": 0.19672076404094696,
+      "learning_rate": 0.001,
+      "loss": 1.9485,
+      "step": 12267
+    },
+    {
+      "epoch": 0.518994838818851,
+      "grad_norm": 0.16662020981311798,
+      "learning_rate": 0.001,
+      "loss": 1.9421,
+      "step": 12268
+    },
+    {
+      "epoch": 0.5190371435823674,
+      "grad_norm": 0.20337186753749847,
+      "learning_rate": 0.001,
+      "loss": 2.6949,
+      "step": 12269
+    },
+    {
+      "epoch": 0.5190794483458837,
+      "grad_norm": 3.0193428993225098,
+      "learning_rate": 0.001,
+      "loss": 1.9591,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5191217531094001,
+      "grad_norm": 0.2497682422399521,
+      "learning_rate": 0.001,
+      "loss": 2.3117,
+      "step": 12271
+    },
+    {
+      "epoch": 0.5191640578729165,
+      "grad_norm": 0.16508491337299347,
+      "learning_rate": 0.001,
+      "loss": 2.6591,
+      "step": 12272
+    },
+    {
+      "epoch": 0.5192063626364328,
+      "grad_norm": 0.17598336935043335,
+      "learning_rate": 0.001,
+      "loss": 1.634,
+      "step": 12273
+    },
+    {
+      "epoch": 0.5192486673999492,
+      "grad_norm": 0.22913537919521332,
+      "learning_rate": 0.001,
+      "loss": 1.7489,
+      "step": 12274
+    },
+    {
+      "epoch": 0.5192909721634656,
+      "grad_norm": 0.19798220694065094,
+      "learning_rate": 0.001,
+      "loss": 1.8771,
+      "step": 12275
+    },
+    {
+      "epoch": 0.519333276926982,
+      "grad_norm": 0.181406632065773,
+      "learning_rate": 0.001,
+      "loss": 1.9688,
+      "step": 12276
+    },
+    {
+      "epoch": 0.5193755816904984,
+      "grad_norm": 0.19419100880622864,
+      "learning_rate": 0.001,
+      "loss": 3.1894,
+      "step": 12277
+    },
+    {
+      "epoch": 0.5194178864540148,
+      "grad_norm": 0.17426498234272003,
+      "learning_rate": 0.001,
+      "loss": 2.2271,
+      "step": 12278
+    },
+    {
+      "epoch": 0.5194601912175311,
+      "grad_norm": 0.2602224349975586,
+      "learning_rate": 0.001,
+      "loss": 1.8576,
+      "step": 12279
+    },
+    {
+      "epoch": 0.5195024959810475,
+      "grad_norm": 0.16941271722316742,
+      "learning_rate": 0.001,
+      "loss": 2.652,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5195448007445639,
+      "grad_norm": 2.4823379516601562,
+      "learning_rate": 0.001,
+      "loss": 3.1498,
+      "step": 12281
+    },
+    {
+      "epoch": 0.5195871055080802,
+      "grad_norm": 0.16280913352966309,
+      "learning_rate": 0.001,
+      "loss": 2.2317,
+      "step": 12282
+    },
+    {
+      "epoch": 0.5196294102715966,
+      "grad_norm": 0.2174593210220337,
+      "learning_rate": 0.001,
+      "loss": 2.4013,
+      "step": 12283
+    },
+    {
+      "epoch": 0.519671715035113,
+      "grad_norm": 0.2575055956840515,
+      "learning_rate": 0.001,
+      "loss": 3.4478,
+      "step": 12284
+    },
+    {
+      "epoch": 0.5197140197986293,
+      "grad_norm": 0.18891672790050507,
+      "learning_rate": 0.001,
+      "loss": 3.3149,
+      "step": 12285
+    },
+    {
+      "epoch": 0.5197563245621457,
+      "grad_norm": 0.17338892817497253,
+      "learning_rate": 0.001,
+      "loss": 1.7199,
+      "step": 12286
+    },
+    {
+      "epoch": 0.5197986293256621,
+      "grad_norm": 0.4391571283340454,
+      "learning_rate": 0.001,
+      "loss": 1.8649,
+      "step": 12287
+    },
+    {
+      "epoch": 0.5198409340891784,
+      "grad_norm": 0.17235347628593445,
+      "learning_rate": 0.001,
+      "loss": 2.0456,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5198832388526948,
+      "grad_norm": 0.2375982701778412,
+      "learning_rate": 0.001,
+      "loss": 3.4219,
+      "step": 12289
+    },
+    {
+      "epoch": 0.5199255436162112,
+      "grad_norm": 0.201025128364563,
+      "learning_rate": 0.001,
+      "loss": 1.732,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5199678483797275,
+      "grad_norm": 0.3689859211444855,
+      "learning_rate": 0.001,
+      "loss": 3.6554,
+      "step": 12291
+    },
+    {
+      "epoch": 0.520010153143244,
+      "grad_norm": 0.4115980267524719,
+      "learning_rate": 0.001,
+      "loss": 3.1115,
+      "step": 12292
+    },
+    {
+      "epoch": 0.5200524579067602,
+      "grad_norm": 0.20673608779907227,
+      "learning_rate": 0.001,
+      "loss": 2.0847,
+      "step": 12293
+    },
+    {
+      "epoch": 0.5200947626702767,
+      "grad_norm": 0.18936006724834442,
+      "learning_rate": 0.001,
+      "loss": 1.8575,
+      "step": 12294
+    },
+    {
+      "epoch": 0.5201370674337931,
+      "grad_norm": 0.18387524783611298,
+      "learning_rate": 0.001,
+      "loss": 2.73,
+      "step": 12295
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.1719491332769394,
+      "learning_rate": 0.001,
+      "loss": 2.1344,
+      "step": 12296
+    },
+    {
+      "epoch": 0.5202216769608258,
+      "grad_norm": 0.18453410267829895,
+      "learning_rate": 0.001,
+      "loss": 1.9734,
+      "step": 12297
+    },
+    {
+      "epoch": 0.5202639817243422,
+      "grad_norm": 0.1657096892595291,
+      "learning_rate": 0.001,
+      "loss": 2.1985,
+      "step": 12298
+    },
+    {
+      "epoch": 0.5203062864878585,
+      "grad_norm": 0.16189830005168915,
+      "learning_rate": 0.001,
+      "loss": 1.5063,
+      "step": 12299
+    },
+    {
+      "epoch": 0.5203485912513749,
+      "grad_norm": 1.7175111770629883,
+      "learning_rate": 0.001,
+      "loss": 2.564,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5203908960148913,
+      "grad_norm": 0.5387628674507141,
+      "learning_rate": 0.001,
+      "loss": 2.2738,
+      "step": 12301
+    },
+    {
+      "epoch": 0.5204332007784076,
+      "grad_norm": 0.16023603081703186,
+      "learning_rate": 0.001,
+      "loss": 2.5196,
+      "step": 12302
+    },
+    {
+      "epoch": 0.520475505541924,
+      "grad_norm": 0.21019423007965088,
+      "learning_rate": 0.001,
+      "loss": 1.763,
+      "step": 12303
+    },
+    {
+      "epoch": 0.5205178103054404,
+      "grad_norm": 0.5185931921005249,
+      "learning_rate": 0.001,
+      "loss": 1.9254,
+      "step": 12304
+    },
+    {
+      "epoch": 0.5205601150689567,
+      "grad_norm": 0.16839724779129028,
+      "learning_rate": 0.001,
+      "loss": 1.5031,
+      "step": 12305
+    },
+    {
+      "epoch": 0.5206024198324731,
+      "grad_norm": 0.22643537819385529,
+      "learning_rate": 0.001,
+      "loss": 2.4948,
+      "step": 12306
+    },
+    {
+      "epoch": 0.5206447245959895,
+      "grad_norm": 0.40977147221565247,
+      "learning_rate": 0.001,
+      "loss": 2.2601,
+      "step": 12307
+    },
+    {
+      "epoch": 0.5206870293595058,
+      "grad_norm": 0.1929093897342682,
+      "learning_rate": 0.001,
+      "loss": 1.6832,
+      "step": 12308
+    },
+    {
+      "epoch": 0.5207293341230222,
+      "grad_norm": 0.1705029159784317,
+      "learning_rate": 0.001,
+      "loss": 2.3835,
+      "step": 12309
+    },
+    {
+      "epoch": 0.5207716388865387,
+      "grad_norm": 0.17721432447433472,
+      "learning_rate": 0.001,
+      "loss": 2.0197,
+      "step": 12310
+    },
+    {
+      "epoch": 0.520813943650055,
+      "grad_norm": 0.17962828278541565,
+      "learning_rate": 0.001,
+      "loss": 2.0358,
+      "step": 12311
+    },
+    {
+      "epoch": 0.5208562484135714,
+      "grad_norm": 0.17918899655342102,
+      "learning_rate": 0.001,
+      "loss": 2.6547,
+      "step": 12312
+    },
+    {
+      "epoch": 0.5208985531770878,
+      "grad_norm": 0.18718057870864868,
+      "learning_rate": 0.001,
+      "loss": 2.2771,
+      "step": 12313
+    },
+    {
+      "epoch": 0.5209408579406041,
+      "grad_norm": 0.21974588930606842,
+      "learning_rate": 0.001,
+      "loss": 1.9895,
+      "step": 12314
+    },
+    {
+      "epoch": 0.5209831627041205,
+      "grad_norm": 0.2035195678472519,
+      "learning_rate": 0.001,
+      "loss": 2.0884,
+      "step": 12315
+    },
+    {
+      "epoch": 0.5210254674676369,
+      "grad_norm": 0.8761468529701233,
+      "learning_rate": 0.001,
+      "loss": 3.3274,
+      "step": 12316
+    },
+    {
+      "epoch": 0.5210677722311532,
+      "grad_norm": 0.20555326342582703,
+      "learning_rate": 0.001,
+      "loss": 2.0817,
+      "step": 12317
+    },
+    {
+      "epoch": 0.5211100769946696,
+      "grad_norm": 0.18637920916080475,
+      "learning_rate": 0.001,
+      "loss": 2.6589,
+      "step": 12318
+    },
+    {
+      "epoch": 0.521152381758186,
+      "grad_norm": 0.1818656623363495,
+      "learning_rate": 0.001,
+      "loss": 2.0427,
+      "step": 12319
+    },
+    {
+      "epoch": 0.5211946865217023,
+      "grad_norm": 2.5001955032348633,
+      "learning_rate": 0.001,
+      "loss": 1.729,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5212369912852187,
+      "grad_norm": 0.3406553566455841,
+      "learning_rate": 0.001,
+      "loss": 1.5588,
+      "step": 12321
+    },
+    {
+      "epoch": 0.5212792960487351,
+      "grad_norm": 0.2558114528656006,
+      "learning_rate": 0.001,
+      "loss": 2.0768,
+      "step": 12322
+    },
+    {
+      "epoch": 0.5213216008122514,
+      "grad_norm": 7.105432033538818,
+      "learning_rate": 0.001,
+      "loss": 2.6181,
+      "step": 12323
+    },
+    {
+      "epoch": 0.5213639055757678,
+      "grad_norm": 0.18078719079494476,
+      "learning_rate": 0.001,
+      "loss": 2.1346,
+      "step": 12324
+    },
+    {
+      "epoch": 0.5214062103392842,
+      "grad_norm": 0.2121170312166214,
+      "learning_rate": 0.001,
+      "loss": 2.2811,
+      "step": 12325
+    },
+    {
+      "epoch": 0.5214485151028005,
+      "grad_norm": 0.21996238827705383,
+      "learning_rate": 0.001,
+      "loss": 2.6754,
+      "step": 12326
+    },
+    {
+      "epoch": 0.521490819866317,
+      "grad_norm": 0.1870933473110199,
+      "learning_rate": 0.001,
+      "loss": 2.6883,
+      "step": 12327
+    },
+    {
+      "epoch": 0.5215331246298334,
+      "grad_norm": 0.2098461240530014,
+      "learning_rate": 0.001,
+      "loss": 2.1624,
+      "step": 12328
+    },
+    {
+      "epoch": 0.5215754293933497,
+      "grad_norm": 0.1886197030544281,
+      "learning_rate": 0.001,
+      "loss": 1.961,
+      "step": 12329
+    },
+    {
+      "epoch": 0.5216177341568661,
+      "grad_norm": 0.2189781665802002,
+      "learning_rate": 0.001,
+      "loss": 2.2462,
+      "step": 12330
+    },
+    {
+      "epoch": 0.5216600389203825,
+      "grad_norm": 0.1763230264186859,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 12331
+    },
+    {
+      "epoch": 0.5217023436838988,
+      "grad_norm": 1.2748078107833862,
+      "learning_rate": 0.001,
+      "loss": 2.7384,
+      "step": 12332
+    },
+    {
+      "epoch": 0.5217446484474152,
+      "grad_norm": 0.18424801528453827,
+      "learning_rate": 0.001,
+      "loss": 2.4212,
+      "step": 12333
+    },
+    {
+      "epoch": 0.5217869532109316,
+      "grad_norm": 0.8379740715026855,
+      "learning_rate": 0.001,
+      "loss": 2.8512,
+      "step": 12334
+    },
+    {
+      "epoch": 0.5218292579744479,
+      "grad_norm": 0.19673675298690796,
+      "learning_rate": 0.001,
+      "loss": 1.5325,
+      "step": 12335
+    },
+    {
+      "epoch": 0.5218715627379643,
+      "grad_norm": 1.2396926879882812,
+      "learning_rate": 0.001,
+      "loss": 3.2144,
+      "step": 12336
+    },
+    {
+      "epoch": 0.5219138675014807,
+      "grad_norm": 0.1718030571937561,
+      "learning_rate": 0.001,
+      "loss": 2.4473,
+      "step": 12337
+    },
+    {
+      "epoch": 0.521956172264997,
+      "grad_norm": 0.1788032203912735,
+      "learning_rate": 0.001,
+      "loss": 3.1653,
+      "step": 12338
+    },
+    {
+      "epoch": 0.5219984770285134,
+      "grad_norm": 0.17052021622657776,
+      "learning_rate": 0.001,
+      "loss": 3.2152,
+      "step": 12339
+    },
+    {
+      "epoch": 0.5220407817920297,
+      "grad_norm": 2.6638343334198,
+      "learning_rate": 0.001,
+      "loss": 2.5477,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5220830865555461,
+      "grad_norm": 0.20238353312015533,
+      "learning_rate": 0.001,
+      "loss": 2.1136,
+      "step": 12341
+    },
+    {
+      "epoch": 0.5221253913190625,
+      "grad_norm": 0.17800480127334595,
+      "learning_rate": 0.001,
+      "loss": 1.9329,
+      "step": 12342
+    },
+    {
+      "epoch": 0.5221676960825788,
+      "grad_norm": 0.1630057841539383,
+      "learning_rate": 0.001,
+      "loss": 2.7349,
+      "step": 12343
+    },
+    {
+      "epoch": 0.5222100008460953,
+      "grad_norm": 0.20257359743118286,
+      "learning_rate": 0.001,
+      "loss": 2.3163,
+      "step": 12344
+    },
+    {
+      "epoch": 0.5222523056096117,
+      "grad_norm": 0.2002708911895752,
+      "learning_rate": 0.001,
+      "loss": 1.8127,
+      "step": 12345
+    },
+    {
+      "epoch": 0.522294610373128,
+      "grad_norm": 0.6013284921646118,
+      "learning_rate": 0.001,
+      "loss": 1.9079,
+      "step": 12346
+    },
+    {
+      "epoch": 0.5223369151366444,
+      "grad_norm": 0.1917499452829361,
+      "learning_rate": 0.001,
+      "loss": 2.3516,
+      "step": 12347
+    },
+    {
+      "epoch": 0.5223792199001608,
+      "grad_norm": 0.27086690068244934,
+      "learning_rate": 0.001,
+      "loss": 1.9564,
+      "step": 12348
+    },
+    {
+      "epoch": 0.5224215246636771,
+      "grad_norm": 0.22160665690898895,
+      "learning_rate": 0.001,
+      "loss": 2.3388,
+      "step": 12349
+    },
+    {
+      "epoch": 0.5224638294271935,
+      "grad_norm": 0.17383019626140594,
+      "learning_rate": 0.001,
+      "loss": 2.104,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5225061341907099,
+      "grad_norm": 12.725953102111816,
+      "learning_rate": 0.001,
+      "loss": 1.8924,
+      "step": 12351
+    },
+    {
+      "epoch": 0.5225484389542262,
+      "grad_norm": 0.20004230737686157,
+      "learning_rate": 0.001,
+      "loss": 1.8242,
+      "step": 12352
+    },
+    {
+      "epoch": 0.5225907437177426,
+      "grad_norm": 0.20954293012619019,
+      "learning_rate": 0.001,
+      "loss": 2.2262,
+      "step": 12353
+    },
+    {
+      "epoch": 0.522633048481259,
+      "grad_norm": 0.2141389548778534,
+      "learning_rate": 0.001,
+      "loss": 2.7329,
+      "step": 12354
+    },
+    {
+      "epoch": 0.5226753532447753,
+      "grad_norm": 0.21340107917785645,
+      "learning_rate": 0.001,
+      "loss": 2.121,
+      "step": 12355
+    },
+    {
+      "epoch": 0.5227176580082917,
+      "grad_norm": 0.17848850786685944,
+      "learning_rate": 0.001,
+      "loss": 1.7691,
+      "step": 12356
+    },
+    {
+      "epoch": 0.5227599627718081,
+      "grad_norm": 0.21579593420028687,
+      "learning_rate": 0.001,
+      "loss": 4.1183,
+      "step": 12357
+    },
+    {
+      "epoch": 0.5228022675353244,
+      "grad_norm": 0.3494706451892853,
+      "learning_rate": 0.001,
+      "loss": 2.0613,
+      "step": 12358
+    },
+    {
+      "epoch": 0.5228445722988408,
+      "grad_norm": 0.1930229514837265,
+      "learning_rate": 0.001,
+      "loss": 1.9758,
+      "step": 12359
+    },
+    {
+      "epoch": 0.5228868770623573,
+      "grad_norm": 0.18557684123516083,
+      "learning_rate": 0.001,
+      "loss": 1.7799,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5229291818258736,
+      "grad_norm": 0.6549309492111206,
+      "learning_rate": 0.001,
+      "loss": 2.3321,
+      "step": 12361
+    },
+    {
+      "epoch": 0.52297148658939,
+      "grad_norm": 0.4236035645008087,
+      "learning_rate": 0.001,
+      "loss": 1.793,
+      "step": 12362
+    },
+    {
+      "epoch": 0.5230137913529064,
+      "grad_norm": 0.21107986569404602,
+      "learning_rate": 0.001,
+      "loss": 2.1026,
+      "step": 12363
+    },
+    {
+      "epoch": 0.5230560961164227,
+      "grad_norm": 0.2254447042942047,
+      "learning_rate": 0.001,
+      "loss": 3.0301,
+      "step": 12364
+    },
+    {
+      "epoch": 0.5230984008799391,
+      "grad_norm": 0.19465897977352142,
+      "learning_rate": 0.001,
+      "loss": 2.5741,
+      "step": 12365
+    },
+    {
+      "epoch": 0.5231407056434555,
+      "grad_norm": 0.2160634696483612,
+      "learning_rate": 0.001,
+      "loss": 2.5059,
+      "step": 12366
+    },
+    {
+      "epoch": 0.5231830104069718,
+      "grad_norm": 0.20978963375091553,
+      "learning_rate": 0.001,
+      "loss": 2.9964,
+      "step": 12367
+    },
+    {
+      "epoch": 0.5232253151704882,
+      "grad_norm": 0.22652214765548706,
+      "learning_rate": 0.001,
+      "loss": 2.0517,
+      "step": 12368
+    },
+    {
+      "epoch": 0.5232676199340046,
+      "grad_norm": 0.20549549162387848,
+      "learning_rate": 0.001,
+      "loss": 2.668,
+      "step": 12369
+    },
+    {
+      "epoch": 0.5233099246975209,
+      "grad_norm": 0.16967639327049255,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5233522294610373,
+      "grad_norm": 0.21468955278396606,
+      "learning_rate": 0.001,
+      "loss": 2.0383,
+      "step": 12371
+    },
+    {
+      "epoch": 0.5233945342245537,
+      "grad_norm": 0.1573602259159088,
+      "learning_rate": 0.001,
+      "loss": 1.9794,
+      "step": 12372
+    },
+    {
+      "epoch": 0.52343683898807,
+      "grad_norm": 0.8761491775512695,
+      "learning_rate": 0.001,
+      "loss": 2.5732,
+      "step": 12373
+    },
+    {
+      "epoch": 0.5234791437515864,
+      "grad_norm": 0.3122228980064392,
+      "learning_rate": 0.001,
+      "loss": 3.2694,
+      "step": 12374
+    },
+    {
+      "epoch": 0.5235214485151028,
+      "grad_norm": 0.19162549078464508,
+      "learning_rate": 0.001,
+      "loss": 2.2557,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5235637532786191,
+      "grad_norm": 0.22822991013526917,
+      "learning_rate": 0.001,
+      "loss": 2.1365,
+      "step": 12376
+    },
+    {
+      "epoch": 0.5236060580421356,
+      "grad_norm": 0.16961225867271423,
+      "learning_rate": 0.001,
+      "loss": 1.7234,
+      "step": 12377
+    },
+    {
+      "epoch": 0.523648362805652,
+      "grad_norm": 0.1936628669500351,
+      "learning_rate": 0.001,
+      "loss": 1.4112,
+      "step": 12378
+    },
+    {
+      "epoch": 0.5236906675691683,
+      "grad_norm": 0.1944187730550766,
+      "learning_rate": 0.001,
+      "loss": 1.7775,
+      "step": 12379
+    },
+    {
+      "epoch": 0.5237329723326847,
+      "grad_norm": 0.2097603678703308,
+      "learning_rate": 0.001,
+      "loss": 2.2723,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5237752770962011,
+      "grad_norm": 0.4582602381706238,
+      "learning_rate": 0.001,
+      "loss": 2.1067,
+      "step": 12381
+    },
+    {
+      "epoch": 0.5238175818597174,
+      "grad_norm": 0.15353432297706604,
+      "learning_rate": 0.001,
+      "loss": 2.1572,
+      "step": 12382
+    },
+    {
+      "epoch": 0.5238598866232338,
+      "grad_norm": 0.2179485708475113,
+      "learning_rate": 0.001,
+      "loss": 2.1966,
+      "step": 12383
+    },
+    {
+      "epoch": 0.5239021913867501,
+      "grad_norm": 0.2067425549030304,
+      "learning_rate": 0.001,
+      "loss": 1.6399,
+      "step": 12384
+    },
+    {
+      "epoch": 0.5239444961502665,
+      "grad_norm": 0.19505983591079712,
+      "learning_rate": 0.001,
+      "loss": 2.661,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5239868009137829,
+      "grad_norm": 0.17970894277095795,
+      "learning_rate": 0.001,
+      "loss": 1.9452,
+      "step": 12386
+    },
+    {
+      "epoch": 0.5240291056772992,
+      "grad_norm": 1.098179817199707,
+      "learning_rate": 0.001,
+      "loss": 2.2812,
+      "step": 12387
+    },
+    {
+      "epoch": 0.5240714104408156,
+      "grad_norm": 6.329110622406006,
+      "learning_rate": 0.001,
+      "loss": 3.141,
+      "step": 12388
+    },
+    {
+      "epoch": 0.524113715204332,
+      "grad_norm": 0.18318776786327362,
+      "learning_rate": 0.001,
+      "loss": 2.4143,
+      "step": 12389
+    },
+    {
+      "epoch": 0.5241560199678483,
+      "grad_norm": 0.22181005775928497,
+      "learning_rate": 0.001,
+      "loss": 2.1807,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5241983247313647,
+      "grad_norm": 0.17904549837112427,
+      "learning_rate": 0.001,
+      "loss": 1.9226,
+      "step": 12391
+    },
+    {
+      "epoch": 0.5242406294948811,
+      "grad_norm": 0.21022500097751617,
+      "learning_rate": 0.001,
+      "loss": 2.4073,
+      "step": 12392
+    },
+    {
+      "epoch": 0.5242829342583974,
+      "grad_norm": 0.17983204126358032,
+      "learning_rate": 0.001,
+      "loss": 2.1061,
+      "step": 12393
+    },
+    {
+      "epoch": 0.5243252390219139,
+      "grad_norm": 0.29281267523765564,
+      "learning_rate": 0.001,
+      "loss": 3.1756,
+      "step": 12394
+    },
+    {
+      "epoch": 0.5243675437854303,
+      "grad_norm": 1.9023278951644897,
+      "learning_rate": 0.001,
+      "loss": 2.1109,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5244098485489466,
+      "grad_norm": 0.2698245942592621,
+      "learning_rate": 0.001,
+      "loss": 1.7295,
+      "step": 12396
+    },
+    {
+      "epoch": 0.524452153312463,
+      "grad_norm": 0.22318899631500244,
+      "learning_rate": 0.001,
+      "loss": 2.1475,
+      "step": 12397
+    },
+    {
+      "epoch": 0.5244944580759794,
+      "grad_norm": 0.1895742565393448,
+      "learning_rate": 0.001,
+      "loss": 2.3827,
+      "step": 12398
+    },
+    {
+      "epoch": 0.5245367628394957,
+      "grad_norm": 0.22232870757579803,
+      "learning_rate": 0.001,
+      "loss": 2.3418,
+      "step": 12399
+    },
+    {
+      "epoch": 0.5245790676030121,
+      "grad_norm": 0.2738935947418213,
+      "learning_rate": 0.001,
+      "loss": 4.0447,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5246213723665285,
+      "grad_norm": 0.18201978504657745,
+      "learning_rate": 0.001,
+      "loss": 2.4426,
+      "step": 12401
+    },
+    {
+      "epoch": 0.5246636771300448,
+      "grad_norm": 0.26819124817848206,
+      "learning_rate": 0.001,
+      "loss": 2.2201,
+      "step": 12402
+    },
+    {
+      "epoch": 0.5247059818935612,
+      "grad_norm": 0.40238019824028015,
+      "learning_rate": 0.001,
+      "loss": 1.9465,
+      "step": 12403
+    },
+    {
+      "epoch": 0.5247482866570776,
+      "grad_norm": 0.18625056743621826,
+      "learning_rate": 0.001,
+      "loss": 2.6789,
+      "step": 12404
+    },
+    {
+      "epoch": 0.5247905914205939,
+      "grad_norm": 0.19995659589767456,
+      "learning_rate": 0.001,
+      "loss": 1.4283,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5248328961841103,
+      "grad_norm": 0.19528824090957642,
+      "learning_rate": 0.001,
+      "loss": 1.6882,
+      "step": 12406
+    },
+    {
+      "epoch": 0.5248752009476267,
+      "grad_norm": 0.17695745825767517,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 12407
+    },
+    {
+      "epoch": 0.524917505711143,
+      "grad_norm": 0.1534528285264969,
+      "learning_rate": 0.001,
+      "loss": 1.7225,
+      "step": 12408
+    },
+    {
+      "epoch": 0.5249598104746594,
+      "grad_norm": 0.21268439292907715,
+      "learning_rate": 0.001,
+      "loss": 2.0355,
+      "step": 12409
+    },
+    {
+      "epoch": 0.5250021152381759,
+      "grad_norm": 0.18775634467601776,
+      "learning_rate": 0.001,
+      "loss": 2.0029,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5250444200016922,
+      "grad_norm": 1.877071738243103,
+      "learning_rate": 0.001,
+      "loss": 2.0333,
+      "step": 12411
+    },
+    {
+      "epoch": 0.5250867247652086,
+      "grad_norm": 0.19529949128627777,
+      "learning_rate": 0.001,
+      "loss": 2.5546,
+      "step": 12412
+    },
+    {
+      "epoch": 0.525129029528725,
+      "grad_norm": 0.16856685280799866,
+      "learning_rate": 0.001,
+      "loss": 2.258,
+      "step": 12413
+    },
+    {
+      "epoch": 0.5251713342922413,
+      "grad_norm": 0.16465047001838684,
+      "learning_rate": 0.001,
+      "loss": 2.2241,
+      "step": 12414
+    },
+    {
+      "epoch": 0.5252136390557577,
+      "grad_norm": 0.16923588514328003,
+      "learning_rate": 0.001,
+      "loss": 1.7148,
+      "step": 12415
+    },
+    {
+      "epoch": 0.5252559438192741,
+      "grad_norm": 0.15065397322177887,
+      "learning_rate": 0.001,
+      "loss": 1.7362,
+      "step": 12416
+    },
+    {
+      "epoch": 0.5252982485827904,
+      "grad_norm": 0.18638429045677185,
+      "learning_rate": 0.001,
+      "loss": 2.3777,
+      "step": 12417
+    },
+    {
+      "epoch": 0.5253405533463068,
+      "grad_norm": 0.16178961098194122,
+      "learning_rate": 0.001,
+      "loss": 2.0762,
+      "step": 12418
+    },
+    {
+      "epoch": 0.5253828581098232,
+      "grad_norm": 0.1650550216436386,
+      "learning_rate": 0.001,
+      "loss": 2.1552,
+      "step": 12419
+    },
+    {
+      "epoch": 0.5254251628733395,
+      "grad_norm": 1.2916520833969116,
+      "learning_rate": 0.001,
+      "loss": 3.1018,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5254674676368559,
+      "grad_norm": 0.17819340527057648,
+      "learning_rate": 0.001,
+      "loss": 2.2599,
+      "step": 12421
+    },
+    {
+      "epoch": 0.5255097724003723,
+      "grad_norm": 0.18574517965316772,
+      "learning_rate": 0.001,
+      "loss": 2.1839,
+      "step": 12422
+    },
+    {
+      "epoch": 0.5255520771638886,
+      "grad_norm": 0.15550824999809265,
+      "learning_rate": 0.001,
+      "loss": 1.4498,
+      "step": 12423
+    },
+    {
+      "epoch": 0.525594381927405,
+      "grad_norm": 0.2876248061656952,
+      "learning_rate": 0.001,
+      "loss": 2.4418,
+      "step": 12424
+    },
+    {
+      "epoch": 0.5256366866909215,
+      "grad_norm": 0.5879155397415161,
+      "learning_rate": 0.001,
+      "loss": 2.641,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5256789914544377,
+      "grad_norm": 0.29421496391296387,
+      "learning_rate": 0.001,
+      "loss": 3.3343,
+      "step": 12426
+    },
+    {
+      "epoch": 0.5257212962179542,
+      "grad_norm": 0.17267128825187683,
+      "learning_rate": 0.001,
+      "loss": 1.9214,
+      "step": 12427
+    },
+    {
+      "epoch": 0.5257636009814705,
+      "grad_norm": 0.24222494661808014,
+      "learning_rate": 0.001,
+      "loss": 2.4036,
+      "step": 12428
+    },
+    {
+      "epoch": 0.5258059057449869,
+      "grad_norm": 0.171603262424469,
+      "learning_rate": 0.001,
+      "loss": 2.1059,
+      "step": 12429
+    },
+    {
+      "epoch": 0.5258482105085033,
+      "grad_norm": 0.2517464756965637,
+      "learning_rate": 0.001,
+      "loss": 2.3314,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5258905152720196,
+      "grad_norm": 0.19529391825199127,
+      "learning_rate": 0.001,
+      "loss": 2.3075,
+      "step": 12431
+    },
+    {
+      "epoch": 0.525932820035536,
+      "grad_norm": 3.9220962524414062,
+      "learning_rate": 0.001,
+      "loss": 2.0059,
+      "step": 12432
+    },
+    {
+      "epoch": 0.5259751247990524,
+      "grad_norm": 0.2157333493232727,
+      "learning_rate": 0.001,
+      "loss": 2.0792,
+      "step": 12433
+    },
+    {
+      "epoch": 0.5260174295625687,
+      "grad_norm": 0.7744253277778625,
+      "learning_rate": 0.001,
+      "loss": 1.6057,
+      "step": 12434
+    },
+    {
+      "epoch": 0.5260597343260851,
+      "grad_norm": 0.18592911958694458,
+      "learning_rate": 0.001,
+      "loss": 1.5647,
+      "step": 12435
+    },
+    {
+      "epoch": 0.5261020390896015,
+      "grad_norm": 0.15915216505527496,
+      "learning_rate": 0.001,
+      "loss": 1.7973,
+      "step": 12436
+    },
+    {
+      "epoch": 0.5261443438531178,
+      "grad_norm": 26.635772705078125,
+      "learning_rate": 0.001,
+      "loss": 2.6575,
+      "step": 12437
+    },
+    {
+      "epoch": 0.5261866486166342,
+      "grad_norm": 0.17832812666893005,
+      "learning_rate": 0.001,
+      "loss": 1.701,
+      "step": 12438
+    },
+    {
+      "epoch": 0.5262289533801506,
+      "grad_norm": 0.3404991328716278,
+      "learning_rate": 0.001,
+      "loss": 1.7622,
+      "step": 12439
+    },
+    {
+      "epoch": 0.5262712581436669,
+      "grad_norm": 0.17686522006988525,
+      "learning_rate": 0.001,
+      "loss": 1.5594,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5263135629071833,
+      "grad_norm": 0.17130149900913239,
+      "learning_rate": 0.001,
+      "loss": 2.038,
+      "step": 12441
+    },
+    {
+      "epoch": 0.5263558676706998,
+      "grad_norm": 0.7331603169441223,
+      "learning_rate": 0.001,
+      "loss": 1.3103,
+      "step": 12442
+    },
+    {
+      "epoch": 0.526398172434216,
+      "grad_norm": 0.1968250870704651,
+      "learning_rate": 0.001,
+      "loss": 1.8041,
+      "step": 12443
+    },
+    {
+      "epoch": 0.5264404771977325,
+      "grad_norm": 0.2132941633462906,
+      "learning_rate": 0.001,
+      "loss": 1.9928,
+      "step": 12444
+    },
+    {
+      "epoch": 0.5264827819612489,
+      "grad_norm": 0.36532631516456604,
+      "learning_rate": 0.001,
+      "loss": 2.7363,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5265250867247652,
+      "grad_norm": 0.2068646103143692,
+      "learning_rate": 0.001,
+      "loss": 1.6601,
+      "step": 12446
+    },
+    {
+      "epoch": 0.5265673914882816,
+      "grad_norm": 0.20064789056777954,
+      "learning_rate": 0.001,
+      "loss": 2.081,
+      "step": 12447
+    },
+    {
+      "epoch": 0.526609696251798,
+      "grad_norm": 0.23239631950855255,
+      "learning_rate": 0.001,
+      "loss": 1.8456,
+      "step": 12448
+    },
+    {
+      "epoch": 0.5266520010153143,
+      "grad_norm": 0.186894029378891,
+      "learning_rate": 0.001,
+      "loss": 2.5725,
+      "step": 12449
+    },
+    {
+      "epoch": 0.5266943057788307,
+      "grad_norm": 0.16730260848999023,
+      "learning_rate": 0.001,
+      "loss": 1.9904,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5267366105423471,
+      "grad_norm": 0.2060299515724182,
+      "learning_rate": 0.001,
+      "loss": 2.014,
+      "step": 12451
+    },
+    {
+      "epoch": 0.5267789153058634,
+      "grad_norm": 0.5314529538154602,
+      "learning_rate": 0.001,
+      "loss": 2.1433,
+      "step": 12452
+    },
+    {
+      "epoch": 0.5268212200693798,
+      "grad_norm": 0.18238548934459686,
+      "learning_rate": 0.001,
+      "loss": 1.9902,
+      "step": 12453
+    },
+    {
+      "epoch": 0.5268635248328962,
+      "grad_norm": 0.235755056142807,
+      "learning_rate": 0.001,
+      "loss": 1.8839,
+      "step": 12454
+    },
+    {
+      "epoch": 0.5269058295964125,
+      "grad_norm": 0.19867976009845734,
+      "learning_rate": 0.001,
+      "loss": 3.061,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5269481343599289,
+      "grad_norm": 0.1549888551235199,
+      "learning_rate": 0.001,
+      "loss": 2.0597,
+      "step": 12456
+    },
+    {
+      "epoch": 0.5269904391234453,
+      "grad_norm": 0.1728786826133728,
+      "learning_rate": 0.001,
+      "loss": 2.26,
+      "step": 12457
+    },
+    {
+      "epoch": 0.5270327438869616,
+      "grad_norm": 0.20742489397525787,
+      "learning_rate": 0.001,
+      "loss": 2.246,
+      "step": 12458
+    },
+    {
+      "epoch": 0.527075048650478,
+      "grad_norm": 0.15079474449157715,
+      "learning_rate": 0.001,
+      "loss": 1.3225,
+      "step": 12459
+    },
+    {
+      "epoch": 0.5271173534139945,
+      "grad_norm": 0.18330729007720947,
+      "learning_rate": 0.001,
+      "loss": 1.6184,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5271596581775108,
+      "grad_norm": 0.17771340906620026,
+      "learning_rate": 0.001,
+      "loss": 2.4973,
+      "step": 12461
+    },
+    {
+      "epoch": 0.5272019629410272,
+      "grad_norm": 6.8827996253967285,
+      "learning_rate": 0.001,
+      "loss": 2.6216,
+      "step": 12462
+    },
+    {
+      "epoch": 0.5272442677045436,
+      "grad_norm": 0.1736050844192505,
+      "learning_rate": 0.001,
+      "loss": 1.9857,
+      "step": 12463
+    },
+    {
+      "epoch": 0.5272865724680599,
+      "grad_norm": 0.17495718598365784,
+      "learning_rate": 0.001,
+      "loss": 1.9657,
+      "step": 12464
+    },
+    {
+      "epoch": 0.5273288772315763,
+      "grad_norm": 0.16420039534568787,
+      "learning_rate": 0.001,
+      "loss": 1.7354,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5273711819950927,
+      "grad_norm": 0.4817570447921753,
+      "learning_rate": 0.001,
+      "loss": 1.7285,
+      "step": 12466
+    },
+    {
+      "epoch": 0.527413486758609,
+      "grad_norm": 0.17338165640830994,
+      "learning_rate": 0.001,
+      "loss": 2.617,
+      "step": 12467
+    },
+    {
+      "epoch": 0.5274557915221254,
+      "grad_norm": 0.19620858132839203,
+      "learning_rate": 0.001,
+      "loss": 3.0908,
+      "step": 12468
+    },
+    {
+      "epoch": 0.5274980962856418,
+      "grad_norm": 0.2547888457775116,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 12469
+    },
+    {
+      "epoch": 0.5275404010491581,
+      "grad_norm": 0.20097443461418152,
+      "learning_rate": 0.001,
+      "loss": 2.0749,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5275827058126745,
+      "grad_norm": 0.16553282737731934,
+      "learning_rate": 0.001,
+      "loss": 2.3331,
+      "step": 12471
+    },
+    {
+      "epoch": 0.5276250105761909,
+      "grad_norm": 0.3828006982803345,
+      "learning_rate": 0.001,
+      "loss": 1.2995,
+      "step": 12472
+    },
+    {
+      "epoch": 0.5276673153397072,
+      "grad_norm": 0.1714460253715515,
+      "learning_rate": 0.001,
+      "loss": 1.5497,
+      "step": 12473
+    },
+    {
+      "epoch": 0.5277096201032236,
+      "grad_norm": 0.38405609130859375,
+      "learning_rate": 0.001,
+      "loss": 2.9096,
+      "step": 12474
+    },
+    {
+      "epoch": 0.5277519248667399,
+      "grad_norm": 0.19936303794384003,
+      "learning_rate": 0.001,
+      "loss": 1.9383,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5277942296302564,
+      "grad_norm": 0.20733779668807983,
+      "learning_rate": 0.001,
+      "loss": 2.1244,
+      "step": 12476
+    },
+    {
+      "epoch": 0.5278365343937728,
+      "grad_norm": 0.19483071565628052,
+      "learning_rate": 0.001,
+      "loss": 2.6053,
+      "step": 12477
+    },
+    {
+      "epoch": 0.5278788391572891,
+      "grad_norm": 0.1924906224012375,
+      "learning_rate": 0.001,
+      "loss": 2.3413,
+      "step": 12478
+    },
+    {
+      "epoch": 0.5279211439208055,
+      "grad_norm": 0.15672799944877625,
+      "learning_rate": 0.001,
+      "loss": 2.4487,
+      "step": 12479
+    },
+    {
+      "epoch": 0.5279634486843219,
+      "grad_norm": 4.736910820007324,
+      "learning_rate": 0.001,
+      "loss": 1.852,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5280057534478382,
+      "grad_norm": 0.17773562669754028,
+      "learning_rate": 0.001,
+      "loss": 2.2705,
+      "step": 12481
+    },
+    {
+      "epoch": 0.5280480582113546,
+      "grad_norm": 0.17934168875217438,
+      "learning_rate": 0.001,
+      "loss": 2.0574,
+      "step": 12482
+    },
+    {
+      "epoch": 0.528090362974871,
+      "grad_norm": 0.5093991160392761,
+      "learning_rate": 0.001,
+      "loss": 3.1038,
+      "step": 12483
+    },
+    {
+      "epoch": 0.5281326677383873,
+      "grad_norm": 0.18139603734016418,
+      "learning_rate": 0.001,
+      "loss": 1.9695,
+      "step": 12484
+    },
+    {
+      "epoch": 0.5281749725019037,
+      "grad_norm": 0.17985884845256805,
+      "learning_rate": 0.001,
+      "loss": 2.5194,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5282172772654201,
+      "grad_norm": 0.3463243842124939,
+      "learning_rate": 0.001,
+      "loss": 2.4281,
+      "step": 12486
+    },
+    {
+      "epoch": 0.5282595820289364,
+      "grad_norm": 0.21633853018283844,
+      "learning_rate": 0.001,
+      "loss": 3.2592,
+      "step": 12487
+    },
+    {
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.16759130358695984,
+      "learning_rate": 0.001,
+      "loss": 2.6158,
+      "step": 12488
+    },
+    {
+      "epoch": 0.5283441915559692,
+      "grad_norm": 0.20517773926258087,
+      "learning_rate": 0.001,
+      "loss": 1.9567,
+      "step": 12489
+    },
+    {
+      "epoch": 0.5283864963194855,
+      "grad_norm": 1.2722563743591309,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5284288010830019,
+      "grad_norm": 0.21564523875713348,
+      "learning_rate": 0.001,
+      "loss": 2.731,
+      "step": 12491
+    },
+    {
+      "epoch": 0.5284711058465184,
+      "grad_norm": 0.18117859959602356,
+      "learning_rate": 0.001,
+      "loss": 1.7917,
+      "step": 12492
+    },
+    {
+      "epoch": 0.5285134106100347,
+      "grad_norm": 0.18480683863162994,
+      "learning_rate": 0.001,
+      "loss": 1.7639,
+      "step": 12493
+    },
+    {
+      "epoch": 0.5285557153735511,
+      "grad_norm": 0.18521404266357422,
+      "learning_rate": 0.001,
+      "loss": 1.7627,
+      "step": 12494
+    },
+    {
+      "epoch": 0.5285980201370675,
+      "grad_norm": 0.15031592547893524,
+      "learning_rate": 0.001,
+      "loss": 1.4082,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5286403249005838,
+      "grad_norm": 0.2322620302438736,
+      "learning_rate": 0.001,
+      "loss": 2.1222,
+      "step": 12496
+    },
+    {
+      "epoch": 0.5286826296641002,
+      "grad_norm": 0.21710190176963806,
+      "learning_rate": 0.001,
+      "loss": 2.4748,
+      "step": 12497
+    },
+    {
+      "epoch": 0.5287249344276166,
+      "grad_norm": 2.9545514583587646,
+      "learning_rate": 0.001,
+      "loss": 2.2093,
+      "step": 12498
+    },
+    {
+      "epoch": 0.5287672391911329,
+      "grad_norm": 0.20980048179626465,
+      "learning_rate": 0.001,
+      "loss": 2.3927,
+      "step": 12499
+    },
+    {
+      "epoch": 0.5288095439546493,
+      "grad_norm": 0.2512090802192688,
+      "learning_rate": 0.001,
+      "loss": 1.5181,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5288518487181657,
+      "grad_norm": 0.19579851627349854,
+      "learning_rate": 0.001,
+      "loss": 2.7172,
+      "step": 12501
+    },
+    {
+      "epoch": 0.528894153481682,
+      "grad_norm": 0.16972453892230988,
+      "learning_rate": 0.001,
+      "loss": 2.4574,
+      "step": 12502
+    },
+    {
+      "epoch": 0.5289364582451984,
+      "grad_norm": 2.3002378940582275,
+      "learning_rate": 0.001,
+      "loss": 2.8155,
+      "step": 12503
+    },
+    {
+      "epoch": 0.5289787630087148,
+      "grad_norm": 0.15065890550613403,
+      "learning_rate": 0.001,
+      "loss": 1.7237,
+      "step": 12504
+    },
+    {
+      "epoch": 0.5290210677722311,
+      "grad_norm": 0.1834668517112732,
+      "learning_rate": 0.001,
+      "loss": 2.3786,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5290633725357475,
+      "grad_norm": 0.37765857577323914,
+      "learning_rate": 0.001,
+      "loss": 2.7428,
+      "step": 12506
+    },
+    {
+      "epoch": 0.5291056772992639,
+      "grad_norm": 0.27641263604164124,
+      "learning_rate": 0.001,
+      "loss": 2.5696,
+      "step": 12507
+    },
+    {
+      "epoch": 0.5291479820627802,
+      "grad_norm": 0.2046346813440323,
+      "learning_rate": 0.001,
+      "loss": 2.0424,
+      "step": 12508
+    },
+    {
+      "epoch": 0.5291902868262967,
+      "grad_norm": 2.7164547443389893,
+      "learning_rate": 0.001,
+      "loss": 3.0047,
+      "step": 12509
+    },
+    {
+      "epoch": 0.5292325915898131,
+      "grad_norm": 0.24318911135196686,
+      "learning_rate": 0.001,
+      "loss": 2.8722,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5292748963533294,
+      "grad_norm": 0.20161695778369904,
+      "learning_rate": 0.001,
+      "loss": 3.1377,
+      "step": 12511
+    },
+    {
+      "epoch": 0.5293172011168458,
+      "grad_norm": 0.18776634335517883,
+      "learning_rate": 0.001,
+      "loss": 3.193,
+      "step": 12512
+    },
+    {
+      "epoch": 0.5293595058803622,
+      "grad_norm": 0.17002810537815094,
+      "learning_rate": 0.001,
+      "loss": 2.015,
+      "step": 12513
+    },
+    {
+      "epoch": 0.5294018106438785,
+      "grad_norm": 0.16018223762512207,
+      "learning_rate": 0.001,
+      "loss": 2.0451,
+      "step": 12514
+    },
+    {
+      "epoch": 0.5294441154073949,
+      "grad_norm": 0.2891288101673126,
+      "learning_rate": 0.001,
+      "loss": 1.7617,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5294864201709113,
+      "grad_norm": 0.19036170840263367,
+      "learning_rate": 0.001,
+      "loss": 2.1966,
+      "step": 12516
+    },
+    {
+      "epoch": 0.5295287249344276,
+      "grad_norm": 0.16855932772159576,
+      "learning_rate": 0.001,
+      "loss": 3.1857,
+      "step": 12517
+    },
+    {
+      "epoch": 0.529571029697944,
+      "grad_norm": 0.1747637689113617,
+      "learning_rate": 0.001,
+      "loss": 1.8404,
+      "step": 12518
+    },
+    {
+      "epoch": 0.5296133344614603,
+      "grad_norm": 0.572921633720398,
+      "learning_rate": 0.001,
+      "loss": 2.1041,
+      "step": 12519
+    },
+    {
+      "epoch": 0.5296556392249767,
+      "grad_norm": 0.17884446680545807,
+      "learning_rate": 0.001,
+      "loss": 2.4167,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5296979439884931,
+      "grad_norm": 0.15950831770896912,
+      "learning_rate": 0.001,
+      "loss": 1.9228,
+      "step": 12521
+    },
+    {
+      "epoch": 0.5297402487520094,
+      "grad_norm": 0.18645338714122772,
+      "learning_rate": 0.001,
+      "loss": 1.7911,
+      "step": 12522
+    },
+    {
+      "epoch": 0.5297825535155258,
+      "grad_norm": 0.1739381104707718,
+      "learning_rate": 0.001,
+      "loss": 2.4198,
+      "step": 12523
+    },
+    {
+      "epoch": 0.5298248582790422,
+      "grad_norm": 0.2655837833881378,
+      "learning_rate": 0.001,
+      "loss": 3.0543,
+      "step": 12524
+    },
+    {
+      "epoch": 0.5298671630425585,
+      "grad_norm": 0.16320541501045227,
+      "learning_rate": 0.001,
+      "loss": 1.8807,
+      "step": 12525
+    },
+    {
+      "epoch": 0.529909467806075,
+      "grad_norm": 0.6725846529006958,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 12526
+    },
+    {
+      "epoch": 0.5299517725695914,
+      "grad_norm": 0.18599937856197357,
+      "learning_rate": 0.001,
+      "loss": 3.1388,
+      "step": 12527
+    },
+    {
+      "epoch": 0.5299940773331077,
+      "grad_norm": 0.17061804234981537,
+      "learning_rate": 0.001,
+      "loss": 1.4874,
+      "step": 12528
+    },
+    {
+      "epoch": 0.5300363820966241,
+      "grad_norm": 0.15029898285865784,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 12529
+    },
+    {
+      "epoch": 0.5300786868601405,
+      "grad_norm": 0.16090351343154907,
+      "learning_rate": 0.001,
+      "loss": 2.6116,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5301209916236568,
+      "grad_norm": 0.19355669617652893,
+      "learning_rate": 0.001,
+      "loss": 1.8321,
+      "step": 12531
+    },
+    {
+      "epoch": 0.5301632963871732,
+      "grad_norm": 0.15436525642871857,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 12532
+    },
+    {
+      "epoch": 0.5302056011506896,
+      "grad_norm": 0.20881301164627075,
+      "learning_rate": 0.001,
+      "loss": 3.927,
+      "step": 12533
+    },
+    {
+      "epoch": 0.5302479059142059,
+      "grad_norm": 1.6116613149642944,
+      "learning_rate": 0.001,
+      "loss": 2.1104,
+      "step": 12534
+    },
+    {
+      "epoch": 0.5302902106777223,
+      "grad_norm": 0.4252428710460663,
+      "learning_rate": 0.001,
+      "loss": 2.0893,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5303325154412387,
+      "grad_norm": 3.8983938694000244,
+      "learning_rate": 0.001,
+      "loss": 2.5195,
+      "step": 12536
+    },
+    {
+      "epoch": 0.530374820204755,
+      "grad_norm": 0.23319566249847412,
+      "learning_rate": 0.001,
+      "loss": 2.5801,
+      "step": 12537
+    },
+    {
+      "epoch": 0.5304171249682714,
+      "grad_norm": 2.187912940979004,
+      "learning_rate": 0.001,
+      "loss": 1.8214,
+      "step": 12538
+    },
+    {
+      "epoch": 0.5304594297317878,
+      "grad_norm": 0.18814197182655334,
+      "learning_rate": 0.001,
+      "loss": 2.2121,
+      "step": 12539
+    },
+    {
+      "epoch": 0.5305017344953041,
+      "grad_norm": 0.22966532409191132,
+      "learning_rate": 0.001,
+      "loss": 1.7224,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5305440392588205,
+      "grad_norm": 0.9131950736045837,
+      "learning_rate": 0.001,
+      "loss": 1.7284,
+      "step": 12541
+    },
+    {
+      "epoch": 0.530586344022337,
+      "grad_norm": 0.2272689938545227,
+      "learning_rate": 0.001,
+      "loss": 2.2831,
+      "step": 12542
+    },
+    {
+      "epoch": 0.5306286487858533,
+      "grad_norm": 0.23645274341106415,
+      "learning_rate": 0.001,
+      "loss": 1.8332,
+      "step": 12543
+    },
+    {
+      "epoch": 0.5306709535493697,
+      "grad_norm": 0.2579519748687744,
+      "learning_rate": 0.001,
+      "loss": 2.0468,
+      "step": 12544
+    },
+    {
+      "epoch": 0.5307132583128861,
+      "grad_norm": 0.2276112288236618,
+      "learning_rate": 0.001,
+      "loss": 2.3614,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5307555630764024,
+      "grad_norm": 0.23741212487220764,
+      "learning_rate": 0.001,
+      "loss": 2.4542,
+      "step": 12546
+    },
+    {
+      "epoch": 0.5307978678399188,
+      "grad_norm": 0.2526029050350189,
+      "learning_rate": 0.001,
+      "loss": 2.5269,
+      "step": 12547
+    },
+    {
+      "epoch": 0.5308401726034352,
+      "grad_norm": 0.20093029737472534,
+      "learning_rate": 0.001,
+      "loss": 1.6386,
+      "step": 12548
+    },
+    {
+      "epoch": 0.5308824773669515,
+      "grad_norm": 0.20511755347251892,
+      "learning_rate": 0.001,
+      "loss": 2.6388,
+      "step": 12549
+    },
+    {
+      "epoch": 0.5309247821304679,
+      "grad_norm": 0.2155929058790207,
+      "learning_rate": 0.001,
+      "loss": 2.2045,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5309670868939843,
+      "grad_norm": 0.20179694890975952,
+      "learning_rate": 0.001,
+      "loss": 2.0887,
+      "step": 12551
+    },
+    {
+      "epoch": 0.5310093916575006,
+      "grad_norm": 0.7059963941574097,
+      "learning_rate": 0.001,
+      "loss": 1.7561,
+      "step": 12552
+    },
+    {
+      "epoch": 0.531051696421017,
+      "grad_norm": 0.2252509593963623,
+      "learning_rate": 0.001,
+      "loss": 3.4456,
+      "step": 12553
+    },
+    {
+      "epoch": 0.5310940011845334,
+      "grad_norm": 0.15286624431610107,
+      "learning_rate": 0.001,
+      "loss": 2.5038,
+      "step": 12554
+    },
+    {
+      "epoch": 0.5311363059480497,
+      "grad_norm": 0.17845967411994934,
+      "learning_rate": 0.001,
+      "loss": 1.9147,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5311786107115661,
+      "grad_norm": 0.2328542023897171,
+      "learning_rate": 0.001,
+      "loss": 2.0514,
+      "step": 12556
+    },
+    {
+      "epoch": 0.5312209154750825,
+      "grad_norm": 0.661899745464325,
+      "learning_rate": 0.001,
+      "loss": 2.1049,
+      "step": 12557
+    },
+    {
+      "epoch": 0.5312632202385988,
+      "grad_norm": 0.5505738258361816,
+      "learning_rate": 0.001,
+      "loss": 2.063,
+      "step": 12558
+    },
+    {
+      "epoch": 0.5313055250021153,
+      "grad_norm": 0.14693965017795563,
+      "learning_rate": 0.001,
+      "loss": 2.2266,
+      "step": 12559
+    },
+    {
+      "epoch": 0.5313478297656317,
+      "grad_norm": 0.14494706690311432,
+      "learning_rate": 0.001,
+      "loss": 2.0662,
+      "step": 12560
+    },
+    {
+      "epoch": 0.531390134529148,
+      "grad_norm": 0.4453897178173065,
+      "learning_rate": 0.001,
+      "loss": 2.1152,
+      "step": 12561
+    },
+    {
+      "epoch": 0.5314324392926644,
+      "grad_norm": 0.23559346795082092,
+      "learning_rate": 0.001,
+      "loss": 1.9954,
+      "step": 12562
+    },
+    {
+      "epoch": 0.5314747440561808,
+      "grad_norm": 0.17883579432964325,
+      "learning_rate": 0.001,
+      "loss": 2.1284,
+      "step": 12563
+    },
+    {
+      "epoch": 0.5315170488196971,
+      "grad_norm": 0.16119042038917542,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 12564
+    },
+    {
+      "epoch": 0.5315593535832135,
+      "grad_norm": 0.16220824420452118,
+      "learning_rate": 0.001,
+      "loss": 1.5702,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5316016583467298,
+      "grad_norm": 0.8208723068237305,
+      "learning_rate": 0.001,
+      "loss": 2.2647,
+      "step": 12566
+    },
+    {
+      "epoch": 0.5316439631102462,
+      "grad_norm": 0.17101448774337769,
+      "learning_rate": 0.001,
+      "loss": 1.8805,
+      "step": 12567
+    },
+    {
+      "epoch": 0.5316862678737626,
+      "grad_norm": 0.18567869067192078,
+      "learning_rate": 0.001,
+      "loss": 1.8957,
+      "step": 12568
+    },
+    {
+      "epoch": 0.5317285726372789,
+      "grad_norm": 0.15984505414962769,
+      "learning_rate": 0.001,
+      "loss": 2.0752,
+      "step": 12569
+    },
+    {
+      "epoch": 0.5317708774007953,
+      "grad_norm": 0.1836562305688858,
+      "learning_rate": 0.001,
+      "loss": 2.0664,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5318131821643117,
+      "grad_norm": 0.2100599855184555,
+      "learning_rate": 0.001,
+      "loss": 3.0159,
+      "step": 12571
+    },
+    {
+      "epoch": 0.531855486927828,
+      "grad_norm": 0.16221459209918976,
+      "learning_rate": 0.001,
+      "loss": 1.8475,
+      "step": 12572
+    },
+    {
+      "epoch": 0.5318977916913444,
+      "grad_norm": 0.18409012258052826,
+      "learning_rate": 0.001,
+      "loss": 1.9189,
+      "step": 12573
+    },
+    {
+      "epoch": 0.5319400964548608,
+      "grad_norm": 0.2355443686246872,
+      "learning_rate": 0.001,
+      "loss": 2.4434,
+      "step": 12574
+    },
+    {
+      "epoch": 0.5319824012183771,
+      "grad_norm": 1.7131577730178833,
+      "learning_rate": 0.001,
+      "loss": 2.3271,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5320247059818936,
+      "grad_norm": 0.16390523314476013,
+      "learning_rate": 0.001,
+      "loss": 3.1412,
+      "step": 12576
+    },
+    {
+      "epoch": 0.53206701074541,
+      "grad_norm": 0.17767739295959473,
+      "learning_rate": 0.001,
+      "loss": 3.1423,
+      "step": 12577
+    },
+    {
+      "epoch": 0.5321093155089263,
+      "grad_norm": 0.28065043687820435,
+      "learning_rate": 0.001,
+      "loss": 1.9327,
+      "step": 12578
+    },
+    {
+      "epoch": 0.5321516202724427,
+      "grad_norm": 0.15590040385723114,
+      "learning_rate": 0.001,
+      "loss": 3.3144,
+      "step": 12579
+    },
+    {
+      "epoch": 0.5321939250359591,
+      "grad_norm": 0.18487408757209778,
+      "learning_rate": 0.001,
+      "loss": 2.1503,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5322362297994754,
+      "grad_norm": 0.4396058917045593,
+      "learning_rate": 0.001,
+      "loss": 2.4607,
+      "step": 12581
+    },
+    {
+      "epoch": 0.5322785345629918,
+      "grad_norm": 1.2496229410171509,
+      "learning_rate": 0.001,
+      "loss": 2.3387,
+      "step": 12582
+    },
+    {
+      "epoch": 0.5323208393265082,
+      "grad_norm": 0.19142375886440277,
+      "learning_rate": 0.001,
+      "loss": 1.9137,
+      "step": 12583
+    },
+    {
+      "epoch": 0.5323631440900245,
+      "grad_norm": 1.4793001413345337,
+      "learning_rate": 0.001,
+      "loss": 2.1602,
+      "step": 12584
+    },
+    {
+      "epoch": 0.5324054488535409,
+      "grad_norm": 0.4164004921913147,
+      "learning_rate": 0.001,
+      "loss": 2.9867,
+      "step": 12585
+    },
+    {
+      "epoch": 0.5324477536170573,
+      "grad_norm": 0.17239639163017273,
+      "learning_rate": 0.001,
+      "loss": 2.5442,
+      "step": 12586
+    },
+    {
+      "epoch": 0.5324900583805736,
+      "grad_norm": 0.19600702822208405,
+      "learning_rate": 0.001,
+      "loss": 1.9799,
+      "step": 12587
+    },
+    {
+      "epoch": 0.53253236314409,
+      "grad_norm": 0.1782451719045639,
+      "learning_rate": 0.001,
+      "loss": 2.5812,
+      "step": 12588
+    },
+    {
+      "epoch": 0.5325746679076064,
+      "grad_norm": 0.18545618653297424,
+      "learning_rate": 0.001,
+      "loss": 2.3101,
+      "step": 12589
+    },
+    {
+      "epoch": 0.5326169726711227,
+      "grad_norm": 0.21294978260993958,
+      "learning_rate": 0.001,
+      "loss": 1.6784,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5326592774346391,
+      "grad_norm": 0.16218768060207367,
+      "learning_rate": 0.001,
+      "loss": 2.438,
+      "step": 12591
+    },
+    {
+      "epoch": 0.5327015821981556,
+      "grad_norm": 1.644671082496643,
+      "learning_rate": 0.001,
+      "loss": 2.0642,
+      "step": 12592
+    },
+    {
+      "epoch": 0.5327438869616719,
+      "grad_norm": 0.2141799032688141,
+      "learning_rate": 0.001,
+      "loss": 3.1325,
+      "step": 12593
+    },
+    {
+      "epoch": 0.5327861917251883,
+      "grad_norm": 0.17733773589134216,
+      "learning_rate": 0.001,
+      "loss": 2.3464,
+      "step": 12594
+    },
+    {
+      "epoch": 0.5328284964887047,
+      "grad_norm": 0.1840038150548935,
+      "learning_rate": 0.001,
+      "loss": 1.9205,
+      "step": 12595
+    },
+    {
+      "epoch": 0.532870801252221,
+      "grad_norm": 0.3170630931854248,
+      "learning_rate": 0.001,
+      "loss": 2.405,
+      "step": 12596
+    },
+    {
+      "epoch": 0.5329131060157374,
+      "grad_norm": 0.24258072674274445,
+      "learning_rate": 0.001,
+      "loss": 1.855,
+      "step": 12597
+    },
+    {
+      "epoch": 0.5329554107792538,
+      "grad_norm": 0.9366814494132996,
+      "learning_rate": 0.001,
+      "loss": 2.0033,
+      "step": 12598
+    },
+    {
+      "epoch": 0.5329977155427701,
+      "grad_norm": 0.8258916139602661,
+      "learning_rate": 0.001,
+      "loss": 2.5754,
+      "step": 12599
+    },
+    {
+      "epoch": 0.5330400203062865,
+      "grad_norm": 1.7923530340194702,
+      "learning_rate": 0.001,
+      "loss": 2.3026,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5330823250698029,
+      "grad_norm": 0.5301435589790344,
+      "learning_rate": 0.001,
+      "loss": 2.2184,
+      "step": 12601
+    },
+    {
+      "epoch": 0.5331246298333192,
+      "grad_norm": 0.1785147339105606,
+      "learning_rate": 0.001,
+      "loss": 2.0238,
+      "step": 12602
+    },
+    {
+      "epoch": 0.5331669345968356,
+      "grad_norm": 0.15722912549972534,
+      "learning_rate": 0.001,
+      "loss": 1.7873,
+      "step": 12603
+    },
+    {
+      "epoch": 0.533209239360352,
+      "grad_norm": 0.17702187597751617,
+      "learning_rate": 0.001,
+      "loss": 3.0993,
+      "step": 12604
+    },
+    {
+      "epoch": 0.5332515441238683,
+      "grad_norm": 0.21016950905323029,
+      "learning_rate": 0.001,
+      "loss": 2.2791,
+      "step": 12605
+    },
+    {
+      "epoch": 0.5332938488873847,
+      "grad_norm": 0.19180504977703094,
+      "learning_rate": 0.001,
+      "loss": 1.7831,
+      "step": 12606
+    },
+    {
+      "epoch": 0.5333361536509011,
+      "grad_norm": 0.17286181449890137,
+      "learning_rate": 0.001,
+      "loss": 1.94,
+      "step": 12607
+    },
+    {
+      "epoch": 0.5333784584144174,
+      "grad_norm": 0.25898101925849915,
+      "learning_rate": 0.001,
+      "loss": 2.2634,
+      "step": 12608
+    },
+    {
+      "epoch": 0.5334207631779339,
+      "grad_norm": 0.2569015324115753,
+      "learning_rate": 0.001,
+      "loss": 2.2876,
+      "step": 12609
+    },
+    {
+      "epoch": 0.5334630679414502,
+      "grad_norm": 0.32984477281570435,
+      "learning_rate": 0.001,
+      "loss": 2.9021,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5335053727049666,
+      "grad_norm": 0.2177520990371704,
+      "learning_rate": 0.001,
+      "loss": 2.2496,
+      "step": 12611
+    },
+    {
+      "epoch": 0.533547677468483,
+      "grad_norm": 0.18992137908935547,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 12612
+    },
+    {
+      "epoch": 0.5335899822319993,
+      "grad_norm": 0.20471316576004028,
+      "learning_rate": 0.001,
+      "loss": 2.4037,
+      "step": 12613
+    },
+    {
+      "epoch": 0.5336322869955157,
+      "grad_norm": 0.21054227650165558,
+      "learning_rate": 0.001,
+      "loss": 1.9332,
+      "step": 12614
+    },
+    {
+      "epoch": 0.5336745917590321,
+      "grad_norm": 0.20003412663936615,
+      "learning_rate": 0.001,
+      "loss": 2.4779,
+      "step": 12615
+    },
+    {
+      "epoch": 0.5337168965225484,
+      "grad_norm": 0.17985932528972626,
+      "learning_rate": 0.001,
+      "loss": 1.5322,
+      "step": 12616
+    },
+    {
+      "epoch": 0.5337592012860648,
+      "grad_norm": 2.2871792316436768,
+      "learning_rate": 0.001,
+      "loss": 2.4277,
+      "step": 12617
+    },
+    {
+      "epoch": 0.5338015060495812,
+      "grad_norm": 0.16966228187084198,
+      "learning_rate": 0.001,
+      "loss": 2.4963,
+      "step": 12618
+    },
+    {
+      "epoch": 0.5338438108130975,
+      "grad_norm": 0.19583900272846222,
+      "learning_rate": 0.001,
+      "loss": 2.4291,
+      "step": 12619
+    },
+    {
+      "epoch": 0.5338861155766139,
+      "grad_norm": 0.18535350263118744,
+      "learning_rate": 0.001,
+      "loss": 2.1341,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5339284203401303,
+      "grad_norm": 0.8092800378799438,
+      "learning_rate": 0.001,
+      "loss": 2.5207,
+      "step": 12621
+    },
+    {
+      "epoch": 0.5339707251036466,
+      "grad_norm": 0.19231988489627838,
+      "learning_rate": 0.001,
+      "loss": 2.378,
+      "step": 12622
+    },
+    {
+      "epoch": 0.534013029867163,
+      "grad_norm": 0.22486920654773712,
+      "learning_rate": 0.001,
+      "loss": 2.1152,
+      "step": 12623
+    },
+    {
+      "epoch": 0.5340553346306794,
+      "grad_norm": 0.32526126503944397,
+      "learning_rate": 0.001,
+      "loss": 3.1697,
+      "step": 12624
+    },
+    {
+      "epoch": 0.5340976393941957,
+      "grad_norm": 0.20092487335205078,
+      "learning_rate": 0.001,
+      "loss": 2.8525,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5341399441577122,
+      "grad_norm": 0.2700296640396118,
+      "learning_rate": 0.001,
+      "loss": 2.0498,
+      "step": 12626
+    },
+    {
+      "epoch": 0.5341822489212286,
+      "grad_norm": 0.48722365498542786,
+      "learning_rate": 0.001,
+      "loss": 1.8687,
+      "step": 12627
+    },
+    {
+      "epoch": 0.5342245536847449,
+      "grad_norm": 0.16998623311519623,
+      "learning_rate": 0.001,
+      "loss": 2.2888,
+      "step": 12628
+    },
+    {
+      "epoch": 0.5342668584482613,
+      "grad_norm": 0.15864147245883942,
+      "learning_rate": 0.001,
+      "loss": 1.7734,
+      "step": 12629
+    },
+    {
+      "epoch": 0.5343091632117777,
+      "grad_norm": 0.31583741307258606,
+      "learning_rate": 0.001,
+      "loss": 2.4233,
+      "step": 12630
+    },
+    {
+      "epoch": 0.534351467975294,
+      "grad_norm": 0.19196642935276031,
+      "learning_rate": 0.001,
+      "loss": 1.7955,
+      "step": 12631
+    },
+    {
+      "epoch": 0.5343937727388104,
+      "grad_norm": 0.1647808700799942,
+      "learning_rate": 0.001,
+      "loss": 2.5027,
+      "step": 12632
+    },
+    {
+      "epoch": 0.5344360775023268,
+      "grad_norm": 0.5048976540565491,
+      "learning_rate": 0.001,
+      "loss": 2.1264,
+      "step": 12633
+    },
+    {
+      "epoch": 0.5344783822658431,
+      "grad_norm": 0.22664372622966766,
+      "learning_rate": 0.001,
+      "loss": 2.9306,
+      "step": 12634
+    },
+    {
+      "epoch": 0.5345206870293595,
+      "grad_norm": 0.20378440618515015,
+      "learning_rate": 0.001,
+      "loss": 2.8485,
+      "step": 12635
+    },
+    {
+      "epoch": 0.5345629917928759,
+      "grad_norm": 0.4030454158782959,
+      "learning_rate": 0.001,
+      "loss": 2.3177,
+      "step": 12636
+    },
+    {
+      "epoch": 0.5346052965563922,
+      "grad_norm": 0.7921542525291443,
+      "learning_rate": 0.001,
+      "loss": 3.1555,
+      "step": 12637
+    },
+    {
+      "epoch": 0.5346476013199086,
+      "grad_norm": 0.7897359728813171,
+      "learning_rate": 0.001,
+      "loss": 2.6089,
+      "step": 12638
+    },
+    {
+      "epoch": 0.534689906083425,
+      "grad_norm": 0.2273802012205124,
+      "learning_rate": 0.001,
+      "loss": 2.2542,
+      "step": 12639
+    },
+    {
+      "epoch": 0.5347322108469413,
+      "grad_norm": 0.27411505579948425,
+      "learning_rate": 0.001,
+      "loss": 2.8019,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5347745156104577,
+      "grad_norm": 0.26895782351493835,
+      "learning_rate": 0.001,
+      "loss": 2.6258,
+      "step": 12641
+    },
+    {
+      "epoch": 0.5348168203739742,
+      "grad_norm": 0.17039963603019714,
+      "learning_rate": 0.001,
+      "loss": 2.5587,
+      "step": 12642
+    },
+    {
+      "epoch": 0.5348591251374905,
+      "grad_norm": 0.2543392777442932,
+      "learning_rate": 0.001,
+      "loss": 2.3699,
+      "step": 12643
+    },
+    {
+      "epoch": 0.5349014299010069,
+      "grad_norm": 0.19515450298786163,
+      "learning_rate": 0.001,
+      "loss": 2.1001,
+      "step": 12644
+    },
+    {
+      "epoch": 0.5349437346645233,
+      "grad_norm": 0.16606023907661438,
+      "learning_rate": 0.001,
+      "loss": 1.7925,
+      "step": 12645
+    },
+    {
+      "epoch": 0.5349860394280396,
+      "grad_norm": 0.4345617890357971,
+      "learning_rate": 0.001,
+      "loss": 1.9236,
+      "step": 12646
+    },
+    {
+      "epoch": 0.535028344191556,
+      "grad_norm": 0.18390463292598724,
+      "learning_rate": 0.001,
+      "loss": 2.1396,
+      "step": 12647
+    },
+    {
+      "epoch": 0.5350706489550724,
+      "grad_norm": 0.17461326718330383,
+      "learning_rate": 0.001,
+      "loss": 1.7379,
+      "step": 12648
+    },
+    {
+      "epoch": 0.5351129537185887,
+      "grad_norm": 0.17248356342315674,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 12649
+    },
+    {
+      "epoch": 0.5351552584821051,
+      "grad_norm": 0.17978371679782867,
+      "learning_rate": 0.001,
+      "loss": 2.7611,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5351975632456215,
+      "grad_norm": 0.4586908221244812,
+      "learning_rate": 0.001,
+      "loss": 1.4555,
+      "step": 12651
+    },
+    {
+      "epoch": 0.5352398680091378,
+      "grad_norm": 0.233989417552948,
+      "learning_rate": 0.001,
+      "loss": 1.9321,
+      "step": 12652
+    },
+    {
+      "epoch": 0.5352821727726542,
+      "grad_norm": 0.1826629787683487,
+      "learning_rate": 0.001,
+      "loss": 2.3415,
+      "step": 12653
+    },
+    {
+      "epoch": 0.5353244775361705,
+      "grad_norm": 0.15267974138259888,
+      "learning_rate": 0.001,
+      "loss": 1.4848,
+      "step": 12654
+    },
+    {
+      "epoch": 0.5353667822996869,
+      "grad_norm": 4.387269020080566,
+      "learning_rate": 0.001,
+      "loss": 2.716,
+      "step": 12655
+    },
+    {
+      "epoch": 0.5354090870632033,
+      "grad_norm": 0.26923081278800964,
+      "learning_rate": 0.001,
+      "loss": 3.246,
+      "step": 12656
+    },
+    {
+      "epoch": 0.5354513918267196,
+      "grad_norm": 0.15483248233795166,
+      "learning_rate": 0.001,
+      "loss": 1.5195,
+      "step": 12657
+    },
+    {
+      "epoch": 0.535493696590236,
+      "grad_norm": 0.18110813200473785,
+      "learning_rate": 0.001,
+      "loss": 1.7063,
+      "step": 12658
+    },
+    {
+      "epoch": 0.5355360013537525,
+      "grad_norm": 0.1986570507287979,
+      "learning_rate": 0.001,
+      "loss": 2.1137,
+      "step": 12659
+    },
+    {
+      "epoch": 0.5355783061172688,
+      "grad_norm": 0.16241492331027985,
+      "learning_rate": 0.001,
+      "loss": 1.7866,
+      "step": 12660
+    },
+    {
+      "epoch": 0.5356206108807852,
+      "grad_norm": 0.22161731123924255,
+      "learning_rate": 0.001,
+      "loss": 2.3589,
+      "step": 12661
+    },
+    {
+      "epoch": 0.5356629156443016,
+      "grad_norm": 0.17555823922157288,
+      "learning_rate": 0.001,
+      "loss": 1.9163,
+      "step": 12662
+    },
+    {
+      "epoch": 0.5357052204078179,
+      "grad_norm": 0.20892333984375,
+      "learning_rate": 0.001,
+      "loss": 2.7506,
+      "step": 12663
+    },
+    {
+      "epoch": 0.5357475251713343,
+      "grad_norm": 0.1600860357284546,
+      "learning_rate": 0.001,
+      "loss": 1.9907,
+      "step": 12664
+    },
+    {
+      "epoch": 0.5357898299348507,
+      "grad_norm": 0.20017333328723907,
+      "learning_rate": 0.001,
+      "loss": 2.6324,
+      "step": 12665
+    },
+    {
+      "epoch": 0.535832134698367,
+      "grad_norm": 0.18923135101795197,
+      "learning_rate": 0.001,
+      "loss": 2.1908,
+      "step": 12666
+    },
+    {
+      "epoch": 0.5358744394618834,
+      "grad_norm": 0.24769580364227295,
+      "learning_rate": 0.001,
+      "loss": 1.9219,
+      "step": 12667
+    },
+    {
+      "epoch": 0.5359167442253998,
+      "grad_norm": 0.43653157353401184,
+      "learning_rate": 0.001,
+      "loss": 2.1292,
+      "step": 12668
+    },
+    {
+      "epoch": 0.5359590489889161,
+      "grad_norm": 0.18919190764427185,
+      "learning_rate": 0.001,
+      "loss": 2.4091,
+      "step": 12669
+    },
+    {
+      "epoch": 0.5360013537524325,
+      "grad_norm": 16.16994857788086,
+      "learning_rate": 0.001,
+      "loss": 2.1238,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5360436585159489,
+      "grad_norm": 0.16744960844516754,
+      "learning_rate": 0.001,
+      "loss": 2.7491,
+      "step": 12671
+    },
+    {
+      "epoch": 0.5360859632794652,
+      "grad_norm": 0.1780613660812378,
+      "learning_rate": 0.001,
+      "loss": 1.724,
+      "step": 12672
+    },
+    {
+      "epoch": 0.5361282680429816,
+      "grad_norm": 0.16598299145698547,
+      "learning_rate": 0.001,
+      "loss": 1.8861,
+      "step": 12673
+    },
+    {
+      "epoch": 0.536170572806498,
+      "grad_norm": 0.38400644063949585,
+      "learning_rate": 0.001,
+      "loss": 2.247,
+      "step": 12674
+    },
+    {
+      "epoch": 0.5362128775700143,
+      "grad_norm": 0.24128788709640503,
+      "learning_rate": 0.001,
+      "loss": 2.6687,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5362551823335308,
+      "grad_norm": 0.17561288177967072,
+      "learning_rate": 0.001,
+      "loss": 1.8747,
+      "step": 12676
+    },
+    {
+      "epoch": 0.5362974870970472,
+      "grad_norm": 0.2160193771123886,
+      "learning_rate": 0.001,
+      "loss": 1.8482,
+      "step": 12677
+    },
+    {
+      "epoch": 0.5363397918605635,
+      "grad_norm": 0.2257080227136612,
+      "learning_rate": 0.001,
+      "loss": 2.167,
+      "step": 12678
+    },
+    {
+      "epoch": 0.5363820966240799,
+      "grad_norm": 0.16601672768592834,
+      "learning_rate": 0.001,
+      "loss": 1.2827,
+      "step": 12679
+    },
+    {
+      "epoch": 0.5364244013875963,
+      "grad_norm": 0.20976103842258453,
+      "learning_rate": 0.001,
+      "loss": 2.4844,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5364667061511126,
+      "grad_norm": 0.18916988372802734,
+      "learning_rate": 0.001,
+      "loss": 1.8666,
+      "step": 12681
+    },
+    {
+      "epoch": 0.536509010914629,
+      "grad_norm": 0.20311769843101501,
+      "learning_rate": 0.001,
+      "loss": 2.84,
+      "step": 12682
+    },
+    {
+      "epoch": 0.5365513156781454,
+      "grad_norm": 0.21866144239902496,
+      "learning_rate": 0.001,
+      "loss": 2.5776,
+      "step": 12683
+    },
+    {
+      "epoch": 0.5365936204416617,
+      "grad_norm": 0.2240811288356781,
+      "learning_rate": 0.001,
+      "loss": 2.7133,
+      "step": 12684
+    },
+    {
+      "epoch": 0.5366359252051781,
+      "grad_norm": 0.13089242577552795,
+      "learning_rate": 0.001,
+      "loss": 1.3683,
+      "step": 12685
+    },
+    {
+      "epoch": 0.5366782299686945,
+      "grad_norm": 0.16819548606872559,
+      "learning_rate": 0.001,
+      "loss": 2.8725,
+      "step": 12686
+    },
+    {
+      "epoch": 0.5367205347322108,
+      "grad_norm": 0.2232055813074112,
+      "learning_rate": 0.001,
+      "loss": 2.0103,
+      "step": 12687
+    },
+    {
+      "epoch": 0.5367628394957272,
+      "grad_norm": 0.1957560032606125,
+      "learning_rate": 0.001,
+      "loss": 2.4051,
+      "step": 12688
+    },
+    {
+      "epoch": 0.5368051442592436,
+      "grad_norm": 0.18882183730602264,
+      "learning_rate": 0.001,
+      "loss": 2.382,
+      "step": 12689
+    },
+    {
+      "epoch": 0.5368474490227599,
+      "grad_norm": 0.211518332362175,
+      "learning_rate": 0.001,
+      "loss": 2.4219,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5368897537862763,
+      "grad_norm": 0.4028577506542206,
+      "learning_rate": 0.001,
+      "loss": 1.6842,
+      "step": 12691
+    },
+    {
+      "epoch": 0.5369320585497928,
+      "grad_norm": 0.5534846782684326,
+      "learning_rate": 0.001,
+      "loss": 2.8452,
+      "step": 12692
+    },
+    {
+      "epoch": 0.536974363313309,
+      "grad_norm": 0.6162401437759399,
+      "learning_rate": 0.001,
+      "loss": 1.8964,
+      "step": 12693
+    },
+    {
+      "epoch": 0.5370166680768255,
+      "grad_norm": 0.2093658596277237,
+      "learning_rate": 0.001,
+      "loss": 2.2681,
+      "step": 12694
+    },
+    {
+      "epoch": 0.5370589728403419,
+      "grad_norm": 0.17829060554504395,
+      "learning_rate": 0.001,
+      "loss": 1.7986,
+      "step": 12695
+    },
+    {
+      "epoch": 0.5371012776038582,
+      "grad_norm": 0.29665249586105347,
+      "learning_rate": 0.001,
+      "loss": 2.194,
+      "step": 12696
+    },
+    {
+      "epoch": 0.5371435823673746,
+      "grad_norm": 0.19808493554592133,
+      "learning_rate": 0.001,
+      "loss": 2.5451,
+      "step": 12697
+    },
+    {
+      "epoch": 0.537185887130891,
+      "grad_norm": 0.14953912794589996,
+      "learning_rate": 0.001,
+      "loss": 1.6499,
+      "step": 12698
+    },
+    {
+      "epoch": 0.5372281918944073,
+      "grad_norm": 0.25920218229293823,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 12699
+    },
+    {
+      "epoch": 0.5372704966579237,
+      "grad_norm": 0.2867739200592041,
+      "learning_rate": 0.001,
+      "loss": 2.2908,
+      "step": 12700
+    },
+    {
+      "epoch": 0.53731280142144,
+      "grad_norm": 0.17412187159061432,
+      "learning_rate": 0.001,
+      "loss": 1.7965,
+      "step": 12701
+    },
+    {
+      "epoch": 0.5373551061849564,
+      "grad_norm": 0.19530300796031952,
+      "learning_rate": 0.001,
+      "loss": 2.733,
+      "step": 12702
+    },
+    {
+      "epoch": 0.5373974109484728,
+      "grad_norm": 0.8992458581924438,
+      "learning_rate": 0.001,
+      "loss": 1.9335,
+      "step": 12703
+    },
+    {
+      "epoch": 0.5374397157119891,
+      "grad_norm": 0.16157057881355286,
+      "learning_rate": 0.001,
+      "loss": 1.4152,
+      "step": 12704
+    },
+    {
+      "epoch": 0.5374820204755055,
+      "grad_norm": 0.7205559611320496,
+      "learning_rate": 0.001,
+      "loss": 1.7546,
+      "step": 12705
+    },
+    {
+      "epoch": 0.5375243252390219,
+      "grad_norm": 0.16520078480243683,
+      "learning_rate": 0.001,
+      "loss": 1.9752,
+      "step": 12706
+    },
+    {
+      "epoch": 0.5375666300025382,
+      "grad_norm": 1.2817127704620361,
+      "learning_rate": 0.001,
+      "loss": 2.4836,
+      "step": 12707
+    },
+    {
+      "epoch": 0.5376089347660546,
+      "grad_norm": 0.576266348361969,
+      "learning_rate": 0.001,
+      "loss": 1.898,
+      "step": 12708
+    },
+    {
+      "epoch": 0.5376512395295711,
+      "grad_norm": 0.25358226895332336,
+      "learning_rate": 0.001,
+      "loss": 2.1221,
+      "step": 12709
+    },
+    {
+      "epoch": 0.5376935442930874,
+      "grad_norm": 0.23752950131893158,
+      "learning_rate": 0.001,
+      "loss": 2.4297,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5377358490566038,
+      "grad_norm": 0.17121052742004395,
+      "learning_rate": 0.001,
+      "loss": 2.2977,
+      "step": 12711
+    },
+    {
+      "epoch": 0.5377781538201202,
+      "grad_norm": 5.213639736175537,
+      "learning_rate": 0.001,
+      "loss": 1.5537,
+      "step": 12712
+    },
+    {
+      "epoch": 0.5378204585836365,
+      "grad_norm": 0.15050987899303436,
+      "learning_rate": 0.001,
+      "loss": 2.1932,
+      "step": 12713
+    },
+    {
+      "epoch": 0.5378627633471529,
+      "grad_norm": 0.2271626889705658,
+      "learning_rate": 0.001,
+      "loss": 2.2467,
+      "step": 12714
+    },
+    {
+      "epoch": 0.5379050681106693,
+      "grad_norm": 0.20240074396133423,
+      "learning_rate": 0.001,
+      "loss": 2.7468,
+      "step": 12715
+    },
+    {
+      "epoch": 0.5379473728741856,
+      "grad_norm": 4.134634971618652,
+      "learning_rate": 0.001,
+      "loss": 1.7431,
+      "step": 12716
+    },
+    {
+      "epoch": 0.537989677637702,
+      "grad_norm": 0.23283101618289948,
+      "learning_rate": 0.001,
+      "loss": 1.8593,
+      "step": 12717
+    },
+    {
+      "epoch": 0.5380319824012184,
+      "grad_norm": 0.1952308714389801,
+      "learning_rate": 0.001,
+      "loss": 2.5033,
+      "step": 12718
+    },
+    {
+      "epoch": 0.5380742871647347,
+      "grad_norm": 0.17305073142051697,
+      "learning_rate": 0.001,
+      "loss": 2.3579,
+      "step": 12719
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.18969833850860596,
+      "learning_rate": 0.001,
+      "loss": 1.8591,
+      "step": 12720
+    },
+    {
+      "epoch": 0.5381588966917675,
+      "grad_norm": 0.19060935080051422,
+      "learning_rate": 0.001,
+      "loss": 1.6669,
+      "step": 12721
+    },
+    {
+      "epoch": 0.5382012014552838,
+      "grad_norm": 10.8906888961792,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 12722
+    },
+    {
+      "epoch": 0.5382435062188002,
+      "grad_norm": 0.2497670203447342,
+      "learning_rate": 0.001,
+      "loss": 2.2148,
+      "step": 12723
+    },
+    {
+      "epoch": 0.5382858109823166,
+      "grad_norm": 0.2164415568113327,
+      "learning_rate": 0.001,
+      "loss": 2.146,
+      "step": 12724
+    },
+    {
+      "epoch": 0.538328115745833,
+      "grad_norm": 0.2605139911174774,
+      "learning_rate": 0.001,
+      "loss": 1.9347,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5383704205093494,
+      "grad_norm": 23.029006958007812,
+      "learning_rate": 0.001,
+      "loss": 1.8033,
+      "step": 12726
+    },
+    {
+      "epoch": 0.5384127252728658,
+      "grad_norm": 0.16982685029506683,
+      "learning_rate": 0.001,
+      "loss": 2.4865,
+      "step": 12727
+    },
+    {
+      "epoch": 0.5384550300363821,
+      "grad_norm": 0.525914192199707,
+      "learning_rate": 0.001,
+      "loss": 1.8129,
+      "step": 12728
+    },
+    {
+      "epoch": 0.5384973347998985,
+      "grad_norm": 0.15495622158050537,
+      "learning_rate": 0.001,
+      "loss": 2.5025,
+      "step": 12729
+    },
+    {
+      "epoch": 0.5385396395634149,
+      "grad_norm": 0.19135193526744843,
+      "learning_rate": 0.001,
+      "loss": 1.598,
+      "step": 12730
+    },
+    {
+      "epoch": 0.5385819443269312,
+      "grad_norm": 1.3947174549102783,
+      "learning_rate": 0.001,
+      "loss": 2.7353,
+      "step": 12731
+    },
+    {
+      "epoch": 0.5386242490904476,
+      "grad_norm": 0.16519546508789062,
+      "learning_rate": 0.001,
+      "loss": 1.5702,
+      "step": 12732
+    },
+    {
+      "epoch": 0.538666553853964,
+      "grad_norm": 0.19213086366653442,
+      "learning_rate": 0.001,
+      "loss": 2.4648,
+      "step": 12733
+    },
+    {
+      "epoch": 0.5387088586174803,
+      "grad_norm": 1.1840908527374268,
+      "learning_rate": 0.001,
+      "loss": 2.0265,
+      "step": 12734
+    },
+    {
+      "epoch": 0.5387511633809967,
+      "grad_norm": 0.32493284344673157,
+      "learning_rate": 0.001,
+      "loss": 2.6341,
+      "step": 12735
+    },
+    {
+      "epoch": 0.5387934681445131,
+      "grad_norm": 0.1699162721633911,
+      "learning_rate": 0.001,
+      "loss": 2.3637,
+      "step": 12736
+    },
+    {
+      "epoch": 0.5388357729080294,
+      "grad_norm": 0.40613803267478943,
+      "learning_rate": 0.001,
+      "loss": 1.9464,
+      "step": 12737
+    },
+    {
+      "epoch": 0.5388780776715458,
+      "grad_norm": 0.2300586849451065,
+      "learning_rate": 0.001,
+      "loss": 1.9922,
+      "step": 12738
+    },
+    {
+      "epoch": 0.5389203824350622,
+      "grad_norm": 0.23353518545627594,
+      "learning_rate": 0.001,
+      "loss": 1.4807,
+      "step": 12739
+    },
+    {
+      "epoch": 0.5389626871985785,
+      "grad_norm": 0.20910601317882538,
+      "learning_rate": 0.001,
+      "loss": 2.0366,
+      "step": 12740
+    },
+    {
+      "epoch": 0.539004991962095,
+      "grad_norm": 0.17505596578121185,
+      "learning_rate": 0.001,
+      "loss": 1.9472,
+      "step": 12741
+    },
+    {
+      "epoch": 0.5390472967256114,
+      "grad_norm": 0.2388429343700409,
+      "learning_rate": 0.001,
+      "loss": 1.8028,
+      "step": 12742
+    },
+    {
+      "epoch": 0.5390896014891277,
+      "grad_norm": 0.211119145154953,
+      "learning_rate": 0.001,
+      "loss": 3.0593,
+      "step": 12743
+    },
+    {
+      "epoch": 0.5391319062526441,
+      "grad_norm": 0.1894591599702835,
+      "learning_rate": 0.001,
+      "loss": 3.2891,
+      "step": 12744
+    },
+    {
+      "epoch": 0.5391742110161604,
+      "grad_norm": 0.20445877313613892,
+      "learning_rate": 0.001,
+      "loss": 1.9584,
+      "step": 12745
+    },
+    {
+      "epoch": 0.5392165157796768,
+      "grad_norm": 0.2004939168691635,
+      "learning_rate": 0.001,
+      "loss": 2.1161,
+      "step": 12746
+    },
+    {
+      "epoch": 0.5392588205431932,
+      "grad_norm": 0.20584918558597565,
+      "learning_rate": 0.001,
+      "loss": 2.124,
+      "step": 12747
+    },
+    {
+      "epoch": 0.5393011253067095,
+      "grad_norm": 0.4080961048603058,
+      "learning_rate": 0.001,
+      "loss": 2.1308,
+      "step": 12748
+    },
+    {
+      "epoch": 0.5393434300702259,
+      "grad_norm": 0.261042982339859,
+      "learning_rate": 0.001,
+      "loss": 2.3502,
+      "step": 12749
+    },
+    {
+      "epoch": 0.5393857348337423,
+      "grad_norm": 0.18766193091869354,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5394280395972586,
+      "grad_norm": 0.178992360830307,
+      "learning_rate": 0.001,
+      "loss": 1.9826,
+      "step": 12751
+    },
+    {
+      "epoch": 0.539470344360775,
+      "grad_norm": 0.22362402081489563,
+      "learning_rate": 0.001,
+      "loss": 2.6882,
+      "step": 12752
+    },
+    {
+      "epoch": 0.5395126491242914,
+      "grad_norm": 0.6612523198127747,
+      "learning_rate": 0.001,
+      "loss": 2.3889,
+      "step": 12753
+    },
+    {
+      "epoch": 0.5395549538878077,
+      "grad_norm": 0.24479064345359802,
+      "learning_rate": 0.001,
+      "loss": 2.4542,
+      "step": 12754
+    },
+    {
+      "epoch": 0.5395972586513241,
+      "grad_norm": 0.3169039189815521,
+      "learning_rate": 0.001,
+      "loss": 2.1717,
+      "step": 12755
+    },
+    {
+      "epoch": 0.5396395634148405,
+      "grad_norm": 0.22589461505413055,
+      "learning_rate": 0.001,
+      "loss": 3.304,
+      "step": 12756
+    },
+    {
+      "epoch": 0.5396818681783568,
+      "grad_norm": 0.22644683718681335,
+      "learning_rate": 0.001,
+      "loss": 1.7265,
+      "step": 12757
+    },
+    {
+      "epoch": 0.5397241729418732,
+      "grad_norm": 0.19094765186309814,
+      "learning_rate": 0.001,
+      "loss": 1.681,
+      "step": 12758
+    },
+    {
+      "epoch": 0.5397664777053897,
+      "grad_norm": 0.17020857334136963,
+      "learning_rate": 0.001,
+      "loss": 2.3251,
+      "step": 12759
+    },
+    {
+      "epoch": 0.539808782468906,
+      "grad_norm": 9.065228462219238,
+      "learning_rate": 0.001,
+      "loss": 2.7484,
+      "step": 12760
+    },
+    {
+      "epoch": 0.5398510872324224,
+      "grad_norm": 0.32355996966362,
+      "learning_rate": 0.001,
+      "loss": 2.1063,
+      "step": 12761
+    },
+    {
+      "epoch": 0.5398933919959388,
+      "grad_norm": 0.22074800729751587,
+      "learning_rate": 0.001,
+      "loss": 2.3476,
+      "step": 12762
+    },
+    {
+      "epoch": 0.5399356967594551,
+      "grad_norm": 1.8396928310394287,
+      "learning_rate": 0.001,
+      "loss": 2.3019,
+      "step": 12763
+    },
+    {
+      "epoch": 0.5399780015229715,
+      "grad_norm": 0.18278691172599792,
+      "learning_rate": 0.001,
+      "loss": 2.0778,
+      "step": 12764
+    },
+    {
+      "epoch": 0.5400203062864879,
+      "grad_norm": 0.2510671019554138,
+      "learning_rate": 0.001,
+      "loss": 1.8773,
+      "step": 12765
+    },
+    {
+      "epoch": 0.5400626110500042,
+      "grad_norm": 0.17996838688850403,
+      "learning_rate": 0.001,
+      "loss": 2.517,
+      "step": 12766
+    },
+    {
+      "epoch": 0.5401049158135206,
+      "grad_norm": 0.15889018774032593,
+      "learning_rate": 0.001,
+      "loss": 1.7955,
+      "step": 12767
+    },
+    {
+      "epoch": 0.540147220577037,
+      "grad_norm": 0.18059539794921875,
+      "learning_rate": 0.001,
+      "loss": 2.4617,
+      "step": 12768
+    },
+    {
+      "epoch": 0.5401895253405533,
+      "grad_norm": 1.934308648109436,
+      "learning_rate": 0.001,
+      "loss": 2.4551,
+      "step": 12769
+    },
+    {
+      "epoch": 0.5402318301040697,
+      "grad_norm": 0.764201283454895,
+      "learning_rate": 0.001,
+      "loss": 2.2693,
+      "step": 12770
+    },
+    {
+      "epoch": 0.5402741348675861,
+      "grad_norm": 0.16094563901424408,
+      "learning_rate": 0.001,
+      "loss": 3.8751,
+      "step": 12771
+    },
+    {
+      "epoch": 0.5403164396311024,
+      "grad_norm": 0.20989876985549927,
+      "learning_rate": 0.001,
+      "loss": 2.4648,
+      "step": 12772
+    },
+    {
+      "epoch": 0.5403587443946188,
+      "grad_norm": 0.4561876058578491,
+      "learning_rate": 0.001,
+      "loss": 2.0358,
+      "step": 12773
+    },
+    {
+      "epoch": 0.5404010491581352,
+      "grad_norm": 0.19272848963737488,
+      "learning_rate": 0.001,
+      "loss": 2.3598,
+      "step": 12774
+    },
+    {
+      "epoch": 0.5404433539216515,
+      "grad_norm": 0.1573980748653412,
+      "learning_rate": 0.001,
+      "loss": 3.2602,
+      "step": 12775
+    },
+    {
+      "epoch": 0.540485658685168,
+      "grad_norm": 0.20343995094299316,
+      "learning_rate": 0.001,
+      "loss": 2.9864,
+      "step": 12776
+    },
+    {
+      "epoch": 0.5405279634486844,
+      "grad_norm": 1.2975257635116577,
+      "learning_rate": 0.001,
+      "loss": 2.7595,
+      "step": 12777
+    },
+    {
+      "epoch": 0.5405702682122007,
+      "grad_norm": 0.20765641331672668,
+      "learning_rate": 0.001,
+      "loss": 2.8354,
+      "step": 12778
+    },
+    {
+      "epoch": 0.5406125729757171,
+      "grad_norm": 0.1908862143754959,
+      "learning_rate": 0.001,
+      "loss": 1.833,
+      "step": 12779
+    },
+    {
+      "epoch": 0.5406548777392335,
+      "grad_norm": 0.15055324137210846,
+      "learning_rate": 0.001,
+      "loss": 1.4764,
+      "step": 12780
+    },
+    {
+      "epoch": 0.5406971825027498,
+      "grad_norm": 0.19887371361255646,
+      "learning_rate": 0.001,
+      "loss": 2.6805,
+      "step": 12781
+    },
+    {
+      "epoch": 0.5407394872662662,
+      "grad_norm": 0.19075161218643188,
+      "learning_rate": 0.001,
+      "loss": 2.7236,
+      "step": 12782
+    },
+    {
+      "epoch": 0.5407817920297826,
+      "grad_norm": 0.14533232152462006,
+      "learning_rate": 0.001,
+      "loss": 1.5061,
+      "step": 12783
+    },
+    {
+      "epoch": 0.5408240967932989,
+      "grad_norm": 0.26192668080329895,
+      "learning_rate": 0.001,
+      "loss": 2.8406,
+      "step": 12784
+    },
+    {
+      "epoch": 0.5408664015568153,
+      "grad_norm": 0.1515062302350998,
+      "learning_rate": 0.001,
+      "loss": 2.997,
+      "step": 12785
+    },
+    {
+      "epoch": 0.5409087063203317,
+      "grad_norm": 4.155677795410156,
+      "learning_rate": 0.001,
+      "loss": 2.7793,
+      "step": 12786
+    },
+    {
+      "epoch": 0.540951011083848,
+      "grad_norm": 0.24522151052951813,
+      "learning_rate": 0.001,
+      "loss": 1.8206,
+      "step": 12787
+    },
+    {
+      "epoch": 0.5409933158473644,
+      "grad_norm": 1.2809727191925049,
+      "learning_rate": 0.001,
+      "loss": 3.2282,
+      "step": 12788
+    },
+    {
+      "epoch": 0.5410356206108807,
+      "grad_norm": 0.21048471331596375,
+      "learning_rate": 0.001,
+      "loss": 2.7134,
+      "step": 12789
+    },
+    {
+      "epoch": 0.5410779253743971,
+      "grad_norm": 0.1910332590341568,
+      "learning_rate": 0.001,
+      "loss": 1.9444,
+      "step": 12790
+    },
+    {
+      "epoch": 0.5411202301379135,
+      "grad_norm": 0.20288752019405365,
+      "learning_rate": 0.001,
+      "loss": 1.7156,
+      "step": 12791
+    },
+    {
+      "epoch": 0.5411625349014298,
+      "grad_norm": 0.1752224564552307,
+      "learning_rate": 0.001,
+      "loss": 2.0678,
+      "step": 12792
+    },
+    {
+      "epoch": 0.5412048396649463,
+      "grad_norm": 0.18332380056381226,
+      "learning_rate": 0.001,
+      "loss": 1.6786,
+      "step": 12793
+    },
+    {
+      "epoch": 0.5412471444284627,
+      "grad_norm": 0.16607828438282013,
+      "learning_rate": 0.001,
+      "loss": 2.0846,
+      "step": 12794
+    },
+    {
+      "epoch": 0.541289449191979,
+      "grad_norm": 0.2198660671710968,
+      "learning_rate": 0.001,
+      "loss": 2.083,
+      "step": 12795
+    },
+    {
+      "epoch": 0.5413317539554954,
+      "grad_norm": 0.19361701607704163,
+      "learning_rate": 0.001,
+      "loss": 1.8434,
+      "step": 12796
+    },
+    {
+      "epoch": 0.5413740587190118,
+      "grad_norm": 0.17056864500045776,
+      "learning_rate": 0.001,
+      "loss": 1.7502,
+      "step": 12797
+    },
+    {
+      "epoch": 0.5414163634825281,
+      "grad_norm": 0.2116164267063141,
+      "learning_rate": 0.001,
+      "loss": 2.3516,
+      "step": 12798
+    },
+    {
+      "epoch": 0.5414586682460445,
+      "grad_norm": 0.1802276223897934,
+      "learning_rate": 0.001,
+      "loss": 2.6387,
+      "step": 12799
+    },
+    {
+      "epoch": 0.5415009730095609,
+      "grad_norm": 0.1857595592737198,
+      "learning_rate": 0.001,
+      "loss": 2.052,
+      "step": 12800
+    },
+    {
+      "epoch": 0.5415432777730772,
+      "grad_norm": 0.16181547939777374,
+      "learning_rate": 0.001,
+      "loss": 2.9238,
+      "step": 12801
+    },
+    {
+      "epoch": 0.5415855825365936,
+      "grad_norm": 0.17999279499053955,
+      "learning_rate": 0.001,
+      "loss": 2.1358,
+      "step": 12802
+    },
+    {
+      "epoch": 0.54162788730011,
+      "grad_norm": 0.17646092176437378,
+      "learning_rate": 0.001,
+      "loss": 1.9901,
+      "step": 12803
+    },
+    {
+      "epoch": 0.5416701920636263,
+      "grad_norm": 0.19860303401947021,
+      "learning_rate": 0.001,
+      "loss": 2.5992,
+      "step": 12804
+    },
+    {
+      "epoch": 0.5417124968271427,
+      "grad_norm": 0.43836483359336853,
+      "learning_rate": 0.001,
+      "loss": 2.0902,
+      "step": 12805
+    },
+    {
+      "epoch": 0.5417548015906591,
+      "grad_norm": 0.32996633648872375,
+      "learning_rate": 0.001,
+      "loss": 1.8199,
+      "step": 12806
+    },
+    {
+      "epoch": 0.5417971063541754,
+      "grad_norm": 0.16782180964946747,
+      "learning_rate": 0.001,
+      "loss": 1.8684,
+      "step": 12807
+    },
+    {
+      "epoch": 0.5418394111176918,
+      "grad_norm": 0.29726442694664,
+      "learning_rate": 0.001,
+      "loss": 2.1423,
+      "step": 12808
+    },
+    {
+      "epoch": 0.5418817158812083,
+      "grad_norm": 0.19097116589546204,
+      "learning_rate": 0.001,
+      "loss": 3.4195,
+      "step": 12809
+    },
+    {
+      "epoch": 0.5419240206447246,
+      "grad_norm": 0.41679999232292175,
+      "learning_rate": 0.001,
+      "loss": 1.9064,
+      "step": 12810
+    },
+    {
+      "epoch": 0.541966325408241,
+      "grad_norm": 0.15182755887508392,
+      "learning_rate": 0.001,
+      "loss": 2.4639,
+      "step": 12811
+    },
+    {
+      "epoch": 0.5420086301717574,
+      "grad_norm": 0.16130530834197998,
+      "learning_rate": 0.001,
+      "loss": 1.777,
+      "step": 12812
+    },
+    {
+      "epoch": 0.5420509349352737,
+      "grad_norm": 0.15703986585140228,
+      "learning_rate": 0.001,
+      "loss": 2.4515,
+      "step": 12813
+    },
+    {
+      "epoch": 0.5420932396987901,
+      "grad_norm": 22.553680419921875,
+      "learning_rate": 0.001,
+      "loss": 2.077,
+      "step": 12814
+    },
+    {
+      "epoch": 0.5421355444623065,
+      "grad_norm": 0.17222054302692413,
+      "learning_rate": 0.001,
+      "loss": 1.8372,
+      "step": 12815
+    },
+    {
+      "epoch": 0.5421778492258228,
+      "grad_norm": 0.21091894805431366,
+      "learning_rate": 0.001,
+      "loss": 2.3193,
+      "step": 12816
+    },
+    {
+      "epoch": 0.5422201539893392,
+      "grad_norm": 0.17758257687091827,
+      "learning_rate": 0.001,
+      "loss": 2.2889,
+      "step": 12817
+    },
+    {
+      "epoch": 0.5422624587528556,
+      "grad_norm": 0.2192198932170868,
+      "learning_rate": 0.001,
+      "loss": 2.8897,
+      "step": 12818
+    },
+    {
+      "epoch": 0.5423047635163719,
+      "grad_norm": 0.16406521201133728,
+      "learning_rate": 0.001,
+      "loss": 2.1214,
+      "step": 12819
+    },
+    {
+      "epoch": 0.5423470682798883,
+      "grad_norm": 0.29036909341812134,
+      "learning_rate": 0.001,
+      "loss": 1.6994,
+      "step": 12820
+    },
+    {
+      "epoch": 0.5423893730434047,
+      "grad_norm": 0.16406984627246857,
+      "learning_rate": 0.001,
+      "loss": 2.4504,
+      "step": 12821
+    },
+    {
+      "epoch": 0.542431677806921,
+      "grad_norm": 0.310654878616333,
+      "learning_rate": 0.001,
+      "loss": 2.7693,
+      "step": 12822
+    },
+    {
+      "epoch": 0.5424739825704374,
+      "grad_norm": 0.3856427073478699,
+      "learning_rate": 0.001,
+      "loss": 1.8619,
+      "step": 12823
+    },
+    {
+      "epoch": 0.5425162873339539,
+      "grad_norm": 0.15692728757858276,
+      "learning_rate": 0.001,
+      "loss": 2.5887,
+      "step": 12824
+    },
+    {
+      "epoch": 0.5425585920974701,
+      "grad_norm": 0.1917227804660797,
+      "learning_rate": 0.001,
+      "loss": 2.4362,
+      "step": 12825
+    },
+    {
+      "epoch": 0.5426008968609866,
+      "grad_norm": 0.16880708932876587,
+      "learning_rate": 0.001,
+      "loss": 2.6024,
+      "step": 12826
+    },
+    {
+      "epoch": 0.542643201624503,
+      "grad_norm": 0.22414904832839966,
+      "learning_rate": 0.001,
+      "loss": 1.8351,
+      "step": 12827
+    },
+    {
+      "epoch": 0.5426855063880193,
+      "grad_norm": 0.19142870604991913,
+      "learning_rate": 0.001,
+      "loss": 2.1463,
+      "step": 12828
+    },
+    {
+      "epoch": 0.5427278111515357,
+      "grad_norm": 0.9796238541603088,
+      "learning_rate": 0.001,
+      "loss": 2.4191,
+      "step": 12829
+    },
+    {
+      "epoch": 0.5427701159150521,
+      "grad_norm": 0.18335986137390137,
+      "learning_rate": 0.001,
+      "loss": 2.8576,
+      "step": 12830
+    },
+    {
+      "epoch": 0.5428124206785684,
+      "grad_norm": 0.162856325507164,
+      "learning_rate": 0.001,
+      "loss": 1.473,
+      "step": 12831
+    },
+    {
+      "epoch": 0.5428547254420848,
+      "grad_norm": 0.162824347615242,
+      "learning_rate": 0.001,
+      "loss": 1.9001,
+      "step": 12832
+    },
+    {
+      "epoch": 0.5428970302056012,
+      "grad_norm": 0.1477918028831482,
+      "learning_rate": 0.001,
+      "loss": 1.5069,
+      "step": 12833
+    },
+    {
+      "epoch": 0.5429393349691175,
+      "grad_norm": 0.21265248954296112,
+      "learning_rate": 0.001,
+      "loss": 2.1842,
+      "step": 12834
+    },
+    {
+      "epoch": 0.5429816397326339,
+      "grad_norm": 0.19099631905555725,
+      "learning_rate": 0.001,
+      "loss": 1.1751,
+      "step": 12835
+    },
+    {
+      "epoch": 0.5430239444961502,
+      "grad_norm": 8.667723655700684,
+      "learning_rate": 0.001,
+      "loss": 1.6845,
+      "step": 12836
+    },
+    {
+      "epoch": 0.5430662492596666,
+      "grad_norm": 0.22022154927253723,
+      "learning_rate": 0.001,
+      "loss": 2.5801,
+      "step": 12837
+    },
+    {
+      "epoch": 0.543108554023183,
+      "grad_norm": 0.17374145984649658,
+      "learning_rate": 0.001,
+      "loss": 2.4922,
+      "step": 12838
+    },
+    {
+      "epoch": 0.5431508587866993,
+      "grad_norm": 0.16898944973945618,
+      "learning_rate": 0.001,
+      "loss": 1.8262,
+      "step": 12839
+    },
+    {
+      "epoch": 0.5431931635502157,
+      "grad_norm": 0.18354932963848114,
+      "learning_rate": 0.001,
+      "loss": 2.2251,
+      "step": 12840
+    },
+    {
+      "epoch": 0.5432354683137322,
+      "grad_norm": 0.16196799278259277,
+      "learning_rate": 0.001,
+      "loss": 2.6641,
+      "step": 12841
+    },
+    {
+      "epoch": 0.5432777730772484,
+      "grad_norm": 0.2222384810447693,
+      "learning_rate": 0.001,
+      "loss": 2.1457,
+      "step": 12842
+    },
+    {
+      "epoch": 0.5433200778407649,
+      "grad_norm": 0.19985118508338928,
+      "learning_rate": 0.001,
+      "loss": 2.302,
+      "step": 12843
+    },
+    {
+      "epoch": 0.5433623826042813,
+      "grad_norm": 0.1899694949388504,
+      "learning_rate": 0.001,
+      "loss": 3.4242,
+      "step": 12844
+    },
+    {
+      "epoch": 0.5434046873677976,
+      "grad_norm": 0.19545935094356537,
+      "learning_rate": 0.001,
+      "loss": 1.9181,
+      "step": 12845
+    },
+    {
+      "epoch": 0.543446992131314,
+      "grad_norm": 0.16950981318950653,
+      "learning_rate": 0.001,
+      "loss": 2.4149,
+      "step": 12846
+    },
+    {
+      "epoch": 0.5434892968948304,
+      "grad_norm": 0.19809971749782562,
+      "learning_rate": 0.001,
+      "loss": 1.6052,
+      "step": 12847
+    },
+    {
+      "epoch": 0.5435316016583467,
+      "grad_norm": 0.2023109793663025,
+      "learning_rate": 0.001,
+      "loss": 2.1515,
+      "step": 12848
+    },
+    {
+      "epoch": 0.5435739064218631,
+      "grad_norm": 0.889703631401062,
+      "learning_rate": 0.001,
+      "loss": 2.0461,
+      "step": 12849
+    },
+    {
+      "epoch": 0.5436162111853795,
+      "grad_norm": 0.18343132734298706,
+      "learning_rate": 0.001,
+      "loss": 2.4869,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5436585159488958,
+      "grad_norm": 0.16023565828800201,
+      "learning_rate": 0.001,
+      "loss": 2.1514,
+      "step": 12851
+    },
+    {
+      "epoch": 0.5437008207124122,
+      "grad_norm": 0.16939814388751984,
+      "learning_rate": 0.001,
+      "loss": 1.9151,
+      "step": 12852
+    },
+    {
+      "epoch": 0.5437431254759286,
+      "grad_norm": 0.16399548947811127,
+      "learning_rate": 0.001,
+      "loss": 2.4465,
+      "step": 12853
+    },
+    {
+      "epoch": 0.5437854302394449,
+      "grad_norm": 0.1430288702249527,
+      "learning_rate": 0.001,
+      "loss": 1.7558,
+      "step": 12854
+    },
+    {
+      "epoch": 0.5438277350029613,
+      "grad_norm": 0.5551788210868835,
+      "learning_rate": 0.001,
+      "loss": 1.9211,
+      "step": 12855
+    },
+    {
+      "epoch": 0.5438700397664777,
+      "grad_norm": 0.15507280826568604,
+      "learning_rate": 0.001,
+      "loss": 1.4507,
+      "step": 12856
+    },
+    {
+      "epoch": 0.543912344529994,
+      "grad_norm": 0.27142950892448425,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 12857
+    },
+    {
+      "epoch": 0.5439546492935105,
+      "grad_norm": 0.7331579327583313,
+      "learning_rate": 0.001,
+      "loss": 3.3138,
+      "step": 12858
+    },
+    {
+      "epoch": 0.5439969540570269,
+      "grad_norm": 1.1282596588134766,
+      "learning_rate": 0.001,
+      "loss": 2.4492,
+      "step": 12859
+    },
+    {
+      "epoch": 0.5440392588205432,
+      "grad_norm": 0.1771765947341919,
+      "learning_rate": 0.001,
+      "loss": 2.3039,
+      "step": 12860
+    },
+    {
+      "epoch": 0.5440815635840596,
+      "grad_norm": 0.24000783264636993,
+      "learning_rate": 0.001,
+      "loss": 1.6899,
+      "step": 12861
+    },
+    {
+      "epoch": 0.544123868347576,
+      "grad_norm": 0.23145051300525665,
+      "learning_rate": 0.001,
+      "loss": 2.4085,
+      "step": 12862
+    },
+    {
+      "epoch": 0.5441661731110923,
+      "grad_norm": 0.21521174907684326,
+      "learning_rate": 0.001,
+      "loss": 3.2071,
+      "step": 12863
+    },
+    {
+      "epoch": 0.5442084778746087,
+      "grad_norm": 0.16820016503334045,
+      "learning_rate": 0.001,
+      "loss": 2.267,
+      "step": 12864
+    },
+    {
+      "epoch": 0.5442507826381251,
+      "grad_norm": 0.21583041548728943,
+      "learning_rate": 0.001,
+      "loss": 2.1649,
+      "step": 12865
+    },
+    {
+      "epoch": 0.5442930874016414,
+      "grad_norm": 0.15462155640125275,
+      "learning_rate": 0.001,
+      "loss": 2.2514,
+      "step": 12866
+    },
+    {
+      "epoch": 0.5443353921651578,
+      "grad_norm": 0.30019253492355347,
+      "learning_rate": 0.001,
+      "loss": 2.4805,
+      "step": 12867
+    },
+    {
+      "epoch": 0.5443776969286742,
+      "grad_norm": 0.17040373384952545,
+      "learning_rate": 0.001,
+      "loss": 2.1466,
+      "step": 12868
+    },
+    {
+      "epoch": 0.5444200016921905,
+      "grad_norm": 0.2993798553943634,
+      "learning_rate": 0.001,
+      "loss": 1.5786,
+      "step": 12869
+    },
+    {
+      "epoch": 0.5444623064557069,
+      "grad_norm": 0.17798520624637604,
+      "learning_rate": 0.001,
+      "loss": 1.7204,
+      "step": 12870
+    },
+    {
+      "epoch": 0.5445046112192233,
+      "grad_norm": 0.18030913174152374,
+      "learning_rate": 0.001,
+      "loss": 2.9747,
+      "step": 12871
+    },
+    {
+      "epoch": 0.5445469159827396,
+      "grad_norm": 0.22262489795684814,
+      "learning_rate": 0.001,
+      "loss": 2.4943,
+      "step": 12872
+    },
+    {
+      "epoch": 0.544589220746256,
+      "grad_norm": 0.16953717172145844,
+      "learning_rate": 0.001,
+      "loss": 3.0568,
+      "step": 12873
+    },
+    {
+      "epoch": 0.5446315255097725,
+      "grad_norm": 0.34046247601509094,
+      "learning_rate": 0.001,
+      "loss": 2.8593,
+      "step": 12874
+    },
+    {
+      "epoch": 0.5446738302732888,
+      "grad_norm": 0.1570599377155304,
+      "learning_rate": 0.001,
+      "loss": 2.5897,
+      "step": 12875
+    },
+    {
+      "epoch": 0.5447161350368052,
+      "grad_norm": 0.17176446318626404,
+      "learning_rate": 0.001,
+      "loss": 1.9241,
+      "step": 12876
+    },
+    {
+      "epoch": 0.5447584398003216,
+      "grad_norm": 0.17753027379512787,
+      "learning_rate": 0.001,
+      "loss": 2.161,
+      "step": 12877
+    },
+    {
+      "epoch": 0.5448007445638379,
+      "grad_norm": 0.1912051886320114,
+      "learning_rate": 0.001,
+      "loss": 1.8007,
+      "step": 12878
+    },
+    {
+      "epoch": 0.5448430493273543,
+      "grad_norm": 0.49159884452819824,
+      "learning_rate": 0.001,
+      "loss": 1.7072,
+      "step": 12879
+    },
+    {
+      "epoch": 0.5448853540908706,
+      "grad_norm": 0.23397985100746155,
+      "learning_rate": 0.001,
+      "loss": 1.9658,
+      "step": 12880
+    },
+    {
+      "epoch": 0.544927658854387,
+      "grad_norm": 0.17725947499275208,
+      "learning_rate": 0.001,
+      "loss": 1.8988,
+      "step": 12881
+    },
+    {
+      "epoch": 0.5449699636179034,
+      "grad_norm": 0.14953695237636566,
+      "learning_rate": 0.001,
+      "loss": 2.4378,
+      "step": 12882
+    },
+    {
+      "epoch": 0.5450122683814197,
+      "grad_norm": 0.1612466275691986,
+      "learning_rate": 0.001,
+      "loss": 2.0907,
+      "step": 12883
+    },
+    {
+      "epoch": 0.5450545731449361,
+      "grad_norm": 0.2178923636674881,
+      "learning_rate": 0.001,
+      "loss": 2.9344,
+      "step": 12884
+    },
+    {
+      "epoch": 0.5450968779084525,
+      "grad_norm": 0.19458921253681183,
+      "learning_rate": 0.001,
+      "loss": 2.2585,
+      "step": 12885
+    },
+    {
+      "epoch": 0.5451391826719688,
+      "grad_norm": 0.24837756156921387,
+      "learning_rate": 0.001,
+      "loss": 2.0981,
+      "step": 12886
+    },
+    {
+      "epoch": 0.5451814874354852,
+      "grad_norm": 0.16423627734184265,
+      "learning_rate": 0.001,
+      "loss": 1.4668,
+      "step": 12887
+    },
+    {
+      "epoch": 0.5452237921990016,
+      "grad_norm": 0.17279313504695892,
+      "learning_rate": 0.001,
+      "loss": 1.8586,
+      "step": 12888
+    },
+    {
+      "epoch": 0.5452660969625179,
+      "grad_norm": 0.1528485268354416,
+      "learning_rate": 0.001,
+      "loss": 2.2937,
+      "step": 12889
+    },
+    {
+      "epoch": 0.5453084017260343,
+      "grad_norm": 0.14856481552124023,
+      "learning_rate": 0.001,
+      "loss": 2.3376,
+      "step": 12890
+    },
+    {
+      "epoch": 0.5453507064895508,
+      "grad_norm": 0.1969471573829651,
+      "learning_rate": 0.001,
+      "loss": 2.8098,
+      "step": 12891
+    },
+    {
+      "epoch": 0.545393011253067,
+      "grad_norm": 0.17648574709892273,
+      "learning_rate": 0.001,
+      "loss": 2.3516,
+      "step": 12892
+    },
+    {
+      "epoch": 0.5454353160165835,
+      "grad_norm": 0.2679540812969208,
+      "learning_rate": 0.001,
+      "loss": 2.5151,
+      "step": 12893
+    },
+    {
+      "epoch": 0.5454776207800999,
+      "grad_norm": 0.17323081195354462,
+      "learning_rate": 0.001,
+      "loss": 1.472,
+      "step": 12894
+    },
+    {
+      "epoch": 0.5455199255436162,
+      "grad_norm": 0.7942979335784912,
+      "learning_rate": 0.001,
+      "loss": 2.2294,
+      "step": 12895
+    },
+    {
+      "epoch": 0.5455622303071326,
+      "grad_norm": 0.2997882664203644,
+      "learning_rate": 0.001,
+      "loss": 2.7284,
+      "step": 12896
+    },
+    {
+      "epoch": 0.545604535070649,
+      "grad_norm": 0.1583070456981659,
+      "learning_rate": 0.001,
+      "loss": 2.6989,
+      "step": 12897
+    },
+    {
+      "epoch": 0.5456468398341653,
+      "grad_norm": 0.19625523686408997,
+      "learning_rate": 0.001,
+      "loss": 1.7012,
+      "step": 12898
+    },
+    {
+      "epoch": 0.5456891445976817,
+      "grad_norm": 0.16871203482151031,
+      "learning_rate": 0.001,
+      "loss": 1.7467,
+      "step": 12899
+    },
+    {
+      "epoch": 0.5457314493611981,
+      "grad_norm": 0.17078301310539246,
+      "learning_rate": 0.001,
+      "loss": 2.7648,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5457737541247144,
+      "grad_norm": 0.1832430213689804,
+      "learning_rate": 0.001,
+      "loss": 2.109,
+      "step": 12901
+    },
+    {
+      "epoch": 0.5458160588882308,
+      "grad_norm": 0.18720830976963043,
+      "learning_rate": 0.001,
+      "loss": 2.4266,
+      "step": 12902
+    },
+    {
+      "epoch": 0.5458583636517472,
+      "grad_norm": 0.7275272011756897,
+      "learning_rate": 0.001,
+      "loss": 2.3496,
+      "step": 12903
+    },
+    {
+      "epoch": 0.5459006684152635,
+      "grad_norm": 0.1972072273492813,
+      "learning_rate": 0.001,
+      "loss": 1.8186,
+      "step": 12904
+    },
+    {
+      "epoch": 0.5459429731787799,
+      "grad_norm": 0.17827208340168,
+      "learning_rate": 0.001,
+      "loss": 2.7438,
+      "step": 12905
+    },
+    {
+      "epoch": 0.5459852779422963,
+      "grad_norm": 7.450066089630127,
+      "learning_rate": 0.001,
+      "loss": 2.246,
+      "step": 12906
+    },
+    {
+      "epoch": 0.5460275827058126,
+      "grad_norm": 0.15848639607429504,
+      "learning_rate": 0.001,
+      "loss": 1.7138,
+      "step": 12907
+    },
+    {
+      "epoch": 0.546069887469329,
+      "grad_norm": 0.1964915245771408,
+      "learning_rate": 0.001,
+      "loss": 1.9687,
+      "step": 12908
+    },
+    {
+      "epoch": 0.5461121922328455,
+      "grad_norm": 0.1862536370754242,
+      "learning_rate": 0.001,
+      "loss": 1.9072,
+      "step": 12909
+    },
+    {
+      "epoch": 0.5461544969963618,
+      "grad_norm": 0.18136167526245117,
+      "learning_rate": 0.001,
+      "loss": 1.7342,
+      "step": 12910
+    },
+    {
+      "epoch": 0.5461968017598782,
+      "grad_norm": 0.16859842836856842,
+      "learning_rate": 0.001,
+      "loss": 2.6862,
+      "step": 12911
+    },
+    {
+      "epoch": 0.5462391065233946,
+      "grad_norm": 0.4886167049407959,
+      "learning_rate": 0.001,
+      "loss": 2.0441,
+      "step": 12912
+    },
+    {
+      "epoch": 0.5462814112869109,
+      "grad_norm": 0.25893867015838623,
+      "learning_rate": 0.001,
+      "loss": 3.6045,
+      "step": 12913
+    },
+    {
+      "epoch": 0.5463237160504273,
+      "grad_norm": 0.2573549747467041,
+      "learning_rate": 0.001,
+      "loss": 1.9222,
+      "step": 12914
+    },
+    {
+      "epoch": 0.5463660208139437,
+      "grad_norm": 0.17247262597084045,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 12915
+    },
+    {
+      "epoch": 0.54640832557746,
+      "grad_norm": 2.4677224159240723,
+      "learning_rate": 0.001,
+      "loss": 2.0424,
+      "step": 12916
+    },
+    {
+      "epoch": 0.5464506303409764,
+      "grad_norm": 0.20727023482322693,
+      "learning_rate": 0.001,
+      "loss": 1.9061,
+      "step": 12917
+    },
+    {
+      "epoch": 0.5464929351044928,
+      "grad_norm": 0.1556202620267868,
+      "learning_rate": 0.001,
+      "loss": 2.4613,
+      "step": 12918
+    },
+    {
+      "epoch": 0.5465352398680091,
+      "grad_norm": 0.18038994073867798,
+      "learning_rate": 0.001,
+      "loss": 2.001,
+      "step": 12919
+    },
+    {
+      "epoch": 0.5465775446315255,
+      "grad_norm": 0.20001617074012756,
+      "learning_rate": 0.001,
+      "loss": 2.1099,
+      "step": 12920
+    },
+    {
+      "epoch": 0.5466198493950419,
+      "grad_norm": 0.15489211678504944,
+      "learning_rate": 0.001,
+      "loss": 1.7065,
+      "step": 12921
+    },
+    {
+      "epoch": 0.5466621541585582,
+      "grad_norm": 0.41907620429992676,
+      "learning_rate": 0.001,
+      "loss": 2.915,
+      "step": 12922
+    },
+    {
+      "epoch": 0.5467044589220746,
+      "grad_norm": 0.15193894505500793,
+      "learning_rate": 0.001,
+      "loss": 2.4002,
+      "step": 12923
+    },
+    {
+      "epoch": 0.546746763685591,
+      "grad_norm": 0.16042934358119965,
+      "learning_rate": 0.001,
+      "loss": 1.7455,
+      "step": 12924
+    },
+    {
+      "epoch": 0.5467890684491074,
+      "grad_norm": 8.326644897460938,
+      "learning_rate": 0.001,
+      "loss": 2.3913,
+      "step": 12925
+    },
+    {
+      "epoch": 0.5468313732126238,
+      "grad_norm": 0.22436167299747467,
+      "learning_rate": 0.001,
+      "loss": 2.9849,
+      "step": 12926
+    },
+    {
+      "epoch": 0.5468736779761401,
+      "grad_norm": 0.17266489565372467,
+      "learning_rate": 0.001,
+      "loss": 1.7618,
+      "step": 12927
+    },
+    {
+      "epoch": 0.5469159827396565,
+      "grad_norm": 0.21502335369586945,
+      "learning_rate": 0.001,
+      "loss": 1.632,
+      "step": 12928
+    },
+    {
+      "epoch": 0.5469582875031729,
+      "grad_norm": 0.18782548606395721,
+      "learning_rate": 0.001,
+      "loss": 2.7248,
+      "step": 12929
+    },
+    {
+      "epoch": 0.5470005922666892,
+      "grad_norm": 0.44364723563194275,
+      "learning_rate": 0.001,
+      "loss": 2.0572,
+      "step": 12930
+    },
+    {
+      "epoch": 0.5470428970302056,
+      "grad_norm": 0.22250796854496002,
+      "learning_rate": 0.001,
+      "loss": 1.8936,
+      "step": 12931
+    },
+    {
+      "epoch": 0.547085201793722,
+      "grad_norm": 0.19908872246742249,
+      "learning_rate": 0.001,
+      "loss": 2.5246,
+      "step": 12932
+    },
+    {
+      "epoch": 0.5471275065572383,
+      "grad_norm": 0.15950053930282593,
+      "learning_rate": 0.001,
+      "loss": 2.6949,
+      "step": 12933
+    },
+    {
+      "epoch": 0.5471698113207547,
+      "grad_norm": 0.45760655403137207,
+      "learning_rate": 0.001,
+      "loss": 1.5259,
+      "step": 12934
+    },
+    {
+      "epoch": 0.5472121160842711,
+      "grad_norm": 1.588517427444458,
+      "learning_rate": 0.001,
+      "loss": 3.1214,
+      "step": 12935
+    },
+    {
+      "epoch": 0.5472544208477874,
+      "grad_norm": 0.22262723743915558,
+      "learning_rate": 0.001,
+      "loss": 3.0762,
+      "step": 12936
+    },
+    {
+      "epoch": 0.5472967256113038,
+      "grad_norm": 0.1947306990623474,
+      "learning_rate": 0.001,
+      "loss": 2.2629,
+      "step": 12937
+    },
+    {
+      "epoch": 0.5473390303748202,
+      "grad_norm": 0.31676188111305237,
+      "learning_rate": 0.001,
+      "loss": 1.8793,
+      "step": 12938
+    },
+    {
+      "epoch": 0.5473813351383365,
+      "grad_norm": 0.2185761034488678,
+      "learning_rate": 0.001,
+      "loss": 2.5543,
+      "step": 12939
+    },
+    {
+      "epoch": 0.5474236399018529,
+      "grad_norm": 0.21403570473194122,
+      "learning_rate": 0.001,
+      "loss": 2.5567,
+      "step": 12940
+    },
+    {
+      "epoch": 0.5474659446653694,
+      "grad_norm": 0.23452875018119812,
+      "learning_rate": 0.001,
+      "loss": 2.3212,
+      "step": 12941
+    },
+    {
+      "epoch": 0.5475082494288857,
+      "grad_norm": 0.1822385936975479,
+      "learning_rate": 0.001,
+      "loss": 1.6779,
+      "step": 12942
+    },
+    {
+      "epoch": 0.5475505541924021,
+      "grad_norm": 1.615887999534607,
+      "learning_rate": 0.001,
+      "loss": 1.8927,
+      "step": 12943
+    },
+    {
+      "epoch": 0.5475928589559185,
+      "grad_norm": 0.2374815195798874,
+      "learning_rate": 0.001,
+      "loss": 1.8798,
+      "step": 12944
+    },
+    {
+      "epoch": 0.5476351637194348,
+      "grad_norm": 0.3294655680656433,
+      "learning_rate": 0.001,
+      "loss": 3.5583,
+      "step": 12945
+    },
+    {
+      "epoch": 0.5476774684829512,
+      "grad_norm": 0.17319463193416595,
+      "learning_rate": 0.001,
+      "loss": 2.3341,
+      "step": 12946
+    },
+    {
+      "epoch": 0.5477197732464676,
+      "grad_norm": 0.18420511484146118,
+      "learning_rate": 0.001,
+      "loss": 2.6946,
+      "step": 12947
+    },
+    {
+      "epoch": 0.5477620780099839,
+      "grad_norm": 0.5846876502037048,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 12948
+    },
+    {
+      "epoch": 0.5478043827735003,
+      "grad_norm": 0.23985953629016876,
+      "learning_rate": 0.001,
+      "loss": 2.0719,
+      "step": 12949
+    },
+    {
+      "epoch": 0.5478466875370167,
+      "grad_norm": 0.1828928142786026,
+      "learning_rate": 0.001,
+      "loss": 2.1034,
+      "step": 12950
+    },
+    {
+      "epoch": 0.547888992300533,
+      "grad_norm": 0.2998861074447632,
+      "learning_rate": 0.001,
+      "loss": 1.9081,
+      "step": 12951
+    },
+    {
+      "epoch": 0.5479312970640494,
+      "grad_norm": 0.22397345304489136,
+      "learning_rate": 0.001,
+      "loss": 2.3908,
+      "step": 12952
+    },
+    {
+      "epoch": 0.5479736018275658,
+      "grad_norm": 0.19218890368938446,
+      "learning_rate": 0.001,
+      "loss": 2.7397,
+      "step": 12953
+    },
+    {
+      "epoch": 0.5480159065910821,
+      "grad_norm": 0.1939002424478531,
+      "learning_rate": 0.001,
+      "loss": 2.3556,
+      "step": 12954
+    },
+    {
+      "epoch": 0.5480582113545985,
+      "grad_norm": 0.18285121023654938,
+      "learning_rate": 0.001,
+      "loss": 2.6213,
+      "step": 12955
+    },
+    {
+      "epoch": 0.5481005161181149,
+      "grad_norm": 0.21203845739364624,
+      "learning_rate": 0.001,
+      "loss": 2.7509,
+      "step": 12956
+    },
+    {
+      "epoch": 0.5481428208816312,
+      "grad_norm": 0.306901752948761,
+      "learning_rate": 0.001,
+      "loss": 2.286,
+      "step": 12957
+    },
+    {
+      "epoch": 0.5481851256451477,
+      "grad_norm": 0.18208357691764832,
+      "learning_rate": 0.001,
+      "loss": 2.0693,
+      "step": 12958
+    },
+    {
+      "epoch": 0.5482274304086641,
+      "grad_norm": 0.22899079322814941,
+      "learning_rate": 0.001,
+      "loss": 2.8397,
+      "step": 12959
+    },
+    {
+      "epoch": 0.5482697351721804,
+      "grad_norm": 0.16618044674396515,
+      "learning_rate": 0.001,
+      "loss": 2.9579,
+      "step": 12960
+    },
+    {
+      "epoch": 0.5483120399356968,
+      "grad_norm": 0.17978453636169434,
+      "learning_rate": 0.001,
+      "loss": 1.903,
+      "step": 12961
+    },
+    {
+      "epoch": 0.5483543446992132,
+      "grad_norm": 0.18377989530563354,
+      "learning_rate": 0.001,
+      "loss": 2.4393,
+      "step": 12962
+    },
+    {
+      "epoch": 0.5483966494627295,
+      "grad_norm": 0.21771173179149628,
+      "learning_rate": 0.001,
+      "loss": 3.4764,
+      "step": 12963
+    },
+    {
+      "epoch": 0.5484389542262459,
+      "grad_norm": 0.16839885711669922,
+      "learning_rate": 0.001,
+      "loss": 1.7648,
+      "step": 12964
+    },
+    {
+      "epoch": 0.5484812589897623,
+      "grad_norm": 0.17164848744869232,
+      "learning_rate": 0.001,
+      "loss": 1.8755,
+      "step": 12965
+    },
+    {
+      "epoch": 0.5485235637532786,
+      "grad_norm": 0.25105589628219604,
+      "learning_rate": 0.001,
+      "loss": 2.2976,
+      "step": 12966
+    },
+    {
+      "epoch": 0.548565868516795,
+      "grad_norm": 0.17218393087387085,
+      "learning_rate": 0.001,
+      "loss": 1.842,
+      "step": 12967
+    },
+    {
+      "epoch": 0.5486081732803114,
+      "grad_norm": 0.2835632264614105,
+      "learning_rate": 0.001,
+      "loss": 2.5885,
+      "step": 12968
+    },
+    {
+      "epoch": 0.5486504780438277,
+      "grad_norm": 0.2169799655675888,
+      "learning_rate": 0.001,
+      "loss": 2.1249,
+      "step": 12969
+    },
+    {
+      "epoch": 0.5486927828073441,
+      "grad_norm": 0.28637203574180603,
+      "learning_rate": 0.001,
+      "loss": 1.6262,
+      "step": 12970
+    },
+    {
+      "epoch": 0.5487350875708604,
+      "grad_norm": 0.2180042862892151,
+      "learning_rate": 0.001,
+      "loss": 1.9191,
+      "step": 12971
+    },
+    {
+      "epoch": 0.5487773923343768,
+      "grad_norm": 0.20677872002124786,
+      "learning_rate": 0.001,
+      "loss": 3.2069,
+      "step": 12972
+    },
+    {
+      "epoch": 0.5488196970978932,
+      "grad_norm": 2.7703301906585693,
+      "learning_rate": 0.001,
+      "loss": 3.6456,
+      "step": 12973
+    },
+    {
+      "epoch": 0.5488620018614095,
+      "grad_norm": 0.4277131259441376,
+      "learning_rate": 0.001,
+      "loss": 3.3904,
+      "step": 12974
+    },
+    {
+      "epoch": 0.548904306624926,
+      "grad_norm": 0.19730469584465027,
+      "learning_rate": 0.001,
+      "loss": 2.4932,
+      "step": 12975
+    },
+    {
+      "epoch": 0.5489466113884424,
+      "grad_norm": 7.952031135559082,
+      "learning_rate": 0.001,
+      "loss": 1.7476,
+      "step": 12976
+    },
+    {
+      "epoch": 0.5489889161519587,
+      "grad_norm": 0.3109131157398224,
+      "learning_rate": 0.001,
+      "loss": 2.152,
+      "step": 12977
+    },
+    {
+      "epoch": 0.5490312209154751,
+      "grad_norm": 0.36644262075424194,
+      "learning_rate": 0.001,
+      "loss": 1.6802,
+      "step": 12978
+    },
+    {
+      "epoch": 0.5490735256789915,
+      "grad_norm": 0.24093574285507202,
+      "learning_rate": 0.001,
+      "loss": 1.4931,
+      "step": 12979
+    },
+    {
+      "epoch": 0.5491158304425078,
+      "grad_norm": 0.22954730689525604,
+      "learning_rate": 0.001,
+      "loss": 1.6246,
+      "step": 12980
+    },
+    {
+      "epoch": 0.5491581352060242,
+      "grad_norm": 0.21042773127555847,
+      "learning_rate": 0.001,
+      "loss": 2.835,
+      "step": 12981
+    },
+    {
+      "epoch": 0.5492004399695406,
+      "grad_norm": 0.1865551769733429,
+      "learning_rate": 0.001,
+      "loss": 2.28,
+      "step": 12982
+    },
+    {
+      "epoch": 0.5492427447330569,
+      "grad_norm": 0.15189926326274872,
+      "learning_rate": 0.001,
+      "loss": 1.629,
+      "step": 12983
+    },
+    {
+      "epoch": 0.5492850494965733,
+      "grad_norm": 0.3665443956851959,
+      "learning_rate": 0.001,
+      "loss": 2.1248,
+      "step": 12984
+    },
+    {
+      "epoch": 0.5493273542600897,
+      "grad_norm": 0.13915681838989258,
+      "learning_rate": 0.001,
+      "loss": 3.073,
+      "step": 12985
+    },
+    {
+      "epoch": 0.549369659023606,
+      "grad_norm": 0.1796284019947052,
+      "learning_rate": 0.001,
+      "loss": 1.1191,
+      "step": 12986
+    },
+    {
+      "epoch": 0.5494119637871224,
+      "grad_norm": 0.3686416745185852,
+      "learning_rate": 0.001,
+      "loss": 2.0302,
+      "step": 12987
+    },
+    {
+      "epoch": 0.5494542685506388,
+      "grad_norm": 0.685794472694397,
+      "learning_rate": 0.001,
+      "loss": 2.1283,
+      "step": 12988
+    },
+    {
+      "epoch": 0.5494965733141551,
+      "grad_norm": 0.20203375816345215,
+      "learning_rate": 0.001,
+      "loss": 3.4477,
+      "step": 12989
+    },
+    {
+      "epoch": 0.5495388780776715,
+      "grad_norm": 0.23336085677146912,
+      "learning_rate": 0.001,
+      "loss": 2.3422,
+      "step": 12990
+    },
+    {
+      "epoch": 0.549581182841188,
+      "grad_norm": 0.1661679744720459,
+      "learning_rate": 0.001,
+      "loss": 2.7796,
+      "step": 12991
+    },
+    {
+      "epoch": 0.5496234876047043,
+      "grad_norm": 0.17486825585365295,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 12992
+    },
+    {
+      "epoch": 0.5496657923682207,
+      "grad_norm": 0.1599670946598053,
+      "learning_rate": 0.001,
+      "loss": 2.1133,
+      "step": 12993
+    },
+    {
+      "epoch": 0.5497080971317371,
+      "grad_norm": 0.17942267656326294,
+      "learning_rate": 0.001,
+      "loss": 2.227,
+      "step": 12994
+    },
+    {
+      "epoch": 0.5497504018952534,
+      "grad_norm": 0.18072116374969482,
+      "learning_rate": 0.001,
+      "loss": 1.7973,
+      "step": 12995
+    },
+    {
+      "epoch": 0.5497927066587698,
+      "grad_norm": 0.17400150001049042,
+      "learning_rate": 0.001,
+      "loss": 1.676,
+      "step": 12996
+    },
+    {
+      "epoch": 0.5498350114222862,
+      "grad_norm": 0.2297709584236145,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 12997
+    },
+    {
+      "epoch": 0.5498773161858025,
+      "grad_norm": 0.2160550206899643,
+      "learning_rate": 0.001,
+      "loss": 2.0131,
+      "step": 12998
+    },
+    {
+      "epoch": 0.5499196209493189,
+      "grad_norm": 0.22192339599132538,
+      "learning_rate": 0.001,
+      "loss": 3.1413,
+      "step": 12999
+    },
+    {
+      "epoch": 0.5499619257128353,
+      "grad_norm": 0.2530875504016876,
+      "learning_rate": 0.001,
+      "loss": 2.6736,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5500042304763516,
+      "grad_norm": 0.17102450132369995,
+      "learning_rate": 0.001,
+      "loss": 2.3392,
+      "step": 13001
+    },
+    {
+      "epoch": 0.550046535239868,
+      "grad_norm": 0.17439766228199005,
+      "learning_rate": 0.001,
+      "loss": 2.1533,
+      "step": 13002
+    },
+    {
+      "epoch": 0.5500888400033844,
+      "grad_norm": 0.21379339694976807,
+      "learning_rate": 0.001,
+      "loss": 2.4908,
+      "step": 13003
+    },
+    {
+      "epoch": 0.5501311447669007,
+      "grad_norm": 0.19133144617080688,
+      "learning_rate": 0.001,
+      "loss": 1.8931,
+      "step": 13004
+    },
+    {
+      "epoch": 0.5501734495304171,
+      "grad_norm": 0.1691633015871048,
+      "learning_rate": 0.001,
+      "loss": 1.3374,
+      "step": 13005
+    },
+    {
+      "epoch": 0.5502157542939335,
+      "grad_norm": 0.24116800725460052,
+      "learning_rate": 0.001,
+      "loss": 2.4278,
+      "step": 13006
+    },
+    {
+      "epoch": 0.5502580590574498,
+      "grad_norm": 0.19122330844402313,
+      "learning_rate": 0.001,
+      "loss": 2.0226,
+      "step": 13007
+    },
+    {
+      "epoch": 0.5503003638209663,
+      "grad_norm": 0.15570446848869324,
+      "learning_rate": 0.001,
+      "loss": 1.7573,
+      "step": 13008
+    },
+    {
+      "epoch": 0.5503426685844827,
+      "grad_norm": 0.20239442586898804,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 13009
+    },
+    {
+      "epoch": 0.550384973347999,
+      "grad_norm": 0.1555495709180832,
+      "learning_rate": 0.001,
+      "loss": 3.2426,
+      "step": 13010
+    },
+    {
+      "epoch": 0.5504272781115154,
+      "grad_norm": 0.16427169740200043,
+      "learning_rate": 0.001,
+      "loss": 1.7691,
+      "step": 13011
+    },
+    {
+      "epoch": 0.5504695828750318,
+      "grad_norm": 0.16692587733268738,
+      "learning_rate": 0.001,
+      "loss": 2.0761,
+      "step": 13012
+    },
+    {
+      "epoch": 0.5505118876385481,
+      "grad_norm": 0.1881280094385147,
+      "learning_rate": 0.001,
+      "loss": 2.1607,
+      "step": 13013
+    },
+    {
+      "epoch": 0.5505541924020645,
+      "grad_norm": 0.1543162316083908,
+      "learning_rate": 0.001,
+      "loss": 2.4762,
+      "step": 13014
+    },
+    {
+      "epoch": 0.5505964971655808,
+      "grad_norm": 0.1498258113861084,
+      "learning_rate": 0.001,
+      "loss": 2.1568,
+      "step": 13015
+    },
+    {
+      "epoch": 0.5506388019290972,
+      "grad_norm": 0.15180440247058868,
+      "learning_rate": 0.001,
+      "loss": 2.5071,
+      "step": 13016
+    },
+    {
+      "epoch": 0.5506811066926136,
+      "grad_norm": 0.16765782237052917,
+      "learning_rate": 0.001,
+      "loss": 2.642,
+      "step": 13017
+    },
+    {
+      "epoch": 0.5507234114561299,
+      "grad_norm": 0.16857920587062836,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 13018
+    },
+    {
+      "epoch": 0.5507657162196463,
+      "grad_norm": 0.1566455215215683,
+      "learning_rate": 0.001,
+      "loss": 2.5403,
+      "step": 13019
+    },
+    {
+      "epoch": 0.5508080209831627,
+      "grad_norm": 0.1734788417816162,
+      "learning_rate": 0.001,
+      "loss": 2.1921,
+      "step": 13020
+    },
+    {
+      "epoch": 0.550850325746679,
+      "grad_norm": 0.15811264514923096,
+      "learning_rate": 0.001,
+      "loss": 1.4409,
+      "step": 13021
+    },
+    {
+      "epoch": 0.5508926305101954,
+      "grad_norm": 0.25248953700065613,
+      "learning_rate": 0.001,
+      "loss": 1.9164,
+      "step": 13022
+    },
+    {
+      "epoch": 0.5509349352737118,
+      "grad_norm": 0.19516262412071228,
+      "learning_rate": 0.001,
+      "loss": 2.1415,
+      "step": 13023
+    },
+    {
+      "epoch": 0.5509772400372281,
+      "grad_norm": 0.16840478777885437,
+      "learning_rate": 0.001,
+      "loss": 2.2028,
+      "step": 13024
+    },
+    {
+      "epoch": 0.5510195448007446,
+      "grad_norm": 6.247500896453857,
+      "learning_rate": 0.001,
+      "loss": 2.3796,
+      "step": 13025
+    },
+    {
+      "epoch": 0.551061849564261,
+      "grad_norm": 2.1479499340057373,
+      "learning_rate": 0.001,
+      "loss": 2.4607,
+      "step": 13026
+    },
+    {
+      "epoch": 0.5511041543277773,
+      "grad_norm": 0.1944126933813095,
+      "learning_rate": 0.001,
+      "loss": 1.8729,
+      "step": 13027
+    },
+    {
+      "epoch": 0.5511464590912937,
+      "grad_norm": 0.16638629138469696,
+      "learning_rate": 0.001,
+      "loss": 1.7599,
+      "step": 13028
+    },
+    {
+      "epoch": 0.5511887638548101,
+      "grad_norm": 0.17286452651023865,
+      "learning_rate": 0.001,
+      "loss": 3.7786,
+      "step": 13029
+    },
+    {
+      "epoch": 0.5512310686183264,
+      "grad_norm": 0.18785692751407623,
+      "learning_rate": 0.001,
+      "loss": 1.624,
+      "step": 13030
+    },
+    {
+      "epoch": 0.5512733733818428,
+      "grad_norm": 0.2681181728839874,
+      "learning_rate": 0.001,
+      "loss": 2.4162,
+      "step": 13031
+    },
+    {
+      "epoch": 0.5513156781453592,
+      "grad_norm": 1.382765531539917,
+      "learning_rate": 0.001,
+      "loss": 2.8352,
+      "step": 13032
+    },
+    {
+      "epoch": 0.5513579829088755,
+      "grad_norm": 0.1791725754737854,
+      "learning_rate": 0.001,
+      "loss": 1.5965,
+      "step": 13033
+    },
+    {
+      "epoch": 0.5514002876723919,
+      "grad_norm": 0.22276152670383453,
+      "learning_rate": 0.001,
+      "loss": 2.2998,
+      "step": 13034
+    },
+    {
+      "epoch": 0.5514425924359083,
+      "grad_norm": 0.21398457884788513,
+      "learning_rate": 0.001,
+      "loss": 1.873,
+      "step": 13035
+    },
+    {
+      "epoch": 0.5514848971994246,
+      "grad_norm": 0.19248297810554504,
+      "learning_rate": 0.001,
+      "loss": 2.0772,
+      "step": 13036
+    },
+    {
+      "epoch": 0.551527201962941,
+      "grad_norm": 0.19134873151779175,
+      "learning_rate": 0.001,
+      "loss": 2.2879,
+      "step": 13037
+    },
+    {
+      "epoch": 0.5515695067264574,
+      "grad_norm": 0.2077552080154419,
+      "learning_rate": 0.001,
+      "loss": 2.2834,
+      "step": 13038
+    },
+    {
+      "epoch": 0.5516118114899737,
+      "grad_norm": 0.18088361620903015,
+      "learning_rate": 0.001,
+      "loss": 1.7441,
+      "step": 13039
+    },
+    {
+      "epoch": 0.5516541162534901,
+      "grad_norm": 0.19229553639888763,
+      "learning_rate": 0.001,
+      "loss": 2.3027,
+      "step": 13040
+    },
+    {
+      "epoch": 0.5516964210170066,
+      "grad_norm": 0.20166753232479095,
+      "learning_rate": 0.001,
+      "loss": 2.6727,
+      "step": 13041
+    },
+    {
+      "epoch": 0.5517387257805229,
+      "grad_norm": 0.24196840822696686,
+      "learning_rate": 0.001,
+      "loss": 2.5962,
+      "step": 13042
+    },
+    {
+      "epoch": 0.5517810305440393,
+      "grad_norm": 0.18725115060806274,
+      "learning_rate": 0.001,
+      "loss": 2.1676,
+      "step": 13043
+    },
+    {
+      "epoch": 0.5518233353075557,
+      "grad_norm": 0.20191504061222076,
+      "learning_rate": 0.001,
+      "loss": 2.0362,
+      "step": 13044
+    },
+    {
+      "epoch": 0.551865640071072,
+      "grad_norm": 0.1696188747882843,
+      "learning_rate": 0.001,
+      "loss": 2.5537,
+      "step": 13045
+    },
+    {
+      "epoch": 0.5519079448345884,
+      "grad_norm": 0.28044790029525757,
+      "learning_rate": 0.001,
+      "loss": 1.6491,
+      "step": 13046
+    },
+    {
+      "epoch": 0.5519502495981048,
+      "grad_norm": 0.15965206921100616,
+      "learning_rate": 0.001,
+      "loss": 2.5324,
+      "step": 13047
+    },
+    {
+      "epoch": 0.5519925543616211,
+      "grad_norm": 0.16759942471981049,
+      "learning_rate": 0.001,
+      "loss": 2.6668,
+      "step": 13048
+    },
+    {
+      "epoch": 0.5520348591251375,
+      "grad_norm": 0.16884449124336243,
+      "learning_rate": 0.001,
+      "loss": 1.9035,
+      "step": 13049
+    },
+    {
+      "epoch": 0.5520771638886539,
+      "grad_norm": 0.39007458090782166,
+      "learning_rate": 0.001,
+      "loss": 1.7417,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5521194686521702,
+      "grad_norm": 0.15368978679180145,
+      "learning_rate": 0.001,
+      "loss": 3.4138,
+      "step": 13051
+    },
+    {
+      "epoch": 0.5521617734156866,
+      "grad_norm": 0.443830281496048,
+      "learning_rate": 0.001,
+      "loss": 1.5381,
+      "step": 13052
+    },
+    {
+      "epoch": 0.552204078179203,
+      "grad_norm": 0.29350072145462036,
+      "learning_rate": 0.001,
+      "loss": 2.4967,
+      "step": 13053
+    },
+    {
+      "epoch": 0.5522463829427193,
+      "grad_norm": 0.14899539947509766,
+      "learning_rate": 0.001,
+      "loss": 2.6561,
+      "step": 13054
+    },
+    {
+      "epoch": 0.5522886877062357,
+      "grad_norm": 0.18092837929725647,
+      "learning_rate": 0.001,
+      "loss": 1.9984,
+      "step": 13055
+    },
+    {
+      "epoch": 0.5523309924697521,
+      "grad_norm": 0.21982833743095398,
+      "learning_rate": 0.001,
+      "loss": 3.1229,
+      "step": 13056
+    },
+    {
+      "epoch": 0.5523732972332684,
+      "grad_norm": 1.6211110353469849,
+      "learning_rate": 0.001,
+      "loss": 1.9592,
+      "step": 13057
+    },
+    {
+      "epoch": 0.5524156019967849,
+      "grad_norm": 0.16090841591358185,
+      "learning_rate": 0.001,
+      "loss": 2.6147,
+      "step": 13058
+    },
+    {
+      "epoch": 0.5524579067603013,
+      "grad_norm": 0.20968280732631683,
+      "learning_rate": 0.001,
+      "loss": 2.6148,
+      "step": 13059
+    },
+    {
+      "epoch": 0.5525002115238176,
+      "grad_norm": 0.16014662384986877,
+      "learning_rate": 0.001,
+      "loss": 2.2449,
+      "step": 13060
+    },
+    {
+      "epoch": 0.552542516287334,
+      "grad_norm": 0.4493609070777893,
+      "learning_rate": 0.001,
+      "loss": 1.5071,
+      "step": 13061
+    },
+    {
+      "epoch": 0.5525848210508503,
+      "grad_norm": 0.20664609968662262,
+      "learning_rate": 0.001,
+      "loss": 2.2177,
+      "step": 13062
+    },
+    {
+      "epoch": 0.5526271258143667,
+      "grad_norm": 0.1800384223461151,
+      "learning_rate": 0.001,
+      "loss": 1.9051,
+      "step": 13063
+    },
+    {
+      "epoch": 0.5526694305778831,
+      "grad_norm": 0.16999885439872742,
+      "learning_rate": 0.001,
+      "loss": 2.2448,
+      "step": 13064
+    },
+    {
+      "epoch": 0.5527117353413994,
+      "grad_norm": 0.1717803031206131,
+      "learning_rate": 0.001,
+      "loss": 1.679,
+      "step": 13065
+    },
+    {
+      "epoch": 0.5527540401049158,
+      "grad_norm": 0.3044344186782837,
+      "learning_rate": 0.001,
+      "loss": 2.5296,
+      "step": 13066
+    },
+    {
+      "epoch": 0.5527963448684322,
+      "grad_norm": 0.15595632791519165,
+      "learning_rate": 0.001,
+      "loss": 1.7021,
+      "step": 13067
+    },
+    {
+      "epoch": 0.5528386496319485,
+      "grad_norm": 0.15000808238983154,
+      "learning_rate": 0.001,
+      "loss": 1.5892,
+      "step": 13068
+    },
+    {
+      "epoch": 0.5528809543954649,
+      "grad_norm": 1.946815848350525,
+      "learning_rate": 0.001,
+      "loss": 1.4456,
+      "step": 13069
+    },
+    {
+      "epoch": 0.5529232591589813,
+      "grad_norm": 0.16129404306411743,
+      "learning_rate": 0.001,
+      "loss": 1.5363,
+      "step": 13070
+    },
+    {
+      "epoch": 0.5529655639224976,
+      "grad_norm": 0.22747275233268738,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 13071
+    },
+    {
+      "epoch": 0.553007868686014,
+      "grad_norm": 0.16564899682998657,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 13072
+    },
+    {
+      "epoch": 0.5530501734495304,
+      "grad_norm": 0.1645679473876953,
+      "learning_rate": 0.001,
+      "loss": 2.7037,
+      "step": 13073
+    },
+    {
+      "epoch": 0.5530924782130467,
+      "grad_norm": 0.18701903522014618,
+      "learning_rate": 0.001,
+      "loss": 1.9455,
+      "step": 13074
+    },
+    {
+      "epoch": 0.5531347829765632,
+      "grad_norm": 0.16058051586151123,
+      "learning_rate": 0.001,
+      "loss": 2.1071,
+      "step": 13075
+    },
+    {
+      "epoch": 0.5531770877400796,
+      "grad_norm": 0.16248644888401031,
+      "learning_rate": 0.001,
+      "loss": 2.7373,
+      "step": 13076
+    },
+    {
+      "epoch": 0.5532193925035959,
+      "grad_norm": 0.20399704575538635,
+      "learning_rate": 0.001,
+      "loss": 2.0442,
+      "step": 13077
+    },
+    {
+      "epoch": 0.5532616972671123,
+      "grad_norm": 0.14781156182289124,
+      "learning_rate": 0.001,
+      "loss": 1.7802,
+      "step": 13078
+    },
+    {
+      "epoch": 0.5533040020306287,
+      "grad_norm": 0.3947334587574005,
+      "learning_rate": 0.001,
+      "loss": 1.9844,
+      "step": 13079
+    },
+    {
+      "epoch": 0.553346306794145,
+      "grad_norm": 0.16816608607769012,
+      "learning_rate": 0.001,
+      "loss": 2.2254,
+      "step": 13080
+    },
+    {
+      "epoch": 0.5533886115576614,
+      "grad_norm": 0.548812210559845,
+      "learning_rate": 0.001,
+      "loss": 2.331,
+      "step": 13081
+    },
+    {
+      "epoch": 0.5534309163211778,
+      "grad_norm": 0.16253416240215302,
+      "learning_rate": 0.001,
+      "loss": 2.1435,
+      "step": 13082
+    },
+    {
+      "epoch": 0.5534732210846941,
+      "grad_norm": 0.33334457874298096,
+      "learning_rate": 0.001,
+      "loss": 1.7767,
+      "step": 13083
+    },
+    {
+      "epoch": 0.5535155258482105,
+      "grad_norm": 0.7046263813972473,
+      "learning_rate": 0.001,
+      "loss": 2.3232,
+      "step": 13084
+    },
+    {
+      "epoch": 0.5535578306117269,
+      "grad_norm": 0.18305683135986328,
+      "learning_rate": 0.001,
+      "loss": 2.4841,
+      "step": 13085
+    },
+    {
+      "epoch": 0.5536001353752432,
+      "grad_norm": 0.9522390365600586,
+      "learning_rate": 0.001,
+      "loss": 3.2129,
+      "step": 13086
+    },
+    {
+      "epoch": 0.5536424401387596,
+      "grad_norm": 0.16383132338523865,
+      "learning_rate": 0.001,
+      "loss": 1.8029,
+      "step": 13087
+    },
+    {
+      "epoch": 0.553684744902276,
+      "grad_norm": 0.16644108295440674,
+      "learning_rate": 0.001,
+      "loss": 1.9539,
+      "step": 13088
+    },
+    {
+      "epoch": 0.5537270496657923,
+      "grad_norm": 0.16594970226287842,
+      "learning_rate": 0.001,
+      "loss": 2.1514,
+      "step": 13089
+    },
+    {
+      "epoch": 0.5537693544293087,
+      "grad_norm": 0.19843560457229614,
+      "learning_rate": 0.001,
+      "loss": 2.751,
+      "step": 13090
+    },
+    {
+      "epoch": 0.5538116591928252,
+      "grad_norm": 0.166713684797287,
+      "learning_rate": 0.001,
+      "loss": 1.9202,
+      "step": 13091
+    },
+    {
+      "epoch": 0.5538539639563415,
+      "grad_norm": 0.18431521952152252,
+      "learning_rate": 0.001,
+      "loss": 2.3885,
+      "step": 13092
+    },
+    {
+      "epoch": 0.5538962687198579,
+      "grad_norm": 0.16817611455917358,
+      "learning_rate": 0.001,
+      "loss": 2.1286,
+      "step": 13093
+    },
+    {
+      "epoch": 0.5539385734833743,
+      "grad_norm": 0.15801174938678741,
+      "learning_rate": 0.001,
+      "loss": 1.8729,
+      "step": 13094
+    },
+    {
+      "epoch": 0.5539808782468906,
+      "grad_norm": 0.3943849802017212,
+      "learning_rate": 0.001,
+      "loss": 2.5361,
+      "step": 13095
+    },
+    {
+      "epoch": 0.554023183010407,
+      "grad_norm": 0.14820386469364166,
+      "learning_rate": 0.001,
+      "loss": 1.7892,
+      "step": 13096
+    },
+    {
+      "epoch": 0.5540654877739234,
+      "grad_norm": 0.15670961141586304,
+      "learning_rate": 0.001,
+      "loss": 2.8515,
+      "step": 13097
+    },
+    {
+      "epoch": 0.5541077925374397,
+      "grad_norm": 0.1757040172815323,
+      "learning_rate": 0.001,
+      "loss": 2.1255,
+      "step": 13098
+    },
+    {
+      "epoch": 0.5541500973009561,
+      "grad_norm": 0.5804983377456665,
+      "learning_rate": 0.001,
+      "loss": 1.6095,
+      "step": 13099
+    },
+    {
+      "epoch": 0.5541924020644725,
+      "grad_norm": 0.18276961147785187,
+      "learning_rate": 0.001,
+      "loss": 1.8125,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5542347068279888,
+      "grad_norm": 0.24642734229564667,
+      "learning_rate": 0.001,
+      "loss": 2.0911,
+      "step": 13101
+    },
+    {
+      "epoch": 0.5542770115915052,
+      "grad_norm": 0.16965211927890778,
+      "learning_rate": 0.001,
+      "loss": 1.8867,
+      "step": 13102
+    },
+    {
+      "epoch": 0.5543193163550216,
+      "grad_norm": 0.1579209566116333,
+      "learning_rate": 0.001,
+      "loss": 1.7393,
+      "step": 13103
+    },
+    {
+      "epoch": 0.5543616211185379,
+      "grad_norm": 0.2219139188528061,
+      "learning_rate": 0.001,
+      "loss": 1.8149,
+      "step": 13104
+    },
+    {
+      "epoch": 0.5544039258820543,
+      "grad_norm": 0.17188802361488342,
+      "learning_rate": 0.001,
+      "loss": 1.8126,
+      "step": 13105
+    },
+    {
+      "epoch": 0.5544462306455706,
+      "grad_norm": 0.14392749965190887,
+      "learning_rate": 0.001,
+      "loss": 2.0954,
+      "step": 13106
+    },
+    {
+      "epoch": 0.554488535409087,
+      "grad_norm": 0.19999627768993378,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 13107
+    },
+    {
+      "epoch": 0.5545308401726035,
+      "grad_norm": 0.20926088094711304,
+      "learning_rate": 0.001,
+      "loss": 3.1452,
+      "step": 13108
+    },
+    {
+      "epoch": 0.5545731449361198,
+      "grad_norm": 0.19249233603477478,
+      "learning_rate": 0.001,
+      "loss": 2.3985,
+      "step": 13109
+    },
+    {
+      "epoch": 0.5546154496996362,
+      "grad_norm": 0.1850895881652832,
+      "learning_rate": 0.001,
+      "loss": 1.7581,
+      "step": 13110
+    },
+    {
+      "epoch": 0.5546577544631526,
+      "grad_norm": 0.1713310182094574,
+      "learning_rate": 0.001,
+      "loss": 3.1031,
+      "step": 13111
+    },
+    {
+      "epoch": 0.5547000592266689,
+      "grad_norm": 0.21461325883865356,
+      "learning_rate": 0.001,
+      "loss": 1.462,
+      "step": 13112
+    },
+    {
+      "epoch": 0.5547423639901853,
+      "grad_norm": 0.19272612035274506,
+      "learning_rate": 0.001,
+      "loss": 2.0605,
+      "step": 13113
+    },
+    {
+      "epoch": 0.5547846687537017,
+      "grad_norm": 0.5416992902755737,
+      "learning_rate": 0.001,
+      "loss": 2.4492,
+      "step": 13114
+    },
+    {
+      "epoch": 0.554826973517218,
+      "grad_norm": 0.16244624555110931,
+      "learning_rate": 0.001,
+      "loss": 1.9301,
+      "step": 13115
+    },
+    {
+      "epoch": 0.5548692782807344,
+      "grad_norm": 0.1623430848121643,
+      "learning_rate": 0.001,
+      "loss": 1.8153,
+      "step": 13116
+    },
+    {
+      "epoch": 0.5549115830442508,
+      "grad_norm": 0.15452533960342407,
+      "learning_rate": 0.001,
+      "loss": 2.2119,
+      "step": 13117
+    },
+    {
+      "epoch": 0.5549538878077671,
+      "grad_norm": 0.1652129739522934,
+      "learning_rate": 0.001,
+      "loss": 2.0223,
+      "step": 13118
+    },
+    {
+      "epoch": 0.5549961925712835,
+      "grad_norm": 0.19139006733894348,
+      "learning_rate": 0.001,
+      "loss": 2.2582,
+      "step": 13119
+    },
+    {
+      "epoch": 0.5550384973347999,
+      "grad_norm": 0.21414639055728912,
+      "learning_rate": 0.001,
+      "loss": 1.978,
+      "step": 13120
+    },
+    {
+      "epoch": 0.5550808020983162,
+      "grad_norm": 0.18957240879535675,
+      "learning_rate": 0.001,
+      "loss": 2.3123,
+      "step": 13121
+    },
+    {
+      "epoch": 0.5551231068618326,
+      "grad_norm": 0.16492891311645508,
+      "learning_rate": 0.001,
+      "loss": 2.2614,
+      "step": 13122
+    },
+    {
+      "epoch": 0.555165411625349,
+      "grad_norm": 0.17315563559532166,
+      "learning_rate": 0.001,
+      "loss": 1.4405,
+      "step": 13123
+    },
+    {
+      "epoch": 0.5552077163888653,
+      "grad_norm": 0.18722382187843323,
+      "learning_rate": 0.001,
+      "loss": 2.4344,
+      "step": 13124
+    },
+    {
+      "epoch": 0.5552500211523818,
+      "grad_norm": 0.13166561722755432,
+      "learning_rate": 0.001,
+      "loss": 2.4014,
+      "step": 13125
+    },
+    {
+      "epoch": 0.5552923259158982,
+      "grad_norm": 0.22480559349060059,
+      "learning_rate": 0.001,
+      "loss": 1.9176,
+      "step": 13126
+    },
+    {
+      "epoch": 0.5553346306794145,
+      "grad_norm": 0.15770910680294037,
+      "learning_rate": 0.001,
+      "loss": 2.2783,
+      "step": 13127
+    },
+    {
+      "epoch": 0.5553769354429309,
+      "grad_norm": 0.17504677176475525,
+      "learning_rate": 0.001,
+      "loss": 1.8905,
+      "step": 13128
+    },
+    {
+      "epoch": 0.5554192402064473,
+      "grad_norm": 0.18048499524593353,
+      "learning_rate": 0.001,
+      "loss": 2.557,
+      "step": 13129
+    },
+    {
+      "epoch": 0.5554615449699636,
+      "grad_norm": 0.19142648577690125,
+      "learning_rate": 0.001,
+      "loss": 1.8773,
+      "step": 13130
+    },
+    {
+      "epoch": 0.55550384973348,
+      "grad_norm": 0.16051210463047028,
+      "learning_rate": 0.001,
+      "loss": 1.8186,
+      "step": 13131
+    },
+    {
+      "epoch": 0.5555461544969964,
+      "grad_norm": 0.21467453241348267,
+      "learning_rate": 0.001,
+      "loss": 2.09,
+      "step": 13132
+    },
+    {
+      "epoch": 0.5555884592605127,
+      "grad_norm": 0.7346461415290833,
+      "learning_rate": 0.001,
+      "loss": 2.0679,
+      "step": 13133
+    },
+    {
+      "epoch": 0.5556307640240291,
+      "grad_norm": 0.1747002899646759,
+      "learning_rate": 0.001,
+      "loss": 1.8083,
+      "step": 13134
+    },
+    {
+      "epoch": 0.5556730687875455,
+      "grad_norm": 0.17838482558727264,
+      "learning_rate": 0.001,
+      "loss": 2.428,
+      "step": 13135
+    },
+    {
+      "epoch": 0.5557153735510618,
+      "grad_norm": 0.15864208340644836,
+      "learning_rate": 0.001,
+      "loss": 2.0233,
+      "step": 13136
+    },
+    {
+      "epoch": 0.5557576783145782,
+      "grad_norm": 0.27574223279953003,
+      "learning_rate": 0.001,
+      "loss": 2.9788,
+      "step": 13137
+    },
+    {
+      "epoch": 0.5557999830780946,
+      "grad_norm": 0.18483395874500275,
+      "learning_rate": 0.001,
+      "loss": 2.0938,
+      "step": 13138
+    },
+    {
+      "epoch": 0.5558422878416109,
+      "grad_norm": 0.1585381031036377,
+      "learning_rate": 0.001,
+      "loss": 1.4932,
+      "step": 13139
+    },
+    {
+      "epoch": 0.5558845926051273,
+      "grad_norm": 0.19436490535736084,
+      "learning_rate": 0.001,
+      "loss": 2.827,
+      "step": 13140
+    },
+    {
+      "epoch": 0.5559268973686438,
+      "grad_norm": 0.9730831384658813,
+      "learning_rate": 0.001,
+      "loss": 1.8887,
+      "step": 13141
+    },
+    {
+      "epoch": 0.5559692021321601,
+      "grad_norm": 0.16187086701393127,
+      "learning_rate": 0.001,
+      "loss": 2.15,
+      "step": 13142
+    },
+    {
+      "epoch": 0.5560115068956765,
+      "grad_norm": 0.17080490291118622,
+      "learning_rate": 0.001,
+      "loss": 2.7272,
+      "step": 13143
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.4234422445297241,
+      "learning_rate": 0.001,
+      "loss": 2.6932,
+      "step": 13144
+    },
+    {
+      "epoch": 0.5560961164227092,
+      "grad_norm": 0.16992883384227753,
+      "learning_rate": 0.001,
+      "loss": 1.6274,
+      "step": 13145
+    },
+    {
+      "epoch": 0.5561384211862256,
+      "grad_norm": 0.1784851998090744,
+      "learning_rate": 0.001,
+      "loss": 1.9868,
+      "step": 13146
+    },
+    {
+      "epoch": 0.556180725949742,
+      "grad_norm": 0.18666186928749084,
+      "learning_rate": 0.001,
+      "loss": 2.8875,
+      "step": 13147
+    },
+    {
+      "epoch": 0.5562230307132583,
+      "grad_norm": 0.17807726562023163,
+      "learning_rate": 0.001,
+      "loss": 1.6625,
+      "step": 13148
+    },
+    {
+      "epoch": 0.5562653354767747,
+      "grad_norm": 0.14291433990001678,
+      "learning_rate": 0.001,
+      "loss": 2.1721,
+      "step": 13149
+    },
+    {
+      "epoch": 0.556307640240291,
+      "grad_norm": 0.18871572613716125,
+      "learning_rate": 0.001,
+      "loss": 1.5161,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5563499450038074,
+      "grad_norm": 0.2232998013496399,
+      "learning_rate": 0.001,
+      "loss": 1.7724,
+      "step": 13151
+    },
+    {
+      "epoch": 0.5563922497673238,
+      "grad_norm": 0.21239492297172546,
+      "learning_rate": 0.001,
+      "loss": 2.0644,
+      "step": 13152
+    },
+    {
+      "epoch": 0.5564345545308401,
+      "grad_norm": 0.177618145942688,
+      "learning_rate": 0.001,
+      "loss": 2.4754,
+      "step": 13153
+    },
+    {
+      "epoch": 0.5564768592943565,
+      "grad_norm": 0.2010020613670349,
+      "learning_rate": 0.001,
+      "loss": 1.8863,
+      "step": 13154
+    },
+    {
+      "epoch": 0.5565191640578729,
+      "grad_norm": 0.15301641821861267,
+      "learning_rate": 0.001,
+      "loss": 2.6225,
+      "step": 13155
+    },
+    {
+      "epoch": 0.5565614688213892,
+      "grad_norm": 0.16343499720096588,
+      "learning_rate": 0.001,
+      "loss": 2.5127,
+      "step": 13156
+    },
+    {
+      "epoch": 0.5566037735849056,
+      "grad_norm": 0.15531091392040253,
+      "learning_rate": 0.001,
+      "loss": 2.0179,
+      "step": 13157
+    },
+    {
+      "epoch": 0.5566460783484221,
+      "grad_norm": 0.18499448895454407,
+      "learning_rate": 0.001,
+      "loss": 1.4872,
+      "step": 13158
+    },
+    {
+      "epoch": 0.5566883831119384,
+      "grad_norm": 0.21716418862342834,
+      "learning_rate": 0.001,
+      "loss": 1.7633,
+      "step": 13159
+    },
+    {
+      "epoch": 0.5567306878754548,
+      "grad_norm": 0.3633074462413788,
+      "learning_rate": 0.001,
+      "loss": 1.7639,
+      "step": 13160
+    },
+    {
+      "epoch": 0.5567729926389712,
+      "grad_norm": 0.20026670396327972,
+      "learning_rate": 0.001,
+      "loss": 1.4217,
+      "step": 13161
+    },
+    {
+      "epoch": 0.5568152974024875,
+      "grad_norm": 0.18606965243816376,
+      "learning_rate": 0.001,
+      "loss": 1.8908,
+      "step": 13162
+    },
+    {
+      "epoch": 0.5568576021660039,
+      "grad_norm": 0.17996345460414886,
+      "learning_rate": 0.001,
+      "loss": 1.7892,
+      "step": 13163
+    },
+    {
+      "epoch": 0.5568999069295203,
+      "grad_norm": 0.1515737622976303,
+      "learning_rate": 0.001,
+      "loss": 1.8319,
+      "step": 13164
+    },
+    {
+      "epoch": 0.5569422116930366,
+      "grad_norm": 0.21763929724693298,
+      "learning_rate": 0.001,
+      "loss": 3.2495,
+      "step": 13165
+    },
+    {
+      "epoch": 0.556984516456553,
+      "grad_norm": 0.18336108326911926,
+      "learning_rate": 0.001,
+      "loss": 1.6638,
+      "step": 13166
+    },
+    {
+      "epoch": 0.5570268212200694,
+      "grad_norm": 0.1941053867340088,
+      "learning_rate": 0.001,
+      "loss": 2.2389,
+      "step": 13167
+    },
+    {
+      "epoch": 0.5570691259835857,
+      "grad_norm": 0.2293504923582077,
+      "learning_rate": 0.001,
+      "loss": 1.7683,
+      "step": 13168
+    },
+    {
+      "epoch": 0.5571114307471021,
+      "grad_norm": 0.18686215579509735,
+      "learning_rate": 0.001,
+      "loss": 1.6258,
+      "step": 13169
+    },
+    {
+      "epoch": 0.5571537355106185,
+      "grad_norm": 0.17129628360271454,
+      "learning_rate": 0.001,
+      "loss": 1.8111,
+      "step": 13170
+    },
+    {
+      "epoch": 0.5571960402741348,
+      "grad_norm": 0.20387300848960876,
+      "learning_rate": 0.001,
+      "loss": 2.2676,
+      "step": 13171
+    },
+    {
+      "epoch": 0.5572383450376512,
+      "grad_norm": 0.15360292792320251,
+      "learning_rate": 0.001,
+      "loss": 1.8776,
+      "step": 13172
+    },
+    {
+      "epoch": 0.5572806498011676,
+      "grad_norm": 0.20021335780620575,
+      "learning_rate": 0.001,
+      "loss": 1.9061,
+      "step": 13173
+    },
+    {
+      "epoch": 0.557322954564684,
+      "grad_norm": 0.19312505424022675,
+      "learning_rate": 0.001,
+      "loss": 1.8143,
+      "step": 13174
+    },
+    {
+      "epoch": 0.5573652593282004,
+      "grad_norm": 0.19307121634483337,
+      "learning_rate": 0.001,
+      "loss": 4.1747,
+      "step": 13175
+    },
+    {
+      "epoch": 0.5574075640917168,
+      "grad_norm": 0.16115133464336395,
+      "learning_rate": 0.001,
+      "loss": 2.7506,
+      "step": 13176
+    },
+    {
+      "epoch": 0.5574498688552331,
+      "grad_norm": 0.1843666285276413,
+      "learning_rate": 0.001,
+      "loss": 2.6517,
+      "step": 13177
+    },
+    {
+      "epoch": 0.5574921736187495,
+      "grad_norm": 0.21303428709506989,
+      "learning_rate": 0.001,
+      "loss": 2.7585,
+      "step": 13178
+    },
+    {
+      "epoch": 0.5575344783822659,
+      "grad_norm": 0.16230140626430511,
+      "learning_rate": 0.001,
+      "loss": 1.822,
+      "step": 13179
+    },
+    {
+      "epoch": 0.5575767831457822,
+      "grad_norm": 0.20180179178714752,
+      "learning_rate": 0.001,
+      "loss": 2.4642,
+      "step": 13180
+    },
+    {
+      "epoch": 0.5576190879092986,
+      "grad_norm": 0.1860133558511734,
+      "learning_rate": 0.001,
+      "loss": 2.4435,
+      "step": 13181
+    },
+    {
+      "epoch": 0.557661392672815,
+      "grad_norm": 0.33960339426994324,
+      "learning_rate": 0.001,
+      "loss": 3.8256,
+      "step": 13182
+    },
+    {
+      "epoch": 0.5577036974363313,
+      "grad_norm": 0.1435151845216751,
+      "learning_rate": 0.001,
+      "loss": 1.4867,
+      "step": 13183
+    },
+    {
+      "epoch": 0.5577460021998477,
+      "grad_norm": 0.1442885398864746,
+      "learning_rate": 0.001,
+      "loss": 3.0296,
+      "step": 13184
+    },
+    {
+      "epoch": 0.5577883069633641,
+      "grad_norm": 0.3879304528236389,
+      "learning_rate": 0.001,
+      "loss": 2.4309,
+      "step": 13185
+    },
+    {
+      "epoch": 0.5578306117268804,
+      "grad_norm": 0.15823771059513092,
+      "learning_rate": 0.001,
+      "loss": 2.0006,
+      "step": 13186
+    },
+    {
+      "epoch": 0.5578729164903968,
+      "grad_norm": 0.2137526571750641,
+      "learning_rate": 0.001,
+      "loss": 2.8213,
+      "step": 13187
+    },
+    {
+      "epoch": 0.5579152212539132,
+      "grad_norm": 0.2220454066991806,
+      "learning_rate": 0.001,
+      "loss": 1.751,
+      "step": 13188
+    },
+    {
+      "epoch": 0.5579575260174295,
+      "grad_norm": 0.15580366551876068,
+      "learning_rate": 0.001,
+      "loss": 3.3767,
+      "step": 13189
+    },
+    {
+      "epoch": 0.557999830780946,
+      "grad_norm": 0.1863831877708435,
+      "learning_rate": 0.001,
+      "loss": 2.8947,
+      "step": 13190
+    },
+    {
+      "epoch": 0.5580421355444624,
+      "grad_norm": 0.17505452036857605,
+      "learning_rate": 0.001,
+      "loss": 2.7388,
+      "step": 13191
+    },
+    {
+      "epoch": 0.5580844403079787,
+      "grad_norm": 0.5289129018783569,
+      "learning_rate": 0.001,
+      "loss": 1.7719,
+      "step": 13192
+    },
+    {
+      "epoch": 0.5581267450714951,
+      "grad_norm": 0.15774008631706238,
+      "learning_rate": 0.001,
+      "loss": 1.9727,
+      "step": 13193
+    },
+    {
+      "epoch": 0.5581690498350115,
+      "grad_norm": 0.16783976554870605,
+      "learning_rate": 0.001,
+      "loss": 1.9035,
+      "step": 13194
+    },
+    {
+      "epoch": 0.5582113545985278,
+      "grad_norm": 0.19274777173995972,
+      "learning_rate": 0.001,
+      "loss": 2.3897,
+      "step": 13195
+    },
+    {
+      "epoch": 0.5582536593620442,
+      "grad_norm": 0.15515726804733276,
+      "learning_rate": 0.001,
+      "loss": 2.1614,
+      "step": 13196
+    },
+    {
+      "epoch": 0.5582959641255605,
+      "grad_norm": 0.20335707068443298,
+      "learning_rate": 0.001,
+      "loss": 2.5368,
+      "step": 13197
+    },
+    {
+      "epoch": 0.5583382688890769,
+      "grad_norm": 0.17299920320510864,
+      "learning_rate": 0.001,
+      "loss": 1.9999,
+      "step": 13198
+    },
+    {
+      "epoch": 0.5583805736525933,
+      "grad_norm": 0.167136088013649,
+      "learning_rate": 0.001,
+      "loss": 1.841,
+      "step": 13199
+    },
+    {
+      "epoch": 0.5584228784161096,
+      "grad_norm": 0.19228960573673248,
+      "learning_rate": 0.001,
+      "loss": 3.3171,
+      "step": 13200
+    },
+    {
+      "epoch": 0.558465183179626,
+      "grad_norm": 0.13816307485103607,
+      "learning_rate": 0.001,
+      "loss": 3.1786,
+      "step": 13201
+    },
+    {
+      "epoch": 0.5585074879431424,
+      "grad_norm": 1.117954969406128,
+      "learning_rate": 0.001,
+      "loss": 2.6146,
+      "step": 13202
+    },
+    {
+      "epoch": 0.5585497927066587,
+      "grad_norm": 0.1563483625650406,
+      "learning_rate": 0.001,
+      "loss": 2.3211,
+      "step": 13203
+    },
+    {
+      "epoch": 0.5585920974701751,
+      "grad_norm": 0.15023623406887054,
+      "learning_rate": 0.001,
+      "loss": 1.7047,
+      "step": 13204
+    },
+    {
+      "epoch": 0.5586344022336915,
+      "grad_norm": 0.21180753409862518,
+      "learning_rate": 0.001,
+      "loss": 3.0361,
+      "step": 13205
+    },
+    {
+      "epoch": 0.5586767069972078,
+      "grad_norm": 1.6476328372955322,
+      "learning_rate": 0.001,
+      "loss": 2.0756,
+      "step": 13206
+    },
+    {
+      "epoch": 0.5587190117607242,
+      "grad_norm": 0.1770959347486496,
+      "learning_rate": 0.001,
+      "loss": 2.0876,
+      "step": 13207
+    },
+    {
+      "epoch": 0.5587613165242407,
+      "grad_norm": 0.16517281532287598,
+      "learning_rate": 0.001,
+      "loss": 2.9517,
+      "step": 13208
+    },
+    {
+      "epoch": 0.558803621287757,
+      "grad_norm": 0.24339810013771057,
+      "learning_rate": 0.001,
+      "loss": 2.227,
+      "step": 13209
+    },
+    {
+      "epoch": 0.5588459260512734,
+      "grad_norm": 0.2662486732006073,
+      "learning_rate": 0.001,
+      "loss": 1.6548,
+      "step": 13210
+    },
+    {
+      "epoch": 0.5588882308147898,
+      "grad_norm": 2.719054698944092,
+      "learning_rate": 0.001,
+      "loss": 2.0465,
+      "step": 13211
+    },
+    {
+      "epoch": 0.5589305355783061,
+      "grad_norm": 0.4585936367511749,
+      "learning_rate": 0.001,
+      "loss": 2.4569,
+      "step": 13212
+    },
+    {
+      "epoch": 0.5589728403418225,
+      "grad_norm": 0.2304200530052185,
+      "learning_rate": 0.001,
+      "loss": 2.2496,
+      "step": 13213
+    },
+    {
+      "epoch": 0.5590151451053389,
+      "grad_norm": 0.4977359175682068,
+      "learning_rate": 0.001,
+      "loss": 2.5334,
+      "step": 13214
+    },
+    {
+      "epoch": 0.5590574498688552,
+      "grad_norm": 0.1636125147342682,
+      "learning_rate": 0.001,
+      "loss": 2.0662,
+      "step": 13215
+    },
+    {
+      "epoch": 0.5590997546323716,
+      "grad_norm": 0.18210989236831665,
+      "learning_rate": 0.001,
+      "loss": 1.9016,
+      "step": 13216
+    },
+    {
+      "epoch": 0.559142059395888,
+      "grad_norm": 0.5983842611312866,
+      "learning_rate": 0.001,
+      "loss": 3.717,
+      "step": 13217
+    },
+    {
+      "epoch": 0.5591843641594043,
+      "grad_norm": 0.20817320048809052,
+      "learning_rate": 0.001,
+      "loss": 1.5124,
+      "step": 13218
+    },
+    {
+      "epoch": 0.5592266689229207,
+      "grad_norm": 0.16508392989635468,
+      "learning_rate": 0.001,
+      "loss": 1.7414,
+      "step": 13219
+    },
+    {
+      "epoch": 0.5592689736864371,
+      "grad_norm": 0.19145143032073975,
+      "learning_rate": 0.001,
+      "loss": 2.4444,
+      "step": 13220
+    },
+    {
+      "epoch": 0.5593112784499534,
+      "grad_norm": 0.2849757969379425,
+      "learning_rate": 0.001,
+      "loss": 4.1417,
+      "step": 13221
+    },
+    {
+      "epoch": 0.5593535832134698,
+      "grad_norm": 0.20053541660308838,
+      "learning_rate": 0.001,
+      "loss": 2.7225,
+      "step": 13222
+    },
+    {
+      "epoch": 0.5593958879769863,
+      "grad_norm": 0.19015778601169586,
+      "learning_rate": 0.001,
+      "loss": 1.9326,
+      "step": 13223
+    },
+    {
+      "epoch": 0.5594381927405025,
+      "grad_norm": 0.22139129042625427,
+      "learning_rate": 0.001,
+      "loss": 2.0288,
+      "step": 13224
+    },
+    {
+      "epoch": 0.559480497504019,
+      "grad_norm": 0.18793657422065735,
+      "learning_rate": 0.001,
+      "loss": 2.234,
+      "step": 13225
+    },
+    {
+      "epoch": 0.5595228022675354,
+      "grad_norm": 0.3711003363132477,
+      "learning_rate": 0.001,
+      "loss": 2.8216,
+      "step": 13226
+    },
+    {
+      "epoch": 0.5595651070310517,
+      "grad_norm": 0.35137489438056946,
+      "learning_rate": 0.001,
+      "loss": 2.3047,
+      "step": 13227
+    },
+    {
+      "epoch": 0.5596074117945681,
+      "grad_norm": 0.3833855092525482,
+      "learning_rate": 0.001,
+      "loss": 2.2983,
+      "step": 13228
+    },
+    {
+      "epoch": 0.5596497165580845,
+      "grad_norm": 0.19461855292320251,
+      "learning_rate": 0.001,
+      "loss": 2.4572,
+      "step": 13229
+    },
+    {
+      "epoch": 0.5596920213216008,
+      "grad_norm": 0.22493231296539307,
+      "learning_rate": 0.001,
+      "loss": 2.1651,
+      "step": 13230
+    },
+    {
+      "epoch": 0.5597343260851172,
+      "grad_norm": 0.19130484759807587,
+      "learning_rate": 0.001,
+      "loss": 2.1431,
+      "step": 13231
+    },
+    {
+      "epoch": 0.5597766308486336,
+      "grad_norm": 0.1570371687412262,
+      "learning_rate": 0.001,
+      "loss": 1.8944,
+      "step": 13232
+    },
+    {
+      "epoch": 0.5598189356121499,
+      "grad_norm": 0.16099748015403748,
+      "learning_rate": 0.001,
+      "loss": 2.277,
+      "step": 13233
+    },
+    {
+      "epoch": 0.5598612403756663,
+      "grad_norm": 0.18128852546215057,
+      "learning_rate": 0.001,
+      "loss": 1.8642,
+      "step": 13234
+    },
+    {
+      "epoch": 0.5599035451391827,
+      "grad_norm": 0.2672935426235199,
+      "learning_rate": 0.001,
+      "loss": 1.5749,
+      "step": 13235
+    },
+    {
+      "epoch": 0.559945849902699,
+      "grad_norm": 0.1762719750404358,
+      "learning_rate": 0.001,
+      "loss": 2.1202,
+      "step": 13236
+    },
+    {
+      "epoch": 0.5599881546662154,
+      "grad_norm": 0.17224456369876862,
+      "learning_rate": 0.001,
+      "loss": 2.2378,
+      "step": 13237
+    },
+    {
+      "epoch": 0.5600304594297318,
+      "grad_norm": 0.15539632737636566,
+      "learning_rate": 0.001,
+      "loss": 2.3729,
+      "step": 13238
+    },
+    {
+      "epoch": 0.5600727641932481,
+      "grad_norm": 0.16019120812416077,
+      "learning_rate": 0.001,
+      "loss": 2.0313,
+      "step": 13239
+    },
+    {
+      "epoch": 0.5601150689567646,
+      "grad_norm": 0.21855637431144714,
+      "learning_rate": 0.001,
+      "loss": 2.3144,
+      "step": 13240
+    },
+    {
+      "epoch": 0.5601573737202808,
+      "grad_norm": 0.20648352801799774,
+      "learning_rate": 0.001,
+      "loss": 1.9184,
+      "step": 13241
+    },
+    {
+      "epoch": 0.5601996784837973,
+      "grad_norm": 0.1866380274295807,
+      "learning_rate": 0.001,
+      "loss": 2.8954,
+      "step": 13242
+    },
+    {
+      "epoch": 0.5602419832473137,
+      "grad_norm": 4.305062294006348,
+      "learning_rate": 0.001,
+      "loss": 2.8966,
+      "step": 13243
+    },
+    {
+      "epoch": 0.56028428801083,
+      "grad_norm": 0.18245016038417816,
+      "learning_rate": 0.001,
+      "loss": 2.472,
+      "step": 13244
+    },
+    {
+      "epoch": 0.5603265927743464,
+      "grad_norm": 0.15738457441329956,
+      "learning_rate": 0.001,
+      "loss": 1.5529,
+      "step": 13245
+    },
+    {
+      "epoch": 0.5603688975378628,
+      "grad_norm": 0.1629572957754135,
+      "learning_rate": 0.001,
+      "loss": 1.5193,
+      "step": 13246
+    },
+    {
+      "epoch": 0.5604112023013791,
+      "grad_norm": 0.46464836597442627,
+      "learning_rate": 0.001,
+      "loss": 3.0418,
+      "step": 13247
+    },
+    {
+      "epoch": 0.5604535070648955,
+      "grad_norm": 0.2348056435585022,
+      "learning_rate": 0.001,
+      "loss": 2.3288,
+      "step": 13248
+    },
+    {
+      "epoch": 0.5604958118284119,
+      "grad_norm": 0.9263020753860474,
+      "learning_rate": 0.001,
+      "loss": 2.1417,
+      "step": 13249
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 0.32688605785369873,
+      "learning_rate": 0.001,
+      "loss": 2.0048,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5605804213554446,
+      "grad_norm": 0.226731076836586,
+      "learning_rate": 0.001,
+      "loss": 2.6983,
+      "step": 13251
+    },
+    {
+      "epoch": 0.560622726118961,
+      "grad_norm": 0.40565672516822815,
+      "learning_rate": 0.001,
+      "loss": 2.6863,
+      "step": 13252
+    },
+    {
+      "epoch": 0.5606650308824773,
+      "grad_norm": 0.21357344090938568,
+      "learning_rate": 0.001,
+      "loss": 1.5538,
+      "step": 13253
+    },
+    {
+      "epoch": 0.5607073356459937,
+      "grad_norm": 0.20866970717906952,
+      "learning_rate": 0.001,
+      "loss": 1.9627,
+      "step": 13254
+    },
+    {
+      "epoch": 0.5607496404095101,
+      "grad_norm": 0.20830655097961426,
+      "learning_rate": 0.001,
+      "loss": 2.1551,
+      "step": 13255
+    },
+    {
+      "epoch": 0.5607919451730264,
+      "grad_norm": 0.19615496695041656,
+      "learning_rate": 0.001,
+      "loss": 2.1498,
+      "step": 13256
+    },
+    {
+      "epoch": 0.5608342499365429,
+      "grad_norm": 0.32191988825798035,
+      "learning_rate": 0.001,
+      "loss": 3.5976,
+      "step": 13257
+    },
+    {
+      "epoch": 0.5608765547000593,
+      "grad_norm": 0.22609533369541168,
+      "learning_rate": 0.001,
+      "loss": 3.2048,
+      "step": 13258
+    },
+    {
+      "epoch": 0.5609188594635756,
+      "grad_norm": 0.45731955766677856,
+      "learning_rate": 0.001,
+      "loss": 1.5562,
+      "step": 13259
+    },
+    {
+      "epoch": 0.560961164227092,
+      "grad_norm": 0.19741539657115936,
+      "learning_rate": 0.001,
+      "loss": 2.3421,
+      "step": 13260
+    },
+    {
+      "epoch": 0.5610034689906084,
+      "grad_norm": 0.24196267127990723,
+      "learning_rate": 0.001,
+      "loss": 3.386,
+      "step": 13261
+    },
+    {
+      "epoch": 0.5610457737541247,
+      "grad_norm": 0.24868476390838623,
+      "learning_rate": 0.001,
+      "loss": 1.7664,
+      "step": 13262
+    },
+    {
+      "epoch": 0.5610880785176411,
+      "grad_norm": 0.17724378407001495,
+      "learning_rate": 0.001,
+      "loss": 2.038,
+      "step": 13263
+    },
+    {
+      "epoch": 0.5611303832811575,
+      "grad_norm": 0.17900416254997253,
+      "learning_rate": 0.001,
+      "loss": 2.5304,
+      "step": 13264
+    },
+    {
+      "epoch": 0.5611726880446738,
+      "grad_norm": 0.2126455008983612,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 13265
+    },
+    {
+      "epoch": 0.5612149928081902,
+      "grad_norm": 0.20829962193965912,
+      "learning_rate": 0.001,
+      "loss": 2.4063,
+      "step": 13266
+    },
+    {
+      "epoch": 0.5612572975717066,
+      "grad_norm": 0.22738134860992432,
+      "learning_rate": 0.001,
+      "loss": 1.7318,
+      "step": 13267
+    },
+    {
+      "epoch": 0.5612996023352229,
+      "grad_norm": 0.18738828599452972,
+      "learning_rate": 0.001,
+      "loss": 2.0383,
+      "step": 13268
+    },
+    {
+      "epoch": 0.5613419070987393,
+      "grad_norm": 0.2207876741886139,
+      "learning_rate": 0.001,
+      "loss": 2.9998,
+      "step": 13269
+    },
+    {
+      "epoch": 0.5613842118622557,
+      "grad_norm": 0.1774303913116455,
+      "learning_rate": 0.001,
+      "loss": 2.2946,
+      "step": 13270
+    },
+    {
+      "epoch": 0.561426516625772,
+      "grad_norm": 0.16445757448673248,
+      "learning_rate": 0.001,
+      "loss": 2.1301,
+      "step": 13271
+    },
+    {
+      "epoch": 0.5614688213892884,
+      "grad_norm": 0.1810140758752823,
+      "learning_rate": 0.001,
+      "loss": 1.7674,
+      "step": 13272
+    },
+    {
+      "epoch": 0.5615111261528049,
+      "grad_norm": 0.2058965414762497,
+      "learning_rate": 0.001,
+      "loss": 1.5423,
+      "step": 13273
+    },
+    {
+      "epoch": 0.5615534309163212,
+      "grad_norm": 0.20573607087135315,
+      "learning_rate": 0.001,
+      "loss": 3.1317,
+      "step": 13274
+    },
+    {
+      "epoch": 0.5615957356798376,
+      "grad_norm": 0.2003343552350998,
+      "learning_rate": 0.001,
+      "loss": 3.3971,
+      "step": 13275
+    },
+    {
+      "epoch": 0.561638040443354,
+      "grad_norm": 0.16482414305210114,
+      "learning_rate": 0.001,
+      "loss": 2.5649,
+      "step": 13276
+    },
+    {
+      "epoch": 0.5616803452068703,
+      "grad_norm": 0.16140007972717285,
+      "learning_rate": 0.001,
+      "loss": 2.7057,
+      "step": 13277
+    },
+    {
+      "epoch": 0.5617226499703867,
+      "grad_norm": 2.4391520023345947,
+      "learning_rate": 0.001,
+      "loss": 1.9083,
+      "step": 13278
+    },
+    {
+      "epoch": 0.5617649547339031,
+      "grad_norm": 0.17537763714790344,
+      "learning_rate": 0.001,
+      "loss": 2.0083,
+      "step": 13279
+    },
+    {
+      "epoch": 0.5618072594974194,
+      "grad_norm": 0.17112372815608978,
+      "learning_rate": 0.001,
+      "loss": 1.7442,
+      "step": 13280
+    },
+    {
+      "epoch": 0.5618495642609358,
+      "grad_norm": 0.17031919956207275,
+      "learning_rate": 0.001,
+      "loss": 1.8762,
+      "step": 13281
+    },
+    {
+      "epoch": 0.5618918690244522,
+      "grad_norm": 0.15143336355686188,
+      "learning_rate": 0.001,
+      "loss": 2.3816,
+      "step": 13282
+    },
+    {
+      "epoch": 0.5619341737879685,
+      "grad_norm": 0.15341715514659882,
+      "learning_rate": 0.001,
+      "loss": 2.9387,
+      "step": 13283
+    },
+    {
+      "epoch": 0.5619764785514849,
+      "grad_norm": 0.18952025473117828,
+      "learning_rate": 0.001,
+      "loss": 2.3547,
+      "step": 13284
+    },
+    {
+      "epoch": 0.5620187833150013,
+      "grad_norm": 0.18509429693222046,
+      "learning_rate": 0.001,
+      "loss": 2.503,
+      "step": 13285
+    },
+    {
+      "epoch": 0.5620610880785176,
+      "grad_norm": 0.16015632450580597,
+      "learning_rate": 0.001,
+      "loss": 2.008,
+      "step": 13286
+    },
+    {
+      "epoch": 0.562103392842034,
+      "grad_norm": 0.1822974979877472,
+      "learning_rate": 0.001,
+      "loss": 1.8725,
+      "step": 13287
+    },
+    {
+      "epoch": 0.5621456976055503,
+      "grad_norm": 0.20895791053771973,
+      "learning_rate": 0.001,
+      "loss": 3.7383,
+      "step": 13288
+    },
+    {
+      "epoch": 0.5621880023690667,
+      "grad_norm": 0.2861064076423645,
+      "learning_rate": 0.001,
+      "loss": 1.7759,
+      "step": 13289
+    },
+    {
+      "epoch": 0.5622303071325832,
+      "grad_norm": 0.1581949144601822,
+      "learning_rate": 0.001,
+      "loss": 2.9572,
+      "step": 13290
+    },
+    {
+      "epoch": 0.5622726118960995,
+      "grad_norm": 0.1519918292760849,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 13291
+    },
+    {
+      "epoch": 0.5623149166596159,
+      "grad_norm": 1.3050894737243652,
+      "learning_rate": 0.001,
+      "loss": 1.6955,
+      "step": 13292
+    },
+    {
+      "epoch": 0.5623572214231323,
+      "grad_norm": 0.16207565367221832,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 13293
+    },
+    {
+      "epoch": 0.5623995261866486,
+      "grad_norm": 0.21494990587234497,
+      "learning_rate": 0.001,
+      "loss": 1.8584,
+      "step": 13294
+    },
+    {
+      "epoch": 0.562441830950165,
+      "grad_norm": 0.1507209986448288,
+      "learning_rate": 0.001,
+      "loss": 2.1614,
+      "step": 13295
+    },
+    {
+      "epoch": 0.5624841357136814,
+      "grad_norm": 0.5043923854827881,
+      "learning_rate": 0.001,
+      "loss": 2.4189,
+      "step": 13296
+    },
+    {
+      "epoch": 0.5625264404771977,
+      "grad_norm": 0.15872474014759064,
+      "learning_rate": 0.001,
+      "loss": 2.1197,
+      "step": 13297
+    },
+    {
+      "epoch": 0.5625687452407141,
+      "grad_norm": 0.18348495662212372,
+      "learning_rate": 0.001,
+      "loss": 1.9253,
+      "step": 13298
+    },
+    {
+      "epoch": 0.5626110500042305,
+      "grad_norm": 0.22080762684345245,
+      "learning_rate": 0.001,
+      "loss": 2.7428,
+      "step": 13299
+    },
+    {
+      "epoch": 0.5626533547677468,
+      "grad_norm": 0.21116049587726593,
+      "learning_rate": 0.001,
+      "loss": 2.4198,
+      "step": 13300
+    },
+    {
+      "epoch": 0.5626956595312632,
+      "grad_norm": 0.26892927289009094,
+      "learning_rate": 0.001,
+      "loss": 2.1072,
+      "step": 13301
+    },
+    {
+      "epoch": 0.5627379642947796,
+      "grad_norm": 0.17305731773376465,
+      "learning_rate": 0.001,
+      "loss": 2.569,
+      "step": 13302
+    },
+    {
+      "epoch": 0.5627802690582959,
+      "grad_norm": 0.18649710714817047,
+      "learning_rate": 0.001,
+      "loss": 3.0606,
+      "step": 13303
+    },
+    {
+      "epoch": 0.5628225738218123,
+      "grad_norm": 0.15958692133426666,
+      "learning_rate": 0.001,
+      "loss": 2.2803,
+      "step": 13304
+    },
+    {
+      "epoch": 0.5628648785853287,
+      "grad_norm": 0.20183780789375305,
+      "learning_rate": 0.001,
+      "loss": 3.1339,
+      "step": 13305
+    },
+    {
+      "epoch": 0.562907183348845,
+      "grad_norm": 0.23811282217502594,
+      "learning_rate": 0.001,
+      "loss": 2.0551,
+      "step": 13306
+    },
+    {
+      "epoch": 0.5629494881123615,
+      "grad_norm": 0.2011602222919464,
+      "learning_rate": 0.001,
+      "loss": 2.1267,
+      "step": 13307
+    },
+    {
+      "epoch": 0.5629917928758779,
+      "grad_norm": 0.18026520311832428,
+      "learning_rate": 0.001,
+      "loss": 2.5384,
+      "step": 13308
+    },
+    {
+      "epoch": 0.5630340976393942,
+      "grad_norm": 0.19247384369373322,
+      "learning_rate": 0.001,
+      "loss": 2.0309,
+      "step": 13309
+    },
+    {
+      "epoch": 0.5630764024029106,
+      "grad_norm": 0.23704922199249268,
+      "learning_rate": 0.001,
+      "loss": 2.4209,
+      "step": 13310
+    },
+    {
+      "epoch": 0.563118707166427,
+      "grad_norm": 0.19546206295490265,
+      "learning_rate": 0.001,
+      "loss": 2.3817,
+      "step": 13311
+    },
+    {
+      "epoch": 0.5631610119299433,
+      "grad_norm": 0.1445315182209015,
+      "learning_rate": 0.001,
+      "loss": 1.8728,
+      "step": 13312
+    },
+    {
+      "epoch": 0.5632033166934597,
+      "grad_norm": 0.14706316590309143,
+      "learning_rate": 0.001,
+      "loss": 1.7439,
+      "step": 13313
+    },
+    {
+      "epoch": 0.5632456214569761,
+      "grad_norm": 0.19955837726593018,
+      "learning_rate": 0.001,
+      "loss": 2.1125,
+      "step": 13314
+    },
+    {
+      "epoch": 0.5632879262204924,
+      "grad_norm": 0.19193534553050995,
+      "learning_rate": 0.001,
+      "loss": 3.1798,
+      "step": 13315
+    },
+    {
+      "epoch": 0.5633302309840088,
+      "grad_norm": 0.19683898985385895,
+      "learning_rate": 0.001,
+      "loss": 1.9713,
+      "step": 13316
+    },
+    {
+      "epoch": 0.5633725357475252,
+      "grad_norm": 0.182073175907135,
+      "learning_rate": 0.001,
+      "loss": 3.2368,
+      "step": 13317
+    },
+    {
+      "epoch": 0.5634148405110415,
+      "grad_norm": 0.3637740910053253,
+      "learning_rate": 0.001,
+      "loss": 2.0353,
+      "step": 13318
+    },
+    {
+      "epoch": 0.5634571452745579,
+      "grad_norm": 0.1640586256980896,
+      "learning_rate": 0.001,
+      "loss": 1.5906,
+      "step": 13319
+    },
+    {
+      "epoch": 0.5634994500380743,
+      "grad_norm": 0.2922709584236145,
+      "learning_rate": 0.001,
+      "loss": 2.5071,
+      "step": 13320
+    },
+    {
+      "epoch": 0.5635417548015906,
+      "grad_norm": 0.1738395392894745,
+      "learning_rate": 0.001,
+      "loss": 1.973,
+      "step": 13321
+    },
+    {
+      "epoch": 0.563584059565107,
+      "grad_norm": 0.18015141785144806,
+      "learning_rate": 0.001,
+      "loss": 1.9577,
+      "step": 13322
+    },
+    {
+      "epoch": 0.5636263643286235,
+      "grad_norm": 0.17254801094532013,
+      "learning_rate": 0.001,
+      "loss": 2.3277,
+      "step": 13323
+    },
+    {
+      "epoch": 0.5636686690921398,
+      "grad_norm": 0.20676442980766296,
+      "learning_rate": 0.001,
+      "loss": 1.898,
+      "step": 13324
+    },
+    {
+      "epoch": 0.5637109738556562,
+      "grad_norm": 0.19672460854053497,
+      "learning_rate": 0.001,
+      "loss": 2.5557,
+      "step": 13325
+    },
+    {
+      "epoch": 0.5637532786191726,
+      "grad_norm": 0.16016629338264465,
+      "learning_rate": 0.001,
+      "loss": 1.7529,
+      "step": 13326
+    },
+    {
+      "epoch": 0.5637955833826889,
+      "grad_norm": 0.1732088327407837,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 13327
+    },
+    {
+      "epoch": 0.5638378881462053,
+      "grad_norm": 0.1827797144651413,
+      "learning_rate": 0.001,
+      "loss": 1.7933,
+      "step": 13328
+    },
+    {
+      "epoch": 0.5638801929097217,
+      "grad_norm": 0.16698625683784485,
+      "learning_rate": 0.001,
+      "loss": 2.7892,
+      "step": 13329
+    },
+    {
+      "epoch": 0.563922497673238,
+      "grad_norm": 0.1709325909614563,
+      "learning_rate": 0.001,
+      "loss": 2.5153,
+      "step": 13330
+    },
+    {
+      "epoch": 0.5639648024367544,
+      "grad_norm": 0.25048211216926575,
+      "learning_rate": 0.001,
+      "loss": 2.6259,
+      "step": 13331
+    },
+    {
+      "epoch": 0.5640071072002707,
+      "grad_norm": 0.17622262239456177,
+      "learning_rate": 0.001,
+      "loss": 2.1055,
+      "step": 13332
+    },
+    {
+      "epoch": 0.5640494119637871,
+      "grad_norm": 0.166434183716774,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 13333
+    },
+    {
+      "epoch": 0.5640917167273035,
+      "grad_norm": 0.7903834581375122,
+      "learning_rate": 0.001,
+      "loss": 1.9703,
+      "step": 13334
+    },
+    {
+      "epoch": 0.5641340214908198,
+      "grad_norm": 0.15010859072208405,
+      "learning_rate": 0.001,
+      "loss": 1.7071,
+      "step": 13335
+    },
+    {
+      "epoch": 0.5641763262543362,
+      "grad_norm": 0.1890597641468048,
+      "learning_rate": 0.001,
+      "loss": 2.995,
+      "step": 13336
+    },
+    {
+      "epoch": 0.5642186310178526,
+      "grad_norm": 0.19205980002880096,
+      "learning_rate": 0.001,
+      "loss": 2.0444,
+      "step": 13337
+    },
+    {
+      "epoch": 0.5642609357813689,
+      "grad_norm": 0.1756296306848526,
+      "learning_rate": 0.001,
+      "loss": 2.3472,
+      "step": 13338
+    },
+    {
+      "epoch": 0.5643032405448853,
+      "grad_norm": 0.22155724465847015,
+      "learning_rate": 0.001,
+      "loss": 3.1698,
+      "step": 13339
+    },
+    {
+      "epoch": 0.5643455453084018,
+      "grad_norm": 0.18253694474697113,
+      "learning_rate": 0.001,
+      "loss": 1.7616,
+      "step": 13340
+    },
+    {
+      "epoch": 0.564387850071918,
+      "grad_norm": 0.22829587757587433,
+      "learning_rate": 0.001,
+      "loss": 1.5545,
+      "step": 13341
+    },
+    {
+      "epoch": 0.5644301548354345,
+      "grad_norm": 2.0366973876953125,
+      "learning_rate": 0.001,
+      "loss": 1.7702,
+      "step": 13342
+    },
+    {
+      "epoch": 0.5644724595989509,
+      "grad_norm": 0.7380493879318237,
+      "learning_rate": 0.001,
+      "loss": 2.1662,
+      "step": 13343
+    },
+    {
+      "epoch": 0.5645147643624672,
+      "grad_norm": 0.19614994525909424,
+      "learning_rate": 0.001,
+      "loss": 2.5599,
+      "step": 13344
+    },
+    {
+      "epoch": 0.5645570691259836,
+      "grad_norm": 0.39196836948394775,
+      "learning_rate": 0.001,
+      "loss": 1.6906,
+      "step": 13345
+    },
+    {
+      "epoch": 0.5645993738895,
+      "grad_norm": 0.3456530272960663,
+      "learning_rate": 0.001,
+      "loss": 1.9598,
+      "step": 13346
+    },
+    {
+      "epoch": 0.5646416786530163,
+      "grad_norm": 0.20277409255504608,
+      "learning_rate": 0.001,
+      "loss": 2.0331,
+      "step": 13347
+    },
+    {
+      "epoch": 0.5646839834165327,
+      "grad_norm": 0.4468050003051758,
+      "learning_rate": 0.001,
+      "loss": 1.8206,
+      "step": 13348
+    },
+    {
+      "epoch": 0.5647262881800491,
+      "grad_norm": 0.3284309208393097,
+      "learning_rate": 0.001,
+      "loss": 2.2595,
+      "step": 13349
+    },
+    {
+      "epoch": 0.5647685929435654,
+      "grad_norm": 0.16732022166252136,
+      "learning_rate": 0.001,
+      "loss": 1.9984,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5648108977070818,
+      "grad_norm": 3.110480785369873,
+      "learning_rate": 0.001,
+      "loss": 2.6128,
+      "step": 13351
+    },
+    {
+      "epoch": 0.5648532024705982,
+      "grad_norm": 0.18067099153995514,
+      "learning_rate": 0.001,
+      "loss": 2.0611,
+      "step": 13352
+    },
+    {
+      "epoch": 0.5648955072341145,
+      "grad_norm": 0.17390722036361694,
+      "learning_rate": 0.001,
+      "loss": 2.0928,
+      "step": 13353
+    },
+    {
+      "epoch": 0.5649378119976309,
+      "grad_norm": 0.18560457229614258,
+      "learning_rate": 0.001,
+      "loss": 3.1778,
+      "step": 13354
+    },
+    {
+      "epoch": 0.5649801167611473,
+      "grad_norm": 0.18010523915290833,
+      "learning_rate": 0.001,
+      "loss": 2.5076,
+      "step": 13355
+    },
+    {
+      "epoch": 0.5650224215246636,
+      "grad_norm": 0.18280434608459473,
+      "learning_rate": 0.001,
+      "loss": 2.2038,
+      "step": 13356
+    },
+    {
+      "epoch": 0.56506472628818,
+      "grad_norm": 0.9968026280403137,
+      "learning_rate": 0.001,
+      "loss": 2.715,
+      "step": 13357
+    },
+    {
+      "epoch": 0.5651070310516965,
+      "grad_norm": 1.5389236211776733,
+      "learning_rate": 0.001,
+      "loss": 2.5267,
+      "step": 13358
+    },
+    {
+      "epoch": 0.5651493358152128,
+      "grad_norm": 0.8743659257888794,
+      "learning_rate": 0.001,
+      "loss": 2.5618,
+      "step": 13359
+    },
+    {
+      "epoch": 0.5651916405787292,
+      "grad_norm": 0.20591340959072113,
+      "learning_rate": 0.001,
+      "loss": 2.3677,
+      "step": 13360
+    },
+    {
+      "epoch": 0.5652339453422456,
+      "grad_norm": 0.20366302132606506,
+      "learning_rate": 0.001,
+      "loss": 2.3497,
+      "step": 13361
+    },
+    {
+      "epoch": 0.5652762501057619,
+      "grad_norm": 0.18198545277118683,
+      "learning_rate": 0.001,
+      "loss": 2.1656,
+      "step": 13362
+    },
+    {
+      "epoch": 0.5653185548692783,
+      "grad_norm": 0.1899123340845108,
+      "learning_rate": 0.001,
+      "loss": 1.8184,
+      "step": 13363
+    },
+    {
+      "epoch": 0.5653608596327947,
+      "grad_norm": 0.21096497774124146,
+      "learning_rate": 0.001,
+      "loss": 2.3129,
+      "step": 13364
+    },
+    {
+      "epoch": 0.565403164396311,
+      "grad_norm": 0.17829644680023193,
+      "learning_rate": 0.001,
+      "loss": 2.4071,
+      "step": 13365
+    },
+    {
+      "epoch": 0.5654454691598274,
+      "grad_norm": 0.1842295229434967,
+      "learning_rate": 0.001,
+      "loss": 3.0018,
+      "step": 13366
+    },
+    {
+      "epoch": 0.5654877739233438,
+      "grad_norm": 0.2008989155292511,
+      "learning_rate": 0.001,
+      "loss": 2.1184,
+      "step": 13367
+    },
+    {
+      "epoch": 0.5655300786868601,
+      "grad_norm": 0.22309176623821259,
+      "learning_rate": 0.001,
+      "loss": 2.2204,
+      "step": 13368
+    },
+    {
+      "epoch": 0.5655723834503765,
+      "grad_norm": 0.4550027847290039,
+      "learning_rate": 0.001,
+      "loss": 2.8608,
+      "step": 13369
+    },
+    {
+      "epoch": 0.5656146882138929,
+      "grad_norm": 0.16846232116222382,
+      "learning_rate": 0.001,
+      "loss": 1.7115,
+      "step": 13370
+    },
+    {
+      "epoch": 0.5656569929774092,
+      "grad_norm": 0.9170604944229126,
+      "learning_rate": 0.001,
+      "loss": 2.3162,
+      "step": 13371
+    },
+    {
+      "epoch": 0.5656992977409256,
+      "grad_norm": 0.21079829335212708,
+      "learning_rate": 0.001,
+      "loss": 2.2757,
+      "step": 13372
+    },
+    {
+      "epoch": 0.565741602504442,
+      "grad_norm": 0.15155363082885742,
+      "learning_rate": 0.001,
+      "loss": 1.9875,
+      "step": 13373
+    },
+    {
+      "epoch": 0.5657839072679584,
+      "grad_norm": 0.5003759860992432,
+      "learning_rate": 0.001,
+      "loss": 2.126,
+      "step": 13374
+    },
+    {
+      "epoch": 0.5658262120314748,
+      "grad_norm": 0.2025512456893921,
+      "learning_rate": 0.001,
+      "loss": 1.8468,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5658685167949911,
+      "grad_norm": 0.45572540163993835,
+      "learning_rate": 0.001,
+      "loss": 2.3299,
+      "step": 13376
+    },
+    {
+      "epoch": 0.5659108215585075,
+      "grad_norm": 0.17576457560062408,
+      "learning_rate": 0.001,
+      "loss": 2.3811,
+      "step": 13377
+    },
+    {
+      "epoch": 0.5659531263220239,
+      "grad_norm": 0.28992700576782227,
+      "learning_rate": 0.001,
+      "loss": 2.4098,
+      "step": 13378
+    },
+    {
+      "epoch": 0.5659954310855402,
+      "grad_norm": 0.6396428942680359,
+      "learning_rate": 0.001,
+      "loss": 2.3187,
+      "step": 13379
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.18657970428466797,
+      "learning_rate": 0.001,
+      "loss": 2.3154,
+      "step": 13380
+    },
+    {
+      "epoch": 0.566080040612573,
+      "grad_norm": 0.4546518623828888,
+      "learning_rate": 0.001,
+      "loss": 1.7812,
+      "step": 13381
+    },
+    {
+      "epoch": 0.5661223453760893,
+      "grad_norm": 0.4422479271888733,
+      "learning_rate": 0.001,
+      "loss": 2.0153,
+      "step": 13382
+    },
+    {
+      "epoch": 0.5661646501396057,
+      "grad_norm": 0.19912591576576233,
+      "learning_rate": 0.001,
+      "loss": 1.6749,
+      "step": 13383
+    },
+    {
+      "epoch": 0.5662069549031221,
+      "grad_norm": 0.20137451589107513,
+      "learning_rate": 0.001,
+      "loss": 2.3742,
+      "step": 13384
+    },
+    {
+      "epoch": 0.5662492596666384,
+      "grad_norm": 0.22245512902736664,
+      "learning_rate": 0.001,
+      "loss": 3.0519,
+      "step": 13385
+    },
+    {
+      "epoch": 0.5662915644301548,
+      "grad_norm": 0.179260715842247,
+      "learning_rate": 0.001,
+      "loss": 2.305,
+      "step": 13386
+    },
+    {
+      "epoch": 0.5663338691936712,
+      "grad_norm": 0.19476951658725739,
+      "learning_rate": 0.001,
+      "loss": 2.0913,
+      "step": 13387
+    },
+    {
+      "epoch": 0.5663761739571875,
+      "grad_norm": 0.1848563849925995,
+      "learning_rate": 0.001,
+      "loss": 2.4601,
+      "step": 13388
+    },
+    {
+      "epoch": 0.5664184787207039,
+      "grad_norm": 0.26256316900253296,
+      "learning_rate": 0.001,
+      "loss": 2.8465,
+      "step": 13389
+    },
+    {
+      "epoch": 0.5664607834842204,
+      "grad_norm": 0.1553989052772522,
+      "learning_rate": 0.001,
+      "loss": 1.6542,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5665030882477367,
+      "grad_norm": 0.17743252217769623,
+      "learning_rate": 0.001,
+      "loss": 1.7768,
+      "step": 13391
+    },
+    {
+      "epoch": 0.5665453930112531,
+      "grad_norm": 0.19140616059303284,
+      "learning_rate": 0.001,
+      "loss": 1.461,
+      "step": 13392
+    },
+    {
+      "epoch": 0.5665876977747695,
+      "grad_norm": 0.19345830380916595,
+      "learning_rate": 0.001,
+      "loss": 2.0676,
+      "step": 13393
+    },
+    {
+      "epoch": 0.5666300025382858,
+      "grad_norm": 0.2207154482603073,
+      "learning_rate": 0.001,
+      "loss": 2.4146,
+      "step": 13394
+    },
+    {
+      "epoch": 0.5666723073018022,
+      "grad_norm": 0.1724683791399002,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 13395
+    },
+    {
+      "epoch": 0.5667146120653186,
+      "grad_norm": 0.16311654448509216,
+      "learning_rate": 0.001,
+      "loss": 3.1914,
+      "step": 13396
+    },
+    {
+      "epoch": 0.5667569168288349,
+      "grad_norm": 0.16985690593719482,
+      "learning_rate": 0.001,
+      "loss": 2.3727,
+      "step": 13397
+    },
+    {
+      "epoch": 0.5667992215923513,
+      "grad_norm": 0.1899961233139038,
+      "learning_rate": 0.001,
+      "loss": 2.2874,
+      "step": 13398
+    },
+    {
+      "epoch": 0.5668415263558677,
+      "grad_norm": 0.18617494404315948,
+      "learning_rate": 0.001,
+      "loss": 1.4395,
+      "step": 13399
+    },
+    {
+      "epoch": 0.566883831119384,
+      "grad_norm": 0.18079866468906403,
+      "learning_rate": 0.001,
+      "loss": 3.0455,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5669261358829004,
+      "grad_norm": 0.18550261855125427,
+      "learning_rate": 0.001,
+      "loss": 3.1314,
+      "step": 13401
+    },
+    {
+      "epoch": 0.5669684406464168,
+      "grad_norm": 0.17485053837299347,
+      "learning_rate": 0.001,
+      "loss": 2.1399,
+      "step": 13402
+    },
+    {
+      "epoch": 0.5670107454099331,
+      "grad_norm": 0.15552884340286255,
+      "learning_rate": 0.001,
+      "loss": 1.5707,
+      "step": 13403
+    },
+    {
+      "epoch": 0.5670530501734495,
+      "grad_norm": 0.14861054718494415,
+      "learning_rate": 0.001,
+      "loss": 1.7765,
+      "step": 13404
+    },
+    {
+      "epoch": 0.567095354936966,
+      "grad_norm": 0.4289642572402954,
+      "learning_rate": 0.001,
+      "loss": 1.7516,
+      "step": 13405
+    },
+    {
+      "epoch": 0.5671376597004822,
+      "grad_norm": 0.14437785744667053,
+      "learning_rate": 0.001,
+      "loss": 2.3381,
+      "step": 13406
+    },
+    {
+      "epoch": 0.5671799644639987,
+      "grad_norm": 0.1592496633529663,
+      "learning_rate": 0.001,
+      "loss": 1.7272,
+      "step": 13407
+    },
+    {
+      "epoch": 0.5672222692275151,
+      "grad_norm": 0.2676011323928833,
+      "learning_rate": 0.001,
+      "loss": 2.5001,
+      "step": 13408
+    },
+    {
+      "epoch": 0.5672645739910314,
+      "grad_norm": 0.1811024695634842,
+      "learning_rate": 0.001,
+      "loss": 2.3746,
+      "step": 13409
+    },
+    {
+      "epoch": 0.5673068787545478,
+      "grad_norm": 0.16653041541576385,
+      "learning_rate": 0.001,
+      "loss": 1.9266,
+      "step": 13410
+    },
+    {
+      "epoch": 0.5673491835180642,
+      "grad_norm": 0.18817272782325745,
+      "learning_rate": 0.001,
+      "loss": 2.2231,
+      "step": 13411
+    },
+    {
+      "epoch": 0.5673914882815805,
+      "grad_norm": 0.381883442401886,
+      "learning_rate": 0.001,
+      "loss": 1.8577,
+      "step": 13412
+    },
+    {
+      "epoch": 0.5674337930450969,
+      "grad_norm": 0.1743682324886322,
+      "learning_rate": 0.001,
+      "loss": 2.6966,
+      "step": 13413
+    },
+    {
+      "epoch": 0.5674760978086133,
+      "grad_norm": 0.18651573359966278,
+      "learning_rate": 0.001,
+      "loss": 2.3024,
+      "step": 13414
+    },
+    {
+      "epoch": 0.5675184025721296,
+      "grad_norm": 2.339829206466675,
+      "learning_rate": 0.001,
+      "loss": 2.0363,
+      "step": 13415
+    },
+    {
+      "epoch": 0.567560707335646,
+      "grad_norm": 0.28542375564575195,
+      "learning_rate": 0.001,
+      "loss": 2.1393,
+      "step": 13416
+    },
+    {
+      "epoch": 0.5676030120991624,
+      "grad_norm": 0.17619076371192932,
+      "learning_rate": 0.001,
+      "loss": 1.4569,
+      "step": 13417
+    },
+    {
+      "epoch": 0.5676453168626787,
+      "grad_norm": 0.171681746840477,
+      "learning_rate": 0.001,
+      "loss": 2.2856,
+      "step": 13418
+    },
+    {
+      "epoch": 0.5676876216261951,
+      "grad_norm": 0.23256142437458038,
+      "learning_rate": 0.001,
+      "loss": 1.973,
+      "step": 13419
+    },
+    {
+      "epoch": 0.5677299263897115,
+      "grad_norm": 0.22075031697750092,
+      "learning_rate": 0.001,
+      "loss": 1.9528,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5677722311532278,
+      "grad_norm": 0.12937584519386292,
+      "learning_rate": 0.001,
+      "loss": 2.196,
+      "step": 13421
+    },
+    {
+      "epoch": 0.5678145359167442,
+      "grad_norm": 0.17965421080589294,
+      "learning_rate": 0.001,
+      "loss": 1.5309,
+      "step": 13422
+    },
+    {
+      "epoch": 0.5678568406802605,
+      "grad_norm": 2.4008188247680664,
+      "learning_rate": 0.001,
+      "loss": 2.3106,
+      "step": 13423
+    },
+    {
+      "epoch": 0.567899145443777,
+      "grad_norm": 0.16604921221733093,
+      "learning_rate": 0.001,
+      "loss": 2.2181,
+      "step": 13424
+    },
+    {
+      "epoch": 0.5679414502072934,
+      "grad_norm": 0.2204952985048294,
+      "learning_rate": 0.001,
+      "loss": 2.9946,
+      "step": 13425
+    },
+    {
+      "epoch": 0.5679837549708097,
+      "grad_norm": 0.2380129098892212,
+      "learning_rate": 0.001,
+      "loss": 1.8916,
+      "step": 13426
+    },
+    {
+      "epoch": 0.5680260597343261,
+      "grad_norm": 0.20674805343151093,
+      "learning_rate": 0.001,
+      "loss": 2.224,
+      "step": 13427
+    },
+    {
+      "epoch": 0.5680683644978425,
+      "grad_norm": 0.2611851990222931,
+      "learning_rate": 0.001,
+      "loss": 2.041,
+      "step": 13428
+    },
+    {
+      "epoch": 0.5681106692613588,
+      "grad_norm": 4.238038063049316,
+      "learning_rate": 0.001,
+      "loss": 3.423,
+      "step": 13429
+    },
+    {
+      "epoch": 0.5681529740248752,
+      "grad_norm": 0.26452019810676575,
+      "learning_rate": 0.001,
+      "loss": 2.9858,
+      "step": 13430
+    },
+    {
+      "epoch": 0.5681952787883916,
+      "grad_norm": 0.2289876937866211,
+      "learning_rate": 0.001,
+      "loss": 2.6998,
+      "step": 13431
+    },
+    {
+      "epoch": 0.5682375835519079,
+      "grad_norm": 0.3570123314857483,
+      "learning_rate": 0.001,
+      "loss": 2.1574,
+      "step": 13432
+    },
+    {
+      "epoch": 0.5682798883154243,
+      "grad_norm": 0.33323732018470764,
+      "learning_rate": 0.001,
+      "loss": 1.9346,
+      "step": 13433
+    },
+    {
+      "epoch": 0.5683221930789407,
+      "grad_norm": 0.2315545380115509,
+      "learning_rate": 0.001,
+      "loss": 1.9166,
+      "step": 13434
+    },
+    {
+      "epoch": 0.568364497842457,
+      "grad_norm": 0.42233508825302124,
+      "learning_rate": 0.001,
+      "loss": 2.1505,
+      "step": 13435
+    },
+    {
+      "epoch": 0.5684068026059734,
+      "grad_norm": 0.28535693883895874,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 13436
+    },
+    {
+      "epoch": 0.5684491073694898,
+      "grad_norm": 0.6629549264907837,
+      "learning_rate": 0.001,
+      "loss": 2.3038,
+      "step": 13437
+    },
+    {
+      "epoch": 0.5684914121330061,
+      "grad_norm": 0.18485631048679352,
+      "learning_rate": 0.001,
+      "loss": 2.8315,
+      "step": 13438
+    },
+    {
+      "epoch": 0.5685337168965225,
+      "grad_norm": 1.2029775381088257,
+      "learning_rate": 0.001,
+      "loss": 2.7915,
+      "step": 13439
+    },
+    {
+      "epoch": 0.568576021660039,
+      "grad_norm": 0.16780279576778412,
+      "learning_rate": 0.001,
+      "loss": 2.0964,
+      "step": 13440
+    },
+    {
+      "epoch": 0.5686183264235553,
+      "grad_norm": 0.4147586524486542,
+      "learning_rate": 0.001,
+      "loss": 2.8658,
+      "step": 13441
+    },
+    {
+      "epoch": 0.5686606311870717,
+      "grad_norm": 0.21042154729366302,
+      "learning_rate": 0.001,
+      "loss": 2.2641,
+      "step": 13442
+    },
+    {
+      "epoch": 0.5687029359505881,
+      "grad_norm": 0.2242194414138794,
+      "learning_rate": 0.001,
+      "loss": 2.0489,
+      "step": 13443
+    },
+    {
+      "epoch": 0.5687452407141044,
+      "grad_norm": 0.27875545620918274,
+      "learning_rate": 0.001,
+      "loss": 1.592,
+      "step": 13444
+    },
+    {
+      "epoch": 0.5687875454776208,
+      "grad_norm": 0.1999659240245819,
+      "learning_rate": 0.001,
+      "loss": 2.1463,
+      "step": 13445
+    },
+    {
+      "epoch": 0.5688298502411372,
+      "grad_norm": 0.15846888720989227,
+      "learning_rate": 0.001,
+      "loss": 2.1381,
+      "step": 13446
+    },
+    {
+      "epoch": 0.5688721550046535,
+      "grad_norm": 0.21760839223861694,
+      "learning_rate": 0.001,
+      "loss": 2.4506,
+      "step": 13447
+    },
+    {
+      "epoch": 0.5689144597681699,
+      "grad_norm": 0.15946030616760254,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 13448
+    },
+    {
+      "epoch": 0.5689567645316863,
+      "grad_norm": 0.21263648569583893,
+      "learning_rate": 0.001,
+      "loss": 1.999,
+      "step": 13449
+    },
+    {
+      "epoch": 0.5689990692952026,
+      "grad_norm": 0.1755702942609787,
+      "learning_rate": 0.001,
+      "loss": 1.845,
+      "step": 13450
+    },
+    {
+      "epoch": 0.569041374058719,
+      "grad_norm": 0.14761324226856232,
+      "learning_rate": 0.001,
+      "loss": 1.5729,
+      "step": 13451
+    },
+    {
+      "epoch": 0.5690836788222354,
+      "grad_norm": 0.16154663264751434,
+      "learning_rate": 0.001,
+      "loss": 1.4929,
+      "step": 13452
+    },
+    {
+      "epoch": 0.5691259835857517,
+      "grad_norm": 0.15619386732578278,
+      "learning_rate": 0.001,
+      "loss": 2.8627,
+      "step": 13453
+    },
+    {
+      "epoch": 0.5691682883492681,
+      "grad_norm": 0.2875896394252777,
+      "learning_rate": 0.001,
+      "loss": 3.1241,
+      "step": 13454
+    },
+    {
+      "epoch": 0.5692105931127845,
+      "grad_norm": 0.20529961585998535,
+      "learning_rate": 0.001,
+      "loss": 2.3149,
+      "step": 13455
+    },
+    {
+      "epoch": 0.5692528978763008,
+      "grad_norm": 0.134473517537117,
+      "learning_rate": 0.001,
+      "loss": 1.909,
+      "step": 13456
+    },
+    {
+      "epoch": 0.5692952026398173,
+      "grad_norm": 0.15408681333065033,
+      "learning_rate": 0.001,
+      "loss": 2.2475,
+      "step": 13457
+    },
+    {
+      "epoch": 0.5693375074033337,
+      "grad_norm": 0.17341335117816925,
+      "learning_rate": 0.001,
+      "loss": 2.7923,
+      "step": 13458
+    },
+    {
+      "epoch": 0.56937981216685,
+      "grad_norm": 0.1393430083990097,
+      "learning_rate": 0.001,
+      "loss": 2.5653,
+      "step": 13459
+    },
+    {
+      "epoch": 0.5694221169303664,
+      "grad_norm": 0.20593148469924927,
+      "learning_rate": 0.001,
+      "loss": 2.7352,
+      "step": 13460
+    },
+    {
+      "epoch": 0.5694644216938828,
+      "grad_norm": 0.19775381684303284,
+      "learning_rate": 0.001,
+      "loss": 1.8592,
+      "step": 13461
+    },
+    {
+      "epoch": 0.5695067264573991,
+      "grad_norm": 0.17005833983421326,
+      "learning_rate": 0.001,
+      "loss": 1.5063,
+      "step": 13462
+    },
+    {
+      "epoch": 0.5695490312209155,
+      "grad_norm": 0.1618228405714035,
+      "learning_rate": 0.001,
+      "loss": 2.7133,
+      "step": 13463
+    },
+    {
+      "epoch": 0.5695913359844319,
+      "grad_norm": 0.19688403606414795,
+      "learning_rate": 0.001,
+      "loss": 3.1128,
+      "step": 13464
+    },
+    {
+      "epoch": 0.5696336407479482,
+      "grad_norm": 0.184058278799057,
+      "learning_rate": 0.001,
+      "loss": 2.203,
+      "step": 13465
+    },
+    {
+      "epoch": 0.5696759455114646,
+      "grad_norm": 0.4234599769115448,
+      "learning_rate": 0.001,
+      "loss": 1.4275,
+      "step": 13466
+    },
+    {
+      "epoch": 0.5697182502749809,
+      "grad_norm": 0.610440194606781,
+      "learning_rate": 0.001,
+      "loss": 1.8219,
+      "step": 13467
+    },
+    {
+      "epoch": 0.5697605550384973,
+      "grad_norm": 0.6861364841461182,
+      "learning_rate": 0.001,
+      "loss": 1.8298,
+      "step": 13468
+    },
+    {
+      "epoch": 0.5698028598020137,
+      "grad_norm": 0.17145289480686188,
+      "learning_rate": 0.001,
+      "loss": 2.5483,
+      "step": 13469
+    },
+    {
+      "epoch": 0.56984516456553,
+      "grad_norm": 0.2071324735879898,
+      "learning_rate": 0.001,
+      "loss": 1.5715,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5698874693290464,
+      "grad_norm": 0.1772976666688919,
+      "learning_rate": 0.001,
+      "loss": 2.9208,
+      "step": 13471
+    },
+    {
+      "epoch": 0.5699297740925628,
+      "grad_norm": 0.14574064314365387,
+      "learning_rate": 0.001,
+      "loss": 2.1653,
+      "step": 13472
+    },
+    {
+      "epoch": 0.5699720788560791,
+      "grad_norm": 0.16666589677333832,
+      "learning_rate": 0.001,
+      "loss": 2.4713,
+      "step": 13473
+    },
+    {
+      "epoch": 0.5700143836195956,
+      "grad_norm": 0.19138103723526,
+      "learning_rate": 0.001,
+      "loss": 2.0317,
+      "step": 13474
+    },
+    {
+      "epoch": 0.570056688383112,
+      "grad_norm": 0.21919706463813782,
+      "learning_rate": 0.001,
+      "loss": 1.5688,
+      "step": 13475
+    },
+    {
+      "epoch": 0.5700989931466283,
+      "grad_norm": 0.6911723613739014,
+      "learning_rate": 0.001,
+      "loss": 2.5346,
+      "step": 13476
+    },
+    {
+      "epoch": 0.5701412979101447,
+      "grad_norm": 1.9358234405517578,
+      "learning_rate": 0.001,
+      "loss": 2.5255,
+      "step": 13477
+    },
+    {
+      "epoch": 0.5701836026736611,
+      "grad_norm": 0.43052294850349426,
+      "learning_rate": 0.001,
+      "loss": 2.2194,
+      "step": 13478
+    },
+    {
+      "epoch": 0.5702259074371774,
+      "grad_norm": 0.1567905694246292,
+      "learning_rate": 0.001,
+      "loss": 1.5455,
+      "step": 13479
+    },
+    {
+      "epoch": 0.5702682122006938,
+      "grad_norm": 0.16955433785915375,
+      "learning_rate": 0.001,
+      "loss": 2.2044,
+      "step": 13480
+    },
+    {
+      "epoch": 0.5703105169642102,
+      "grad_norm": 0.21759982407093048,
+      "learning_rate": 0.001,
+      "loss": 1.6764,
+      "step": 13481
+    },
+    {
+      "epoch": 0.5703528217277265,
+      "grad_norm": 0.26957741379737854,
+      "learning_rate": 0.001,
+      "loss": 3.0424,
+      "step": 13482
+    },
+    {
+      "epoch": 0.5703951264912429,
+      "grad_norm": 0.23791447281837463,
+      "learning_rate": 0.001,
+      "loss": 2.1794,
+      "step": 13483
+    },
+    {
+      "epoch": 0.5704374312547593,
+      "grad_norm": 0.17288313806056976,
+      "learning_rate": 0.001,
+      "loss": 2.8003,
+      "step": 13484
+    },
+    {
+      "epoch": 0.5704797360182756,
+      "grad_norm": 0.186492919921875,
+      "learning_rate": 0.001,
+      "loss": 2.5023,
+      "step": 13485
+    },
+    {
+      "epoch": 0.570522040781792,
+      "grad_norm": 0.17396779358386993,
+      "learning_rate": 0.001,
+      "loss": 2.3465,
+      "step": 13486
+    },
+    {
+      "epoch": 0.5705643455453084,
+      "grad_norm": 0.20612338185310364,
+      "learning_rate": 0.001,
+      "loss": 1.9371,
+      "step": 13487
+    },
+    {
+      "epoch": 0.5706066503088247,
+      "grad_norm": 0.16367042064666748,
+      "learning_rate": 0.001,
+      "loss": 1.9677,
+      "step": 13488
+    },
+    {
+      "epoch": 0.5706489550723411,
+      "grad_norm": 1.0515310764312744,
+      "learning_rate": 0.001,
+      "loss": 1.7665,
+      "step": 13489
+    },
+    {
+      "epoch": 0.5706912598358576,
+      "grad_norm": 0.4732486307621002,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5707335645993739,
+      "grad_norm": 0.14785124361515045,
+      "learning_rate": 0.001,
+      "loss": 1.9295,
+      "step": 13491
+    },
+    {
+      "epoch": 0.5707758693628903,
+      "grad_norm": 0.1713053435087204,
+      "learning_rate": 0.001,
+      "loss": 1.8555,
+      "step": 13492
+    },
+    {
+      "epoch": 0.5708181741264067,
+      "grad_norm": 0.17723537981510162,
+      "learning_rate": 0.001,
+      "loss": 2.1308,
+      "step": 13493
+    },
+    {
+      "epoch": 0.570860478889923,
+      "grad_norm": 1.3990591764450073,
+      "learning_rate": 0.001,
+      "loss": 2.5823,
+      "step": 13494
+    },
+    {
+      "epoch": 0.5709027836534394,
+      "grad_norm": 0.6561245918273926,
+      "learning_rate": 0.001,
+      "loss": 2.6915,
+      "step": 13495
+    },
+    {
+      "epoch": 0.5709450884169558,
+      "grad_norm": 1.7256704568862915,
+      "learning_rate": 0.001,
+      "loss": 1.4257,
+      "step": 13496
+    },
+    {
+      "epoch": 0.5709873931804721,
+      "grad_norm": 0.22391989827156067,
+      "learning_rate": 0.001,
+      "loss": 2.3335,
+      "step": 13497
+    },
+    {
+      "epoch": 0.5710296979439885,
+      "grad_norm": 0.1710107922554016,
+      "learning_rate": 0.001,
+      "loss": 1.8255,
+      "step": 13498
+    },
+    {
+      "epoch": 0.5710720027075049,
+      "grad_norm": 0.21606822311878204,
+      "learning_rate": 0.001,
+      "loss": 2.894,
+      "step": 13499
+    },
+    {
+      "epoch": 0.5711143074710212,
+      "grad_norm": 0.19133202731609344,
+      "learning_rate": 0.001,
+      "loss": 1.5382,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5711566122345376,
+      "grad_norm": 0.2007291465997696,
+      "learning_rate": 0.001,
+      "loss": 2.0458,
+      "step": 13501
+    },
+    {
+      "epoch": 0.571198916998054,
+      "grad_norm": 0.21422770619392395,
+      "learning_rate": 0.001,
+      "loss": 2.758,
+      "step": 13502
+    },
+    {
+      "epoch": 0.5712412217615703,
+      "grad_norm": 0.1938318908214569,
+      "learning_rate": 0.001,
+      "loss": 2.7117,
+      "step": 13503
+    },
+    {
+      "epoch": 0.5712835265250867,
+      "grad_norm": 0.26207903027534485,
+      "learning_rate": 0.001,
+      "loss": 1.8618,
+      "step": 13504
+    },
+    {
+      "epoch": 0.5713258312886031,
+      "grad_norm": 0.2131361961364746,
+      "learning_rate": 0.001,
+      "loss": 1.9149,
+      "step": 13505
+    },
+    {
+      "epoch": 0.5713681360521194,
+      "grad_norm": 0.1706126481294632,
+      "learning_rate": 0.001,
+      "loss": 2.3218,
+      "step": 13506
+    },
+    {
+      "epoch": 0.5714104408156359,
+      "grad_norm": 0.20523306727409363,
+      "learning_rate": 0.001,
+      "loss": 2.2526,
+      "step": 13507
+    },
+    {
+      "epoch": 0.5714527455791523,
+      "grad_norm": 0.25251221656799316,
+      "learning_rate": 0.001,
+      "loss": 2.3534,
+      "step": 13508
+    },
+    {
+      "epoch": 0.5714950503426686,
+      "grad_norm": 0.3562532663345337,
+      "learning_rate": 0.001,
+      "loss": 1.6221,
+      "step": 13509
+    },
+    {
+      "epoch": 0.571537355106185,
+      "grad_norm": 0.19563472270965576,
+      "learning_rate": 0.001,
+      "loss": 2.2984,
+      "step": 13510
+    },
+    {
+      "epoch": 0.5715796598697013,
+      "grad_norm": 0.19130732119083405,
+      "learning_rate": 0.001,
+      "loss": 3.6665,
+      "step": 13511
+    },
+    {
+      "epoch": 0.5716219646332177,
+      "grad_norm": 0.2187919318675995,
+      "learning_rate": 0.001,
+      "loss": 2.7956,
+      "step": 13512
+    },
+    {
+      "epoch": 0.5716642693967341,
+      "grad_norm": 0.7064900994300842,
+      "learning_rate": 0.001,
+      "loss": 1.7473,
+      "step": 13513
+    },
+    {
+      "epoch": 0.5717065741602504,
+      "grad_norm": 0.17418193817138672,
+      "learning_rate": 0.001,
+      "loss": 2.1856,
+      "step": 13514
+    },
+    {
+      "epoch": 0.5717488789237668,
+      "grad_norm": 0.1919695883989334,
+      "learning_rate": 0.001,
+      "loss": 2.6144,
+      "step": 13515
+    },
+    {
+      "epoch": 0.5717911836872832,
+      "grad_norm": 0.1946498155593872,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 13516
+    },
+    {
+      "epoch": 0.5718334884507995,
+      "grad_norm": 0.18168438971042633,
+      "learning_rate": 0.001,
+      "loss": 2.7927,
+      "step": 13517
+    },
+    {
+      "epoch": 0.5718757932143159,
+      "grad_norm": 0.410906583070755,
+      "learning_rate": 0.001,
+      "loss": 2.8412,
+      "step": 13518
+    },
+    {
+      "epoch": 0.5719180979778323,
+      "grad_norm": 0.3952532708644867,
+      "learning_rate": 0.001,
+      "loss": 1.6598,
+      "step": 13519
+    },
+    {
+      "epoch": 0.5719604027413486,
+      "grad_norm": 0.7694887518882751,
+      "learning_rate": 0.001,
+      "loss": 2.2268,
+      "step": 13520
+    },
+    {
+      "epoch": 0.572002707504865,
+      "grad_norm": 0.2468140423297882,
+      "learning_rate": 0.001,
+      "loss": 2.8834,
+      "step": 13521
+    },
+    {
+      "epoch": 0.5720450122683814,
+      "grad_norm": 0.23892976343631744,
+      "learning_rate": 0.001,
+      "loss": 1.4797,
+      "step": 13522
+    },
+    {
+      "epoch": 0.5720873170318977,
+      "grad_norm": 0.2801859974861145,
+      "learning_rate": 0.001,
+      "loss": 2.2426,
+      "step": 13523
+    },
+    {
+      "epoch": 0.5721296217954142,
+      "grad_norm": 0.1985364854335785,
+      "learning_rate": 0.001,
+      "loss": 2.5463,
+      "step": 13524
+    },
+    {
+      "epoch": 0.5721719265589306,
+      "grad_norm": 0.24735181033611298,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 13525
+    },
+    {
+      "epoch": 0.5722142313224469,
+      "grad_norm": 0.18506476283073425,
+      "learning_rate": 0.001,
+      "loss": 1.8858,
+      "step": 13526
+    },
+    {
+      "epoch": 0.5722565360859633,
+      "grad_norm": 0.2307346612215042,
+      "learning_rate": 0.001,
+      "loss": 2.1989,
+      "step": 13527
+    },
+    {
+      "epoch": 0.5722988408494797,
+      "grad_norm": 0.16953586041927338,
+      "learning_rate": 0.001,
+      "loss": 2.0845,
+      "step": 13528
+    },
+    {
+      "epoch": 0.572341145612996,
+      "grad_norm": 0.18418249487876892,
+      "learning_rate": 0.001,
+      "loss": 2.2964,
+      "step": 13529
+    },
+    {
+      "epoch": 0.5723834503765124,
+      "grad_norm": 0.2580213248729706,
+      "learning_rate": 0.001,
+      "loss": 2.1883,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5724257551400288,
+      "grad_norm": 2.6409804821014404,
+      "learning_rate": 0.001,
+      "loss": 2.2781,
+      "step": 13531
+    },
+    {
+      "epoch": 0.5724680599035451,
+      "grad_norm": 0.15479686856269836,
+      "learning_rate": 0.001,
+      "loss": 2.2066,
+      "step": 13532
+    },
+    {
+      "epoch": 0.5725103646670615,
+      "grad_norm": 0.8504530787467957,
+      "learning_rate": 0.001,
+      "loss": 1.5421,
+      "step": 13533
+    },
+    {
+      "epoch": 0.5725526694305779,
+      "grad_norm": 0.17100943624973297,
+      "learning_rate": 0.001,
+      "loss": 3.2645,
+      "step": 13534
+    },
+    {
+      "epoch": 0.5725949741940942,
+      "grad_norm": 0.6614644527435303,
+      "learning_rate": 0.001,
+      "loss": 3.2304,
+      "step": 13535
+    },
+    {
+      "epoch": 0.5726372789576106,
+      "grad_norm": 1.6653027534484863,
+      "learning_rate": 0.001,
+      "loss": 2.8675,
+      "step": 13536
+    },
+    {
+      "epoch": 0.572679583721127,
+      "grad_norm": 0.16103801131248474,
+      "learning_rate": 0.001,
+      "loss": 1.9109,
+      "step": 13537
+    },
+    {
+      "epoch": 0.5727218884846433,
+      "grad_norm": 0.16726185381412506,
+      "learning_rate": 0.001,
+      "loss": 3.5696,
+      "step": 13538
+    },
+    {
+      "epoch": 0.5727641932481597,
+      "grad_norm": 0.3862936496734619,
+      "learning_rate": 0.001,
+      "loss": 2.0125,
+      "step": 13539
+    },
+    {
+      "epoch": 0.5728064980116762,
+      "grad_norm": 1.980115532875061,
+      "learning_rate": 0.001,
+      "loss": 2.5362,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5728488027751925,
+      "grad_norm": 0.1421581506729126,
+      "learning_rate": 0.001,
+      "loss": 2.1459,
+      "step": 13541
+    },
+    {
+      "epoch": 0.5728911075387089,
+      "grad_norm": 0.18831764161586761,
+      "learning_rate": 0.001,
+      "loss": 2.523,
+      "step": 13542
+    },
+    {
+      "epoch": 0.5729334123022253,
+      "grad_norm": 0.15375645458698273,
+      "learning_rate": 0.001,
+      "loss": 1.82,
+      "step": 13543
+    },
+    {
+      "epoch": 0.5729757170657416,
+      "grad_norm": 0.14795835316181183,
+      "learning_rate": 0.001,
+      "loss": 2.3571,
+      "step": 13544
+    },
+    {
+      "epoch": 0.573018021829258,
+      "grad_norm": 0.15851891040802002,
+      "learning_rate": 0.001,
+      "loss": 2.5942,
+      "step": 13545
+    },
+    {
+      "epoch": 0.5730603265927744,
+      "grad_norm": 0.15028858184814453,
+      "learning_rate": 0.001,
+      "loss": 3.2621,
+      "step": 13546
+    },
+    {
+      "epoch": 0.5731026313562907,
+      "grad_norm": 0.15741689503192902,
+      "learning_rate": 0.001,
+      "loss": 1.9997,
+      "step": 13547
+    },
+    {
+      "epoch": 0.5731449361198071,
+      "grad_norm": 0.1891627013683319,
+      "learning_rate": 0.001,
+      "loss": 2.2628,
+      "step": 13548
+    },
+    {
+      "epoch": 0.5731872408833235,
+      "grad_norm": 0.16026534140110016,
+      "learning_rate": 0.001,
+      "loss": 2.6502,
+      "step": 13549
+    },
+    {
+      "epoch": 0.5732295456468398,
+      "grad_norm": 0.18979184329509735,
+      "learning_rate": 0.001,
+      "loss": 2.1462,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5732718504103562,
+      "grad_norm": 0.17325901985168457,
+      "learning_rate": 0.001,
+      "loss": 1.8367,
+      "step": 13551
+    },
+    {
+      "epoch": 0.5733141551738726,
+      "grad_norm": 0.2838257849216461,
+      "learning_rate": 0.001,
+      "loss": 2.2564,
+      "step": 13552
+    },
+    {
+      "epoch": 0.5733564599373889,
+      "grad_norm": 0.4795995056629181,
+      "learning_rate": 0.001,
+      "loss": 1.9584,
+      "step": 13553
+    },
+    {
+      "epoch": 0.5733987647009053,
+      "grad_norm": 1.5541175603866577,
+      "learning_rate": 0.001,
+      "loss": 3.0344,
+      "step": 13554
+    },
+    {
+      "epoch": 0.5734410694644217,
+      "grad_norm": 0.1850278079509735,
+      "learning_rate": 0.001,
+      "loss": 2.1873,
+      "step": 13555
+    },
+    {
+      "epoch": 0.573483374227938,
+      "grad_norm": 0.20095832645893097,
+      "learning_rate": 0.001,
+      "loss": 2.0309,
+      "step": 13556
+    },
+    {
+      "epoch": 0.5735256789914545,
+      "grad_norm": 0.212057963013649,
+      "learning_rate": 0.001,
+      "loss": 2.0179,
+      "step": 13557
+    },
+    {
+      "epoch": 0.5735679837549708,
+      "grad_norm": 0.18375861644744873,
+      "learning_rate": 0.001,
+      "loss": 1.7909,
+      "step": 13558
+    },
+    {
+      "epoch": 0.5736102885184872,
+      "grad_norm": 0.1902046501636505,
+      "learning_rate": 0.001,
+      "loss": 1.7385,
+      "step": 13559
+    },
+    {
+      "epoch": 0.5736525932820036,
+      "grad_norm": 0.1839946061372757,
+      "learning_rate": 0.001,
+      "loss": 1.9316,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5736948980455199,
+      "grad_norm": 0.16849452257156372,
+      "learning_rate": 0.001,
+      "loss": 2.4155,
+      "step": 13561
+    },
+    {
+      "epoch": 0.5737372028090363,
+      "grad_norm": 0.18684113025665283,
+      "learning_rate": 0.001,
+      "loss": 2.9065,
+      "step": 13562
+    },
+    {
+      "epoch": 0.5737795075725527,
+      "grad_norm": 0.325171560049057,
+      "learning_rate": 0.001,
+      "loss": 3.7306,
+      "step": 13563
+    },
+    {
+      "epoch": 0.573821812336069,
+      "grad_norm": 0.1625150591135025,
+      "learning_rate": 0.001,
+      "loss": 1.9709,
+      "step": 13564
+    },
+    {
+      "epoch": 0.5738641170995854,
+      "grad_norm": 0.5446557402610779,
+      "learning_rate": 0.001,
+      "loss": 2.5266,
+      "step": 13565
+    },
+    {
+      "epoch": 0.5739064218631018,
+      "grad_norm": 0.1913897693157196,
+      "learning_rate": 0.001,
+      "loss": 2.5541,
+      "step": 13566
+    },
+    {
+      "epoch": 0.5739487266266181,
+      "grad_norm": 0.23677568137645721,
+      "learning_rate": 0.001,
+      "loss": 3.3124,
+      "step": 13567
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.18259602785110474,
+      "learning_rate": 0.001,
+      "loss": 2.398,
+      "step": 13568
+    },
+    {
+      "epoch": 0.5740333361536509,
+      "grad_norm": 0.166291281580925,
+      "learning_rate": 0.001,
+      "loss": 1.7677,
+      "step": 13569
+    },
+    {
+      "epoch": 0.5740756409171672,
+      "grad_norm": 0.19032956659793854,
+      "learning_rate": 0.001,
+      "loss": 2.0508,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5741179456806836,
+      "grad_norm": 0.303861528635025,
+      "learning_rate": 0.001,
+      "loss": 1.8316,
+      "step": 13571
+    },
+    {
+      "epoch": 0.5741602504442,
+      "grad_norm": 0.41305580735206604,
+      "learning_rate": 0.001,
+      "loss": 2.0014,
+      "step": 13572
+    },
+    {
+      "epoch": 0.5742025552077163,
+      "grad_norm": 0.1796369105577469,
+      "learning_rate": 0.001,
+      "loss": 2.7648,
+      "step": 13573
+    },
+    {
+      "epoch": 0.5742448599712328,
+      "grad_norm": 0.15992271900177002,
+      "learning_rate": 0.001,
+      "loss": 1.8547,
+      "step": 13574
+    },
+    {
+      "epoch": 0.5742871647347492,
+      "grad_norm": 0.195455402135849,
+      "learning_rate": 0.001,
+      "loss": 1.5503,
+      "step": 13575
+    },
+    {
+      "epoch": 0.5743294694982655,
+      "grad_norm": 0.17384329438209534,
+      "learning_rate": 0.001,
+      "loss": 2.1787,
+      "step": 13576
+    },
+    {
+      "epoch": 0.5743717742617819,
+      "grad_norm": 0.3714596629142761,
+      "learning_rate": 0.001,
+      "loss": 2.782,
+      "step": 13577
+    },
+    {
+      "epoch": 0.5744140790252983,
+      "grad_norm": 0.9288785457611084,
+      "learning_rate": 0.001,
+      "loss": 2.5938,
+      "step": 13578
+    },
+    {
+      "epoch": 0.5744563837888146,
+      "grad_norm": 0.28560522198677063,
+      "learning_rate": 0.001,
+      "loss": 2.3935,
+      "step": 13579
+    },
+    {
+      "epoch": 0.574498688552331,
+      "grad_norm": 0.23299963772296906,
+      "learning_rate": 0.001,
+      "loss": 2.4586,
+      "step": 13580
+    },
+    {
+      "epoch": 0.5745409933158474,
+      "grad_norm": 0.19017741084098816,
+      "learning_rate": 0.001,
+      "loss": 2.3019,
+      "step": 13581
+    },
+    {
+      "epoch": 0.5745832980793637,
+      "grad_norm": 0.18954819440841675,
+      "learning_rate": 0.001,
+      "loss": 1.8131,
+      "step": 13582
+    },
+    {
+      "epoch": 0.5746256028428801,
+      "grad_norm": 0.1803303062915802,
+      "learning_rate": 0.001,
+      "loss": 1.9121,
+      "step": 13583
+    },
+    {
+      "epoch": 0.5746679076063965,
+      "grad_norm": 4.713398456573486,
+      "learning_rate": 0.001,
+      "loss": 1.9541,
+      "step": 13584
+    },
+    {
+      "epoch": 0.5747102123699128,
+      "grad_norm": 0.2334275096654892,
+      "learning_rate": 0.001,
+      "loss": 1.7124,
+      "step": 13585
+    },
+    {
+      "epoch": 0.5747525171334292,
+      "grad_norm": 0.1691213995218277,
+      "learning_rate": 0.001,
+      "loss": 1.6664,
+      "step": 13586
+    },
+    {
+      "epoch": 0.5747948218969456,
+      "grad_norm": 0.21941591799259186,
+      "learning_rate": 0.001,
+      "loss": 2.0986,
+      "step": 13587
+    },
+    {
+      "epoch": 0.5748371266604619,
+      "grad_norm": 0.30102136731147766,
+      "learning_rate": 0.001,
+      "loss": 2.5321,
+      "step": 13588
+    },
+    {
+      "epoch": 0.5748794314239783,
+      "grad_norm": 0.21132823824882507,
+      "learning_rate": 0.001,
+      "loss": 2.5585,
+      "step": 13589
+    },
+    {
+      "epoch": 0.5749217361874948,
+      "grad_norm": 3.1371676921844482,
+      "learning_rate": 0.001,
+      "loss": 3.0142,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5749640409510111,
+      "grad_norm": 0.3118325471878052,
+      "learning_rate": 0.001,
+      "loss": 1.7528,
+      "step": 13591
+    },
+    {
+      "epoch": 0.5750063457145275,
+      "grad_norm": 0.26828131079673767,
+      "learning_rate": 0.001,
+      "loss": 2.4065,
+      "step": 13592
+    },
+    {
+      "epoch": 0.5750486504780439,
+      "grad_norm": 0.21891672909259796,
+      "learning_rate": 0.001,
+      "loss": 1.8595,
+      "step": 13593
+    },
+    {
+      "epoch": 0.5750909552415602,
+      "grad_norm": 0.20417436957359314,
+      "learning_rate": 0.001,
+      "loss": 1.8951,
+      "step": 13594
+    },
+    {
+      "epoch": 0.5751332600050766,
+      "grad_norm": 0.4515736997127533,
+      "learning_rate": 0.001,
+      "loss": 2.1664,
+      "step": 13595
+    },
+    {
+      "epoch": 0.575175564768593,
+      "grad_norm": 0.17579565942287445,
+      "learning_rate": 0.001,
+      "loss": 2.4499,
+      "step": 13596
+    },
+    {
+      "epoch": 0.5752178695321093,
+      "grad_norm": 0.18956947326660156,
+      "learning_rate": 0.001,
+      "loss": 1.7775,
+      "step": 13597
+    },
+    {
+      "epoch": 0.5752601742956257,
+      "grad_norm": 0.18791532516479492,
+      "learning_rate": 0.001,
+      "loss": 1.858,
+      "step": 13598
+    },
+    {
+      "epoch": 0.5753024790591421,
+      "grad_norm": 0.18148577213287354,
+      "learning_rate": 0.001,
+      "loss": 1.8695,
+      "step": 13599
+    },
+    {
+      "epoch": 0.5753447838226584,
+      "grad_norm": 0.16978557407855988,
+      "learning_rate": 0.001,
+      "loss": 2.0121,
+      "step": 13600
+    },
+    {
+      "epoch": 0.5753870885861748,
+      "grad_norm": 0.1755099594593048,
+      "learning_rate": 0.001,
+      "loss": 1.9807,
+      "step": 13601
+    },
+    {
+      "epoch": 0.5754293933496911,
+      "grad_norm": 0.22287164628505707,
+      "learning_rate": 0.001,
+      "loss": 2.9232,
+      "step": 13602
+    },
+    {
+      "epoch": 0.5754716981132075,
+      "grad_norm": 0.2872997224330902,
+      "learning_rate": 0.001,
+      "loss": 2.2583,
+      "step": 13603
+    },
+    {
+      "epoch": 0.5755140028767239,
+      "grad_norm": 0.43780237436294556,
+      "learning_rate": 0.001,
+      "loss": 1.995,
+      "step": 13604
+    },
+    {
+      "epoch": 0.5755563076402402,
+      "grad_norm": 0.20227646827697754,
+      "learning_rate": 0.001,
+      "loss": 2.8379,
+      "step": 13605
+    },
+    {
+      "epoch": 0.5755986124037566,
+      "grad_norm": 1.0031191110610962,
+      "learning_rate": 0.001,
+      "loss": 2.8675,
+      "step": 13606
+    },
+    {
+      "epoch": 0.5756409171672731,
+      "grad_norm": 0.1701931655406952,
+      "learning_rate": 0.001,
+      "loss": 2.2968,
+      "step": 13607
+    },
+    {
+      "epoch": 0.5756832219307894,
+      "grad_norm": 0.7706162929534912,
+      "learning_rate": 0.001,
+      "loss": 2.6568,
+      "step": 13608
+    },
+    {
+      "epoch": 0.5757255266943058,
+      "grad_norm": 0.22482161223888397,
+      "learning_rate": 0.001,
+      "loss": 1.7453,
+      "step": 13609
+    },
+    {
+      "epoch": 0.5757678314578222,
+      "grad_norm": 1.2364606857299805,
+      "learning_rate": 0.001,
+      "loss": 1.8909,
+      "step": 13610
+    },
+    {
+      "epoch": 0.5758101362213385,
+      "grad_norm": 0.154103621840477,
+      "learning_rate": 0.001,
+      "loss": 1.7456,
+      "step": 13611
+    },
+    {
+      "epoch": 0.5758524409848549,
+      "grad_norm": 0.15440334379673004,
+      "learning_rate": 0.001,
+      "loss": 2.0969,
+      "step": 13612
+    },
+    {
+      "epoch": 0.5758947457483713,
+      "grad_norm": 0.21539804339408875,
+      "learning_rate": 0.001,
+      "loss": 2.4492,
+      "step": 13613
+    },
+    {
+      "epoch": 0.5759370505118876,
+      "grad_norm": 0.17441478371620178,
+      "learning_rate": 0.001,
+      "loss": 1.9286,
+      "step": 13614
+    },
+    {
+      "epoch": 0.575979355275404,
+      "grad_norm": 0.16577987372875214,
+      "learning_rate": 0.001,
+      "loss": 1.7983,
+      "step": 13615
+    },
+    {
+      "epoch": 0.5760216600389204,
+      "grad_norm": 0.6091001033782959,
+      "learning_rate": 0.001,
+      "loss": 2.4681,
+      "step": 13616
+    },
+    {
+      "epoch": 0.5760639648024367,
+      "grad_norm": 0.7010548114776611,
+      "learning_rate": 0.001,
+      "loss": 2.991,
+      "step": 13617
+    },
+    {
+      "epoch": 0.5761062695659531,
+      "grad_norm": 0.18760816752910614,
+      "learning_rate": 0.001,
+      "loss": 2.7225,
+      "step": 13618
+    },
+    {
+      "epoch": 0.5761485743294695,
+      "grad_norm": 0.2499050348997116,
+      "learning_rate": 0.001,
+      "loss": 1.889,
+      "step": 13619
+    },
+    {
+      "epoch": 0.5761908790929858,
+      "grad_norm": 0.19262200593948364,
+      "learning_rate": 0.001,
+      "loss": 1.9694,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5762331838565022,
+      "grad_norm": 0.2874251902103424,
+      "learning_rate": 0.001,
+      "loss": 2.0702,
+      "step": 13621
+    },
+    {
+      "epoch": 0.5762754886200186,
+      "grad_norm": 0.2215861827135086,
+      "learning_rate": 0.001,
+      "loss": 3.0974,
+      "step": 13622
+    },
+    {
+      "epoch": 0.576317793383535,
+      "grad_norm": 1.4504274129867554,
+      "learning_rate": 0.001,
+      "loss": 2.6667,
+      "step": 13623
+    },
+    {
+      "epoch": 0.5763600981470514,
+      "grad_norm": 0.3368513286113739,
+      "learning_rate": 0.001,
+      "loss": 2.2946,
+      "step": 13624
+    },
+    {
+      "epoch": 0.5764024029105678,
+      "grad_norm": 0.5865402817726135,
+      "learning_rate": 0.001,
+      "loss": 2.1963,
+      "step": 13625
+    },
+    {
+      "epoch": 0.5764447076740841,
+      "grad_norm": 0.16736286878585815,
+      "learning_rate": 0.001,
+      "loss": 2.4853,
+      "step": 13626
+    },
+    {
+      "epoch": 0.5764870124376005,
+      "grad_norm": 0.18231581151485443,
+      "learning_rate": 0.001,
+      "loss": 1.9809,
+      "step": 13627
+    },
+    {
+      "epoch": 0.5765293172011169,
+      "grad_norm": 0.16837336122989655,
+      "learning_rate": 0.001,
+      "loss": 3.887,
+      "step": 13628
+    },
+    {
+      "epoch": 0.5765716219646332,
+      "grad_norm": 0.4993027150630951,
+      "learning_rate": 0.001,
+      "loss": 2.3137,
+      "step": 13629
+    },
+    {
+      "epoch": 0.5766139267281496,
+      "grad_norm": 0.2104516625404358,
+      "learning_rate": 0.001,
+      "loss": 1.6916,
+      "step": 13630
+    },
+    {
+      "epoch": 0.576656231491666,
+      "grad_norm": 0.20687484741210938,
+      "learning_rate": 0.001,
+      "loss": 2.2164,
+      "step": 13631
+    },
+    {
+      "epoch": 0.5766985362551823,
+      "grad_norm": 0.18873409926891327,
+      "learning_rate": 0.001,
+      "loss": 2.7869,
+      "step": 13632
+    },
+    {
+      "epoch": 0.5767408410186987,
+      "grad_norm": 0.16937416791915894,
+      "learning_rate": 0.001,
+      "loss": 1.9121,
+      "step": 13633
+    },
+    {
+      "epoch": 0.5767831457822151,
+      "grad_norm": 0.15023858845233917,
+      "learning_rate": 0.001,
+      "loss": 2.6665,
+      "step": 13634
+    },
+    {
+      "epoch": 0.5768254505457314,
+      "grad_norm": 0.15277251601219177,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 13635
+    },
+    {
+      "epoch": 0.5768677553092478,
+      "grad_norm": 0.19467462599277496,
+      "learning_rate": 0.001,
+      "loss": 1.6275,
+      "step": 13636
+    },
+    {
+      "epoch": 0.5769100600727642,
+      "grad_norm": 0.1640617549419403,
+      "learning_rate": 0.001,
+      "loss": 1.331,
+      "step": 13637
+    },
+    {
+      "epoch": 0.5769523648362805,
+      "grad_norm": 0.17704744637012482,
+      "learning_rate": 0.001,
+      "loss": 2.0327,
+      "step": 13638
+    },
+    {
+      "epoch": 0.576994669599797,
+      "grad_norm": 0.16801083087921143,
+      "learning_rate": 0.001,
+      "loss": 2.6699,
+      "step": 13639
+    },
+    {
+      "epoch": 0.5770369743633134,
+      "grad_norm": 0.22247737646102905,
+      "learning_rate": 0.001,
+      "loss": 1.7275,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5770792791268297,
+      "grad_norm": 0.2886200249195099,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 13641
+    },
+    {
+      "epoch": 0.5771215838903461,
+      "grad_norm": 2.0032057762145996,
+      "learning_rate": 0.001,
+      "loss": 1.4168,
+      "step": 13642
+    },
+    {
+      "epoch": 0.5771638886538625,
+      "grad_norm": 0.1862766593694687,
+      "learning_rate": 0.001,
+      "loss": 3.6991,
+      "step": 13643
+    },
+    {
+      "epoch": 0.5772061934173788,
+      "grad_norm": 0.27933868765830994,
+      "learning_rate": 0.001,
+      "loss": 2.5766,
+      "step": 13644
+    },
+    {
+      "epoch": 0.5772484981808952,
+      "grad_norm": 0.15503652393817902,
+      "learning_rate": 0.001,
+      "loss": 1.4434,
+      "step": 13645
+    },
+    {
+      "epoch": 0.5772908029444115,
+      "grad_norm": 0.1637222021818161,
+      "learning_rate": 0.001,
+      "loss": 1.8226,
+      "step": 13646
+    },
+    {
+      "epoch": 0.5773331077079279,
+      "grad_norm": 0.1900179535150528,
+      "learning_rate": 0.001,
+      "loss": 2.8638,
+      "step": 13647
+    },
+    {
+      "epoch": 0.5773754124714443,
+      "grad_norm": 5.069095134735107,
+      "learning_rate": 0.001,
+      "loss": 2.3441,
+      "step": 13648
+    },
+    {
+      "epoch": 0.5774177172349606,
+      "grad_norm": 0.21330921351909637,
+      "learning_rate": 0.001,
+      "loss": 2.6057,
+      "step": 13649
+    },
+    {
+      "epoch": 0.577460021998477,
+      "grad_norm": 0.3153631389141083,
+      "learning_rate": 0.001,
+      "loss": 3.2548,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5775023267619934,
+      "grad_norm": 0.18381884694099426,
+      "learning_rate": 0.001,
+      "loss": 1.7146,
+      "step": 13651
+    },
+    {
+      "epoch": 0.5775446315255097,
+      "grad_norm": 0.2848173677921295,
+      "learning_rate": 0.001,
+      "loss": 1.6744,
+      "step": 13652
+    },
+    {
+      "epoch": 0.5775869362890261,
+      "grad_norm": 0.16625621914863586,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 13653
+    },
+    {
+      "epoch": 0.5776292410525425,
+      "grad_norm": 0.14194847643375397,
+      "learning_rate": 0.001,
+      "loss": 1.6789,
+      "step": 13654
+    },
+    {
+      "epoch": 0.5776715458160588,
+      "grad_norm": 0.24383410811424255,
+      "learning_rate": 0.001,
+      "loss": 2.5261,
+      "step": 13655
+    },
+    {
+      "epoch": 0.5777138505795752,
+      "grad_norm": 0.18297232687473297,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 13656
+    },
+    {
+      "epoch": 0.5777561553430917,
+      "grad_norm": 0.16804887354373932,
+      "learning_rate": 0.001,
+      "loss": 2.0048,
+      "step": 13657
+    },
+    {
+      "epoch": 0.577798460106608,
+      "grad_norm": 0.20523540675640106,
+      "learning_rate": 0.001,
+      "loss": 3.0154,
+      "step": 13658
+    },
+    {
+      "epoch": 0.5778407648701244,
+      "grad_norm": 0.9247041344642639,
+      "learning_rate": 0.001,
+      "loss": 2.0177,
+      "step": 13659
+    },
+    {
+      "epoch": 0.5778830696336408,
+      "grad_norm": 0.27602866291999817,
+      "learning_rate": 0.001,
+      "loss": 1.8518,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5779253743971571,
+      "grad_norm": 0.1811681091785431,
+      "learning_rate": 0.001,
+      "loss": 2.2486,
+      "step": 13661
+    },
+    {
+      "epoch": 0.5779676791606735,
+      "grad_norm": 1.9261819124221802,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 13662
+    },
+    {
+      "epoch": 0.5780099839241899,
+      "grad_norm": 0.23115941882133484,
+      "learning_rate": 0.001,
+      "loss": 2.0086,
+      "step": 13663
+    },
+    {
+      "epoch": 0.5780522886877062,
+      "grad_norm": 0.20311318337917328,
+      "learning_rate": 0.001,
+      "loss": 2.3334,
+      "step": 13664
+    },
+    {
+      "epoch": 0.5780945934512226,
+      "grad_norm": 0.1666504442691803,
+      "learning_rate": 0.001,
+      "loss": 2.0804,
+      "step": 13665
+    },
+    {
+      "epoch": 0.578136898214739,
+      "grad_norm": 0.20418089628219604,
+      "learning_rate": 0.001,
+      "loss": 2.2807,
+      "step": 13666
+    },
+    {
+      "epoch": 0.5781792029782553,
+      "grad_norm": 0.1729002743959427,
+      "learning_rate": 0.001,
+      "loss": 1.6245,
+      "step": 13667
+    },
+    {
+      "epoch": 0.5782215077417717,
+      "grad_norm": 0.1773560643196106,
+      "learning_rate": 0.001,
+      "loss": 1.6606,
+      "step": 13668
+    },
+    {
+      "epoch": 0.5782638125052881,
+      "grad_norm": 0.17324134707450867,
+      "learning_rate": 0.001,
+      "loss": 1.5963,
+      "step": 13669
+    },
+    {
+      "epoch": 0.5783061172688044,
+      "grad_norm": 0.503576397895813,
+      "learning_rate": 0.001,
+      "loss": 2.0075,
+      "step": 13670
+    },
+    {
+      "epoch": 0.5783484220323208,
+      "grad_norm": 0.15948913991451263,
+      "learning_rate": 0.001,
+      "loss": 1.7914,
+      "step": 13671
+    },
+    {
+      "epoch": 0.5783907267958373,
+      "grad_norm": 0.2122723013162613,
+      "learning_rate": 0.001,
+      "loss": 2.4304,
+      "step": 13672
+    },
+    {
+      "epoch": 0.5784330315593535,
+      "grad_norm": 0.9828127026557922,
+      "learning_rate": 0.001,
+      "loss": 2.3654,
+      "step": 13673
+    },
+    {
+      "epoch": 0.57847533632287,
+      "grad_norm": 0.1700863391160965,
+      "learning_rate": 0.001,
+      "loss": 2.0107,
+      "step": 13674
+    },
+    {
+      "epoch": 0.5785176410863864,
+      "grad_norm": 0.3214263617992401,
+      "learning_rate": 0.001,
+      "loss": 2.5081,
+      "step": 13675
+    },
+    {
+      "epoch": 0.5785599458499027,
+      "grad_norm": 0.21781599521636963,
+      "learning_rate": 0.001,
+      "loss": 2.0888,
+      "step": 13676
+    },
+    {
+      "epoch": 0.5786022506134191,
+      "grad_norm": 0.16866712272167206,
+      "learning_rate": 0.001,
+      "loss": 2.407,
+      "step": 13677
+    },
+    {
+      "epoch": 0.5786445553769355,
+      "grad_norm": 0.15926438570022583,
+      "learning_rate": 0.001,
+      "loss": 1.3379,
+      "step": 13678
+    },
+    {
+      "epoch": 0.5786868601404518,
+      "grad_norm": 0.23510025441646576,
+      "learning_rate": 0.001,
+      "loss": 1.7714,
+      "step": 13679
+    },
+    {
+      "epoch": 0.5787291649039682,
+      "grad_norm": 0.14508269727230072,
+      "learning_rate": 0.001,
+      "loss": 2.1146,
+      "step": 13680
+    },
+    {
+      "epoch": 0.5787714696674846,
+      "grad_norm": 2.956667423248291,
+      "learning_rate": 0.001,
+      "loss": 2.8911,
+      "step": 13681
+    },
+    {
+      "epoch": 0.5788137744310009,
+      "grad_norm": 0.24240057170391083,
+      "learning_rate": 0.001,
+      "loss": 3.1488,
+      "step": 13682
+    },
+    {
+      "epoch": 0.5788560791945173,
+      "grad_norm": 0.16761580109596252,
+      "learning_rate": 0.001,
+      "loss": 3.1758,
+      "step": 13683
+    },
+    {
+      "epoch": 0.5788983839580337,
+      "grad_norm": 0.15230026841163635,
+      "learning_rate": 0.001,
+      "loss": 1.5823,
+      "step": 13684
+    },
+    {
+      "epoch": 0.57894068872155,
+      "grad_norm": 0.14217911660671234,
+      "learning_rate": 0.001,
+      "loss": 1.3058,
+      "step": 13685
+    },
+    {
+      "epoch": 0.5789829934850664,
+      "grad_norm": 0.23708081245422363,
+      "learning_rate": 0.001,
+      "loss": 3.0425,
+      "step": 13686
+    },
+    {
+      "epoch": 0.5790252982485828,
+      "grad_norm": 0.19518715143203735,
+      "learning_rate": 0.001,
+      "loss": 2.1443,
+      "step": 13687
+    },
+    {
+      "epoch": 0.5790676030120991,
+      "grad_norm": 0.2842262089252472,
+      "learning_rate": 0.001,
+      "loss": 2.324,
+      "step": 13688
+    },
+    {
+      "epoch": 0.5791099077756156,
+      "grad_norm": 0.2659071683883667,
+      "learning_rate": 0.001,
+      "loss": 2.3996,
+      "step": 13689
+    },
+    {
+      "epoch": 0.579152212539132,
+      "grad_norm": 0.16496817767620087,
+      "learning_rate": 0.001,
+      "loss": 1.9989,
+      "step": 13690
+    },
+    {
+      "epoch": 0.5791945173026483,
+      "grad_norm": 0.19301360845565796,
+      "learning_rate": 0.001,
+      "loss": 3.5109,
+      "step": 13691
+    },
+    {
+      "epoch": 0.5792368220661647,
+      "grad_norm": 0.3519555628299713,
+      "learning_rate": 0.001,
+      "loss": 3.3803,
+      "step": 13692
+    },
+    {
+      "epoch": 0.579279126829681,
+      "grad_norm": 0.26763901114463806,
+      "learning_rate": 0.001,
+      "loss": 2.4265,
+      "step": 13693
+    },
+    {
+      "epoch": 0.5793214315931974,
+      "grad_norm": 0.1800110638141632,
+      "learning_rate": 0.001,
+      "loss": 1.8322,
+      "step": 13694
+    },
+    {
+      "epoch": 0.5793637363567138,
+      "grad_norm": 0.16732731461524963,
+      "learning_rate": 0.001,
+      "loss": 1.8233,
+      "step": 13695
+    },
+    {
+      "epoch": 0.5794060411202301,
+      "grad_norm": 0.17104704678058624,
+      "learning_rate": 0.001,
+      "loss": 2.2859,
+      "step": 13696
+    },
+    {
+      "epoch": 0.5794483458837465,
+      "grad_norm": 0.18005624413490295,
+      "learning_rate": 0.001,
+      "loss": 3.1283,
+      "step": 13697
+    },
+    {
+      "epoch": 0.5794906506472629,
+      "grad_norm": 0.15262584388256073,
+      "learning_rate": 0.001,
+      "loss": 2.445,
+      "step": 13698
+    },
+    {
+      "epoch": 0.5795329554107792,
+      "grad_norm": 0.13767917454242706,
+      "learning_rate": 0.001,
+      "loss": 1.5765,
+      "step": 13699
+    },
+    {
+      "epoch": 0.5795752601742956,
+      "grad_norm": 0.18696613609790802,
+      "learning_rate": 0.001,
+      "loss": 2.0157,
+      "step": 13700
+    },
+    {
+      "epoch": 0.579617564937812,
+      "grad_norm": 0.21265771985054016,
+      "learning_rate": 0.001,
+      "loss": 2.1951,
+      "step": 13701
+    },
+    {
+      "epoch": 0.5796598697013283,
+      "grad_norm": 0.154225155711174,
+      "learning_rate": 0.001,
+      "loss": 2.6487,
+      "step": 13702
+    },
+    {
+      "epoch": 0.5797021744648447,
+      "grad_norm": 0.3607092797756195,
+      "learning_rate": 0.001,
+      "loss": 1.9679,
+      "step": 13703
+    },
+    {
+      "epoch": 0.5797444792283611,
+      "grad_norm": 0.30119165778160095,
+      "learning_rate": 0.001,
+      "loss": 2.7281,
+      "step": 13704
+    },
+    {
+      "epoch": 0.5797867839918774,
+      "grad_norm": 0.14737650752067566,
+      "learning_rate": 0.001,
+      "loss": 1.6039,
+      "step": 13705
+    },
+    {
+      "epoch": 0.5798290887553939,
+      "grad_norm": 0.1710781455039978,
+      "learning_rate": 0.001,
+      "loss": 1.7936,
+      "step": 13706
+    },
+    {
+      "epoch": 0.5798713935189103,
+      "grad_norm": 0.21716709434986115,
+      "learning_rate": 0.001,
+      "loss": 2.3436,
+      "step": 13707
+    },
+    {
+      "epoch": 0.5799136982824266,
+      "grad_norm": 0.13513155281543732,
+      "learning_rate": 0.001,
+      "loss": 2.2419,
+      "step": 13708
+    },
+    {
+      "epoch": 0.579956003045943,
+      "grad_norm": 0.14529554545879364,
+      "learning_rate": 0.001,
+      "loss": 2.1327,
+      "step": 13709
+    },
+    {
+      "epoch": 0.5799983078094594,
+      "grad_norm": 0.14698375761508942,
+      "learning_rate": 0.001,
+      "loss": 1.6831,
+      "step": 13710
+    },
+    {
+      "epoch": 0.5800406125729757,
+      "grad_norm": 0.1601586788892746,
+      "learning_rate": 0.001,
+      "loss": 2.5013,
+      "step": 13711
+    },
+    {
+      "epoch": 0.5800829173364921,
+      "grad_norm": 2.6871955394744873,
+      "learning_rate": 0.001,
+      "loss": 2.3585,
+      "step": 13712
+    },
+    {
+      "epoch": 0.5801252221000085,
+      "grad_norm": 1.6209683418273926,
+      "learning_rate": 0.001,
+      "loss": 2.7192,
+      "step": 13713
+    },
+    {
+      "epoch": 0.5801675268635248,
+      "grad_norm": 0.4082275927066803,
+      "learning_rate": 0.001,
+      "loss": 2.3748,
+      "step": 13714
+    },
+    {
+      "epoch": 0.5802098316270412,
+      "grad_norm": 0.8170319199562073,
+      "learning_rate": 0.001,
+      "loss": 1.9599,
+      "step": 13715
+    },
+    {
+      "epoch": 0.5802521363905576,
+      "grad_norm": 0.15276405215263367,
+      "learning_rate": 0.001,
+      "loss": 3.2077,
+      "step": 13716
+    },
+    {
+      "epoch": 0.5802944411540739,
+      "grad_norm": 0.16727223992347717,
+      "learning_rate": 0.001,
+      "loss": 3.1438,
+      "step": 13717
+    },
+    {
+      "epoch": 0.5803367459175903,
+      "grad_norm": 0.1725165992975235,
+      "learning_rate": 0.001,
+      "loss": 2.3128,
+      "step": 13718
+    },
+    {
+      "epoch": 0.5803790506811067,
+      "grad_norm": 0.16121014952659607,
+      "learning_rate": 0.001,
+      "loss": 2.9876,
+      "step": 13719
+    },
+    {
+      "epoch": 0.580421355444623,
+      "grad_norm": 0.27277088165283203,
+      "learning_rate": 0.001,
+      "loss": 2.1281,
+      "step": 13720
+    },
+    {
+      "epoch": 0.5804636602081394,
+      "grad_norm": 0.2123079150915146,
+      "learning_rate": 0.001,
+      "loss": 2.2785,
+      "step": 13721
+    },
+    {
+      "epoch": 0.5805059649716559,
+      "grad_norm": 0.7211734056472778,
+      "learning_rate": 0.001,
+      "loss": 2.1592,
+      "step": 13722
+    },
+    {
+      "epoch": 0.5805482697351722,
+      "grad_norm": 0.5063629746437073,
+      "learning_rate": 0.001,
+      "loss": 1.605,
+      "step": 13723
+    },
+    {
+      "epoch": 0.5805905744986886,
+      "grad_norm": 0.15953439474105835,
+      "learning_rate": 0.001,
+      "loss": 2.3575,
+      "step": 13724
+    },
+    {
+      "epoch": 0.580632879262205,
+      "grad_norm": 0.2004745602607727,
+      "learning_rate": 0.001,
+      "loss": 2.5765,
+      "step": 13725
+    },
+    {
+      "epoch": 0.5806751840257213,
+      "grad_norm": 0.17336571216583252,
+      "learning_rate": 0.001,
+      "loss": 2.242,
+      "step": 13726
+    },
+    {
+      "epoch": 0.5807174887892377,
+      "grad_norm": 0.15672898292541504,
+      "learning_rate": 0.001,
+      "loss": 1.6338,
+      "step": 13727
+    },
+    {
+      "epoch": 0.5807597935527541,
+      "grad_norm": 0.16324591636657715,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 13728
+    },
+    {
+      "epoch": 0.5808020983162704,
+      "grad_norm": 0.8709842562675476,
+      "learning_rate": 0.001,
+      "loss": 1.8762,
+      "step": 13729
+    },
+    {
+      "epoch": 0.5808444030797868,
+      "grad_norm": 0.16312599182128906,
+      "learning_rate": 0.001,
+      "loss": 1.8523,
+      "step": 13730
+    },
+    {
+      "epoch": 0.5808867078433032,
+      "grad_norm": 0.6291779279708862,
+      "learning_rate": 0.001,
+      "loss": 3.0103,
+      "step": 13731
+    },
+    {
+      "epoch": 0.5809290126068195,
+      "grad_norm": 0.18601246178150177,
+      "learning_rate": 0.001,
+      "loss": 2.3727,
+      "step": 13732
+    },
+    {
+      "epoch": 0.5809713173703359,
+      "grad_norm": 0.22245633602142334,
+      "learning_rate": 0.001,
+      "loss": 1.926,
+      "step": 13733
+    },
+    {
+      "epoch": 0.5810136221338523,
+      "grad_norm": 0.1883118599653244,
+      "learning_rate": 0.001,
+      "loss": 2.1428,
+      "step": 13734
+    },
+    {
+      "epoch": 0.5810559268973686,
+      "grad_norm": 0.15982557833194733,
+      "learning_rate": 0.001,
+      "loss": 2.1728,
+      "step": 13735
+    },
+    {
+      "epoch": 0.581098231660885,
+      "grad_norm": 0.18497587740421295,
+      "learning_rate": 0.001,
+      "loss": 1.7981,
+      "step": 13736
+    },
+    {
+      "epoch": 0.5811405364244013,
+      "grad_norm": 0.1280389130115509,
+      "learning_rate": 0.001,
+      "loss": 1.5147,
+      "step": 13737
+    },
+    {
+      "epoch": 0.5811828411879177,
+      "grad_norm": 0.2510825991630554,
+      "learning_rate": 0.001,
+      "loss": 1.5612,
+      "step": 13738
+    },
+    {
+      "epoch": 0.5812251459514342,
+      "grad_norm": 0.1742805689573288,
+      "learning_rate": 0.001,
+      "loss": 2.0385,
+      "step": 13739
+    },
+    {
+      "epoch": 0.5812674507149505,
+      "grad_norm": 0.14577478170394897,
+      "learning_rate": 0.001,
+      "loss": 2.1657,
+      "step": 13740
+    },
+    {
+      "epoch": 0.5813097554784669,
+      "grad_norm": 0.17133548855781555,
+      "learning_rate": 0.001,
+      "loss": 2.2744,
+      "step": 13741
+    },
+    {
+      "epoch": 0.5813520602419833,
+      "grad_norm": 0.16263654828071594,
+      "learning_rate": 0.001,
+      "loss": 1.8197,
+      "step": 13742
+    },
+    {
+      "epoch": 0.5813943650054996,
+      "grad_norm": 0.19162042438983917,
+      "learning_rate": 0.001,
+      "loss": 1.8052,
+      "step": 13743
+    },
+    {
+      "epoch": 0.581436669769016,
+      "grad_norm": 0.17110483348369598,
+      "learning_rate": 0.001,
+      "loss": 2.1316,
+      "step": 13744
+    },
+    {
+      "epoch": 0.5814789745325324,
+      "grad_norm": 0.4235318601131439,
+      "learning_rate": 0.001,
+      "loss": 2.5278,
+      "step": 13745
+    },
+    {
+      "epoch": 0.5815212792960487,
+      "grad_norm": 0.15516316890716553,
+      "learning_rate": 0.001,
+      "loss": 2.2714,
+      "step": 13746
+    },
+    {
+      "epoch": 0.5815635840595651,
+      "grad_norm": 1.130749225616455,
+      "learning_rate": 0.001,
+      "loss": 1.8972,
+      "step": 13747
+    },
+    {
+      "epoch": 0.5816058888230815,
+      "grad_norm": 0.41215425729751587,
+      "learning_rate": 0.001,
+      "loss": 3.452,
+      "step": 13748
+    },
+    {
+      "epoch": 0.5816481935865978,
+      "grad_norm": 4.2232441902160645,
+      "learning_rate": 0.001,
+      "loss": 1.7344,
+      "step": 13749
+    },
+    {
+      "epoch": 0.5816904983501142,
+      "grad_norm": 0.19343996047973633,
+      "learning_rate": 0.001,
+      "loss": 2.1452,
+      "step": 13750
+    },
+    {
+      "epoch": 0.5817328031136306,
+      "grad_norm": 36.49468231201172,
+      "learning_rate": 0.001,
+      "loss": 1.8784,
+      "step": 13751
+    },
+    {
+      "epoch": 0.5817751078771469,
+      "grad_norm": 0.2389318197965622,
+      "learning_rate": 0.001,
+      "loss": 2.6143,
+      "step": 13752
+    },
+    {
+      "epoch": 0.5818174126406633,
+      "grad_norm": 0.24423779547214508,
+      "learning_rate": 0.001,
+      "loss": 2.6543,
+      "step": 13753
+    },
+    {
+      "epoch": 0.5818597174041797,
+      "grad_norm": 0.2429848611354828,
+      "learning_rate": 0.001,
+      "loss": 2.2033,
+      "step": 13754
+    },
+    {
+      "epoch": 0.581902022167696,
+      "grad_norm": 0.2506393790245056,
+      "learning_rate": 0.001,
+      "loss": 2.3673,
+      "step": 13755
+    },
+    {
+      "epoch": 0.5819443269312125,
+      "grad_norm": 0.1692809909582138,
+      "learning_rate": 0.001,
+      "loss": 2.2108,
+      "step": 13756
+    },
+    {
+      "epoch": 0.5819866316947289,
+      "grad_norm": 5.47391939163208,
+      "learning_rate": 0.001,
+      "loss": 2.5862,
+      "step": 13757
+    },
+    {
+      "epoch": 0.5820289364582452,
+      "grad_norm": 0.17192316055297852,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 13758
+    },
+    {
+      "epoch": 0.5820712412217616,
+      "grad_norm": 0.22438155114650726,
+      "learning_rate": 0.001,
+      "loss": 2.4762,
+      "step": 13759
+    },
+    {
+      "epoch": 0.582113545985278,
+      "grad_norm": 1.2624096870422363,
+      "learning_rate": 0.001,
+      "loss": 2.1616,
+      "step": 13760
+    },
+    {
+      "epoch": 0.5821558507487943,
+      "grad_norm": 0.202859029173851,
+      "learning_rate": 0.001,
+      "loss": 2.0482,
+      "step": 13761
+    },
+    {
+      "epoch": 0.5821981555123107,
+      "grad_norm": 0.22284016013145447,
+      "learning_rate": 0.001,
+      "loss": 2.1997,
+      "step": 13762
+    },
+    {
+      "epoch": 0.5822404602758271,
+      "grad_norm": 0.20142649114131927,
+      "learning_rate": 0.001,
+      "loss": 1.8527,
+      "step": 13763
+    },
+    {
+      "epoch": 0.5822827650393434,
+      "grad_norm": 0.1934656798839569,
+      "learning_rate": 0.001,
+      "loss": 2.7891,
+      "step": 13764
+    },
+    {
+      "epoch": 0.5823250698028598,
+      "grad_norm": 0.25359898805618286,
+      "learning_rate": 0.001,
+      "loss": 2.1556,
+      "step": 13765
+    },
+    {
+      "epoch": 0.5823673745663762,
+      "grad_norm": 0.43819230794906616,
+      "learning_rate": 0.001,
+      "loss": 2.4958,
+      "step": 13766
+    },
+    {
+      "epoch": 0.5824096793298925,
+      "grad_norm": 0.17032918334007263,
+      "learning_rate": 0.001,
+      "loss": 1.8058,
+      "step": 13767
+    },
+    {
+      "epoch": 0.5824519840934089,
+      "grad_norm": 2.558115005493164,
+      "learning_rate": 0.001,
+      "loss": 3.1201,
+      "step": 13768
+    },
+    {
+      "epoch": 0.5824942888569253,
+      "grad_norm": 0.20715180039405823,
+      "learning_rate": 0.001,
+      "loss": 2.2601,
+      "step": 13769
+    },
+    {
+      "epoch": 0.5825365936204416,
+      "grad_norm": 0.5378583669662476,
+      "learning_rate": 0.001,
+      "loss": 2.006,
+      "step": 13770
+    },
+    {
+      "epoch": 0.582578898383958,
+      "grad_norm": 0.28567495942115784,
+      "learning_rate": 0.001,
+      "loss": 1.7223,
+      "step": 13771
+    },
+    {
+      "epoch": 0.5826212031474745,
+      "grad_norm": 0.18178150057792664,
+      "learning_rate": 0.001,
+      "loss": 1.3784,
+      "step": 13772
+    },
+    {
+      "epoch": 0.5826635079109908,
+      "grad_norm": 0.23406149446964264,
+      "learning_rate": 0.001,
+      "loss": 4.444,
+      "step": 13773
+    },
+    {
+      "epoch": 0.5827058126745072,
+      "grad_norm": 0.18092291057109833,
+      "learning_rate": 0.001,
+      "loss": 3.0016,
+      "step": 13774
+    },
+    {
+      "epoch": 0.5827481174380236,
+      "grad_norm": 0.3098083436489105,
+      "learning_rate": 0.001,
+      "loss": 2.7088,
+      "step": 13775
+    },
+    {
+      "epoch": 0.5827904222015399,
+      "grad_norm": 0.18877199292182922,
+      "learning_rate": 0.001,
+      "loss": 1.6503,
+      "step": 13776
+    },
+    {
+      "epoch": 0.5828327269650563,
+      "grad_norm": 0.2062119096517563,
+      "learning_rate": 0.001,
+      "loss": 2.1285,
+      "step": 13777
+    },
+    {
+      "epoch": 0.5828750317285727,
+      "grad_norm": 0.3584443926811218,
+      "learning_rate": 0.001,
+      "loss": 1.7564,
+      "step": 13778
+    },
+    {
+      "epoch": 0.582917336492089,
+      "grad_norm": 0.17341378331184387,
+      "learning_rate": 0.001,
+      "loss": 1.9815,
+      "step": 13779
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 0.17184588313102722,
+      "learning_rate": 0.001,
+      "loss": 1.8823,
+      "step": 13780
+    },
+    {
+      "epoch": 0.5830019460191218,
+      "grad_norm": 0.16942469775676727,
+      "learning_rate": 0.001,
+      "loss": 1.9993,
+      "step": 13781
+    },
+    {
+      "epoch": 0.5830442507826381,
+      "grad_norm": 0.17933522164821625,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 13782
+    },
+    {
+      "epoch": 0.5830865555461545,
+      "grad_norm": 0.15469792485237122,
+      "learning_rate": 0.001,
+      "loss": 2.1212,
+      "step": 13783
+    },
+    {
+      "epoch": 0.5831288603096708,
+      "grad_norm": 0.1891220211982727,
+      "learning_rate": 0.001,
+      "loss": 3.1775,
+      "step": 13784
+    },
+    {
+      "epoch": 0.5831711650731872,
+      "grad_norm": 0.1707414835691452,
+      "learning_rate": 0.001,
+      "loss": 1.991,
+      "step": 13785
+    },
+    {
+      "epoch": 0.5832134698367036,
+      "grad_norm": 1.5627321004867554,
+      "learning_rate": 0.001,
+      "loss": 1.999,
+      "step": 13786
+    },
+    {
+      "epoch": 0.5832557746002199,
+      "grad_norm": 0.15185093879699707,
+      "learning_rate": 0.001,
+      "loss": 2.8022,
+      "step": 13787
+    },
+    {
+      "epoch": 0.5832980793637363,
+      "grad_norm": 0.1897820383310318,
+      "learning_rate": 0.001,
+      "loss": 2.0066,
+      "step": 13788
+    },
+    {
+      "epoch": 0.5833403841272528,
+      "grad_norm": 0.4447472393512726,
+      "learning_rate": 0.001,
+      "loss": 3.3133,
+      "step": 13789
+    },
+    {
+      "epoch": 0.583382688890769,
+      "grad_norm": 0.15921324491500854,
+      "learning_rate": 0.001,
+      "loss": 2.5172,
+      "step": 13790
+    },
+    {
+      "epoch": 0.5834249936542855,
+      "grad_norm": 0.22110402584075928,
+      "learning_rate": 0.001,
+      "loss": 2.0695,
+      "step": 13791
+    },
+    {
+      "epoch": 0.5834672984178019,
+      "grad_norm": 0.34887969493865967,
+      "learning_rate": 0.001,
+      "loss": 2.6305,
+      "step": 13792
+    },
+    {
+      "epoch": 0.5835096031813182,
+      "grad_norm": 0.16820944845676422,
+      "learning_rate": 0.001,
+      "loss": 2.5623,
+      "step": 13793
+    },
+    {
+      "epoch": 0.5835519079448346,
+      "grad_norm": 0.4705934226512909,
+      "learning_rate": 0.001,
+      "loss": 2.2158,
+      "step": 13794
+    },
+    {
+      "epoch": 0.583594212708351,
+      "grad_norm": 0.24807946383953094,
+      "learning_rate": 0.001,
+      "loss": 1.9532,
+      "step": 13795
+    },
+    {
+      "epoch": 0.5836365174718673,
+      "grad_norm": 1.3096846342086792,
+      "learning_rate": 0.001,
+      "loss": 1.9417,
+      "step": 13796
+    },
+    {
+      "epoch": 0.5836788222353837,
+      "grad_norm": 0.1883528232574463,
+      "learning_rate": 0.001,
+      "loss": 2.1403,
+      "step": 13797
+    },
+    {
+      "epoch": 0.5837211269989001,
+      "grad_norm": 0.20582067966461182,
+      "learning_rate": 0.001,
+      "loss": 2.3299,
+      "step": 13798
+    },
+    {
+      "epoch": 0.5837634317624164,
+      "grad_norm": 2.3853557109832764,
+      "learning_rate": 0.001,
+      "loss": 1.6175,
+      "step": 13799
+    },
+    {
+      "epoch": 0.5838057365259328,
+      "grad_norm": 0.21598750352859497,
+      "learning_rate": 0.001,
+      "loss": 2.3283,
+      "step": 13800
+    },
+    {
+      "epoch": 0.5838480412894492,
+      "grad_norm": 0.41916608810424805,
+      "learning_rate": 0.001,
+      "loss": 2.8298,
+      "step": 13801
+    },
+    {
+      "epoch": 0.5838903460529655,
+      "grad_norm": 0.18645887076854706,
+      "learning_rate": 0.001,
+      "loss": 1.4992,
+      "step": 13802
+    },
+    {
+      "epoch": 0.5839326508164819,
+      "grad_norm": 0.15576979517936707,
+      "learning_rate": 0.001,
+      "loss": 1.6169,
+      "step": 13803
+    },
+    {
+      "epoch": 0.5839749555799983,
+      "grad_norm": 0.1782311350107193,
+      "learning_rate": 0.001,
+      "loss": 1.7768,
+      "step": 13804
+    },
+    {
+      "epoch": 0.5840172603435146,
+      "grad_norm": 0.8121781349182129,
+      "learning_rate": 0.001,
+      "loss": 2.2347,
+      "step": 13805
+    },
+    {
+      "epoch": 0.584059565107031,
+      "grad_norm": 0.20430848002433777,
+      "learning_rate": 0.001,
+      "loss": 2.1171,
+      "step": 13806
+    },
+    {
+      "epoch": 0.5841018698705475,
+      "grad_norm": 0.5235466361045837,
+      "learning_rate": 0.001,
+      "loss": 1.7292,
+      "step": 13807
+    },
+    {
+      "epoch": 0.5841441746340638,
+      "grad_norm": 0.17465122044086456,
+      "learning_rate": 0.001,
+      "loss": 1.7302,
+      "step": 13808
+    },
+    {
+      "epoch": 0.5841864793975802,
+      "grad_norm": 0.18919281661510468,
+      "learning_rate": 0.001,
+      "loss": 1.8968,
+      "step": 13809
+    },
+    {
+      "epoch": 0.5842287841610966,
+      "grad_norm": 0.17904122173786163,
+      "learning_rate": 0.001,
+      "loss": 1.8885,
+      "step": 13810
+    },
+    {
+      "epoch": 0.5842710889246129,
+      "grad_norm": 3.1780359745025635,
+      "learning_rate": 0.001,
+      "loss": 2.5823,
+      "step": 13811
+    },
+    {
+      "epoch": 0.5843133936881293,
+      "grad_norm": 0.4842691123485565,
+      "learning_rate": 0.001,
+      "loss": 3.0183,
+      "step": 13812
+    },
+    {
+      "epoch": 0.5843556984516457,
+      "grad_norm": 0.21552209556102753,
+      "learning_rate": 0.001,
+      "loss": 2.5317,
+      "step": 13813
+    },
+    {
+      "epoch": 0.584398003215162,
+      "grad_norm": 2.484487533569336,
+      "learning_rate": 0.001,
+      "loss": 1.738,
+      "step": 13814
+    },
+    {
+      "epoch": 0.5844403079786784,
+      "grad_norm": 0.3187903165817261,
+      "learning_rate": 0.001,
+      "loss": 3.056,
+      "step": 13815
+    },
+    {
+      "epoch": 0.5844826127421948,
+      "grad_norm": 2.3998584747314453,
+      "learning_rate": 0.001,
+      "loss": 1.7838,
+      "step": 13816
+    },
+    {
+      "epoch": 0.5845249175057111,
+      "grad_norm": 0.17576216161251068,
+      "learning_rate": 0.001,
+      "loss": 2.3258,
+      "step": 13817
+    },
+    {
+      "epoch": 0.5845672222692275,
+      "grad_norm": 0.6371609568595886,
+      "learning_rate": 0.001,
+      "loss": 2.2197,
+      "step": 13818
+    },
+    {
+      "epoch": 0.5846095270327439,
+      "grad_norm": 0.20169895887374878,
+      "learning_rate": 0.001,
+      "loss": 2.5768,
+      "step": 13819
+    },
+    {
+      "epoch": 0.5846518317962602,
+      "grad_norm": 0.20052120089530945,
+      "learning_rate": 0.001,
+      "loss": 1.7194,
+      "step": 13820
+    },
+    {
+      "epoch": 0.5846941365597766,
+      "grad_norm": 0.1984112709760666,
+      "learning_rate": 0.001,
+      "loss": 2.0484,
+      "step": 13821
+    },
+    {
+      "epoch": 0.584736441323293,
+      "grad_norm": 0.16683201491832733,
+      "learning_rate": 0.001,
+      "loss": 1.9458,
+      "step": 13822
+    },
+    {
+      "epoch": 0.5847787460868094,
+      "grad_norm": 0.19755448400974274,
+      "learning_rate": 0.001,
+      "loss": 2.6561,
+      "step": 13823
+    },
+    {
+      "epoch": 0.5848210508503258,
+      "grad_norm": 0.18185539543628693,
+      "learning_rate": 0.001,
+      "loss": 2.2397,
+      "step": 13824
+    },
+    {
+      "epoch": 0.5848633556138422,
+      "grad_norm": 0.17746347188949585,
+      "learning_rate": 0.001,
+      "loss": 2.9947,
+      "step": 13825
+    },
+    {
+      "epoch": 0.5849056603773585,
+      "grad_norm": 0.22636628150939941,
+      "learning_rate": 0.001,
+      "loss": 2.5098,
+      "step": 13826
+    },
+    {
+      "epoch": 0.5849479651408749,
+      "grad_norm": 0.5232028365135193,
+      "learning_rate": 0.001,
+      "loss": 3.0739,
+      "step": 13827
+    },
+    {
+      "epoch": 0.5849902699043912,
+      "grad_norm": 0.20873093605041504,
+      "learning_rate": 0.001,
+      "loss": 2.4017,
+      "step": 13828
+    },
+    {
+      "epoch": 0.5850325746679076,
+      "grad_norm": 0.2317352294921875,
+      "learning_rate": 0.001,
+      "loss": 1.5189,
+      "step": 13829
+    },
+    {
+      "epoch": 0.585074879431424,
+      "grad_norm": 2.4872546195983887,
+      "learning_rate": 0.001,
+      "loss": 1.9005,
+      "step": 13830
+    },
+    {
+      "epoch": 0.5851171841949403,
+      "grad_norm": 0.20402368903160095,
+      "learning_rate": 0.001,
+      "loss": 1.6404,
+      "step": 13831
+    },
+    {
+      "epoch": 0.5851594889584567,
+      "grad_norm": 0.16960665583610535,
+      "learning_rate": 0.001,
+      "loss": 1.7397,
+      "step": 13832
+    },
+    {
+      "epoch": 0.5852017937219731,
+      "grad_norm": 0.16686907410621643,
+      "learning_rate": 0.001,
+      "loss": 2.6307,
+      "step": 13833
+    },
+    {
+      "epoch": 0.5852440984854894,
+      "grad_norm": 0.20668341219425201,
+      "learning_rate": 0.001,
+      "loss": 2.2207,
+      "step": 13834
+    },
+    {
+      "epoch": 0.5852864032490058,
+      "grad_norm": 0.17005237936973572,
+      "learning_rate": 0.001,
+      "loss": 2.3992,
+      "step": 13835
+    },
+    {
+      "epoch": 0.5853287080125222,
+      "grad_norm": 0.16195209324359894,
+      "learning_rate": 0.001,
+      "loss": 1.6082,
+      "step": 13836
+    },
+    {
+      "epoch": 0.5853710127760385,
+      "grad_norm": 0.3230375349521637,
+      "learning_rate": 0.001,
+      "loss": 2.2868,
+      "step": 13837
+    },
+    {
+      "epoch": 0.585413317539555,
+      "grad_norm": 0.16415375471115112,
+      "learning_rate": 0.001,
+      "loss": 2.2974,
+      "step": 13838
+    },
+    {
+      "epoch": 0.5854556223030714,
+      "grad_norm": 0.1570240557193756,
+      "learning_rate": 0.001,
+      "loss": 2.3329,
+      "step": 13839
+    },
+    {
+      "epoch": 0.5854979270665877,
+      "grad_norm": 0.17124520242214203,
+      "learning_rate": 0.001,
+      "loss": 1.725,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5855402318301041,
+      "grad_norm": 0.2079612761735916,
+      "learning_rate": 0.001,
+      "loss": 2.2798,
+      "step": 13841
+    },
+    {
+      "epoch": 0.5855825365936205,
+      "grad_norm": 0.20434881746768951,
+      "learning_rate": 0.001,
+      "loss": 2.7584,
+      "step": 13842
+    },
+    {
+      "epoch": 0.5856248413571368,
+      "grad_norm": 0.17397215962409973,
+      "learning_rate": 0.001,
+      "loss": 2.5361,
+      "step": 13843
+    },
+    {
+      "epoch": 0.5856671461206532,
+      "grad_norm": 0.14827987551689148,
+      "learning_rate": 0.001,
+      "loss": 3.5882,
+      "step": 13844
+    },
+    {
+      "epoch": 0.5857094508841696,
+      "grad_norm": 0.16601242125034332,
+      "learning_rate": 0.001,
+      "loss": 1.7631,
+      "step": 13845
+    },
+    {
+      "epoch": 0.5857517556476859,
+      "grad_norm": 0.1840457022190094,
+      "learning_rate": 0.001,
+      "loss": 3.2754,
+      "step": 13846
+    },
+    {
+      "epoch": 0.5857940604112023,
+      "grad_norm": 0.19872917234897614,
+      "learning_rate": 0.001,
+      "loss": 2.24,
+      "step": 13847
+    },
+    {
+      "epoch": 0.5858363651747187,
+      "grad_norm": 0.18587267398834229,
+      "learning_rate": 0.001,
+      "loss": 2.8884,
+      "step": 13848
+    },
+    {
+      "epoch": 0.585878669938235,
+      "grad_norm": 0.5582289695739746,
+      "learning_rate": 0.001,
+      "loss": 2.3321,
+      "step": 13849
+    },
+    {
+      "epoch": 0.5859209747017514,
+      "grad_norm": 0.21963296830654144,
+      "learning_rate": 0.001,
+      "loss": 2.3329,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5859632794652678,
+      "grad_norm": 0.1582070291042328,
+      "learning_rate": 0.001,
+      "loss": 2.5041,
+      "step": 13851
+    },
+    {
+      "epoch": 0.5860055842287841,
+      "grad_norm": 0.1862238198518753,
+      "learning_rate": 0.001,
+      "loss": 1.9722,
+      "step": 13852
+    },
+    {
+      "epoch": 0.5860478889923005,
+      "grad_norm": 0.24395392835140228,
+      "learning_rate": 0.001,
+      "loss": 2.1698,
+      "step": 13853
+    },
+    {
+      "epoch": 0.586090193755817,
+      "grad_norm": 0.19407311081886292,
+      "learning_rate": 0.001,
+      "loss": 2.0222,
+      "step": 13854
+    },
+    {
+      "epoch": 0.5861324985193332,
+      "grad_norm": 0.18697842955589294,
+      "learning_rate": 0.001,
+      "loss": 2.0685,
+      "step": 13855
+    },
+    {
+      "epoch": 0.5861748032828497,
+      "grad_norm": 0.15743017196655273,
+      "learning_rate": 0.001,
+      "loss": 2.9768,
+      "step": 13856
+    },
+    {
+      "epoch": 0.5862171080463661,
+      "grad_norm": 0.1737719476222992,
+      "learning_rate": 0.001,
+      "loss": 1.6942,
+      "step": 13857
+    },
+    {
+      "epoch": 0.5862594128098824,
+      "grad_norm": 0.17090411484241486,
+      "learning_rate": 0.001,
+      "loss": 1.5764,
+      "step": 13858
+    },
+    {
+      "epoch": 0.5863017175733988,
+      "grad_norm": 0.21270331740379333,
+      "learning_rate": 0.001,
+      "loss": 1.7628,
+      "step": 13859
+    },
+    {
+      "epoch": 0.5863440223369152,
+      "grad_norm": 0.16474924981594086,
+      "learning_rate": 0.001,
+      "loss": 2.702,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5863863271004315,
+      "grad_norm": 0.18944130837917328,
+      "learning_rate": 0.001,
+      "loss": 2.5051,
+      "step": 13861
+    },
+    {
+      "epoch": 0.5864286318639479,
+      "grad_norm": 0.1647646725177765,
+      "learning_rate": 0.001,
+      "loss": 2.3672,
+      "step": 13862
+    },
+    {
+      "epoch": 0.5864709366274643,
+      "grad_norm": 0.2273533195257187,
+      "learning_rate": 0.001,
+      "loss": 2.4584,
+      "step": 13863
+    },
+    {
+      "epoch": 0.5865132413909806,
+      "grad_norm": 0.14887486398220062,
+      "learning_rate": 0.001,
+      "loss": 2.4992,
+      "step": 13864
+    },
+    {
+      "epoch": 0.586555546154497,
+      "grad_norm": 0.20371198654174805,
+      "learning_rate": 0.001,
+      "loss": 2.2056,
+      "step": 13865
+    },
+    {
+      "epoch": 0.5865978509180134,
+      "grad_norm": 0.9710887670516968,
+      "learning_rate": 0.001,
+      "loss": 1.9499,
+      "step": 13866
+    },
+    {
+      "epoch": 0.5866401556815297,
+      "grad_norm": 0.16241279244422913,
+      "learning_rate": 0.001,
+      "loss": 1.9251,
+      "step": 13867
+    },
+    {
+      "epoch": 0.5866824604450461,
+      "grad_norm": 0.401077002286911,
+      "learning_rate": 0.001,
+      "loss": 2.3491,
+      "step": 13868
+    },
+    {
+      "epoch": 0.5867247652085625,
+      "grad_norm": 0.16556090116500854,
+      "learning_rate": 0.001,
+      "loss": 1.7719,
+      "step": 13869
+    },
+    {
+      "epoch": 0.5867670699720788,
+      "grad_norm": 0.14938072860240936,
+      "learning_rate": 0.001,
+      "loss": 1.8205,
+      "step": 13870
+    },
+    {
+      "epoch": 0.5868093747355952,
+      "grad_norm": 0.1616201102733612,
+      "learning_rate": 0.001,
+      "loss": 2.3958,
+      "step": 13871
+    },
+    {
+      "epoch": 0.5868516794991115,
+      "grad_norm": 0.19175079464912415,
+      "learning_rate": 0.001,
+      "loss": 1.9595,
+      "step": 13872
+    },
+    {
+      "epoch": 0.586893984262628,
+      "grad_norm": 0.21373534202575684,
+      "learning_rate": 0.001,
+      "loss": 2.2095,
+      "step": 13873
+    },
+    {
+      "epoch": 0.5869362890261444,
+      "grad_norm": 0.20551612973213196,
+      "learning_rate": 0.001,
+      "loss": 2.4731,
+      "step": 13874
+    },
+    {
+      "epoch": 0.5869785937896607,
+      "grad_norm": 0.2638375759124756,
+      "learning_rate": 0.001,
+      "loss": 2.8643,
+      "step": 13875
+    },
+    {
+      "epoch": 0.5870208985531771,
+      "grad_norm": 0.1855308711528778,
+      "learning_rate": 0.001,
+      "loss": 1.8964,
+      "step": 13876
+    },
+    {
+      "epoch": 0.5870632033166935,
+      "grad_norm": 0.2503952383995056,
+      "learning_rate": 0.001,
+      "loss": 2.9562,
+      "step": 13877
+    },
+    {
+      "epoch": 0.5871055080802098,
+      "grad_norm": 0.15639376640319824,
+      "learning_rate": 0.001,
+      "loss": 2.6518,
+      "step": 13878
+    },
+    {
+      "epoch": 0.5871478128437262,
+      "grad_norm": 0.3068271279335022,
+      "learning_rate": 0.001,
+      "loss": 2.9139,
+      "step": 13879
+    },
+    {
+      "epoch": 0.5871901176072426,
+      "grad_norm": 2.3085885047912598,
+      "learning_rate": 0.001,
+      "loss": 1.9774,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5872324223707589,
+      "grad_norm": 0.17612521350383759,
+      "learning_rate": 0.001,
+      "loss": 2.3076,
+      "step": 13881
+    },
+    {
+      "epoch": 0.5872747271342753,
+      "grad_norm": 0.21990317106246948,
+      "learning_rate": 0.001,
+      "loss": 2.7583,
+      "step": 13882
+    },
+    {
+      "epoch": 0.5873170318977917,
+      "grad_norm": 0.15076623857021332,
+      "learning_rate": 0.001,
+      "loss": 1.9021,
+      "step": 13883
+    },
+    {
+      "epoch": 0.587359336661308,
+      "grad_norm": 0.17355187237262726,
+      "learning_rate": 0.001,
+      "loss": 2.1959,
+      "step": 13884
+    },
+    {
+      "epoch": 0.5874016414248244,
+      "grad_norm": 0.16059938073158264,
+      "learning_rate": 0.001,
+      "loss": 2.5057,
+      "step": 13885
+    },
+    {
+      "epoch": 0.5874439461883408,
+      "grad_norm": 0.17203792929649353,
+      "learning_rate": 0.001,
+      "loss": 2.2947,
+      "step": 13886
+    },
+    {
+      "epoch": 0.5874862509518571,
+      "grad_norm": 0.15990357100963593,
+      "learning_rate": 0.001,
+      "loss": 1.7769,
+      "step": 13887
+    },
+    {
+      "epoch": 0.5875285557153735,
+      "grad_norm": 0.1762009710073471,
+      "learning_rate": 0.001,
+      "loss": 3.1737,
+      "step": 13888
+    },
+    {
+      "epoch": 0.58757086047889,
+      "grad_norm": 0.14753255248069763,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 13889
+    },
+    {
+      "epoch": 0.5876131652424063,
+      "grad_norm": 0.20030535757541656,
+      "learning_rate": 0.001,
+      "loss": 2.094,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5876554700059227,
+      "grad_norm": 0.1795814037322998,
+      "learning_rate": 0.001,
+      "loss": 2.438,
+      "step": 13891
+    },
+    {
+      "epoch": 0.5876977747694391,
+      "grad_norm": 0.2042204886674881,
+      "learning_rate": 0.001,
+      "loss": 1.9378,
+      "step": 13892
+    },
+    {
+      "epoch": 0.5877400795329554,
+      "grad_norm": 0.18718861043453217,
+      "learning_rate": 0.001,
+      "loss": 3.3314,
+      "step": 13893
+    },
+    {
+      "epoch": 0.5877823842964718,
+      "grad_norm": 0.18534649908542633,
+      "learning_rate": 0.001,
+      "loss": 1.9572,
+      "step": 13894
+    },
+    {
+      "epoch": 0.5878246890599882,
+      "grad_norm": 0.14680497348308563,
+      "learning_rate": 0.001,
+      "loss": 1.6453,
+      "step": 13895
+    },
+    {
+      "epoch": 0.5878669938235045,
+      "grad_norm": 0.22413283586502075,
+      "learning_rate": 0.001,
+      "loss": 2.7833,
+      "step": 13896
+    },
+    {
+      "epoch": 0.5879092985870209,
+      "grad_norm": 0.20671682059764862,
+      "learning_rate": 0.001,
+      "loss": 3.5421,
+      "step": 13897
+    },
+    {
+      "epoch": 0.5879516033505373,
+      "grad_norm": 0.3078271448612213,
+      "learning_rate": 0.001,
+      "loss": 3.6566,
+      "step": 13898
+    },
+    {
+      "epoch": 0.5879939081140536,
+      "grad_norm": 0.3358747661113739,
+      "learning_rate": 0.001,
+      "loss": 1.9084,
+      "step": 13899
+    },
+    {
+      "epoch": 0.58803621287757,
+      "grad_norm": 0.15469489991664886,
+      "learning_rate": 0.001,
+      "loss": 1.4976,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5880785176410864,
+      "grad_norm": 0.19083096086978912,
+      "learning_rate": 0.001,
+      "loss": 2.5376,
+      "step": 13901
+    },
+    {
+      "epoch": 0.5881208224046027,
+      "grad_norm": 0.16537901759147644,
+      "learning_rate": 0.001,
+      "loss": 1.4644,
+      "step": 13902
+    },
+    {
+      "epoch": 0.5881631271681191,
+      "grad_norm": 0.1859930157661438,
+      "learning_rate": 0.001,
+      "loss": 2.2167,
+      "step": 13903
+    },
+    {
+      "epoch": 0.5882054319316355,
+      "grad_norm": 0.17841875553131104,
+      "learning_rate": 0.001,
+      "loss": 1.7491,
+      "step": 13904
+    },
+    {
+      "epoch": 0.5882477366951518,
+      "grad_norm": 0.17977175116539001,
+      "learning_rate": 0.001,
+      "loss": 2.1064,
+      "step": 13905
+    },
+    {
+      "epoch": 0.5882900414586683,
+      "grad_norm": 0.33689969778060913,
+      "learning_rate": 0.001,
+      "loss": 2.7081,
+      "step": 13906
+    },
+    {
+      "epoch": 0.5883323462221847,
+      "grad_norm": 0.19339759647846222,
+      "learning_rate": 0.001,
+      "loss": 2.108,
+      "step": 13907
+    },
+    {
+      "epoch": 0.588374650985701,
+      "grad_norm": 0.14537258446216583,
+      "learning_rate": 0.001,
+      "loss": 2.1058,
+      "step": 13908
+    },
+    {
+      "epoch": 0.5884169557492174,
+      "grad_norm": 0.19922007620334625,
+      "learning_rate": 0.001,
+      "loss": 1.4782,
+      "step": 13909
+    },
+    {
+      "epoch": 0.5884592605127338,
+      "grad_norm": 0.1968841850757599,
+      "learning_rate": 0.001,
+      "loss": 2.6562,
+      "step": 13910
+    },
+    {
+      "epoch": 0.5885015652762501,
+      "grad_norm": 0.15195704996585846,
+      "learning_rate": 0.001,
+      "loss": 2.0136,
+      "step": 13911
+    },
+    {
+      "epoch": 0.5885438700397665,
+      "grad_norm": 0.17894305288791656,
+      "learning_rate": 0.001,
+      "loss": 3.5737,
+      "step": 13912
+    },
+    {
+      "epoch": 0.5885861748032829,
+      "grad_norm": 0.2585073411464691,
+      "learning_rate": 0.001,
+      "loss": 2.9576,
+      "step": 13913
+    },
+    {
+      "epoch": 0.5886284795667992,
+      "grad_norm": 0.240971177816391,
+      "learning_rate": 0.001,
+      "loss": 2.4537,
+      "step": 13914
+    },
+    {
+      "epoch": 0.5886707843303156,
+      "grad_norm": 0.1822543740272522,
+      "learning_rate": 0.001,
+      "loss": 2.2145,
+      "step": 13915
+    },
+    {
+      "epoch": 0.588713089093832,
+      "grad_norm": 0.1564810425043106,
+      "learning_rate": 0.001,
+      "loss": 2.5292,
+      "step": 13916
+    },
+    {
+      "epoch": 0.5887553938573483,
+      "grad_norm": 0.16015750169754028,
+      "learning_rate": 0.001,
+      "loss": 2.2064,
+      "step": 13917
+    },
+    {
+      "epoch": 0.5887976986208647,
+      "grad_norm": 0.2196504920721054,
+      "learning_rate": 0.001,
+      "loss": 2.7055,
+      "step": 13918
+    },
+    {
+      "epoch": 0.588840003384381,
+      "grad_norm": 0.19382864236831665,
+      "learning_rate": 0.001,
+      "loss": 3.4921,
+      "step": 13919
+    },
+    {
+      "epoch": 0.5888823081478974,
+      "grad_norm": 0.1581316739320755,
+      "learning_rate": 0.001,
+      "loss": 2.4825,
+      "step": 13920
+    },
+    {
+      "epoch": 0.5889246129114138,
+      "grad_norm": 0.168742835521698,
+      "learning_rate": 0.001,
+      "loss": 1.8284,
+      "step": 13921
+    },
+    {
+      "epoch": 0.5889669176749301,
+      "grad_norm": 0.17295318841934204,
+      "learning_rate": 0.001,
+      "loss": 2.2552,
+      "step": 13922
+    },
+    {
+      "epoch": 0.5890092224384466,
+      "grad_norm": 0.14867152273654938,
+      "learning_rate": 0.001,
+      "loss": 1.4105,
+      "step": 13923
+    },
+    {
+      "epoch": 0.589051527201963,
+      "grad_norm": 0.19675691425800323,
+      "learning_rate": 0.001,
+      "loss": 1.9656,
+      "step": 13924
+    },
+    {
+      "epoch": 0.5890938319654793,
+      "grad_norm": 0.17681030929088593,
+      "learning_rate": 0.001,
+      "loss": 3.0477,
+      "step": 13925
+    },
+    {
+      "epoch": 0.5891361367289957,
+      "grad_norm": 0.5007891654968262,
+      "learning_rate": 0.001,
+      "loss": 2.2316,
+      "step": 13926
+    },
+    {
+      "epoch": 0.5891784414925121,
+      "grad_norm": 0.6708447337150574,
+      "learning_rate": 0.001,
+      "loss": 1.6344,
+      "step": 13927
+    },
+    {
+      "epoch": 0.5892207462560284,
+      "grad_norm": 0.19077329337596893,
+      "learning_rate": 0.001,
+      "loss": 2.792,
+      "step": 13928
+    },
+    {
+      "epoch": 0.5892630510195448,
+      "grad_norm": 0.28526973724365234,
+      "learning_rate": 0.001,
+      "loss": 2.514,
+      "step": 13929
+    },
+    {
+      "epoch": 0.5893053557830612,
+      "grad_norm": 0.34212541580200195,
+      "learning_rate": 0.001,
+      "loss": 1.6885,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5893476605465775,
+      "grad_norm": 0.17507655918598175,
+      "learning_rate": 0.001,
+      "loss": 2.1797,
+      "step": 13931
+    },
+    {
+      "epoch": 0.5893899653100939,
+      "grad_norm": 1.8926448822021484,
+      "learning_rate": 0.001,
+      "loss": 1.746,
+      "step": 13932
+    },
+    {
+      "epoch": 0.5894322700736103,
+      "grad_norm": 0.1368320882320404,
+      "learning_rate": 0.001,
+      "loss": 1.5929,
+      "step": 13933
+    },
+    {
+      "epoch": 0.5894745748371266,
+      "grad_norm": 0.20840121805667877,
+      "learning_rate": 0.001,
+      "loss": 2.0687,
+      "step": 13934
+    },
+    {
+      "epoch": 0.589516879600643,
+      "grad_norm": 0.3458685576915741,
+      "learning_rate": 0.001,
+      "loss": 1.7513,
+      "step": 13935
+    },
+    {
+      "epoch": 0.5895591843641594,
+      "grad_norm": 0.17107515037059784,
+      "learning_rate": 0.001,
+      "loss": 1.8047,
+      "step": 13936
+    },
+    {
+      "epoch": 0.5896014891276757,
+      "grad_norm": 0.37925609946250916,
+      "learning_rate": 0.001,
+      "loss": 2.0818,
+      "step": 13937
+    },
+    {
+      "epoch": 0.5896437938911921,
+      "grad_norm": 0.31357237696647644,
+      "learning_rate": 0.001,
+      "loss": 2.9934,
+      "step": 13938
+    },
+    {
+      "epoch": 0.5896860986547086,
+      "grad_norm": 0.16676288843154907,
+      "learning_rate": 0.001,
+      "loss": 2.134,
+      "step": 13939
+    },
+    {
+      "epoch": 0.5897284034182249,
+      "grad_norm": 0.17158526182174683,
+      "learning_rate": 0.001,
+      "loss": 2.1719,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5897707081817413,
+      "grad_norm": 0.2076321244239807,
+      "learning_rate": 0.001,
+      "loss": 2.6721,
+      "step": 13941
+    },
+    {
+      "epoch": 0.5898130129452577,
+      "grad_norm": 0.14922165870666504,
+      "learning_rate": 0.001,
+      "loss": 2.6749,
+      "step": 13942
+    },
+    {
+      "epoch": 0.589855317708774,
+      "grad_norm": 0.3216572403907776,
+      "learning_rate": 0.001,
+      "loss": 2.0285,
+      "step": 13943
+    },
+    {
+      "epoch": 0.5898976224722904,
+      "grad_norm": 0.2103135734796524,
+      "learning_rate": 0.001,
+      "loss": 1.9216,
+      "step": 13944
+    },
+    {
+      "epoch": 0.5899399272358068,
+      "grad_norm": 0.16740508377552032,
+      "learning_rate": 0.001,
+      "loss": 2.4028,
+      "step": 13945
+    },
+    {
+      "epoch": 0.5899822319993231,
+      "grad_norm": 0.2918654680252075,
+      "learning_rate": 0.001,
+      "loss": 2.7537,
+      "step": 13946
+    },
+    {
+      "epoch": 0.5900245367628395,
+      "grad_norm": 0.18662185966968536,
+      "learning_rate": 0.001,
+      "loss": 1.6938,
+      "step": 13947
+    },
+    {
+      "epoch": 0.5900668415263559,
+      "grad_norm": 0.19983160495758057,
+      "learning_rate": 0.001,
+      "loss": 2.2847,
+      "step": 13948
+    },
+    {
+      "epoch": 0.5901091462898722,
+      "grad_norm": 0.21256953477859497,
+      "learning_rate": 0.001,
+      "loss": 3.3585,
+      "step": 13949
+    },
+    {
+      "epoch": 0.5901514510533886,
+      "grad_norm": 0.1589345932006836,
+      "learning_rate": 0.001,
+      "loss": 3.151,
+      "step": 13950
+    },
+    {
+      "epoch": 0.590193755816905,
+      "grad_norm": 0.5611705183982849,
+      "learning_rate": 0.001,
+      "loss": 2.2215,
+      "step": 13951
+    },
+    {
+      "epoch": 0.5902360605804213,
+      "grad_norm": 0.19331638514995575,
+      "learning_rate": 0.001,
+      "loss": 3.0221,
+      "step": 13952
+    },
+    {
+      "epoch": 0.5902783653439377,
+      "grad_norm": 0.17588597536087036,
+      "learning_rate": 0.001,
+      "loss": 1.8878,
+      "step": 13953
+    },
+    {
+      "epoch": 0.5903206701074541,
+      "grad_norm": 0.14944405853748322,
+      "learning_rate": 0.001,
+      "loss": 2.3562,
+      "step": 13954
+    },
+    {
+      "epoch": 0.5903629748709704,
+      "grad_norm": 0.15163281559944153,
+      "learning_rate": 0.001,
+      "loss": 1.2766,
+      "step": 13955
+    },
+    {
+      "epoch": 0.5904052796344869,
+      "grad_norm": 0.1893131136894226,
+      "learning_rate": 0.001,
+      "loss": 2.3026,
+      "step": 13956
+    },
+    {
+      "epoch": 0.5904475843980033,
+      "grad_norm": 0.17503425478935242,
+      "learning_rate": 0.001,
+      "loss": 2.9459,
+      "step": 13957
+    },
+    {
+      "epoch": 0.5904898891615196,
+      "grad_norm": 0.2079678177833557,
+      "learning_rate": 0.001,
+      "loss": 1.8288,
+      "step": 13958
+    },
+    {
+      "epoch": 0.590532193925036,
+      "grad_norm": 0.4229949116706848,
+      "learning_rate": 0.001,
+      "loss": 2.3859,
+      "step": 13959
+    },
+    {
+      "epoch": 0.5905744986885524,
+      "grad_norm": 0.1723388284444809,
+      "learning_rate": 0.001,
+      "loss": 2.6868,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5906168034520687,
+      "grad_norm": 0.16546562314033508,
+      "learning_rate": 0.001,
+      "loss": 1.7051,
+      "step": 13961
+    },
+    {
+      "epoch": 0.5906591082155851,
+      "grad_norm": 0.1712961196899414,
+      "learning_rate": 0.001,
+      "loss": 1.9134,
+      "step": 13962
+    },
+    {
+      "epoch": 0.5907014129791014,
+      "grad_norm": 0.18814264237880707,
+      "learning_rate": 0.001,
+      "loss": 2.4536,
+      "step": 13963
+    },
+    {
+      "epoch": 0.5907437177426178,
+      "grad_norm": 1.0132697820663452,
+      "learning_rate": 0.001,
+      "loss": 2.2008,
+      "step": 13964
+    },
+    {
+      "epoch": 0.5907860225061342,
+      "grad_norm": 0.16948258876800537,
+      "learning_rate": 0.001,
+      "loss": 1.9644,
+      "step": 13965
+    },
+    {
+      "epoch": 0.5908283272696505,
+      "grad_norm": 0.17865702509880066,
+      "learning_rate": 0.001,
+      "loss": 3.5171,
+      "step": 13966
+    },
+    {
+      "epoch": 0.5908706320331669,
+      "grad_norm": 0.1568206250667572,
+      "learning_rate": 0.001,
+      "loss": 2.2852,
+      "step": 13967
+    },
+    {
+      "epoch": 0.5909129367966833,
+      "grad_norm": 0.18641190230846405,
+      "learning_rate": 0.001,
+      "loss": 1.5695,
+      "step": 13968
+    },
+    {
+      "epoch": 0.5909552415601996,
+      "grad_norm": 0.484352707862854,
+      "learning_rate": 0.001,
+      "loss": 2.8019,
+      "step": 13969
+    },
+    {
+      "epoch": 0.590997546323716,
+      "grad_norm": 0.1573459804058075,
+      "learning_rate": 0.001,
+      "loss": 1.5844,
+      "step": 13970
+    },
+    {
+      "epoch": 0.5910398510872324,
+      "grad_norm": 0.236774280667305,
+      "learning_rate": 0.001,
+      "loss": 2.7873,
+      "step": 13971
+    },
+    {
+      "epoch": 0.5910821558507487,
+      "grad_norm": 1.0169886350631714,
+      "learning_rate": 0.001,
+      "loss": 1.9265,
+      "step": 13972
+    },
+    {
+      "epoch": 0.5911244606142652,
+      "grad_norm": 0.19387690722942352,
+      "learning_rate": 0.001,
+      "loss": 1.6925,
+      "step": 13973
+    },
+    {
+      "epoch": 0.5911667653777816,
+      "grad_norm": 0.17645257711410522,
+      "learning_rate": 0.001,
+      "loss": 2.2239,
+      "step": 13974
+    },
+    {
+      "epoch": 0.5912090701412979,
+      "grad_norm": 0.1641792207956314,
+      "learning_rate": 0.001,
+      "loss": 2.3921,
+      "step": 13975
+    },
+    {
+      "epoch": 0.5912513749048143,
+      "grad_norm": 0.20244446396827698,
+      "learning_rate": 0.001,
+      "loss": 2.133,
+      "step": 13976
+    },
+    {
+      "epoch": 0.5912936796683307,
+      "grad_norm": 0.2241334766149521,
+      "learning_rate": 0.001,
+      "loss": 2.7806,
+      "step": 13977
+    },
+    {
+      "epoch": 0.591335984431847,
+      "grad_norm": 0.16886314749717712,
+      "learning_rate": 0.001,
+      "loss": 2.3937,
+      "step": 13978
+    },
+    {
+      "epoch": 0.5913782891953634,
+      "grad_norm": 0.17606092989444733,
+      "learning_rate": 0.001,
+      "loss": 1.5095,
+      "step": 13979
+    },
+    {
+      "epoch": 0.5914205939588798,
+      "grad_norm": 0.16961275041103363,
+      "learning_rate": 0.001,
+      "loss": 2.3644,
+      "step": 13980
+    },
+    {
+      "epoch": 0.5914628987223961,
+      "grad_norm": 0.16681556403636932,
+      "learning_rate": 0.001,
+      "loss": 2.0659,
+      "step": 13981
+    },
+    {
+      "epoch": 0.5915052034859125,
+      "grad_norm": 0.2830229699611664,
+      "learning_rate": 0.001,
+      "loss": 1.8581,
+      "step": 13982
+    },
+    {
+      "epoch": 0.5915475082494289,
+      "grad_norm": 0.6050325036048889,
+      "learning_rate": 0.001,
+      "loss": 2.2993,
+      "step": 13983
+    },
+    {
+      "epoch": 0.5915898130129452,
+      "grad_norm": 0.17912669479846954,
+      "learning_rate": 0.001,
+      "loss": 2.4188,
+      "step": 13984
+    },
+    {
+      "epoch": 0.5916321177764616,
+      "grad_norm": 0.19286015629768372,
+      "learning_rate": 0.001,
+      "loss": 2.7553,
+      "step": 13985
+    },
+    {
+      "epoch": 0.591674422539978,
+      "grad_norm": 0.1747298538684845,
+      "learning_rate": 0.001,
+      "loss": 1.9951,
+      "step": 13986
+    },
+    {
+      "epoch": 0.5917167273034943,
+      "grad_norm": 0.24552187323570251,
+      "learning_rate": 0.001,
+      "loss": 2.9972,
+      "step": 13987
+    },
+    {
+      "epoch": 0.5917590320670107,
+      "grad_norm": 0.3392343521118164,
+      "learning_rate": 0.001,
+      "loss": 2.7581,
+      "step": 13988
+    },
+    {
+      "epoch": 0.5918013368305272,
+      "grad_norm": 0.22997502982616425,
+      "learning_rate": 0.001,
+      "loss": 2.4476,
+      "step": 13989
+    },
+    {
+      "epoch": 0.5918436415940435,
+      "grad_norm": 0.17258018255233765,
+      "learning_rate": 0.001,
+      "loss": 1.8377,
+      "step": 13990
+    },
+    {
+      "epoch": 0.5918859463575599,
+      "grad_norm": 0.6078843474388123,
+      "learning_rate": 0.001,
+      "loss": 2.6106,
+      "step": 13991
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.19402997195720673,
+      "learning_rate": 0.001,
+      "loss": 2.5247,
+      "step": 13992
+    },
+    {
+      "epoch": 0.5919705558845926,
+      "grad_norm": 0.1636909693479538,
+      "learning_rate": 0.001,
+      "loss": 2.0067,
+      "step": 13993
+    },
+    {
+      "epoch": 0.592012860648109,
+      "grad_norm": 0.18842186033725739,
+      "learning_rate": 0.001,
+      "loss": 1.6082,
+      "step": 13994
+    },
+    {
+      "epoch": 0.5920551654116254,
+      "grad_norm": 0.1705595999956131,
+      "learning_rate": 0.001,
+      "loss": 1.6876,
+      "step": 13995
+    },
+    {
+      "epoch": 0.5920974701751417,
+      "grad_norm": 0.23243148624897003,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 13996
+    },
+    {
+      "epoch": 0.5921397749386581,
+      "grad_norm": 0.18063674867153168,
+      "learning_rate": 0.001,
+      "loss": 2.2572,
+      "step": 13997
+    },
+    {
+      "epoch": 0.5921820797021745,
+      "grad_norm": 0.21057967841625214,
+      "learning_rate": 0.001,
+      "loss": 1.7763,
+      "step": 13998
+    },
+    {
+      "epoch": 0.5922243844656908,
+      "grad_norm": 0.19632886350154877,
+      "learning_rate": 0.001,
+      "loss": 1.9372,
+      "step": 13999
+    },
+    {
+      "epoch": 0.5922666892292072,
+      "grad_norm": 0.16427557170391083,
+      "learning_rate": 0.001,
+      "loss": 1.974,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5923089939927236,
+      "grad_norm": 0.15664273500442505,
+      "learning_rate": 0.001,
+      "loss": 1.7737,
+      "step": 14001
+    },
+    {
+      "epoch": 0.5923512987562399,
+      "grad_norm": 15.016242027282715,
+      "learning_rate": 0.001,
+      "loss": 2.0841,
+      "step": 14002
+    },
+    {
+      "epoch": 0.5923936035197563,
+      "grad_norm": 0.14031845331192017,
+      "learning_rate": 0.001,
+      "loss": 2.2924,
+      "step": 14003
+    },
+    {
+      "epoch": 0.5924359082832727,
+      "grad_norm": 0.17053398489952087,
+      "learning_rate": 0.001,
+      "loss": 2.4202,
+      "step": 14004
+    },
+    {
+      "epoch": 0.592478213046789,
+      "grad_norm": 0.1829700618982315,
+      "learning_rate": 0.001,
+      "loss": 2.0318,
+      "step": 14005
+    },
+    {
+      "epoch": 0.5925205178103055,
+      "grad_norm": 0.16647842526435852,
+      "learning_rate": 0.001,
+      "loss": 2.2545,
+      "step": 14006
+    },
+    {
+      "epoch": 0.5925628225738218,
+      "grad_norm": 0.1678646355867386,
+      "learning_rate": 0.001,
+      "loss": 1.8475,
+      "step": 14007
+    },
+    {
+      "epoch": 0.5926051273373382,
+      "grad_norm": 0.2268977016210556,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 14008
+    },
+    {
+      "epoch": 0.5926474321008546,
+      "grad_norm": 0.16663521528244019,
+      "learning_rate": 0.001,
+      "loss": 1.7448,
+      "step": 14009
+    },
+    {
+      "epoch": 0.5926897368643709,
+      "grad_norm": 0.33862343430519104,
+      "learning_rate": 0.001,
+      "loss": 1.9905,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5927320416278873,
+      "grad_norm": 0.5181235671043396,
+      "learning_rate": 0.001,
+      "loss": 2.3188,
+      "step": 14011
+    },
+    {
+      "epoch": 0.5927743463914037,
+      "grad_norm": 0.15368396043777466,
+      "learning_rate": 0.001,
+      "loss": 2.6473,
+      "step": 14012
+    },
+    {
+      "epoch": 0.59281665115492,
+      "grad_norm": 0.18206775188446045,
+      "learning_rate": 0.001,
+      "loss": 1.3312,
+      "step": 14013
+    },
+    {
+      "epoch": 0.5928589559184364,
+      "grad_norm": 0.21230578422546387,
+      "learning_rate": 0.001,
+      "loss": 1.8037,
+      "step": 14014
+    },
+    {
+      "epoch": 0.5929012606819528,
+      "grad_norm": 0.1920211762189865,
+      "learning_rate": 0.001,
+      "loss": 2.3031,
+      "step": 14015
+    },
+    {
+      "epoch": 0.5929435654454691,
+      "grad_norm": 0.1721087396144867,
+      "learning_rate": 0.001,
+      "loss": 2.0063,
+      "step": 14016
+    },
+    {
+      "epoch": 0.5929858702089855,
+      "grad_norm": 0.17139104008674622,
+      "learning_rate": 0.001,
+      "loss": 1.8707,
+      "step": 14017
+    },
+    {
+      "epoch": 0.5930281749725019,
+      "grad_norm": 0.2000657021999359,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 14018
+    },
+    {
+      "epoch": 0.5930704797360182,
+      "grad_norm": 2.25368595123291,
+      "learning_rate": 0.001,
+      "loss": 2.2605,
+      "step": 14019
+    },
+    {
+      "epoch": 0.5931127844995346,
+      "grad_norm": 0.17565739154815674,
+      "learning_rate": 0.001,
+      "loss": 2.4521,
+      "step": 14020
+    },
+    {
+      "epoch": 0.593155089263051,
+      "grad_norm": 0.7381449341773987,
+      "learning_rate": 0.001,
+      "loss": 2.9677,
+      "step": 14021
+    },
+    {
+      "epoch": 0.5931973940265673,
+      "grad_norm": 0.17680025100708008,
+      "learning_rate": 0.001,
+      "loss": 2.6527,
+      "step": 14022
+    },
+    {
+      "epoch": 0.5932396987900838,
+      "grad_norm": 0.6374967694282532,
+      "learning_rate": 0.001,
+      "loss": 2.7785,
+      "step": 14023
+    },
+    {
+      "epoch": 0.5932820035536002,
+      "grad_norm": 0.18539991974830627,
+      "learning_rate": 0.001,
+      "loss": 2.1156,
+      "step": 14024
+    },
+    {
+      "epoch": 0.5933243083171165,
+      "grad_norm": 0.17498497664928436,
+      "learning_rate": 0.001,
+      "loss": 1.9663,
+      "step": 14025
+    },
+    {
+      "epoch": 0.5933666130806329,
+      "grad_norm": 0.16935625672340393,
+      "learning_rate": 0.001,
+      "loss": 1.7092,
+      "step": 14026
+    },
+    {
+      "epoch": 0.5934089178441493,
+      "grad_norm": 0.1708284169435501,
+      "learning_rate": 0.001,
+      "loss": 1.619,
+      "step": 14027
+    },
+    {
+      "epoch": 0.5934512226076656,
+      "grad_norm": 0.1765231490135193,
+      "learning_rate": 0.001,
+      "loss": 1.8157,
+      "step": 14028
+    },
+    {
+      "epoch": 0.593493527371182,
+      "grad_norm": 0.21221309900283813,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 14029
+    },
+    {
+      "epoch": 0.5935358321346984,
+      "grad_norm": 16.430587768554688,
+      "learning_rate": 0.001,
+      "loss": 2.7351,
+      "step": 14030
+    },
+    {
+      "epoch": 0.5935781368982147,
+      "grad_norm": 0.15817879140377045,
+      "learning_rate": 0.001,
+      "loss": 2.0115,
+      "step": 14031
+    },
+    {
+      "epoch": 0.5936204416617311,
+      "grad_norm": 0.2005200982093811,
+      "learning_rate": 0.001,
+      "loss": 1.8476,
+      "step": 14032
+    },
+    {
+      "epoch": 0.5936627464252475,
+      "grad_norm": 0.18343733251094818,
+      "learning_rate": 0.001,
+      "loss": 1.9199,
+      "step": 14033
+    },
+    {
+      "epoch": 0.5937050511887638,
+      "grad_norm": 13.979016304016113,
+      "learning_rate": 0.001,
+      "loss": 1.6957,
+      "step": 14034
+    },
+    {
+      "epoch": 0.5937473559522802,
+      "grad_norm": 1.4430853128433228,
+      "learning_rate": 0.001,
+      "loss": 3.6849,
+      "step": 14035
+    },
+    {
+      "epoch": 0.5937896607157966,
+      "grad_norm": 0.22609612345695496,
+      "learning_rate": 0.001,
+      "loss": 3.7324,
+      "step": 14036
+    },
+    {
+      "epoch": 0.5938319654793129,
+      "grad_norm": 0.3302842080593109,
+      "learning_rate": 0.001,
+      "loss": 2.7874,
+      "step": 14037
+    },
+    {
+      "epoch": 0.5938742702428293,
+      "grad_norm": 0.39011886715888977,
+      "learning_rate": 0.001,
+      "loss": 2.6145,
+      "step": 14038
+    },
+    {
+      "epoch": 0.5939165750063458,
+      "grad_norm": 0.21780306100845337,
+      "learning_rate": 0.001,
+      "loss": 1.8593,
+      "step": 14039
+    },
+    {
+      "epoch": 0.5939588797698621,
+      "grad_norm": 0.2653607130050659,
+      "learning_rate": 0.001,
+      "loss": 2.3802,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5940011845333785,
+      "grad_norm": 0.18377161026000977,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 14041
+    },
+    {
+      "epoch": 0.5940434892968949,
+      "grad_norm": 0.22068673372268677,
+      "learning_rate": 0.001,
+      "loss": 3.5105,
+      "step": 14042
+    },
+    {
+      "epoch": 0.5940857940604112,
+      "grad_norm": 0.2928449213504791,
+      "learning_rate": 0.001,
+      "loss": 2.2201,
+      "step": 14043
+    },
+    {
+      "epoch": 0.5941280988239276,
+      "grad_norm": 5.771327495574951,
+      "learning_rate": 0.001,
+      "loss": 3.2931,
+      "step": 14044
+    },
+    {
+      "epoch": 0.594170403587444,
+      "grad_norm": 0.24995769560337067,
+      "learning_rate": 0.001,
+      "loss": 3.6736,
+      "step": 14045
+    },
+    {
+      "epoch": 0.5942127083509603,
+      "grad_norm": 1.1308993101119995,
+      "learning_rate": 0.001,
+      "loss": 2.1198,
+      "step": 14046
+    },
+    {
+      "epoch": 0.5942550131144767,
+      "grad_norm": 0.16081666946411133,
+      "learning_rate": 0.001,
+      "loss": 1.9559,
+      "step": 14047
+    },
+    {
+      "epoch": 0.5942973178779931,
+      "grad_norm": 0.8098224997520447,
+      "learning_rate": 0.001,
+      "loss": 3.7266,
+      "step": 14048
+    },
+    {
+      "epoch": 0.5943396226415094,
+      "grad_norm": 6.794950008392334,
+      "learning_rate": 0.001,
+      "loss": 2.869,
+      "step": 14049
+    },
+    {
+      "epoch": 0.5943819274050258,
+      "grad_norm": 0.2231823056936264,
+      "learning_rate": 0.001,
+      "loss": 2.455,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5944242321685422,
+      "grad_norm": 0.23370391130447388,
+      "learning_rate": 0.001,
+      "loss": 2.4393,
+      "step": 14051
+    },
+    {
+      "epoch": 0.5944665369320585,
+      "grad_norm": 0.300771564245224,
+      "learning_rate": 0.001,
+      "loss": 2.7126,
+      "step": 14052
+    },
+    {
+      "epoch": 0.5945088416955749,
+      "grad_norm": 0.9046683311462402,
+      "learning_rate": 0.001,
+      "loss": 3.1607,
+      "step": 14053
+    },
+    {
+      "epoch": 0.5945511464590912,
+      "grad_norm": 0.2239486575126648,
+      "learning_rate": 0.001,
+      "loss": 2.566,
+      "step": 14054
+    },
+    {
+      "epoch": 0.5945934512226076,
+      "grad_norm": 0.20858706533908844,
+      "learning_rate": 0.001,
+      "loss": 1.9689,
+      "step": 14055
+    },
+    {
+      "epoch": 0.5946357559861241,
+      "grad_norm": 0.24165865778923035,
+      "learning_rate": 0.001,
+      "loss": 2.2796,
+      "step": 14056
+    },
+    {
+      "epoch": 0.5946780607496404,
+      "grad_norm": 0.18632391095161438,
+      "learning_rate": 0.001,
+      "loss": 2.541,
+      "step": 14057
+    },
+    {
+      "epoch": 0.5947203655131568,
+      "grad_norm": 0.21690063178539276,
+      "learning_rate": 0.001,
+      "loss": 2.8571,
+      "step": 14058
+    },
+    {
+      "epoch": 0.5947626702766732,
+      "grad_norm": 0.2051665484905243,
+      "learning_rate": 0.001,
+      "loss": 3.3445,
+      "step": 14059
+    },
+    {
+      "epoch": 0.5948049750401895,
+      "grad_norm": 0.2277742475271225,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5948472798037059,
+      "grad_norm": 0.16131742298603058,
+      "learning_rate": 0.001,
+      "loss": 2.1966,
+      "step": 14061
+    },
+    {
+      "epoch": 0.5948895845672223,
+      "grad_norm": 0.20708271861076355,
+      "learning_rate": 0.001,
+      "loss": 1.7722,
+      "step": 14062
+    },
+    {
+      "epoch": 0.5949318893307386,
+      "grad_norm": 0.17444756627082825,
+      "learning_rate": 0.001,
+      "loss": 2.1499,
+      "step": 14063
+    },
+    {
+      "epoch": 0.594974194094255,
+      "grad_norm": 0.8531075119972229,
+      "learning_rate": 0.001,
+      "loss": 2.6364,
+      "step": 14064
+    },
+    {
+      "epoch": 0.5950164988577714,
+      "grad_norm": 0.3456467092037201,
+      "learning_rate": 0.001,
+      "loss": 2.8462,
+      "step": 14065
+    },
+    {
+      "epoch": 0.5950588036212877,
+      "grad_norm": 0.2058996707201004,
+      "learning_rate": 0.001,
+      "loss": 1.8034,
+      "step": 14066
+    },
+    {
+      "epoch": 0.5951011083848041,
+      "grad_norm": 0.22642020881175995,
+      "learning_rate": 0.001,
+      "loss": 2.8038,
+      "step": 14067
+    },
+    {
+      "epoch": 0.5951434131483205,
+      "grad_norm": 4.5695719718933105,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 14068
+    },
+    {
+      "epoch": 0.5951857179118368,
+      "grad_norm": 0.23832251131534576,
+      "learning_rate": 0.001,
+      "loss": 2.3427,
+      "step": 14069
+    },
+    {
+      "epoch": 0.5952280226753532,
+      "grad_norm": 0.19538003206253052,
+      "learning_rate": 0.001,
+      "loss": 3.9116,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5952703274388697,
+      "grad_norm": 0.53928542137146,
+      "learning_rate": 0.001,
+      "loss": 1.8973,
+      "step": 14071
+    },
+    {
+      "epoch": 0.595312632202386,
+      "grad_norm": 0.26821160316467285,
+      "learning_rate": 0.001,
+      "loss": 2.5537,
+      "step": 14072
+    },
+    {
+      "epoch": 0.5953549369659024,
+      "grad_norm": 0.20947696268558502,
+      "learning_rate": 0.001,
+      "loss": 2.5992,
+      "step": 14073
+    },
+    {
+      "epoch": 0.5953972417294188,
+      "grad_norm": 0.2498164176940918,
+      "learning_rate": 0.001,
+      "loss": 2.0545,
+      "step": 14074
+    },
+    {
+      "epoch": 0.5954395464929351,
+      "grad_norm": 0.22430068254470825,
+      "learning_rate": 0.001,
+      "loss": 2.6415,
+      "step": 14075
+    },
+    {
+      "epoch": 0.5954818512564515,
+      "grad_norm": 0.39491817355155945,
+      "learning_rate": 0.001,
+      "loss": 2.3454,
+      "step": 14076
+    },
+    {
+      "epoch": 0.5955241560199679,
+      "grad_norm": 0.21011623740196228,
+      "learning_rate": 0.001,
+      "loss": 3.1634,
+      "step": 14077
+    },
+    {
+      "epoch": 0.5955664607834842,
+      "grad_norm": 0.20442980527877808,
+      "learning_rate": 0.001,
+      "loss": 2.2365,
+      "step": 14078
+    },
+    {
+      "epoch": 0.5956087655470006,
+      "grad_norm": 0.18820206820964813,
+      "learning_rate": 0.001,
+      "loss": 2.0868,
+      "step": 14079
+    },
+    {
+      "epoch": 0.595651070310517,
+      "grad_norm": 0.21796266734600067,
+      "learning_rate": 0.001,
+      "loss": 2.4075,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5956933750740333,
+      "grad_norm": 0.16265445947647095,
+      "learning_rate": 0.001,
+      "loss": 2.2618,
+      "step": 14081
+    },
+    {
+      "epoch": 0.5957356798375497,
+      "grad_norm": 0.17943935096263885,
+      "learning_rate": 0.001,
+      "loss": 1.7516,
+      "step": 14082
+    },
+    {
+      "epoch": 0.5957779846010661,
+      "grad_norm": 0.3553915321826935,
+      "learning_rate": 0.001,
+      "loss": 1.9473,
+      "step": 14083
+    },
+    {
+      "epoch": 0.5958202893645824,
+      "grad_norm": 0.1593002825975418,
+      "learning_rate": 0.001,
+      "loss": 1.6759,
+      "step": 14084
+    },
+    {
+      "epoch": 0.5958625941280988,
+      "grad_norm": 0.18188300728797913,
+      "learning_rate": 0.001,
+      "loss": 2.2562,
+      "step": 14085
+    },
+    {
+      "epoch": 0.5959048988916152,
+      "grad_norm": 0.16794385015964508,
+      "learning_rate": 0.001,
+      "loss": 2.2645,
+      "step": 14086
+    },
+    {
+      "epoch": 0.5959472036551315,
+      "grad_norm": 0.16501161456108093,
+      "learning_rate": 0.001,
+      "loss": 2.7828,
+      "step": 14087
+    },
+    {
+      "epoch": 0.595989508418648,
+      "grad_norm": 0.22263823449611664,
+      "learning_rate": 0.001,
+      "loss": 2.0619,
+      "step": 14088
+    },
+    {
+      "epoch": 0.5960318131821644,
+      "grad_norm": 0.4712727963924408,
+      "learning_rate": 0.001,
+      "loss": 3.4105,
+      "step": 14089
+    },
+    {
+      "epoch": 0.5960741179456807,
+      "grad_norm": 0.19112852215766907,
+      "learning_rate": 0.001,
+      "loss": 2.1327,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5961164227091971,
+      "grad_norm": 0.1422937512397766,
+      "learning_rate": 0.001,
+      "loss": 2.0298,
+      "step": 14091
+    },
+    {
+      "epoch": 0.5961587274727135,
+      "grad_norm": 0.14151273667812347,
+      "learning_rate": 0.001,
+      "loss": 1.967,
+      "step": 14092
+    },
+    {
+      "epoch": 0.5962010322362298,
+      "grad_norm": 0.1558687388896942,
+      "learning_rate": 0.001,
+      "loss": 2.4991,
+      "step": 14093
+    },
+    {
+      "epoch": 0.5962433369997462,
+      "grad_norm": 0.18624737858772278,
+      "learning_rate": 0.001,
+      "loss": 3.1451,
+      "step": 14094
+    },
+    {
+      "epoch": 0.5962856417632626,
+      "grad_norm": 0.15436244010925293,
+      "learning_rate": 0.001,
+      "loss": 2.2889,
+      "step": 14095
+    },
+    {
+      "epoch": 0.5963279465267789,
+      "grad_norm": 0.15433412790298462,
+      "learning_rate": 0.001,
+      "loss": 1.7584,
+      "step": 14096
+    },
+    {
+      "epoch": 0.5963702512902953,
+      "grad_norm": 0.24146470427513123,
+      "learning_rate": 0.001,
+      "loss": 3.3554,
+      "step": 14097
+    },
+    {
+      "epoch": 0.5964125560538116,
+      "grad_norm": 0.15554895997047424,
+      "learning_rate": 0.001,
+      "loss": 1.8556,
+      "step": 14098
+    },
+    {
+      "epoch": 0.596454860817328,
+      "grad_norm": 0.17149004340171814,
+      "learning_rate": 0.001,
+      "loss": 2.3022,
+      "step": 14099
+    },
+    {
+      "epoch": 0.5964971655808444,
+      "grad_norm": 0.2116393893957138,
+      "learning_rate": 0.001,
+      "loss": 2.0259,
+      "step": 14100
+    },
+    {
+      "epoch": 0.5965394703443607,
+      "grad_norm": 0.15278328955173492,
+      "learning_rate": 0.001,
+      "loss": 2.0596,
+      "step": 14101
+    },
+    {
+      "epoch": 0.5965817751078771,
+      "grad_norm": 0.14205598831176758,
+      "learning_rate": 0.001,
+      "loss": 2.5957,
+      "step": 14102
+    },
+    {
+      "epoch": 0.5966240798713935,
+      "grad_norm": 0.14854857325553894,
+      "learning_rate": 0.001,
+      "loss": 2.006,
+      "step": 14103
+    },
+    {
+      "epoch": 0.5966663846349098,
+      "grad_norm": 0.5435006022453308,
+      "learning_rate": 0.001,
+      "loss": 2.9184,
+      "step": 14104
+    },
+    {
+      "epoch": 0.5967086893984263,
+      "grad_norm": 0.15318304300308228,
+      "learning_rate": 0.001,
+      "loss": 2.3277,
+      "step": 14105
+    },
+    {
+      "epoch": 0.5967509941619427,
+      "grad_norm": 0.3546973168849945,
+      "learning_rate": 0.001,
+      "loss": 1.9591,
+      "step": 14106
+    },
+    {
+      "epoch": 0.596793298925459,
+      "grad_norm": 0.15295492112636566,
+      "learning_rate": 0.001,
+      "loss": 2.7593,
+      "step": 14107
+    },
+    {
+      "epoch": 0.5968356036889754,
+      "grad_norm": 0.13746538758277893,
+      "learning_rate": 0.001,
+      "loss": 2.6047,
+      "step": 14108
+    },
+    {
+      "epoch": 0.5968779084524918,
+      "grad_norm": 0.18742793798446655,
+      "learning_rate": 0.001,
+      "loss": 1.4895,
+      "step": 14109
+    },
+    {
+      "epoch": 0.5969202132160081,
+      "grad_norm": 0.28040406107902527,
+      "learning_rate": 0.001,
+      "loss": 3.4037,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5969625179795245,
+      "grad_norm": 0.15165936946868896,
+      "learning_rate": 0.001,
+      "loss": 2.2845,
+      "step": 14111
+    },
+    {
+      "epoch": 0.5970048227430409,
+      "grad_norm": 0.1565185934305191,
+      "learning_rate": 0.001,
+      "loss": 2.1824,
+      "step": 14112
+    },
+    {
+      "epoch": 0.5970471275065572,
+      "grad_norm": 0.16770948469638824,
+      "learning_rate": 0.001,
+      "loss": 2.4891,
+      "step": 14113
+    },
+    {
+      "epoch": 0.5970894322700736,
+      "grad_norm": 0.14534291625022888,
+      "learning_rate": 0.001,
+      "loss": 2.0148,
+      "step": 14114
+    },
+    {
+      "epoch": 0.59713173703359,
+      "grad_norm": 0.157785564661026,
+      "learning_rate": 0.001,
+      "loss": 2.0671,
+      "step": 14115
+    },
+    {
+      "epoch": 0.5971740417971063,
+      "grad_norm": 0.15590552985668182,
+      "learning_rate": 0.001,
+      "loss": 3.3419,
+      "step": 14116
+    },
+    {
+      "epoch": 0.5972163465606227,
+      "grad_norm": 0.24124379456043243,
+      "learning_rate": 0.001,
+      "loss": 3.057,
+      "step": 14117
+    },
+    {
+      "epoch": 0.5972586513241391,
+      "grad_norm": 4.952909469604492,
+      "learning_rate": 0.001,
+      "loss": 2.278,
+      "step": 14118
+    },
+    {
+      "epoch": 0.5973009560876554,
+      "grad_norm": 0.3283090889453888,
+      "learning_rate": 0.001,
+      "loss": 2.4805,
+      "step": 14119
+    },
+    {
+      "epoch": 0.5973432608511718,
+      "grad_norm": 0.14943337440490723,
+      "learning_rate": 0.001,
+      "loss": 1.8459,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5973855656146883,
+      "grad_norm": 0.1828925907611847,
+      "learning_rate": 0.001,
+      "loss": 2.2262,
+      "step": 14121
+    },
+    {
+      "epoch": 0.5974278703782046,
+      "grad_norm": 0.21698467433452606,
+      "learning_rate": 0.001,
+      "loss": 1.5742,
+      "step": 14122
+    },
+    {
+      "epoch": 0.597470175141721,
+      "grad_norm": 0.17111243307590485,
+      "learning_rate": 0.001,
+      "loss": 1.6804,
+      "step": 14123
+    },
+    {
+      "epoch": 0.5975124799052374,
+      "grad_norm": 0.1583365499973297,
+      "learning_rate": 0.001,
+      "loss": 1.6393,
+      "step": 14124
+    },
+    {
+      "epoch": 0.5975547846687537,
+      "grad_norm": 0.18235273659229279,
+      "learning_rate": 0.001,
+      "loss": 2.8478,
+      "step": 14125
+    },
+    {
+      "epoch": 0.5975970894322701,
+      "grad_norm": 0.16986776888370514,
+      "learning_rate": 0.001,
+      "loss": 1.6388,
+      "step": 14126
+    },
+    {
+      "epoch": 0.5976393941957865,
+      "grad_norm": 0.39957574009895325,
+      "learning_rate": 0.001,
+      "loss": 2.1505,
+      "step": 14127
+    },
+    {
+      "epoch": 0.5976816989593028,
+      "grad_norm": 0.15922756493091583,
+      "learning_rate": 0.001,
+      "loss": 1.7598,
+      "step": 14128
+    },
+    {
+      "epoch": 0.5977240037228192,
+      "grad_norm": 6.040116786956787,
+      "learning_rate": 0.001,
+      "loss": 3.9252,
+      "step": 14129
+    },
+    {
+      "epoch": 0.5977663084863356,
+      "grad_norm": 0.2572100758552551,
+      "learning_rate": 0.001,
+      "loss": 2.1899,
+      "step": 14130
+    },
+    {
+      "epoch": 0.5978086132498519,
+      "grad_norm": 7.241512298583984,
+      "learning_rate": 0.001,
+      "loss": 2.238,
+      "step": 14131
+    },
+    {
+      "epoch": 0.5978509180133683,
+      "grad_norm": 0.21074838936328888,
+      "learning_rate": 0.001,
+      "loss": 1.8177,
+      "step": 14132
+    },
+    {
+      "epoch": 0.5978932227768847,
+      "grad_norm": 0.20617999136447906,
+      "learning_rate": 0.001,
+      "loss": 2.7263,
+      "step": 14133
+    },
+    {
+      "epoch": 0.597935527540401,
+      "grad_norm": 8.84471321105957,
+      "learning_rate": 0.001,
+      "loss": 2.3485,
+      "step": 14134
+    },
+    {
+      "epoch": 0.5979778323039174,
+      "grad_norm": 0.21541795134544373,
+      "learning_rate": 0.001,
+      "loss": 2.4785,
+      "step": 14135
+    },
+    {
+      "epoch": 0.5980201370674338,
+      "grad_norm": 0.17976322770118713,
+      "learning_rate": 0.001,
+      "loss": 1.9674,
+      "step": 14136
+    },
+    {
+      "epoch": 0.5980624418309501,
+      "grad_norm": 0.18282859027385712,
+      "learning_rate": 0.001,
+      "loss": 2.8773,
+      "step": 14137
+    },
+    {
+      "epoch": 0.5981047465944666,
+      "grad_norm": 0.29796525835990906,
+      "learning_rate": 0.001,
+      "loss": 2.8154,
+      "step": 14138
+    },
+    {
+      "epoch": 0.598147051357983,
+      "grad_norm": 0.19228790700435638,
+      "learning_rate": 0.001,
+      "loss": 2.666,
+      "step": 14139
+    },
+    {
+      "epoch": 0.5981893561214993,
+      "grad_norm": 0.3605715334415436,
+      "learning_rate": 0.001,
+      "loss": 2.8384,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5982316608850157,
+      "grad_norm": 0.28930771350860596,
+      "learning_rate": 0.001,
+      "loss": 1.6602,
+      "step": 14141
+    },
+    {
+      "epoch": 0.5982739656485321,
+      "grad_norm": 0.21113309264183044,
+      "learning_rate": 0.001,
+      "loss": 1.586,
+      "step": 14142
+    },
+    {
+      "epoch": 0.5983162704120484,
+      "grad_norm": 0.16728419065475464,
+      "learning_rate": 0.001,
+      "loss": 3.1558,
+      "step": 14143
+    },
+    {
+      "epoch": 0.5983585751755648,
+      "grad_norm": 0.19090363383293152,
+      "learning_rate": 0.001,
+      "loss": 2.0699,
+      "step": 14144
+    },
+    {
+      "epoch": 0.5984008799390811,
+      "grad_norm": 0.23593130707740784,
+      "learning_rate": 0.001,
+      "loss": 2.9676,
+      "step": 14145
+    },
+    {
+      "epoch": 0.5984431847025975,
+      "grad_norm": 0.1566046178340912,
+      "learning_rate": 0.001,
+      "loss": 1.873,
+      "step": 14146
+    },
+    {
+      "epoch": 0.5984854894661139,
+      "grad_norm": 0.14671561121940613,
+      "learning_rate": 0.001,
+      "loss": 1.6314,
+      "step": 14147
+    },
+    {
+      "epoch": 0.5985277942296302,
+      "grad_norm": 0.17109639942646027,
+      "learning_rate": 0.001,
+      "loss": 1.7268,
+      "step": 14148
+    },
+    {
+      "epoch": 0.5985700989931466,
+      "grad_norm": 0.13763944804668427,
+      "learning_rate": 0.001,
+      "loss": 2.2571,
+      "step": 14149
+    },
+    {
+      "epoch": 0.598612403756663,
+      "grad_norm": 0.14401958882808685,
+      "learning_rate": 0.001,
+      "loss": 1.5655,
+      "step": 14150
+    },
+    {
+      "epoch": 0.5986547085201793,
+      "grad_norm": 0.21010830998420715,
+      "learning_rate": 0.001,
+      "loss": 1.9039,
+      "step": 14151
+    },
+    {
+      "epoch": 0.5986970132836957,
+      "grad_norm": 1.928266167640686,
+      "learning_rate": 0.001,
+      "loss": 2.4406,
+      "step": 14152
+    },
+    {
+      "epoch": 0.5987393180472121,
+      "grad_norm": 0.1703515201807022,
+      "learning_rate": 0.001,
+      "loss": 2.6037,
+      "step": 14153
+    },
+    {
+      "epoch": 0.5987816228107284,
+      "grad_norm": 0.2860715687274933,
+      "learning_rate": 0.001,
+      "loss": 2.0231,
+      "step": 14154
+    },
+    {
+      "epoch": 0.5988239275742449,
+      "grad_norm": 0.19897747039794922,
+      "learning_rate": 0.001,
+      "loss": 2.8199,
+      "step": 14155
+    },
+    {
+      "epoch": 0.5988662323377613,
+      "grad_norm": 0.18251550197601318,
+      "learning_rate": 0.001,
+      "loss": 1.7923,
+      "step": 14156
+    },
+    {
+      "epoch": 0.5989085371012776,
+      "grad_norm": 1.5346297025680542,
+      "learning_rate": 0.001,
+      "loss": 1.7232,
+      "step": 14157
+    },
+    {
+      "epoch": 0.598950841864794,
+      "grad_norm": 0.16617514193058014,
+      "learning_rate": 0.001,
+      "loss": 1.4424,
+      "step": 14158
+    },
+    {
+      "epoch": 0.5989931466283104,
+      "grad_norm": 0.14995898306369781,
+      "learning_rate": 0.001,
+      "loss": 1.8629,
+      "step": 14159
+    },
+    {
+      "epoch": 0.5990354513918267,
+      "grad_norm": 0.1728651225566864,
+      "learning_rate": 0.001,
+      "loss": 2.0475,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5990777561553431,
+      "grad_norm": 0.27386367321014404,
+      "learning_rate": 0.001,
+      "loss": 1.6863,
+      "step": 14161
+    },
+    {
+      "epoch": 0.5991200609188595,
+      "grad_norm": 0.21479547023773193,
+      "learning_rate": 0.001,
+      "loss": 1.8521,
+      "step": 14162
+    },
+    {
+      "epoch": 0.5991623656823758,
+      "grad_norm": 0.2000337839126587,
+      "learning_rate": 0.001,
+      "loss": 2.8076,
+      "step": 14163
+    },
+    {
+      "epoch": 0.5992046704458922,
+      "grad_norm": 0.16316631436347961,
+      "learning_rate": 0.001,
+      "loss": 3.3434,
+      "step": 14164
+    },
+    {
+      "epoch": 0.5992469752094086,
+      "grad_norm": 0.18021291494369507,
+      "learning_rate": 0.001,
+      "loss": 2.0387,
+      "step": 14165
+    },
+    {
+      "epoch": 0.5992892799729249,
+      "grad_norm": 0.24870742857456207,
+      "learning_rate": 0.001,
+      "loss": 2.6437,
+      "step": 14166
+    },
+    {
+      "epoch": 0.5993315847364413,
+      "grad_norm": 0.17932534217834473,
+      "learning_rate": 0.001,
+      "loss": 2.412,
+      "step": 14167
+    },
+    {
+      "epoch": 0.5993738894999577,
+      "grad_norm": 0.22996637225151062,
+      "learning_rate": 0.001,
+      "loss": 3.117,
+      "step": 14168
+    },
+    {
+      "epoch": 0.599416194263474,
+      "grad_norm": 13.875000953674316,
+      "learning_rate": 0.001,
+      "loss": 2.2331,
+      "step": 14169
+    },
+    {
+      "epoch": 0.5994584990269904,
+      "grad_norm": 0.8741011619567871,
+      "learning_rate": 0.001,
+      "loss": 2.2017,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5995008037905069,
+      "grad_norm": 0.3972923159599304,
+      "learning_rate": 0.001,
+      "loss": 2.4348,
+      "step": 14171
+    },
+    {
+      "epoch": 0.5995431085540232,
+      "grad_norm": 0.18050576746463776,
+      "learning_rate": 0.001,
+      "loss": 2.2719,
+      "step": 14172
+    },
+    {
+      "epoch": 0.5995854133175396,
+      "grad_norm": 0.6231819987297058,
+      "learning_rate": 0.001,
+      "loss": 1.9272,
+      "step": 14173
+    },
+    {
+      "epoch": 0.599627718081056,
+      "grad_norm": 0.2706356942653656,
+      "learning_rate": 0.001,
+      "loss": 1.8771,
+      "step": 14174
+    },
+    {
+      "epoch": 0.5996700228445723,
+      "grad_norm": 0.2066756784915924,
+      "learning_rate": 0.001,
+      "loss": 1.9065,
+      "step": 14175
+    },
+    {
+      "epoch": 0.5997123276080887,
+      "grad_norm": 0.2722444236278534,
+      "learning_rate": 0.001,
+      "loss": 2.5702,
+      "step": 14176
+    },
+    {
+      "epoch": 0.5997546323716051,
+      "grad_norm": 0.18113920092582703,
+      "learning_rate": 0.001,
+      "loss": 1.8397,
+      "step": 14177
+    },
+    {
+      "epoch": 0.5997969371351214,
+      "grad_norm": 0.2111581563949585,
+      "learning_rate": 0.001,
+      "loss": 2.0402,
+      "step": 14178
+    },
+    {
+      "epoch": 0.5998392418986378,
+      "grad_norm": 0.1754538118839264,
+      "learning_rate": 0.001,
+      "loss": 1.4719,
+      "step": 14179
+    },
+    {
+      "epoch": 0.5998815466621542,
+      "grad_norm": 0.15258760750293732,
+      "learning_rate": 0.001,
+      "loss": 2.0344,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5999238514256705,
+      "grad_norm": 0.18332518637180328,
+      "learning_rate": 0.001,
+      "loss": 2.2107,
+      "step": 14181
+    },
+    {
+      "epoch": 0.5999661561891869,
+      "grad_norm": 0.23446984589099884,
+      "learning_rate": 0.001,
+      "loss": 2.8876,
+      "step": 14182
+    },
+    {
+      "epoch": 0.6000084609527033,
+      "grad_norm": 0.20823650062084198,
+      "learning_rate": 0.001,
+      "loss": 2.5374,
+      "step": 14183
+    },
+    {
+      "epoch": 0.6000507657162196,
+      "grad_norm": 0.16053082048892975,
+      "learning_rate": 0.001,
+      "loss": 2.1142,
+      "step": 14184
+    },
+    {
+      "epoch": 0.600093070479736,
+      "grad_norm": 0.2412220984697342,
+      "learning_rate": 0.001,
+      "loss": 3.0947,
+      "step": 14185
+    },
+    {
+      "epoch": 0.6001353752432524,
+      "grad_norm": 0.17646367847919464,
+      "learning_rate": 0.001,
+      "loss": 2.4277,
+      "step": 14186
+    },
+    {
+      "epoch": 0.6001776800067687,
+      "grad_norm": 0.16096143424510956,
+      "learning_rate": 0.001,
+      "loss": 2.2449,
+      "step": 14187
+    },
+    {
+      "epoch": 0.6002199847702852,
+      "grad_norm": 0.2696020007133484,
+      "learning_rate": 0.001,
+      "loss": 1.9888,
+      "step": 14188
+    },
+    {
+      "epoch": 0.6002622895338015,
+      "grad_norm": 0.14469857513904572,
+      "learning_rate": 0.001,
+      "loss": 2.14,
+      "step": 14189
+    },
+    {
+      "epoch": 0.6003045942973179,
+      "grad_norm": 0.17882224917411804,
+      "learning_rate": 0.001,
+      "loss": 2.0161,
+      "step": 14190
+    },
+    {
+      "epoch": 0.6003468990608343,
+      "grad_norm": 0.1592990756034851,
+      "learning_rate": 0.001,
+      "loss": 2.067,
+      "step": 14191
+    },
+    {
+      "epoch": 0.6003892038243506,
+      "grad_norm": 0.8799763321876526,
+      "learning_rate": 0.001,
+      "loss": 2.3578,
+      "step": 14192
+    },
+    {
+      "epoch": 0.600431508587867,
+      "grad_norm": 0.1751909852027893,
+      "learning_rate": 0.001,
+      "loss": 1.8269,
+      "step": 14193
+    },
+    {
+      "epoch": 0.6004738133513834,
+      "grad_norm": 0.22344686090946198,
+      "learning_rate": 0.001,
+      "loss": 2.8203,
+      "step": 14194
+    },
+    {
+      "epoch": 0.6005161181148997,
+      "grad_norm": 0.19454553723335266,
+      "learning_rate": 0.001,
+      "loss": 2.5996,
+      "step": 14195
+    },
+    {
+      "epoch": 0.6005584228784161,
+      "grad_norm": 0.16790246963500977,
+      "learning_rate": 0.001,
+      "loss": 2.8472,
+      "step": 14196
+    },
+    {
+      "epoch": 0.6006007276419325,
+      "grad_norm": 0.16107873618602753,
+      "learning_rate": 0.001,
+      "loss": 1.7427,
+      "step": 14197
+    },
+    {
+      "epoch": 0.6006430324054488,
+      "grad_norm": 22.60145378112793,
+      "learning_rate": 0.001,
+      "loss": 1.6772,
+      "step": 14198
+    },
+    {
+      "epoch": 0.6006853371689652,
+      "grad_norm": 0.16331073641777039,
+      "learning_rate": 0.001,
+      "loss": 2.0363,
+      "step": 14199
+    },
+    {
+      "epoch": 0.6007276419324816,
+      "grad_norm": 1.7857904434204102,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 14200
+    },
+    {
+      "epoch": 0.6007699466959979,
+      "grad_norm": 0.1780909299850464,
+      "learning_rate": 0.001,
+      "loss": 2.9103,
+      "step": 14201
+    },
+    {
+      "epoch": 0.6008122514595143,
+      "grad_norm": 0.3058990240097046,
+      "learning_rate": 0.001,
+      "loss": 2.3492,
+      "step": 14202
+    },
+    {
+      "epoch": 0.6008545562230307,
+      "grad_norm": 0.1861615926027298,
+      "learning_rate": 0.001,
+      "loss": 2.2742,
+      "step": 14203
+    },
+    {
+      "epoch": 0.600896860986547,
+      "grad_norm": 0.21202732622623444,
+      "learning_rate": 0.001,
+      "loss": 1.6728,
+      "step": 14204
+    },
+    {
+      "epoch": 0.6009391657500635,
+      "grad_norm": 0.17166270315647125,
+      "learning_rate": 0.001,
+      "loss": 2.293,
+      "step": 14205
+    },
+    {
+      "epoch": 0.6009814705135799,
+      "grad_norm": 0.1747630089521408,
+      "learning_rate": 0.001,
+      "loss": 3.0309,
+      "step": 14206
+    },
+    {
+      "epoch": 0.6010237752770962,
+      "grad_norm": 0.1900450885295868,
+      "learning_rate": 0.001,
+      "loss": 1.6363,
+      "step": 14207
+    },
+    {
+      "epoch": 0.6010660800406126,
+      "grad_norm": 6.871549129486084,
+      "learning_rate": 0.001,
+      "loss": 1.5597,
+      "step": 14208
+    },
+    {
+      "epoch": 0.601108384804129,
+      "grad_norm": 0.20221640169620514,
+      "learning_rate": 0.001,
+      "loss": 1.6391,
+      "step": 14209
+    },
+    {
+      "epoch": 0.6011506895676453,
+      "grad_norm": 0.5754346251487732,
+      "learning_rate": 0.001,
+      "loss": 4.2668,
+      "step": 14210
+    },
+    {
+      "epoch": 0.6011929943311617,
+      "grad_norm": 0.2078094184398651,
+      "learning_rate": 0.001,
+      "loss": 1.8344,
+      "step": 14211
+    },
+    {
+      "epoch": 0.6012352990946781,
+      "grad_norm": 0.19465067982673645,
+      "learning_rate": 0.001,
+      "loss": 2.707,
+      "step": 14212
+    },
+    {
+      "epoch": 0.6012776038581944,
+      "grad_norm": 0.1583552062511444,
+      "learning_rate": 0.001,
+      "loss": 1.791,
+      "step": 14213
+    },
+    {
+      "epoch": 0.6013199086217108,
+      "grad_norm": 0.6577909588813782,
+      "learning_rate": 0.001,
+      "loss": 1.8294,
+      "step": 14214
+    },
+    {
+      "epoch": 0.6013622133852272,
+      "grad_norm": 0.16228260099887848,
+      "learning_rate": 0.001,
+      "loss": 2.0963,
+      "step": 14215
+    },
+    {
+      "epoch": 0.6014045181487435,
+      "grad_norm": 0.21815434098243713,
+      "learning_rate": 0.001,
+      "loss": 1.9625,
+      "step": 14216
+    },
+    {
+      "epoch": 0.6014468229122599,
+      "grad_norm": 0.168915793299675,
+      "learning_rate": 0.001,
+      "loss": 1.7163,
+      "step": 14217
+    },
+    {
+      "epoch": 0.6014891276757763,
+      "grad_norm": 0.21465511620044708,
+      "learning_rate": 0.001,
+      "loss": 2.9348,
+      "step": 14218
+    },
+    {
+      "epoch": 0.6015314324392926,
+      "grad_norm": 1.094494104385376,
+      "learning_rate": 0.001,
+      "loss": 2.2373,
+      "step": 14219
+    },
+    {
+      "epoch": 0.601573737202809,
+      "grad_norm": 0.21323548257350922,
+      "learning_rate": 0.001,
+      "loss": 2.9424,
+      "step": 14220
+    },
+    {
+      "epoch": 0.6016160419663255,
+      "grad_norm": 4.479442596435547,
+      "learning_rate": 0.001,
+      "loss": 1.9656,
+      "step": 14221
+    },
+    {
+      "epoch": 0.6016583467298418,
+      "grad_norm": 0.13365942239761353,
+      "learning_rate": 0.001,
+      "loss": 1.264,
+      "step": 14222
+    },
+    {
+      "epoch": 0.6017006514933582,
+      "grad_norm": 1.7394766807556152,
+      "learning_rate": 0.001,
+      "loss": 2.1951,
+      "step": 14223
+    },
+    {
+      "epoch": 0.6017429562568746,
+      "grad_norm": 3.335594415664673,
+      "learning_rate": 0.001,
+      "loss": 2.4577,
+      "step": 14224
+    },
+    {
+      "epoch": 0.6017852610203909,
+      "grad_norm": 0.574901282787323,
+      "learning_rate": 0.001,
+      "loss": 2.5492,
+      "step": 14225
+    },
+    {
+      "epoch": 0.6018275657839073,
+      "grad_norm": 0.16462647914886475,
+      "learning_rate": 0.001,
+      "loss": 2.7077,
+      "step": 14226
+    },
+    {
+      "epoch": 0.6018698705474237,
+      "grad_norm": 0.36625564098358154,
+      "learning_rate": 0.001,
+      "loss": 2.6481,
+      "step": 14227
+    },
+    {
+      "epoch": 0.60191217531094,
+      "grad_norm": 0.33818289637565613,
+      "learning_rate": 0.001,
+      "loss": 2.2215,
+      "step": 14228
+    },
+    {
+      "epoch": 0.6019544800744564,
+      "grad_norm": 0.18577156960964203,
+      "learning_rate": 0.001,
+      "loss": 2.0114,
+      "step": 14229
+    },
+    {
+      "epoch": 0.6019967848379728,
+      "grad_norm": 0.14699004590511322,
+      "learning_rate": 0.001,
+      "loss": 1.6804,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6020390896014891,
+      "grad_norm": 0.17045558989048004,
+      "learning_rate": 0.001,
+      "loss": 2.3435,
+      "step": 14231
+    },
+    {
+      "epoch": 0.6020813943650055,
+      "grad_norm": 0.1529216319322586,
+      "learning_rate": 0.001,
+      "loss": 2.1237,
+      "step": 14232
+    },
+    {
+      "epoch": 0.6021236991285218,
+      "grad_norm": 0.1529701054096222,
+      "learning_rate": 0.001,
+      "loss": 3.5144,
+      "step": 14233
+    },
+    {
+      "epoch": 0.6021660038920382,
+      "grad_norm": 0.18276235461235046,
+      "learning_rate": 0.001,
+      "loss": 2.1834,
+      "step": 14234
+    },
+    {
+      "epoch": 0.6022083086555546,
+      "grad_norm": 0.2533540725708008,
+      "learning_rate": 0.001,
+      "loss": 3.0911,
+      "step": 14235
+    },
+    {
+      "epoch": 0.6022506134190709,
+      "grad_norm": 0.1999509334564209,
+      "learning_rate": 0.001,
+      "loss": 1.7524,
+      "step": 14236
+    },
+    {
+      "epoch": 0.6022929181825873,
+      "grad_norm": 0.1922556310892105,
+      "learning_rate": 0.001,
+      "loss": 2.1238,
+      "step": 14237
+    },
+    {
+      "epoch": 0.6023352229461038,
+      "grad_norm": 1.1301320791244507,
+      "learning_rate": 0.001,
+      "loss": 1.7954,
+      "step": 14238
+    },
+    {
+      "epoch": 0.60237752770962,
+      "grad_norm": 0.14961791038513184,
+      "learning_rate": 0.001,
+      "loss": 1.6322,
+      "step": 14239
+    },
+    {
+      "epoch": 0.6024198324731365,
+      "grad_norm": 0.5575346350669861,
+      "learning_rate": 0.001,
+      "loss": 2.2569,
+      "step": 14240
+    },
+    {
+      "epoch": 0.6024621372366529,
+      "grad_norm": 0.19425736367702484,
+      "learning_rate": 0.001,
+      "loss": 1.8248,
+      "step": 14241
+    },
+    {
+      "epoch": 0.6025044420001692,
+      "grad_norm": 0.1504928469657898,
+      "learning_rate": 0.001,
+      "loss": 2.2506,
+      "step": 14242
+    },
+    {
+      "epoch": 0.6025467467636856,
+      "grad_norm": 1.183660626411438,
+      "learning_rate": 0.001,
+      "loss": 1.9673,
+      "step": 14243
+    },
+    {
+      "epoch": 0.602589051527202,
+      "grad_norm": 0.3906610608100891,
+      "learning_rate": 0.001,
+      "loss": 2.8927,
+      "step": 14244
+    },
+    {
+      "epoch": 0.6026313562907183,
+      "grad_norm": 0.20628994703292847,
+      "learning_rate": 0.001,
+      "loss": 2.1268,
+      "step": 14245
+    },
+    {
+      "epoch": 0.6026736610542347,
+      "grad_norm": 4.261792182922363,
+      "learning_rate": 0.001,
+      "loss": 4.2053,
+      "step": 14246
+    },
+    {
+      "epoch": 0.6027159658177511,
+      "grad_norm": 0.15113836526870728,
+      "learning_rate": 0.001,
+      "loss": 1.5448,
+      "step": 14247
+    },
+    {
+      "epoch": 0.6027582705812674,
+      "grad_norm": 0.20235134661197662,
+      "learning_rate": 0.001,
+      "loss": 2.3315,
+      "step": 14248
+    },
+    {
+      "epoch": 0.6028005753447838,
+      "grad_norm": 0.15247638523578644,
+      "learning_rate": 0.001,
+      "loss": 2.15,
+      "step": 14249
+    },
+    {
+      "epoch": 0.6028428801083002,
+      "grad_norm": 0.18794965744018555,
+      "learning_rate": 0.001,
+      "loss": 2.3342,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6028851848718165,
+      "grad_norm": 4.161924839019775,
+      "learning_rate": 0.001,
+      "loss": 1.9379,
+      "step": 14251
+    },
+    {
+      "epoch": 0.6029274896353329,
+      "grad_norm": 0.1974867433309555,
+      "learning_rate": 0.001,
+      "loss": 2.1114,
+      "step": 14252
+    },
+    {
+      "epoch": 0.6029697943988493,
+      "grad_norm": 0.15145453810691833,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 14253
+    },
+    {
+      "epoch": 0.6030120991623656,
+      "grad_norm": 0.1826041340827942,
+      "learning_rate": 0.001,
+      "loss": 1.8654,
+      "step": 14254
+    },
+    {
+      "epoch": 0.603054403925882,
+      "grad_norm": 0.33202704787254333,
+      "learning_rate": 0.001,
+      "loss": 2.8525,
+      "step": 14255
+    },
+    {
+      "epoch": 0.6030967086893985,
+      "grad_norm": 0.17979615926742554,
+      "learning_rate": 0.001,
+      "loss": 2.1171,
+      "step": 14256
+    },
+    {
+      "epoch": 0.6031390134529148,
+      "grad_norm": 0.33477500081062317,
+      "learning_rate": 0.001,
+      "loss": 2.4972,
+      "step": 14257
+    },
+    {
+      "epoch": 0.6031813182164312,
+      "grad_norm": 0.1628679484128952,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 14258
+    },
+    {
+      "epoch": 0.6032236229799476,
+      "grad_norm": 0.15380176901817322,
+      "learning_rate": 0.001,
+      "loss": 1.6094,
+      "step": 14259
+    },
+    {
+      "epoch": 0.6032659277434639,
+      "grad_norm": 0.15169599652290344,
+      "learning_rate": 0.001,
+      "loss": 1.3153,
+      "step": 14260
+    },
+    {
+      "epoch": 0.6033082325069803,
+      "grad_norm": 1.1661324501037598,
+      "learning_rate": 0.001,
+      "loss": 1.8356,
+      "step": 14261
+    },
+    {
+      "epoch": 0.6033505372704967,
+      "grad_norm": 0.14348934590816498,
+      "learning_rate": 0.001,
+      "loss": 1.4043,
+      "step": 14262
+    },
+    {
+      "epoch": 0.603392842034013,
+      "grad_norm": 0.35176733136177063,
+      "learning_rate": 0.001,
+      "loss": 2.327,
+      "step": 14263
+    },
+    {
+      "epoch": 0.6034351467975294,
+      "grad_norm": 2.9565463066101074,
+      "learning_rate": 0.001,
+      "loss": 2.2605,
+      "step": 14264
+    },
+    {
+      "epoch": 0.6034774515610458,
+      "grad_norm": 0.20699867606163025,
+      "learning_rate": 0.001,
+      "loss": 2.5949,
+      "step": 14265
+    },
+    {
+      "epoch": 0.6035197563245621,
+      "grad_norm": 0.6101794242858887,
+      "learning_rate": 0.001,
+      "loss": 2.4005,
+      "step": 14266
+    },
+    {
+      "epoch": 0.6035620610880785,
+      "grad_norm": 0.2099284827709198,
+      "learning_rate": 0.001,
+      "loss": 1.7885,
+      "step": 14267
+    },
+    {
+      "epoch": 0.6036043658515949,
+      "grad_norm": 0.20473699271678925,
+      "learning_rate": 0.001,
+      "loss": 2.0055,
+      "step": 14268
+    },
+    {
+      "epoch": 0.6036466706151112,
+      "grad_norm": 0.2228740006685257,
+      "learning_rate": 0.001,
+      "loss": 2.4903,
+      "step": 14269
+    },
+    {
+      "epoch": 0.6036889753786276,
+      "grad_norm": 0.6952749490737915,
+      "learning_rate": 0.001,
+      "loss": 2.8113,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6037312801421441,
+      "grad_norm": 0.16827528178691864,
+      "learning_rate": 0.001,
+      "loss": 1.7197,
+      "step": 14271
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.23891712725162506,
+      "learning_rate": 0.001,
+      "loss": 2.2606,
+      "step": 14272
+    },
+    {
+      "epoch": 0.6038158896691768,
+      "grad_norm": 0.220239520072937,
+      "learning_rate": 0.001,
+      "loss": 1.4048,
+      "step": 14273
+    },
+    {
+      "epoch": 0.6038581944326932,
+      "grad_norm": 0.16221195459365845,
+      "learning_rate": 0.001,
+      "loss": 1.7971,
+      "step": 14274
+    },
+    {
+      "epoch": 0.6039004991962095,
+      "grad_norm": 0.31431055068969727,
+      "learning_rate": 0.001,
+      "loss": 3.3171,
+      "step": 14275
+    },
+    {
+      "epoch": 0.6039428039597259,
+      "grad_norm": 1.185067057609558,
+      "learning_rate": 0.001,
+      "loss": 2.7166,
+      "step": 14276
+    },
+    {
+      "epoch": 0.6039851087232423,
+      "grad_norm": 0.15773341059684753,
+      "learning_rate": 0.001,
+      "loss": 2.0959,
+      "step": 14277
+    },
+    {
+      "epoch": 0.6040274134867586,
+      "grad_norm": 0.16446535289287567,
+      "learning_rate": 0.001,
+      "loss": 1.6047,
+      "step": 14278
+    },
+    {
+      "epoch": 0.604069718250275,
+      "grad_norm": 0.20500445365905762,
+      "learning_rate": 0.001,
+      "loss": 2.4833,
+      "step": 14279
+    },
+    {
+      "epoch": 0.6041120230137913,
+      "grad_norm": 0.1654626429080963,
+      "learning_rate": 0.001,
+      "loss": 1.7549,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6041543277773077,
+      "grad_norm": 0.16922923922538757,
+      "learning_rate": 0.001,
+      "loss": 2.0515,
+      "step": 14281
+    },
+    {
+      "epoch": 0.6041966325408241,
+      "grad_norm": 0.17011629045009613,
+      "learning_rate": 0.001,
+      "loss": 2.5793,
+      "step": 14282
+    },
+    {
+      "epoch": 0.6042389373043404,
+      "grad_norm": 0.25761881470680237,
+      "learning_rate": 0.001,
+      "loss": 1.9069,
+      "step": 14283
+    },
+    {
+      "epoch": 0.6042812420678568,
+      "grad_norm": 0.18269698321819305,
+      "learning_rate": 0.001,
+      "loss": 3.0221,
+      "step": 14284
+    },
+    {
+      "epoch": 0.6043235468313732,
+      "grad_norm": 85.4771728515625,
+      "learning_rate": 0.001,
+      "loss": 2.3729,
+      "step": 14285
+    },
+    {
+      "epoch": 0.6043658515948895,
+      "grad_norm": 0.7515902519226074,
+      "learning_rate": 0.001,
+      "loss": 2.3315,
+      "step": 14286
+    },
+    {
+      "epoch": 0.604408156358406,
+      "grad_norm": 0.2449134737253189,
+      "learning_rate": 0.001,
+      "loss": 1.72,
+      "step": 14287
+    },
+    {
+      "epoch": 0.6044504611219224,
+      "grad_norm": 0.15557745099067688,
+      "learning_rate": 0.001,
+      "loss": 1.9485,
+      "step": 14288
+    },
+    {
+      "epoch": 0.6044927658854387,
+      "grad_norm": 0.18973484635353088,
+      "learning_rate": 0.001,
+      "loss": 2.1576,
+      "step": 14289
+    },
+    {
+      "epoch": 0.6045350706489551,
+      "grad_norm": 0.2871471643447876,
+      "learning_rate": 0.001,
+      "loss": 1.7434,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6045773754124715,
+      "grad_norm": 0.4421120882034302,
+      "learning_rate": 0.001,
+      "loss": 3.4448,
+      "step": 14291
+    },
+    {
+      "epoch": 0.6046196801759878,
+      "grad_norm": 0.15245676040649414,
+      "learning_rate": 0.001,
+      "loss": 1.6375,
+      "step": 14292
+    },
+    {
+      "epoch": 0.6046619849395042,
+      "grad_norm": 0.22273124754428864,
+      "learning_rate": 0.001,
+      "loss": 3.4129,
+      "step": 14293
+    },
+    {
+      "epoch": 0.6047042897030206,
+      "grad_norm": 0.1912602186203003,
+      "learning_rate": 0.001,
+      "loss": 2.368,
+      "step": 14294
+    },
+    {
+      "epoch": 0.6047465944665369,
+      "grad_norm": 0.18390634655952454,
+      "learning_rate": 0.001,
+      "loss": 2.33,
+      "step": 14295
+    },
+    {
+      "epoch": 0.6047888992300533,
+      "grad_norm": 0.2985355854034424,
+      "learning_rate": 0.001,
+      "loss": 2.6485,
+      "step": 14296
+    },
+    {
+      "epoch": 0.6048312039935697,
+      "grad_norm": 0.1954064518213272,
+      "learning_rate": 0.001,
+      "loss": 2.0261,
+      "step": 14297
+    },
+    {
+      "epoch": 0.604873508757086,
+      "grad_norm": 0.16802330315113068,
+      "learning_rate": 0.001,
+      "loss": 2.1737,
+      "step": 14298
+    },
+    {
+      "epoch": 0.6049158135206024,
+      "grad_norm": 0.1490180343389511,
+      "learning_rate": 0.001,
+      "loss": 1.6823,
+      "step": 14299
+    },
+    {
+      "epoch": 0.6049581182841188,
+      "grad_norm": 0.1882055103778839,
+      "learning_rate": 0.001,
+      "loss": 2.7867,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6050004230476351,
+      "grad_norm": 0.18217292428016663,
+      "learning_rate": 0.001,
+      "loss": 2.2182,
+      "step": 14301
+    },
+    {
+      "epoch": 0.6050427278111515,
+      "grad_norm": 0.16061586141586304,
+      "learning_rate": 0.001,
+      "loss": 1.2033,
+      "step": 14302
+    },
+    {
+      "epoch": 0.605085032574668,
+      "grad_norm": 0.17680132389068604,
+      "learning_rate": 0.001,
+      "loss": 2.3674,
+      "step": 14303
+    },
+    {
+      "epoch": 0.6051273373381842,
+      "grad_norm": 0.21177402138710022,
+      "learning_rate": 0.001,
+      "loss": 2.0609,
+      "step": 14304
+    },
+    {
+      "epoch": 0.6051696421017007,
+      "grad_norm": 0.8895161151885986,
+      "learning_rate": 0.001,
+      "loss": 2.1776,
+      "step": 14305
+    },
+    {
+      "epoch": 0.6052119468652171,
+      "grad_norm": 0.18167245388031006,
+      "learning_rate": 0.001,
+      "loss": 1.8625,
+      "step": 14306
+    },
+    {
+      "epoch": 0.6052542516287334,
+      "grad_norm": 7.46832799911499,
+      "learning_rate": 0.001,
+      "loss": 1.9056,
+      "step": 14307
+    },
+    {
+      "epoch": 0.6052965563922498,
+      "grad_norm": 0.17370212078094482,
+      "learning_rate": 0.001,
+      "loss": 2.3242,
+      "step": 14308
+    },
+    {
+      "epoch": 0.6053388611557662,
+      "grad_norm": 0.1907428652048111,
+      "learning_rate": 0.001,
+      "loss": 2.38,
+      "step": 14309
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 0.5028619766235352,
+      "learning_rate": 0.001,
+      "loss": 2.3635,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6054234706827989,
+      "grad_norm": 0.15247510373592377,
+      "learning_rate": 0.001,
+      "loss": 1.7588,
+      "step": 14311
+    },
+    {
+      "epoch": 0.6054657754463153,
+      "grad_norm": 0.2253142148256302,
+      "learning_rate": 0.001,
+      "loss": 2.7246,
+      "step": 14312
+    },
+    {
+      "epoch": 0.6055080802098316,
+      "grad_norm": 0.2002718299627304,
+      "learning_rate": 0.001,
+      "loss": 2.9367,
+      "step": 14313
+    },
+    {
+      "epoch": 0.605550384973348,
+      "grad_norm": 1.097982406616211,
+      "learning_rate": 0.001,
+      "loss": 1.6434,
+      "step": 14314
+    },
+    {
+      "epoch": 0.6055926897368644,
+      "grad_norm": 0.20762449502944946,
+      "learning_rate": 0.001,
+      "loss": 2.281,
+      "step": 14315
+    },
+    {
+      "epoch": 0.6056349945003807,
+      "grad_norm": 0.8287076950073242,
+      "learning_rate": 0.001,
+      "loss": 2.1989,
+      "step": 14316
+    },
+    {
+      "epoch": 0.6056772992638971,
+      "grad_norm": 0.19646909832954407,
+      "learning_rate": 0.001,
+      "loss": 3.1902,
+      "step": 14317
+    },
+    {
+      "epoch": 0.6057196040274135,
+      "grad_norm": 0.21229322254657745,
+      "learning_rate": 0.001,
+      "loss": 2.3791,
+      "step": 14318
+    },
+    {
+      "epoch": 0.6057619087909298,
+      "grad_norm": 0.216338649392128,
+      "learning_rate": 0.001,
+      "loss": 2.6643,
+      "step": 14319
+    },
+    {
+      "epoch": 0.6058042135544462,
+      "grad_norm": 0.177630215883255,
+      "learning_rate": 0.001,
+      "loss": 3.0275,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6058465183179627,
+      "grad_norm": 0.173061802983284,
+      "learning_rate": 0.001,
+      "loss": 1.5222,
+      "step": 14321
+    },
+    {
+      "epoch": 0.605888823081479,
+      "grad_norm": 0.2353350967168808,
+      "learning_rate": 0.001,
+      "loss": 2.1331,
+      "step": 14322
+    },
+    {
+      "epoch": 0.6059311278449954,
+      "grad_norm": 0.1542205810546875,
+      "learning_rate": 0.001,
+      "loss": 1.6273,
+      "step": 14323
+    },
+    {
+      "epoch": 0.6059734326085117,
+      "grad_norm": 0.14759258925914764,
+      "learning_rate": 0.001,
+      "loss": 2.9241,
+      "step": 14324
+    },
+    {
+      "epoch": 0.6060157373720281,
+      "grad_norm": 1.2428350448608398,
+      "learning_rate": 0.001,
+      "loss": 1.851,
+      "step": 14325
+    },
+    {
+      "epoch": 0.6060580421355445,
+      "grad_norm": 0.3622968792915344,
+      "learning_rate": 0.001,
+      "loss": 2.0143,
+      "step": 14326
+    },
+    {
+      "epoch": 0.6061003468990608,
+      "grad_norm": 0.1762288361787796,
+      "learning_rate": 0.001,
+      "loss": 2.4346,
+      "step": 14327
+    },
+    {
+      "epoch": 0.6061426516625772,
+      "grad_norm": 0.24503590166568756,
+      "learning_rate": 0.001,
+      "loss": 1.8936,
+      "step": 14328
+    },
+    {
+      "epoch": 0.6061849564260936,
+      "grad_norm": 0.16428449749946594,
+      "learning_rate": 0.001,
+      "loss": 1.829,
+      "step": 14329
+    },
+    {
+      "epoch": 0.6062272611896099,
+      "grad_norm": 0.16530753672122955,
+      "learning_rate": 0.001,
+      "loss": 2.2527,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6062695659531263,
+      "grad_norm": 0.20797306299209595,
+      "learning_rate": 0.001,
+      "loss": 2.2176,
+      "step": 14331
+    },
+    {
+      "epoch": 0.6063118707166427,
+      "grad_norm": 0.27849939465522766,
+      "learning_rate": 0.001,
+      "loss": 2.2585,
+      "step": 14332
+    },
+    {
+      "epoch": 0.606354175480159,
+      "grad_norm": 1.6591880321502686,
+      "learning_rate": 0.001,
+      "loss": 3.7181,
+      "step": 14333
+    },
+    {
+      "epoch": 0.6063964802436754,
+      "grad_norm": 0.15750765800476074,
+      "learning_rate": 0.001,
+      "loss": 2.2277,
+      "step": 14334
+    },
+    {
+      "epoch": 0.6064387850071918,
+      "grad_norm": 0.5017919540405273,
+      "learning_rate": 0.001,
+      "loss": 2.991,
+      "step": 14335
+    },
+    {
+      "epoch": 0.6064810897707081,
+      "grad_norm": 0.37937721610069275,
+      "learning_rate": 0.001,
+      "loss": 2.2904,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6065233945342245,
+      "grad_norm": 1.156523585319519,
+      "learning_rate": 0.001,
+      "loss": 2.0803,
+      "step": 14337
+    },
+    {
+      "epoch": 0.606565699297741,
+      "grad_norm": 0.26391497254371643,
+      "learning_rate": 0.001,
+      "loss": 2.2633,
+      "step": 14338
+    },
+    {
+      "epoch": 0.6066080040612573,
+      "grad_norm": 0.17540274560451508,
+      "learning_rate": 0.001,
+      "loss": 2.4428,
+      "step": 14339
+    },
+    {
+      "epoch": 0.6066503088247737,
+      "grad_norm": 0.1678505390882492,
+      "learning_rate": 0.001,
+      "loss": 2.1558,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6066926135882901,
+      "grad_norm": 0.1621529459953308,
+      "learning_rate": 0.001,
+      "loss": 2.3497,
+      "step": 14341
+    },
+    {
+      "epoch": 0.6067349183518064,
+      "grad_norm": 0.2335866242647171,
+      "learning_rate": 0.001,
+      "loss": 2.0897,
+      "step": 14342
+    },
+    {
+      "epoch": 0.6067772231153228,
+      "grad_norm": 0.2002691626548767,
+      "learning_rate": 0.001,
+      "loss": 2.613,
+      "step": 14343
+    },
+    {
+      "epoch": 0.6068195278788392,
+      "grad_norm": 13.141894340515137,
+      "learning_rate": 0.001,
+      "loss": 1.9083,
+      "step": 14344
+    },
+    {
+      "epoch": 0.6068618326423555,
+      "grad_norm": 0.19172821938991547,
+      "learning_rate": 0.001,
+      "loss": 1.8666,
+      "step": 14345
+    },
+    {
+      "epoch": 0.6069041374058719,
+      "grad_norm": 0.2590673267841339,
+      "learning_rate": 0.001,
+      "loss": 1.9539,
+      "step": 14346
+    },
+    {
+      "epoch": 0.6069464421693883,
+      "grad_norm": 0.17630375921726227,
+      "learning_rate": 0.001,
+      "loss": 2.6305,
+      "step": 14347
+    },
+    {
+      "epoch": 0.6069887469329046,
+      "grad_norm": 0.14956524968147278,
+      "learning_rate": 0.001,
+      "loss": 2.8753,
+      "step": 14348
+    },
+    {
+      "epoch": 0.607031051696421,
+      "grad_norm": 0.1801750808954239,
+      "learning_rate": 0.001,
+      "loss": 1.9418,
+      "step": 14349
+    },
+    {
+      "epoch": 0.6070733564599374,
+      "grad_norm": 0.1958351731300354,
+      "learning_rate": 0.001,
+      "loss": 3.2227,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6071156612234537,
+      "grad_norm": 0.24762704968452454,
+      "learning_rate": 0.001,
+      "loss": 2.6021,
+      "step": 14351
+    },
+    {
+      "epoch": 0.6071579659869701,
+      "grad_norm": 0.19234824180603027,
+      "learning_rate": 0.001,
+      "loss": 3.2872,
+      "step": 14352
+    },
+    {
+      "epoch": 0.6072002707504865,
+      "grad_norm": 0.1839028298854828,
+      "learning_rate": 0.001,
+      "loss": 2.8299,
+      "step": 14353
+    },
+    {
+      "epoch": 0.6072425755140028,
+      "grad_norm": 0.18792131543159485,
+      "learning_rate": 0.001,
+      "loss": 1.9494,
+      "step": 14354
+    },
+    {
+      "epoch": 0.6072848802775193,
+      "grad_norm": 0.14730499684810638,
+      "learning_rate": 0.001,
+      "loss": 1.4713,
+      "step": 14355
+    },
+    {
+      "epoch": 0.6073271850410357,
+      "grad_norm": 0.5098087787628174,
+      "learning_rate": 0.001,
+      "loss": 1.9097,
+      "step": 14356
+    },
+    {
+      "epoch": 0.607369489804552,
+      "grad_norm": 0.15701937675476074,
+      "learning_rate": 0.001,
+      "loss": 2.2055,
+      "step": 14357
+    },
+    {
+      "epoch": 0.6074117945680684,
+      "grad_norm": 0.16516199707984924,
+      "learning_rate": 0.001,
+      "loss": 2.4988,
+      "step": 14358
+    },
+    {
+      "epoch": 0.6074540993315848,
+      "grad_norm": 0.2014375776052475,
+      "learning_rate": 0.001,
+      "loss": 2.3594,
+      "step": 14359
+    },
+    {
+      "epoch": 0.6074964040951011,
+      "grad_norm": 0.17397524416446686,
+      "learning_rate": 0.001,
+      "loss": 2.3691,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6075387088586175,
+      "grad_norm": 8.400336265563965,
+      "learning_rate": 0.001,
+      "loss": 3.4756,
+      "step": 14361
+    },
+    {
+      "epoch": 0.6075810136221339,
+      "grad_norm": 0.18245413899421692,
+      "learning_rate": 0.001,
+      "loss": 1.7794,
+      "step": 14362
+    },
+    {
+      "epoch": 0.6076233183856502,
+      "grad_norm": 0.19193901121616364,
+      "learning_rate": 0.001,
+      "loss": 1.7134,
+      "step": 14363
+    },
+    {
+      "epoch": 0.6076656231491666,
+      "grad_norm": 0.19857852160930634,
+      "learning_rate": 0.001,
+      "loss": 2.7933,
+      "step": 14364
+    },
+    {
+      "epoch": 0.607707927912683,
+      "grad_norm": 0.25610992312431335,
+      "learning_rate": 0.001,
+      "loss": 1.4993,
+      "step": 14365
+    },
+    {
+      "epoch": 0.6077502326761993,
+      "grad_norm": 0.268586128950119,
+      "learning_rate": 0.001,
+      "loss": 3.1649,
+      "step": 14366
+    },
+    {
+      "epoch": 0.6077925374397157,
+      "grad_norm": 0.30467700958251953,
+      "learning_rate": 0.001,
+      "loss": 2.0349,
+      "step": 14367
+    },
+    {
+      "epoch": 0.607834842203232,
+      "grad_norm": 0.38377171754837036,
+      "learning_rate": 0.001,
+      "loss": 2.3225,
+      "step": 14368
+    },
+    {
+      "epoch": 0.6078771469667484,
+      "grad_norm": 0.16598057746887207,
+      "learning_rate": 0.001,
+      "loss": 2.5853,
+      "step": 14369
+    },
+    {
+      "epoch": 0.6079194517302648,
+      "grad_norm": 0.2181134968996048,
+      "learning_rate": 0.001,
+      "loss": 3.0605,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6079617564937811,
+      "grad_norm": 0.182630717754364,
+      "learning_rate": 0.001,
+      "loss": 1.6975,
+      "step": 14371
+    },
+    {
+      "epoch": 0.6080040612572976,
+      "grad_norm": 0.7057227492332458,
+      "learning_rate": 0.001,
+      "loss": 2.5886,
+      "step": 14372
+    },
+    {
+      "epoch": 0.608046366020814,
+      "grad_norm": 0.1978188157081604,
+      "learning_rate": 0.001,
+      "loss": 2.124,
+      "step": 14373
+    },
+    {
+      "epoch": 0.6080886707843303,
+      "grad_norm": 0.2088436782360077,
+      "learning_rate": 0.001,
+      "loss": 3.0534,
+      "step": 14374
+    },
+    {
+      "epoch": 0.6081309755478467,
+      "grad_norm": 0.15031234920024872,
+      "learning_rate": 0.001,
+      "loss": 2.396,
+      "step": 14375
+    },
+    {
+      "epoch": 0.6081732803113631,
+      "grad_norm": 0.2140101194381714,
+      "learning_rate": 0.001,
+      "loss": 2.1995,
+      "step": 14376
+    },
+    {
+      "epoch": 0.6082155850748794,
+      "grad_norm": 0.15244650840759277,
+      "learning_rate": 0.001,
+      "loss": 2.3434,
+      "step": 14377
+    },
+    {
+      "epoch": 0.6082578898383958,
+      "grad_norm": 0.19588300585746765,
+      "learning_rate": 0.001,
+      "loss": 1.9308,
+      "step": 14378
+    },
+    {
+      "epoch": 0.6083001946019122,
+      "grad_norm": 0.1862742006778717,
+      "learning_rate": 0.001,
+      "loss": 2.6186,
+      "step": 14379
+    },
+    {
+      "epoch": 0.6083424993654285,
+      "grad_norm": 0.16094960272312164,
+      "learning_rate": 0.001,
+      "loss": 1.7264,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6083848041289449,
+      "grad_norm": 0.17328019440174103,
+      "learning_rate": 0.001,
+      "loss": 1.566,
+      "step": 14381
+    },
+    {
+      "epoch": 0.6084271088924613,
+      "grad_norm": 0.15607021749019623,
+      "learning_rate": 0.001,
+      "loss": 2.6352,
+      "step": 14382
+    },
+    {
+      "epoch": 0.6084694136559776,
+      "grad_norm": 0.15530943870544434,
+      "learning_rate": 0.001,
+      "loss": 2.0045,
+      "step": 14383
+    },
+    {
+      "epoch": 0.608511718419494,
+      "grad_norm": 0.16822437942028046,
+      "learning_rate": 0.001,
+      "loss": 3.1795,
+      "step": 14384
+    },
+    {
+      "epoch": 0.6085540231830104,
+      "grad_norm": 0.15865309536457062,
+      "learning_rate": 0.001,
+      "loss": 2.3704,
+      "step": 14385
+    },
+    {
+      "epoch": 0.6085963279465267,
+      "grad_norm": 0.30900880694389343,
+      "learning_rate": 0.001,
+      "loss": 1.8696,
+      "step": 14386
+    },
+    {
+      "epoch": 0.6086386327100431,
+      "grad_norm": 0.1799941211938858,
+      "learning_rate": 0.001,
+      "loss": 2.3696,
+      "step": 14387
+    },
+    {
+      "epoch": 0.6086809374735596,
+      "grad_norm": 0.20973291993141174,
+      "learning_rate": 0.001,
+      "loss": 2.8554,
+      "step": 14388
+    },
+    {
+      "epoch": 0.6087232422370759,
+      "grad_norm": 0.15292273461818695,
+      "learning_rate": 0.001,
+      "loss": 1.5335,
+      "step": 14389
+    },
+    {
+      "epoch": 0.6087655470005923,
+      "grad_norm": 0.1537332534790039,
+      "learning_rate": 0.001,
+      "loss": 2.7656,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6088078517641087,
+      "grad_norm": 0.1706329882144928,
+      "learning_rate": 0.001,
+      "loss": 2.221,
+      "step": 14391
+    },
+    {
+      "epoch": 0.608850156527625,
+      "grad_norm": 0.15366674959659576,
+      "learning_rate": 0.001,
+      "loss": 1.4947,
+      "step": 14392
+    },
+    {
+      "epoch": 0.6088924612911414,
+      "grad_norm": 0.16146250069141388,
+      "learning_rate": 0.001,
+      "loss": 2.2607,
+      "step": 14393
+    },
+    {
+      "epoch": 0.6089347660546578,
+      "grad_norm": 0.1696520894765854,
+      "learning_rate": 0.001,
+      "loss": 2.8822,
+      "step": 14394
+    },
+    {
+      "epoch": 0.6089770708181741,
+      "grad_norm": 0.7845591306686401,
+      "learning_rate": 0.001,
+      "loss": 2.0901,
+      "step": 14395
+    },
+    {
+      "epoch": 0.6090193755816905,
+      "grad_norm": 0.20241577923297882,
+      "learning_rate": 0.001,
+      "loss": 3.0825,
+      "step": 14396
+    },
+    {
+      "epoch": 0.6090616803452069,
+      "grad_norm": 0.16126592457294464,
+      "learning_rate": 0.001,
+      "loss": 1.6057,
+      "step": 14397
+    },
+    {
+      "epoch": 0.6091039851087232,
+      "grad_norm": 0.18037404119968414,
+      "learning_rate": 0.001,
+      "loss": 2.864,
+      "step": 14398
+    },
+    {
+      "epoch": 0.6091462898722396,
+      "grad_norm": 0.1628550887107849,
+      "learning_rate": 0.001,
+      "loss": 3.1012,
+      "step": 14399
+    },
+    {
+      "epoch": 0.609188594635756,
+      "grad_norm": 0.20835568010807037,
+      "learning_rate": 0.001,
+      "loss": 2.0582,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6092308993992723,
+      "grad_norm": 0.18450294435024261,
+      "learning_rate": 0.001,
+      "loss": 2.5508,
+      "step": 14401
+    },
+    {
+      "epoch": 0.6092732041627887,
+      "grad_norm": 0.33053863048553467,
+      "learning_rate": 0.001,
+      "loss": 1.758,
+      "step": 14402
+    },
+    {
+      "epoch": 0.6093155089263051,
+      "grad_norm": 1.6624572277069092,
+      "learning_rate": 0.001,
+      "loss": 2.8896,
+      "step": 14403
+    },
+    {
+      "epoch": 0.6093578136898214,
+      "grad_norm": 0.14562225341796875,
+      "learning_rate": 0.001,
+      "loss": 1.592,
+      "step": 14404
+    },
+    {
+      "epoch": 0.6094001184533379,
+      "grad_norm": 0.45143476128578186,
+      "learning_rate": 0.001,
+      "loss": 1.8523,
+      "step": 14405
+    },
+    {
+      "epoch": 0.6094424232168543,
+      "grad_norm": 0.47654953598976135,
+      "learning_rate": 0.001,
+      "loss": 2.7751,
+      "step": 14406
+    },
+    {
+      "epoch": 0.6094847279803706,
+      "grad_norm": 8.705418586730957,
+      "learning_rate": 0.001,
+      "loss": 1.8628,
+      "step": 14407
+    },
+    {
+      "epoch": 0.609527032743887,
+      "grad_norm": 0.17469455301761627,
+      "learning_rate": 0.001,
+      "loss": 2.4936,
+      "step": 14408
+    },
+    {
+      "epoch": 0.6095693375074034,
+      "grad_norm": 0.20742036402225494,
+      "learning_rate": 0.001,
+      "loss": 2.1476,
+      "step": 14409
+    },
+    {
+      "epoch": 0.6096116422709197,
+      "grad_norm": 0.19691161811351776,
+      "learning_rate": 0.001,
+      "loss": 1.8967,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6096539470344361,
+      "grad_norm": 0.16114525496959686,
+      "learning_rate": 0.001,
+      "loss": 1.6115,
+      "step": 14411
+    },
+    {
+      "epoch": 0.6096962517979525,
+      "grad_norm": 1.256961703300476,
+      "learning_rate": 0.001,
+      "loss": 2.0185,
+      "step": 14412
+    },
+    {
+      "epoch": 0.6097385565614688,
+      "grad_norm": 0.16019225120544434,
+      "learning_rate": 0.001,
+      "loss": 1.7158,
+      "step": 14413
+    },
+    {
+      "epoch": 0.6097808613249852,
+      "grad_norm": 0.23191338777542114,
+      "learning_rate": 0.001,
+      "loss": 2.2885,
+      "step": 14414
+    },
+    {
+      "epoch": 0.6098231660885015,
+      "grad_norm": 0.17342033982276917,
+      "learning_rate": 0.001,
+      "loss": 1.9171,
+      "step": 14415
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.14773350954055786,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 14416
+    },
+    {
+      "epoch": 0.6099077756155343,
+      "grad_norm": 0.23442687094211578,
+      "learning_rate": 0.001,
+      "loss": 2.2754,
+      "step": 14417
+    },
+    {
+      "epoch": 0.6099500803790506,
+      "grad_norm": 0.8946742415428162,
+      "learning_rate": 0.001,
+      "loss": 2.0776,
+      "step": 14418
+    },
+    {
+      "epoch": 0.609992385142567,
+      "grad_norm": 0.22962956130504608,
+      "learning_rate": 0.001,
+      "loss": 2.4009,
+      "step": 14419
+    },
+    {
+      "epoch": 0.6100346899060834,
+      "grad_norm": 0.15502995252609253,
+      "learning_rate": 0.001,
+      "loss": 1.824,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6100769946695997,
+      "grad_norm": 0.15531884133815765,
+      "learning_rate": 0.001,
+      "loss": 1.7114,
+      "step": 14421
+    },
+    {
+      "epoch": 0.6101192994331162,
+      "grad_norm": 0.18697033822536469,
+      "learning_rate": 0.001,
+      "loss": 1.8181,
+      "step": 14422
+    },
+    {
+      "epoch": 0.6101616041966326,
+      "grad_norm": 0.14635103940963745,
+      "learning_rate": 0.001,
+      "loss": 2.2397,
+      "step": 14423
+    },
+    {
+      "epoch": 0.6102039089601489,
+      "grad_norm": 0.15549708902835846,
+      "learning_rate": 0.001,
+      "loss": 2.2291,
+      "step": 14424
+    },
+    {
+      "epoch": 0.6102462137236653,
+      "grad_norm": 0.15069544315338135,
+      "learning_rate": 0.001,
+      "loss": 2.1368,
+      "step": 14425
+    },
+    {
+      "epoch": 0.6102885184871817,
+      "grad_norm": 0.15826652944087982,
+      "learning_rate": 0.001,
+      "loss": 2.4466,
+      "step": 14426
+    },
+    {
+      "epoch": 0.610330823250698,
+      "grad_norm": 0.18682676553726196,
+      "learning_rate": 0.001,
+      "loss": 1.9024,
+      "step": 14427
+    },
+    {
+      "epoch": 0.6103731280142144,
+      "grad_norm": 0.40561962127685547,
+      "learning_rate": 0.001,
+      "loss": 1.8962,
+      "step": 14428
+    },
+    {
+      "epoch": 0.6104154327777308,
+      "grad_norm": 0.20110554993152618,
+      "learning_rate": 0.001,
+      "loss": 2.3748,
+      "step": 14429
+    },
+    {
+      "epoch": 0.6104577375412471,
+      "grad_norm": 0.15364859998226166,
+      "learning_rate": 0.001,
+      "loss": 2.1927,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6105000423047635,
+      "grad_norm": 0.17697322368621826,
+      "learning_rate": 0.001,
+      "loss": 1.8882,
+      "step": 14431
+    },
+    {
+      "epoch": 0.6105423470682799,
+      "grad_norm": 0.7132118344306946,
+      "learning_rate": 0.001,
+      "loss": 3.1219,
+      "step": 14432
+    },
+    {
+      "epoch": 0.6105846518317962,
+      "grad_norm": 0.19553466141223907,
+      "learning_rate": 0.001,
+      "loss": 2.4858,
+      "step": 14433
+    },
+    {
+      "epoch": 0.6106269565953126,
+      "grad_norm": 0.15867097675800323,
+      "learning_rate": 0.001,
+      "loss": 2.2074,
+      "step": 14434
+    },
+    {
+      "epoch": 0.610669261358829,
+      "grad_norm": 0.8470575213432312,
+      "learning_rate": 0.001,
+      "loss": 3.0073,
+      "step": 14435
+    },
+    {
+      "epoch": 0.6107115661223453,
+      "grad_norm": 0.16488543152809143,
+      "learning_rate": 0.001,
+      "loss": 1.6394,
+      "step": 14436
+    },
+    {
+      "epoch": 0.6107538708858617,
+      "grad_norm": 0.2117576003074646,
+      "learning_rate": 0.001,
+      "loss": 2.5382,
+      "step": 14437
+    },
+    {
+      "epoch": 0.6107961756493782,
+      "grad_norm": 0.2346462607383728,
+      "learning_rate": 0.001,
+      "loss": 1.9488,
+      "step": 14438
+    },
+    {
+      "epoch": 0.6108384804128945,
+      "grad_norm": 1.9085814952850342,
+      "learning_rate": 0.001,
+      "loss": 2.3393,
+      "step": 14439
+    },
+    {
+      "epoch": 0.6108807851764109,
+      "grad_norm": 0.15879637002944946,
+      "learning_rate": 0.001,
+      "loss": 2.3031,
+      "step": 14440
+    },
+    {
+      "epoch": 0.6109230899399273,
+      "grad_norm": 0.13453783094882965,
+      "learning_rate": 0.001,
+      "loss": 1.7157,
+      "step": 14441
+    },
+    {
+      "epoch": 0.6109653947034436,
+      "grad_norm": 0.3421229422092438,
+      "learning_rate": 0.001,
+      "loss": 2.4225,
+      "step": 14442
+    },
+    {
+      "epoch": 0.61100769946696,
+      "grad_norm": 0.175466388463974,
+      "learning_rate": 0.001,
+      "loss": 2.3742,
+      "step": 14443
+    },
+    {
+      "epoch": 0.6110500042304764,
+      "grad_norm": 5.125629901885986,
+      "learning_rate": 0.001,
+      "loss": 2.022,
+      "step": 14444
+    },
+    {
+      "epoch": 0.6110923089939927,
+      "grad_norm": 0.15601593255996704,
+      "learning_rate": 0.001,
+      "loss": 3.9727,
+      "step": 14445
+    },
+    {
+      "epoch": 0.6111346137575091,
+      "grad_norm": 0.16740910708904266,
+      "learning_rate": 0.001,
+      "loss": 2.8391,
+      "step": 14446
+    },
+    {
+      "epoch": 0.6111769185210255,
+      "grad_norm": 0.18118330836296082,
+      "learning_rate": 0.001,
+      "loss": 2.154,
+      "step": 14447
+    },
+    {
+      "epoch": 0.6112192232845418,
+      "grad_norm": 0.17362359166145325,
+      "learning_rate": 0.001,
+      "loss": 2.4111,
+      "step": 14448
+    },
+    {
+      "epoch": 0.6112615280480582,
+      "grad_norm": 0.20023968815803528,
+      "learning_rate": 0.001,
+      "loss": 2.0913,
+      "step": 14449
+    },
+    {
+      "epoch": 0.6113038328115746,
+      "grad_norm": 0.954468846321106,
+      "learning_rate": 0.001,
+      "loss": 2.5481,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6113461375750909,
+      "grad_norm": 0.2559969127178192,
+      "learning_rate": 0.001,
+      "loss": 2.5936,
+      "step": 14451
+    },
+    {
+      "epoch": 0.6113884423386073,
+      "grad_norm": 0.1946243792772293,
+      "learning_rate": 0.001,
+      "loss": 2.4807,
+      "step": 14452
+    },
+    {
+      "epoch": 0.6114307471021238,
+      "grad_norm": 0.30189049243927,
+      "learning_rate": 0.001,
+      "loss": 1.8112,
+      "step": 14453
+    },
+    {
+      "epoch": 0.61147305186564,
+      "grad_norm": 0.20965829491615295,
+      "learning_rate": 0.001,
+      "loss": 2.1163,
+      "step": 14454
+    },
+    {
+      "epoch": 0.6115153566291565,
+      "grad_norm": 0.2877630889415741,
+      "learning_rate": 0.001,
+      "loss": 1.9946,
+      "step": 14455
+    },
+    {
+      "epoch": 0.6115576613926729,
+      "grad_norm": 0.23159538209438324,
+      "learning_rate": 0.001,
+      "loss": 2.35,
+      "step": 14456
+    },
+    {
+      "epoch": 0.6115999661561892,
+      "grad_norm": 1.8251463174819946,
+      "learning_rate": 0.001,
+      "loss": 2.014,
+      "step": 14457
+    },
+    {
+      "epoch": 0.6116422709197056,
+      "grad_norm": 0.2677466869354248,
+      "learning_rate": 0.001,
+      "loss": 2.1856,
+      "step": 14458
+    },
+    {
+      "epoch": 0.6116845756832219,
+      "grad_norm": 0.291852742433548,
+      "learning_rate": 0.001,
+      "loss": 3.0096,
+      "step": 14459
+    },
+    {
+      "epoch": 0.6117268804467383,
+      "grad_norm": 0.22112785279750824,
+      "learning_rate": 0.001,
+      "loss": 1.8697,
+      "step": 14460
+    },
+    {
+      "epoch": 0.6117691852102547,
+      "grad_norm": 0.1902538239955902,
+      "learning_rate": 0.001,
+      "loss": 3.42,
+      "step": 14461
+    },
+    {
+      "epoch": 0.611811489973771,
+      "grad_norm": 0.17315338551998138,
+      "learning_rate": 0.001,
+      "loss": 2.4311,
+      "step": 14462
+    },
+    {
+      "epoch": 0.6118537947372874,
+      "grad_norm": 0.17455525696277618,
+      "learning_rate": 0.001,
+      "loss": 2.2537,
+      "step": 14463
+    },
+    {
+      "epoch": 0.6118960995008038,
+      "grad_norm": 0.22085832059383392,
+      "learning_rate": 0.001,
+      "loss": 3.21,
+      "step": 14464
+    },
+    {
+      "epoch": 0.6119384042643201,
+      "grad_norm": 0.16527900099754333,
+      "learning_rate": 0.001,
+      "loss": 1.7905,
+      "step": 14465
+    },
+    {
+      "epoch": 0.6119807090278365,
+      "grad_norm": 0.18405689299106598,
+      "learning_rate": 0.001,
+      "loss": 2.2348,
+      "step": 14466
+    },
+    {
+      "epoch": 0.6120230137913529,
+      "grad_norm": 0.19731445610523224,
+      "learning_rate": 0.001,
+      "loss": 2.0635,
+      "step": 14467
+    },
+    {
+      "epoch": 0.6120653185548692,
+      "grad_norm": 0.20120853185653687,
+      "learning_rate": 0.001,
+      "loss": 3.1794,
+      "step": 14468
+    },
+    {
+      "epoch": 0.6121076233183856,
+      "grad_norm": 0.18305185437202454,
+      "learning_rate": 0.001,
+      "loss": 1.4547,
+      "step": 14469
+    },
+    {
+      "epoch": 0.612149928081902,
+      "grad_norm": 0.18967366218566895,
+      "learning_rate": 0.001,
+      "loss": 1.6065,
+      "step": 14470
+    },
+    {
+      "epoch": 0.6121922328454183,
+      "grad_norm": 0.19936278462409973,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 14471
+    },
+    {
+      "epoch": 0.6122345376089348,
+      "grad_norm": 0.17981190979480743,
+      "learning_rate": 0.001,
+      "loss": 1.6158,
+      "step": 14472
+    },
+    {
+      "epoch": 0.6122768423724512,
+      "grad_norm": 0.24982218444347382,
+      "learning_rate": 0.001,
+      "loss": 2.276,
+      "step": 14473
+    },
+    {
+      "epoch": 0.6123191471359675,
+      "grad_norm": 0.2008954882621765,
+      "learning_rate": 0.001,
+      "loss": 1.9837,
+      "step": 14474
+    },
+    {
+      "epoch": 0.6123614518994839,
+      "grad_norm": 0.1441406011581421,
+      "learning_rate": 0.001,
+      "loss": 2.0582,
+      "step": 14475
+    },
+    {
+      "epoch": 0.6124037566630003,
+      "grad_norm": 0.6654687523841858,
+      "learning_rate": 0.001,
+      "loss": 1.4793,
+      "step": 14476
+    },
+    {
+      "epoch": 0.6124460614265166,
+      "grad_norm": 0.17162881791591644,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 14477
+    },
+    {
+      "epoch": 0.612488366190033,
+      "grad_norm": 0.1826455295085907,
+      "learning_rate": 0.001,
+      "loss": 1.9079,
+      "step": 14478
+    },
+    {
+      "epoch": 0.6125306709535494,
+      "grad_norm": 2.3016321659088135,
+      "learning_rate": 0.001,
+      "loss": 3.0657,
+      "step": 14479
+    },
+    {
+      "epoch": 0.6125729757170657,
+      "grad_norm": 0.24080024659633636,
+      "learning_rate": 0.001,
+      "loss": 2.9867,
+      "step": 14480
+    },
+    {
+      "epoch": 0.6126152804805821,
+      "grad_norm": 0.2750950753688812,
+      "learning_rate": 0.001,
+      "loss": 1.9209,
+      "step": 14481
+    },
+    {
+      "epoch": 0.6126575852440985,
+      "grad_norm": 0.22558815777301788,
+      "learning_rate": 0.001,
+      "loss": 2.6685,
+      "step": 14482
+    },
+    {
+      "epoch": 0.6126998900076148,
+      "grad_norm": 0.1902596652507782,
+      "learning_rate": 0.001,
+      "loss": 2.2843,
+      "step": 14483
+    },
+    {
+      "epoch": 0.6127421947711312,
+      "grad_norm": 0.18809862434864044,
+      "learning_rate": 0.001,
+      "loss": 2.212,
+      "step": 14484
+    },
+    {
+      "epoch": 0.6127844995346476,
+      "grad_norm": 0.1640164852142334,
+      "learning_rate": 0.001,
+      "loss": 2.0967,
+      "step": 14485
+    },
+    {
+      "epoch": 0.6128268042981639,
+      "grad_norm": 0.1704949587583542,
+      "learning_rate": 0.001,
+      "loss": 2.6479,
+      "step": 14486
+    },
+    {
+      "epoch": 0.6128691090616804,
+      "grad_norm": 2.7232117652893066,
+      "learning_rate": 0.001,
+      "loss": 1.8047,
+      "step": 14487
+    },
+    {
+      "epoch": 0.6129114138251968,
+      "grad_norm": 0.16258461773395538,
+      "learning_rate": 0.001,
+      "loss": 2.556,
+      "step": 14488
+    },
+    {
+      "epoch": 0.6129537185887131,
+      "grad_norm": 0.22606851160526276,
+      "learning_rate": 0.001,
+      "loss": 1.8643,
+      "step": 14489
+    },
+    {
+      "epoch": 0.6129960233522295,
+      "grad_norm": 0.39778846502304077,
+      "learning_rate": 0.001,
+      "loss": 2.8814,
+      "step": 14490
+    },
+    {
+      "epoch": 0.6130383281157459,
+      "grad_norm": 0.1924993395805359,
+      "learning_rate": 0.001,
+      "loss": 2.1945,
+      "step": 14491
+    },
+    {
+      "epoch": 0.6130806328792622,
+      "grad_norm": 0.1684507429599762,
+      "learning_rate": 0.001,
+      "loss": 2.0614,
+      "step": 14492
+    },
+    {
+      "epoch": 0.6131229376427786,
+      "grad_norm": 0.14290301501750946,
+      "learning_rate": 0.001,
+      "loss": 1.9804,
+      "step": 14493
+    },
+    {
+      "epoch": 0.613165242406295,
+      "grad_norm": 0.4323841333389282,
+      "learning_rate": 0.001,
+      "loss": 1.4766,
+      "step": 14494
+    },
+    {
+      "epoch": 0.6132075471698113,
+      "grad_norm": 0.15192975103855133,
+      "learning_rate": 0.001,
+      "loss": 1.5979,
+      "step": 14495
+    },
+    {
+      "epoch": 0.6132498519333277,
+      "grad_norm": 0.23368826508522034,
+      "learning_rate": 0.001,
+      "loss": 3.5542,
+      "step": 14496
+    },
+    {
+      "epoch": 0.6132921566968441,
+      "grad_norm": 0.2155388742685318,
+      "learning_rate": 0.001,
+      "loss": 2.2724,
+      "step": 14497
+    },
+    {
+      "epoch": 0.6133344614603604,
+      "grad_norm": 0.15829609334468842,
+      "learning_rate": 0.001,
+      "loss": 3.0361,
+      "step": 14498
+    },
+    {
+      "epoch": 0.6133767662238768,
+      "grad_norm": 0.16065742075443268,
+      "learning_rate": 0.001,
+      "loss": 1.3002,
+      "step": 14499
+    },
+    {
+      "epoch": 0.6134190709873932,
+      "grad_norm": 1.1077728271484375,
+      "learning_rate": 0.001,
+      "loss": 2.6874,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6134613757509095,
+      "grad_norm": 0.27375224232673645,
+      "learning_rate": 0.001,
+      "loss": 2.4451,
+      "step": 14501
+    },
+    {
+      "epoch": 0.6135036805144259,
+      "grad_norm": 0.14725899696350098,
+      "learning_rate": 0.001,
+      "loss": 2.2295,
+      "step": 14502
+    },
+    {
+      "epoch": 0.6135459852779424,
+      "grad_norm": 0.15817123651504517,
+      "learning_rate": 0.001,
+      "loss": 2.0776,
+      "step": 14503
+    },
+    {
+      "epoch": 0.6135882900414587,
+      "grad_norm": 0.16688252985477448,
+      "learning_rate": 0.001,
+      "loss": 1.6458,
+      "step": 14504
+    },
+    {
+      "epoch": 0.6136305948049751,
+      "grad_norm": 0.1742338240146637,
+      "learning_rate": 0.001,
+      "loss": 2.1804,
+      "step": 14505
+    },
+    {
+      "epoch": 0.6136728995684914,
+      "grad_norm": 0.17663906514644623,
+      "learning_rate": 0.001,
+      "loss": 2.1758,
+      "step": 14506
+    },
+    {
+      "epoch": 0.6137152043320078,
+      "grad_norm": 0.19644849002361298,
+      "learning_rate": 0.001,
+      "loss": 2.0111,
+      "step": 14507
+    },
+    {
+      "epoch": 0.6137575090955242,
+      "grad_norm": 0.2182653397321701,
+      "learning_rate": 0.001,
+      "loss": 2.7332,
+      "step": 14508
+    },
+    {
+      "epoch": 0.6137998138590405,
+      "grad_norm": 1.093945026397705,
+      "learning_rate": 0.001,
+      "loss": 2.3192,
+      "step": 14509
+    },
+    {
+      "epoch": 0.6138421186225569,
+      "grad_norm": 0.20383012294769287,
+      "learning_rate": 0.001,
+      "loss": 2.4847,
+      "step": 14510
+    },
+    {
+      "epoch": 0.6138844233860733,
+      "grad_norm": 0.1353861391544342,
+      "learning_rate": 0.001,
+      "loss": 1.9635,
+      "step": 14511
+    },
+    {
+      "epoch": 0.6139267281495896,
+      "grad_norm": 0.2178875207901001,
+      "learning_rate": 0.001,
+      "loss": 1.7267,
+      "step": 14512
+    },
+    {
+      "epoch": 0.613969032913106,
+      "grad_norm": 0.4006967544555664,
+      "learning_rate": 0.001,
+      "loss": 1.8074,
+      "step": 14513
+    },
+    {
+      "epoch": 0.6140113376766224,
+      "grad_norm": 0.16235609352588654,
+      "learning_rate": 0.001,
+      "loss": 3.2147,
+      "step": 14514
+    },
+    {
+      "epoch": 0.6140536424401387,
+      "grad_norm": 0.21571922302246094,
+      "learning_rate": 0.001,
+      "loss": 2.0414,
+      "step": 14515
+    },
+    {
+      "epoch": 0.6140959472036551,
+      "grad_norm": 2.157762289047241,
+      "learning_rate": 0.001,
+      "loss": 2.4857,
+      "step": 14516
+    },
+    {
+      "epoch": 0.6141382519671715,
+      "grad_norm": 0.20253095030784607,
+      "learning_rate": 0.001,
+      "loss": 1.9104,
+      "step": 14517
+    },
+    {
+      "epoch": 0.6141805567306878,
+      "grad_norm": 0.14087221026420593,
+      "learning_rate": 0.001,
+      "loss": 1.518,
+      "step": 14518
+    },
+    {
+      "epoch": 0.6142228614942042,
+      "grad_norm": 0.2999531924724579,
+      "learning_rate": 0.001,
+      "loss": 2.6567,
+      "step": 14519
+    },
+    {
+      "epoch": 0.6142651662577207,
+      "grad_norm": 3.447448253631592,
+      "learning_rate": 0.001,
+      "loss": 1.865,
+      "step": 14520
+    },
+    {
+      "epoch": 0.614307471021237,
+      "grad_norm": 0.217742919921875,
+      "learning_rate": 0.001,
+      "loss": 2.1669,
+      "step": 14521
+    },
+    {
+      "epoch": 0.6143497757847534,
+      "grad_norm": 0.22795531153678894,
+      "learning_rate": 0.001,
+      "loss": 2.0143,
+      "step": 14522
+    },
+    {
+      "epoch": 0.6143920805482698,
+      "grad_norm": 0.15495827794075012,
+      "learning_rate": 0.001,
+      "loss": 1.879,
+      "step": 14523
+    },
+    {
+      "epoch": 0.6144343853117861,
+      "grad_norm": 0.1813746690750122,
+      "learning_rate": 0.001,
+      "loss": 3.0125,
+      "step": 14524
+    },
+    {
+      "epoch": 0.6144766900753025,
+      "grad_norm": 0.33635345101356506,
+      "learning_rate": 0.001,
+      "loss": 2.0765,
+      "step": 14525
+    },
+    {
+      "epoch": 0.6145189948388189,
+      "grad_norm": 0.2094181776046753,
+      "learning_rate": 0.001,
+      "loss": 2.7443,
+      "step": 14526
+    },
+    {
+      "epoch": 0.6145612996023352,
+      "grad_norm": 0.18319903314113617,
+      "learning_rate": 0.001,
+      "loss": 1.7433,
+      "step": 14527
+    },
+    {
+      "epoch": 0.6146036043658516,
+      "grad_norm": 0.25899675488471985,
+      "learning_rate": 0.001,
+      "loss": 1.8407,
+      "step": 14528
+    },
+    {
+      "epoch": 0.614645909129368,
+      "grad_norm": 0.17044375836849213,
+      "learning_rate": 0.001,
+      "loss": 3.0131,
+      "step": 14529
+    },
+    {
+      "epoch": 0.6146882138928843,
+      "grad_norm": 0.19204241037368774,
+      "learning_rate": 0.001,
+      "loss": 1.6593,
+      "step": 14530
+    },
+    {
+      "epoch": 0.6147305186564007,
+      "grad_norm": 0.15827490389347076,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 14531
+    },
+    {
+      "epoch": 0.6147728234199171,
+      "grad_norm": 0.16154128313064575,
+      "learning_rate": 0.001,
+      "loss": 2.025,
+      "step": 14532
+    },
+    {
+      "epoch": 0.6148151281834334,
+      "grad_norm": 0.18466542661190033,
+      "learning_rate": 0.001,
+      "loss": 1.822,
+      "step": 14533
+    },
+    {
+      "epoch": 0.6148574329469498,
+      "grad_norm": 0.15772341191768646,
+      "learning_rate": 0.001,
+      "loss": 2.458,
+      "step": 14534
+    },
+    {
+      "epoch": 0.6148997377104662,
+      "grad_norm": 0.2001129388809204,
+      "learning_rate": 0.001,
+      "loss": 2.7717,
+      "step": 14535
+    },
+    {
+      "epoch": 0.6149420424739825,
+      "grad_norm": 0.49999427795410156,
+      "learning_rate": 0.001,
+      "loss": 1.7162,
+      "step": 14536
+    },
+    {
+      "epoch": 0.614984347237499,
+      "grad_norm": 0.18692375719547272,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 14537
+    },
+    {
+      "epoch": 0.6150266520010154,
+      "grad_norm": 0.43777942657470703,
+      "learning_rate": 0.001,
+      "loss": 1.8468,
+      "step": 14538
+    },
+    {
+      "epoch": 0.6150689567645317,
+      "grad_norm": 0.18012431263923645,
+      "learning_rate": 0.001,
+      "loss": 2.0141,
+      "step": 14539
+    },
+    {
+      "epoch": 0.6151112615280481,
+      "grad_norm": 0.17725875973701477,
+      "learning_rate": 0.001,
+      "loss": 1.9882,
+      "step": 14540
+    },
+    {
+      "epoch": 0.6151535662915645,
+      "grad_norm": 0.19504068791866302,
+      "learning_rate": 0.001,
+      "loss": 2.7564,
+      "step": 14541
+    },
+    {
+      "epoch": 0.6151958710550808,
+      "grad_norm": 0.18181535601615906,
+      "learning_rate": 0.001,
+      "loss": 1.6593,
+      "step": 14542
+    },
+    {
+      "epoch": 0.6152381758185972,
+      "grad_norm": 0.1605805605649948,
+      "learning_rate": 0.001,
+      "loss": 2.4811,
+      "step": 14543
+    },
+    {
+      "epoch": 0.6152804805821136,
+      "grad_norm": 0.1837165355682373,
+      "learning_rate": 0.001,
+      "loss": 2.5541,
+      "step": 14544
+    },
+    {
+      "epoch": 0.6153227853456299,
+      "grad_norm": 0.1661083698272705,
+      "learning_rate": 0.001,
+      "loss": 2.745,
+      "step": 14545
+    },
+    {
+      "epoch": 0.6153650901091463,
+      "grad_norm": 0.1631242036819458,
+      "learning_rate": 0.001,
+      "loss": 3.1111,
+      "step": 14546
+    },
+    {
+      "epoch": 0.6154073948726627,
+      "grad_norm": 0.1594124138355255,
+      "learning_rate": 0.001,
+      "loss": 2.1363,
+      "step": 14547
+    },
+    {
+      "epoch": 0.615449699636179,
+      "grad_norm": 0.1624990850687027,
+      "learning_rate": 0.001,
+      "loss": 1.9491,
+      "step": 14548
+    },
+    {
+      "epoch": 0.6154920043996954,
+      "grad_norm": 0.15371422469615936,
+      "learning_rate": 0.001,
+      "loss": 1.8865,
+      "step": 14549
+    },
+    {
+      "epoch": 0.6155343091632117,
+      "grad_norm": 0.1542615443468094,
+      "learning_rate": 0.001,
+      "loss": 1.8565,
+      "step": 14550
+    },
+    {
+      "epoch": 0.6155766139267281,
+      "grad_norm": 0.18403269350528717,
+      "learning_rate": 0.001,
+      "loss": 1.5119,
+      "step": 14551
+    },
+    {
+      "epoch": 0.6156189186902445,
+      "grad_norm": 0.19757772982120514,
+      "learning_rate": 0.001,
+      "loss": 2.071,
+      "step": 14552
+    },
+    {
+      "epoch": 0.6156612234537608,
+      "grad_norm": 0.21193090081214905,
+      "learning_rate": 0.001,
+      "loss": 1.5183,
+      "step": 14553
+    },
+    {
+      "epoch": 0.6157035282172773,
+      "grad_norm": 0.15698926150798798,
+      "learning_rate": 0.001,
+      "loss": 1.9242,
+      "step": 14554
+    },
+    {
+      "epoch": 0.6157458329807937,
+      "grad_norm": 0.18835414946079254,
+      "learning_rate": 0.001,
+      "loss": 2.5247,
+      "step": 14555
+    },
+    {
+      "epoch": 0.61578813774431,
+      "grad_norm": 0.1583268940448761,
+      "learning_rate": 0.001,
+      "loss": 1.9729,
+      "step": 14556
+    },
+    {
+      "epoch": 0.6158304425078264,
+      "grad_norm": 0.16368068754673004,
+      "learning_rate": 0.001,
+      "loss": 2.4374,
+      "step": 14557
+    },
+    {
+      "epoch": 0.6158727472713428,
+      "grad_norm": 0.1612701714038849,
+      "learning_rate": 0.001,
+      "loss": 1.6765,
+      "step": 14558
+    },
+    {
+      "epoch": 0.6159150520348591,
+      "grad_norm": 0.14190715551376343,
+      "learning_rate": 0.001,
+      "loss": 1.9572,
+      "step": 14559
+    },
+    {
+      "epoch": 0.6159573567983755,
+      "grad_norm": 0.15973453223705292,
+      "learning_rate": 0.001,
+      "loss": 2.3115,
+      "step": 14560
+    },
+    {
+      "epoch": 0.6159996615618919,
+      "grad_norm": 0.15088126063346863,
+      "learning_rate": 0.001,
+      "loss": 2.3368,
+      "step": 14561
+    },
+    {
+      "epoch": 0.6160419663254082,
+      "grad_norm": 0.1798139065504074,
+      "learning_rate": 0.001,
+      "loss": 2.6461,
+      "step": 14562
+    },
+    {
+      "epoch": 0.6160842710889246,
+      "grad_norm": 2.4017934799194336,
+      "learning_rate": 0.001,
+      "loss": 2.928,
+      "step": 14563
+    },
+    {
+      "epoch": 0.616126575852441,
+      "grad_norm": 0.14553676545619965,
+      "learning_rate": 0.001,
+      "loss": 1.5844,
+      "step": 14564
+    },
+    {
+      "epoch": 0.6161688806159573,
+      "grad_norm": 0.18981827795505524,
+      "learning_rate": 0.001,
+      "loss": 1.7026,
+      "step": 14565
+    },
+    {
+      "epoch": 0.6162111853794737,
+      "grad_norm": 0.1583603173494339,
+      "learning_rate": 0.001,
+      "loss": 1.93,
+      "step": 14566
+    },
+    {
+      "epoch": 0.6162534901429901,
+      "grad_norm": 0.19451354444026947,
+      "learning_rate": 0.001,
+      "loss": 2.2567,
+      "step": 14567
+    },
+    {
+      "epoch": 0.6162957949065064,
+      "grad_norm": 0.18572446703910828,
+      "learning_rate": 0.001,
+      "loss": 3.772,
+      "step": 14568
+    },
+    {
+      "epoch": 0.6163380996700228,
+      "grad_norm": 0.16187942028045654,
+      "learning_rate": 0.001,
+      "loss": 2.2992,
+      "step": 14569
+    },
+    {
+      "epoch": 0.6163804044335393,
+      "grad_norm": 0.1688256412744522,
+      "learning_rate": 0.001,
+      "loss": 1.6916,
+      "step": 14570
+    },
+    {
+      "epoch": 0.6164227091970556,
+      "grad_norm": 0.16254852712154388,
+      "learning_rate": 0.001,
+      "loss": 1.4337,
+      "step": 14571
+    },
+    {
+      "epoch": 0.616465013960572,
+      "grad_norm": 0.14653462171554565,
+      "learning_rate": 0.001,
+      "loss": 2.2641,
+      "step": 14572
+    },
+    {
+      "epoch": 0.6165073187240884,
+      "grad_norm": 0.18988440930843353,
+      "learning_rate": 0.001,
+      "loss": 2.2461,
+      "step": 14573
+    },
+    {
+      "epoch": 0.6165496234876047,
+      "grad_norm": 0.19180841743946075,
+      "learning_rate": 0.001,
+      "loss": 2.1642,
+      "step": 14574
+    },
+    {
+      "epoch": 0.6165919282511211,
+      "grad_norm": 0.17922373116016388,
+      "learning_rate": 0.001,
+      "loss": 2.0952,
+      "step": 14575
+    },
+    {
+      "epoch": 0.6166342330146375,
+      "grad_norm": 0.17525149881839752,
+      "learning_rate": 0.001,
+      "loss": 2.2221,
+      "step": 14576
+    },
+    {
+      "epoch": 0.6166765377781538,
+      "grad_norm": 0.18129895627498627,
+      "learning_rate": 0.001,
+      "loss": 2.5061,
+      "step": 14577
+    },
+    {
+      "epoch": 0.6167188425416702,
+      "grad_norm": 0.3474824130535126,
+      "learning_rate": 0.001,
+      "loss": 2.2037,
+      "step": 14578
+    },
+    {
+      "epoch": 0.6167611473051866,
+      "grad_norm": 0.2225697934627533,
+      "learning_rate": 0.001,
+      "loss": 2.9076,
+      "step": 14579
+    },
+    {
+      "epoch": 0.6168034520687029,
+      "grad_norm": 0.15336523950099945,
+      "learning_rate": 0.001,
+      "loss": 2.0626,
+      "step": 14580
+    },
+    {
+      "epoch": 0.6168457568322193,
+      "grad_norm": 0.22657287120819092,
+      "learning_rate": 0.001,
+      "loss": 1.8477,
+      "step": 14581
+    },
+    {
+      "epoch": 0.6168880615957357,
+      "grad_norm": 0.19697429239749908,
+      "learning_rate": 0.001,
+      "loss": 2.6378,
+      "step": 14582
+    },
+    {
+      "epoch": 0.616930366359252,
+      "grad_norm": 0.783888041973114,
+      "learning_rate": 0.001,
+      "loss": 2.097,
+      "step": 14583
+    },
+    {
+      "epoch": 0.6169726711227684,
+      "grad_norm": 0.2075197994709015,
+      "learning_rate": 0.001,
+      "loss": 2.4211,
+      "step": 14584
+    },
+    {
+      "epoch": 0.6170149758862848,
+      "grad_norm": 0.1552828848361969,
+      "learning_rate": 0.001,
+      "loss": 2.6256,
+      "step": 14585
+    },
+    {
+      "epoch": 0.6170572806498011,
+      "grad_norm": 0.1605806201696396,
+      "learning_rate": 0.001,
+      "loss": 1.842,
+      "step": 14586
+    },
+    {
+      "epoch": 0.6170995854133176,
+      "grad_norm": 0.18926933407783508,
+      "learning_rate": 0.001,
+      "loss": 2.7009,
+      "step": 14587
+    },
+    {
+      "epoch": 0.617141890176834,
+      "grad_norm": 0.17639969289302826,
+      "learning_rate": 0.001,
+      "loss": 2.0845,
+      "step": 14588
+    },
+    {
+      "epoch": 0.6171841949403503,
+      "grad_norm": 0.16132144629955292,
+      "learning_rate": 0.001,
+      "loss": 2.5827,
+      "step": 14589
+    },
+    {
+      "epoch": 0.6172264997038667,
+      "grad_norm": 0.1768246591091156,
+      "learning_rate": 0.001,
+      "loss": 2.6096,
+      "step": 14590
+    },
+    {
+      "epoch": 0.6172688044673831,
+      "grad_norm": 0.1509474366903305,
+      "learning_rate": 0.001,
+      "loss": 3.6571,
+      "step": 14591
+    },
+    {
+      "epoch": 0.6173111092308994,
+      "grad_norm": 0.17288728058338165,
+      "learning_rate": 0.001,
+      "loss": 2.6354,
+      "step": 14592
+    },
+    {
+      "epoch": 0.6173534139944158,
+      "grad_norm": 0.1312096118927002,
+      "learning_rate": 0.001,
+      "loss": 2.0085,
+      "step": 14593
+    },
+    {
+      "epoch": 0.6173957187579321,
+      "grad_norm": 0.2396790236234665,
+      "learning_rate": 0.001,
+      "loss": 2.3952,
+      "step": 14594
+    },
+    {
+      "epoch": 0.6174380235214485,
+      "grad_norm": 0.1728651076555252,
+      "learning_rate": 0.001,
+      "loss": 1.4488,
+      "step": 14595
+    },
+    {
+      "epoch": 0.6174803282849649,
+      "grad_norm": 0.5452121496200562,
+      "learning_rate": 0.001,
+      "loss": 1.9965,
+      "step": 14596
+    },
+    {
+      "epoch": 0.6175226330484812,
+      "grad_norm": 0.8020671010017395,
+      "learning_rate": 0.001,
+      "loss": 1.8593,
+      "step": 14597
+    },
+    {
+      "epoch": 0.6175649378119976,
+      "grad_norm": 0.15915119647979736,
+      "learning_rate": 0.001,
+      "loss": 2.0414,
+      "step": 14598
+    },
+    {
+      "epoch": 0.617607242575514,
+      "grad_norm": 0.13803361356258392,
+      "learning_rate": 0.001,
+      "loss": 2.0599,
+      "step": 14599
+    },
+    {
+      "epoch": 0.6176495473390303,
+      "grad_norm": 0.22241367399692535,
+      "learning_rate": 0.001,
+      "loss": 2.8024,
+      "step": 14600
+    },
+    {
+      "epoch": 0.6176918521025467,
+      "grad_norm": 1.1794174909591675,
+      "learning_rate": 0.001,
+      "loss": 2.3638,
+      "step": 14601
+    },
+    {
+      "epoch": 0.6177341568660631,
+      "grad_norm": 0.20139962434768677,
+      "learning_rate": 0.001,
+      "loss": 2.7807,
+      "step": 14602
+    },
+    {
+      "epoch": 0.6177764616295794,
+      "grad_norm": 0.19287584722042084,
+      "learning_rate": 0.001,
+      "loss": 2.6512,
+      "step": 14603
+    },
+    {
+      "epoch": 0.6178187663930959,
+      "grad_norm": 0.18711356818675995,
+      "learning_rate": 0.001,
+      "loss": 2.3137,
+      "step": 14604
+    },
+    {
+      "epoch": 0.6178610711566123,
+      "grad_norm": 0.18138305842876434,
+      "learning_rate": 0.001,
+      "loss": 2.7517,
+      "step": 14605
+    },
+    {
+      "epoch": 0.6179033759201286,
+      "grad_norm": 0.1796899139881134,
+      "learning_rate": 0.001,
+      "loss": 2.9958,
+      "step": 14606
+    },
+    {
+      "epoch": 0.617945680683645,
+      "grad_norm": 0.20048393309116364,
+      "learning_rate": 0.001,
+      "loss": 3.1626,
+      "step": 14607
+    },
+    {
+      "epoch": 0.6179879854471614,
+      "grad_norm": 0.1451716423034668,
+      "learning_rate": 0.001,
+      "loss": 2.6609,
+      "step": 14608
+    },
+    {
+      "epoch": 0.6180302902106777,
+      "grad_norm": 0.4020373225212097,
+      "learning_rate": 0.001,
+      "loss": 2.3252,
+      "step": 14609
+    },
+    {
+      "epoch": 0.6180725949741941,
+      "grad_norm": 0.2272024154663086,
+      "learning_rate": 0.001,
+      "loss": 2.3946,
+      "step": 14610
+    },
+    {
+      "epoch": 0.6181148997377105,
+      "grad_norm": 0.24336324632167816,
+      "learning_rate": 0.001,
+      "loss": 2.0187,
+      "step": 14611
+    },
+    {
+      "epoch": 0.6181572045012268,
+      "grad_norm": 3.6941826343536377,
+      "learning_rate": 0.001,
+      "loss": 1.9722,
+      "step": 14612
+    },
+    {
+      "epoch": 0.6181995092647432,
+      "grad_norm": 0.1938718855381012,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 14613
+    },
+    {
+      "epoch": 0.6182418140282596,
+      "grad_norm": 0.6641799211502075,
+      "learning_rate": 0.001,
+      "loss": 2.0153,
+      "step": 14614
+    },
+    {
+      "epoch": 0.6182841187917759,
+      "grad_norm": 0.4109531342983246,
+      "learning_rate": 0.001,
+      "loss": 1.9137,
+      "step": 14615
+    },
+    {
+      "epoch": 0.6183264235552923,
+      "grad_norm": 0.14402462542057037,
+      "learning_rate": 0.001,
+      "loss": 2.4357,
+      "step": 14616
+    },
+    {
+      "epoch": 0.6183687283188087,
+      "grad_norm": 0.14747104048728943,
+      "learning_rate": 0.001,
+      "loss": 2.7109,
+      "step": 14617
+    },
+    {
+      "epoch": 0.618411033082325,
+      "grad_norm": 0.14934441447257996,
+      "learning_rate": 0.001,
+      "loss": 1.6499,
+      "step": 14618
+    },
+    {
+      "epoch": 0.6184533378458414,
+      "grad_norm": 0.3803541362285614,
+      "learning_rate": 0.001,
+      "loss": 2.4184,
+      "step": 14619
+    },
+    {
+      "epoch": 0.6184956426093579,
+      "grad_norm": 0.18346597254276276,
+      "learning_rate": 0.001,
+      "loss": 2.0202,
+      "step": 14620
+    },
+    {
+      "epoch": 0.6185379473728742,
+      "grad_norm": 2.4871697425842285,
+      "learning_rate": 0.001,
+      "loss": 1.9789,
+      "step": 14621
+    },
+    {
+      "epoch": 0.6185802521363906,
+      "grad_norm": 0.15989036858081818,
+      "learning_rate": 0.001,
+      "loss": 1.9835,
+      "step": 14622
+    },
+    {
+      "epoch": 0.618622556899907,
+      "grad_norm": 0.17353324592113495,
+      "learning_rate": 0.001,
+      "loss": 1.9959,
+      "step": 14623
+    },
+    {
+      "epoch": 0.6186648616634233,
+      "grad_norm": 0.18269836902618408,
+      "learning_rate": 0.001,
+      "loss": 2.9858,
+      "step": 14624
+    },
+    {
+      "epoch": 0.6187071664269397,
+      "grad_norm": 0.19032901525497437,
+      "learning_rate": 0.001,
+      "loss": 3.0415,
+      "step": 14625
+    },
+    {
+      "epoch": 0.6187494711904561,
+      "grad_norm": 0.17497460544109344,
+      "learning_rate": 0.001,
+      "loss": 2.0095,
+      "step": 14626
+    },
+    {
+      "epoch": 0.6187917759539724,
+      "grad_norm": 0.14147375524044037,
+      "learning_rate": 0.001,
+      "loss": 1.6253,
+      "step": 14627
+    },
+    {
+      "epoch": 0.6188340807174888,
+      "grad_norm": 0.1689339280128479,
+      "learning_rate": 0.001,
+      "loss": 2.3544,
+      "step": 14628
+    },
+    {
+      "epoch": 0.6188763854810052,
+      "grad_norm": 0.14351142942905426,
+      "learning_rate": 0.001,
+      "loss": 2.6217,
+      "step": 14629
+    },
+    {
+      "epoch": 0.6189186902445215,
+      "grad_norm": 0.30054372549057007,
+      "learning_rate": 0.001,
+      "loss": 2.5242,
+      "step": 14630
+    },
+    {
+      "epoch": 0.6189609950080379,
+      "grad_norm": 0.1856541931629181,
+      "learning_rate": 0.001,
+      "loss": 2.0127,
+      "step": 14631
+    },
+    {
+      "epoch": 0.6190032997715543,
+      "grad_norm": 0.2573411762714386,
+      "learning_rate": 0.001,
+      "loss": 2.2372,
+      "step": 14632
+    },
+    {
+      "epoch": 0.6190456045350706,
+      "grad_norm": 0.1990809440612793,
+      "learning_rate": 0.001,
+      "loss": 2.2482,
+      "step": 14633
+    },
+    {
+      "epoch": 0.619087909298587,
+      "grad_norm": 0.1465701162815094,
+      "learning_rate": 0.001,
+      "loss": 1.8241,
+      "step": 14634
+    },
+    {
+      "epoch": 0.6191302140621034,
+      "grad_norm": 0.1555158942937851,
+      "learning_rate": 0.001,
+      "loss": 2.1623,
+      "step": 14635
+    },
+    {
+      "epoch": 0.6191725188256197,
+      "grad_norm": 0.14374671876430511,
+      "learning_rate": 0.001,
+      "loss": 2.4611,
+      "step": 14636
+    },
+    {
+      "epoch": 0.6192148235891362,
+      "grad_norm": 0.19940204918384552,
+      "learning_rate": 0.001,
+      "loss": 2.1044,
+      "step": 14637
+    },
+    {
+      "epoch": 0.6192571283526526,
+      "grad_norm": 0.14661625027656555,
+      "learning_rate": 0.001,
+      "loss": 1.6882,
+      "step": 14638
+    },
+    {
+      "epoch": 0.6192994331161689,
+      "grad_norm": 0.17937296628952026,
+      "learning_rate": 0.001,
+      "loss": 2.2717,
+      "step": 14639
+    },
+    {
+      "epoch": 0.6193417378796853,
+      "grad_norm": 0.15153121948242188,
+      "learning_rate": 0.001,
+      "loss": 1.7106,
+      "step": 14640
+    },
+    {
+      "epoch": 0.6193840426432016,
+      "grad_norm": 0.1418657898902893,
+      "learning_rate": 0.001,
+      "loss": 1.6178,
+      "step": 14641
+    },
+    {
+      "epoch": 0.619426347406718,
+      "grad_norm": 0.19973745942115784,
+      "learning_rate": 0.001,
+      "loss": 1.786,
+      "step": 14642
+    },
+    {
+      "epoch": 0.6194686521702344,
+      "grad_norm": 0.16791045665740967,
+      "learning_rate": 0.001,
+      "loss": 1.7425,
+      "step": 14643
+    },
+    {
+      "epoch": 0.6195109569337507,
+      "grad_norm": 1.9201226234436035,
+      "learning_rate": 0.001,
+      "loss": 3.1445,
+      "step": 14644
+    },
+    {
+      "epoch": 0.6195532616972671,
+      "grad_norm": 0.1937858760356903,
+      "learning_rate": 0.001,
+      "loss": 3.0628,
+      "step": 14645
+    },
+    {
+      "epoch": 0.6195955664607835,
+      "grad_norm": 0.8908295631408691,
+      "learning_rate": 0.001,
+      "loss": 1.4817,
+      "step": 14646
+    },
+    {
+      "epoch": 0.6196378712242998,
+      "grad_norm": 0.18158625066280365,
+      "learning_rate": 0.001,
+      "loss": 2.1148,
+      "step": 14647
+    },
+    {
+      "epoch": 0.6196801759878162,
+      "grad_norm": 0.18421104550361633,
+      "learning_rate": 0.001,
+      "loss": 1.7709,
+      "step": 14648
+    },
+    {
+      "epoch": 0.6197224807513326,
+      "grad_norm": 0.19580382108688354,
+      "learning_rate": 0.001,
+      "loss": 2.2813,
+      "step": 14649
+    },
+    {
+      "epoch": 0.6197647855148489,
+      "grad_norm": 0.23519310355186462,
+      "learning_rate": 0.001,
+      "loss": 1.6237,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6198070902783653,
+      "grad_norm": 0.27454647421836853,
+      "learning_rate": 0.001,
+      "loss": 2.3794,
+      "step": 14651
+    },
+    {
+      "epoch": 0.6198493950418817,
+      "grad_norm": 0.23078219592571259,
+      "learning_rate": 0.001,
+      "loss": 2.2003,
+      "step": 14652
+    },
+    {
+      "epoch": 0.619891699805398,
+      "grad_norm": 2.1767003536224365,
+      "learning_rate": 0.001,
+      "loss": 2.1137,
+      "step": 14653
+    },
+    {
+      "epoch": 0.6199340045689145,
+      "grad_norm": 0.7693164944648743,
+      "learning_rate": 0.001,
+      "loss": 1.6826,
+      "step": 14654
+    },
+    {
+      "epoch": 0.6199763093324309,
+      "grad_norm": 0.18725281953811646,
+      "learning_rate": 0.001,
+      "loss": 1.6722,
+      "step": 14655
+    },
+    {
+      "epoch": 0.6200186140959472,
+      "grad_norm": 0.17474079132080078,
+      "learning_rate": 0.001,
+      "loss": 1.6512,
+      "step": 14656
+    },
+    {
+      "epoch": 0.6200609188594636,
+      "grad_norm": 0.2864467203617096,
+      "learning_rate": 0.001,
+      "loss": 1.657,
+      "step": 14657
+    },
+    {
+      "epoch": 0.62010322362298,
+      "grad_norm": 0.16446319222450256,
+      "learning_rate": 0.001,
+      "loss": 2.22,
+      "step": 14658
+    },
+    {
+      "epoch": 0.6201455283864963,
+      "grad_norm": 0.21445626020431519,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 14659
+    },
+    {
+      "epoch": 0.6201878331500127,
+      "grad_norm": 0.18971355259418488,
+      "learning_rate": 0.001,
+      "loss": 1.6979,
+      "step": 14660
+    },
+    {
+      "epoch": 0.6202301379135291,
+      "grad_norm": 0.1920858919620514,
+      "learning_rate": 0.001,
+      "loss": 2.9302,
+      "step": 14661
+    },
+    {
+      "epoch": 0.6202724426770454,
+      "grad_norm": 2.1202192306518555,
+      "learning_rate": 0.001,
+      "loss": 2.9636,
+      "step": 14662
+    },
+    {
+      "epoch": 0.6203147474405618,
+      "grad_norm": 0.1833968460559845,
+      "learning_rate": 0.001,
+      "loss": 2.9163,
+      "step": 14663
+    },
+    {
+      "epoch": 0.6203570522040782,
+      "grad_norm": 1.7556793689727783,
+      "learning_rate": 0.001,
+      "loss": 2.1679,
+      "step": 14664
+    },
+    {
+      "epoch": 0.6203993569675945,
+      "grad_norm": 0.1881345808506012,
+      "learning_rate": 0.001,
+      "loss": 1.4928,
+      "step": 14665
+    },
+    {
+      "epoch": 0.6204416617311109,
+      "grad_norm": 0.23631784319877625,
+      "learning_rate": 0.001,
+      "loss": 2.231,
+      "step": 14666
+    },
+    {
+      "epoch": 0.6204839664946273,
+      "grad_norm": 0.23086075484752655,
+      "learning_rate": 0.001,
+      "loss": 2.1282,
+      "step": 14667
+    },
+    {
+      "epoch": 0.6205262712581436,
+      "grad_norm": 0.23201145231723785,
+      "learning_rate": 0.001,
+      "loss": 2.5473,
+      "step": 14668
+    },
+    {
+      "epoch": 0.62056857602166,
+      "grad_norm": 0.20039266347885132,
+      "learning_rate": 0.001,
+      "loss": 2.1767,
+      "step": 14669
+    },
+    {
+      "epoch": 0.6206108807851765,
+      "grad_norm": 0.2201448231935501,
+      "learning_rate": 0.001,
+      "loss": 2.3714,
+      "step": 14670
+    },
+    {
+      "epoch": 0.6206531855486928,
+      "grad_norm": 0.25646641850471497,
+      "learning_rate": 0.001,
+      "loss": 2.2821,
+      "step": 14671
+    },
+    {
+      "epoch": 0.6206954903122092,
+      "grad_norm": 0.2379496693611145,
+      "learning_rate": 0.001,
+      "loss": 2.428,
+      "step": 14672
+    },
+    {
+      "epoch": 0.6207377950757256,
+      "grad_norm": 0.6765031218528748,
+      "learning_rate": 0.001,
+      "loss": 2.5439,
+      "step": 14673
+    },
+    {
+      "epoch": 0.6207800998392419,
+      "grad_norm": 0.23694269359111786,
+      "learning_rate": 0.001,
+      "loss": 3.0224,
+      "step": 14674
+    },
+    {
+      "epoch": 0.6208224046027583,
+      "grad_norm": 0.22863154113292694,
+      "learning_rate": 0.001,
+      "loss": 2.0422,
+      "step": 14675
+    },
+    {
+      "epoch": 0.6208647093662747,
+      "grad_norm": 0.2052396535873413,
+      "learning_rate": 0.001,
+      "loss": 2.1184,
+      "step": 14676
+    },
+    {
+      "epoch": 0.620907014129791,
+      "grad_norm": 0.42548587918281555,
+      "learning_rate": 0.001,
+      "loss": 2.3153,
+      "step": 14677
+    },
+    {
+      "epoch": 0.6209493188933074,
+      "grad_norm": 0.20352311432361603,
+      "learning_rate": 0.001,
+      "loss": 2.7797,
+      "step": 14678
+    },
+    {
+      "epoch": 0.6209916236568238,
+      "grad_norm": 1.041845440864563,
+      "learning_rate": 0.001,
+      "loss": 1.669,
+      "step": 14679
+    },
+    {
+      "epoch": 0.6210339284203401,
+      "grad_norm": 0.19072070717811584,
+      "learning_rate": 0.001,
+      "loss": 2.6243,
+      "step": 14680
+    },
+    {
+      "epoch": 0.6210762331838565,
+      "grad_norm": 0.17468595504760742,
+      "learning_rate": 0.001,
+      "loss": 2.0063,
+      "step": 14681
+    },
+    {
+      "epoch": 0.6211185379473729,
+      "grad_norm": 0.18021397292613983,
+      "learning_rate": 0.001,
+      "loss": 1.8637,
+      "step": 14682
+    },
+    {
+      "epoch": 0.6211608427108892,
+      "grad_norm": 0.1784118413925171,
+      "learning_rate": 0.001,
+      "loss": 2.5515,
+      "step": 14683
+    },
+    {
+      "epoch": 0.6212031474744056,
+      "grad_norm": 0.17213745415210724,
+      "learning_rate": 0.001,
+      "loss": 1.9208,
+      "step": 14684
+    },
+    {
+      "epoch": 0.6212454522379219,
+      "grad_norm": 0.16187311708927155,
+      "learning_rate": 0.001,
+      "loss": 2.6264,
+      "step": 14685
+    },
+    {
+      "epoch": 0.6212877570014383,
+      "grad_norm": 0.1690584123134613,
+      "learning_rate": 0.001,
+      "loss": 2.0176,
+      "step": 14686
+    },
+    {
+      "epoch": 0.6213300617649548,
+      "grad_norm": 0.22212959825992584,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 14687
+    },
+    {
+      "epoch": 0.621372366528471,
+      "grad_norm": 0.19889722764492035,
+      "learning_rate": 0.001,
+      "loss": 2.108,
+      "step": 14688
+    },
+    {
+      "epoch": 0.6214146712919875,
+      "grad_norm": 0.20588736236095428,
+      "learning_rate": 0.001,
+      "loss": 2.604,
+      "step": 14689
+    },
+    {
+      "epoch": 0.6214569760555039,
+      "grad_norm": 0.15710745751857758,
+      "learning_rate": 0.001,
+      "loss": 1.642,
+      "step": 14690
+    },
+    {
+      "epoch": 0.6214992808190202,
+      "grad_norm": 1.2365772724151611,
+      "learning_rate": 0.001,
+      "loss": 1.5983,
+      "step": 14691
+    },
+    {
+      "epoch": 0.6215415855825366,
+      "grad_norm": 0.1592063158750534,
+      "learning_rate": 0.001,
+      "loss": 1.7246,
+      "step": 14692
+    },
+    {
+      "epoch": 0.621583890346053,
+      "grad_norm": 0.2230633795261383,
+      "learning_rate": 0.001,
+      "loss": 2.5722,
+      "step": 14693
+    },
+    {
+      "epoch": 0.6216261951095693,
+      "grad_norm": 0.17436636984348297,
+      "learning_rate": 0.001,
+      "loss": 1.6434,
+      "step": 14694
+    },
+    {
+      "epoch": 0.6216684998730857,
+      "grad_norm": 0.1729450523853302,
+      "learning_rate": 0.001,
+      "loss": 2.0217,
+      "step": 14695
+    },
+    {
+      "epoch": 0.6217108046366021,
+      "grad_norm": 0.16045428812503815,
+      "learning_rate": 0.001,
+      "loss": 1.4028,
+      "step": 14696
+    },
+    {
+      "epoch": 0.6217531094001184,
+      "grad_norm": 0.1945531815290451,
+      "learning_rate": 0.001,
+      "loss": 2.3836,
+      "step": 14697
+    },
+    {
+      "epoch": 0.6217954141636348,
+      "grad_norm": 3.1854636669158936,
+      "learning_rate": 0.001,
+      "loss": 3.0657,
+      "step": 14698
+    },
+    {
+      "epoch": 0.6218377189271512,
+      "grad_norm": 0.15348710119724274,
+      "learning_rate": 0.001,
+      "loss": 2.0376,
+      "step": 14699
+    },
+    {
+      "epoch": 0.6218800236906675,
+      "grad_norm": 0.16753935813903809,
+      "learning_rate": 0.001,
+      "loss": 1.6844,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6219223284541839,
+      "grad_norm": 0.1476292461156845,
+      "learning_rate": 0.001,
+      "loss": 1.878,
+      "step": 14701
+    },
+    {
+      "epoch": 0.6219646332177003,
+      "grad_norm": 0.21797721087932587,
+      "learning_rate": 0.001,
+      "loss": 1.692,
+      "step": 14702
+    },
+    {
+      "epoch": 0.6220069379812166,
+      "grad_norm": 0.176117941737175,
+      "learning_rate": 0.001,
+      "loss": 2.2716,
+      "step": 14703
+    },
+    {
+      "epoch": 0.622049242744733,
+      "grad_norm": 4.9479660987854,
+      "learning_rate": 0.001,
+      "loss": 1.8148,
+      "step": 14704
+    },
+    {
+      "epoch": 0.6220915475082495,
+      "grad_norm": 0.16278578341007233,
+      "learning_rate": 0.001,
+      "loss": 1.9686,
+      "step": 14705
+    },
+    {
+      "epoch": 0.6221338522717658,
+      "grad_norm": 0.19134145975112915,
+      "learning_rate": 0.001,
+      "loss": 2.3014,
+      "step": 14706
+    },
+    {
+      "epoch": 0.6221761570352822,
+      "grad_norm": 0.1470852494239807,
+      "learning_rate": 0.001,
+      "loss": 2.4502,
+      "step": 14707
+    },
+    {
+      "epoch": 0.6222184617987986,
+      "grad_norm": 0.3955981731414795,
+      "learning_rate": 0.001,
+      "loss": 2.0912,
+      "step": 14708
+    },
+    {
+      "epoch": 0.6222607665623149,
+      "grad_norm": 0.2723364233970642,
+      "learning_rate": 0.001,
+      "loss": 1.8479,
+      "step": 14709
+    },
+    {
+      "epoch": 0.6223030713258313,
+      "grad_norm": 0.15754322707653046,
+      "learning_rate": 0.001,
+      "loss": 2.4443,
+      "step": 14710
+    },
+    {
+      "epoch": 0.6223453760893477,
+      "grad_norm": 0.21072664856910706,
+      "learning_rate": 0.001,
+      "loss": 1.9223,
+      "step": 14711
+    },
+    {
+      "epoch": 0.622387680852864,
+      "grad_norm": 0.17302967607975006,
+      "learning_rate": 0.001,
+      "loss": 2.3223,
+      "step": 14712
+    },
+    {
+      "epoch": 0.6224299856163804,
+      "grad_norm": 0.17941580712795258,
+      "learning_rate": 0.001,
+      "loss": 1.7731,
+      "step": 14713
+    },
+    {
+      "epoch": 0.6224722903798968,
+      "grad_norm": 0.325706422328949,
+      "learning_rate": 0.001,
+      "loss": 2.6128,
+      "step": 14714
+    },
+    {
+      "epoch": 0.6225145951434131,
+      "grad_norm": 1.440018653869629,
+      "learning_rate": 0.001,
+      "loss": 1.9563,
+      "step": 14715
+    },
+    {
+      "epoch": 0.6225568999069295,
+      "grad_norm": 0.1713675558567047,
+      "learning_rate": 0.001,
+      "loss": 2.3617,
+      "step": 14716
+    },
+    {
+      "epoch": 0.6225992046704459,
+      "grad_norm": 0.19308139383792877,
+      "learning_rate": 0.001,
+      "loss": 1.6729,
+      "step": 14717
+    },
+    {
+      "epoch": 0.6226415094339622,
+      "grad_norm": 0.17297999560832977,
+      "learning_rate": 0.001,
+      "loss": 3.6168,
+      "step": 14718
+    },
+    {
+      "epoch": 0.6226838141974786,
+      "grad_norm": 0.27014705538749695,
+      "learning_rate": 0.001,
+      "loss": 3.2252,
+      "step": 14719
+    },
+    {
+      "epoch": 0.6227261189609951,
+      "grad_norm": 0.15865100920200348,
+      "learning_rate": 0.001,
+      "loss": 1.4966,
+      "step": 14720
+    },
+    {
+      "epoch": 0.6227684237245114,
+      "grad_norm": 0.14343230426311493,
+      "learning_rate": 0.001,
+      "loss": 1.711,
+      "step": 14721
+    },
+    {
+      "epoch": 0.6228107284880278,
+      "grad_norm": 0.1730983704328537,
+      "learning_rate": 0.001,
+      "loss": 2.159,
+      "step": 14722
+    },
+    {
+      "epoch": 0.6228530332515442,
+      "grad_norm": 0.20309610664844513,
+      "learning_rate": 0.001,
+      "loss": 2.1347,
+      "step": 14723
+    },
+    {
+      "epoch": 0.6228953380150605,
+      "grad_norm": 0.9017519950866699,
+      "learning_rate": 0.001,
+      "loss": 1.8289,
+      "step": 14724
+    },
+    {
+      "epoch": 0.6229376427785769,
+      "grad_norm": 0.17153169214725494,
+      "learning_rate": 0.001,
+      "loss": 2.527,
+      "step": 14725
+    },
+    {
+      "epoch": 0.6229799475420933,
+      "grad_norm": 0.1731494665145874,
+      "learning_rate": 0.001,
+      "loss": 1.3847,
+      "step": 14726
+    },
+    {
+      "epoch": 0.6230222523056096,
+      "grad_norm": 0.20264741778373718,
+      "learning_rate": 0.001,
+      "loss": 2.6843,
+      "step": 14727
+    },
+    {
+      "epoch": 0.623064557069126,
+      "grad_norm": 0.1755369007587433,
+      "learning_rate": 0.001,
+      "loss": 1.8534,
+      "step": 14728
+    },
+    {
+      "epoch": 0.6231068618326423,
+      "grad_norm": 0.1698158234357834,
+      "learning_rate": 0.001,
+      "loss": 2.4937,
+      "step": 14729
+    },
+    {
+      "epoch": 0.6231491665961587,
+      "grad_norm": 0.3321411907672882,
+      "learning_rate": 0.001,
+      "loss": 1.902,
+      "step": 14730
+    },
+    {
+      "epoch": 0.6231914713596751,
+      "grad_norm": 0.1981087177991867,
+      "learning_rate": 0.001,
+      "loss": 2.086,
+      "step": 14731
+    },
+    {
+      "epoch": 0.6232337761231914,
+      "grad_norm": 0.14606229960918427,
+      "learning_rate": 0.001,
+      "loss": 1.7585,
+      "step": 14732
+    },
+    {
+      "epoch": 0.6232760808867078,
+      "grad_norm": 0.15275610983371735,
+      "learning_rate": 0.001,
+      "loss": 2.0629,
+      "step": 14733
+    },
+    {
+      "epoch": 0.6233183856502242,
+      "grad_norm": 0.3921429514884949,
+      "learning_rate": 0.001,
+      "loss": 2.0281,
+      "step": 14734
+    },
+    {
+      "epoch": 0.6233606904137405,
+      "grad_norm": 0.15665315091609955,
+      "learning_rate": 0.001,
+      "loss": 1.9623,
+      "step": 14735
+    },
+    {
+      "epoch": 0.623402995177257,
+      "grad_norm": 0.1723744422197342,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 14736
+    },
+    {
+      "epoch": 0.6234452999407734,
+      "grad_norm": 0.16472692787647247,
+      "learning_rate": 0.001,
+      "loss": 2.3605,
+      "step": 14737
+    },
+    {
+      "epoch": 0.6234876047042897,
+      "grad_norm": 0.22719836235046387,
+      "learning_rate": 0.001,
+      "loss": 2.9966,
+      "step": 14738
+    },
+    {
+      "epoch": 0.6235299094678061,
+      "grad_norm": 0.5726194381713867,
+      "learning_rate": 0.001,
+      "loss": 2.0207,
+      "step": 14739
+    },
+    {
+      "epoch": 0.6235722142313225,
+      "grad_norm": 0.16665321588516235,
+      "learning_rate": 0.001,
+      "loss": 2.1726,
+      "step": 14740
+    },
+    {
+      "epoch": 0.6236145189948388,
+      "grad_norm": 0.15230652689933777,
+      "learning_rate": 0.001,
+      "loss": 2.2088,
+      "step": 14741
+    },
+    {
+      "epoch": 0.6236568237583552,
+      "grad_norm": 0.17324359714984894,
+      "learning_rate": 0.001,
+      "loss": 2.0516,
+      "step": 14742
+    },
+    {
+      "epoch": 0.6236991285218716,
+      "grad_norm": 0.24778017401695251,
+      "learning_rate": 0.001,
+      "loss": 1.798,
+      "step": 14743
+    },
+    {
+      "epoch": 0.6237414332853879,
+      "grad_norm": 0.18029212951660156,
+      "learning_rate": 0.001,
+      "loss": 1.5368,
+      "step": 14744
+    },
+    {
+      "epoch": 0.6237837380489043,
+      "grad_norm": 0.17106720805168152,
+      "learning_rate": 0.001,
+      "loss": 2.5262,
+      "step": 14745
+    },
+    {
+      "epoch": 0.6238260428124207,
+      "grad_norm": 0.14900583028793335,
+      "learning_rate": 0.001,
+      "loss": 1.2427,
+      "step": 14746
+    },
+    {
+      "epoch": 0.623868347575937,
+      "grad_norm": 0.2502407133579254,
+      "learning_rate": 0.001,
+      "loss": 3.2134,
+      "step": 14747
+    },
+    {
+      "epoch": 0.6239106523394534,
+      "grad_norm": 0.21255750954151154,
+      "learning_rate": 0.001,
+      "loss": 2.0656,
+      "step": 14748
+    },
+    {
+      "epoch": 0.6239529571029698,
+      "grad_norm": 0.15253281593322754,
+      "learning_rate": 0.001,
+      "loss": 3.1515,
+      "step": 14749
+    },
+    {
+      "epoch": 0.6239952618664861,
+      "grad_norm": 0.15575817227363586,
+      "learning_rate": 0.001,
+      "loss": 2.1155,
+      "step": 14750
+    },
+    {
+      "epoch": 0.6240375666300025,
+      "grad_norm": 0.26133993268013,
+      "learning_rate": 0.001,
+      "loss": 2.7806,
+      "step": 14751
+    },
+    {
+      "epoch": 0.624079871393519,
+      "grad_norm": 0.15048684179782867,
+      "learning_rate": 0.001,
+      "loss": 1.8556,
+      "step": 14752
+    },
+    {
+      "epoch": 0.6241221761570352,
+      "grad_norm": 0.2583177387714386,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 14753
+    },
+    {
+      "epoch": 0.6241644809205517,
+      "grad_norm": 0.15635612607002258,
+      "learning_rate": 0.001,
+      "loss": 1.8436,
+      "step": 14754
+    },
+    {
+      "epoch": 0.6242067856840681,
+      "grad_norm": 0.3417166769504547,
+      "learning_rate": 0.001,
+      "loss": 2.3692,
+      "step": 14755
+    },
+    {
+      "epoch": 0.6242490904475844,
+      "grad_norm": 0.29285457730293274,
+      "learning_rate": 0.001,
+      "loss": 2.2755,
+      "step": 14756
+    },
+    {
+      "epoch": 0.6242913952111008,
+      "grad_norm": 0.2324390560388565,
+      "learning_rate": 0.001,
+      "loss": 2.2066,
+      "step": 14757
+    },
+    {
+      "epoch": 0.6243336999746172,
+      "grad_norm": 0.15262608230113983,
+      "learning_rate": 0.001,
+      "loss": 1.5537,
+      "step": 14758
+    },
+    {
+      "epoch": 0.6243760047381335,
+      "grad_norm": 0.1738661378622055,
+      "learning_rate": 0.001,
+      "loss": 2.2254,
+      "step": 14759
+    },
+    {
+      "epoch": 0.6244183095016499,
+      "grad_norm": 0.17284923791885376,
+      "learning_rate": 0.001,
+      "loss": 2.619,
+      "step": 14760
+    },
+    {
+      "epoch": 0.6244606142651663,
+      "grad_norm": 0.16344395279884338,
+      "learning_rate": 0.001,
+      "loss": 1.8052,
+      "step": 14761
+    },
+    {
+      "epoch": 0.6245029190286826,
+      "grad_norm": 1.0341507196426392,
+      "learning_rate": 0.001,
+      "loss": 2.4716,
+      "step": 14762
+    },
+    {
+      "epoch": 0.624545223792199,
+      "grad_norm": 0.17319472134113312,
+      "learning_rate": 0.001,
+      "loss": 1.7911,
+      "step": 14763
+    },
+    {
+      "epoch": 0.6245875285557154,
+      "grad_norm": 0.1723223626613617,
+      "learning_rate": 0.001,
+      "loss": 2.2801,
+      "step": 14764
+    },
+    {
+      "epoch": 0.6246298333192317,
+      "grad_norm": 0.16963432729244232,
+      "learning_rate": 0.001,
+      "loss": 1.8165,
+      "step": 14765
+    },
+    {
+      "epoch": 0.6246721380827481,
+      "grad_norm": 0.1635103076696396,
+      "learning_rate": 0.001,
+      "loss": 2.3229,
+      "step": 14766
+    },
+    {
+      "epoch": 0.6247144428462645,
+      "grad_norm": 0.14711347222328186,
+      "learning_rate": 0.001,
+      "loss": 2.0542,
+      "step": 14767
+    },
+    {
+      "epoch": 0.6247567476097808,
+      "grad_norm": 0.15598388016223907,
+      "learning_rate": 0.001,
+      "loss": 3.1843,
+      "step": 14768
+    },
+    {
+      "epoch": 0.6247990523732972,
+      "grad_norm": 0.15948522090911865,
+      "learning_rate": 0.001,
+      "loss": 3.8822,
+      "step": 14769
+    },
+    {
+      "epoch": 0.6248413571368137,
+      "grad_norm": 0.1707606315612793,
+      "learning_rate": 0.001,
+      "loss": 1.5173,
+      "step": 14770
+    },
+    {
+      "epoch": 0.62488366190033,
+      "grad_norm": 0.20455005764961243,
+      "learning_rate": 0.001,
+      "loss": 2.8538,
+      "step": 14771
+    },
+    {
+      "epoch": 0.6249259666638464,
+      "grad_norm": 0.18177711963653564,
+      "learning_rate": 0.001,
+      "loss": 1.576,
+      "step": 14772
+    },
+    {
+      "epoch": 0.6249682714273628,
+      "grad_norm": 0.6637758016586304,
+      "learning_rate": 0.001,
+      "loss": 3.9036,
+      "step": 14773
+    },
+    {
+      "epoch": 0.6250105761908791,
+      "grad_norm": 0.14505170285701752,
+      "learning_rate": 0.001,
+      "loss": 1.3907,
+      "step": 14774
+    },
+    {
+      "epoch": 0.6250528809543955,
+      "grad_norm": 0.1670999825000763,
+      "learning_rate": 0.001,
+      "loss": 2.2654,
+      "step": 14775
+    },
+    {
+      "epoch": 0.6250951857179118,
+      "grad_norm": 0.15915502607822418,
+      "learning_rate": 0.001,
+      "loss": 1.5619,
+      "step": 14776
+    },
+    {
+      "epoch": 0.6251374904814282,
+      "grad_norm": 0.24664057791233063,
+      "learning_rate": 0.001,
+      "loss": 2.0549,
+      "step": 14777
+    },
+    {
+      "epoch": 0.6251797952449446,
+      "grad_norm": 0.19590286910533905,
+      "learning_rate": 0.001,
+      "loss": 1.9566,
+      "step": 14778
+    },
+    {
+      "epoch": 0.6252221000084609,
+      "grad_norm": 0.1494143158197403,
+      "learning_rate": 0.001,
+      "loss": 2.6541,
+      "step": 14779
+    },
+    {
+      "epoch": 0.6252644047719773,
+      "grad_norm": 0.1772175133228302,
+      "learning_rate": 0.001,
+      "loss": 2.077,
+      "step": 14780
+    },
+    {
+      "epoch": 0.6253067095354937,
+      "grad_norm": 0.16731050610542297,
+      "learning_rate": 0.001,
+      "loss": 1.8391,
+      "step": 14781
+    },
+    {
+      "epoch": 0.62534901429901,
+      "grad_norm": 0.18583473563194275,
+      "learning_rate": 0.001,
+      "loss": 2.561,
+      "step": 14782
+    },
+    {
+      "epoch": 0.6253913190625264,
+      "grad_norm": 0.15559391677379608,
+      "learning_rate": 0.001,
+      "loss": 1.7544,
+      "step": 14783
+    },
+    {
+      "epoch": 0.6254336238260428,
+      "grad_norm": 0.16811911761760712,
+      "learning_rate": 0.001,
+      "loss": 2.3595,
+      "step": 14784
+    },
+    {
+      "epoch": 0.6254759285895591,
+      "grad_norm": 0.18134218454360962,
+      "learning_rate": 0.001,
+      "loss": 1.9538,
+      "step": 14785
+    },
+    {
+      "epoch": 0.6255182333530755,
+      "grad_norm": 0.8683392405509949,
+      "learning_rate": 0.001,
+      "loss": 2.0525,
+      "step": 14786
+    },
+    {
+      "epoch": 0.625560538116592,
+      "grad_norm": 0.17296668887138367,
+      "learning_rate": 0.001,
+      "loss": 2.1946,
+      "step": 14787
+    },
+    {
+      "epoch": 0.6256028428801083,
+      "grad_norm": 0.24003466963768005,
+      "learning_rate": 0.001,
+      "loss": 2.5681,
+      "step": 14788
+    },
+    {
+      "epoch": 0.6256451476436247,
+      "grad_norm": 0.1762605756521225,
+      "learning_rate": 0.001,
+      "loss": 2.2416,
+      "step": 14789
+    },
+    {
+      "epoch": 0.6256874524071411,
+      "grad_norm": 0.262921541929245,
+      "learning_rate": 0.001,
+      "loss": 3.2333,
+      "step": 14790
+    },
+    {
+      "epoch": 0.6257297571706574,
+      "grad_norm": 0.2003992646932602,
+      "learning_rate": 0.001,
+      "loss": 2.1344,
+      "step": 14791
+    },
+    {
+      "epoch": 0.6257720619341738,
+      "grad_norm": 0.1967916488647461,
+      "learning_rate": 0.001,
+      "loss": 1.5399,
+      "step": 14792
+    },
+    {
+      "epoch": 0.6258143666976902,
+      "grad_norm": 0.18930691480636597,
+      "learning_rate": 0.001,
+      "loss": 1.3929,
+      "step": 14793
+    },
+    {
+      "epoch": 0.6258566714612065,
+      "grad_norm": 0.6461501121520996,
+      "learning_rate": 0.001,
+      "loss": 1.8684,
+      "step": 14794
+    },
+    {
+      "epoch": 0.6258989762247229,
+      "grad_norm": 0.16277913749217987,
+      "learning_rate": 0.001,
+      "loss": 2.1545,
+      "step": 14795
+    },
+    {
+      "epoch": 0.6259412809882393,
+      "grad_norm": 0.13681641221046448,
+      "learning_rate": 0.001,
+      "loss": 1.4387,
+      "step": 14796
+    },
+    {
+      "epoch": 0.6259835857517556,
+      "grad_norm": 0.49204564094543457,
+      "learning_rate": 0.001,
+      "loss": 1.716,
+      "step": 14797
+    },
+    {
+      "epoch": 0.626025890515272,
+      "grad_norm": 0.2248888909816742,
+      "learning_rate": 0.001,
+      "loss": 2.0344,
+      "step": 14798
+    },
+    {
+      "epoch": 0.6260681952787884,
+      "grad_norm": 0.1486063003540039,
+      "learning_rate": 0.001,
+      "loss": 1.7528,
+      "step": 14799
+    },
+    {
+      "epoch": 0.6261105000423047,
+      "grad_norm": 0.16557568311691284,
+      "learning_rate": 0.001,
+      "loss": 1.7937,
+      "step": 14800
+    },
+    {
+      "epoch": 0.6261528048058211,
+      "grad_norm": 0.2470521777868271,
+      "learning_rate": 0.001,
+      "loss": 1.88,
+      "step": 14801
+    },
+    {
+      "epoch": 0.6261951095693375,
+      "grad_norm": 1.031522274017334,
+      "learning_rate": 0.001,
+      "loss": 2.0462,
+      "step": 14802
+    },
+    {
+      "epoch": 0.6262374143328538,
+      "grad_norm": 0.39598196744918823,
+      "learning_rate": 0.001,
+      "loss": 1.4901,
+      "step": 14803
+    },
+    {
+      "epoch": 0.6262797190963703,
+      "grad_norm": 0.1830819547176361,
+      "learning_rate": 0.001,
+      "loss": 1.7788,
+      "step": 14804
+    },
+    {
+      "epoch": 0.6263220238598867,
+      "grad_norm": 0.18388625979423523,
+      "learning_rate": 0.001,
+      "loss": 2.308,
+      "step": 14805
+    },
+    {
+      "epoch": 0.626364328623403,
+      "grad_norm": 1.8769688606262207,
+      "learning_rate": 0.001,
+      "loss": 2.2326,
+      "step": 14806
+    },
+    {
+      "epoch": 0.6264066333869194,
+      "grad_norm": 0.1682252585887909,
+      "learning_rate": 0.001,
+      "loss": 1.7469,
+      "step": 14807
+    },
+    {
+      "epoch": 0.6264489381504358,
+      "grad_norm": 0.1951892077922821,
+      "learning_rate": 0.001,
+      "loss": 3.3156,
+      "step": 14808
+    },
+    {
+      "epoch": 0.6264912429139521,
+      "grad_norm": 0.20327278971672058,
+      "learning_rate": 0.001,
+      "loss": 2.7878,
+      "step": 14809
+    },
+    {
+      "epoch": 0.6265335476774685,
+      "grad_norm": 7.379302024841309,
+      "learning_rate": 0.001,
+      "loss": 2.2043,
+      "step": 14810
+    },
+    {
+      "epoch": 0.6265758524409849,
+      "grad_norm": 0.19853001832962036,
+      "learning_rate": 0.001,
+      "loss": 2.6433,
+      "step": 14811
+    },
+    {
+      "epoch": 0.6266181572045012,
+      "grad_norm": 0.4619123637676239,
+      "learning_rate": 0.001,
+      "loss": 1.9174,
+      "step": 14812
+    },
+    {
+      "epoch": 0.6266604619680176,
+      "grad_norm": 0.167628213763237,
+      "learning_rate": 0.001,
+      "loss": 1.3258,
+      "step": 14813
+    },
+    {
+      "epoch": 0.626702766731534,
+      "grad_norm": 1.1932356357574463,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 14814
+    },
+    {
+      "epoch": 0.6267450714950503,
+      "grad_norm": 0.21053379774093628,
+      "learning_rate": 0.001,
+      "loss": 3.2038,
+      "step": 14815
+    },
+    {
+      "epoch": 0.6267873762585667,
+      "grad_norm": 1.0147972106933594,
+      "learning_rate": 0.001,
+      "loss": 3.0992,
+      "step": 14816
+    },
+    {
+      "epoch": 0.6268296810220831,
+      "grad_norm": 0.2153492271900177,
+      "learning_rate": 0.001,
+      "loss": 2.7228,
+      "step": 14817
+    },
+    {
+      "epoch": 0.6268719857855994,
+      "grad_norm": 0.4768437147140503,
+      "learning_rate": 0.001,
+      "loss": 1.9515,
+      "step": 14818
+    },
+    {
+      "epoch": 0.6269142905491158,
+      "grad_norm": 0.6105790138244629,
+      "learning_rate": 0.001,
+      "loss": 2.5721,
+      "step": 14819
+    },
+    {
+      "epoch": 0.6269565953126321,
+      "grad_norm": 0.17983876168727875,
+      "learning_rate": 0.001,
+      "loss": 1.896,
+      "step": 14820
+    },
+    {
+      "epoch": 0.6269989000761486,
+      "grad_norm": 0.23579160869121552,
+      "learning_rate": 0.001,
+      "loss": 1.8257,
+      "step": 14821
+    },
+    {
+      "epoch": 0.627041204839665,
+      "grad_norm": 0.14642800390720367,
+      "learning_rate": 0.001,
+      "loss": 1.7842,
+      "step": 14822
+    },
+    {
+      "epoch": 0.6270835096031813,
+      "grad_norm": 0.23658250272274017,
+      "learning_rate": 0.001,
+      "loss": 2.5395,
+      "step": 14823
+    },
+    {
+      "epoch": 0.6271258143666977,
+      "grad_norm": 0.18231017887592316,
+      "learning_rate": 0.001,
+      "loss": 2.6851,
+      "step": 14824
+    },
+    {
+      "epoch": 0.6271681191302141,
+      "grad_norm": 0.22886481881141663,
+      "learning_rate": 0.001,
+      "loss": 2.5188,
+      "step": 14825
+    },
+    {
+      "epoch": 0.6272104238937304,
+      "grad_norm": 0.32207468152046204,
+      "learning_rate": 0.001,
+      "loss": 2.9632,
+      "step": 14826
+    },
+    {
+      "epoch": 0.6272527286572468,
+      "grad_norm": 0.2477169632911682,
+      "learning_rate": 0.001,
+      "loss": 2.8172,
+      "step": 14827
+    },
+    {
+      "epoch": 0.6272950334207632,
+      "grad_norm": 0.2553689479827881,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 14828
+    },
+    {
+      "epoch": 0.6273373381842795,
+      "grad_norm": 0.16400131583213806,
+      "learning_rate": 0.001,
+      "loss": 2.1881,
+      "step": 14829
+    },
+    {
+      "epoch": 0.6273796429477959,
+      "grad_norm": 0.47019270062446594,
+      "learning_rate": 0.001,
+      "loss": 2.8494,
+      "step": 14830
+    },
+    {
+      "epoch": 0.6274219477113123,
+      "grad_norm": 0.3059365153312683,
+      "learning_rate": 0.001,
+      "loss": 2.2702,
+      "step": 14831
+    },
+    {
+      "epoch": 0.6274642524748286,
+      "grad_norm": 0.21573302149772644,
+      "learning_rate": 0.001,
+      "loss": 2.0846,
+      "step": 14832
+    },
+    {
+      "epoch": 0.627506557238345,
+      "grad_norm": 0.168625608086586,
+      "learning_rate": 0.001,
+      "loss": 1.7721,
+      "step": 14833
+    },
+    {
+      "epoch": 0.6275488620018614,
+      "grad_norm": 0.15803949534893036,
+      "learning_rate": 0.001,
+      "loss": 1.42,
+      "step": 14834
+    },
+    {
+      "epoch": 0.6275911667653777,
+      "grad_norm": 0.22716844081878662,
+      "learning_rate": 0.001,
+      "loss": 1.655,
+      "step": 14835
+    },
+    {
+      "epoch": 0.6276334715288941,
+      "grad_norm": 0.2481343150138855,
+      "learning_rate": 0.001,
+      "loss": 2.1238,
+      "step": 14836
+    },
+    {
+      "epoch": 0.6276757762924106,
+      "grad_norm": 0.21885570883750916,
+      "learning_rate": 0.001,
+      "loss": 2.5259,
+      "step": 14837
+    },
+    {
+      "epoch": 0.6277180810559269,
+      "grad_norm": 0.22164210677146912,
+      "learning_rate": 0.001,
+      "loss": 2.5285,
+      "step": 14838
+    },
+    {
+      "epoch": 0.6277603858194433,
+      "grad_norm": 0.17463375627994537,
+      "learning_rate": 0.001,
+      "loss": 2.1484,
+      "step": 14839
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.38111189007759094,
+      "learning_rate": 0.001,
+      "loss": 3.1991,
+      "step": 14840
+    },
+    {
+      "epoch": 0.627844995346476,
+      "grad_norm": 0.2066563367843628,
+      "learning_rate": 0.001,
+      "loss": 1.804,
+      "step": 14841
+    },
+    {
+      "epoch": 0.6278873001099924,
+      "grad_norm": 0.23109540343284607,
+      "learning_rate": 0.001,
+      "loss": 1.7063,
+      "step": 14842
+    },
+    {
+      "epoch": 0.6279296048735088,
+      "grad_norm": 0.1922856718301773,
+      "learning_rate": 0.001,
+      "loss": 1.6542,
+      "step": 14843
+    },
+    {
+      "epoch": 0.6279719096370251,
+      "grad_norm": 0.1745460331439972,
+      "learning_rate": 0.001,
+      "loss": 2.2544,
+      "step": 14844
+    },
+    {
+      "epoch": 0.6280142144005415,
+      "grad_norm": 0.16005033254623413,
+      "learning_rate": 0.001,
+      "loss": 1.6839,
+      "step": 14845
+    },
+    {
+      "epoch": 0.6280565191640579,
+      "grad_norm": 0.14515602588653564,
+      "learning_rate": 0.001,
+      "loss": 2.7744,
+      "step": 14846
+    },
+    {
+      "epoch": 0.6280988239275742,
+      "grad_norm": 0.1645001471042633,
+      "learning_rate": 0.001,
+      "loss": 1.6844,
+      "step": 14847
+    },
+    {
+      "epoch": 0.6281411286910906,
+      "grad_norm": 0.19100764393806458,
+      "learning_rate": 0.001,
+      "loss": 2.6849,
+      "step": 14848
+    },
+    {
+      "epoch": 0.628183433454607,
+      "grad_norm": 0.1547725945711136,
+      "learning_rate": 0.001,
+      "loss": 1.6892,
+      "step": 14849
+    },
+    {
+      "epoch": 0.6282257382181233,
+      "grad_norm": 0.15877199172973633,
+      "learning_rate": 0.001,
+      "loss": 2.3236,
+      "step": 14850
+    },
+    {
+      "epoch": 0.6282680429816397,
+      "grad_norm": 0.858564555644989,
+      "learning_rate": 0.001,
+      "loss": 2.6084,
+      "step": 14851
+    },
+    {
+      "epoch": 0.6283103477451562,
+      "grad_norm": 0.15130288898944855,
+      "learning_rate": 0.001,
+      "loss": 2.6856,
+      "step": 14852
+    },
+    {
+      "epoch": 0.6283526525086724,
+      "grad_norm": 0.18292021751403809,
+      "learning_rate": 0.001,
+      "loss": 1.7743,
+      "step": 14853
+    },
+    {
+      "epoch": 0.6283949572721889,
+      "grad_norm": 0.17332005500793457,
+      "learning_rate": 0.001,
+      "loss": 2.0246,
+      "step": 14854
+    },
+    {
+      "epoch": 0.6284372620357053,
+      "grad_norm": 0.2719305455684662,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 14855
+    },
+    {
+      "epoch": 0.6284795667992216,
+      "grad_norm": 0.17727939784526825,
+      "learning_rate": 0.001,
+      "loss": 2.3099,
+      "step": 14856
+    },
+    {
+      "epoch": 0.628521871562738,
+      "grad_norm": 0.21192631125450134,
+      "learning_rate": 0.001,
+      "loss": 2.0069,
+      "step": 14857
+    },
+    {
+      "epoch": 0.6285641763262544,
+      "grad_norm": 0.16656994819641113,
+      "learning_rate": 0.001,
+      "loss": 2.1416,
+      "step": 14858
+    },
+    {
+      "epoch": 0.6286064810897707,
+      "grad_norm": 0.14432191848754883,
+      "learning_rate": 0.001,
+      "loss": 1.9373,
+      "step": 14859
+    },
+    {
+      "epoch": 0.6286487858532871,
+      "grad_norm": 0.16526567935943604,
+      "learning_rate": 0.001,
+      "loss": 2.9011,
+      "step": 14860
+    },
+    {
+      "epoch": 0.6286910906168035,
+      "grad_norm": 0.35414940118789673,
+      "learning_rate": 0.001,
+      "loss": 1.6422,
+      "step": 14861
+    },
+    {
+      "epoch": 0.6287333953803198,
+      "grad_norm": 1.0360305309295654,
+      "learning_rate": 0.001,
+      "loss": 1.7833,
+      "step": 14862
+    },
+    {
+      "epoch": 0.6287757001438362,
+      "grad_norm": 0.18930722773075104,
+      "learning_rate": 0.001,
+      "loss": 1.7108,
+      "step": 14863
+    },
+    {
+      "epoch": 0.6288180049073525,
+      "grad_norm": 0.23029504716396332,
+      "learning_rate": 0.001,
+      "loss": 2.3319,
+      "step": 14864
+    },
+    {
+      "epoch": 0.6288603096708689,
+      "grad_norm": 0.1663587987422943,
+      "learning_rate": 0.001,
+      "loss": 2.4982,
+      "step": 14865
+    },
+    {
+      "epoch": 0.6289026144343853,
+      "grad_norm": 0.14894288778305054,
+      "learning_rate": 0.001,
+      "loss": 2.4744,
+      "step": 14866
+    },
+    {
+      "epoch": 0.6289449191979016,
+      "grad_norm": 0.1686108112335205,
+      "learning_rate": 0.001,
+      "loss": 2.1875,
+      "step": 14867
+    },
+    {
+      "epoch": 0.628987223961418,
+      "grad_norm": 0.28346070647239685,
+      "learning_rate": 0.001,
+      "loss": 1.7888,
+      "step": 14868
+    },
+    {
+      "epoch": 0.6290295287249345,
+      "grad_norm": 0.2675085961818695,
+      "learning_rate": 0.001,
+      "loss": 2.2643,
+      "step": 14869
+    },
+    {
+      "epoch": 0.6290718334884507,
+      "grad_norm": 0.5835021734237671,
+      "learning_rate": 0.001,
+      "loss": 2.4189,
+      "step": 14870
+    },
+    {
+      "epoch": 0.6291141382519672,
+      "grad_norm": 0.1684013456106186,
+      "learning_rate": 0.001,
+      "loss": 2.6277,
+      "step": 14871
+    },
+    {
+      "epoch": 0.6291564430154836,
+      "grad_norm": 0.26715943217277527,
+      "learning_rate": 0.001,
+      "loss": 2.1086,
+      "step": 14872
+    },
+    {
+      "epoch": 0.6291987477789999,
+      "grad_norm": 0.1863265186548233,
+      "learning_rate": 0.001,
+      "loss": 2.702,
+      "step": 14873
+    },
+    {
+      "epoch": 0.6292410525425163,
+      "grad_norm": 0.17459061741828918,
+      "learning_rate": 0.001,
+      "loss": 1.8087,
+      "step": 14874
+    },
+    {
+      "epoch": 0.6292833573060327,
+      "grad_norm": 0.19673362374305725,
+      "learning_rate": 0.001,
+      "loss": 3.9467,
+      "step": 14875
+    },
+    {
+      "epoch": 0.629325662069549,
+      "grad_norm": 0.19001983106136322,
+      "learning_rate": 0.001,
+      "loss": 1.4823,
+      "step": 14876
+    },
+    {
+      "epoch": 0.6293679668330654,
+      "grad_norm": 0.24293777346611023,
+      "learning_rate": 0.001,
+      "loss": 3.0979,
+      "step": 14877
+    },
+    {
+      "epoch": 0.6294102715965818,
+      "grad_norm": 0.2093208283185959,
+      "learning_rate": 0.001,
+      "loss": 2.5133,
+      "step": 14878
+    },
+    {
+      "epoch": 0.6294525763600981,
+      "grad_norm": 0.1860685646533966,
+      "learning_rate": 0.001,
+      "loss": 1.69,
+      "step": 14879
+    },
+    {
+      "epoch": 0.6294948811236145,
+      "grad_norm": 2.451101303100586,
+      "learning_rate": 0.001,
+      "loss": 1.933,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6295371858871309,
+      "grad_norm": 0.16000764071941376,
+      "learning_rate": 0.001,
+      "loss": 1.6052,
+      "step": 14881
+    },
+    {
+      "epoch": 0.6295794906506472,
+      "grad_norm": 0.17317312955856323,
+      "learning_rate": 0.001,
+      "loss": 1.7292,
+      "step": 14882
+    },
+    {
+      "epoch": 0.6296217954141636,
+      "grad_norm": 0.15874236822128296,
+      "learning_rate": 0.001,
+      "loss": 2.214,
+      "step": 14883
+    },
+    {
+      "epoch": 0.62966410017768,
+      "grad_norm": 0.16798686981201172,
+      "learning_rate": 0.001,
+      "loss": 1.7421,
+      "step": 14884
+    },
+    {
+      "epoch": 0.6297064049411963,
+      "grad_norm": 8.583555221557617,
+      "learning_rate": 0.001,
+      "loss": 2.7575,
+      "step": 14885
+    },
+    {
+      "epoch": 0.6297487097047128,
+      "grad_norm": 0.20271947979927063,
+      "learning_rate": 0.001,
+      "loss": 2.0891,
+      "step": 14886
+    },
+    {
+      "epoch": 0.6297910144682292,
+      "grad_norm": 0.19663158059120178,
+      "learning_rate": 0.001,
+      "loss": 2.0016,
+      "step": 14887
+    },
+    {
+      "epoch": 0.6298333192317455,
+      "grad_norm": 0.15687520802021027,
+      "learning_rate": 0.001,
+      "loss": 1.4679,
+      "step": 14888
+    },
+    {
+      "epoch": 0.6298756239952619,
+      "grad_norm": 0.6638527512550354,
+      "learning_rate": 0.001,
+      "loss": 2.6785,
+      "step": 14889
+    },
+    {
+      "epoch": 0.6299179287587783,
+      "grad_norm": 0.1655256301164627,
+      "learning_rate": 0.001,
+      "loss": 1.9314,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6299602335222946,
+      "grad_norm": 0.21192145347595215,
+      "learning_rate": 0.001,
+      "loss": 2.146,
+      "step": 14891
+    },
+    {
+      "epoch": 0.630002538285811,
+      "grad_norm": 0.1955963522195816,
+      "learning_rate": 0.001,
+      "loss": 1.6481,
+      "step": 14892
+    },
+    {
+      "epoch": 0.6300448430493274,
+      "grad_norm": 30.6153621673584,
+      "learning_rate": 0.001,
+      "loss": 3.6619,
+      "step": 14893
+    },
+    {
+      "epoch": 0.6300871478128437,
+      "grad_norm": 0.2149381786584854,
+      "learning_rate": 0.001,
+      "loss": 3.3041,
+      "step": 14894
+    },
+    {
+      "epoch": 0.6301294525763601,
+      "grad_norm": 0.19510112702846527,
+      "learning_rate": 0.001,
+      "loss": 1.7846,
+      "step": 14895
+    },
+    {
+      "epoch": 0.6301717573398765,
+      "grad_norm": 1.3764687776565552,
+      "learning_rate": 0.001,
+      "loss": 2.012,
+      "step": 14896
+    },
+    {
+      "epoch": 0.6302140621033928,
+      "grad_norm": 0.15694308280944824,
+      "learning_rate": 0.001,
+      "loss": 1.8974,
+      "step": 14897
+    },
+    {
+      "epoch": 0.6302563668669092,
+      "grad_norm": 0.2188582718372345,
+      "learning_rate": 0.001,
+      "loss": 3.2371,
+      "step": 14898
+    },
+    {
+      "epoch": 0.6302986716304256,
+      "grad_norm": 0.1506149172782898,
+      "learning_rate": 0.001,
+      "loss": 1.5915,
+      "step": 14899
+    },
+    {
+      "epoch": 0.6303409763939419,
+      "grad_norm": 0.3987337648868561,
+      "learning_rate": 0.001,
+      "loss": 2.1704,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6303832811574583,
+      "grad_norm": 0.19197428226470947,
+      "learning_rate": 0.001,
+      "loss": 2.7042,
+      "step": 14901
+    },
+    {
+      "epoch": 0.6304255859209748,
+      "grad_norm": 0.16541145741939545,
+      "learning_rate": 0.001,
+      "loss": 1.7064,
+      "step": 14902
+    },
+    {
+      "epoch": 0.630467890684491,
+      "grad_norm": 0.3559930622577667,
+      "learning_rate": 0.001,
+      "loss": 1.8449,
+      "step": 14903
+    },
+    {
+      "epoch": 0.6305101954480075,
+      "grad_norm": 0.1497167944908142,
+      "learning_rate": 0.001,
+      "loss": 2.1508,
+      "step": 14904
+    },
+    {
+      "epoch": 0.6305525002115239,
+      "grad_norm": 3.9234423637390137,
+      "learning_rate": 0.001,
+      "loss": 2.4774,
+      "step": 14905
+    },
+    {
+      "epoch": 0.6305948049750402,
+      "grad_norm": 1.5417382717132568,
+      "learning_rate": 0.001,
+      "loss": 2.181,
+      "step": 14906
+    },
+    {
+      "epoch": 0.6306371097385566,
+      "grad_norm": 0.2059686779975891,
+      "learning_rate": 0.001,
+      "loss": 1.5838,
+      "step": 14907
+    },
+    {
+      "epoch": 0.630679414502073,
+      "grad_norm": 0.17090870440006256,
+      "learning_rate": 0.001,
+      "loss": 2.7388,
+      "step": 14908
+    },
+    {
+      "epoch": 0.6307217192655893,
+      "grad_norm": 0.1849461793899536,
+      "learning_rate": 0.001,
+      "loss": 1.9216,
+      "step": 14909
+    },
+    {
+      "epoch": 0.6307640240291057,
+      "grad_norm": 0.22465984523296356,
+      "learning_rate": 0.001,
+      "loss": 2.0343,
+      "step": 14910
+    },
+    {
+      "epoch": 0.630806328792622,
+      "grad_norm": 0.6882874369621277,
+      "learning_rate": 0.001,
+      "loss": 2.4926,
+      "step": 14911
+    },
+    {
+      "epoch": 0.6308486335561384,
+      "grad_norm": 0.43035462498664856,
+      "learning_rate": 0.001,
+      "loss": 1.9783,
+      "step": 14912
+    },
+    {
+      "epoch": 0.6308909383196548,
+      "grad_norm": 0.35573610663414,
+      "learning_rate": 0.001,
+      "loss": 2.3259,
+      "step": 14913
+    },
+    {
+      "epoch": 0.6309332430831711,
+      "grad_norm": 0.18853552639484406,
+      "learning_rate": 0.001,
+      "loss": 2.3126,
+      "step": 14914
+    },
+    {
+      "epoch": 0.6309755478466875,
+      "grad_norm": 0.17522744834423065,
+      "learning_rate": 0.001,
+      "loss": 2.5944,
+      "step": 14915
+    },
+    {
+      "epoch": 0.6310178526102039,
+      "grad_norm": 0.5187592506408691,
+      "learning_rate": 0.001,
+      "loss": 2.4785,
+      "step": 14916
+    },
+    {
+      "epoch": 0.6310601573737202,
+      "grad_norm": 0.184511199593544,
+      "learning_rate": 0.001,
+      "loss": 1.9519,
+      "step": 14917
+    },
+    {
+      "epoch": 0.6311024621372366,
+      "grad_norm": 2.1163647174835205,
+      "learning_rate": 0.001,
+      "loss": 2.9251,
+      "step": 14918
+    },
+    {
+      "epoch": 0.631144766900753,
+      "grad_norm": 0.17871823906898499,
+      "learning_rate": 0.001,
+      "loss": 2.1804,
+      "step": 14919
+    },
+    {
+      "epoch": 0.6311870716642694,
+      "grad_norm": 0.49142828583717346,
+      "learning_rate": 0.001,
+      "loss": 2.1875,
+      "step": 14920
+    },
+    {
+      "epoch": 0.6312293764277858,
+      "grad_norm": 0.294746458530426,
+      "learning_rate": 0.001,
+      "loss": 2.2112,
+      "step": 14921
+    },
+    {
+      "epoch": 0.6312716811913022,
+      "grad_norm": 0.23569050431251526,
+      "learning_rate": 0.001,
+      "loss": 2.9051,
+      "step": 14922
+    },
+    {
+      "epoch": 0.6313139859548185,
+      "grad_norm": 0.20902414619922638,
+      "learning_rate": 0.001,
+      "loss": 1.7777,
+      "step": 14923
+    },
+    {
+      "epoch": 0.6313562907183349,
+      "grad_norm": 0.30001044273376465,
+      "learning_rate": 0.001,
+      "loss": 2.2292,
+      "step": 14924
+    },
+    {
+      "epoch": 0.6313985954818513,
+      "grad_norm": 0.20105008780956268,
+      "learning_rate": 0.001,
+      "loss": 2.2452,
+      "step": 14925
+    },
+    {
+      "epoch": 0.6314409002453676,
+      "grad_norm": 0.4256247282028198,
+      "learning_rate": 0.001,
+      "loss": 2.3232,
+      "step": 14926
+    },
+    {
+      "epoch": 0.631483205008884,
+      "grad_norm": 0.1731926053762436,
+      "learning_rate": 0.001,
+      "loss": 1.7504,
+      "step": 14927
+    },
+    {
+      "epoch": 0.6315255097724004,
+      "grad_norm": 0.16395363211631775,
+      "learning_rate": 0.001,
+      "loss": 1.5929,
+      "step": 14928
+    },
+    {
+      "epoch": 0.6315678145359167,
+      "grad_norm": 0.395588755607605,
+      "learning_rate": 0.001,
+      "loss": 2.7332,
+      "step": 14929
+    },
+    {
+      "epoch": 0.6316101192994331,
+      "grad_norm": 0.1726456880569458,
+      "learning_rate": 0.001,
+      "loss": 1.8693,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6316524240629495,
+      "grad_norm": 0.1603616327047348,
+      "learning_rate": 0.001,
+      "loss": 2.5375,
+      "step": 14931
+    },
+    {
+      "epoch": 0.6316947288264658,
+      "grad_norm": 0.21942928433418274,
+      "learning_rate": 0.001,
+      "loss": 2.6368,
+      "step": 14932
+    },
+    {
+      "epoch": 0.6317370335899822,
+      "grad_norm": 0.20922864973545074,
+      "learning_rate": 0.001,
+      "loss": 2.3417,
+      "step": 14933
+    },
+    {
+      "epoch": 0.6317793383534986,
+      "grad_norm": 0.142195925116539,
+      "learning_rate": 0.001,
+      "loss": 2.0823,
+      "step": 14934
+    },
+    {
+      "epoch": 0.6318216431170149,
+      "grad_norm": 0.4264514744281769,
+      "learning_rate": 0.001,
+      "loss": 2.3226,
+      "step": 14935
+    },
+    {
+      "epoch": 0.6318639478805314,
+      "grad_norm": 0.17566858232021332,
+      "learning_rate": 0.001,
+      "loss": 1.4364,
+      "step": 14936
+    },
+    {
+      "epoch": 0.6319062526440478,
+      "grad_norm": 0.5592827200889587,
+      "learning_rate": 0.001,
+      "loss": 2.4408,
+      "step": 14937
+    },
+    {
+      "epoch": 0.6319485574075641,
+      "grad_norm": 0.4113415479660034,
+      "learning_rate": 0.001,
+      "loss": 2.6319,
+      "step": 14938
+    },
+    {
+      "epoch": 0.6319908621710805,
+      "grad_norm": 0.1990566849708557,
+      "learning_rate": 0.001,
+      "loss": 2.4111,
+      "step": 14939
+    },
+    {
+      "epoch": 0.6320331669345969,
+      "grad_norm": 0.7872762084007263,
+      "learning_rate": 0.001,
+      "loss": 2.0307,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6320754716981132,
+      "grad_norm": 0.18138566613197327,
+      "learning_rate": 0.001,
+      "loss": 2.3526,
+      "step": 14941
+    },
+    {
+      "epoch": 0.6321177764616296,
+      "grad_norm": 0.1856798529624939,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 14942
+    },
+    {
+      "epoch": 0.632160081225146,
+      "grad_norm": 0.1658208668231964,
+      "learning_rate": 0.001,
+      "loss": 1.8302,
+      "step": 14943
+    },
+    {
+      "epoch": 0.6322023859886623,
+      "grad_norm": 0.48950570821762085,
+      "learning_rate": 0.001,
+      "loss": 1.6563,
+      "step": 14944
+    },
+    {
+      "epoch": 0.6322446907521787,
+      "grad_norm": 1.1566777229309082,
+      "learning_rate": 0.001,
+      "loss": 2.7926,
+      "step": 14945
+    },
+    {
+      "epoch": 0.6322869955156951,
+      "grad_norm": 0.20745162665843964,
+      "learning_rate": 0.001,
+      "loss": 2.2428,
+      "step": 14946
+    },
+    {
+      "epoch": 0.6323293002792114,
+      "grad_norm": 0.2187420278787613,
+      "learning_rate": 0.001,
+      "loss": 2.3801,
+      "step": 14947
+    },
+    {
+      "epoch": 0.6323716050427278,
+      "grad_norm": 101.63371276855469,
+      "learning_rate": 0.001,
+      "loss": 2.0414,
+      "step": 14948
+    },
+    {
+      "epoch": 0.6324139098062442,
+      "grad_norm": 0.1789313405752182,
+      "learning_rate": 0.001,
+      "loss": 1.9108,
+      "step": 14949
+    },
+    {
+      "epoch": 0.6324562145697605,
+      "grad_norm": 0.2938269078731537,
+      "learning_rate": 0.001,
+      "loss": 1.9316,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6324985193332769,
+      "grad_norm": 0.32403501868247986,
+      "learning_rate": 0.001,
+      "loss": 1.9873,
+      "step": 14951
+    },
+    {
+      "epoch": 0.6325408240967934,
+      "grad_norm": 0.1873176246881485,
+      "learning_rate": 0.001,
+      "loss": 1.8201,
+      "step": 14952
+    },
+    {
+      "epoch": 0.6325831288603097,
+      "grad_norm": 0.19360357522964478,
+      "learning_rate": 0.001,
+      "loss": 2.7862,
+      "step": 14953
+    },
+    {
+      "epoch": 0.6326254336238261,
+      "grad_norm": 0.17896825075149536,
+      "learning_rate": 0.001,
+      "loss": 2.0564,
+      "step": 14954
+    },
+    {
+      "epoch": 0.6326677383873424,
+      "grad_norm": 1.2869054079055786,
+      "learning_rate": 0.001,
+      "loss": 1.5205,
+      "step": 14955
+    },
+    {
+      "epoch": 0.6327100431508588,
+      "grad_norm": 0.18864065408706665,
+      "learning_rate": 0.001,
+      "loss": 3.0444,
+      "step": 14956
+    },
+    {
+      "epoch": 0.6327523479143752,
+      "grad_norm": 0.880074143409729,
+      "learning_rate": 0.001,
+      "loss": 1.9113,
+      "step": 14957
+    },
+    {
+      "epoch": 0.6327946526778915,
+      "grad_norm": 0.1550513505935669,
+      "learning_rate": 0.001,
+      "loss": 2.146,
+      "step": 14958
+    },
+    {
+      "epoch": 0.6328369574414079,
+      "grad_norm": 0.20279888808727264,
+      "learning_rate": 0.001,
+      "loss": 2.5939,
+      "step": 14959
+    },
+    {
+      "epoch": 0.6328792622049243,
+      "grad_norm": 0.4904308617115021,
+      "learning_rate": 0.001,
+      "loss": 2.1548,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6329215669684406,
+      "grad_norm": 0.15668341517448425,
+      "learning_rate": 0.001,
+      "loss": 3.1503,
+      "step": 14961
+    },
+    {
+      "epoch": 0.632963871731957,
+      "grad_norm": 1.097795009613037,
+      "learning_rate": 0.001,
+      "loss": 1.9487,
+      "step": 14962
+    },
+    {
+      "epoch": 0.6330061764954734,
+      "grad_norm": 0.19955521821975708,
+      "learning_rate": 0.001,
+      "loss": 1.4885,
+      "step": 14963
+    },
+    {
+      "epoch": 0.6330484812589897,
+      "grad_norm": 0.15636754035949707,
+      "learning_rate": 0.001,
+      "loss": 2.508,
+      "step": 14964
+    },
+    {
+      "epoch": 0.6330907860225061,
+      "grad_norm": 2.758397340774536,
+      "learning_rate": 0.001,
+      "loss": 1.6241,
+      "step": 14965
+    },
+    {
+      "epoch": 0.6331330907860225,
+      "grad_norm": 2.932732582092285,
+      "learning_rate": 0.001,
+      "loss": 1.9931,
+      "step": 14966
+    },
+    {
+      "epoch": 0.6331753955495388,
+      "grad_norm": 0.21314847469329834,
+      "learning_rate": 0.001,
+      "loss": 2.2011,
+      "step": 14967
+    },
+    {
+      "epoch": 0.6332177003130552,
+      "grad_norm": 0.23298072814941406,
+      "learning_rate": 0.001,
+      "loss": 3.0595,
+      "step": 14968
+    },
+    {
+      "epoch": 0.6332600050765717,
+      "grad_norm": 0.246332049369812,
+      "learning_rate": 0.001,
+      "loss": 1.9456,
+      "step": 14969
+    },
+    {
+      "epoch": 0.633302309840088,
+      "grad_norm": 0.18252138793468475,
+      "learning_rate": 0.001,
+      "loss": 3.1042,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6333446146036044,
+      "grad_norm": 0.20896947383880615,
+      "learning_rate": 0.001,
+      "loss": 2.6753,
+      "step": 14971
+    },
+    {
+      "epoch": 0.6333869193671208,
+      "grad_norm": 25.508777618408203,
+      "learning_rate": 0.001,
+      "loss": 3.5983,
+      "step": 14972
+    },
+    {
+      "epoch": 0.6334292241306371,
+      "grad_norm": 0.8212241530418396,
+      "learning_rate": 0.001,
+      "loss": 2.4581,
+      "step": 14973
+    },
+    {
+      "epoch": 0.6334715288941535,
+      "grad_norm": 0.36662164330482483,
+      "learning_rate": 0.001,
+      "loss": 1.8713,
+      "step": 14974
+    },
+    {
+      "epoch": 0.6335138336576699,
+      "grad_norm": 0.14960072934627533,
+      "learning_rate": 0.001,
+      "loss": 2.7025,
+      "step": 14975
+    },
+    {
+      "epoch": 0.6335561384211862,
+      "grad_norm": 0.17821753025054932,
+      "learning_rate": 0.001,
+      "loss": 2.1402,
+      "step": 14976
+    },
+    {
+      "epoch": 0.6335984431847026,
+      "grad_norm": 0.17542670667171478,
+      "learning_rate": 0.001,
+      "loss": 1.9822,
+      "step": 14977
+    },
+    {
+      "epoch": 0.633640747948219,
+      "grad_norm": 0.25928542017936707,
+      "learning_rate": 0.001,
+      "loss": 3.1297,
+      "step": 14978
+    },
+    {
+      "epoch": 0.6336830527117353,
+      "grad_norm": 0.1568516343832016,
+      "learning_rate": 0.001,
+      "loss": 1.6175,
+      "step": 14979
+    },
+    {
+      "epoch": 0.6337253574752517,
+      "grad_norm": 0.18374250829219818,
+      "learning_rate": 0.001,
+      "loss": 3.5063,
+      "step": 14980
+    },
+    {
+      "epoch": 0.6337676622387681,
+      "grad_norm": 0.23148831725120544,
+      "learning_rate": 0.001,
+      "loss": 2.3223,
+      "step": 14981
+    },
+    {
+      "epoch": 0.6338099670022844,
+      "grad_norm": 0.23832428455352783,
+      "learning_rate": 0.001,
+      "loss": 2.036,
+      "step": 14982
+    },
+    {
+      "epoch": 0.6338522717658008,
+      "grad_norm": 0.19005969166755676,
+      "learning_rate": 0.001,
+      "loss": 3.0384,
+      "step": 14983
+    },
+    {
+      "epoch": 0.6338945765293172,
+      "grad_norm": 2.5607242584228516,
+      "learning_rate": 0.001,
+      "loss": 2.0137,
+      "step": 14984
+    },
+    {
+      "epoch": 0.6339368812928335,
+      "grad_norm": 0.4837716817855835,
+      "learning_rate": 0.001,
+      "loss": 3.2616,
+      "step": 14985
+    },
+    {
+      "epoch": 0.63397918605635,
+      "grad_norm": 0.5023155212402344,
+      "learning_rate": 0.001,
+      "loss": 2.6116,
+      "step": 14986
+    },
+    {
+      "epoch": 0.6340214908198664,
+      "grad_norm": 0.21957671642303467,
+      "learning_rate": 0.001,
+      "loss": 2.8777,
+      "step": 14987
+    },
+    {
+      "epoch": 0.6340637955833827,
+      "grad_norm": 0.15454959869384766,
+      "learning_rate": 0.001,
+      "loss": 2.8105,
+      "step": 14988
+    },
+    {
+      "epoch": 0.6341061003468991,
+      "grad_norm": 7.386536121368408,
+      "learning_rate": 0.001,
+      "loss": 1.749,
+      "step": 14989
+    },
+    {
+      "epoch": 0.6341484051104155,
+      "grad_norm": 0.22474679350852966,
+      "learning_rate": 0.001,
+      "loss": 2.1073,
+      "step": 14990
+    },
+    {
+      "epoch": 0.6341907098739318,
+      "grad_norm": 0.22885270416736603,
+      "learning_rate": 0.001,
+      "loss": 2.722,
+      "step": 14991
+    },
+    {
+      "epoch": 0.6342330146374482,
+      "grad_norm": 0.1935967653989792,
+      "learning_rate": 0.001,
+      "loss": 1.5022,
+      "step": 14992
+    },
+    {
+      "epoch": 0.6342753194009646,
+      "grad_norm": 1.1925420761108398,
+      "learning_rate": 0.001,
+      "loss": 3.1638,
+      "step": 14993
+    },
+    {
+      "epoch": 0.6343176241644809,
+      "grad_norm": 0.16157661378383636,
+      "learning_rate": 0.001,
+      "loss": 1.8381,
+      "step": 14994
+    },
+    {
+      "epoch": 0.6343599289279973,
+      "grad_norm": 0.2140708863735199,
+      "learning_rate": 0.001,
+      "loss": 2.2911,
+      "step": 14995
+    },
+    {
+      "epoch": 0.6344022336915137,
+      "grad_norm": 0.15380524098873138,
+      "learning_rate": 0.001,
+      "loss": 2.2116,
+      "step": 14996
+    },
+    {
+      "epoch": 0.63444453845503,
+      "grad_norm": 0.19003191590309143,
+      "learning_rate": 0.001,
+      "loss": 2.0538,
+      "step": 14997
+    },
+    {
+      "epoch": 0.6344868432185464,
+      "grad_norm": 0.17649468779563904,
+      "learning_rate": 0.001,
+      "loss": 1.9866,
+      "step": 14998
+    },
+    {
+      "epoch": 0.6345291479820628,
+      "grad_norm": 1.8466432094573975,
+      "learning_rate": 0.001,
+      "loss": 2.4553,
+      "step": 14999
+    },
+    {
+      "epoch": 0.6345714527455791,
+      "grad_norm": 0.1778542697429657,
+      "learning_rate": 0.001,
+      "loss": 2.5347,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6346137575090955,
+      "grad_norm": 0.19804713129997253,
+      "learning_rate": 0.001,
+      "loss": 1.5805,
+      "step": 15001
+    },
+    {
+      "epoch": 0.6346560622726118,
+      "grad_norm": 0.14694872498512268,
+      "learning_rate": 0.001,
+      "loss": 2.0717,
+      "step": 15002
+    },
+    {
+      "epoch": 0.6346983670361283,
+      "grad_norm": 1.054595708847046,
+      "learning_rate": 0.001,
+      "loss": 3.0874,
+      "step": 15003
+    },
+    {
+      "epoch": 0.6347406717996447,
+      "grad_norm": 0.1948278397321701,
+      "learning_rate": 0.001,
+      "loss": 1.9202,
+      "step": 15004
+    },
+    {
+      "epoch": 0.634782976563161,
+      "grad_norm": 0.23619548976421356,
+      "learning_rate": 0.001,
+      "loss": 2.0307,
+      "step": 15005
+    },
+    {
+      "epoch": 0.6348252813266774,
+      "grad_norm": 0.21441768109798431,
+      "learning_rate": 0.001,
+      "loss": 2.4843,
+      "step": 15006
+    },
+    {
+      "epoch": 0.6348675860901938,
+      "grad_norm": 0.19178879261016846,
+      "learning_rate": 0.001,
+      "loss": 2.0039,
+      "step": 15007
+    },
+    {
+      "epoch": 0.6349098908537101,
+      "grad_norm": 0.19128453731536865,
+      "learning_rate": 0.001,
+      "loss": 1.8224,
+      "step": 15008
+    },
+    {
+      "epoch": 0.6349521956172265,
+      "grad_norm": 0.30151814222335815,
+      "learning_rate": 0.001,
+      "loss": 2.6074,
+      "step": 15009
+    },
+    {
+      "epoch": 0.6349945003807429,
+      "grad_norm": 0.1635221242904663,
+      "learning_rate": 0.001,
+      "loss": 1.5166,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6350368051442592,
+      "grad_norm": 0.15775498747825623,
+      "learning_rate": 0.001,
+      "loss": 1.501,
+      "step": 15011
+    },
+    {
+      "epoch": 0.6350791099077756,
+      "grad_norm": 0.19926244020462036,
+      "learning_rate": 0.001,
+      "loss": 2.0229,
+      "step": 15012
+    },
+    {
+      "epoch": 0.635121414671292,
+      "grad_norm": 0.2127702683210373,
+      "learning_rate": 0.001,
+      "loss": 1.9368,
+      "step": 15013
+    },
+    {
+      "epoch": 0.6351637194348083,
+      "grad_norm": 0.19976593554019928,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 15014
+    },
+    {
+      "epoch": 0.6352060241983247,
+      "grad_norm": 0.15875552594661713,
+      "learning_rate": 0.001,
+      "loss": 2.2435,
+      "step": 15015
+    },
+    {
+      "epoch": 0.6352483289618411,
+      "grad_norm": 0.20924918353557587,
+      "learning_rate": 0.001,
+      "loss": 1.889,
+      "step": 15016
+    },
+    {
+      "epoch": 0.6352906337253574,
+      "grad_norm": 0.16609390079975128,
+      "learning_rate": 0.001,
+      "loss": 2.3809,
+      "step": 15017
+    },
+    {
+      "epoch": 0.6353329384888738,
+      "grad_norm": 0.17234043776988983,
+      "learning_rate": 0.001,
+      "loss": 1.9048,
+      "step": 15018
+    },
+    {
+      "epoch": 0.6353752432523903,
+      "grad_norm": 0.18904206156730652,
+      "learning_rate": 0.001,
+      "loss": 2.9954,
+      "step": 15019
+    },
+    {
+      "epoch": 0.6354175480159066,
+      "grad_norm": 0.2591887414455414,
+      "learning_rate": 0.001,
+      "loss": 1.4825,
+      "step": 15020
+    },
+    {
+      "epoch": 0.635459852779423,
+      "grad_norm": 0.21162384748458862,
+      "learning_rate": 0.001,
+      "loss": 2.1042,
+      "step": 15021
+    },
+    {
+      "epoch": 0.6355021575429394,
+      "grad_norm": 0.1585322916507721,
+      "learning_rate": 0.001,
+      "loss": 2.2755,
+      "step": 15022
+    },
+    {
+      "epoch": 0.6355444623064557,
+      "grad_norm": 0.3131169080734253,
+      "learning_rate": 0.001,
+      "loss": 2.6273,
+      "step": 15023
+    },
+    {
+      "epoch": 0.6355867670699721,
+      "grad_norm": 0.1742488294839859,
+      "learning_rate": 0.001,
+      "loss": 2.3599,
+      "step": 15024
+    },
+    {
+      "epoch": 0.6356290718334885,
+      "grad_norm": 0.37338918447494507,
+      "learning_rate": 0.001,
+      "loss": 2.752,
+      "step": 15025
+    },
+    {
+      "epoch": 0.6356713765970048,
+      "grad_norm": 0.20515532791614532,
+      "learning_rate": 0.001,
+      "loss": 2.4337,
+      "step": 15026
+    },
+    {
+      "epoch": 0.6357136813605212,
+      "grad_norm": 0.5772908926010132,
+      "learning_rate": 0.001,
+      "loss": 2.1753,
+      "step": 15027
+    },
+    {
+      "epoch": 0.6357559861240376,
+      "grad_norm": 1.2035185098648071,
+      "learning_rate": 0.001,
+      "loss": 2.298,
+      "step": 15028
+    },
+    {
+      "epoch": 0.6357982908875539,
+      "grad_norm": 0.3183322250843048,
+      "learning_rate": 0.001,
+      "loss": 2.7047,
+      "step": 15029
+    },
+    {
+      "epoch": 0.6358405956510703,
+      "grad_norm": 0.15647853910923004,
+      "learning_rate": 0.001,
+      "loss": 1.6478,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6358829004145867,
+      "grad_norm": 0.16988950967788696,
+      "learning_rate": 0.001,
+      "loss": 1.7595,
+      "step": 15031
+    },
+    {
+      "epoch": 0.635925205178103,
+      "grad_norm": 1.6382477283477783,
+      "learning_rate": 0.001,
+      "loss": 1.7299,
+      "step": 15032
+    },
+    {
+      "epoch": 0.6359675099416194,
+      "grad_norm": 0.3181857466697693,
+      "learning_rate": 0.001,
+      "loss": 1.609,
+      "step": 15033
+    },
+    {
+      "epoch": 0.6360098147051358,
+      "grad_norm": 0.20994387567043304,
+      "learning_rate": 0.001,
+      "loss": 2.6348,
+      "step": 15034
+    },
+    {
+      "epoch": 0.6360521194686521,
+      "grad_norm": 0.17180518805980682,
+      "learning_rate": 0.001,
+      "loss": 1.4551,
+      "step": 15035
+    },
+    {
+      "epoch": 0.6360944242321686,
+      "grad_norm": 0.15223318338394165,
+      "learning_rate": 0.001,
+      "loss": 2.0057,
+      "step": 15036
+    },
+    {
+      "epoch": 0.636136728995685,
+      "grad_norm": 0.14394080638885498,
+      "learning_rate": 0.001,
+      "loss": 1.8883,
+      "step": 15037
+    },
+    {
+      "epoch": 0.6361790337592013,
+      "grad_norm": 0.8304606080055237,
+      "learning_rate": 0.001,
+      "loss": 2.4011,
+      "step": 15038
+    },
+    {
+      "epoch": 0.6362213385227177,
+      "grad_norm": 0.1558523178100586,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 15039
+    },
+    {
+      "epoch": 0.6362636432862341,
+      "grad_norm": 0.19884271919727325,
+      "learning_rate": 0.001,
+      "loss": 3.0935,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6363059480497504,
+      "grad_norm": 0.19426396489143372,
+      "learning_rate": 0.001,
+      "loss": 2.1119,
+      "step": 15041
+    },
+    {
+      "epoch": 0.6363482528132668,
+      "grad_norm": 0.1493634283542633,
+      "learning_rate": 0.001,
+      "loss": 2.2015,
+      "step": 15042
+    },
+    {
+      "epoch": 0.6363905575767832,
+      "grad_norm": 0.1937769651412964,
+      "learning_rate": 0.001,
+      "loss": 2.2157,
+      "step": 15043
+    },
+    {
+      "epoch": 0.6364328623402995,
+      "grad_norm": 0.1691095232963562,
+      "learning_rate": 0.001,
+      "loss": 1.7524,
+      "step": 15044
+    },
+    {
+      "epoch": 0.6364751671038159,
+      "grad_norm": 0.2379046529531479,
+      "learning_rate": 0.001,
+      "loss": 1.9738,
+      "step": 15045
+    },
+    {
+      "epoch": 0.6365174718673322,
+      "grad_norm": 0.4896133244037628,
+      "learning_rate": 0.001,
+      "loss": 1.2294,
+      "step": 15046
+    },
+    {
+      "epoch": 0.6365597766308486,
+      "grad_norm": 0.18665045499801636,
+      "learning_rate": 0.001,
+      "loss": 2.2706,
+      "step": 15047
+    },
+    {
+      "epoch": 0.636602081394365,
+      "grad_norm": 0.1756897121667862,
+      "learning_rate": 0.001,
+      "loss": 2.0685,
+      "step": 15048
+    },
+    {
+      "epoch": 0.6366443861578813,
+      "grad_norm": 0.46474409103393555,
+      "learning_rate": 0.001,
+      "loss": 1.7136,
+      "step": 15049
+    },
+    {
+      "epoch": 0.6366866909213977,
+      "grad_norm": 0.7416086196899414,
+      "learning_rate": 0.001,
+      "loss": 2.623,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6367289956849141,
+      "grad_norm": 0.21350111067295074,
+      "learning_rate": 0.001,
+      "loss": 1.6577,
+      "step": 15051
+    },
+    {
+      "epoch": 0.6367713004484304,
+      "grad_norm": 0.6124548316001892,
+      "learning_rate": 0.001,
+      "loss": 3.5035,
+      "step": 15052
+    },
+    {
+      "epoch": 0.6368136052119469,
+      "grad_norm": 0.19088469445705414,
+      "learning_rate": 0.001,
+      "loss": 2.3181,
+      "step": 15053
+    },
+    {
+      "epoch": 0.6368559099754633,
+      "grad_norm": 0.14974898099899292,
+      "learning_rate": 0.001,
+      "loss": 2.0188,
+      "step": 15054
+    },
+    {
+      "epoch": 0.6368982147389796,
+      "grad_norm": 0.17937543988227844,
+      "learning_rate": 0.001,
+      "loss": 2.6969,
+      "step": 15055
+    },
+    {
+      "epoch": 0.636940519502496,
+      "grad_norm": 0.22011379897594452,
+      "learning_rate": 0.001,
+      "loss": 2.7894,
+      "step": 15056
+    },
+    {
+      "epoch": 0.6369828242660124,
+      "grad_norm": 0.18520912528038025,
+      "learning_rate": 0.001,
+      "loss": 1.7704,
+      "step": 15057
+    },
+    {
+      "epoch": 0.6370251290295287,
+      "grad_norm": 0.1443547159433365,
+      "learning_rate": 0.001,
+      "loss": 2.0866,
+      "step": 15058
+    },
+    {
+      "epoch": 0.6370674337930451,
+      "grad_norm": 0.1495741754770279,
+      "learning_rate": 0.001,
+      "loss": 1.5056,
+      "step": 15059
+    },
+    {
+      "epoch": 0.6371097385565615,
+      "grad_norm": 9.01001262664795,
+      "learning_rate": 0.001,
+      "loss": 1.8438,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6371520433200778,
+      "grad_norm": 0.46700048446655273,
+      "learning_rate": 0.001,
+      "loss": 2.1663,
+      "step": 15061
+    },
+    {
+      "epoch": 0.6371943480835942,
+      "grad_norm": 0.20175836980342865,
+      "learning_rate": 0.001,
+      "loss": 2.9477,
+      "step": 15062
+    },
+    {
+      "epoch": 0.6372366528471106,
+      "grad_norm": 0.162751242518425,
+      "learning_rate": 0.001,
+      "loss": 1.6253,
+      "step": 15063
+    },
+    {
+      "epoch": 0.6372789576106269,
+      "grad_norm": 0.1852269321680069,
+      "learning_rate": 0.001,
+      "loss": 1.4472,
+      "step": 15064
+    },
+    {
+      "epoch": 0.6373212623741433,
+      "grad_norm": 0.17682723701000214,
+      "learning_rate": 0.001,
+      "loss": 1.6896,
+      "step": 15065
+    },
+    {
+      "epoch": 0.6373635671376597,
+      "grad_norm": 0.8060948848724365,
+      "learning_rate": 0.001,
+      "loss": 1.6884,
+      "step": 15066
+    },
+    {
+      "epoch": 0.637405871901176,
+      "grad_norm": 0.30199164152145386,
+      "learning_rate": 0.001,
+      "loss": 2.0973,
+      "step": 15067
+    },
+    {
+      "epoch": 0.6374481766646924,
+      "grad_norm": 0.20303098857402802,
+      "learning_rate": 0.001,
+      "loss": 2.1595,
+      "step": 15068
+    },
+    {
+      "epoch": 0.6374904814282089,
+      "grad_norm": 0.20386897027492523,
+      "learning_rate": 0.001,
+      "loss": 1.6897,
+      "step": 15069
+    },
+    {
+      "epoch": 0.6375327861917252,
+      "grad_norm": 0.233832448720932,
+      "learning_rate": 0.001,
+      "loss": 2.1659,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6375750909552416,
+      "grad_norm": 2.2971856594085693,
+      "learning_rate": 0.001,
+      "loss": 4.1125,
+      "step": 15071
+    },
+    {
+      "epoch": 0.637617395718758,
+      "grad_norm": 0.19406281411647797,
+      "learning_rate": 0.001,
+      "loss": 2.3436,
+      "step": 15072
+    },
+    {
+      "epoch": 0.6376597004822743,
+      "grad_norm": 0.16026239097118378,
+      "learning_rate": 0.001,
+      "loss": 2.5325,
+      "step": 15073
+    },
+    {
+      "epoch": 0.6377020052457907,
+      "grad_norm": 0.21731014549732208,
+      "learning_rate": 0.001,
+      "loss": 2.6568,
+      "step": 15074
+    },
+    {
+      "epoch": 0.6377443100093071,
+      "grad_norm": 0.1755847930908203,
+      "learning_rate": 0.001,
+      "loss": 2.48,
+      "step": 15075
+    },
+    {
+      "epoch": 0.6377866147728234,
+      "grad_norm": 0.16319885849952698,
+      "learning_rate": 0.001,
+      "loss": 2.152,
+      "step": 15076
+    },
+    {
+      "epoch": 0.6378289195363398,
+      "grad_norm": 0.1832021176815033,
+      "learning_rate": 0.001,
+      "loss": 2.4498,
+      "step": 15077
+    },
+    {
+      "epoch": 0.6378712242998562,
+      "grad_norm": 0.17077642679214478,
+      "learning_rate": 0.001,
+      "loss": 2.2088,
+      "step": 15078
+    },
+    {
+      "epoch": 0.6379135290633725,
+      "grad_norm": 0.2068411260843277,
+      "learning_rate": 0.001,
+      "loss": 2.3452,
+      "step": 15079
+    },
+    {
+      "epoch": 0.6379558338268889,
+      "grad_norm": 0.1923973709344864,
+      "learning_rate": 0.001,
+      "loss": 2.124,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6379981385904053,
+      "grad_norm": 7.424524784088135,
+      "learning_rate": 0.001,
+      "loss": 2.9359,
+      "step": 15081
+    },
+    {
+      "epoch": 0.6380404433539216,
+      "grad_norm": 0.2013271301984787,
+      "learning_rate": 0.001,
+      "loss": 1.4738,
+      "step": 15082
+    },
+    {
+      "epoch": 0.638082748117438,
+      "grad_norm": 0.7266159653663635,
+      "learning_rate": 0.001,
+      "loss": 2.4914,
+      "step": 15083
+    },
+    {
+      "epoch": 0.6381250528809544,
+      "grad_norm": 0.7086105346679688,
+      "learning_rate": 0.001,
+      "loss": 2.7849,
+      "step": 15084
+    },
+    {
+      "epoch": 0.6381673576444707,
+      "grad_norm": 1.527816891670227,
+      "learning_rate": 0.001,
+      "loss": 3.3425,
+      "step": 15085
+    },
+    {
+      "epoch": 0.6382096624079872,
+      "grad_norm": 0.30964910984039307,
+      "learning_rate": 0.001,
+      "loss": 2.8131,
+      "step": 15086
+    },
+    {
+      "epoch": 0.6382519671715036,
+      "grad_norm": 0.18733543157577515,
+      "learning_rate": 0.001,
+      "loss": 2.89,
+      "step": 15087
+    },
+    {
+      "epoch": 0.6382942719350199,
+      "grad_norm": 0.32117486000061035,
+      "learning_rate": 0.001,
+      "loss": 3.6968,
+      "step": 15088
+    },
+    {
+      "epoch": 0.6383365766985363,
+      "grad_norm": 0.1650446653366089,
+      "learning_rate": 0.001,
+      "loss": 1.6528,
+      "step": 15089
+    },
+    {
+      "epoch": 0.6383788814620526,
+      "grad_norm": 1.424420714378357,
+      "learning_rate": 0.001,
+      "loss": 2.9049,
+      "step": 15090
+    },
+    {
+      "epoch": 0.638421186225569,
+      "grad_norm": 0.18598754703998566,
+      "learning_rate": 0.001,
+      "loss": 2.5301,
+      "step": 15091
+    },
+    {
+      "epoch": 0.6384634909890854,
+      "grad_norm": 0.24407409131526947,
+      "learning_rate": 0.001,
+      "loss": 1.614,
+      "step": 15092
+    },
+    {
+      "epoch": 0.6385057957526017,
+      "grad_norm": 0.18757480382919312,
+      "learning_rate": 0.001,
+      "loss": 2.3472,
+      "step": 15093
+    },
+    {
+      "epoch": 0.6385481005161181,
+      "grad_norm": 0.17683260142803192,
+      "learning_rate": 0.001,
+      "loss": 2.2612,
+      "step": 15094
+    },
+    {
+      "epoch": 0.6385904052796345,
+      "grad_norm": 0.2939551770687103,
+      "learning_rate": 0.001,
+      "loss": 1.9736,
+      "step": 15095
+    },
+    {
+      "epoch": 0.6386327100431508,
+      "grad_norm": 0.1915324181318283,
+      "learning_rate": 0.001,
+      "loss": 1.6365,
+      "step": 15096
+    },
+    {
+      "epoch": 0.6386750148066672,
+      "grad_norm": 0.21610991656780243,
+      "learning_rate": 0.001,
+      "loss": 2.0241,
+      "step": 15097
+    },
+    {
+      "epoch": 0.6387173195701836,
+      "grad_norm": 0.17032171785831451,
+      "learning_rate": 0.001,
+      "loss": 2.7106,
+      "step": 15098
+    },
+    {
+      "epoch": 0.6387596243336999,
+      "grad_norm": 0.33380383253097534,
+      "learning_rate": 0.001,
+      "loss": 1.8032,
+      "step": 15099
+    },
+    {
+      "epoch": 0.6388019290972163,
+      "grad_norm": 0.15979735553264618,
+      "learning_rate": 0.001,
+      "loss": 2.0711,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6388442338607327,
+      "grad_norm": 0.21214862167835236,
+      "learning_rate": 0.001,
+      "loss": 2.4475,
+      "step": 15101
+    },
+    {
+      "epoch": 0.638886538624249,
+      "grad_norm": 0.19921685755252838,
+      "learning_rate": 0.001,
+      "loss": 2.1466,
+      "step": 15102
+    },
+    {
+      "epoch": 0.6389288433877655,
+      "grad_norm": 0.18050815165042877,
+      "learning_rate": 0.001,
+      "loss": 2.3784,
+      "step": 15103
+    },
+    {
+      "epoch": 0.6389711481512819,
+      "grad_norm": 1.592774748802185,
+      "learning_rate": 0.001,
+      "loss": 2.568,
+      "step": 15104
+    },
+    {
+      "epoch": 0.6390134529147982,
+      "grad_norm": 0.18110394477844238,
+      "learning_rate": 0.001,
+      "loss": 2.3632,
+      "step": 15105
+    },
+    {
+      "epoch": 0.6390557576783146,
+      "grad_norm": 0.13899345695972443,
+      "learning_rate": 0.001,
+      "loss": 2.49,
+      "step": 15106
+    },
+    {
+      "epoch": 0.639098062441831,
+      "grad_norm": 0.16161954402923584,
+      "learning_rate": 0.001,
+      "loss": 1.9518,
+      "step": 15107
+    },
+    {
+      "epoch": 0.6391403672053473,
+      "grad_norm": 0.2258518636226654,
+      "learning_rate": 0.001,
+      "loss": 1.8021,
+      "step": 15108
+    },
+    {
+      "epoch": 0.6391826719688637,
+      "grad_norm": 0.19789251685142517,
+      "learning_rate": 0.001,
+      "loss": 2.7393,
+      "step": 15109
+    },
+    {
+      "epoch": 0.6392249767323801,
+      "grad_norm": 0.37085291743278503,
+      "learning_rate": 0.001,
+      "loss": 1.7739,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6392672814958964,
+      "grad_norm": 1.3115873336791992,
+      "learning_rate": 0.001,
+      "loss": 2.4295,
+      "step": 15111
+    },
+    {
+      "epoch": 0.6393095862594128,
+      "grad_norm": 0.44819048047065735,
+      "learning_rate": 0.001,
+      "loss": 2.1524,
+      "step": 15112
+    },
+    {
+      "epoch": 0.6393518910229292,
+      "grad_norm": 0.1929224282503128,
+      "learning_rate": 0.001,
+      "loss": 1.8331,
+      "step": 15113
+    },
+    {
+      "epoch": 0.6393941957864455,
+      "grad_norm": 0.17352215945720673,
+      "learning_rate": 0.001,
+      "loss": 1.9057,
+      "step": 15114
+    },
+    {
+      "epoch": 0.6394365005499619,
+      "grad_norm": 0.32852211594581604,
+      "learning_rate": 0.001,
+      "loss": 1.5222,
+      "step": 15115
+    },
+    {
+      "epoch": 0.6394788053134783,
+      "grad_norm": 0.17034880816936493,
+      "learning_rate": 0.001,
+      "loss": 2.1697,
+      "step": 15116
+    },
+    {
+      "epoch": 0.6395211100769946,
+      "grad_norm": 0.17381560802459717,
+      "learning_rate": 0.001,
+      "loss": 1.616,
+      "step": 15117
+    },
+    {
+      "epoch": 0.639563414840511,
+      "grad_norm": 1.7698090076446533,
+      "learning_rate": 0.001,
+      "loss": 2.1559,
+      "step": 15118
+    },
+    {
+      "epoch": 0.6396057196040275,
+      "grad_norm": 0.2830827534198761,
+      "learning_rate": 0.001,
+      "loss": 2.1308,
+      "step": 15119
+    },
+    {
+      "epoch": 0.6396480243675438,
+      "grad_norm": 0.25885289907455444,
+      "learning_rate": 0.001,
+      "loss": 1.9705,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6396903291310602,
+      "grad_norm": 0.2354385405778885,
+      "learning_rate": 0.001,
+      "loss": 2.0764,
+      "step": 15121
+    },
+    {
+      "epoch": 0.6397326338945766,
+      "grad_norm": 0.2808350920677185,
+      "learning_rate": 0.001,
+      "loss": 2.703,
+      "step": 15122
+    },
+    {
+      "epoch": 0.6397749386580929,
+      "grad_norm": 0.1770741045475006,
+      "learning_rate": 0.001,
+      "loss": 2.2367,
+      "step": 15123
+    },
+    {
+      "epoch": 0.6398172434216093,
+      "grad_norm": 0.1734161674976349,
+      "learning_rate": 0.001,
+      "loss": 1.4621,
+      "step": 15124
+    },
+    {
+      "epoch": 0.6398595481851257,
+      "grad_norm": 0.2023870348930359,
+      "learning_rate": 0.001,
+      "loss": 2.4376,
+      "step": 15125
+    },
+    {
+      "epoch": 0.639901852948642,
+      "grad_norm": 0.7190189361572266,
+      "learning_rate": 0.001,
+      "loss": 3.3754,
+      "step": 15126
+    },
+    {
+      "epoch": 0.6399441577121584,
+      "grad_norm": 0.18912023305892944,
+      "learning_rate": 0.001,
+      "loss": 3.4364,
+      "step": 15127
+    },
+    {
+      "epoch": 0.6399864624756748,
+      "grad_norm": 0.16115638613700867,
+      "learning_rate": 0.001,
+      "loss": 1.8208,
+      "step": 15128
+    },
+    {
+      "epoch": 0.6400287672391911,
+      "grad_norm": 0.17985692620277405,
+      "learning_rate": 0.001,
+      "loss": 1.921,
+      "step": 15129
+    },
+    {
+      "epoch": 0.6400710720027075,
+      "grad_norm": 1.403349757194519,
+      "learning_rate": 0.001,
+      "loss": 3.553,
+      "step": 15130
+    },
+    {
+      "epoch": 0.6401133767662239,
+      "grad_norm": 0.14458277821540833,
+      "learning_rate": 0.001,
+      "loss": 1.8948,
+      "step": 15131
+    },
+    {
+      "epoch": 0.6401556815297402,
+      "grad_norm": 0.21987377107143402,
+      "learning_rate": 0.001,
+      "loss": 2.8338,
+      "step": 15132
+    },
+    {
+      "epoch": 0.6401979862932566,
+      "grad_norm": 0.3733513355255127,
+      "learning_rate": 0.001,
+      "loss": 2.4485,
+      "step": 15133
+    },
+    {
+      "epoch": 0.640240291056773,
+      "grad_norm": 21.617033004760742,
+      "learning_rate": 0.001,
+      "loss": 2.7377,
+      "step": 15134
+    },
+    {
+      "epoch": 0.6402825958202893,
+      "grad_norm": 0.4703228771686554,
+      "learning_rate": 0.001,
+      "loss": 2.7947,
+      "step": 15135
+    },
+    {
+      "epoch": 0.6403249005838058,
+      "grad_norm": 1.1009413003921509,
+      "learning_rate": 0.001,
+      "loss": 2.0846,
+      "step": 15136
+    },
+    {
+      "epoch": 0.640367205347322,
+      "grad_norm": 0.16309663653373718,
+      "learning_rate": 0.001,
+      "loss": 2.8663,
+      "step": 15137
+    },
+    {
+      "epoch": 0.6404095101108385,
+      "grad_norm": 0.18451957404613495,
+      "learning_rate": 0.001,
+      "loss": 2.7764,
+      "step": 15138
+    },
+    {
+      "epoch": 0.6404518148743549,
+      "grad_norm": 0.20747211575508118,
+      "learning_rate": 0.001,
+      "loss": 1.8173,
+      "step": 15139
+    },
+    {
+      "epoch": 0.6404941196378712,
+      "grad_norm": 0.2003680169582367,
+      "learning_rate": 0.001,
+      "loss": 2.5292,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6405364244013876,
+      "grad_norm": 0.14385853707790375,
+      "learning_rate": 0.001,
+      "loss": 3.4984,
+      "step": 15141
+    },
+    {
+      "epoch": 0.640578729164904,
+      "grad_norm": 0.20833978056907654,
+      "learning_rate": 0.001,
+      "loss": 2.2487,
+      "step": 15142
+    },
+    {
+      "epoch": 0.6406210339284203,
+      "grad_norm": 1.4632526636123657,
+      "learning_rate": 0.001,
+      "loss": 2.327,
+      "step": 15143
+    },
+    {
+      "epoch": 0.6406633386919367,
+      "grad_norm": 0.1615937054157257,
+      "learning_rate": 0.001,
+      "loss": 2.6719,
+      "step": 15144
+    },
+    {
+      "epoch": 0.6407056434554531,
+      "grad_norm": 0.3300820291042328,
+      "learning_rate": 0.001,
+      "loss": 1.7582,
+      "step": 15145
+    },
+    {
+      "epoch": 0.6407479482189694,
+      "grad_norm": 0.40735575556755066,
+      "learning_rate": 0.001,
+      "loss": 1.6912,
+      "step": 15146
+    },
+    {
+      "epoch": 0.6407902529824858,
+      "grad_norm": 0.16738514602184296,
+      "learning_rate": 0.001,
+      "loss": 1.6134,
+      "step": 15147
+    },
+    {
+      "epoch": 0.6408325577460022,
+      "grad_norm": 0.17721709609031677,
+      "learning_rate": 0.001,
+      "loss": 2.4597,
+      "step": 15148
+    },
+    {
+      "epoch": 0.6408748625095185,
+      "grad_norm": 0.2825102210044861,
+      "learning_rate": 0.001,
+      "loss": 1.8421,
+      "step": 15149
+    },
+    {
+      "epoch": 0.6409171672730349,
+      "grad_norm": 0.19013796746730804,
+      "learning_rate": 0.001,
+      "loss": 2.0838,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6409594720365513,
+      "grad_norm": 0.22980915009975433,
+      "learning_rate": 0.001,
+      "loss": 3.4241,
+      "step": 15151
+    },
+    {
+      "epoch": 0.6410017768000676,
+      "grad_norm": 0.4537425935268402,
+      "learning_rate": 0.001,
+      "loss": 2.7787,
+      "step": 15152
+    },
+    {
+      "epoch": 0.6410440815635841,
+      "grad_norm": 0.44562122225761414,
+      "learning_rate": 0.001,
+      "loss": 1.986,
+      "step": 15153
+    },
+    {
+      "epoch": 0.6410863863271005,
+      "grad_norm": 0.1742187738418579,
+      "learning_rate": 0.001,
+      "loss": 1.8076,
+      "step": 15154
+    },
+    {
+      "epoch": 0.6411286910906168,
+      "grad_norm": 0.41731923818588257,
+      "learning_rate": 0.001,
+      "loss": 2.7682,
+      "step": 15155
+    },
+    {
+      "epoch": 0.6411709958541332,
+      "grad_norm": 0.17305545508861542,
+      "learning_rate": 0.001,
+      "loss": 2.2091,
+      "step": 15156
+    },
+    {
+      "epoch": 0.6412133006176496,
+      "grad_norm": 3.8810274600982666,
+      "learning_rate": 0.001,
+      "loss": 2.5479,
+      "step": 15157
+    },
+    {
+      "epoch": 0.6412556053811659,
+      "grad_norm": 0.22583703696727753,
+      "learning_rate": 0.001,
+      "loss": 1.9204,
+      "step": 15158
+    },
+    {
+      "epoch": 0.6412979101446823,
+      "grad_norm": 0.14476729929447174,
+      "learning_rate": 0.001,
+      "loss": 3.0101,
+      "step": 15159
+    },
+    {
+      "epoch": 0.6413402149081987,
+      "grad_norm": 0.19955182075500488,
+      "learning_rate": 0.001,
+      "loss": 2.3063,
+      "step": 15160
+    },
+    {
+      "epoch": 0.641382519671715,
+      "grad_norm": 0.15358324348926544,
+      "learning_rate": 0.001,
+      "loss": 1.4909,
+      "step": 15161
+    },
+    {
+      "epoch": 0.6414248244352314,
+      "grad_norm": 0.2164243459701538,
+      "learning_rate": 0.001,
+      "loss": 1.9155,
+      "step": 15162
+    },
+    {
+      "epoch": 0.6414671291987478,
+      "grad_norm": 0.15397514402866364,
+      "learning_rate": 0.001,
+      "loss": 1.6991,
+      "step": 15163
+    },
+    {
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.14911989867687225,
+      "learning_rate": 0.001,
+      "loss": 1.9703,
+      "step": 15164
+    },
+    {
+      "epoch": 0.6415517387257805,
+      "grad_norm": 0.1464262753725052,
+      "learning_rate": 0.001,
+      "loss": 1.8296,
+      "step": 15165
+    },
+    {
+      "epoch": 0.6415940434892969,
+      "grad_norm": 0.9781848192214966,
+      "learning_rate": 0.001,
+      "loss": 4.401,
+      "step": 15166
+    },
+    {
+      "epoch": 0.6416363482528132,
+      "grad_norm": 0.18489103019237518,
+      "learning_rate": 0.001,
+      "loss": 1.6242,
+      "step": 15167
+    },
+    {
+      "epoch": 0.6416786530163296,
+      "grad_norm": 0.16017593443393707,
+      "learning_rate": 0.001,
+      "loss": 1.9852,
+      "step": 15168
+    },
+    {
+      "epoch": 0.6417209577798461,
+      "grad_norm": 1.0685312747955322,
+      "learning_rate": 0.001,
+      "loss": 2.4017,
+      "step": 15169
+    },
+    {
+      "epoch": 0.6417632625433624,
+      "grad_norm": 0.17249107360839844,
+      "learning_rate": 0.001,
+      "loss": 1.9322,
+      "step": 15170
+    },
+    {
+      "epoch": 0.6418055673068788,
+      "grad_norm": 0.2893502712249756,
+      "learning_rate": 0.001,
+      "loss": 1.7509,
+      "step": 15171
+    },
+    {
+      "epoch": 0.6418478720703952,
+      "grad_norm": 1.8474113941192627,
+      "learning_rate": 0.001,
+      "loss": 1.5001,
+      "step": 15172
+    },
+    {
+      "epoch": 0.6418901768339115,
+      "grad_norm": 0.21686485409736633,
+      "learning_rate": 0.001,
+      "loss": 2.2628,
+      "step": 15173
+    },
+    {
+      "epoch": 0.6419324815974279,
+      "grad_norm": 0.1666136085987091,
+      "learning_rate": 0.001,
+      "loss": 3.1013,
+      "step": 15174
+    },
+    {
+      "epoch": 0.6419747863609443,
+      "grad_norm": 0.20355500280857086,
+      "learning_rate": 0.001,
+      "loss": 1.8449,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6420170911244606,
+      "grad_norm": 0.16820862889289856,
+      "learning_rate": 0.001,
+      "loss": 2.2891,
+      "step": 15176
+    },
+    {
+      "epoch": 0.642059395887977,
+      "grad_norm": 0.16452017426490784,
+      "learning_rate": 0.001,
+      "loss": 1.8697,
+      "step": 15177
+    },
+    {
+      "epoch": 0.6421017006514934,
+      "grad_norm": 0.17088449001312256,
+      "learning_rate": 0.001,
+      "loss": 1.9137,
+      "step": 15178
+    },
+    {
+      "epoch": 0.6421440054150097,
+      "grad_norm": 0.4427139163017273,
+      "learning_rate": 0.001,
+      "loss": 1.8976,
+      "step": 15179
+    },
+    {
+      "epoch": 0.6421863101785261,
+      "grad_norm": 0.16302001476287842,
+      "learning_rate": 0.001,
+      "loss": 2.2036,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6422286149420424,
+      "grad_norm": 0.22736626863479614,
+      "learning_rate": 0.001,
+      "loss": 1.8923,
+      "step": 15181
+    },
+    {
+      "epoch": 0.6422709197055588,
+      "grad_norm": 0.20898133516311646,
+      "learning_rate": 0.001,
+      "loss": 2.1798,
+      "step": 15182
+    },
+    {
+      "epoch": 0.6423132244690752,
+      "grad_norm": 1.8661376237869263,
+      "learning_rate": 0.001,
+      "loss": 2.4555,
+      "step": 15183
+    },
+    {
+      "epoch": 0.6423555292325915,
+      "grad_norm": 0.22298663854599,
+      "learning_rate": 0.001,
+      "loss": 2.2072,
+      "step": 15184
+    },
+    {
+      "epoch": 0.642397833996108,
+      "grad_norm": 3.3507096767425537,
+      "learning_rate": 0.001,
+      "loss": 2.7466,
+      "step": 15185
+    },
+    {
+      "epoch": 0.6424401387596244,
+      "grad_norm": 0.3017289638519287,
+      "learning_rate": 0.001,
+      "loss": 2.6671,
+      "step": 15186
+    },
+    {
+      "epoch": 0.6424824435231407,
+      "grad_norm": 0.24658021330833435,
+      "learning_rate": 0.001,
+      "loss": 1.7703,
+      "step": 15187
+    },
+    {
+      "epoch": 0.6425247482866571,
+      "grad_norm": 2.2778401374816895,
+      "learning_rate": 0.001,
+      "loss": 3.8874,
+      "step": 15188
+    },
+    {
+      "epoch": 0.6425670530501735,
+      "grad_norm": 0.1641712188720703,
+      "learning_rate": 0.001,
+      "loss": 2.0445,
+      "step": 15189
+    },
+    {
+      "epoch": 0.6426093578136898,
+      "grad_norm": 0.26722750067710876,
+      "learning_rate": 0.001,
+      "loss": 2.8895,
+      "step": 15190
+    },
+    {
+      "epoch": 0.6426516625772062,
+      "grad_norm": 0.1975189447402954,
+      "learning_rate": 0.001,
+      "loss": 2.0646,
+      "step": 15191
+    },
+    {
+      "epoch": 0.6426939673407226,
+      "grad_norm": 0.22183305025100708,
+      "learning_rate": 0.001,
+      "loss": 1.8173,
+      "step": 15192
+    },
+    {
+      "epoch": 0.6427362721042389,
+      "grad_norm": 0.17388564348220825,
+      "learning_rate": 0.001,
+      "loss": 1.6142,
+      "step": 15193
+    },
+    {
+      "epoch": 0.6427785768677553,
+      "grad_norm": 0.169900044798851,
+      "learning_rate": 0.001,
+      "loss": 2.7619,
+      "step": 15194
+    },
+    {
+      "epoch": 0.6428208816312717,
+      "grad_norm": 0.2132178544998169,
+      "learning_rate": 0.001,
+      "loss": 2.2104,
+      "step": 15195
+    },
+    {
+      "epoch": 0.642863186394788,
+      "grad_norm": 0.2020595222711563,
+      "learning_rate": 0.001,
+      "loss": 1.8507,
+      "step": 15196
+    },
+    {
+      "epoch": 0.6429054911583044,
+      "grad_norm": 0.18588212132453918,
+      "learning_rate": 0.001,
+      "loss": 2.8819,
+      "step": 15197
+    },
+    {
+      "epoch": 0.6429477959218208,
+      "grad_norm": 0.17926856875419617,
+      "learning_rate": 0.001,
+      "loss": 2.0165,
+      "step": 15198
+    },
+    {
+      "epoch": 0.6429901006853371,
+      "grad_norm": 0.1749180406332016,
+      "learning_rate": 0.001,
+      "loss": 2.548,
+      "step": 15199
+    },
+    {
+      "epoch": 0.6430324054488535,
+      "grad_norm": 0.1662074625492096,
+      "learning_rate": 0.001,
+      "loss": 2.1117,
+      "step": 15200
+    },
+    {
+      "epoch": 0.64307471021237,
+      "grad_norm": 0.5484973788261414,
+      "learning_rate": 0.001,
+      "loss": 1.8429,
+      "step": 15201
+    },
+    {
+      "epoch": 0.6431170149758862,
+      "grad_norm": 0.16047289967536926,
+      "learning_rate": 0.001,
+      "loss": 2.5943,
+      "step": 15202
+    },
+    {
+      "epoch": 0.6431593197394027,
+      "grad_norm": 0.17750583589076996,
+      "learning_rate": 0.001,
+      "loss": 1.8823,
+      "step": 15203
+    },
+    {
+      "epoch": 0.6432016245029191,
+      "grad_norm": 0.17922556400299072,
+      "learning_rate": 0.001,
+      "loss": 1.844,
+      "step": 15204
+    },
+    {
+      "epoch": 0.6432439292664354,
+      "grad_norm": 0.2863519489765167,
+      "learning_rate": 0.001,
+      "loss": 2.2631,
+      "step": 15205
+    },
+    {
+      "epoch": 0.6432862340299518,
+      "grad_norm": 0.19518601894378662,
+      "learning_rate": 0.001,
+      "loss": 1.5934,
+      "step": 15206
+    },
+    {
+      "epoch": 0.6433285387934682,
+      "grad_norm": 0.15284669399261475,
+      "learning_rate": 0.001,
+      "loss": 1.7176,
+      "step": 15207
+    },
+    {
+      "epoch": 0.6433708435569845,
+      "grad_norm": 0.16001272201538086,
+      "learning_rate": 0.001,
+      "loss": 2.4634,
+      "step": 15208
+    },
+    {
+      "epoch": 0.6434131483205009,
+      "grad_norm": 1.5020637512207031,
+      "learning_rate": 0.001,
+      "loss": 2.363,
+      "step": 15209
+    },
+    {
+      "epoch": 0.6434554530840173,
+      "grad_norm": 0.19093799591064453,
+      "learning_rate": 0.001,
+      "loss": 2.8734,
+      "step": 15210
+    },
+    {
+      "epoch": 0.6434977578475336,
+      "grad_norm": 0.1673070341348648,
+      "learning_rate": 0.001,
+      "loss": 2.2707,
+      "step": 15211
+    },
+    {
+      "epoch": 0.64354006261105,
+      "grad_norm": 0.1651526242494583,
+      "learning_rate": 0.001,
+      "loss": 2.457,
+      "step": 15212
+    },
+    {
+      "epoch": 0.6435823673745664,
+      "grad_norm": 0.34426966309547424,
+      "learning_rate": 0.001,
+      "loss": 2.6741,
+      "step": 15213
+    },
+    {
+      "epoch": 0.6436246721380827,
+      "grad_norm": 0.15150640904903412,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 15214
+    },
+    {
+      "epoch": 0.6436669769015991,
+      "grad_norm": 0.1601693034172058,
+      "learning_rate": 0.001,
+      "loss": 1.4741,
+      "step": 15215
+    },
+    {
+      "epoch": 0.6437092816651155,
+      "grad_norm": 0.19948755204677582,
+      "learning_rate": 0.001,
+      "loss": 1.6806,
+      "step": 15216
+    },
+    {
+      "epoch": 0.6437515864286318,
+      "grad_norm": 0.17080558836460114,
+      "learning_rate": 0.001,
+      "loss": 2.2009,
+      "step": 15217
+    },
+    {
+      "epoch": 0.6437938911921482,
+      "grad_norm": 0.1389591544866562,
+      "learning_rate": 0.001,
+      "loss": 1.921,
+      "step": 15218
+    },
+    {
+      "epoch": 0.6438361959556647,
+      "grad_norm": 1.358182668685913,
+      "learning_rate": 0.001,
+      "loss": 2.0884,
+      "step": 15219
+    },
+    {
+      "epoch": 0.643878500719181,
+      "grad_norm": 0.3052426278591156,
+      "learning_rate": 0.001,
+      "loss": 3.247,
+      "step": 15220
+    },
+    {
+      "epoch": 0.6439208054826974,
+      "grad_norm": 0.16330936551094055,
+      "learning_rate": 0.001,
+      "loss": 1.6572,
+      "step": 15221
+    },
+    {
+      "epoch": 0.6439631102462138,
+      "grad_norm": 0.21469064056873322,
+      "learning_rate": 0.001,
+      "loss": 1.9606,
+      "step": 15222
+    },
+    {
+      "epoch": 0.6440054150097301,
+      "grad_norm": 0.17421077191829681,
+      "learning_rate": 0.001,
+      "loss": 2.3584,
+      "step": 15223
+    },
+    {
+      "epoch": 0.6440477197732465,
+      "grad_norm": 0.18720628321170807,
+      "learning_rate": 0.001,
+      "loss": 2.3146,
+      "step": 15224
+    },
+    {
+      "epoch": 0.6440900245367628,
+      "grad_norm": 1.2276394367218018,
+      "learning_rate": 0.001,
+      "loss": 2.6331,
+      "step": 15225
+    },
+    {
+      "epoch": 0.6441323293002792,
+      "grad_norm": 0.17419876158237457,
+      "learning_rate": 0.001,
+      "loss": 2.1331,
+      "step": 15226
+    },
+    {
+      "epoch": 0.6441746340637956,
+      "grad_norm": 0.1927102506160736,
+      "learning_rate": 0.001,
+      "loss": 2.2267,
+      "step": 15227
+    },
+    {
+      "epoch": 0.6442169388273119,
+      "grad_norm": 0.19751717150211334,
+      "learning_rate": 0.001,
+      "loss": 2.5559,
+      "step": 15228
+    },
+    {
+      "epoch": 0.6442592435908283,
+      "grad_norm": 0.14785194396972656,
+      "learning_rate": 0.001,
+      "loss": 3.0032,
+      "step": 15229
+    },
+    {
+      "epoch": 0.6443015483543447,
+      "grad_norm": 0.3033665716648102,
+      "learning_rate": 0.001,
+      "loss": 2.1101,
+      "step": 15230
+    },
+    {
+      "epoch": 0.644343853117861,
+      "grad_norm": 0.1637023538351059,
+      "learning_rate": 0.001,
+      "loss": 2.0958,
+      "step": 15231
+    },
+    {
+      "epoch": 0.6443861578813774,
+      "grad_norm": 0.24798746407032013,
+      "learning_rate": 0.001,
+      "loss": 2.0162,
+      "step": 15232
+    },
+    {
+      "epoch": 0.6444284626448938,
+      "grad_norm": 0.16221873462200165,
+      "learning_rate": 0.001,
+      "loss": 1.8261,
+      "step": 15233
+    },
+    {
+      "epoch": 0.6444707674084101,
+      "grad_norm": 0.1729075014591217,
+      "learning_rate": 0.001,
+      "loss": 2.3098,
+      "step": 15234
+    },
+    {
+      "epoch": 0.6445130721719265,
+      "grad_norm": 0.1991363763809204,
+      "learning_rate": 0.001,
+      "loss": 1.3549,
+      "step": 15235
+    },
+    {
+      "epoch": 0.644555376935443,
+      "grad_norm": 1.0189127922058105,
+      "learning_rate": 0.001,
+      "loss": 2.1717,
+      "step": 15236
+    },
+    {
+      "epoch": 0.6445976816989593,
+      "grad_norm": 0.16026759147644043,
+      "learning_rate": 0.001,
+      "loss": 2.4036,
+      "step": 15237
+    },
+    {
+      "epoch": 0.6446399864624757,
+      "grad_norm": 0.31504741311073303,
+      "learning_rate": 0.001,
+      "loss": 1.429,
+      "step": 15238
+    },
+    {
+      "epoch": 0.6446822912259921,
+      "grad_norm": 0.2236756980419159,
+      "learning_rate": 0.001,
+      "loss": 2.204,
+      "step": 15239
+    },
+    {
+      "epoch": 0.6447245959895084,
+      "grad_norm": 0.18025624752044678,
+      "learning_rate": 0.001,
+      "loss": 2.653,
+      "step": 15240
+    },
+    {
+      "epoch": 0.6447669007530248,
+      "grad_norm": 0.1674138605594635,
+      "learning_rate": 0.001,
+      "loss": 3.3606,
+      "step": 15241
+    },
+    {
+      "epoch": 0.6448092055165412,
+      "grad_norm": 0.1862388402223587,
+      "learning_rate": 0.001,
+      "loss": 3.4166,
+      "step": 15242
+    },
+    {
+      "epoch": 0.6448515102800575,
+      "grad_norm": 0.17802289128303528,
+      "learning_rate": 0.001,
+      "loss": 3.0027,
+      "step": 15243
+    },
+    {
+      "epoch": 0.6448938150435739,
+      "grad_norm": 0.1625264286994934,
+      "learning_rate": 0.001,
+      "loss": 2.5584,
+      "step": 15244
+    },
+    {
+      "epoch": 0.6449361198070903,
+      "grad_norm": 0.293082058429718,
+      "learning_rate": 0.001,
+      "loss": 2.7967,
+      "step": 15245
+    },
+    {
+      "epoch": 0.6449784245706066,
+      "grad_norm": 0.20152679085731506,
+      "learning_rate": 0.001,
+      "loss": 1.4473,
+      "step": 15246
+    },
+    {
+      "epoch": 0.645020729334123,
+      "grad_norm": 0.5746887922286987,
+      "learning_rate": 0.001,
+      "loss": 1.424,
+      "step": 15247
+    },
+    {
+      "epoch": 0.6450630340976394,
+      "grad_norm": 0.19594338536262512,
+      "learning_rate": 0.001,
+      "loss": 2.3114,
+      "step": 15248
+    },
+    {
+      "epoch": 0.6451053388611557,
+      "grad_norm": 0.218429297208786,
+      "learning_rate": 0.001,
+      "loss": 1.768,
+      "step": 15249
+    },
+    {
+      "epoch": 0.6451476436246721,
+      "grad_norm": 0.2682070732116699,
+      "learning_rate": 0.001,
+      "loss": 1.842,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6451899483881885,
+      "grad_norm": 0.34150052070617676,
+      "learning_rate": 0.001,
+      "loss": 1.7695,
+      "step": 15251
+    },
+    {
+      "epoch": 0.6452322531517048,
+      "grad_norm": 0.19166529178619385,
+      "learning_rate": 0.001,
+      "loss": 2.596,
+      "step": 15252
+    },
+    {
+      "epoch": 0.6452745579152213,
+      "grad_norm": 1.010944128036499,
+      "learning_rate": 0.001,
+      "loss": 2.4779,
+      "step": 15253
+    },
+    {
+      "epoch": 0.6453168626787377,
+      "grad_norm": 0.20341774821281433,
+      "learning_rate": 0.001,
+      "loss": 2.7125,
+      "step": 15254
+    },
+    {
+      "epoch": 0.645359167442254,
+      "grad_norm": 0.28562796115875244,
+      "learning_rate": 0.001,
+      "loss": 2.493,
+      "step": 15255
+    },
+    {
+      "epoch": 0.6454014722057704,
+      "grad_norm": 0.19176359474658966,
+      "learning_rate": 0.001,
+      "loss": 2.4811,
+      "step": 15256
+    },
+    {
+      "epoch": 0.6454437769692868,
+      "grad_norm": 0.17995816469192505,
+      "learning_rate": 0.001,
+      "loss": 1.7175,
+      "step": 15257
+    },
+    {
+      "epoch": 0.6454860817328031,
+      "grad_norm": 0.2325253188610077,
+      "learning_rate": 0.001,
+      "loss": 2.5527,
+      "step": 15258
+    },
+    {
+      "epoch": 0.6455283864963195,
+      "grad_norm": 0.7789939641952515,
+      "learning_rate": 0.001,
+      "loss": 1.7572,
+      "step": 15259
+    },
+    {
+      "epoch": 0.6455706912598359,
+      "grad_norm": 0.1507723033428192,
+      "learning_rate": 0.001,
+      "loss": 2.1685,
+      "step": 15260
+    },
+    {
+      "epoch": 0.6456129960233522,
+      "grad_norm": 0.15807674825191498,
+      "learning_rate": 0.001,
+      "loss": 2.5854,
+      "step": 15261
+    },
+    {
+      "epoch": 0.6456553007868686,
+      "grad_norm": 0.1875530183315277,
+      "learning_rate": 0.001,
+      "loss": 2.4665,
+      "step": 15262
+    },
+    {
+      "epoch": 0.645697605550385,
+      "grad_norm": 0.20089443027973175,
+      "learning_rate": 0.001,
+      "loss": 1.8378,
+      "step": 15263
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.40566131472587585,
+      "learning_rate": 0.001,
+      "loss": 1.5999,
+      "step": 15264
+    },
+    {
+      "epoch": 0.6457822150774177,
+      "grad_norm": 2.285637378692627,
+      "learning_rate": 0.001,
+      "loss": 2.1813,
+      "step": 15265
+    },
+    {
+      "epoch": 0.6458245198409341,
+      "grad_norm": 0.1539418250322342,
+      "learning_rate": 0.001,
+      "loss": 1.8228,
+      "step": 15266
+    },
+    {
+      "epoch": 0.6458668246044504,
+      "grad_norm": 5.015846252441406,
+      "learning_rate": 0.001,
+      "loss": 2.1239,
+      "step": 15267
+    },
+    {
+      "epoch": 0.6459091293679668,
+      "grad_norm": 0.2014991044998169,
+      "learning_rate": 0.001,
+      "loss": 2.0453,
+      "step": 15268
+    },
+    {
+      "epoch": 0.6459514341314833,
+      "grad_norm": 0.4300079345703125,
+      "learning_rate": 0.001,
+      "loss": 2.0821,
+      "step": 15269
+    },
+    {
+      "epoch": 0.6459937388949996,
+      "grad_norm": 0.20501923561096191,
+      "learning_rate": 0.001,
+      "loss": 2.1233,
+      "step": 15270
+    },
+    {
+      "epoch": 0.646036043658516,
+      "grad_norm": 0.17998738586902618,
+      "learning_rate": 0.001,
+      "loss": 1.9638,
+      "step": 15271
+    },
+    {
+      "epoch": 0.6460783484220323,
+      "grad_norm": 0.17478275299072266,
+      "learning_rate": 0.001,
+      "loss": 1.6565,
+      "step": 15272
+    },
+    {
+      "epoch": 0.6461206531855487,
+      "grad_norm": 0.17657312750816345,
+      "learning_rate": 0.001,
+      "loss": 1.9822,
+      "step": 15273
+    },
+    {
+      "epoch": 0.6461629579490651,
+      "grad_norm": 2.826846122741699,
+      "learning_rate": 0.001,
+      "loss": 1.6034,
+      "step": 15274
+    },
+    {
+      "epoch": 0.6462052627125814,
+      "grad_norm": 0.9250707626342773,
+      "learning_rate": 0.001,
+      "loss": 3.1013,
+      "step": 15275
+    },
+    {
+      "epoch": 0.6462475674760978,
+      "grad_norm": 0.1742154359817505,
+      "learning_rate": 0.001,
+      "loss": 2.5725,
+      "step": 15276
+    },
+    {
+      "epoch": 0.6462898722396142,
+      "grad_norm": 1.7787930965423584,
+      "learning_rate": 0.001,
+      "loss": 2.0831,
+      "step": 15277
+    },
+    {
+      "epoch": 0.6463321770031305,
+      "grad_norm": 0.17975670099258423,
+      "learning_rate": 0.001,
+      "loss": 1.9179,
+      "step": 15278
+    },
+    {
+      "epoch": 0.6463744817666469,
+      "grad_norm": 0.1871751993894577,
+      "learning_rate": 0.001,
+      "loss": 1.7723,
+      "step": 15279
+    },
+    {
+      "epoch": 0.6464167865301633,
+      "grad_norm": 0.19217629730701447,
+      "learning_rate": 0.001,
+      "loss": 2.0674,
+      "step": 15280
+    },
+    {
+      "epoch": 0.6464590912936796,
+      "grad_norm": 0.2209371030330658,
+      "learning_rate": 0.001,
+      "loss": 1.9014,
+      "step": 15281
+    },
+    {
+      "epoch": 0.646501396057196,
+      "grad_norm": 0.1662231981754303,
+      "learning_rate": 0.001,
+      "loss": 2.1925,
+      "step": 15282
+    },
+    {
+      "epoch": 0.6465437008207124,
+      "grad_norm": 0.40641865134239197,
+      "learning_rate": 0.001,
+      "loss": 2.4045,
+      "step": 15283
+    },
+    {
+      "epoch": 0.6465860055842287,
+      "grad_norm": 0.14980608224868774,
+      "learning_rate": 0.001,
+      "loss": 1.4153,
+      "step": 15284
+    },
+    {
+      "epoch": 0.6466283103477451,
+      "grad_norm": 0.18884730339050293,
+      "learning_rate": 0.001,
+      "loss": 1.945,
+      "step": 15285
+    },
+    {
+      "epoch": 0.6466706151112616,
+      "grad_norm": 0.1964019387960434,
+      "learning_rate": 0.001,
+      "loss": 2.997,
+      "step": 15286
+    },
+    {
+      "epoch": 0.6467129198747779,
+      "grad_norm": 0.1846921443939209,
+      "learning_rate": 0.001,
+      "loss": 2.0583,
+      "step": 15287
+    },
+    {
+      "epoch": 0.6467552246382943,
+      "grad_norm": 0.3572556674480438,
+      "learning_rate": 0.001,
+      "loss": 3.6665,
+      "step": 15288
+    },
+    {
+      "epoch": 0.6467975294018107,
+      "grad_norm": 0.1696724146604538,
+      "learning_rate": 0.001,
+      "loss": 1.6555,
+      "step": 15289
+    },
+    {
+      "epoch": 0.646839834165327,
+      "grad_norm": 0.18197889626026154,
+      "learning_rate": 0.001,
+      "loss": 2.0833,
+      "step": 15290
+    },
+    {
+      "epoch": 0.6468821389288434,
+      "grad_norm": 0.7675547003746033,
+      "learning_rate": 0.001,
+      "loss": 1.9319,
+      "step": 15291
+    },
+    {
+      "epoch": 0.6469244436923598,
+      "grad_norm": 0.16503793001174927,
+      "learning_rate": 0.001,
+      "loss": 2.4541,
+      "step": 15292
+    },
+    {
+      "epoch": 0.6469667484558761,
+      "grad_norm": 0.5148576498031616,
+      "learning_rate": 0.001,
+      "loss": 1.8749,
+      "step": 15293
+    },
+    {
+      "epoch": 0.6470090532193925,
+      "grad_norm": 0.1498442143201828,
+      "learning_rate": 0.001,
+      "loss": 2.6986,
+      "step": 15294
+    },
+    {
+      "epoch": 0.6470513579829089,
+      "grad_norm": 0.22199515998363495,
+      "learning_rate": 0.001,
+      "loss": 2.5812,
+      "step": 15295
+    },
+    {
+      "epoch": 0.6470936627464252,
+      "grad_norm": 0.15802200138568878,
+      "learning_rate": 0.001,
+      "loss": 1.5685,
+      "step": 15296
+    },
+    {
+      "epoch": 0.6471359675099416,
+      "grad_norm": 0.15270139276981354,
+      "learning_rate": 0.001,
+      "loss": 2.1229,
+      "step": 15297
+    },
+    {
+      "epoch": 0.647178272273458,
+      "grad_norm": 0.15476052463054657,
+      "learning_rate": 0.001,
+      "loss": 2.2869,
+      "step": 15298
+    },
+    {
+      "epoch": 0.6472205770369743,
+      "grad_norm": 0.23796048760414124,
+      "learning_rate": 0.001,
+      "loss": 1.6435,
+      "step": 15299
+    },
+    {
+      "epoch": 0.6472628818004907,
+      "grad_norm": 0.2162591516971588,
+      "learning_rate": 0.001,
+      "loss": 2.5024,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6473051865640072,
+      "grad_norm": 0.5299810171127319,
+      "learning_rate": 0.001,
+      "loss": 2.4246,
+      "step": 15301
+    },
+    {
+      "epoch": 0.6473474913275234,
+      "grad_norm": 1.423614740371704,
+      "learning_rate": 0.001,
+      "loss": 2.4182,
+      "step": 15302
+    },
+    {
+      "epoch": 0.6473897960910399,
+      "grad_norm": 0.15965206921100616,
+      "learning_rate": 0.001,
+      "loss": 1.6585,
+      "step": 15303
+    },
+    {
+      "epoch": 0.6474321008545563,
+      "grad_norm": 0.2295125126838684,
+      "learning_rate": 0.001,
+      "loss": 1.5463,
+      "step": 15304
+    },
+    {
+      "epoch": 0.6474744056180726,
+      "grad_norm": 0.16311728954315186,
+      "learning_rate": 0.001,
+      "loss": 2.6348,
+      "step": 15305
+    },
+    {
+      "epoch": 0.647516710381589,
+      "grad_norm": 0.16523398458957672,
+      "learning_rate": 0.001,
+      "loss": 2.221,
+      "step": 15306
+    },
+    {
+      "epoch": 0.6475590151451054,
+      "grad_norm": 0.16306725144386292,
+      "learning_rate": 0.001,
+      "loss": 1.871,
+      "step": 15307
+    },
+    {
+      "epoch": 0.6476013199086217,
+      "grad_norm": 0.24580958485603333,
+      "learning_rate": 0.001,
+      "loss": 3.2734,
+      "step": 15308
+    },
+    {
+      "epoch": 0.6476436246721381,
+      "grad_norm": 0.3764786124229431,
+      "learning_rate": 0.001,
+      "loss": 2.7352,
+      "step": 15309
+    },
+    {
+      "epoch": 0.6476859294356545,
+      "grad_norm": 0.41898685693740845,
+      "learning_rate": 0.001,
+      "loss": 2.2815,
+      "step": 15310
+    },
+    {
+      "epoch": 0.6477282341991708,
+      "grad_norm": 1.0306098461151123,
+      "learning_rate": 0.001,
+      "loss": 2.1651,
+      "step": 15311
+    },
+    {
+      "epoch": 0.6477705389626872,
+      "grad_norm": 0.20003588497638702,
+      "learning_rate": 0.001,
+      "loss": 1.8374,
+      "step": 15312
+    },
+    {
+      "epoch": 0.6478128437262036,
+      "grad_norm": 0.18757633864879608,
+      "learning_rate": 0.001,
+      "loss": 1.9034,
+      "step": 15313
+    },
+    {
+      "epoch": 0.6478551484897199,
+      "grad_norm": 0.2218223512172699,
+      "learning_rate": 0.001,
+      "loss": 1.9381,
+      "step": 15314
+    },
+    {
+      "epoch": 0.6478974532532363,
+      "grad_norm": 0.17475609481334686,
+      "learning_rate": 0.001,
+      "loss": 1.7699,
+      "step": 15315
+    },
+    {
+      "epoch": 0.6479397580167526,
+      "grad_norm": 0.23167279362678528,
+      "learning_rate": 0.001,
+      "loss": 1.9099,
+      "step": 15316
+    },
+    {
+      "epoch": 0.647982062780269,
+      "grad_norm": 0.1843143105506897,
+      "learning_rate": 0.001,
+      "loss": 1.8182,
+      "step": 15317
+    },
+    {
+      "epoch": 0.6480243675437855,
+      "grad_norm": 0.20558197796344757,
+      "learning_rate": 0.001,
+      "loss": 2.5398,
+      "step": 15318
+    },
+    {
+      "epoch": 0.6480666723073017,
+      "grad_norm": 0.17838044464588165,
+      "learning_rate": 0.001,
+      "loss": 1.7194,
+      "step": 15319
+    },
+    {
+      "epoch": 0.6481089770708182,
+      "grad_norm": 0.21093347668647766,
+      "learning_rate": 0.001,
+      "loss": 3.3587,
+      "step": 15320
+    },
+    {
+      "epoch": 0.6481512818343346,
+      "grad_norm": 0.24948959052562714,
+      "learning_rate": 0.001,
+      "loss": 2.1242,
+      "step": 15321
+    },
+    {
+      "epoch": 0.6481935865978509,
+      "grad_norm": 0.17409680783748627,
+      "learning_rate": 0.001,
+      "loss": 1.7667,
+      "step": 15322
+    },
+    {
+      "epoch": 0.6482358913613673,
+      "grad_norm": 0.1577576845884323,
+      "learning_rate": 0.001,
+      "loss": 1.9457,
+      "step": 15323
+    },
+    {
+      "epoch": 0.6482781961248837,
+      "grad_norm": 0.6335095167160034,
+      "learning_rate": 0.001,
+      "loss": 2.0039,
+      "step": 15324
+    },
+    {
+      "epoch": 0.6483205008884,
+      "grad_norm": 0.23187367618083954,
+      "learning_rate": 0.001,
+      "loss": 1.9439,
+      "step": 15325
+    },
+    {
+      "epoch": 0.6483628056519164,
+      "grad_norm": 0.17001605033874512,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 15326
+    },
+    {
+      "epoch": 0.6484051104154328,
+      "grad_norm": 0.1839197725057602,
+      "learning_rate": 0.001,
+      "loss": 3.4747,
+      "step": 15327
+    },
+    {
+      "epoch": 0.6484474151789491,
+      "grad_norm": 0.18075038492679596,
+      "learning_rate": 0.001,
+      "loss": 1.8859,
+      "step": 15328
+    },
+    {
+      "epoch": 0.6484897199424655,
+      "grad_norm": 0.42471522092819214,
+      "learning_rate": 0.001,
+      "loss": 2.1715,
+      "step": 15329
+    },
+    {
+      "epoch": 0.6485320247059819,
+      "grad_norm": 0.3928261995315552,
+      "learning_rate": 0.001,
+      "loss": 3.119,
+      "step": 15330
+    },
+    {
+      "epoch": 0.6485743294694982,
+      "grad_norm": 0.15341652929782867,
+      "learning_rate": 0.001,
+      "loss": 2.5067,
+      "step": 15331
+    },
+    {
+      "epoch": 0.6486166342330146,
+      "grad_norm": 0.16816523671150208,
+      "learning_rate": 0.001,
+      "loss": 2.3909,
+      "step": 15332
+    },
+    {
+      "epoch": 0.648658938996531,
+      "grad_norm": 0.24296429753303528,
+      "learning_rate": 0.001,
+      "loss": 2.4028,
+      "step": 15333
+    },
+    {
+      "epoch": 0.6487012437600473,
+      "grad_norm": 0.3298254907131195,
+      "learning_rate": 0.001,
+      "loss": 2.0308,
+      "step": 15334
+    },
+    {
+      "epoch": 0.6487435485235638,
+      "grad_norm": 0.9870914816856384,
+      "learning_rate": 0.001,
+      "loss": 2.1879,
+      "step": 15335
+    },
+    {
+      "epoch": 0.6487858532870802,
+      "grad_norm": 0.1429598480463028,
+      "learning_rate": 0.001,
+      "loss": 1.3078,
+      "step": 15336
+    },
+    {
+      "epoch": 0.6488281580505965,
+      "grad_norm": 2.075826644897461,
+      "learning_rate": 0.001,
+      "loss": 1.9437,
+      "step": 15337
+    },
+    {
+      "epoch": 0.6488704628141129,
+      "grad_norm": 0.18827824294567108,
+      "learning_rate": 0.001,
+      "loss": 1.5201,
+      "step": 15338
+    },
+    {
+      "epoch": 0.6489127675776293,
+      "grad_norm": 0.16094815731048584,
+      "learning_rate": 0.001,
+      "loss": 2.2602,
+      "step": 15339
+    },
+    {
+      "epoch": 0.6489550723411456,
+      "grad_norm": 0.18486766517162323,
+      "learning_rate": 0.001,
+      "loss": 2.3534,
+      "step": 15340
+    },
+    {
+      "epoch": 0.648997377104662,
+      "grad_norm": 0.22534742951393127,
+      "learning_rate": 0.001,
+      "loss": 2.5483,
+      "step": 15341
+    },
+    {
+      "epoch": 0.6490396818681784,
+      "grad_norm": 0.17582590878009796,
+      "learning_rate": 0.001,
+      "loss": 1.4281,
+      "step": 15342
+    },
+    {
+      "epoch": 0.6490819866316947,
+      "grad_norm": 0.13245414197444916,
+      "learning_rate": 0.001,
+      "loss": 1.9149,
+      "step": 15343
+    },
+    {
+      "epoch": 0.6491242913952111,
+      "grad_norm": 0.4428796172142029,
+      "learning_rate": 0.001,
+      "loss": 2.2785,
+      "step": 15344
+    },
+    {
+      "epoch": 0.6491665961587275,
+      "grad_norm": 0.289799302816391,
+      "learning_rate": 0.001,
+      "loss": 1.9939,
+      "step": 15345
+    },
+    {
+      "epoch": 0.6492089009222438,
+      "grad_norm": 0.16207891702651978,
+      "learning_rate": 0.001,
+      "loss": 2.3712,
+      "step": 15346
+    },
+    {
+      "epoch": 0.6492512056857602,
+      "grad_norm": 0.20188239216804504,
+      "learning_rate": 0.001,
+      "loss": 2.4096,
+      "step": 15347
+    },
+    {
+      "epoch": 0.6492935104492766,
+      "grad_norm": 0.21010519564151764,
+      "learning_rate": 0.001,
+      "loss": 2.6139,
+      "step": 15348
+    },
+    {
+      "epoch": 0.6493358152127929,
+      "grad_norm": 0.3073784112930298,
+      "learning_rate": 0.001,
+      "loss": 2.1791,
+      "step": 15349
+    },
+    {
+      "epoch": 0.6493781199763093,
+      "grad_norm": 0.3525810241699219,
+      "learning_rate": 0.001,
+      "loss": 2.2354,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6494204247398258,
+      "grad_norm": 0.18765120208263397,
+      "learning_rate": 0.001,
+      "loss": 2.6249,
+      "step": 15351
+    },
+    {
+      "epoch": 0.649462729503342,
+      "grad_norm": 0.21588948369026184,
+      "learning_rate": 0.001,
+      "loss": 1.4647,
+      "step": 15352
+    },
+    {
+      "epoch": 0.6495050342668585,
+      "grad_norm": 0.17089973390102386,
+      "learning_rate": 0.001,
+      "loss": 1.6018,
+      "step": 15353
+    },
+    {
+      "epoch": 0.6495473390303749,
+      "grad_norm": 1.505226492881775,
+      "learning_rate": 0.001,
+      "loss": 2.5792,
+      "step": 15354
+    },
+    {
+      "epoch": 0.6495896437938912,
+      "grad_norm": 0.18574289977550507,
+      "learning_rate": 0.001,
+      "loss": 2.5838,
+      "step": 15355
+    },
+    {
+      "epoch": 0.6496319485574076,
+      "grad_norm": 0.9261936545372009,
+      "learning_rate": 0.001,
+      "loss": 3.0768,
+      "step": 15356
+    },
+    {
+      "epoch": 0.649674253320924,
+      "grad_norm": 0.1780634969472885,
+      "learning_rate": 0.001,
+      "loss": 3.5627,
+      "step": 15357
+    },
+    {
+      "epoch": 0.6497165580844403,
+      "grad_norm": 0.5185316205024719,
+      "learning_rate": 0.001,
+      "loss": 2.1678,
+      "step": 15358
+    },
+    {
+      "epoch": 0.6497588628479567,
+      "grad_norm": 0.3285927176475525,
+      "learning_rate": 0.001,
+      "loss": 2.4195,
+      "step": 15359
+    },
+    {
+      "epoch": 0.6498011676114731,
+      "grad_norm": 0.2265651822090149,
+      "learning_rate": 0.001,
+      "loss": 2.5156,
+      "step": 15360
+    },
+    {
+      "epoch": 0.6498434723749894,
+      "grad_norm": 0.14902041852474213,
+      "learning_rate": 0.001,
+      "loss": 2.5237,
+      "step": 15361
+    },
+    {
+      "epoch": 0.6498857771385058,
+      "grad_norm": 0.20233316719532013,
+      "learning_rate": 0.001,
+      "loss": 2.4619,
+      "step": 15362
+    },
+    {
+      "epoch": 0.6499280819020221,
+      "grad_norm": 0.16044741868972778,
+      "learning_rate": 0.001,
+      "loss": 1.9778,
+      "step": 15363
+    },
+    {
+      "epoch": 0.6499703866655385,
+      "grad_norm": 0.22437353432178497,
+      "learning_rate": 0.001,
+      "loss": 2.704,
+      "step": 15364
+    },
+    {
+      "epoch": 0.6500126914290549,
+      "grad_norm": 0.1866353303194046,
+      "learning_rate": 0.001,
+      "loss": 1.7799,
+      "step": 15365
+    },
+    {
+      "epoch": 0.6500549961925712,
+      "grad_norm": 0.25852930545806885,
+      "learning_rate": 0.001,
+      "loss": 1.3778,
+      "step": 15366
+    },
+    {
+      "epoch": 0.6500973009560876,
+      "grad_norm": 0.23204196989536285,
+      "learning_rate": 0.001,
+      "loss": 2.7044,
+      "step": 15367
+    },
+    {
+      "epoch": 0.650139605719604,
+      "grad_norm": 0.19387875497341156,
+      "learning_rate": 0.001,
+      "loss": 3.1026,
+      "step": 15368
+    },
+    {
+      "epoch": 0.6501819104831204,
+      "grad_norm": 0.15012617409229279,
+      "learning_rate": 0.001,
+      "loss": 2.8182,
+      "step": 15369
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 0.1906086802482605,
+      "learning_rate": 0.001,
+      "loss": 2.2236,
+      "step": 15370
+    },
+    {
+      "epoch": 0.6502665200101532,
+      "grad_norm": 0.16298934817314148,
+      "learning_rate": 0.001,
+      "loss": 1.6815,
+      "step": 15371
+    },
+    {
+      "epoch": 0.6503088247736695,
+      "grad_norm": 6.248525142669678,
+      "learning_rate": 0.001,
+      "loss": 1.8015,
+      "step": 15372
+    },
+    {
+      "epoch": 0.6503511295371859,
+      "grad_norm": 0.29866117238998413,
+      "learning_rate": 0.001,
+      "loss": 2.1441,
+      "step": 15373
+    },
+    {
+      "epoch": 0.6503934343007023,
+      "grad_norm": 0.5905458331108093,
+      "learning_rate": 0.001,
+      "loss": 1.629,
+      "step": 15374
+    },
+    {
+      "epoch": 0.6504357390642186,
+      "grad_norm": 0.19776056706905365,
+      "learning_rate": 0.001,
+      "loss": 1.8649,
+      "step": 15375
+    },
+    {
+      "epoch": 0.650478043827735,
+      "grad_norm": 0.19329819083213806,
+      "learning_rate": 0.001,
+      "loss": 2.4082,
+      "step": 15376
+    },
+    {
+      "epoch": 0.6505203485912514,
+      "grad_norm": 0.2139400839805603,
+      "learning_rate": 0.001,
+      "loss": 3.2294,
+      "step": 15377
+    },
+    {
+      "epoch": 0.6505626533547677,
+      "grad_norm": 0.1699124574661255,
+      "learning_rate": 0.001,
+      "loss": 2.2416,
+      "step": 15378
+    },
+    {
+      "epoch": 0.6506049581182841,
+      "grad_norm": 0.17111091315746307,
+      "learning_rate": 0.001,
+      "loss": 1.564,
+      "step": 15379
+    },
+    {
+      "epoch": 0.6506472628818005,
+      "grad_norm": 0.1551768034696579,
+      "learning_rate": 0.001,
+      "loss": 2.4745,
+      "step": 15380
+    },
+    {
+      "epoch": 0.6506895676453168,
+      "grad_norm": 0.3416825830936432,
+      "learning_rate": 0.001,
+      "loss": 2.7327,
+      "step": 15381
+    },
+    {
+      "epoch": 0.6507318724088332,
+      "grad_norm": 0.1836758702993393,
+      "learning_rate": 0.001,
+      "loss": 2.3768,
+      "step": 15382
+    },
+    {
+      "epoch": 0.6507741771723496,
+      "grad_norm": 0.5745386481285095,
+      "learning_rate": 0.001,
+      "loss": 1.7781,
+      "step": 15383
+    },
+    {
+      "epoch": 0.6508164819358659,
+      "grad_norm": 0.21170569956302643,
+      "learning_rate": 0.001,
+      "loss": 2.6607,
+      "step": 15384
+    },
+    {
+      "epoch": 0.6508587866993824,
+      "grad_norm": 0.1727539747953415,
+      "learning_rate": 0.001,
+      "loss": 2.1116,
+      "step": 15385
+    },
+    {
+      "epoch": 0.6509010914628988,
+      "grad_norm": 1.874002456665039,
+      "learning_rate": 0.001,
+      "loss": 2.9078,
+      "step": 15386
+    },
+    {
+      "epoch": 0.6509433962264151,
+      "grad_norm": 0.21820873022079468,
+      "learning_rate": 0.001,
+      "loss": 1.8,
+      "step": 15387
+    },
+    {
+      "epoch": 0.6509857009899315,
+      "grad_norm": 6.197278022766113,
+      "learning_rate": 0.001,
+      "loss": 1.6494,
+      "step": 15388
+    },
+    {
+      "epoch": 0.6510280057534479,
+      "grad_norm": 0.16244280338287354,
+      "learning_rate": 0.001,
+      "loss": 1.9974,
+      "step": 15389
+    },
+    {
+      "epoch": 0.6510703105169642,
+      "grad_norm": 0.2022523730993271,
+      "learning_rate": 0.001,
+      "loss": 2.4231,
+      "step": 15390
+    },
+    {
+      "epoch": 0.6511126152804806,
+      "grad_norm": 0.19882678985595703,
+      "learning_rate": 0.001,
+      "loss": 2.0508,
+      "step": 15391
+    },
+    {
+      "epoch": 0.651154920043997,
+      "grad_norm": 0.5075106024742126,
+      "learning_rate": 0.001,
+      "loss": 1.6093,
+      "step": 15392
+    },
+    {
+      "epoch": 0.6511972248075133,
+      "grad_norm": 0.17667998373508453,
+      "learning_rate": 0.001,
+      "loss": 1.8422,
+      "step": 15393
+    },
+    {
+      "epoch": 0.6512395295710297,
+      "grad_norm": 0.28973835706710815,
+      "learning_rate": 0.001,
+      "loss": 3.3682,
+      "step": 15394
+    },
+    {
+      "epoch": 0.6512818343345461,
+      "grad_norm": 36.70282745361328,
+      "learning_rate": 0.001,
+      "loss": 2.7338,
+      "step": 15395
+    },
+    {
+      "epoch": 0.6513241390980624,
+      "grad_norm": 0.16966314613819122,
+      "learning_rate": 0.001,
+      "loss": 2.356,
+      "step": 15396
+    },
+    {
+      "epoch": 0.6513664438615788,
+      "grad_norm": 0.18758277595043182,
+      "learning_rate": 0.001,
+      "loss": 1.7575,
+      "step": 15397
+    },
+    {
+      "epoch": 0.6514087486250952,
+      "grad_norm": 0.16251905262470245,
+      "learning_rate": 0.001,
+      "loss": 1.4579,
+      "step": 15398
+    },
+    {
+      "epoch": 0.6514510533886115,
+      "grad_norm": 0.24462532997131348,
+      "learning_rate": 0.001,
+      "loss": 2.0842,
+      "step": 15399
+    },
+    {
+      "epoch": 0.6514933581521279,
+      "grad_norm": 0.29812654852867126,
+      "learning_rate": 0.001,
+      "loss": 2.6172,
+      "step": 15400
+    },
+    {
+      "epoch": 0.6515356629156444,
+      "grad_norm": 1.4498318433761597,
+      "learning_rate": 0.001,
+      "loss": 1.6447,
+      "step": 15401
+    },
+    {
+      "epoch": 0.6515779676791607,
+      "grad_norm": 0.19964168965816498,
+      "learning_rate": 0.001,
+      "loss": 1.4949,
+      "step": 15402
+    },
+    {
+      "epoch": 0.6516202724426771,
+      "grad_norm": 0.17301076650619507,
+      "learning_rate": 0.001,
+      "loss": 3.407,
+      "step": 15403
+    },
+    {
+      "epoch": 0.6516625772061935,
+      "grad_norm": 0.18589536845684052,
+      "learning_rate": 0.001,
+      "loss": 2.1407,
+      "step": 15404
+    },
+    {
+      "epoch": 0.6517048819697098,
+      "grad_norm": 0.18076921999454498,
+      "learning_rate": 0.001,
+      "loss": 2.0286,
+      "step": 15405
+    },
+    {
+      "epoch": 0.6517471867332262,
+      "grad_norm": 0.15782049298286438,
+      "learning_rate": 0.001,
+      "loss": 2.6584,
+      "step": 15406
+    },
+    {
+      "epoch": 0.6517894914967425,
+      "grad_norm": 1.653499960899353,
+      "learning_rate": 0.001,
+      "loss": 2.1578,
+      "step": 15407
+    },
+    {
+      "epoch": 0.6518317962602589,
+      "grad_norm": 0.20613858103752136,
+      "learning_rate": 0.001,
+      "loss": 1.7229,
+      "step": 15408
+    },
+    {
+      "epoch": 0.6518741010237753,
+      "grad_norm": 0.16704171895980835,
+      "learning_rate": 0.001,
+      "loss": 2.4252,
+      "step": 15409
+    },
+    {
+      "epoch": 0.6519164057872916,
+      "grad_norm": 0.3319103717803955,
+      "learning_rate": 0.001,
+      "loss": 2.2387,
+      "step": 15410
+    },
+    {
+      "epoch": 0.651958710550808,
+      "grad_norm": 0.1666289120912552,
+      "learning_rate": 0.001,
+      "loss": 1.8471,
+      "step": 15411
+    },
+    {
+      "epoch": 0.6520010153143244,
+      "grad_norm": 0.3717435896396637,
+      "learning_rate": 0.001,
+      "loss": 1.9818,
+      "step": 15412
+    },
+    {
+      "epoch": 0.6520433200778407,
+      "grad_norm": 0.21722985804080963,
+      "learning_rate": 0.001,
+      "loss": 1.9053,
+      "step": 15413
+    },
+    {
+      "epoch": 0.6520856248413571,
+      "grad_norm": 0.21333225071430206,
+      "learning_rate": 0.001,
+      "loss": 2.19,
+      "step": 15414
+    },
+    {
+      "epoch": 0.6521279296048735,
+      "grad_norm": 0.22371824085712433,
+      "learning_rate": 0.001,
+      "loss": 2.5068,
+      "step": 15415
+    },
+    {
+      "epoch": 0.6521702343683898,
+      "grad_norm": 0.1780281811952591,
+      "learning_rate": 0.001,
+      "loss": 1.9781,
+      "step": 15416
+    },
+    {
+      "epoch": 0.6522125391319062,
+      "grad_norm": 0.1755794882774353,
+      "learning_rate": 0.001,
+      "loss": 2.8328,
+      "step": 15417
+    },
+    {
+      "epoch": 0.6522548438954227,
+      "grad_norm": 1.0463812351226807,
+      "learning_rate": 0.001,
+      "loss": 2.8041,
+      "step": 15418
+    },
+    {
+      "epoch": 0.652297148658939,
+      "grad_norm": 0.16094234585762024,
+      "learning_rate": 0.001,
+      "loss": 1.7606,
+      "step": 15419
+    },
+    {
+      "epoch": 0.6523394534224554,
+      "grad_norm": 0.14978618919849396,
+      "learning_rate": 0.001,
+      "loss": 2.2326,
+      "step": 15420
+    },
+    {
+      "epoch": 0.6523817581859718,
+      "grad_norm": 0.19463518261909485,
+      "learning_rate": 0.001,
+      "loss": 2.1057,
+      "step": 15421
+    },
+    {
+      "epoch": 0.6524240629494881,
+      "grad_norm": 0.6183575391769409,
+      "learning_rate": 0.001,
+      "loss": 1.9251,
+      "step": 15422
+    },
+    {
+      "epoch": 0.6524663677130045,
+      "grad_norm": 0.15496648848056793,
+      "learning_rate": 0.001,
+      "loss": 1.7538,
+      "step": 15423
+    },
+    {
+      "epoch": 0.6525086724765209,
+      "grad_norm": 0.17894995212554932,
+      "learning_rate": 0.001,
+      "loss": 1.6924,
+      "step": 15424
+    },
+    {
+      "epoch": 0.6525509772400372,
+      "grad_norm": 0.16409684717655182,
+      "learning_rate": 0.001,
+      "loss": 1.995,
+      "step": 15425
+    },
+    {
+      "epoch": 0.6525932820035536,
+      "grad_norm": 0.15463796257972717,
+      "learning_rate": 0.001,
+      "loss": 1.8611,
+      "step": 15426
+    },
+    {
+      "epoch": 0.65263558676707,
+      "grad_norm": 0.14601118862628937,
+      "learning_rate": 0.001,
+      "loss": 2.8148,
+      "step": 15427
+    },
+    {
+      "epoch": 0.6526778915305863,
+      "grad_norm": 0.19336079061031342,
+      "learning_rate": 0.001,
+      "loss": 2.299,
+      "step": 15428
+    },
+    {
+      "epoch": 0.6527201962941027,
+      "grad_norm": 0.15365414321422577,
+      "learning_rate": 0.001,
+      "loss": 1.8484,
+      "step": 15429
+    },
+    {
+      "epoch": 0.6527625010576191,
+      "grad_norm": 0.17741386592388153,
+      "learning_rate": 0.001,
+      "loss": 1.9603,
+      "step": 15430
+    },
+    {
+      "epoch": 0.6528048058211354,
+      "grad_norm": 0.1832387000322342,
+      "learning_rate": 0.001,
+      "loss": 2.7062,
+      "step": 15431
+    },
+    {
+      "epoch": 0.6528471105846518,
+      "grad_norm": 0.16831880807876587,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 15432
+    },
+    {
+      "epoch": 0.6528894153481682,
+      "grad_norm": 0.2662883698940277,
+      "learning_rate": 0.001,
+      "loss": 1.614,
+      "step": 15433
+    },
+    {
+      "epoch": 0.6529317201116845,
+      "grad_norm": 0.15198548138141632,
+      "learning_rate": 0.001,
+      "loss": 1.7864,
+      "step": 15434
+    },
+    {
+      "epoch": 0.652974024875201,
+      "grad_norm": 0.1819853037595749,
+      "learning_rate": 0.001,
+      "loss": 1.8496,
+      "step": 15435
+    },
+    {
+      "epoch": 0.6530163296387174,
+      "grad_norm": 0.17152895033359528,
+      "learning_rate": 0.001,
+      "loss": 2.5133,
+      "step": 15436
+    },
+    {
+      "epoch": 0.6530586344022337,
+      "grad_norm": 0.16041690111160278,
+      "learning_rate": 0.001,
+      "loss": 1.5841,
+      "step": 15437
+    },
+    {
+      "epoch": 0.6531009391657501,
+      "grad_norm": 0.1594046652317047,
+      "learning_rate": 0.001,
+      "loss": 1.9821,
+      "step": 15438
+    },
+    {
+      "epoch": 0.6531432439292665,
+      "grad_norm": 0.45842236280441284,
+      "learning_rate": 0.001,
+      "loss": 2.6672,
+      "step": 15439
+    },
+    {
+      "epoch": 0.6531855486927828,
+      "grad_norm": 0.6107159852981567,
+      "learning_rate": 0.001,
+      "loss": 2.3053,
+      "step": 15440
+    },
+    {
+      "epoch": 0.6532278534562992,
+      "grad_norm": 0.19147630035877228,
+      "learning_rate": 0.001,
+      "loss": 1.6185,
+      "step": 15441
+    },
+    {
+      "epoch": 0.6532701582198156,
+      "grad_norm": 0.1753721982240677,
+      "learning_rate": 0.001,
+      "loss": 2.134,
+      "step": 15442
+    },
+    {
+      "epoch": 0.6533124629833319,
+      "grad_norm": 0.18597379326820374,
+      "learning_rate": 0.001,
+      "loss": 2.6756,
+      "step": 15443
+    },
+    {
+      "epoch": 0.6533547677468483,
+      "grad_norm": 0.17725729942321777,
+      "learning_rate": 0.001,
+      "loss": 2.1048,
+      "step": 15444
+    },
+    {
+      "epoch": 0.6533970725103647,
+      "grad_norm": 0.13884805142879486,
+      "learning_rate": 0.001,
+      "loss": 1.2568,
+      "step": 15445
+    },
+    {
+      "epoch": 0.653439377273881,
+      "grad_norm": 0.9566793441772461,
+      "learning_rate": 0.001,
+      "loss": 1.6837,
+      "step": 15446
+    },
+    {
+      "epoch": 0.6534816820373974,
+      "grad_norm": 0.2636922597885132,
+      "learning_rate": 0.001,
+      "loss": 2.301,
+      "step": 15447
+    },
+    {
+      "epoch": 0.6535239868009138,
+      "grad_norm": 0.20910915732383728,
+      "learning_rate": 0.001,
+      "loss": 2.7653,
+      "step": 15448
+    },
+    {
+      "epoch": 0.6535662915644301,
+      "grad_norm": 0.38089674711227417,
+      "learning_rate": 0.001,
+      "loss": 1.8104,
+      "step": 15449
+    },
+    {
+      "epoch": 0.6536085963279465,
+      "grad_norm": 0.18689171969890594,
+      "learning_rate": 0.001,
+      "loss": 1.6432,
+      "step": 15450
+    },
+    {
+      "epoch": 0.6536509010914628,
+      "grad_norm": 0.19268843531608582,
+      "learning_rate": 0.001,
+      "loss": 1.7838,
+      "step": 15451
+    },
+    {
+      "epoch": 0.6536932058549793,
+      "grad_norm": 0.17988325655460358,
+      "learning_rate": 0.001,
+      "loss": 2.2731,
+      "step": 15452
+    },
+    {
+      "epoch": 0.6537355106184957,
+      "grad_norm": 0.17474591732025146,
+      "learning_rate": 0.001,
+      "loss": 1.865,
+      "step": 15453
+    },
+    {
+      "epoch": 0.653777815382012,
+      "grad_norm": 0.16267631947994232,
+      "learning_rate": 0.001,
+      "loss": 1.614,
+      "step": 15454
+    },
+    {
+      "epoch": 0.6538201201455284,
+      "grad_norm": 0.27630341053009033,
+      "learning_rate": 0.001,
+      "loss": 1.9749,
+      "step": 15455
+    },
+    {
+      "epoch": 0.6538624249090448,
+      "grad_norm": 0.16530779004096985,
+      "learning_rate": 0.001,
+      "loss": 1.5857,
+      "step": 15456
+    },
+    {
+      "epoch": 0.6539047296725611,
+      "grad_norm": 0.7646900415420532,
+      "learning_rate": 0.001,
+      "loss": 2.128,
+      "step": 15457
+    },
+    {
+      "epoch": 0.6539470344360775,
+      "grad_norm": 0.21980629861354828,
+      "learning_rate": 0.001,
+      "loss": 2.7452,
+      "step": 15458
+    },
+    {
+      "epoch": 0.6539893391995939,
+      "grad_norm": 0.18467316031455994,
+      "learning_rate": 0.001,
+      "loss": 3.4153,
+      "step": 15459
+    },
+    {
+      "epoch": 0.6540316439631102,
+      "grad_norm": 0.1603761464357376,
+      "learning_rate": 0.001,
+      "loss": 2.3351,
+      "step": 15460
+    },
+    {
+      "epoch": 0.6540739487266266,
+      "grad_norm": 0.19741065800189972,
+      "learning_rate": 0.001,
+      "loss": 1.9429,
+      "step": 15461
+    },
+    {
+      "epoch": 0.654116253490143,
+      "grad_norm": 0.2010018527507782,
+      "learning_rate": 0.001,
+      "loss": 2.3873,
+      "step": 15462
+    },
+    {
+      "epoch": 0.6541585582536593,
+      "grad_norm": 0.1678660362958908,
+      "learning_rate": 0.001,
+      "loss": 1.8596,
+      "step": 15463
+    },
+    {
+      "epoch": 0.6542008630171757,
+      "grad_norm": 0.15577636659145355,
+      "learning_rate": 0.001,
+      "loss": 2.438,
+      "step": 15464
+    },
+    {
+      "epoch": 0.6542431677806921,
+      "grad_norm": 0.16869132220745087,
+      "learning_rate": 0.001,
+      "loss": 2.0241,
+      "step": 15465
+    },
+    {
+      "epoch": 0.6542854725442084,
+      "grad_norm": 0.16793528199195862,
+      "learning_rate": 0.001,
+      "loss": 2.3212,
+      "step": 15466
+    },
+    {
+      "epoch": 0.6543277773077248,
+      "grad_norm": 0.14412140846252441,
+      "learning_rate": 0.001,
+      "loss": 2.1674,
+      "step": 15467
+    },
+    {
+      "epoch": 0.6543700820712413,
+      "grad_norm": 0.17492231726646423,
+      "learning_rate": 0.001,
+      "loss": 1.7155,
+      "step": 15468
+    },
+    {
+      "epoch": 0.6544123868347576,
+      "grad_norm": 0.16392190754413605,
+      "learning_rate": 0.001,
+      "loss": 1.5997,
+      "step": 15469
+    },
+    {
+      "epoch": 0.654454691598274,
+      "grad_norm": 0.15477705001831055,
+      "learning_rate": 0.001,
+      "loss": 2.3755,
+      "step": 15470
+    },
+    {
+      "epoch": 0.6544969963617904,
+      "grad_norm": 9.079736709594727,
+      "learning_rate": 0.001,
+      "loss": 2.5364,
+      "step": 15471
+    },
+    {
+      "epoch": 0.6545393011253067,
+      "grad_norm": 0.2828308641910553,
+      "learning_rate": 0.001,
+      "loss": 1.9828,
+      "step": 15472
+    },
+    {
+      "epoch": 0.6545816058888231,
+      "grad_norm": 0.2948327958583832,
+      "learning_rate": 0.001,
+      "loss": 2.5522,
+      "step": 15473
+    },
+    {
+      "epoch": 0.6546239106523395,
+      "grad_norm": 0.2276574969291687,
+      "learning_rate": 0.001,
+      "loss": 2.3777,
+      "step": 15474
+    },
+    {
+      "epoch": 0.6546662154158558,
+      "grad_norm": 0.23238156735897064,
+      "learning_rate": 0.001,
+      "loss": 1.8858,
+      "step": 15475
+    },
+    {
+      "epoch": 0.6547085201793722,
+      "grad_norm": 0.2526158094406128,
+      "learning_rate": 0.001,
+      "loss": 2.4122,
+      "step": 15476
+    },
+    {
+      "epoch": 0.6547508249428886,
+      "grad_norm": 0.1535889059305191,
+      "learning_rate": 0.001,
+      "loss": 1.8641,
+      "step": 15477
+    },
+    {
+      "epoch": 0.6547931297064049,
+      "grad_norm": 0.9799718856811523,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 15478
+    },
+    {
+      "epoch": 0.6548354344699213,
+      "grad_norm": 0.18154305219650269,
+      "learning_rate": 0.001,
+      "loss": 2.3197,
+      "step": 15479
+    },
+    {
+      "epoch": 0.6548777392334377,
+      "grad_norm": 0.1667226254940033,
+      "learning_rate": 0.001,
+      "loss": 2.3554,
+      "step": 15480
+    },
+    {
+      "epoch": 0.654920043996954,
+      "grad_norm": 1.3010785579681396,
+      "learning_rate": 0.001,
+      "loss": 2.7386,
+      "step": 15481
+    },
+    {
+      "epoch": 0.6549623487604704,
+      "grad_norm": 0.19829553365707397,
+      "learning_rate": 0.001,
+      "loss": 1.6301,
+      "step": 15482
+    },
+    {
+      "epoch": 0.6550046535239868,
+      "grad_norm": 0.2689056098461151,
+      "learning_rate": 0.001,
+      "loss": 3.029,
+      "step": 15483
+    },
+    {
+      "epoch": 0.6550469582875031,
+      "grad_norm": 0.33899521827697754,
+      "learning_rate": 0.001,
+      "loss": 2.2413,
+      "step": 15484
+    },
+    {
+      "epoch": 0.6550892630510196,
+      "grad_norm": 0.26710861921310425,
+      "learning_rate": 0.001,
+      "loss": 2.9207,
+      "step": 15485
+    },
+    {
+      "epoch": 0.655131567814536,
+      "grad_norm": 0.22953356802463531,
+      "learning_rate": 0.001,
+      "loss": 2.1988,
+      "step": 15486
+    },
+    {
+      "epoch": 0.6551738725780523,
+      "grad_norm": 0.2150760293006897,
+      "learning_rate": 0.001,
+      "loss": 2.098,
+      "step": 15487
+    },
+    {
+      "epoch": 0.6552161773415687,
+      "grad_norm": 0.20908361673355103,
+      "learning_rate": 0.001,
+      "loss": 2.7129,
+      "step": 15488
+    },
+    {
+      "epoch": 0.6552584821050851,
+      "grad_norm": 0.20225393772125244,
+      "learning_rate": 0.001,
+      "loss": 2.7402,
+      "step": 15489
+    },
+    {
+      "epoch": 0.6553007868686014,
+      "grad_norm": 0.2164647877216339,
+      "learning_rate": 0.001,
+      "loss": 2.245,
+      "step": 15490
+    },
+    {
+      "epoch": 0.6553430916321178,
+      "grad_norm": 0.17477239668369293,
+      "learning_rate": 0.001,
+      "loss": 2.915,
+      "step": 15491
+    },
+    {
+      "epoch": 0.6553853963956342,
+      "grad_norm": 1.133888602256775,
+      "learning_rate": 0.001,
+      "loss": 1.5993,
+      "step": 15492
+    },
+    {
+      "epoch": 0.6554277011591505,
+      "grad_norm": 1.6383650302886963,
+      "learning_rate": 0.001,
+      "loss": 1.6001,
+      "step": 15493
+    },
+    {
+      "epoch": 0.6554700059226669,
+      "grad_norm": 0.419475257396698,
+      "learning_rate": 0.001,
+      "loss": 1.8772,
+      "step": 15494
+    },
+    {
+      "epoch": 0.6555123106861833,
+      "grad_norm": 0.18061023950576782,
+      "learning_rate": 0.001,
+      "loss": 2.0269,
+      "step": 15495
+    },
+    {
+      "epoch": 0.6555546154496996,
+      "grad_norm": 0.3554385304450989,
+      "learning_rate": 0.001,
+      "loss": 2.9187,
+      "step": 15496
+    },
+    {
+      "epoch": 0.655596920213216,
+      "grad_norm": 0.39084017276763916,
+      "learning_rate": 0.001,
+      "loss": 3.1199,
+      "step": 15497
+    },
+    {
+      "epoch": 0.6556392249767323,
+      "grad_norm": 0.168495312333107,
+      "learning_rate": 0.001,
+      "loss": 3.786,
+      "step": 15498
+    },
+    {
+      "epoch": 0.6556815297402487,
+      "grad_norm": 0.6152557134628296,
+      "learning_rate": 0.001,
+      "loss": 2.2492,
+      "step": 15499
+    },
+    {
+      "epoch": 0.6557238345037651,
+      "grad_norm": 0.19804157316684723,
+      "learning_rate": 0.001,
+      "loss": 2.449,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6557661392672814,
+      "grad_norm": 0.1926928460597992,
+      "learning_rate": 0.001,
+      "loss": 2.3249,
+      "step": 15501
+    },
+    {
+      "epoch": 0.6558084440307979,
+      "grad_norm": 0.1641806960105896,
+      "learning_rate": 0.001,
+      "loss": 1.742,
+      "step": 15502
+    },
+    {
+      "epoch": 0.6558507487943143,
+      "grad_norm": 0.1997319757938385,
+      "learning_rate": 0.001,
+      "loss": 1.7896,
+      "step": 15503
+    },
+    {
+      "epoch": 0.6558930535578306,
+      "grad_norm": 0.2186889350414276,
+      "learning_rate": 0.001,
+      "loss": 2.4833,
+      "step": 15504
+    },
+    {
+      "epoch": 0.655935358321347,
+      "grad_norm": 0.16924524307250977,
+      "learning_rate": 0.001,
+      "loss": 2.279,
+      "step": 15505
+    },
+    {
+      "epoch": 0.6559776630848634,
+      "grad_norm": 0.17367133498191833,
+      "learning_rate": 0.001,
+      "loss": 2.3269,
+      "step": 15506
+    },
+    {
+      "epoch": 0.6560199678483797,
+      "grad_norm": 0.1684117615222931,
+      "learning_rate": 0.001,
+      "loss": 2.4659,
+      "step": 15507
+    },
+    {
+      "epoch": 0.6560622726118961,
+      "grad_norm": 8.400760650634766,
+      "learning_rate": 0.001,
+      "loss": 2.1449,
+      "step": 15508
+    },
+    {
+      "epoch": 0.6561045773754125,
+      "grad_norm": 0.17332401871681213,
+      "learning_rate": 0.001,
+      "loss": 1.8829,
+      "step": 15509
+    },
+    {
+      "epoch": 0.6561468821389288,
+      "grad_norm": 0.4135810136795044,
+      "learning_rate": 0.001,
+      "loss": 2.5169,
+      "step": 15510
+    },
+    {
+      "epoch": 0.6561891869024452,
+      "grad_norm": 0.15025272965431213,
+      "learning_rate": 0.001,
+      "loss": 1.9536,
+      "step": 15511
+    },
+    {
+      "epoch": 0.6562314916659616,
+      "grad_norm": 0.21531356871128082,
+      "learning_rate": 0.001,
+      "loss": 2.1685,
+      "step": 15512
+    },
+    {
+      "epoch": 0.6562737964294779,
+      "grad_norm": 0.20930558443069458,
+      "learning_rate": 0.001,
+      "loss": 2.3867,
+      "step": 15513
+    },
+    {
+      "epoch": 0.6563161011929943,
+      "grad_norm": 0.1821804642677307,
+      "learning_rate": 0.001,
+      "loss": 1.4883,
+      "step": 15514
+    },
+    {
+      "epoch": 0.6563584059565107,
+      "grad_norm": 0.2047373205423355,
+      "learning_rate": 0.001,
+      "loss": 3.2604,
+      "step": 15515
+    },
+    {
+      "epoch": 0.656400710720027,
+      "grad_norm": 0.1793907731771469,
+      "learning_rate": 0.001,
+      "loss": 2.5151,
+      "step": 15516
+    },
+    {
+      "epoch": 0.6564430154835434,
+      "grad_norm": 0.20762686431407928,
+      "learning_rate": 0.001,
+      "loss": 3.577,
+      "step": 15517
+    },
+    {
+      "epoch": 0.6564853202470599,
+      "grad_norm": 0.20441284775733948,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 15518
+    },
+    {
+      "epoch": 0.6565276250105762,
+      "grad_norm": 0.3753068447113037,
+      "learning_rate": 0.001,
+      "loss": 3.6612,
+      "step": 15519
+    },
+    {
+      "epoch": 0.6565699297740926,
+      "grad_norm": 0.1667603701353073,
+      "learning_rate": 0.001,
+      "loss": 1.8838,
+      "step": 15520
+    },
+    {
+      "epoch": 0.656612234537609,
+      "grad_norm": 16.620149612426758,
+      "learning_rate": 0.001,
+      "loss": 2.5339,
+      "step": 15521
+    },
+    {
+      "epoch": 0.6566545393011253,
+      "grad_norm": 0.17226529121398926,
+      "learning_rate": 0.001,
+      "loss": 2.9333,
+      "step": 15522
+    },
+    {
+      "epoch": 0.6566968440646417,
+      "grad_norm": 0.1851300597190857,
+      "learning_rate": 0.001,
+      "loss": 2.8224,
+      "step": 15523
+    },
+    {
+      "epoch": 0.6567391488281581,
+      "grad_norm": 0.1961529552936554,
+      "learning_rate": 0.001,
+      "loss": 2.389,
+      "step": 15524
+    },
+    {
+      "epoch": 0.6567814535916744,
+      "grad_norm": 0.2944627106189728,
+      "learning_rate": 0.001,
+      "loss": 2.1613,
+      "step": 15525
+    },
+    {
+      "epoch": 0.6568237583551908,
+      "grad_norm": 0.4502488374710083,
+      "learning_rate": 0.001,
+      "loss": 1.678,
+      "step": 15526
+    },
+    {
+      "epoch": 0.6568660631187072,
+      "grad_norm": 0.19314773380756378,
+      "learning_rate": 0.001,
+      "loss": 2.8654,
+      "step": 15527
+    },
+    {
+      "epoch": 0.6569083678822235,
+      "grad_norm": 0.19277732074260712,
+      "learning_rate": 0.001,
+      "loss": 3.3764,
+      "step": 15528
+    },
+    {
+      "epoch": 0.6569506726457399,
+      "grad_norm": 0.2084587961435318,
+      "learning_rate": 0.001,
+      "loss": 3.0331,
+      "step": 15529
+    },
+    {
+      "epoch": 0.6569929774092563,
+      "grad_norm": 0.24447475373744965,
+      "learning_rate": 0.001,
+      "loss": 2.4765,
+      "step": 15530
+    },
+    {
+      "epoch": 0.6570352821727726,
+      "grad_norm": 0.4201889932155609,
+      "learning_rate": 0.001,
+      "loss": 1.8756,
+      "step": 15531
+    },
+    {
+      "epoch": 0.657077586936289,
+      "grad_norm": 0.3729053735733032,
+      "learning_rate": 0.001,
+      "loss": 1.9558,
+      "step": 15532
+    },
+    {
+      "epoch": 0.6571198916998054,
+      "grad_norm": 0.21242676675319672,
+      "learning_rate": 0.001,
+      "loss": 2.0069,
+      "step": 15533
+    },
+    {
+      "epoch": 0.6571621964633217,
+      "grad_norm": 0.1712314486503601,
+      "learning_rate": 0.001,
+      "loss": 1.3812,
+      "step": 15534
+    },
+    {
+      "epoch": 0.6572045012268382,
+      "grad_norm": 0.1673876792192459,
+      "learning_rate": 0.001,
+      "loss": 1.1909,
+      "step": 15535
+    },
+    {
+      "epoch": 0.6572468059903546,
+      "grad_norm": 0.15203256905078888,
+      "learning_rate": 0.001,
+      "loss": 1.6762,
+      "step": 15536
+    },
+    {
+      "epoch": 0.6572891107538709,
+      "grad_norm": 0.15472784638404846,
+      "learning_rate": 0.001,
+      "loss": 1.5563,
+      "step": 15537
+    },
+    {
+      "epoch": 0.6573314155173873,
+      "grad_norm": 0.22867830097675323,
+      "learning_rate": 0.001,
+      "loss": 2.4964,
+      "step": 15538
+    },
+    {
+      "epoch": 0.6573737202809037,
+      "grad_norm": 1.3349720239639282,
+      "learning_rate": 0.001,
+      "loss": 1.7851,
+      "step": 15539
+    },
+    {
+      "epoch": 0.65741602504442,
+      "grad_norm": 3.2347517013549805,
+      "learning_rate": 0.001,
+      "loss": 1.8873,
+      "step": 15540
+    },
+    {
+      "epoch": 0.6574583298079364,
+      "grad_norm": 1.0401722192764282,
+      "learning_rate": 0.001,
+      "loss": 1.982,
+      "step": 15541
+    },
+    {
+      "epoch": 0.6575006345714527,
+      "grad_norm": 0.4150663912296295,
+      "learning_rate": 0.001,
+      "loss": 2.4127,
+      "step": 15542
+    },
+    {
+      "epoch": 0.6575429393349691,
+      "grad_norm": 0.28380629420280457,
+      "learning_rate": 0.001,
+      "loss": 1.7512,
+      "step": 15543
+    },
+    {
+      "epoch": 0.6575852440984855,
+      "grad_norm": 0.1765417903661728,
+      "learning_rate": 0.001,
+      "loss": 2.7307,
+      "step": 15544
+    },
+    {
+      "epoch": 0.6576275488620018,
+      "grad_norm": 0.22841008007526398,
+      "learning_rate": 0.001,
+      "loss": 2.1964,
+      "step": 15545
+    },
+    {
+      "epoch": 0.6576698536255182,
+      "grad_norm": 0.3355230391025543,
+      "learning_rate": 0.001,
+      "loss": 2.1679,
+      "step": 15546
+    },
+    {
+      "epoch": 0.6577121583890346,
+      "grad_norm": 0.16296561062335968,
+      "learning_rate": 0.001,
+      "loss": 1.8227,
+      "step": 15547
+    },
+    {
+      "epoch": 0.6577544631525509,
+      "grad_norm": 0.2643844485282898,
+      "learning_rate": 0.001,
+      "loss": 2.0391,
+      "step": 15548
+    },
+    {
+      "epoch": 0.6577967679160673,
+      "grad_norm": 0.19168947637081146,
+      "learning_rate": 0.001,
+      "loss": 1.9581,
+      "step": 15549
+    },
+    {
+      "epoch": 0.6578390726795837,
+      "grad_norm": 0.17555919289588928,
+      "learning_rate": 0.001,
+      "loss": 1.7128,
+      "step": 15550
+    },
+    {
+      "epoch": 0.6578813774431,
+      "grad_norm": 0.21022935211658478,
+      "learning_rate": 0.001,
+      "loss": 2.0966,
+      "step": 15551
+    },
+    {
+      "epoch": 0.6579236822066165,
+      "grad_norm": 0.17026807367801666,
+      "learning_rate": 0.001,
+      "loss": 1.7935,
+      "step": 15552
+    },
+    {
+      "epoch": 0.6579659869701329,
+      "grad_norm": 9.642647743225098,
+      "learning_rate": 0.001,
+      "loss": 1.9757,
+      "step": 15553
+    },
+    {
+      "epoch": 0.6580082917336492,
+      "grad_norm": 0.1723700612783432,
+      "learning_rate": 0.001,
+      "loss": 2.0512,
+      "step": 15554
+    },
+    {
+      "epoch": 0.6580505964971656,
+      "grad_norm": 0.16185259819030762,
+      "learning_rate": 0.001,
+      "loss": 2.1963,
+      "step": 15555
+    },
+    {
+      "epoch": 0.658092901260682,
+      "grad_norm": 0.17926618456840515,
+      "learning_rate": 0.001,
+      "loss": 2.3246,
+      "step": 15556
+    },
+    {
+      "epoch": 0.6581352060241983,
+      "grad_norm": 0.31012213230133057,
+      "learning_rate": 0.001,
+      "loss": 2.4251,
+      "step": 15557
+    },
+    {
+      "epoch": 0.6581775107877147,
+      "grad_norm": 0.2613414525985718,
+      "learning_rate": 0.001,
+      "loss": 3.0419,
+      "step": 15558
+    },
+    {
+      "epoch": 0.6582198155512311,
+      "grad_norm": 0.17412632703781128,
+      "learning_rate": 0.001,
+      "loss": 2.2753,
+      "step": 15559
+    },
+    {
+      "epoch": 0.6582621203147474,
+      "grad_norm": 0.40548646450042725,
+      "learning_rate": 0.001,
+      "loss": 1.9838,
+      "step": 15560
+    },
+    {
+      "epoch": 0.6583044250782638,
+      "grad_norm": 0.16524159908294678,
+      "learning_rate": 0.001,
+      "loss": 2.8922,
+      "step": 15561
+    },
+    {
+      "epoch": 0.6583467298417802,
+      "grad_norm": 0.17522113025188446,
+      "learning_rate": 0.001,
+      "loss": 1.704,
+      "step": 15562
+    },
+    {
+      "epoch": 0.6583890346052965,
+      "grad_norm": 0.1847570687532425,
+      "learning_rate": 0.001,
+      "loss": 2.1191,
+      "step": 15563
+    },
+    {
+      "epoch": 0.6584313393688129,
+      "grad_norm": 0.16982965171337128,
+      "learning_rate": 0.001,
+      "loss": 1.4231,
+      "step": 15564
+    },
+    {
+      "epoch": 0.6584736441323293,
+      "grad_norm": 0.16816236078739166,
+      "learning_rate": 0.001,
+      "loss": 2.9246,
+      "step": 15565
+    },
+    {
+      "epoch": 0.6585159488958456,
+      "grad_norm": 0.18339279294013977,
+      "learning_rate": 0.001,
+      "loss": 2.7366,
+      "step": 15566
+    },
+    {
+      "epoch": 0.658558253659362,
+      "grad_norm": 0.15726692974567413,
+      "learning_rate": 0.001,
+      "loss": 1.7157,
+      "step": 15567
+    },
+    {
+      "epoch": 0.6586005584228785,
+      "grad_norm": 0.1590546816587448,
+      "learning_rate": 0.001,
+      "loss": 1.957,
+      "step": 15568
+    },
+    {
+      "epoch": 0.6586428631863948,
+      "grad_norm": 0.28861668705940247,
+      "learning_rate": 0.001,
+      "loss": 2.2768,
+      "step": 15569
+    },
+    {
+      "epoch": 0.6586851679499112,
+      "grad_norm": 0.15033060312271118,
+      "learning_rate": 0.001,
+      "loss": 1.8739,
+      "step": 15570
+    },
+    {
+      "epoch": 0.6587274727134276,
+      "grad_norm": 0.17307975888252258,
+      "learning_rate": 0.001,
+      "loss": 1.7251,
+      "step": 15571
+    },
+    {
+      "epoch": 0.6587697774769439,
+      "grad_norm": 0.4346740245819092,
+      "learning_rate": 0.001,
+      "loss": 3.3548,
+      "step": 15572
+    },
+    {
+      "epoch": 0.6588120822404603,
+      "grad_norm": 0.24258974194526672,
+      "learning_rate": 0.001,
+      "loss": 2.4327,
+      "step": 15573
+    },
+    {
+      "epoch": 0.6588543870039767,
+      "grad_norm": 0.1866195797920227,
+      "learning_rate": 0.001,
+      "loss": 2.1185,
+      "step": 15574
+    },
+    {
+      "epoch": 0.658896691767493,
+      "grad_norm": 0.15456928312778473,
+      "learning_rate": 0.001,
+      "loss": 1.487,
+      "step": 15575
+    },
+    {
+      "epoch": 0.6589389965310094,
+      "grad_norm": 0.18687579035758972,
+      "learning_rate": 0.001,
+      "loss": 2.329,
+      "step": 15576
+    },
+    {
+      "epoch": 0.6589813012945258,
+      "grad_norm": 0.16042332351207733,
+      "learning_rate": 0.001,
+      "loss": 2.1489,
+      "step": 15577
+    },
+    {
+      "epoch": 0.6590236060580421,
+      "grad_norm": 0.15789800882339478,
+      "learning_rate": 0.001,
+      "loss": 2.7708,
+      "step": 15578
+    },
+    {
+      "epoch": 0.6590659108215585,
+      "grad_norm": 0.4976028501987457,
+      "learning_rate": 0.001,
+      "loss": 1.66,
+      "step": 15579
+    },
+    {
+      "epoch": 0.6591082155850749,
+      "grad_norm": 0.18985185027122498,
+      "learning_rate": 0.001,
+      "loss": 1.7931,
+      "step": 15580
+    },
+    {
+      "epoch": 0.6591505203485912,
+      "grad_norm": 0.18553532660007477,
+      "learning_rate": 0.001,
+      "loss": 2.2777,
+      "step": 15581
+    },
+    {
+      "epoch": 0.6591928251121076,
+      "grad_norm": 0.15861479938030243,
+      "learning_rate": 0.001,
+      "loss": 1.9693,
+      "step": 15582
+    },
+    {
+      "epoch": 0.659235129875624,
+      "grad_norm": 0.34126266837120056,
+      "learning_rate": 0.001,
+      "loss": 1.8181,
+      "step": 15583
+    },
+    {
+      "epoch": 0.6592774346391403,
+      "grad_norm": 3.384877920150757,
+      "learning_rate": 0.001,
+      "loss": 1.9545,
+      "step": 15584
+    },
+    {
+      "epoch": 0.6593197394026568,
+      "grad_norm": 0.18254320323467255,
+      "learning_rate": 0.001,
+      "loss": 2.2612,
+      "step": 15585
+    },
+    {
+      "epoch": 0.6593620441661731,
+      "grad_norm": 0.17780011892318726,
+      "learning_rate": 0.001,
+      "loss": 2.3693,
+      "step": 15586
+    },
+    {
+      "epoch": 0.6594043489296895,
+      "grad_norm": 0.2957765460014343,
+      "learning_rate": 0.001,
+      "loss": 2.8228,
+      "step": 15587
+    },
+    {
+      "epoch": 0.6594466536932059,
+      "grad_norm": 0.20004132390022278,
+      "learning_rate": 0.001,
+      "loss": 2.2395,
+      "step": 15588
+    },
+    {
+      "epoch": 0.6594889584567222,
+      "grad_norm": 0.16427592933177948,
+      "learning_rate": 0.001,
+      "loss": 1.9505,
+      "step": 15589
+    },
+    {
+      "epoch": 0.6595312632202386,
+      "grad_norm": 0.3645588755607605,
+      "learning_rate": 0.001,
+      "loss": 2.4585,
+      "step": 15590
+    },
+    {
+      "epoch": 0.659573567983755,
+      "grad_norm": 0.1937217265367508,
+      "learning_rate": 0.001,
+      "loss": 1.9487,
+      "step": 15591
+    },
+    {
+      "epoch": 0.6596158727472713,
+      "grad_norm": 0.5016273260116577,
+      "learning_rate": 0.001,
+      "loss": 3.591,
+      "step": 15592
+    },
+    {
+      "epoch": 0.6596581775107877,
+      "grad_norm": 0.15664246678352356,
+      "learning_rate": 0.001,
+      "loss": 1.5117,
+      "step": 15593
+    },
+    {
+      "epoch": 0.6597004822743041,
+      "grad_norm": 1.7184746265411377,
+      "learning_rate": 0.001,
+      "loss": 2.2226,
+      "step": 15594
+    },
+    {
+      "epoch": 0.6597427870378204,
+      "grad_norm": 1.1707103252410889,
+      "learning_rate": 0.001,
+      "loss": 2.213,
+      "step": 15595
+    },
+    {
+      "epoch": 0.6597850918013368,
+      "grad_norm": 0.1881413757801056,
+      "learning_rate": 0.001,
+      "loss": 2.1988,
+      "step": 15596
+    },
+    {
+      "epoch": 0.6598273965648532,
+      "grad_norm": 0.1818539947271347,
+      "learning_rate": 0.001,
+      "loss": 2.2715,
+      "step": 15597
+    },
+    {
+      "epoch": 0.6598697013283695,
+      "grad_norm": 0.5400058627128601,
+      "learning_rate": 0.001,
+      "loss": 1.6366,
+      "step": 15598
+    },
+    {
+      "epoch": 0.6599120060918859,
+      "grad_norm": 0.3063301742076874,
+      "learning_rate": 0.001,
+      "loss": 2.4906,
+      "step": 15599
+    },
+    {
+      "epoch": 0.6599543108554023,
+      "grad_norm": 0.2518875300884247,
+      "learning_rate": 0.001,
+      "loss": 2.2794,
+      "step": 15600
+    },
+    {
+      "epoch": 0.6599966156189186,
+      "grad_norm": 0.17568255960941315,
+      "learning_rate": 0.001,
+      "loss": 2.1365,
+      "step": 15601
+    },
+    {
+      "epoch": 0.6600389203824351,
+      "grad_norm": 0.17697425186634064,
+      "learning_rate": 0.001,
+      "loss": 2.8336,
+      "step": 15602
+    },
+    {
+      "epoch": 0.6600812251459515,
+      "grad_norm": 1.3957675695419312,
+      "learning_rate": 0.001,
+      "loss": 2.2867,
+      "step": 15603
+    },
+    {
+      "epoch": 0.6601235299094678,
+      "grad_norm": 0.15436089038848877,
+      "learning_rate": 0.001,
+      "loss": 1.9459,
+      "step": 15604
+    },
+    {
+      "epoch": 0.6601658346729842,
+      "grad_norm": 0.16911160945892334,
+      "learning_rate": 0.001,
+      "loss": 2.0371,
+      "step": 15605
+    },
+    {
+      "epoch": 0.6602081394365006,
+      "grad_norm": 0.2687240242958069,
+      "learning_rate": 0.001,
+      "loss": 2.4646,
+      "step": 15606
+    },
+    {
+      "epoch": 0.6602504442000169,
+      "grad_norm": 0.2176024168729782,
+      "learning_rate": 0.001,
+      "loss": 2.0944,
+      "step": 15607
+    },
+    {
+      "epoch": 0.6602927489635333,
+      "grad_norm": 0.4650316536426544,
+      "learning_rate": 0.001,
+      "loss": 1.8368,
+      "step": 15608
+    },
+    {
+      "epoch": 0.6603350537270497,
+      "grad_norm": 0.21563652157783508,
+      "learning_rate": 0.001,
+      "loss": 3.3112,
+      "step": 15609
+    },
+    {
+      "epoch": 0.660377358490566,
+      "grad_norm": 0.14292295277118683,
+      "learning_rate": 0.001,
+      "loss": 2.6444,
+      "step": 15610
+    },
+    {
+      "epoch": 0.6604196632540824,
+      "grad_norm": 0.9657493233680725,
+      "learning_rate": 0.001,
+      "loss": 2.7724,
+      "step": 15611
+    },
+    {
+      "epoch": 0.6604619680175988,
+      "grad_norm": 0.31469717621803284,
+      "learning_rate": 0.001,
+      "loss": 2.1013,
+      "step": 15612
+    },
+    {
+      "epoch": 0.6605042727811151,
+      "grad_norm": 0.16225768625736237,
+      "learning_rate": 0.001,
+      "loss": 1.816,
+      "step": 15613
+    },
+    {
+      "epoch": 0.6605465775446315,
+      "grad_norm": 0.19633138179779053,
+      "learning_rate": 0.001,
+      "loss": 1.9044,
+      "step": 15614
+    },
+    {
+      "epoch": 0.6605888823081479,
+      "grad_norm": 0.18959277868270874,
+      "learning_rate": 0.001,
+      "loss": 2.2034,
+      "step": 15615
+    },
+    {
+      "epoch": 0.6606311870716642,
+      "grad_norm": 0.25994089245796204,
+      "learning_rate": 0.001,
+      "loss": 2.6884,
+      "step": 15616
+    },
+    {
+      "epoch": 0.6606734918351806,
+      "grad_norm": 0.17219536006450653,
+      "learning_rate": 0.001,
+      "loss": 2.4676,
+      "step": 15617
+    },
+    {
+      "epoch": 0.6607157965986971,
+      "grad_norm": 0.15563371777534485,
+      "learning_rate": 0.001,
+      "loss": 2.8265,
+      "step": 15618
+    },
+    {
+      "epoch": 0.6607581013622134,
+      "grad_norm": 0.1527656465768814,
+      "learning_rate": 0.001,
+      "loss": 1.3507,
+      "step": 15619
+    },
+    {
+      "epoch": 0.6608004061257298,
+      "grad_norm": 0.2252880334854126,
+      "learning_rate": 0.001,
+      "loss": 2.4293,
+      "step": 15620
+    },
+    {
+      "epoch": 0.6608427108892462,
+      "grad_norm": 0.20225971937179565,
+      "learning_rate": 0.001,
+      "loss": 2.1757,
+      "step": 15621
+    },
+    {
+      "epoch": 0.6608850156527625,
+      "grad_norm": 0.4528272747993469,
+      "learning_rate": 0.001,
+      "loss": 1.8446,
+      "step": 15622
+    },
+    {
+      "epoch": 0.6609273204162789,
+      "grad_norm": 0.16811969876289368,
+      "learning_rate": 0.001,
+      "loss": 1.8456,
+      "step": 15623
+    },
+    {
+      "epoch": 0.6609696251797953,
+      "grad_norm": 0.16160327196121216,
+      "learning_rate": 0.001,
+      "loss": 2.14,
+      "step": 15624
+    },
+    {
+      "epoch": 0.6610119299433116,
+      "grad_norm": 0.17660216987133026,
+      "learning_rate": 0.001,
+      "loss": 3.2061,
+      "step": 15625
+    },
+    {
+      "epoch": 0.661054234706828,
+      "grad_norm": 0.20479421317577362,
+      "learning_rate": 0.001,
+      "loss": 4.2633,
+      "step": 15626
+    },
+    {
+      "epoch": 0.6610965394703444,
+      "grad_norm": 0.21080619096755981,
+      "learning_rate": 0.001,
+      "loss": 2.2565,
+      "step": 15627
+    },
+    {
+      "epoch": 0.6611388442338607,
+      "grad_norm": 0.180558443069458,
+      "learning_rate": 0.001,
+      "loss": 1.8742,
+      "step": 15628
+    },
+    {
+      "epoch": 0.6611811489973771,
+      "grad_norm": 0.21047769486904144,
+      "learning_rate": 0.001,
+      "loss": 2.6093,
+      "step": 15629
+    },
+    {
+      "epoch": 0.6612234537608935,
+      "grad_norm": 0.1717175841331482,
+      "learning_rate": 0.001,
+      "loss": 1.7606,
+      "step": 15630
+    },
+    {
+      "epoch": 0.6612657585244098,
+      "grad_norm": 0.19754162430763245,
+      "learning_rate": 0.001,
+      "loss": 2.4099,
+      "step": 15631
+    },
+    {
+      "epoch": 0.6613080632879262,
+      "grad_norm": 0.24597451090812683,
+      "learning_rate": 0.001,
+      "loss": 2.0901,
+      "step": 15632
+    },
+    {
+      "epoch": 0.6613503680514425,
+      "grad_norm": 0.1566450297832489,
+      "learning_rate": 0.001,
+      "loss": 1.5422,
+      "step": 15633
+    },
+    {
+      "epoch": 0.661392672814959,
+      "grad_norm": 0.16817563772201538,
+      "learning_rate": 0.001,
+      "loss": 1.6473,
+      "step": 15634
+    },
+    {
+      "epoch": 0.6614349775784754,
+      "grad_norm": 0.16162802278995514,
+      "learning_rate": 0.001,
+      "loss": 2.6791,
+      "step": 15635
+    },
+    {
+      "epoch": 0.6614772823419917,
+      "grad_norm": 0.15925811231136322,
+      "learning_rate": 0.001,
+      "loss": 2.2663,
+      "step": 15636
+    },
+    {
+      "epoch": 0.6615195871055081,
+      "grad_norm": 0.17941713333129883,
+      "learning_rate": 0.001,
+      "loss": 2.1478,
+      "step": 15637
+    },
+    {
+      "epoch": 0.6615618918690245,
+      "grad_norm": 0.14808796346187592,
+      "learning_rate": 0.001,
+      "loss": 2.3699,
+      "step": 15638
+    },
+    {
+      "epoch": 0.6616041966325408,
+      "grad_norm": 0.17554134130477905,
+      "learning_rate": 0.001,
+      "loss": 1.5387,
+      "step": 15639
+    },
+    {
+      "epoch": 0.6616465013960572,
+      "grad_norm": 0.17657719552516937,
+      "learning_rate": 0.001,
+      "loss": 1.8514,
+      "step": 15640
+    },
+    {
+      "epoch": 0.6616888061595736,
+      "grad_norm": 0.9961352348327637,
+      "learning_rate": 0.001,
+      "loss": 2.5427,
+      "step": 15641
+    },
+    {
+      "epoch": 0.6617311109230899,
+      "grad_norm": 0.16566766798496246,
+      "learning_rate": 0.001,
+      "loss": 3.038,
+      "step": 15642
+    },
+    {
+      "epoch": 0.6617734156866063,
+      "grad_norm": 0.3651495575904846,
+      "learning_rate": 0.001,
+      "loss": 2.575,
+      "step": 15643
+    },
+    {
+      "epoch": 0.6618157204501227,
+      "grad_norm": 27.26308250427246,
+      "learning_rate": 0.001,
+      "loss": 1.8942,
+      "step": 15644
+    },
+    {
+      "epoch": 0.661858025213639,
+      "grad_norm": 0.15556779503822327,
+      "learning_rate": 0.001,
+      "loss": 3.0992,
+      "step": 15645
+    },
+    {
+      "epoch": 0.6619003299771554,
+      "grad_norm": 0.2336292266845703,
+      "learning_rate": 0.001,
+      "loss": 2.0776,
+      "step": 15646
+    },
+    {
+      "epoch": 0.6619426347406718,
+      "grad_norm": 0.191901296377182,
+      "learning_rate": 0.001,
+      "loss": 2.344,
+      "step": 15647
+    },
+    {
+      "epoch": 0.6619849395041881,
+      "grad_norm": 0.1957559734582901,
+      "learning_rate": 0.001,
+      "loss": 2.3465,
+      "step": 15648
+    },
+    {
+      "epoch": 0.6620272442677045,
+      "grad_norm": 0.19413861632347107,
+      "learning_rate": 0.001,
+      "loss": 2.6151,
+      "step": 15649
+    },
+    {
+      "epoch": 0.662069549031221,
+      "grad_norm": 0.2169937938451767,
+      "learning_rate": 0.001,
+      "loss": 2.3186,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6621118537947372,
+      "grad_norm": 0.16846176981925964,
+      "learning_rate": 0.001,
+      "loss": 1.4052,
+      "step": 15651
+    },
+    {
+      "epoch": 0.6621541585582537,
+      "grad_norm": 0.23452334105968475,
+      "learning_rate": 0.001,
+      "loss": 2.1938,
+      "step": 15652
+    },
+    {
+      "epoch": 0.6621964633217701,
+      "grad_norm": 0.19128349423408508,
+      "learning_rate": 0.001,
+      "loss": 2.3334,
+      "step": 15653
+    },
+    {
+      "epoch": 0.6622387680852864,
+      "grad_norm": 0.3340997099876404,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 15654
+    },
+    {
+      "epoch": 0.6622810728488028,
+      "grad_norm": 0.2314499020576477,
+      "learning_rate": 0.001,
+      "loss": 1.8396,
+      "step": 15655
+    },
+    {
+      "epoch": 0.6623233776123192,
+      "grad_norm": 0.19297367334365845,
+      "learning_rate": 0.001,
+      "loss": 2.2581,
+      "step": 15656
+    },
+    {
+      "epoch": 0.6623656823758355,
+      "grad_norm": 0.15719622373580933,
+      "learning_rate": 0.001,
+      "loss": 1.741,
+      "step": 15657
+    },
+    {
+      "epoch": 0.6624079871393519,
+      "grad_norm": 0.1912499964237213,
+      "learning_rate": 0.001,
+      "loss": 2.2499,
+      "step": 15658
+    },
+    {
+      "epoch": 0.6624502919028683,
+      "grad_norm": 0.21334213018417358,
+      "learning_rate": 0.001,
+      "loss": 2.5132,
+      "step": 15659
+    },
+    {
+      "epoch": 0.6624925966663846,
+      "grad_norm": 0.215172678232193,
+      "learning_rate": 0.001,
+      "loss": 1.9443,
+      "step": 15660
+    },
+    {
+      "epoch": 0.662534901429901,
+      "grad_norm": 0.18598368763923645,
+      "learning_rate": 0.001,
+      "loss": 2.266,
+      "step": 15661
+    },
+    {
+      "epoch": 0.6625772061934174,
+      "grad_norm": 0.1715102642774582,
+      "learning_rate": 0.001,
+      "loss": 2.4652,
+      "step": 15662
+    },
+    {
+      "epoch": 0.6626195109569337,
+      "grad_norm": 0.19849710166454315,
+      "learning_rate": 0.001,
+      "loss": 1.9286,
+      "step": 15663
+    },
+    {
+      "epoch": 0.6626618157204501,
+      "grad_norm": 0.24974191188812256,
+      "learning_rate": 0.001,
+      "loss": 2.3705,
+      "step": 15664
+    },
+    {
+      "epoch": 0.6627041204839665,
+      "grad_norm": 0.1930912882089615,
+      "learning_rate": 0.001,
+      "loss": 1.6271,
+      "step": 15665
+    },
+    {
+      "epoch": 0.6627464252474828,
+      "grad_norm": 0.18645654618740082,
+      "learning_rate": 0.001,
+      "loss": 1.5376,
+      "step": 15666
+    },
+    {
+      "epoch": 0.6627887300109992,
+      "grad_norm": 0.14447803795337677,
+      "learning_rate": 0.001,
+      "loss": 1.5098,
+      "step": 15667
+    },
+    {
+      "epoch": 0.6628310347745157,
+      "grad_norm": 0.1653042733669281,
+      "learning_rate": 0.001,
+      "loss": 1.52,
+      "step": 15668
+    },
+    {
+      "epoch": 0.662873339538032,
+      "grad_norm": 0.16530010104179382,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 15669
+    },
+    {
+      "epoch": 0.6629156443015484,
+      "grad_norm": 0.18780213594436646,
+      "learning_rate": 0.001,
+      "loss": 1.5614,
+      "step": 15670
+    },
+    {
+      "epoch": 0.6629579490650648,
+      "grad_norm": 3.833364725112915,
+      "learning_rate": 0.001,
+      "loss": 1.9021,
+      "step": 15671
+    },
+    {
+      "epoch": 0.6630002538285811,
+      "grad_norm": 0.16255183517932892,
+      "learning_rate": 0.001,
+      "loss": 2.8819,
+      "step": 15672
+    },
+    {
+      "epoch": 0.6630425585920975,
+      "grad_norm": 0.19622357189655304,
+      "learning_rate": 0.001,
+      "loss": 3.1144,
+      "step": 15673
+    },
+    {
+      "epoch": 0.6630848633556139,
+      "grad_norm": 0.22275784611701965,
+      "learning_rate": 0.001,
+      "loss": 2.0371,
+      "step": 15674
+    },
+    {
+      "epoch": 0.6631271681191302,
+      "grad_norm": 0.17636406421661377,
+      "learning_rate": 0.001,
+      "loss": 2.1601,
+      "step": 15675
+    },
+    {
+      "epoch": 0.6631694728826466,
+      "grad_norm": 0.23526060581207275,
+      "learning_rate": 0.001,
+      "loss": 2.5409,
+      "step": 15676
+    },
+    {
+      "epoch": 0.6632117776461629,
+      "grad_norm": 0.17581601440906525,
+      "learning_rate": 0.001,
+      "loss": 1.7807,
+      "step": 15677
+    },
+    {
+      "epoch": 0.6632540824096793,
+      "grad_norm": 0.14658299088478088,
+      "learning_rate": 0.001,
+      "loss": 2.1031,
+      "step": 15678
+    },
+    {
+      "epoch": 0.6632963871731957,
+      "grad_norm": 0.19805100560188293,
+      "learning_rate": 0.001,
+      "loss": 1.9881,
+      "step": 15679
+    },
+    {
+      "epoch": 0.663338691936712,
+      "grad_norm": 0.1707378327846527,
+      "learning_rate": 0.001,
+      "loss": 2.3714,
+      "step": 15680
+    },
+    {
+      "epoch": 0.6633809967002284,
+      "grad_norm": 0.3700905740261078,
+      "learning_rate": 0.001,
+      "loss": 1.8238,
+      "step": 15681
+    },
+    {
+      "epoch": 0.6634233014637448,
+      "grad_norm": 0.2583630084991455,
+      "learning_rate": 0.001,
+      "loss": 3.4336,
+      "step": 15682
+    },
+    {
+      "epoch": 0.6634656062272611,
+      "grad_norm": 0.1871895045042038,
+      "learning_rate": 0.001,
+      "loss": 2.8345,
+      "step": 15683
+    },
+    {
+      "epoch": 0.6635079109907775,
+      "grad_norm": 0.18821239471435547,
+      "learning_rate": 0.001,
+      "loss": 2.0015,
+      "step": 15684
+    },
+    {
+      "epoch": 0.663550215754294,
+      "grad_norm": 0.1794331818819046,
+      "learning_rate": 0.001,
+      "loss": 2.6855,
+      "step": 15685
+    },
+    {
+      "epoch": 0.6635925205178103,
+      "grad_norm": 0.13422973453998566,
+      "learning_rate": 0.001,
+      "loss": 2.8227,
+      "step": 15686
+    },
+    {
+      "epoch": 0.6636348252813267,
+      "grad_norm": 0.16726963222026825,
+      "learning_rate": 0.001,
+      "loss": 2.4534,
+      "step": 15687
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.19704453647136688,
+      "learning_rate": 0.001,
+      "loss": 3.0143,
+      "step": 15688
+    },
+    {
+      "epoch": 0.6637194348083594,
+      "grad_norm": 0.27630937099456787,
+      "learning_rate": 0.001,
+      "loss": 1.6488,
+      "step": 15689
+    },
+    {
+      "epoch": 0.6637617395718758,
+      "grad_norm": 0.16017358005046844,
+      "learning_rate": 0.001,
+      "loss": 2.288,
+      "step": 15690
+    },
+    {
+      "epoch": 0.6638040443353922,
+      "grad_norm": 0.4271370768547058,
+      "learning_rate": 0.001,
+      "loss": 2.3904,
+      "step": 15691
+    },
+    {
+      "epoch": 0.6638463490989085,
+      "grad_norm": 0.17746147513389587,
+      "learning_rate": 0.001,
+      "loss": 2.5629,
+      "step": 15692
+    },
+    {
+      "epoch": 0.6638886538624249,
+      "grad_norm": 0.22230751812458038,
+      "learning_rate": 0.001,
+      "loss": 3.1131,
+      "step": 15693
+    },
+    {
+      "epoch": 0.6639309586259413,
+      "grad_norm": 0.1868673861026764,
+      "learning_rate": 0.001,
+      "loss": 2.424,
+      "step": 15694
+    },
+    {
+      "epoch": 0.6639732633894576,
+      "grad_norm": 0.1976270079612732,
+      "learning_rate": 0.001,
+      "loss": 2.4694,
+      "step": 15695
+    },
+    {
+      "epoch": 0.664015568152974,
+      "grad_norm": 0.18282292783260345,
+      "learning_rate": 0.001,
+      "loss": 2.1648,
+      "step": 15696
+    },
+    {
+      "epoch": 0.6640578729164904,
+      "grad_norm": 0.1945333033800125,
+      "learning_rate": 0.001,
+      "loss": 2.1718,
+      "step": 15697
+    },
+    {
+      "epoch": 0.6641001776800067,
+      "grad_norm": 0.19967901706695557,
+      "learning_rate": 0.001,
+      "loss": 1.6653,
+      "step": 15698
+    },
+    {
+      "epoch": 0.6641424824435231,
+      "grad_norm": 0.35160374641418457,
+      "learning_rate": 0.001,
+      "loss": 2.9863,
+      "step": 15699
+    },
+    {
+      "epoch": 0.6641847872070396,
+      "grad_norm": 0.18338724970817566,
+      "learning_rate": 0.001,
+      "loss": 2.0818,
+      "step": 15700
+    },
+    {
+      "epoch": 0.6642270919705558,
+      "grad_norm": 0.26480796933174133,
+      "learning_rate": 0.001,
+      "loss": 1.8102,
+      "step": 15701
+    },
+    {
+      "epoch": 0.6642693967340723,
+      "grad_norm": 0.1641346663236618,
+      "learning_rate": 0.001,
+      "loss": 1.8631,
+      "step": 15702
+    },
+    {
+      "epoch": 0.6643117014975887,
+      "grad_norm": 0.19785989820957184,
+      "learning_rate": 0.001,
+      "loss": 2.0509,
+      "step": 15703
+    },
+    {
+      "epoch": 0.664354006261105,
+      "grad_norm": 0.1693398356437683,
+      "learning_rate": 0.001,
+      "loss": 2.2726,
+      "step": 15704
+    },
+    {
+      "epoch": 0.6643963110246214,
+      "grad_norm": 0.20726856589317322,
+      "learning_rate": 0.001,
+      "loss": 2.8105,
+      "step": 15705
+    },
+    {
+      "epoch": 0.6644386157881378,
+      "grad_norm": 0.251668781042099,
+      "learning_rate": 0.001,
+      "loss": 2.1022,
+      "step": 15706
+    },
+    {
+      "epoch": 0.6644809205516541,
+      "grad_norm": 0.1539115458726883,
+      "learning_rate": 0.001,
+      "loss": 2.8067,
+      "step": 15707
+    },
+    {
+      "epoch": 0.6645232253151705,
+      "grad_norm": 0.36175769567489624,
+      "learning_rate": 0.001,
+      "loss": 1.5274,
+      "step": 15708
+    },
+    {
+      "epoch": 0.6645655300786869,
+      "grad_norm": 0.17199809849262238,
+      "learning_rate": 0.001,
+      "loss": 2.5353,
+      "step": 15709
+    },
+    {
+      "epoch": 0.6646078348422032,
+      "grad_norm": 2.3134865760803223,
+      "learning_rate": 0.001,
+      "loss": 1.6276,
+      "step": 15710
+    },
+    {
+      "epoch": 0.6646501396057196,
+      "grad_norm": 0.8231282234191895,
+      "learning_rate": 0.001,
+      "loss": 2.0584,
+      "step": 15711
+    },
+    {
+      "epoch": 0.664692444369236,
+      "grad_norm": 0.5767163038253784,
+      "learning_rate": 0.001,
+      "loss": 1.7341,
+      "step": 15712
+    },
+    {
+      "epoch": 0.6647347491327523,
+      "grad_norm": 0.1625528633594513,
+      "learning_rate": 0.001,
+      "loss": 2.172,
+      "step": 15713
+    },
+    {
+      "epoch": 0.6647770538962687,
+      "grad_norm": 0.8674761056900024,
+      "learning_rate": 0.001,
+      "loss": 2.3215,
+      "step": 15714
+    },
+    {
+      "epoch": 0.6648193586597851,
+      "grad_norm": 0.9063986539840698,
+      "learning_rate": 0.001,
+      "loss": 1.7439,
+      "step": 15715
+    },
+    {
+      "epoch": 0.6648616634233014,
+      "grad_norm": 0.17824086546897888,
+      "learning_rate": 0.001,
+      "loss": 2.0572,
+      "step": 15716
+    },
+    {
+      "epoch": 0.6649039681868179,
+      "grad_norm": 0.2059624344110489,
+      "learning_rate": 0.001,
+      "loss": 2.6742,
+      "step": 15717
+    },
+    {
+      "epoch": 0.6649462729503343,
+      "grad_norm": 0.6646909117698669,
+      "learning_rate": 0.001,
+      "loss": 2.3202,
+      "step": 15718
+    },
+    {
+      "epoch": 0.6649885777138506,
+      "grad_norm": 0.1927110105752945,
+      "learning_rate": 0.001,
+      "loss": 1.8751,
+      "step": 15719
+    },
+    {
+      "epoch": 0.665030882477367,
+      "grad_norm": 0.20116344094276428,
+      "learning_rate": 0.001,
+      "loss": 1.9791,
+      "step": 15720
+    },
+    {
+      "epoch": 0.6650731872408834,
+      "grad_norm": 0.44798147678375244,
+      "learning_rate": 0.001,
+      "loss": 2.8342,
+      "step": 15721
+    },
+    {
+      "epoch": 0.6651154920043997,
+      "grad_norm": 0.239822655916214,
+      "learning_rate": 0.001,
+      "loss": 1.8323,
+      "step": 15722
+    },
+    {
+      "epoch": 0.6651577967679161,
+      "grad_norm": 0.20988096296787262,
+      "learning_rate": 0.001,
+      "loss": 2.2684,
+      "step": 15723
+    },
+    {
+      "epoch": 0.6652001015314324,
+      "grad_norm": 0.22264404594898224,
+      "learning_rate": 0.001,
+      "loss": 2.8356,
+      "step": 15724
+    },
+    {
+      "epoch": 0.6652424062949488,
+      "grad_norm": 0.4639243185520172,
+      "learning_rate": 0.001,
+      "loss": 1.9119,
+      "step": 15725
+    },
+    {
+      "epoch": 0.6652847110584652,
+      "grad_norm": 0.8543448448181152,
+      "learning_rate": 0.001,
+      "loss": 1.6931,
+      "step": 15726
+    },
+    {
+      "epoch": 0.6653270158219815,
+      "grad_norm": 1.1423962116241455,
+      "learning_rate": 0.001,
+      "loss": 1.5227,
+      "step": 15727
+    },
+    {
+      "epoch": 0.6653693205854979,
+      "grad_norm": 0.1901700645685196,
+      "learning_rate": 0.001,
+      "loss": 2.7098,
+      "step": 15728
+    },
+    {
+      "epoch": 0.6654116253490143,
+      "grad_norm": 0.8082202076911926,
+      "learning_rate": 0.001,
+      "loss": 2.3667,
+      "step": 15729
+    },
+    {
+      "epoch": 0.6654539301125306,
+      "grad_norm": 0.3281959891319275,
+      "learning_rate": 0.001,
+      "loss": 1.9087,
+      "step": 15730
+    },
+    {
+      "epoch": 0.665496234876047,
+      "grad_norm": 0.3187538683414459,
+      "learning_rate": 0.001,
+      "loss": 1.9291,
+      "step": 15731
+    },
+    {
+      "epoch": 0.6655385396395634,
+      "grad_norm": 2.2117702960968018,
+      "learning_rate": 0.001,
+      "loss": 2.724,
+      "step": 15732
+    },
+    {
+      "epoch": 0.6655808444030797,
+      "grad_norm": 35.29819107055664,
+      "learning_rate": 0.001,
+      "loss": 2.2991,
+      "step": 15733
+    },
+    {
+      "epoch": 0.6656231491665962,
+      "grad_norm": 0.18186047673225403,
+      "learning_rate": 0.001,
+      "loss": 1.3769,
+      "step": 15734
+    },
+    {
+      "epoch": 0.6656654539301126,
+      "grad_norm": 0.24423609673976898,
+      "learning_rate": 0.001,
+      "loss": 2.32,
+      "step": 15735
+    },
+    {
+      "epoch": 0.6657077586936289,
+      "grad_norm": 0.319638729095459,
+      "learning_rate": 0.001,
+      "loss": 2.4374,
+      "step": 15736
+    },
+    {
+      "epoch": 0.6657500634571453,
+      "grad_norm": 0.1802143007516861,
+      "learning_rate": 0.001,
+      "loss": 2.2174,
+      "step": 15737
+    },
+    {
+      "epoch": 0.6657923682206617,
+      "grad_norm": 0.17544938623905182,
+      "learning_rate": 0.001,
+      "loss": 1.7473,
+      "step": 15738
+    },
+    {
+      "epoch": 0.665834672984178,
+      "grad_norm": 0.9758937954902649,
+      "learning_rate": 0.001,
+      "loss": 2.2743,
+      "step": 15739
+    },
+    {
+      "epoch": 0.6658769777476944,
+      "grad_norm": 0.1803831309080124,
+      "learning_rate": 0.001,
+      "loss": 3.2636,
+      "step": 15740
+    },
+    {
+      "epoch": 0.6659192825112108,
+      "grad_norm": 0.2357131391763687,
+      "learning_rate": 0.001,
+      "loss": 2.4164,
+      "step": 15741
+    },
+    {
+      "epoch": 0.6659615872747271,
+      "grad_norm": 0.16148456931114197,
+      "learning_rate": 0.001,
+      "loss": 1.8385,
+      "step": 15742
+    },
+    {
+      "epoch": 0.6660038920382435,
+      "grad_norm": 0.17655906081199646,
+      "learning_rate": 0.001,
+      "loss": 1.7078,
+      "step": 15743
+    },
+    {
+      "epoch": 0.6660461968017599,
+      "grad_norm": 0.18633557856082916,
+      "learning_rate": 0.001,
+      "loss": 1.5433,
+      "step": 15744
+    },
+    {
+      "epoch": 0.6660885015652762,
+      "grad_norm": 0.16458381712436676,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 15745
+    },
+    {
+      "epoch": 0.6661308063287926,
+      "grad_norm": 0.1529700607061386,
+      "learning_rate": 0.001,
+      "loss": 2.7367,
+      "step": 15746
+    },
+    {
+      "epoch": 0.666173111092309,
+      "grad_norm": 0.16840951144695282,
+      "learning_rate": 0.001,
+      "loss": 2.7347,
+      "step": 15747
+    },
+    {
+      "epoch": 0.6662154158558253,
+      "grad_norm": 0.15616460144519806,
+      "learning_rate": 0.001,
+      "loss": 1.6326,
+      "step": 15748
+    },
+    {
+      "epoch": 0.6662577206193417,
+      "grad_norm": 45.16459274291992,
+      "learning_rate": 0.001,
+      "loss": 2.0402,
+      "step": 15749
+    },
+    {
+      "epoch": 0.6663000253828582,
+      "grad_norm": 0.35248544812202454,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 15750
+    },
+    {
+      "epoch": 0.6663423301463745,
+      "grad_norm": 0.14932596683502197,
+      "learning_rate": 0.001,
+      "loss": 2.1283,
+      "step": 15751
+    },
+    {
+      "epoch": 0.6663846349098909,
+      "grad_norm": 0.22044512629508972,
+      "learning_rate": 0.001,
+      "loss": 2.6434,
+      "step": 15752
+    },
+    {
+      "epoch": 0.6664269396734073,
+      "grad_norm": 0.21503077447414398,
+      "learning_rate": 0.001,
+      "loss": 2.6892,
+      "step": 15753
+    },
+    {
+      "epoch": 0.6664692444369236,
+      "grad_norm": 3.1654529571533203,
+      "learning_rate": 0.001,
+      "loss": 1.6119,
+      "step": 15754
+    },
+    {
+      "epoch": 0.66651154920044,
+      "grad_norm": 0.19109536707401276,
+      "learning_rate": 0.001,
+      "loss": 1.7015,
+      "step": 15755
+    },
+    {
+      "epoch": 0.6665538539639564,
+      "grad_norm": 0.217402383685112,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 15756
+    },
+    {
+      "epoch": 0.6665961587274727,
+      "grad_norm": 0.20222723484039307,
+      "learning_rate": 0.001,
+      "loss": 1.8401,
+      "step": 15757
+    },
+    {
+      "epoch": 0.6666384634909891,
+      "grad_norm": 0.24966546893119812,
+      "learning_rate": 0.001,
+      "loss": 1.984,
+      "step": 15758
+    },
+    {
+      "epoch": 0.6666807682545055,
+      "grad_norm": 0.25226879119873047,
+      "learning_rate": 0.001,
+      "loss": 2.9811,
+      "step": 15759
+    },
+    {
+      "epoch": 0.6667230730180218,
+      "grad_norm": 0.37204667925834656,
+      "learning_rate": 0.001,
+      "loss": 2.5787,
+      "step": 15760
+    },
+    {
+      "epoch": 0.6667653777815382,
+      "grad_norm": 1.5811610221862793,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 15761
+    },
+    {
+      "epoch": 0.6668076825450546,
+      "grad_norm": 0.41911280155181885,
+      "learning_rate": 0.001,
+      "loss": 2.1604,
+      "step": 15762
+    },
+    {
+      "epoch": 0.6668499873085709,
+      "grad_norm": 0.19730260968208313,
+      "learning_rate": 0.001,
+      "loss": 3.0043,
+      "step": 15763
+    },
+    {
+      "epoch": 0.6668922920720873,
+      "grad_norm": 0.23359958827495575,
+      "learning_rate": 0.001,
+      "loss": 3.387,
+      "step": 15764
+    },
+    {
+      "epoch": 0.6669345968356037,
+      "grad_norm": 4.667820930480957,
+      "learning_rate": 0.001,
+      "loss": 2.4877,
+      "step": 15765
+    },
+    {
+      "epoch": 0.66697690159912,
+      "grad_norm": 0.25286656618118286,
+      "learning_rate": 0.001,
+      "loss": 1.9937,
+      "step": 15766
+    },
+    {
+      "epoch": 0.6670192063626365,
+      "grad_norm": 0.23963415622711182,
+      "learning_rate": 0.001,
+      "loss": 1.69,
+      "step": 15767
+    },
+    {
+      "epoch": 0.6670615111261528,
+      "grad_norm": 0.26238617300987244,
+      "learning_rate": 0.001,
+      "loss": 2.352,
+      "step": 15768
+    },
+    {
+      "epoch": 0.6671038158896692,
+      "grad_norm": 83.53499603271484,
+      "learning_rate": 0.001,
+      "loss": 2.5639,
+      "step": 15769
+    },
+    {
+      "epoch": 0.6671461206531856,
+      "grad_norm": 0.325317919254303,
+      "learning_rate": 0.001,
+      "loss": 2.2086,
+      "step": 15770
+    },
+    {
+      "epoch": 0.6671884254167019,
+      "grad_norm": 0.20828962326049805,
+      "learning_rate": 0.001,
+      "loss": 2.473,
+      "step": 15771
+    },
+    {
+      "epoch": 0.6672307301802183,
+      "grad_norm": 0.21652193367481232,
+      "learning_rate": 0.001,
+      "loss": 2.2587,
+      "step": 15772
+    },
+    {
+      "epoch": 0.6672730349437347,
+      "grad_norm": 0.44318485260009766,
+      "learning_rate": 0.001,
+      "loss": 2.0383,
+      "step": 15773
+    },
+    {
+      "epoch": 0.667315339707251,
+      "grad_norm": 0.5616191625595093,
+      "learning_rate": 0.001,
+      "loss": 1.6704,
+      "step": 15774
+    },
+    {
+      "epoch": 0.6673576444707674,
+      "grad_norm": 0.2393038272857666,
+      "learning_rate": 0.001,
+      "loss": 2.2136,
+      "step": 15775
+    },
+    {
+      "epoch": 0.6673999492342838,
+      "grad_norm": 0.2519043982028961,
+      "learning_rate": 0.001,
+      "loss": 2.8899,
+      "step": 15776
+    },
+    {
+      "epoch": 0.6674422539978001,
+      "grad_norm": 0.2502747178077698,
+      "learning_rate": 0.001,
+      "loss": 2.4507,
+      "step": 15777
+    },
+    {
+      "epoch": 0.6674845587613165,
+      "grad_norm": 0.20891942083835602,
+      "learning_rate": 0.001,
+      "loss": 1.7339,
+      "step": 15778
+    },
+    {
+      "epoch": 0.6675268635248329,
+      "grad_norm": 1.0747361183166504,
+      "learning_rate": 0.001,
+      "loss": 2.0077,
+      "step": 15779
+    },
+    {
+      "epoch": 0.6675691682883492,
+      "grad_norm": 0.23666757345199585,
+      "learning_rate": 0.001,
+      "loss": 2.8961,
+      "step": 15780
+    },
+    {
+      "epoch": 0.6676114730518656,
+      "grad_norm": 0.3330824077129364,
+      "learning_rate": 0.001,
+      "loss": 2.9923,
+      "step": 15781
+    },
+    {
+      "epoch": 0.667653777815382,
+      "grad_norm": 0.33238059282302856,
+      "learning_rate": 0.001,
+      "loss": 1.7868,
+      "step": 15782
+    },
+    {
+      "epoch": 0.6676960825788983,
+      "grad_norm": 0.17682193219661713,
+      "learning_rate": 0.001,
+      "loss": 2.7748,
+      "step": 15783
+    },
+    {
+      "epoch": 0.6677383873424148,
+      "grad_norm": 0.2014869600534439,
+      "learning_rate": 0.001,
+      "loss": 1.9485,
+      "step": 15784
+    },
+    {
+      "epoch": 0.6677806921059312,
+      "grad_norm": 0.17806148529052734,
+      "learning_rate": 0.001,
+      "loss": 1.7066,
+      "step": 15785
+    },
+    {
+      "epoch": 0.6678229968694475,
+      "grad_norm": 6.159350395202637,
+      "learning_rate": 0.001,
+      "loss": 2.0137,
+      "step": 15786
+    },
+    {
+      "epoch": 0.6678653016329639,
+      "grad_norm": 0.381747841835022,
+      "learning_rate": 0.001,
+      "loss": 2.2376,
+      "step": 15787
+    },
+    {
+      "epoch": 0.6679076063964803,
+      "grad_norm": 0.2559763789176941,
+      "learning_rate": 0.001,
+      "loss": 2.3698,
+      "step": 15788
+    },
+    {
+      "epoch": 0.6679499111599966,
+      "grad_norm": 0.24722157418727875,
+      "learning_rate": 0.001,
+      "loss": 2.6856,
+      "step": 15789
+    },
+    {
+      "epoch": 0.667992215923513,
+      "grad_norm": 3.812957286834717,
+      "learning_rate": 0.001,
+      "loss": 1.976,
+      "step": 15790
+    },
+    {
+      "epoch": 0.6680345206870294,
+      "grad_norm": 0.3154589533805847,
+      "learning_rate": 0.001,
+      "loss": 4.0476,
+      "step": 15791
+    },
+    {
+      "epoch": 0.6680768254505457,
+      "grad_norm": 0.20563848316669464,
+      "learning_rate": 0.001,
+      "loss": 2.0702,
+      "step": 15792
+    },
+    {
+      "epoch": 0.6681191302140621,
+      "grad_norm": 0.3167200982570648,
+      "learning_rate": 0.001,
+      "loss": 2.8908,
+      "step": 15793
+    },
+    {
+      "epoch": 0.6681614349775785,
+      "grad_norm": 0.2859385907649994,
+      "learning_rate": 0.001,
+      "loss": 2.3652,
+      "step": 15794
+    },
+    {
+      "epoch": 0.6682037397410948,
+      "grad_norm": 4.2778730392456055,
+      "learning_rate": 0.001,
+      "loss": 2.6051,
+      "step": 15795
+    },
+    {
+      "epoch": 0.6682460445046112,
+      "grad_norm": 0.23722970485687256,
+      "learning_rate": 0.001,
+      "loss": 1.4931,
+      "step": 15796
+    },
+    {
+      "epoch": 0.6682883492681276,
+      "grad_norm": 0.19192059338092804,
+      "learning_rate": 0.001,
+      "loss": 2.4992,
+      "step": 15797
+    },
+    {
+      "epoch": 0.6683306540316439,
+      "grad_norm": 0.25284337997436523,
+      "learning_rate": 0.001,
+      "loss": 1.957,
+      "step": 15798
+    },
+    {
+      "epoch": 0.6683729587951603,
+      "grad_norm": 0.23747120797634125,
+      "learning_rate": 0.001,
+      "loss": 1.6552,
+      "step": 15799
+    },
+    {
+      "epoch": 0.6684152635586768,
+      "grad_norm": 0.21687759459018707,
+      "learning_rate": 0.001,
+      "loss": 1.8681,
+      "step": 15800
+    },
+    {
+      "epoch": 0.668457568322193,
+      "grad_norm": 0.2691959738731384,
+      "learning_rate": 0.001,
+      "loss": 2.2289,
+      "step": 15801
+    },
+    {
+      "epoch": 0.6684998730857095,
+      "grad_norm": 0.2524837553501129,
+      "learning_rate": 0.001,
+      "loss": 1.8516,
+      "step": 15802
+    },
+    {
+      "epoch": 0.6685421778492259,
+      "grad_norm": 0.2057633101940155,
+      "learning_rate": 0.001,
+      "loss": 2.509,
+      "step": 15803
+    },
+    {
+      "epoch": 0.6685844826127422,
+      "grad_norm": 1.199033498764038,
+      "learning_rate": 0.001,
+      "loss": 1.5052,
+      "step": 15804
+    },
+    {
+      "epoch": 0.6686267873762586,
+      "grad_norm": 0.3630319833755493,
+      "learning_rate": 0.001,
+      "loss": 2.5884,
+      "step": 15805
+    },
+    {
+      "epoch": 0.668669092139775,
+      "grad_norm": 0.4470106065273285,
+      "learning_rate": 0.001,
+      "loss": 2.7533,
+      "step": 15806
+    },
+    {
+      "epoch": 0.6687113969032913,
+      "grad_norm": 0.18470320105552673,
+      "learning_rate": 0.001,
+      "loss": 2.5,
+      "step": 15807
+    },
+    {
+      "epoch": 0.6687537016668077,
+      "grad_norm": 3.678845167160034,
+      "learning_rate": 0.001,
+      "loss": 2.6681,
+      "step": 15808
+    },
+    {
+      "epoch": 0.6687960064303241,
+      "grad_norm": 0.19354915618896484,
+      "learning_rate": 0.001,
+      "loss": 2.4777,
+      "step": 15809
+    },
+    {
+      "epoch": 0.6688383111938404,
+      "grad_norm": 0.32464343309402466,
+      "learning_rate": 0.001,
+      "loss": 2.7447,
+      "step": 15810
+    },
+    {
+      "epoch": 0.6688806159573568,
+      "grad_norm": 0.1859036386013031,
+      "learning_rate": 0.001,
+      "loss": 1.5693,
+      "step": 15811
+    },
+    {
+      "epoch": 0.6689229207208731,
+      "grad_norm": 0.19540037214756012,
+      "learning_rate": 0.001,
+      "loss": 2.8704,
+      "step": 15812
+    },
+    {
+      "epoch": 0.6689652254843895,
+      "grad_norm": 0.24034158885478973,
+      "learning_rate": 0.001,
+      "loss": 1.6442,
+      "step": 15813
+    },
+    {
+      "epoch": 0.6690075302479059,
+      "grad_norm": 0.20161794126033783,
+      "learning_rate": 0.001,
+      "loss": 1.7371,
+      "step": 15814
+    },
+    {
+      "epoch": 0.6690498350114222,
+      "grad_norm": 0.27223238348960876,
+      "learning_rate": 0.001,
+      "loss": 2.8628,
+      "step": 15815
+    },
+    {
+      "epoch": 0.6690921397749386,
+      "grad_norm": 0.18867754936218262,
+      "learning_rate": 0.001,
+      "loss": 1.8677,
+      "step": 15816
+    },
+    {
+      "epoch": 0.669134444538455,
+      "grad_norm": 0.18999359011650085,
+      "learning_rate": 0.001,
+      "loss": 2.0956,
+      "step": 15817
+    },
+    {
+      "epoch": 0.6691767493019714,
+      "grad_norm": 0.14994925260543823,
+      "learning_rate": 0.001,
+      "loss": 2.6274,
+      "step": 15818
+    },
+    {
+      "epoch": 0.6692190540654878,
+      "grad_norm": 0.16162274777889252,
+      "learning_rate": 0.001,
+      "loss": 2.5421,
+      "step": 15819
+    },
+    {
+      "epoch": 0.6692613588290042,
+      "grad_norm": 1.4299275875091553,
+      "learning_rate": 0.001,
+      "loss": 2.334,
+      "step": 15820
+    },
+    {
+      "epoch": 0.6693036635925205,
+      "grad_norm": 0.9547110199928284,
+      "learning_rate": 0.001,
+      "loss": 2.2495,
+      "step": 15821
+    },
+    {
+      "epoch": 0.6693459683560369,
+      "grad_norm": 0.19420866668224335,
+      "learning_rate": 0.001,
+      "loss": 2.792,
+      "step": 15822
+    },
+    {
+      "epoch": 0.6693882731195533,
+      "grad_norm": 0.15037833154201508,
+      "learning_rate": 0.001,
+      "loss": 2.696,
+      "step": 15823
+    },
+    {
+      "epoch": 0.6694305778830696,
+      "grad_norm": 0.1415679156780243,
+      "learning_rate": 0.001,
+      "loss": 3.4118,
+      "step": 15824
+    },
+    {
+      "epoch": 0.669472882646586,
+      "grad_norm": 0.15855517983436584,
+      "learning_rate": 0.001,
+      "loss": 1.7473,
+      "step": 15825
+    },
+    {
+      "epoch": 0.6695151874101024,
+      "grad_norm": 0.47525739669799805,
+      "learning_rate": 0.001,
+      "loss": 2.8112,
+      "step": 15826
+    },
+    {
+      "epoch": 0.6695574921736187,
+      "grad_norm": 0.35138139128685,
+      "learning_rate": 0.001,
+      "loss": 2.0132,
+      "step": 15827
+    },
+    {
+      "epoch": 0.6695997969371351,
+      "grad_norm": 0.18116721510887146,
+      "learning_rate": 0.001,
+      "loss": 2.488,
+      "step": 15828
+    },
+    {
+      "epoch": 0.6696421017006515,
+      "grad_norm": 0.18540117144584656,
+      "learning_rate": 0.001,
+      "loss": 3.8645,
+      "step": 15829
+    },
+    {
+      "epoch": 0.6696844064641678,
+      "grad_norm": 0.19489508867263794,
+      "learning_rate": 0.001,
+      "loss": 2.7296,
+      "step": 15830
+    },
+    {
+      "epoch": 0.6697267112276842,
+      "grad_norm": 0.19228166341781616,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 15831
+    },
+    {
+      "epoch": 0.6697690159912006,
+      "grad_norm": 0.15116505324840546,
+      "learning_rate": 0.001,
+      "loss": 1.8258,
+      "step": 15832
+    },
+    {
+      "epoch": 0.6698113207547169,
+      "grad_norm": 0.6345039010047913,
+      "learning_rate": 0.001,
+      "loss": 2.1165,
+      "step": 15833
+    },
+    {
+      "epoch": 0.6698536255182334,
+      "grad_norm": 0.15162324905395508,
+      "learning_rate": 0.001,
+      "loss": 2.3055,
+      "step": 15834
+    },
+    {
+      "epoch": 0.6698959302817498,
+      "grad_norm": 1.6587047576904297,
+      "learning_rate": 0.001,
+      "loss": 1.78,
+      "step": 15835
+    },
+    {
+      "epoch": 0.6699382350452661,
+      "grad_norm": 0.1674169898033142,
+      "learning_rate": 0.001,
+      "loss": 1.9546,
+      "step": 15836
+    },
+    {
+      "epoch": 0.6699805398087825,
+      "grad_norm": 0.1696109175682068,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 15837
+    },
+    {
+      "epoch": 0.6700228445722989,
+      "grad_norm": 0.16364048421382904,
+      "learning_rate": 0.001,
+      "loss": 1.7434,
+      "step": 15838
+    },
+    {
+      "epoch": 0.6700651493358152,
+      "grad_norm": 0.18208681046962738,
+      "learning_rate": 0.001,
+      "loss": 2.9249,
+      "step": 15839
+    },
+    {
+      "epoch": 0.6701074540993316,
+      "grad_norm": 0.870050847530365,
+      "learning_rate": 0.001,
+      "loss": 1.8984,
+      "step": 15840
+    },
+    {
+      "epoch": 0.670149758862848,
+      "grad_norm": 0.20092959702014923,
+      "learning_rate": 0.001,
+      "loss": 2.044,
+      "step": 15841
+    },
+    {
+      "epoch": 0.6701920636263643,
+      "grad_norm": 0.24720391631126404,
+      "learning_rate": 0.001,
+      "loss": 2.1473,
+      "step": 15842
+    },
+    {
+      "epoch": 0.6702343683898807,
+      "grad_norm": 0.18767613172531128,
+      "learning_rate": 0.001,
+      "loss": 2.8646,
+      "step": 15843
+    },
+    {
+      "epoch": 0.6702766731533971,
+      "grad_norm": 0.15457838773727417,
+      "learning_rate": 0.001,
+      "loss": 1.7142,
+      "step": 15844
+    },
+    {
+      "epoch": 0.6703189779169134,
+      "grad_norm": 0.1792488992214203,
+      "learning_rate": 0.001,
+      "loss": 2.8613,
+      "step": 15845
+    },
+    {
+      "epoch": 0.6703612826804298,
+      "grad_norm": 0.9255082607269287,
+      "learning_rate": 0.001,
+      "loss": 2.8254,
+      "step": 15846
+    },
+    {
+      "epoch": 0.6704035874439462,
+      "grad_norm": 3.0631420612335205,
+      "learning_rate": 0.001,
+      "loss": 2.5177,
+      "step": 15847
+    },
+    {
+      "epoch": 0.6704458922074625,
+      "grad_norm": 0.1708671599626541,
+      "learning_rate": 0.001,
+      "loss": 2.045,
+      "step": 15848
+    },
+    {
+      "epoch": 0.670488196970979,
+      "grad_norm": 0.1887563019990921,
+      "learning_rate": 0.001,
+      "loss": 1.8023,
+      "step": 15849
+    },
+    {
+      "epoch": 0.6705305017344954,
+      "grad_norm": 0.23792339861392975,
+      "learning_rate": 0.001,
+      "loss": 2.0385,
+      "step": 15850
+    },
+    {
+      "epoch": 0.6705728064980117,
+      "grad_norm": 0.1688668131828308,
+      "learning_rate": 0.001,
+      "loss": 2.0975,
+      "step": 15851
+    },
+    {
+      "epoch": 0.6706151112615281,
+      "grad_norm": 0.21332918107509613,
+      "learning_rate": 0.001,
+      "loss": 2.3229,
+      "step": 15852
+    },
+    {
+      "epoch": 0.6706574160250445,
+      "grad_norm": 0.1556411236524582,
+      "learning_rate": 0.001,
+      "loss": 2.6122,
+      "step": 15853
+    },
+    {
+      "epoch": 0.6706997207885608,
+      "grad_norm": 0.17764410376548767,
+      "learning_rate": 0.001,
+      "loss": 1.339,
+      "step": 15854
+    },
+    {
+      "epoch": 0.6707420255520772,
+      "grad_norm": 0.31873539090156555,
+      "learning_rate": 0.001,
+      "loss": 2.1945,
+      "step": 15855
+    },
+    {
+      "epoch": 0.6707843303155936,
+      "grad_norm": 1.299968957901001,
+      "learning_rate": 0.001,
+      "loss": 1.8488,
+      "step": 15856
+    },
+    {
+      "epoch": 0.6708266350791099,
+      "grad_norm": 0.19831599295139313,
+      "learning_rate": 0.001,
+      "loss": 2.9631,
+      "step": 15857
+    },
+    {
+      "epoch": 0.6708689398426263,
+      "grad_norm": 0.4041413962841034,
+      "learning_rate": 0.001,
+      "loss": 1.7992,
+      "step": 15858
+    },
+    {
+      "epoch": 0.6709112446061426,
+      "grad_norm": 0.19786041975021362,
+      "learning_rate": 0.001,
+      "loss": 2.2841,
+      "step": 15859
+    },
+    {
+      "epoch": 0.670953549369659,
+      "grad_norm": 0.17182038724422455,
+      "learning_rate": 0.001,
+      "loss": 2.1733,
+      "step": 15860
+    },
+    {
+      "epoch": 0.6709958541331754,
+      "grad_norm": 0.21848192811012268,
+      "learning_rate": 0.001,
+      "loss": 2.1951,
+      "step": 15861
+    },
+    {
+      "epoch": 0.6710381588966917,
+      "grad_norm": 0.2034897655248642,
+      "learning_rate": 0.001,
+      "loss": 2.1136,
+      "step": 15862
+    },
+    {
+      "epoch": 0.6710804636602081,
+      "grad_norm": 0.19234777987003326,
+      "learning_rate": 0.001,
+      "loss": 2.0125,
+      "step": 15863
+    },
+    {
+      "epoch": 0.6711227684237245,
+      "grad_norm": 0.2422971874475479,
+      "learning_rate": 0.001,
+      "loss": 3.2963,
+      "step": 15864
+    },
+    {
+      "epoch": 0.6711650731872408,
+      "grad_norm": 0.17505241930484772,
+      "learning_rate": 0.001,
+      "loss": 1.7463,
+      "step": 15865
+    },
+    {
+      "epoch": 0.6712073779507572,
+      "grad_norm": 0.1780707985162735,
+      "learning_rate": 0.001,
+      "loss": 1.6777,
+      "step": 15866
+    },
+    {
+      "epoch": 0.6712496827142737,
+      "grad_norm": 0.1790010929107666,
+      "learning_rate": 0.001,
+      "loss": 2.4067,
+      "step": 15867
+    },
+    {
+      "epoch": 0.67129198747779,
+      "grad_norm": 0.19889087975025177,
+      "learning_rate": 0.001,
+      "loss": 2.5161,
+      "step": 15868
+    },
+    {
+      "epoch": 0.6713342922413064,
+      "grad_norm": 0.9066613912582397,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 15869
+    },
+    {
+      "epoch": 0.6713765970048228,
+      "grad_norm": 0.15774136781692505,
+      "learning_rate": 0.001,
+      "loss": 2.4626,
+      "step": 15870
+    },
+    {
+      "epoch": 0.6714189017683391,
+      "grad_norm": 0.14970077574253082,
+      "learning_rate": 0.001,
+      "loss": 1.6917,
+      "step": 15871
+    },
+    {
+      "epoch": 0.6714612065318555,
+      "grad_norm": 0.1569453924894333,
+      "learning_rate": 0.001,
+      "loss": 2.3071,
+      "step": 15872
+    },
+    {
+      "epoch": 0.6715035112953719,
+      "grad_norm": 0.4611540138721466,
+      "learning_rate": 0.001,
+      "loss": 2.905,
+      "step": 15873
+    },
+    {
+      "epoch": 0.6715458160588882,
+      "grad_norm": 0.15557625889778137,
+      "learning_rate": 0.001,
+      "loss": 2.3258,
+      "step": 15874
+    },
+    {
+      "epoch": 0.6715881208224046,
+      "grad_norm": 0.19848431646823883,
+      "learning_rate": 0.001,
+      "loss": 2.6509,
+      "step": 15875
+    },
+    {
+      "epoch": 0.671630425585921,
+      "grad_norm": 0.16773267090320587,
+      "learning_rate": 0.001,
+      "loss": 2.1152,
+      "step": 15876
+    },
+    {
+      "epoch": 0.6716727303494373,
+      "grad_norm": 0.23722898960113525,
+      "learning_rate": 0.001,
+      "loss": 2.3771,
+      "step": 15877
+    },
+    {
+      "epoch": 0.6717150351129537,
+      "grad_norm": 1.1143704652786255,
+      "learning_rate": 0.001,
+      "loss": 2.607,
+      "step": 15878
+    },
+    {
+      "epoch": 0.6717573398764701,
+      "grad_norm": 0.20059651136398315,
+      "learning_rate": 0.001,
+      "loss": 2.2388,
+      "step": 15879
+    },
+    {
+      "epoch": 0.6717996446399864,
+      "grad_norm": 0.19346950948238373,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 15880
+    },
+    {
+      "epoch": 0.6718419494035028,
+      "grad_norm": 0.1549372375011444,
+      "learning_rate": 0.001,
+      "loss": 1.6958,
+      "step": 15881
+    },
+    {
+      "epoch": 0.6718842541670192,
+      "grad_norm": 0.23537324368953705,
+      "learning_rate": 0.001,
+      "loss": 1.8379,
+      "step": 15882
+    },
+    {
+      "epoch": 0.6719265589305355,
+      "grad_norm": 0.2048342525959015,
+      "learning_rate": 0.001,
+      "loss": 2.3291,
+      "step": 15883
+    },
+    {
+      "epoch": 0.671968863694052,
+      "grad_norm": 0.15993832051753998,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 15884
+    },
+    {
+      "epoch": 0.6720111684575684,
+      "grad_norm": 0.1927027702331543,
+      "learning_rate": 0.001,
+      "loss": 2.0073,
+      "step": 15885
+    },
+    {
+      "epoch": 0.6720534732210847,
+      "grad_norm": 0.2728440463542938,
+      "learning_rate": 0.001,
+      "loss": 2.1145,
+      "step": 15886
+    },
+    {
+      "epoch": 0.6720957779846011,
+      "grad_norm": 0.21112090349197388,
+      "learning_rate": 0.001,
+      "loss": 2.4727,
+      "step": 15887
+    },
+    {
+      "epoch": 0.6721380827481175,
+      "grad_norm": 0.19208763539791107,
+      "learning_rate": 0.001,
+      "loss": 1.6904,
+      "step": 15888
+    },
+    {
+      "epoch": 0.6721803875116338,
+      "grad_norm": 0.18781235814094543,
+      "learning_rate": 0.001,
+      "loss": 2.5108,
+      "step": 15889
+    },
+    {
+      "epoch": 0.6722226922751502,
+      "grad_norm": 0.20023007690906525,
+      "learning_rate": 0.001,
+      "loss": 1.6354,
+      "step": 15890
+    },
+    {
+      "epoch": 0.6722649970386666,
+      "grad_norm": 0.18701715767383575,
+      "learning_rate": 0.001,
+      "loss": 2.3519,
+      "step": 15891
+    },
+    {
+      "epoch": 0.6723073018021829,
+      "grad_norm": 2.3707351684570312,
+      "learning_rate": 0.001,
+      "loss": 1.7433,
+      "step": 15892
+    },
+    {
+      "epoch": 0.6723496065656993,
+      "grad_norm": 1.582403540611267,
+      "learning_rate": 0.001,
+      "loss": 2.2697,
+      "step": 15893
+    },
+    {
+      "epoch": 0.6723919113292157,
+      "grad_norm": 0.17928339540958405,
+      "learning_rate": 0.001,
+      "loss": 2.7044,
+      "step": 15894
+    },
+    {
+      "epoch": 0.672434216092732,
+      "grad_norm": 0.1638590693473816,
+      "learning_rate": 0.001,
+      "loss": 2.5621,
+      "step": 15895
+    },
+    {
+      "epoch": 0.6724765208562484,
+      "grad_norm": 0.25966379046440125,
+      "learning_rate": 0.001,
+      "loss": 2.4395,
+      "step": 15896
+    },
+    {
+      "epoch": 0.6725188256197648,
+      "grad_norm": 0.17949466407299042,
+      "learning_rate": 0.001,
+      "loss": 1.7309,
+      "step": 15897
+    },
+    {
+      "epoch": 0.6725611303832811,
+      "grad_norm": 0.182010680437088,
+      "learning_rate": 0.001,
+      "loss": 1.9328,
+      "step": 15898
+    },
+    {
+      "epoch": 0.6726034351467975,
+      "grad_norm": 0.2181997448205948,
+      "learning_rate": 0.001,
+      "loss": 2.3349,
+      "step": 15899
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 0.28189408779144287,
+      "learning_rate": 0.001,
+      "loss": 2.0677,
+      "step": 15900
+    },
+    {
+      "epoch": 0.6726880446738303,
+      "grad_norm": 0.22430168092250824,
+      "learning_rate": 0.001,
+      "loss": 1.6799,
+      "step": 15901
+    },
+    {
+      "epoch": 0.6727303494373467,
+      "grad_norm": 0.6292093396186829,
+      "learning_rate": 0.001,
+      "loss": 1.8766,
+      "step": 15902
+    },
+    {
+      "epoch": 0.672772654200863,
+      "grad_norm": 4.264910697937012,
+      "learning_rate": 0.001,
+      "loss": 2.1911,
+      "step": 15903
+    },
+    {
+      "epoch": 0.6728149589643794,
+      "grad_norm": 0.2221497744321823,
+      "learning_rate": 0.001,
+      "loss": 2.1785,
+      "step": 15904
+    },
+    {
+      "epoch": 0.6728572637278958,
+      "grad_norm": 0.19245116412639618,
+      "learning_rate": 0.001,
+      "loss": 2.712,
+      "step": 15905
+    },
+    {
+      "epoch": 0.6728995684914121,
+      "grad_norm": 0.25045454502105713,
+      "learning_rate": 0.001,
+      "loss": 3.8263,
+      "step": 15906
+    },
+    {
+      "epoch": 0.6729418732549285,
+      "grad_norm": 0.16638623178005219,
+      "learning_rate": 0.001,
+      "loss": 2.4182,
+      "step": 15907
+    },
+    {
+      "epoch": 0.6729841780184449,
+      "grad_norm": 0.17453812062740326,
+      "learning_rate": 0.001,
+      "loss": 1.8538,
+      "step": 15908
+    },
+    {
+      "epoch": 0.6730264827819612,
+      "grad_norm": 0.27706319093704224,
+      "learning_rate": 0.001,
+      "loss": 2.2582,
+      "step": 15909
+    },
+    {
+      "epoch": 0.6730687875454776,
+      "grad_norm": 0.1942570060491562,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 15910
+    },
+    {
+      "epoch": 0.673111092308994,
+      "grad_norm": 0.18307733535766602,
+      "learning_rate": 0.001,
+      "loss": 2.1524,
+      "step": 15911
+    },
+    {
+      "epoch": 0.6731533970725103,
+      "grad_norm": 0.1911049634218216,
+      "learning_rate": 0.001,
+      "loss": 2.4024,
+      "step": 15912
+    },
+    {
+      "epoch": 0.6731957018360267,
+      "grad_norm": 0.15870538353919983,
+      "learning_rate": 0.001,
+      "loss": 2.1905,
+      "step": 15913
+    },
+    {
+      "epoch": 0.6732380065995431,
+      "grad_norm": 0.14895422756671906,
+      "learning_rate": 0.001,
+      "loss": 1.8064,
+      "step": 15914
+    },
+    {
+      "epoch": 0.6732803113630594,
+      "grad_norm": 0.16844263672828674,
+      "learning_rate": 0.001,
+      "loss": 2.3851,
+      "step": 15915
+    },
+    {
+      "epoch": 0.6733226161265758,
+      "grad_norm": 0.15884919464588165,
+      "learning_rate": 0.001,
+      "loss": 2.3164,
+      "step": 15916
+    },
+    {
+      "epoch": 0.6733649208900923,
+      "grad_norm": 0.20211821794509888,
+      "learning_rate": 0.001,
+      "loss": 2.3867,
+      "step": 15917
+    },
+    {
+      "epoch": 0.6734072256536086,
+      "grad_norm": 0.3958881199359894,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 15918
+    },
+    {
+      "epoch": 0.673449530417125,
+      "grad_norm": 0.4282142221927643,
+      "learning_rate": 0.001,
+      "loss": 2.8803,
+      "step": 15919
+    },
+    {
+      "epoch": 0.6734918351806414,
+      "grad_norm": 0.1638120859861374,
+      "learning_rate": 0.001,
+      "loss": 1.665,
+      "step": 15920
+    },
+    {
+      "epoch": 0.6735341399441577,
+      "grad_norm": 0.15340758860111237,
+      "learning_rate": 0.001,
+      "loss": 2.273,
+      "step": 15921
+    },
+    {
+      "epoch": 0.6735764447076741,
+      "grad_norm": 0.17197485268115997,
+      "learning_rate": 0.001,
+      "loss": 2.0298,
+      "step": 15922
+    },
+    {
+      "epoch": 0.6736187494711905,
+      "grad_norm": 0.1842762529850006,
+      "learning_rate": 0.001,
+      "loss": 2.1636,
+      "step": 15923
+    },
+    {
+      "epoch": 0.6736610542347068,
+      "grad_norm": 0.1678512841463089,
+      "learning_rate": 0.001,
+      "loss": 2.6527,
+      "step": 15924
+    },
+    {
+      "epoch": 0.6737033589982232,
+      "grad_norm": 0.1786356419324875,
+      "learning_rate": 0.001,
+      "loss": 2.7915,
+      "step": 15925
+    },
+    {
+      "epoch": 0.6737456637617396,
+      "grad_norm": 0.14677925407886505,
+      "learning_rate": 0.001,
+      "loss": 1.7595,
+      "step": 15926
+    },
+    {
+      "epoch": 0.6737879685252559,
+      "grad_norm": 0.18858976662158966,
+      "learning_rate": 0.001,
+      "loss": 1.9292,
+      "step": 15927
+    },
+    {
+      "epoch": 0.6738302732887723,
+      "grad_norm": 0.18520209193229675,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 15928
+    },
+    {
+      "epoch": 0.6738725780522887,
+      "grad_norm": 0.9475305676460266,
+      "learning_rate": 0.001,
+      "loss": 2.5929,
+      "step": 15929
+    },
+    {
+      "epoch": 0.673914882815805,
+      "grad_norm": 0.17860578000545502,
+      "learning_rate": 0.001,
+      "loss": 1.4486,
+      "step": 15930
+    },
+    {
+      "epoch": 0.6739571875793214,
+      "grad_norm": 0.1508931964635849,
+      "learning_rate": 0.001,
+      "loss": 1.5058,
+      "step": 15931
+    },
+    {
+      "epoch": 0.6739994923428378,
+      "grad_norm": 0.17374980449676514,
+      "learning_rate": 0.001,
+      "loss": 1.819,
+      "step": 15932
+    },
+    {
+      "epoch": 0.6740417971063541,
+      "grad_norm": 0.1520918607711792,
+      "learning_rate": 0.001,
+      "loss": 1.8909,
+      "step": 15933
+    },
+    {
+      "epoch": 0.6740841018698706,
+      "grad_norm": 1.1293140649795532,
+      "learning_rate": 0.001,
+      "loss": 3.0682,
+      "step": 15934
+    },
+    {
+      "epoch": 0.674126406633387,
+      "grad_norm": 0.17505602538585663,
+      "learning_rate": 0.001,
+      "loss": 3.318,
+      "step": 15935
+    },
+    {
+      "epoch": 0.6741687113969033,
+      "grad_norm": 0.3417392075061798,
+      "learning_rate": 0.001,
+      "loss": 2.3171,
+      "step": 15936
+    },
+    {
+      "epoch": 0.6742110161604197,
+      "grad_norm": 0.39481237530708313,
+      "learning_rate": 0.001,
+      "loss": 1.7587,
+      "step": 15937
+    },
+    {
+      "epoch": 0.6742533209239361,
+      "grad_norm": 0.37343841791152954,
+      "learning_rate": 0.001,
+      "loss": 2.229,
+      "step": 15938
+    },
+    {
+      "epoch": 0.6742956256874524,
+      "grad_norm": 0.17360848188400269,
+      "learning_rate": 0.001,
+      "loss": 2.4422,
+      "step": 15939
+    },
+    {
+      "epoch": 0.6743379304509688,
+      "grad_norm": 0.2295389622449875,
+      "learning_rate": 0.001,
+      "loss": 1.9226,
+      "step": 15940
+    },
+    {
+      "epoch": 0.6743802352144852,
+      "grad_norm": 0.8693257570266724,
+      "learning_rate": 0.001,
+      "loss": 2.2007,
+      "step": 15941
+    },
+    {
+      "epoch": 0.6744225399780015,
+      "grad_norm": 0.19263996183872223,
+      "learning_rate": 0.001,
+      "loss": 3.2355,
+      "step": 15942
+    },
+    {
+      "epoch": 0.6744648447415179,
+      "grad_norm": 0.17120057344436646,
+      "learning_rate": 0.001,
+      "loss": 2.5869,
+      "step": 15943
+    },
+    {
+      "epoch": 0.6745071495050343,
+      "grad_norm": 0.292049378156662,
+      "learning_rate": 0.001,
+      "loss": 1.8235,
+      "step": 15944
+    },
+    {
+      "epoch": 0.6745494542685506,
+      "grad_norm": 0.17973144352436066,
+      "learning_rate": 0.001,
+      "loss": 2.0943,
+      "step": 15945
+    },
+    {
+      "epoch": 0.674591759032067,
+      "grad_norm": 0.30642953515052795,
+      "learning_rate": 0.001,
+      "loss": 3.0754,
+      "step": 15946
+    },
+    {
+      "epoch": 0.6746340637955833,
+      "grad_norm": 0.15908905863761902,
+      "learning_rate": 0.001,
+      "loss": 1.4202,
+      "step": 15947
+    },
+    {
+      "epoch": 0.6746763685590997,
+      "grad_norm": 0.1885390281677246,
+      "learning_rate": 0.001,
+      "loss": 1.8611,
+      "step": 15948
+    },
+    {
+      "epoch": 0.6747186733226161,
+      "grad_norm": 1.8227906227111816,
+      "learning_rate": 0.001,
+      "loss": 3.1409,
+      "step": 15949
+    },
+    {
+      "epoch": 0.6747609780861324,
+      "grad_norm": 0.3331076204776764,
+      "learning_rate": 0.001,
+      "loss": 2.6025,
+      "step": 15950
+    },
+    {
+      "epoch": 0.6748032828496489,
+      "grad_norm": 0.18833647668361664,
+      "learning_rate": 0.001,
+      "loss": 1.7091,
+      "step": 15951
+    },
+    {
+      "epoch": 0.6748455876131653,
+      "grad_norm": 0.8229162693023682,
+      "learning_rate": 0.001,
+      "loss": 1.9747,
+      "step": 15952
+    },
+    {
+      "epoch": 0.6748878923766816,
+      "grad_norm": 0.19028577208518982,
+      "learning_rate": 0.001,
+      "loss": 1.9955,
+      "step": 15953
+    },
+    {
+      "epoch": 0.674930197140198,
+      "grad_norm": 0.19043925404548645,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 15954
+    },
+    {
+      "epoch": 0.6749725019037144,
+      "grad_norm": 0.1884336918592453,
+      "learning_rate": 0.001,
+      "loss": 2.4741,
+      "step": 15955
+    },
+    {
+      "epoch": 0.6750148066672307,
+      "grad_norm": 0.2047235518693924,
+      "learning_rate": 0.001,
+      "loss": 2.729,
+      "step": 15956
+    },
+    {
+      "epoch": 0.6750571114307471,
+      "grad_norm": 1.2655960321426392,
+      "learning_rate": 0.001,
+      "loss": 2.2603,
+      "step": 15957
+    },
+    {
+      "epoch": 0.6750994161942635,
+      "grad_norm": 0.2844761312007904,
+      "learning_rate": 0.001,
+      "loss": 1.8044,
+      "step": 15958
+    },
+    {
+      "epoch": 0.6751417209577798,
+      "grad_norm": 0.19485563039779663,
+      "learning_rate": 0.001,
+      "loss": 2.0,
+      "step": 15959
+    },
+    {
+      "epoch": 0.6751840257212962,
+      "grad_norm": 0.2195066213607788,
+      "learning_rate": 0.001,
+      "loss": 2.6798,
+      "step": 15960
+    },
+    {
+      "epoch": 0.6752263304848126,
+      "grad_norm": 0.18380756676197052,
+      "learning_rate": 0.001,
+      "loss": 1.7459,
+      "step": 15961
+    },
+    {
+      "epoch": 0.6752686352483289,
+      "grad_norm": 0.22318629920482635,
+      "learning_rate": 0.001,
+      "loss": 2.6885,
+      "step": 15962
+    },
+    {
+      "epoch": 0.6753109400118453,
+      "grad_norm": 0.940578818321228,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 15963
+    },
+    {
+      "epoch": 0.6753532447753617,
+      "grad_norm": 0.19312544167041779,
+      "learning_rate": 0.001,
+      "loss": 2.0494,
+      "step": 15964
+    },
+    {
+      "epoch": 0.675395549538878,
+      "grad_norm": 0.24037931859493256,
+      "learning_rate": 0.001,
+      "loss": 2.2044,
+      "step": 15965
+    },
+    {
+      "epoch": 0.6754378543023944,
+      "grad_norm": 0.478103369474411,
+      "learning_rate": 0.001,
+      "loss": 3.5143,
+      "step": 15966
+    },
+    {
+      "epoch": 0.6754801590659109,
+      "grad_norm": 0.976945161819458,
+      "learning_rate": 0.001,
+      "loss": 2.7022,
+      "step": 15967
+    },
+    {
+      "epoch": 0.6755224638294272,
+      "grad_norm": 0.1700313240289688,
+      "learning_rate": 0.001,
+      "loss": 2.0294,
+      "step": 15968
+    },
+    {
+      "epoch": 0.6755647685929436,
+      "grad_norm": 0.36223527789115906,
+      "learning_rate": 0.001,
+      "loss": 2.3376,
+      "step": 15969
+    },
+    {
+      "epoch": 0.67560707335646,
+      "grad_norm": 1.9298176765441895,
+      "learning_rate": 0.001,
+      "loss": 2.5461,
+      "step": 15970
+    },
+    {
+      "epoch": 0.6756493781199763,
+      "grad_norm": 0.23120592534542084,
+      "learning_rate": 0.001,
+      "loss": 2.5119,
+      "step": 15971
+    },
+    {
+      "epoch": 0.6756916828834927,
+      "grad_norm": 0.18629468977451324,
+      "learning_rate": 0.001,
+      "loss": 2.1999,
+      "step": 15972
+    },
+    {
+      "epoch": 0.6757339876470091,
+      "grad_norm": 0.2043001651763916,
+      "learning_rate": 0.001,
+      "loss": 2.3867,
+      "step": 15973
+    },
+    {
+      "epoch": 0.6757762924105254,
+      "grad_norm": 0.1776389479637146,
+      "learning_rate": 0.001,
+      "loss": 1.6796,
+      "step": 15974
+    },
+    {
+      "epoch": 0.6758185971740418,
+      "grad_norm": 0.17264407873153687,
+      "learning_rate": 0.001,
+      "loss": 1.5977,
+      "step": 15975
+    },
+    {
+      "epoch": 0.6758609019375582,
+      "grad_norm": 0.3489270806312561,
+      "learning_rate": 0.001,
+      "loss": 2.4665,
+      "step": 15976
+    },
+    {
+      "epoch": 0.6759032067010745,
+      "grad_norm": 0.18797357380390167,
+      "learning_rate": 0.001,
+      "loss": 2.8661,
+      "step": 15977
+    },
+    {
+      "epoch": 0.6759455114645909,
+      "grad_norm": 0.1839050054550171,
+      "learning_rate": 0.001,
+      "loss": 1.798,
+      "step": 15978
+    },
+    {
+      "epoch": 0.6759878162281073,
+      "grad_norm": 0.2641671299934387,
+      "learning_rate": 0.001,
+      "loss": 2.0923,
+      "step": 15979
+    },
+    {
+      "epoch": 0.6760301209916236,
+      "grad_norm": 0.1782594621181488,
+      "learning_rate": 0.001,
+      "loss": 2.7264,
+      "step": 15980
+    },
+    {
+      "epoch": 0.67607242575514,
+      "grad_norm": 0.3223048150539398,
+      "learning_rate": 0.001,
+      "loss": 1.5769,
+      "step": 15981
+    },
+    {
+      "epoch": 0.6761147305186564,
+      "grad_norm": 0.17407678067684174,
+      "learning_rate": 0.001,
+      "loss": 2.2179,
+      "step": 15982
+    },
+    {
+      "epoch": 0.6761570352821727,
+      "grad_norm": 0.2168140560388565,
+      "learning_rate": 0.001,
+      "loss": 2.9042,
+      "step": 15983
+    },
+    {
+      "epoch": 0.6761993400456892,
+      "grad_norm": 0.24506430327892303,
+      "learning_rate": 0.001,
+      "loss": 2.6773,
+      "step": 15984
+    },
+    {
+      "epoch": 0.6762416448092056,
+      "grad_norm": 3.0023486614227295,
+      "learning_rate": 0.001,
+      "loss": 2.0182,
+      "step": 15985
+    },
+    {
+      "epoch": 0.6762839495727219,
+      "grad_norm": 0.1910613626241684,
+      "learning_rate": 0.001,
+      "loss": 2.047,
+      "step": 15986
+    },
+    {
+      "epoch": 0.6763262543362383,
+      "grad_norm": 0.17066790163516998,
+      "learning_rate": 0.001,
+      "loss": 2.0095,
+      "step": 15987
+    },
+    {
+      "epoch": 0.6763685590997547,
+      "grad_norm": 0.21411702036857605,
+      "learning_rate": 0.001,
+      "loss": 1.883,
+      "step": 15988
+    },
+    {
+      "epoch": 0.676410863863271,
+      "grad_norm": 0.16612938046455383,
+      "learning_rate": 0.001,
+      "loss": 1.7573,
+      "step": 15989
+    },
+    {
+      "epoch": 0.6764531686267874,
+      "grad_norm": 0.2450494021177292,
+      "learning_rate": 0.001,
+      "loss": 2.951,
+      "step": 15990
+    },
+    {
+      "epoch": 0.6764954733903038,
+      "grad_norm": 0.21236656606197357,
+      "learning_rate": 0.001,
+      "loss": 3.2924,
+      "step": 15991
+    },
+    {
+      "epoch": 0.6765377781538201,
+      "grad_norm": 1.4454017877578735,
+      "learning_rate": 0.001,
+      "loss": 2.2104,
+      "step": 15992
+    },
+    {
+      "epoch": 0.6765800829173365,
+      "grad_norm": 0.16844910383224487,
+      "learning_rate": 0.001,
+      "loss": 2.0065,
+      "step": 15993
+    },
+    {
+      "epoch": 0.6766223876808528,
+      "grad_norm": 0.19149447977542877,
+      "learning_rate": 0.001,
+      "loss": 2.4233,
+      "step": 15994
+    },
+    {
+      "epoch": 0.6766646924443692,
+      "grad_norm": 0.15004916489124298,
+      "learning_rate": 0.001,
+      "loss": 1.9877,
+      "step": 15995
+    },
+    {
+      "epoch": 0.6767069972078856,
+      "grad_norm": 0.1823279857635498,
+      "learning_rate": 0.001,
+      "loss": 1.6102,
+      "step": 15996
+    },
+    {
+      "epoch": 0.6767493019714019,
+      "grad_norm": 0.17334629595279694,
+      "learning_rate": 0.001,
+      "loss": 2.3336,
+      "step": 15997
+    },
+    {
+      "epoch": 0.6767916067349183,
+      "grad_norm": 0.19492603838443756,
+      "learning_rate": 0.001,
+      "loss": 2.8301,
+      "step": 15998
+    },
+    {
+      "epoch": 0.6768339114984347,
+      "grad_norm": 0.17607484757900238,
+      "learning_rate": 0.001,
+      "loss": 1.8637,
+      "step": 15999
+    },
+    {
+      "epoch": 0.676876216261951,
+      "grad_norm": 0.15553830564022064,
+      "learning_rate": 0.001,
+      "loss": 2.4753,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6769185210254675,
+      "grad_norm": 0.30437377095222473,
+      "learning_rate": 0.001,
+      "loss": 2.1501,
+      "step": 16001
+    },
+    {
+      "epoch": 0.6769608257889839,
+      "grad_norm": 0.17832662165164948,
+      "learning_rate": 0.001,
+      "loss": 1.9037,
+      "step": 16002
+    },
+    {
+      "epoch": 0.6770031305525002,
+      "grad_norm": 0.25508981943130493,
+      "learning_rate": 0.001,
+      "loss": 1.3602,
+      "step": 16003
+    },
+    {
+      "epoch": 0.6770454353160166,
+      "grad_norm": 0.17589248716831207,
+      "learning_rate": 0.001,
+      "loss": 3.1867,
+      "step": 16004
+    },
+    {
+      "epoch": 0.677087740079533,
+      "grad_norm": 0.1560450941324234,
+      "learning_rate": 0.001,
+      "loss": 1.6277,
+      "step": 16005
+    },
+    {
+      "epoch": 0.6771300448430493,
+      "grad_norm": 25.86013412475586,
+      "learning_rate": 0.001,
+      "loss": 2.1036,
+      "step": 16006
+    },
+    {
+      "epoch": 0.6771723496065657,
+      "grad_norm": 0.153413325548172,
+      "learning_rate": 0.001,
+      "loss": 2.2372,
+      "step": 16007
+    },
+    {
+      "epoch": 0.6772146543700821,
+      "grad_norm": 0.17835178971290588,
+      "learning_rate": 0.001,
+      "loss": 1.8833,
+      "step": 16008
+    },
+    {
+      "epoch": 0.6772569591335984,
+      "grad_norm": 0.15974244475364685,
+      "learning_rate": 0.001,
+      "loss": 1.9219,
+      "step": 16009
+    },
+    {
+      "epoch": 0.6772992638971148,
+      "grad_norm": 0.1814177930355072,
+      "learning_rate": 0.001,
+      "loss": 1.6924,
+      "step": 16010
+    },
+    {
+      "epoch": 0.6773415686606312,
+      "grad_norm": 0.14862357079982758,
+      "learning_rate": 0.001,
+      "loss": 1.2958,
+      "step": 16011
+    },
+    {
+      "epoch": 0.6773838734241475,
+      "grad_norm": 0.2236209362745285,
+      "learning_rate": 0.001,
+      "loss": 2.7953,
+      "step": 16012
+    },
+    {
+      "epoch": 0.6774261781876639,
+      "grad_norm": 0.17278966307640076,
+      "learning_rate": 0.001,
+      "loss": 2.3449,
+      "step": 16013
+    },
+    {
+      "epoch": 0.6774684829511803,
+      "grad_norm": 0.4638851284980774,
+      "learning_rate": 0.001,
+      "loss": 1.7526,
+      "step": 16014
+    },
+    {
+      "epoch": 0.6775107877146966,
+      "grad_norm": 0.1768551915884018,
+      "learning_rate": 0.001,
+      "loss": 3.0475,
+      "step": 16015
+    },
+    {
+      "epoch": 0.677553092478213,
+      "grad_norm": 0.19388507306575775,
+      "learning_rate": 0.001,
+      "loss": 2.2825,
+      "step": 16016
+    },
+    {
+      "epoch": 0.6775953972417295,
+      "grad_norm": 0.20869360864162445,
+      "learning_rate": 0.001,
+      "loss": 2.2376,
+      "step": 16017
+    },
+    {
+      "epoch": 0.6776377020052458,
+      "grad_norm": 0.16567271947860718,
+      "learning_rate": 0.001,
+      "loss": 1.9322,
+      "step": 16018
+    },
+    {
+      "epoch": 0.6776800067687622,
+      "grad_norm": 0.15912829339504242,
+      "learning_rate": 0.001,
+      "loss": 1.5842,
+      "step": 16019
+    },
+    {
+      "epoch": 0.6777223115322786,
+      "grad_norm": 0.14932376146316528,
+      "learning_rate": 0.001,
+      "loss": 1.6747,
+      "step": 16020
+    },
+    {
+      "epoch": 0.6777646162957949,
+      "grad_norm": 1.8247519731521606,
+      "learning_rate": 0.001,
+      "loss": 2.3317,
+      "step": 16021
+    },
+    {
+      "epoch": 0.6778069210593113,
+      "grad_norm": 0.22858451306819916,
+      "learning_rate": 0.001,
+      "loss": 2.5957,
+      "step": 16022
+    },
+    {
+      "epoch": 0.6778492258228277,
+      "grad_norm": 0.1770337074995041,
+      "learning_rate": 0.001,
+      "loss": 1.436,
+      "step": 16023
+    },
+    {
+      "epoch": 0.677891530586344,
+      "grad_norm": 0.1400739699602127,
+      "learning_rate": 0.001,
+      "loss": 1.9365,
+      "step": 16024
+    },
+    {
+      "epoch": 0.6779338353498604,
+      "grad_norm": 0.1432129591703415,
+      "learning_rate": 0.001,
+      "loss": 1.5036,
+      "step": 16025
+    },
+    {
+      "epoch": 0.6779761401133768,
+      "grad_norm": 0.16388866305351257,
+      "learning_rate": 0.001,
+      "loss": 1.8471,
+      "step": 16026
+    },
+    {
+      "epoch": 0.6780184448768931,
+      "grad_norm": 2.311777353286743,
+      "learning_rate": 0.001,
+      "loss": 2.4238,
+      "step": 16027
+    },
+    {
+      "epoch": 0.6780607496404095,
+      "grad_norm": 0.19178903102874756,
+      "learning_rate": 0.001,
+      "loss": 1.9744,
+      "step": 16028
+    },
+    {
+      "epoch": 0.6781030544039259,
+      "grad_norm": 0.15423864126205444,
+      "learning_rate": 0.001,
+      "loss": 2.1231,
+      "step": 16029
+    },
+    {
+      "epoch": 0.6781453591674422,
+      "grad_norm": 0.16429832577705383,
+      "learning_rate": 0.001,
+      "loss": 2.1746,
+      "step": 16030
+    },
+    {
+      "epoch": 0.6781876639309586,
+      "grad_norm": 0.14937609434127808,
+      "learning_rate": 0.001,
+      "loss": 2.2265,
+      "step": 16031
+    },
+    {
+      "epoch": 0.678229968694475,
+      "grad_norm": 0.17871306836605072,
+      "learning_rate": 0.001,
+      "loss": 2.605,
+      "step": 16032
+    },
+    {
+      "epoch": 0.6782722734579913,
+      "grad_norm": 0.1368352472782135,
+      "learning_rate": 0.001,
+      "loss": 1.3931,
+      "step": 16033
+    },
+    {
+      "epoch": 0.6783145782215078,
+      "grad_norm": 0.15225689113140106,
+      "learning_rate": 0.001,
+      "loss": 1.8712,
+      "step": 16034
+    },
+    {
+      "epoch": 0.6783568829850242,
+      "grad_norm": 0.237832173705101,
+      "learning_rate": 0.001,
+      "loss": 2.7011,
+      "step": 16035
+    },
+    {
+      "epoch": 0.6783991877485405,
+      "grad_norm": 0.2749160826206207,
+      "learning_rate": 0.001,
+      "loss": 2.0368,
+      "step": 16036
+    },
+    {
+      "epoch": 0.6784414925120569,
+      "grad_norm": 0.1856054663658142,
+      "learning_rate": 0.001,
+      "loss": 1.5841,
+      "step": 16037
+    },
+    {
+      "epoch": 0.6784837972755732,
+      "grad_norm": 0.35927614569664,
+      "learning_rate": 0.001,
+      "loss": 1.5595,
+      "step": 16038
+    },
+    {
+      "epoch": 0.6785261020390896,
+      "grad_norm": 0.18782657384872437,
+      "learning_rate": 0.001,
+      "loss": 2.4929,
+      "step": 16039
+    },
+    {
+      "epoch": 0.678568406802606,
+      "grad_norm": 0.1664906144142151,
+      "learning_rate": 0.001,
+      "loss": 2.8713,
+      "step": 16040
+    },
+    {
+      "epoch": 0.6786107115661223,
+      "grad_norm": 0.17706331610679626,
+      "learning_rate": 0.001,
+      "loss": 2.2758,
+      "step": 16041
+    },
+    {
+      "epoch": 0.6786530163296387,
+      "grad_norm": 1.698190450668335,
+      "learning_rate": 0.001,
+      "loss": 2.0762,
+      "step": 16042
+    },
+    {
+      "epoch": 0.6786953210931551,
+      "grad_norm": 0.2534584105014801,
+      "learning_rate": 0.001,
+      "loss": 2.2728,
+      "step": 16043
+    },
+    {
+      "epoch": 0.6787376258566714,
+      "grad_norm": 0.14307022094726562,
+      "learning_rate": 0.001,
+      "loss": 2.1301,
+      "step": 16044
+    },
+    {
+      "epoch": 0.6787799306201878,
+      "grad_norm": 0.16313962638378143,
+      "learning_rate": 0.001,
+      "loss": 2.6391,
+      "step": 16045
+    },
+    {
+      "epoch": 0.6788222353837042,
+      "grad_norm": 0.2026868611574173,
+      "learning_rate": 0.001,
+      "loss": 3.0308,
+      "step": 16046
+    },
+    {
+      "epoch": 0.6788645401472205,
+      "grad_norm": 0.1914452612400055,
+      "learning_rate": 0.001,
+      "loss": 2.2605,
+      "step": 16047
+    },
+    {
+      "epoch": 0.6789068449107369,
+      "grad_norm": 0.17033523321151733,
+      "learning_rate": 0.001,
+      "loss": 2.7466,
+      "step": 16048
+    },
+    {
+      "epoch": 0.6789491496742533,
+      "grad_norm": 0.41400858759880066,
+      "learning_rate": 0.001,
+      "loss": 1.8912,
+      "step": 16049
+    },
+    {
+      "epoch": 0.6789914544377696,
+      "grad_norm": 0.14586646854877472,
+      "learning_rate": 0.001,
+      "loss": 2.2457,
+      "step": 16050
+    },
+    {
+      "epoch": 0.6790337592012861,
+      "grad_norm": 0.1477472484111786,
+      "learning_rate": 0.001,
+      "loss": 1.9528,
+      "step": 16051
+    },
+    {
+      "epoch": 0.6790760639648025,
+      "grad_norm": 0.18150284886360168,
+      "learning_rate": 0.001,
+      "loss": 1.8803,
+      "step": 16052
+    },
+    {
+      "epoch": 0.6791183687283188,
+      "grad_norm": 163.43626403808594,
+      "learning_rate": 0.001,
+      "loss": 2.7313,
+      "step": 16053
+    },
+    {
+      "epoch": 0.6791606734918352,
+      "grad_norm": 0.17980308830738068,
+      "learning_rate": 0.001,
+      "loss": 2.8632,
+      "step": 16054
+    },
+    {
+      "epoch": 0.6792029782553516,
+      "grad_norm": 1.3750078678131104,
+      "learning_rate": 0.001,
+      "loss": 2.8298,
+      "step": 16055
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.23904463648796082,
+      "learning_rate": 0.001,
+      "loss": 2.7106,
+      "step": 16056
+    },
+    {
+      "epoch": 0.6792875877823843,
+      "grad_norm": 0.18311509490013123,
+      "learning_rate": 0.001,
+      "loss": 1.9757,
+      "step": 16057
+    },
+    {
+      "epoch": 0.6793298925459007,
+      "grad_norm": 0.15157288312911987,
+      "learning_rate": 0.001,
+      "loss": 2.3797,
+      "step": 16058
+    },
+    {
+      "epoch": 0.679372197309417,
+      "grad_norm": 0.1471908688545227,
+      "learning_rate": 0.001,
+      "loss": 2.9482,
+      "step": 16059
+    },
+    {
+      "epoch": 0.6794145020729334,
+      "grad_norm": 0.16476254165172577,
+      "learning_rate": 0.001,
+      "loss": 3.0529,
+      "step": 16060
+    },
+    {
+      "epoch": 0.6794568068364498,
+      "grad_norm": 0.17069002985954285,
+      "learning_rate": 0.001,
+      "loss": 2.0707,
+      "step": 16061
+    },
+    {
+      "epoch": 0.6794991115999661,
+      "grad_norm": 0.17665056884288788,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 16062
+    },
+    {
+      "epoch": 0.6795414163634825,
+      "grad_norm": 0.1902749389410019,
+      "learning_rate": 0.001,
+      "loss": 2.5135,
+      "step": 16063
+    },
+    {
+      "epoch": 0.6795837211269989,
+      "grad_norm": 0.16366536915302277,
+      "learning_rate": 0.001,
+      "loss": 2.1562,
+      "step": 16064
+    },
+    {
+      "epoch": 0.6796260258905152,
+      "grad_norm": 0.6264306306838989,
+      "learning_rate": 0.001,
+      "loss": 1.8162,
+      "step": 16065
+    },
+    {
+      "epoch": 0.6796683306540316,
+      "grad_norm": 0.2236694097518921,
+      "learning_rate": 0.001,
+      "loss": 2.4581,
+      "step": 16066
+    },
+    {
+      "epoch": 0.6797106354175481,
+      "grad_norm": 0.16119384765625,
+      "learning_rate": 0.001,
+      "loss": 2.1214,
+      "step": 16067
+    },
+    {
+      "epoch": 0.6797529401810644,
+      "grad_norm": 0.16101056337356567,
+      "learning_rate": 0.001,
+      "loss": 1.6174,
+      "step": 16068
+    },
+    {
+      "epoch": 0.6797952449445808,
+      "grad_norm": 0.28261250257492065,
+      "learning_rate": 0.001,
+      "loss": 2.6421,
+      "step": 16069
+    },
+    {
+      "epoch": 0.6798375497080972,
+      "grad_norm": 1.5717463493347168,
+      "learning_rate": 0.001,
+      "loss": 2.7132,
+      "step": 16070
+    },
+    {
+      "epoch": 0.6798798544716135,
+      "grad_norm": 0.15996554493904114,
+      "learning_rate": 0.001,
+      "loss": 2.364,
+      "step": 16071
+    },
+    {
+      "epoch": 0.6799221592351299,
+      "grad_norm": 0.20333285629749298,
+      "learning_rate": 0.001,
+      "loss": 2.6281,
+      "step": 16072
+    },
+    {
+      "epoch": 0.6799644639986463,
+      "grad_norm": 0.17524294555187225,
+      "learning_rate": 0.001,
+      "loss": 2.1776,
+      "step": 16073
+    },
+    {
+      "epoch": 0.6800067687621626,
+      "grad_norm": 0.2101033478975296,
+      "learning_rate": 0.001,
+      "loss": 2.2662,
+      "step": 16074
+    },
+    {
+      "epoch": 0.680049073525679,
+      "grad_norm": 0.1923307329416275,
+      "learning_rate": 0.001,
+      "loss": 1.9986,
+      "step": 16075
+    },
+    {
+      "epoch": 0.6800913782891954,
+      "grad_norm": 0.14267486333847046,
+      "learning_rate": 0.001,
+      "loss": 1.4753,
+      "step": 16076
+    },
+    {
+      "epoch": 0.6801336830527117,
+      "grad_norm": 0.14065372943878174,
+      "learning_rate": 0.001,
+      "loss": 1.4347,
+      "step": 16077
+    },
+    {
+      "epoch": 0.6801759878162281,
+      "grad_norm": 0.20316016674041748,
+      "learning_rate": 0.001,
+      "loss": 1.7941,
+      "step": 16078
+    },
+    {
+      "epoch": 0.6802182925797445,
+      "grad_norm": 0.24780157208442688,
+      "learning_rate": 0.001,
+      "loss": 3.0841,
+      "step": 16079
+    },
+    {
+      "epoch": 0.6802605973432608,
+      "grad_norm": 0.19607360661029816,
+      "learning_rate": 0.001,
+      "loss": 3.1225,
+      "step": 16080
+    },
+    {
+      "epoch": 0.6803029021067772,
+      "grad_norm": 0.16723619401454926,
+      "learning_rate": 0.001,
+      "loss": 1.9696,
+      "step": 16081
+    },
+    {
+      "epoch": 0.6803452068702935,
+      "grad_norm": 10.566794395446777,
+      "learning_rate": 0.001,
+      "loss": 2.347,
+      "step": 16082
+    },
+    {
+      "epoch": 0.68038751163381,
+      "grad_norm": 0.19241277873516083,
+      "learning_rate": 0.001,
+      "loss": 2.1758,
+      "step": 16083
+    },
+    {
+      "epoch": 0.6804298163973264,
+      "grad_norm": 0.3141223192214966,
+      "learning_rate": 0.001,
+      "loss": 2.1376,
+      "step": 16084
+    },
+    {
+      "epoch": 0.6804721211608427,
+      "grad_norm": 0.1563158631324768,
+      "learning_rate": 0.001,
+      "loss": 2.7554,
+      "step": 16085
+    },
+    {
+      "epoch": 0.6805144259243591,
+      "grad_norm": 0.967668354511261,
+      "learning_rate": 0.001,
+      "loss": 2.3854,
+      "step": 16086
+    },
+    {
+      "epoch": 0.6805567306878755,
+      "grad_norm": 0.1570819467306137,
+      "learning_rate": 0.001,
+      "loss": 1.7543,
+      "step": 16087
+    },
+    {
+      "epoch": 0.6805990354513918,
+      "grad_norm": 0.18060454726219177,
+      "learning_rate": 0.001,
+      "loss": 1.3768,
+      "step": 16088
+    },
+    {
+      "epoch": 0.6806413402149082,
+      "grad_norm": 0.1356450766324997,
+      "learning_rate": 0.001,
+      "loss": 1.2979,
+      "step": 16089
+    },
+    {
+      "epoch": 0.6806836449784246,
+      "grad_norm": 0.20549282431602478,
+      "learning_rate": 0.001,
+      "loss": 2.1857,
+      "step": 16090
+    },
+    {
+      "epoch": 0.6807259497419409,
+      "grad_norm": 0.17379418015480042,
+      "learning_rate": 0.001,
+      "loss": 1.7882,
+      "step": 16091
+    },
+    {
+      "epoch": 0.6807682545054573,
+      "grad_norm": 0.17570160329341888,
+      "learning_rate": 0.001,
+      "loss": 2.2534,
+      "step": 16092
+    },
+    {
+      "epoch": 0.6808105592689737,
+      "grad_norm": 0.17414408922195435,
+      "learning_rate": 0.001,
+      "loss": 2.5999,
+      "step": 16093
+    },
+    {
+      "epoch": 0.68085286403249,
+      "grad_norm": 0.9091881513595581,
+      "learning_rate": 0.001,
+      "loss": 2.7196,
+      "step": 16094
+    },
+    {
+      "epoch": 0.6808951687960064,
+      "grad_norm": 0.17393404245376587,
+      "learning_rate": 0.001,
+      "loss": 1.8711,
+      "step": 16095
+    },
+    {
+      "epoch": 0.6809374735595228,
+      "grad_norm": 0.15561740100383759,
+      "learning_rate": 0.001,
+      "loss": 1.4363,
+      "step": 16096
+    },
+    {
+      "epoch": 0.6809797783230391,
+      "grad_norm": 0.16957683861255646,
+      "learning_rate": 0.001,
+      "loss": 2.131,
+      "step": 16097
+    },
+    {
+      "epoch": 0.6810220830865555,
+      "grad_norm": 0.2562768757343292,
+      "learning_rate": 0.001,
+      "loss": 3.0767,
+      "step": 16098
+    },
+    {
+      "epoch": 0.681064387850072,
+      "grad_norm": 1.1886937618255615,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 16099
+    },
+    {
+      "epoch": 0.6811066926135882,
+      "grad_norm": 0.5651175379753113,
+      "learning_rate": 0.001,
+      "loss": 3.2804,
+      "step": 16100
+    },
+    {
+      "epoch": 0.6811489973771047,
+      "grad_norm": 0.407637357711792,
+      "learning_rate": 0.001,
+      "loss": 2.6736,
+      "step": 16101
+    },
+    {
+      "epoch": 0.6811913021406211,
+      "grad_norm": 0.19160562753677368,
+      "learning_rate": 0.001,
+      "loss": 2.064,
+      "step": 16102
+    },
+    {
+      "epoch": 0.6812336069041374,
+      "grad_norm": 0.2181776911020279,
+      "learning_rate": 0.001,
+      "loss": 2.0987,
+      "step": 16103
+    },
+    {
+      "epoch": 0.6812759116676538,
+      "grad_norm": 0.22167150676250458,
+      "learning_rate": 0.001,
+      "loss": 1.9636,
+      "step": 16104
+    },
+    {
+      "epoch": 0.6813182164311702,
+      "grad_norm": 0.4942590892314911,
+      "learning_rate": 0.001,
+      "loss": 3.871,
+      "step": 16105
+    },
+    {
+      "epoch": 0.6813605211946865,
+      "grad_norm": 4.270835876464844,
+      "learning_rate": 0.001,
+      "loss": 2.1507,
+      "step": 16106
+    },
+    {
+      "epoch": 0.6814028259582029,
+      "grad_norm": 2.0456957817077637,
+      "learning_rate": 0.001,
+      "loss": 2.1891,
+      "step": 16107
+    },
+    {
+      "epoch": 0.6814451307217193,
+      "grad_norm": 0.3435433506965637,
+      "learning_rate": 0.001,
+      "loss": 3.4219,
+      "step": 16108
+    },
+    {
+      "epoch": 0.6814874354852356,
+      "grad_norm": 0.3711515963077545,
+      "learning_rate": 0.001,
+      "loss": 3.505,
+      "step": 16109
+    },
+    {
+      "epoch": 0.681529740248752,
+      "grad_norm": 0.40264999866485596,
+      "learning_rate": 0.001,
+      "loss": 4.1824,
+      "step": 16110
+    },
+    {
+      "epoch": 0.6815720450122684,
+      "grad_norm": 0.25708243250846863,
+      "learning_rate": 0.001,
+      "loss": 3.0473,
+      "step": 16111
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.9480053782463074,
+      "learning_rate": 0.001,
+      "loss": 2.4107,
+      "step": 16112
+    },
+    {
+      "epoch": 0.6816566545393011,
+      "grad_norm": 0.49835169315338135,
+      "learning_rate": 0.001,
+      "loss": 2.4002,
+      "step": 16113
+    },
+    {
+      "epoch": 0.6816989593028175,
+      "grad_norm": 0.33678311109542847,
+      "learning_rate": 0.001,
+      "loss": 2.2966,
+      "step": 16114
+    },
+    {
+      "epoch": 0.6817412640663338,
+      "grad_norm": 0.23809859156608582,
+      "learning_rate": 0.001,
+      "loss": 2.1127,
+      "step": 16115
+    },
+    {
+      "epoch": 0.6817835688298503,
+      "grad_norm": 0.26749664545059204,
+      "learning_rate": 0.001,
+      "loss": 3.4136,
+      "step": 16116
+    },
+    {
+      "epoch": 0.6818258735933667,
+      "grad_norm": 0.27437347173690796,
+      "learning_rate": 0.001,
+      "loss": 2.4324,
+      "step": 16117
+    },
+    {
+      "epoch": 0.681868178356883,
+      "grad_norm": 0.26453253626823425,
+      "learning_rate": 0.001,
+      "loss": 2.0741,
+      "step": 16118
+    },
+    {
+      "epoch": 0.6819104831203994,
+      "grad_norm": 0.18699724972248077,
+      "learning_rate": 0.001,
+      "loss": 2.6181,
+      "step": 16119
+    },
+    {
+      "epoch": 0.6819527878839158,
+      "grad_norm": 0.23921284079551697,
+      "learning_rate": 0.001,
+      "loss": 1.9422,
+      "step": 16120
+    },
+    {
+      "epoch": 0.6819950926474321,
+      "grad_norm": 0.27278849482536316,
+      "learning_rate": 0.001,
+      "loss": 2.635,
+      "step": 16121
+    },
+    {
+      "epoch": 0.6820373974109485,
+      "grad_norm": 22.81956672668457,
+      "learning_rate": 0.001,
+      "loss": 2.203,
+      "step": 16122
+    },
+    {
+      "epoch": 0.6820797021744649,
+      "grad_norm": 0.17843838036060333,
+      "learning_rate": 0.001,
+      "loss": 2.2639,
+      "step": 16123
+    },
+    {
+      "epoch": 0.6821220069379812,
+      "grad_norm": 0.17329254746437073,
+      "learning_rate": 0.001,
+      "loss": 2.0456,
+      "step": 16124
+    },
+    {
+      "epoch": 0.6821643117014976,
+      "grad_norm": 0.20157817006111145,
+      "learning_rate": 0.001,
+      "loss": 1.7671,
+      "step": 16125
+    },
+    {
+      "epoch": 0.682206616465014,
+      "grad_norm": 0.3038026988506317,
+      "learning_rate": 0.001,
+      "loss": 2.6014,
+      "step": 16126
+    },
+    {
+      "epoch": 0.6822489212285303,
+      "grad_norm": 0.25918006896972656,
+      "learning_rate": 0.001,
+      "loss": 3.0838,
+      "step": 16127
+    },
+    {
+      "epoch": 0.6822912259920467,
+      "grad_norm": 0.17397001385688782,
+      "learning_rate": 0.001,
+      "loss": 2.6565,
+      "step": 16128
+    },
+    {
+      "epoch": 0.682333530755563,
+      "grad_norm": 0.262084424495697,
+      "learning_rate": 0.001,
+      "loss": 2.5259,
+      "step": 16129
+    },
+    {
+      "epoch": 0.6823758355190794,
+      "grad_norm": 0.21808075904846191,
+      "learning_rate": 0.001,
+      "loss": 2.4083,
+      "step": 16130
+    },
+    {
+      "epoch": 0.6824181402825958,
+      "grad_norm": 0.17491497099399567,
+      "learning_rate": 0.001,
+      "loss": 2.25,
+      "step": 16131
+    },
+    {
+      "epoch": 0.6824604450461121,
+      "grad_norm": 0.1757962703704834,
+      "learning_rate": 0.001,
+      "loss": 2.1881,
+      "step": 16132
+    },
+    {
+      "epoch": 0.6825027498096286,
+      "grad_norm": 0.3888621926307678,
+      "learning_rate": 0.001,
+      "loss": 2.3828,
+      "step": 16133
+    },
+    {
+      "epoch": 0.682545054573145,
+      "grad_norm": 0.15985238552093506,
+      "learning_rate": 0.001,
+      "loss": 2.4909,
+      "step": 16134
+    },
+    {
+      "epoch": 0.6825873593366613,
+      "grad_norm": 0.21272926032543182,
+      "learning_rate": 0.001,
+      "loss": 2.7576,
+      "step": 16135
+    },
+    {
+      "epoch": 0.6826296641001777,
+      "grad_norm": 0.2001175731420517,
+      "learning_rate": 0.001,
+      "loss": 2.2182,
+      "step": 16136
+    },
+    {
+      "epoch": 0.6826719688636941,
+      "grad_norm": 0.18408158421516418,
+      "learning_rate": 0.001,
+      "loss": 1.8213,
+      "step": 16137
+    },
+    {
+      "epoch": 0.6827142736272104,
+      "grad_norm": 0.1814546287059784,
+      "learning_rate": 0.001,
+      "loss": 2.6382,
+      "step": 16138
+    },
+    {
+      "epoch": 0.6827565783907268,
+      "grad_norm": 0.17873550951480865,
+      "learning_rate": 0.001,
+      "loss": 3.3059,
+      "step": 16139
+    },
+    {
+      "epoch": 0.6827988831542432,
+      "grad_norm": 0.20914815366268158,
+      "learning_rate": 0.001,
+      "loss": 2.5146,
+      "step": 16140
+    },
+    {
+      "epoch": 0.6828411879177595,
+      "grad_norm": 9.856534004211426,
+      "learning_rate": 0.001,
+      "loss": 2.0903,
+      "step": 16141
+    },
+    {
+      "epoch": 0.6828834926812759,
+      "grad_norm": 0.19409655034542084,
+      "learning_rate": 0.001,
+      "loss": 2.5403,
+      "step": 16142
+    },
+    {
+      "epoch": 0.6829257974447923,
+      "grad_norm": 0.424080491065979,
+      "learning_rate": 0.001,
+      "loss": 3.0397,
+      "step": 16143
+    },
+    {
+      "epoch": 0.6829681022083086,
+      "grad_norm": 1.3478124141693115,
+      "learning_rate": 0.001,
+      "loss": 2.0678,
+      "step": 16144
+    },
+    {
+      "epoch": 0.683010406971825,
+      "grad_norm": 0.20509791374206543,
+      "learning_rate": 0.001,
+      "loss": 2.4558,
+      "step": 16145
+    },
+    {
+      "epoch": 0.6830527117353414,
+      "grad_norm": 0.19733795523643494,
+      "learning_rate": 0.001,
+      "loss": 2.6852,
+      "step": 16146
+    },
+    {
+      "epoch": 0.6830950164988577,
+      "grad_norm": 0.16524255275726318,
+      "learning_rate": 0.001,
+      "loss": 2.4211,
+      "step": 16147
+    },
+    {
+      "epoch": 0.6831373212623741,
+      "grad_norm": 0.17051273584365845,
+      "learning_rate": 0.001,
+      "loss": 2.6465,
+      "step": 16148
+    },
+    {
+      "epoch": 0.6831796260258906,
+      "grad_norm": 0.4586073160171509,
+      "learning_rate": 0.001,
+      "loss": 3.2835,
+      "step": 16149
+    },
+    {
+      "epoch": 0.6832219307894069,
+      "grad_norm": 0.38520026206970215,
+      "learning_rate": 0.001,
+      "loss": 1.8567,
+      "step": 16150
+    },
+    {
+      "epoch": 0.6832642355529233,
+      "grad_norm": 7.5121941566467285,
+      "learning_rate": 0.001,
+      "loss": 2.7541,
+      "step": 16151
+    },
+    {
+      "epoch": 0.6833065403164397,
+      "grad_norm": 0.19074460864067078,
+      "learning_rate": 0.001,
+      "loss": 2.2912,
+      "step": 16152
+    },
+    {
+      "epoch": 0.683348845079956,
+      "grad_norm": 0.28227201104164124,
+      "learning_rate": 0.001,
+      "loss": 1.4986,
+      "step": 16153
+    },
+    {
+      "epoch": 0.6833911498434724,
+      "grad_norm": 0.1890229731798172,
+      "learning_rate": 0.001,
+      "loss": 2.4198,
+      "step": 16154
+    },
+    {
+      "epoch": 0.6834334546069888,
+      "grad_norm": 0.3094002902507782,
+      "learning_rate": 0.001,
+      "loss": 1.7272,
+      "step": 16155
+    },
+    {
+      "epoch": 0.6834757593705051,
+      "grad_norm": 0.4410035014152527,
+      "learning_rate": 0.001,
+      "loss": 1.7713,
+      "step": 16156
+    },
+    {
+      "epoch": 0.6835180641340215,
+      "grad_norm": 0.20090627670288086,
+      "learning_rate": 0.001,
+      "loss": 2.3903,
+      "step": 16157
+    },
+    {
+      "epoch": 0.6835603688975379,
+      "grad_norm": 0.1748632788658142,
+      "learning_rate": 0.001,
+      "loss": 2.1165,
+      "step": 16158
+    },
+    {
+      "epoch": 0.6836026736610542,
+      "grad_norm": 0.18395616114139557,
+      "learning_rate": 0.001,
+      "loss": 1.9808,
+      "step": 16159
+    },
+    {
+      "epoch": 0.6836449784245706,
+      "grad_norm": 0.18044279515743256,
+      "learning_rate": 0.001,
+      "loss": 2.2385,
+      "step": 16160
+    },
+    {
+      "epoch": 0.683687283188087,
+      "grad_norm": 0.7966620922088623,
+      "learning_rate": 0.001,
+      "loss": 2.1261,
+      "step": 16161
+    },
+    {
+      "epoch": 0.6837295879516033,
+      "grad_norm": 0.5675308108329773,
+      "learning_rate": 0.001,
+      "loss": 1.9956,
+      "step": 16162
+    },
+    {
+      "epoch": 0.6837718927151197,
+      "grad_norm": 0.15210571885108948,
+      "learning_rate": 0.001,
+      "loss": 1.8715,
+      "step": 16163
+    },
+    {
+      "epoch": 0.6838141974786361,
+      "grad_norm": 0.19697412848472595,
+      "learning_rate": 0.001,
+      "loss": 2.2049,
+      "step": 16164
+    },
+    {
+      "epoch": 0.6838565022421524,
+      "grad_norm": 0.2632860243320465,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 16165
+    },
+    {
+      "epoch": 0.6838988070056689,
+      "grad_norm": 0.14289480447769165,
+      "learning_rate": 0.001,
+      "loss": 2.3733,
+      "step": 16166
+    },
+    {
+      "epoch": 0.6839411117691853,
+      "grad_norm": 0.16946500539779663,
+      "learning_rate": 0.001,
+      "loss": 2.1708,
+      "step": 16167
+    },
+    {
+      "epoch": 0.6839834165327016,
+      "grad_norm": 0.1758553683757782,
+      "learning_rate": 0.001,
+      "loss": 1.8589,
+      "step": 16168
+    },
+    {
+      "epoch": 0.684025721296218,
+      "grad_norm": 0.1700044721364975,
+      "learning_rate": 0.001,
+      "loss": 1.6158,
+      "step": 16169
+    },
+    {
+      "epoch": 0.6840680260597344,
+      "grad_norm": 0.15607595443725586,
+      "learning_rate": 0.001,
+      "loss": 1.6295,
+      "step": 16170
+    },
+    {
+      "epoch": 0.6841103308232507,
+      "grad_norm": 0.173630490899086,
+      "learning_rate": 0.001,
+      "loss": 2.871,
+      "step": 16171
+    },
+    {
+      "epoch": 0.6841526355867671,
+      "grad_norm": 0.3207765817642212,
+      "learning_rate": 0.001,
+      "loss": 2.2475,
+      "step": 16172
+    },
+    {
+      "epoch": 0.6841949403502834,
+      "grad_norm": 0.1741725355386734,
+      "learning_rate": 0.001,
+      "loss": 2.2591,
+      "step": 16173
+    },
+    {
+      "epoch": 0.6842372451137998,
+      "grad_norm": 0.17393991351127625,
+      "learning_rate": 0.001,
+      "loss": 2.1785,
+      "step": 16174
+    },
+    {
+      "epoch": 0.6842795498773162,
+      "grad_norm": 0.15135467052459717,
+      "learning_rate": 0.001,
+      "loss": 2.0614,
+      "step": 16175
+    },
+    {
+      "epoch": 0.6843218546408325,
+      "grad_norm": 0.24743154644966125,
+      "learning_rate": 0.001,
+      "loss": 2.5485,
+      "step": 16176
+    },
+    {
+      "epoch": 0.6843641594043489,
+      "grad_norm": 0.28311067819595337,
+      "learning_rate": 0.001,
+      "loss": 2.6526,
+      "step": 16177
+    },
+    {
+      "epoch": 0.6844064641678653,
+      "grad_norm": 0.14594843983650208,
+      "learning_rate": 0.001,
+      "loss": 2.3248,
+      "step": 16178
+    },
+    {
+      "epoch": 0.6844487689313816,
+      "grad_norm": 2.94158935546875,
+      "learning_rate": 0.001,
+      "loss": 2.2385,
+      "step": 16179
+    },
+    {
+      "epoch": 0.684491073694898,
+      "grad_norm": 0.19169901311397552,
+      "learning_rate": 0.001,
+      "loss": 2.3112,
+      "step": 16180
+    },
+    {
+      "epoch": 0.6845333784584144,
+      "grad_norm": 0.6225913763046265,
+      "learning_rate": 0.001,
+      "loss": 2.5312,
+      "step": 16181
+    },
+    {
+      "epoch": 0.6845756832219307,
+      "grad_norm": 0.1315428614616394,
+      "learning_rate": 0.001,
+      "loss": 2.1162,
+      "step": 16182
+    },
+    {
+      "epoch": 0.6846179879854472,
+      "grad_norm": 0.1488700956106186,
+      "learning_rate": 0.001,
+      "loss": 1.6814,
+      "step": 16183
+    },
+    {
+      "epoch": 0.6846602927489636,
+      "grad_norm": 0.1776464730501175,
+      "learning_rate": 0.001,
+      "loss": 2.0227,
+      "step": 16184
+    },
+    {
+      "epoch": 0.6847025975124799,
+      "grad_norm": 0.17623931169509888,
+      "learning_rate": 0.001,
+      "loss": 2.0609,
+      "step": 16185
+    },
+    {
+      "epoch": 0.6847449022759963,
+      "grad_norm": 0.2685874402523041,
+      "learning_rate": 0.001,
+      "loss": 2.3874,
+      "step": 16186
+    },
+    {
+      "epoch": 0.6847872070395127,
+      "grad_norm": 0.20432405173778534,
+      "learning_rate": 0.001,
+      "loss": 1.6907,
+      "step": 16187
+    },
+    {
+      "epoch": 0.684829511803029,
+      "grad_norm": 0.17755457758903503,
+      "learning_rate": 0.001,
+      "loss": 1.7708,
+      "step": 16188
+    },
+    {
+      "epoch": 0.6848718165665454,
+      "grad_norm": 0.3379620611667633,
+      "learning_rate": 0.001,
+      "loss": 2.383,
+      "step": 16189
+    },
+    {
+      "epoch": 0.6849141213300618,
+      "grad_norm": 0.31830352544784546,
+      "learning_rate": 0.001,
+      "loss": 1.7277,
+      "step": 16190
+    },
+    {
+      "epoch": 0.6849564260935781,
+      "grad_norm": 0.18110892176628113,
+      "learning_rate": 0.001,
+      "loss": 3.5317,
+      "step": 16191
+    },
+    {
+      "epoch": 0.6849987308570945,
+      "grad_norm": 0.5665266513824463,
+      "learning_rate": 0.001,
+      "loss": 2.4427,
+      "step": 16192
+    },
+    {
+      "epoch": 0.6850410356206109,
+      "grad_norm": 0.28646814823150635,
+      "learning_rate": 0.001,
+      "loss": 3.2973,
+      "step": 16193
+    },
+    {
+      "epoch": 0.6850833403841272,
+      "grad_norm": 0.19316011667251587,
+      "learning_rate": 0.001,
+      "loss": 2.7649,
+      "step": 16194
+    },
+    {
+      "epoch": 0.6851256451476436,
+      "grad_norm": 0.21821334958076477,
+      "learning_rate": 0.001,
+      "loss": 2.3747,
+      "step": 16195
+    },
+    {
+      "epoch": 0.68516794991116,
+      "grad_norm": 0.18347589671611786,
+      "learning_rate": 0.001,
+      "loss": 2.0008,
+      "step": 16196
+    },
+    {
+      "epoch": 0.6852102546746763,
+      "grad_norm": 0.1854442059993744,
+      "learning_rate": 0.001,
+      "loss": 1.7193,
+      "step": 16197
+    },
+    {
+      "epoch": 0.6852525594381927,
+      "grad_norm": 0.17909593880176544,
+      "learning_rate": 0.001,
+      "loss": 2.0372,
+      "step": 16198
+    },
+    {
+      "epoch": 0.6852948642017092,
+      "grad_norm": 0.1535256803035736,
+      "learning_rate": 0.001,
+      "loss": 1.7345,
+      "step": 16199
+    },
+    {
+      "epoch": 0.6853371689652255,
+      "grad_norm": 0.20372457802295685,
+      "learning_rate": 0.001,
+      "loss": 2.1256,
+      "step": 16200
+    },
+    {
+      "epoch": 0.6853794737287419,
+      "grad_norm": 0.291826456785202,
+      "learning_rate": 0.001,
+      "loss": 2.5165,
+      "step": 16201
+    },
+    {
+      "epoch": 0.6854217784922583,
+      "grad_norm": 0.18575716018676758,
+      "learning_rate": 0.001,
+      "loss": 1.7661,
+      "step": 16202
+    },
+    {
+      "epoch": 0.6854640832557746,
+      "grad_norm": 0.15441353619098663,
+      "learning_rate": 0.001,
+      "loss": 1.672,
+      "step": 16203
+    },
+    {
+      "epoch": 0.685506388019291,
+      "grad_norm": 1.2076762914657593,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 16204
+    },
+    {
+      "epoch": 0.6855486927828074,
+      "grad_norm": 0.17845335602760315,
+      "learning_rate": 0.001,
+      "loss": 2.2695,
+      "step": 16205
+    },
+    {
+      "epoch": 0.6855909975463237,
+      "grad_norm": 0.1648271232843399,
+      "learning_rate": 0.001,
+      "loss": 2.9431,
+      "step": 16206
+    },
+    {
+      "epoch": 0.6856333023098401,
+      "grad_norm": 0.264567494392395,
+      "learning_rate": 0.001,
+      "loss": 2.7659,
+      "step": 16207
+    },
+    {
+      "epoch": 0.6856756070733565,
+      "grad_norm": 0.15774111449718475,
+      "learning_rate": 0.001,
+      "loss": 1.7371,
+      "step": 16208
+    },
+    {
+      "epoch": 0.6857179118368728,
+      "grad_norm": 0.3049740195274353,
+      "learning_rate": 0.001,
+      "loss": 2.4582,
+      "step": 16209
+    },
+    {
+      "epoch": 0.6857602166003892,
+      "grad_norm": 0.371410071849823,
+      "learning_rate": 0.001,
+      "loss": 3.2076,
+      "step": 16210
+    },
+    {
+      "epoch": 0.6858025213639056,
+      "grad_norm": 0.17293484508991241,
+      "learning_rate": 0.001,
+      "loss": 1.4464,
+      "step": 16211
+    },
+    {
+      "epoch": 0.6858448261274219,
+      "grad_norm": 0.3040180802345276,
+      "learning_rate": 0.001,
+      "loss": 2.9824,
+      "step": 16212
+    },
+    {
+      "epoch": 0.6858871308909383,
+      "grad_norm": 0.22977972030639648,
+      "learning_rate": 0.001,
+      "loss": 3.5424,
+      "step": 16213
+    },
+    {
+      "epoch": 0.6859294356544547,
+      "grad_norm": 0.14453327655792236,
+      "learning_rate": 0.001,
+      "loss": 1.4584,
+      "step": 16214
+    },
+    {
+      "epoch": 0.685971740417971,
+      "grad_norm": 0.3432617485523224,
+      "learning_rate": 0.001,
+      "loss": 2.2774,
+      "step": 16215
+    },
+    {
+      "epoch": 0.6860140451814875,
+      "grad_norm": 0.614466667175293,
+      "learning_rate": 0.001,
+      "loss": 2.9487,
+      "step": 16216
+    },
+    {
+      "epoch": 0.6860563499450039,
+      "grad_norm": 0.5213249325752258,
+      "learning_rate": 0.001,
+      "loss": 2.6308,
+      "step": 16217
+    },
+    {
+      "epoch": 0.6860986547085202,
+      "grad_norm": 0.13962678611278534,
+      "learning_rate": 0.001,
+      "loss": 2.3401,
+      "step": 16218
+    },
+    {
+      "epoch": 0.6861409594720366,
+      "grad_norm": 0.13959261775016785,
+      "learning_rate": 0.001,
+      "loss": 2.066,
+      "step": 16219
+    },
+    {
+      "epoch": 0.6861832642355529,
+      "grad_norm": 0.12609069049358368,
+      "learning_rate": 0.001,
+      "loss": 2.7186,
+      "step": 16220
+    },
+    {
+      "epoch": 0.6862255689990693,
+      "grad_norm": 0.1317724585533142,
+      "learning_rate": 0.001,
+      "loss": 1.6887,
+      "step": 16221
+    },
+    {
+      "epoch": 0.6862678737625857,
+      "grad_norm": 0.22671738266944885,
+      "learning_rate": 0.001,
+      "loss": 1.519,
+      "step": 16222
+    },
+    {
+      "epoch": 0.686310178526102,
+      "grad_norm": 0.21472686529159546,
+      "learning_rate": 0.001,
+      "loss": 2.5224,
+      "step": 16223
+    },
+    {
+      "epoch": 0.6863524832896184,
+      "grad_norm": 0.5818791389465332,
+      "learning_rate": 0.001,
+      "loss": 1.7538,
+      "step": 16224
+    },
+    {
+      "epoch": 0.6863947880531348,
+      "grad_norm": 0.22420547902584076,
+      "learning_rate": 0.001,
+      "loss": 1.5812,
+      "step": 16225
+    },
+    {
+      "epoch": 0.6864370928166511,
+      "grad_norm": 0.16626037657260895,
+      "learning_rate": 0.001,
+      "loss": 1.8388,
+      "step": 16226
+    },
+    {
+      "epoch": 0.6864793975801675,
+      "grad_norm": 0.19007837772369385,
+      "learning_rate": 0.001,
+      "loss": 2.5345,
+      "step": 16227
+    },
+    {
+      "epoch": 0.6865217023436839,
+      "grad_norm": 0.3126547932624817,
+      "learning_rate": 0.001,
+      "loss": 2.5063,
+      "step": 16228
+    },
+    {
+      "epoch": 0.6865640071072002,
+      "grad_norm": 0.2687819004058838,
+      "learning_rate": 0.001,
+      "loss": 3.085,
+      "step": 16229
+    },
+    {
+      "epoch": 0.6866063118707166,
+      "grad_norm": 0.9603768587112427,
+      "learning_rate": 0.001,
+      "loss": 1.8578,
+      "step": 16230
+    },
+    {
+      "epoch": 0.686648616634233,
+      "grad_norm": 0.1472010314464569,
+      "learning_rate": 0.001,
+      "loss": 3.0854,
+      "step": 16231
+    },
+    {
+      "epoch": 0.6866909213977493,
+      "grad_norm": 0.20346207916736603,
+      "learning_rate": 0.001,
+      "loss": 1.8353,
+      "step": 16232
+    },
+    {
+      "epoch": 0.6867332261612658,
+      "grad_norm": 0.17270101606845856,
+      "learning_rate": 0.001,
+      "loss": 3.0351,
+      "step": 16233
+    },
+    {
+      "epoch": 0.6867755309247822,
+      "grad_norm": 0.14880718290805817,
+      "learning_rate": 0.001,
+      "loss": 2.1701,
+      "step": 16234
+    },
+    {
+      "epoch": 0.6868178356882985,
+      "grad_norm": 0.23077774047851562,
+      "learning_rate": 0.001,
+      "loss": 3.2428,
+      "step": 16235
+    },
+    {
+      "epoch": 0.6868601404518149,
+      "grad_norm": 0.14143741130828857,
+      "learning_rate": 0.001,
+      "loss": 1.7177,
+      "step": 16236
+    },
+    {
+      "epoch": 0.6869024452153313,
+      "grad_norm": 0.15532909333705902,
+      "learning_rate": 0.001,
+      "loss": 1.4819,
+      "step": 16237
+    },
+    {
+      "epoch": 0.6869447499788476,
+      "grad_norm": 1.5468194484710693,
+      "learning_rate": 0.001,
+      "loss": 1.6917,
+      "step": 16238
+    },
+    {
+      "epoch": 0.686987054742364,
+      "grad_norm": 0.16617412865161896,
+      "learning_rate": 0.001,
+      "loss": 1.4165,
+      "step": 16239
+    },
+    {
+      "epoch": 0.6870293595058804,
+      "grad_norm": 0.25191357731819153,
+      "learning_rate": 0.001,
+      "loss": 2.1047,
+      "step": 16240
+    },
+    {
+      "epoch": 0.6870716642693967,
+      "grad_norm": 0.15333382785320282,
+      "learning_rate": 0.001,
+      "loss": 1.978,
+      "step": 16241
+    },
+    {
+      "epoch": 0.6871139690329131,
+      "grad_norm": 0.16414469480514526,
+      "learning_rate": 0.001,
+      "loss": 2.0745,
+      "step": 16242
+    },
+    {
+      "epoch": 0.6871562737964295,
+      "grad_norm": 1.5129097700119019,
+      "learning_rate": 0.001,
+      "loss": 2.2844,
+      "step": 16243
+    },
+    {
+      "epoch": 0.6871985785599458,
+      "grad_norm": 0.18176978826522827,
+      "learning_rate": 0.001,
+      "loss": 2.1622,
+      "step": 16244
+    },
+    {
+      "epoch": 0.6872408833234622,
+      "grad_norm": 0.17760765552520752,
+      "learning_rate": 0.001,
+      "loss": 2.9898,
+      "step": 16245
+    },
+    {
+      "epoch": 0.6872831880869786,
+      "grad_norm": 0.2571372985839844,
+      "learning_rate": 0.001,
+      "loss": 2.6795,
+      "step": 16246
+    },
+    {
+      "epoch": 0.6873254928504949,
+      "grad_norm": 0.20494753122329712,
+      "learning_rate": 0.001,
+      "loss": 1.6626,
+      "step": 16247
+    },
+    {
+      "epoch": 0.6873677976140113,
+      "grad_norm": 0.9767950773239136,
+      "learning_rate": 0.001,
+      "loss": 1.6601,
+      "step": 16248
+    },
+    {
+      "epoch": 0.6874101023775278,
+      "grad_norm": 0.2790040075778961,
+      "learning_rate": 0.001,
+      "loss": 3.1638,
+      "step": 16249
+    },
+    {
+      "epoch": 0.687452407141044,
+      "grad_norm": 0.2396850734949112,
+      "learning_rate": 0.001,
+      "loss": 1.8693,
+      "step": 16250
+    },
+    {
+      "epoch": 0.6874947119045605,
+      "grad_norm": 1.0322060585021973,
+      "learning_rate": 0.001,
+      "loss": 2.7673,
+      "step": 16251
+    },
+    {
+      "epoch": 0.6875370166680769,
+      "grad_norm": 0.15175770223140717,
+      "learning_rate": 0.001,
+      "loss": 3.0304,
+      "step": 16252
+    },
+    {
+      "epoch": 0.6875793214315932,
+      "grad_norm": 0.18464677035808563,
+      "learning_rate": 0.001,
+      "loss": 2.1256,
+      "step": 16253
+    },
+    {
+      "epoch": 0.6876216261951096,
+      "grad_norm": 0.16536930203437805,
+      "learning_rate": 0.001,
+      "loss": 2.0773,
+      "step": 16254
+    },
+    {
+      "epoch": 0.687663930958626,
+      "grad_norm": 0.20809626579284668,
+      "learning_rate": 0.001,
+      "loss": 1.8316,
+      "step": 16255
+    },
+    {
+      "epoch": 0.6877062357221423,
+      "grad_norm": 0.163310706615448,
+      "learning_rate": 0.001,
+      "loss": 2.1962,
+      "step": 16256
+    },
+    {
+      "epoch": 0.6877485404856587,
+      "grad_norm": 0.39365145564079285,
+      "learning_rate": 0.001,
+      "loss": 2.7742,
+      "step": 16257
+    },
+    {
+      "epoch": 0.6877908452491751,
+      "grad_norm": 0.5424996018409729,
+      "learning_rate": 0.001,
+      "loss": 3.4448,
+      "step": 16258
+    },
+    {
+      "epoch": 0.6878331500126914,
+      "grad_norm": 0.1544448286294937,
+      "learning_rate": 0.001,
+      "loss": 1.9199,
+      "step": 16259
+    },
+    {
+      "epoch": 0.6878754547762078,
+      "grad_norm": 10.53158950805664,
+      "learning_rate": 0.001,
+      "loss": 2.1641,
+      "step": 16260
+    },
+    {
+      "epoch": 0.6879177595397242,
+      "grad_norm": 0.3506162166595459,
+      "learning_rate": 0.001,
+      "loss": 2.6466,
+      "step": 16261
+    },
+    {
+      "epoch": 0.6879600643032405,
+      "grad_norm": 0.16496671736240387,
+      "learning_rate": 0.001,
+      "loss": 2.5371,
+      "step": 16262
+    },
+    {
+      "epoch": 0.6880023690667569,
+      "grad_norm": 0.1795535832643509,
+      "learning_rate": 0.001,
+      "loss": 3.5854,
+      "step": 16263
+    },
+    {
+      "epoch": 0.6880446738302732,
+      "grad_norm": 0.1523294448852539,
+      "learning_rate": 0.001,
+      "loss": 1.3369,
+      "step": 16264
+    },
+    {
+      "epoch": 0.6880869785937896,
+      "grad_norm": 0.16334094107151031,
+      "learning_rate": 0.001,
+      "loss": 1.9804,
+      "step": 16265
+    },
+    {
+      "epoch": 0.688129283357306,
+      "grad_norm": 0.36673104763031006,
+      "learning_rate": 0.001,
+      "loss": 2.0428,
+      "step": 16266
+    },
+    {
+      "epoch": 0.6881715881208224,
+      "grad_norm": 0.23456034064292908,
+      "learning_rate": 0.001,
+      "loss": 1.778,
+      "step": 16267
+    },
+    {
+      "epoch": 0.6882138928843388,
+      "grad_norm": 0.1658594310283661,
+      "learning_rate": 0.001,
+      "loss": 1.8266,
+      "step": 16268
+    },
+    {
+      "epoch": 0.6882561976478552,
+      "grad_norm": 0.17385973036289215,
+      "learning_rate": 0.001,
+      "loss": 2.7395,
+      "step": 16269
+    },
+    {
+      "epoch": 0.6882985024113715,
+      "grad_norm": 0.20876826345920563,
+      "learning_rate": 0.001,
+      "loss": 2.2943,
+      "step": 16270
+    },
+    {
+      "epoch": 0.6883408071748879,
+      "grad_norm": 0.2948185205459595,
+      "learning_rate": 0.001,
+      "loss": 2.295,
+      "step": 16271
+    },
+    {
+      "epoch": 0.6883831119384043,
+      "grad_norm": 0.1359289586544037,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 16272
+    },
+    {
+      "epoch": 0.6884254167019206,
+      "grad_norm": 0.1961379051208496,
+      "learning_rate": 0.001,
+      "loss": 2.3073,
+      "step": 16273
+    },
+    {
+      "epoch": 0.688467721465437,
+      "grad_norm": 0.2041141390800476,
+      "learning_rate": 0.001,
+      "loss": 1.6394,
+      "step": 16274
+    },
+    {
+      "epoch": 0.6885100262289534,
+      "grad_norm": 7.983102798461914,
+      "learning_rate": 0.001,
+      "loss": 1.5184,
+      "step": 16275
+    },
+    {
+      "epoch": 0.6885523309924697,
+      "grad_norm": 0.1904926300048828,
+      "learning_rate": 0.001,
+      "loss": 2.1257,
+      "step": 16276
+    },
+    {
+      "epoch": 0.6885946357559861,
+      "grad_norm": 0.16287410259246826,
+      "learning_rate": 0.001,
+      "loss": 2.0357,
+      "step": 16277
+    },
+    {
+      "epoch": 0.6886369405195025,
+      "grad_norm": 0.17052966356277466,
+      "learning_rate": 0.001,
+      "loss": 2.5068,
+      "step": 16278
+    },
+    {
+      "epoch": 0.6886792452830188,
+      "grad_norm": 0.35149747133255005,
+      "learning_rate": 0.001,
+      "loss": 2.5683,
+      "step": 16279
+    },
+    {
+      "epoch": 0.6887215500465352,
+      "grad_norm": 1.4754235744476318,
+      "learning_rate": 0.001,
+      "loss": 2.9501,
+      "step": 16280
+    },
+    {
+      "epoch": 0.6887638548100516,
+      "grad_norm": 0.1624482125043869,
+      "learning_rate": 0.001,
+      "loss": 1.9048,
+      "step": 16281
+    },
+    {
+      "epoch": 0.6888061595735679,
+      "grad_norm": 1.466571569442749,
+      "learning_rate": 0.001,
+      "loss": 2.5641,
+      "step": 16282
+    },
+    {
+      "epoch": 0.6888484643370844,
+      "grad_norm": 0.18369127810001373,
+      "learning_rate": 0.001,
+      "loss": 1.8939,
+      "step": 16283
+    },
+    {
+      "epoch": 0.6888907691006008,
+      "grad_norm": 0.31858545541763306,
+      "learning_rate": 0.001,
+      "loss": 2.515,
+      "step": 16284
+    },
+    {
+      "epoch": 0.6889330738641171,
+      "grad_norm": 1.7627861499786377,
+      "learning_rate": 0.001,
+      "loss": 2.8145,
+      "step": 16285
+    },
+    {
+      "epoch": 0.6889753786276335,
+      "grad_norm": 0.20102620124816895,
+      "learning_rate": 0.001,
+      "loss": 2.8067,
+      "step": 16286
+    },
+    {
+      "epoch": 0.6890176833911499,
+      "grad_norm": 0.20633406937122345,
+      "learning_rate": 0.001,
+      "loss": 2.5596,
+      "step": 16287
+    },
+    {
+      "epoch": 0.6890599881546662,
+      "grad_norm": 0.2921423017978668,
+      "learning_rate": 0.001,
+      "loss": 2.6385,
+      "step": 16288
+    },
+    {
+      "epoch": 0.6891022929181826,
+      "grad_norm": 1.0566281080245972,
+      "learning_rate": 0.001,
+      "loss": 3.4956,
+      "step": 16289
+    },
+    {
+      "epoch": 0.689144597681699,
+      "grad_norm": 0.23099292814731598,
+      "learning_rate": 0.001,
+      "loss": 2.0018,
+      "step": 16290
+    },
+    {
+      "epoch": 0.6891869024452153,
+      "grad_norm": 0.19271378219127655,
+      "learning_rate": 0.001,
+      "loss": 2.4489,
+      "step": 16291
+    },
+    {
+      "epoch": 0.6892292072087317,
+      "grad_norm": 0.20972347259521484,
+      "learning_rate": 0.001,
+      "loss": 2.3633,
+      "step": 16292
+    },
+    {
+      "epoch": 0.6892715119722481,
+      "grad_norm": 0.17728467285633087,
+      "learning_rate": 0.001,
+      "loss": 1.8417,
+      "step": 16293
+    },
+    {
+      "epoch": 0.6893138167357644,
+      "grad_norm": 1.631922960281372,
+      "learning_rate": 0.001,
+      "loss": 1.903,
+      "step": 16294
+    },
+    {
+      "epoch": 0.6893561214992808,
+      "grad_norm": 0.39803799986839294,
+      "learning_rate": 0.001,
+      "loss": 2.5417,
+      "step": 16295
+    },
+    {
+      "epoch": 0.6893984262627972,
+      "grad_norm": 0.206565260887146,
+      "learning_rate": 0.001,
+      "loss": 2.3135,
+      "step": 16296
+    },
+    {
+      "epoch": 0.6894407310263135,
+      "grad_norm": 0.18916979432106018,
+      "learning_rate": 0.001,
+      "loss": 1.8302,
+      "step": 16297
+    },
+    {
+      "epoch": 0.68948303578983,
+      "grad_norm": 0.21091687679290771,
+      "learning_rate": 0.001,
+      "loss": 2.1947,
+      "step": 16298
+    },
+    {
+      "epoch": 0.6895253405533464,
+      "grad_norm": 0.16629081964492798,
+      "learning_rate": 0.001,
+      "loss": 3.1085,
+      "step": 16299
+    },
+    {
+      "epoch": 0.6895676453168627,
+      "grad_norm": 0.16846491396427155,
+      "learning_rate": 0.001,
+      "loss": 1.5303,
+      "step": 16300
+    },
+    {
+      "epoch": 0.6896099500803791,
+      "grad_norm": 0.1889643669128418,
+      "learning_rate": 0.001,
+      "loss": 2.0213,
+      "step": 16301
+    },
+    {
+      "epoch": 0.6896522548438955,
+      "grad_norm": 0.17257481813430786,
+      "learning_rate": 0.001,
+      "loss": 2.3829,
+      "step": 16302
+    },
+    {
+      "epoch": 0.6896945596074118,
+      "grad_norm": 0.17580455541610718,
+      "learning_rate": 0.001,
+      "loss": 2.7781,
+      "step": 16303
+    },
+    {
+      "epoch": 0.6897368643709282,
+      "grad_norm": 0.1580284833908081,
+      "learning_rate": 0.001,
+      "loss": 2.3906,
+      "step": 16304
+    },
+    {
+      "epoch": 0.6897791691344446,
+      "grad_norm": 0.1577870398759842,
+      "learning_rate": 0.001,
+      "loss": 2.7536,
+      "step": 16305
+    },
+    {
+      "epoch": 0.6898214738979609,
+      "grad_norm": 0.17416058480739594,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 16306
+    },
+    {
+      "epoch": 0.6898637786614773,
+      "grad_norm": 0.7185014486312866,
+      "learning_rate": 0.001,
+      "loss": 1.3881,
+      "step": 16307
+    },
+    {
+      "epoch": 0.6899060834249936,
+      "grad_norm": 0.23386551439762115,
+      "learning_rate": 0.001,
+      "loss": 3.7348,
+      "step": 16308
+    },
+    {
+      "epoch": 0.68994838818851,
+      "grad_norm": 0.1478804051876068,
+      "learning_rate": 0.001,
+      "loss": 1.8103,
+      "step": 16309
+    },
+    {
+      "epoch": 0.6899906929520264,
+      "grad_norm": 0.14378052949905396,
+      "learning_rate": 0.001,
+      "loss": 1.8187,
+      "step": 16310
+    },
+    {
+      "epoch": 0.6900329977155427,
+      "grad_norm": 0.19731678068637848,
+      "learning_rate": 0.001,
+      "loss": 2.5575,
+      "step": 16311
+    },
+    {
+      "epoch": 0.6900753024790591,
+      "grad_norm": 0.6392672657966614,
+      "learning_rate": 0.001,
+      "loss": 2.2473,
+      "step": 16312
+    },
+    {
+      "epoch": 0.6901176072425755,
+      "grad_norm": 0.5831864476203918,
+      "learning_rate": 0.001,
+      "loss": 1.6424,
+      "step": 16313
+    },
+    {
+      "epoch": 0.6901599120060918,
+      "grad_norm": 0.573341965675354,
+      "learning_rate": 0.001,
+      "loss": 1.8507,
+      "step": 16314
+    },
+    {
+      "epoch": 0.6902022167696082,
+      "grad_norm": 0.1696714460849762,
+      "learning_rate": 0.001,
+      "loss": 2.02,
+      "step": 16315
+    },
+    {
+      "epoch": 0.6902445215331247,
+      "grad_norm": 0.8533129692077637,
+      "learning_rate": 0.001,
+      "loss": 2.3024,
+      "step": 16316
+    },
+    {
+      "epoch": 0.690286826296641,
+      "grad_norm": 0.9942197203636169,
+      "learning_rate": 0.001,
+      "loss": 1.8328,
+      "step": 16317
+    },
+    {
+      "epoch": 0.6903291310601574,
+      "grad_norm": 0.587260901927948,
+      "learning_rate": 0.001,
+      "loss": 2.375,
+      "step": 16318
+    },
+    {
+      "epoch": 0.6903714358236738,
+      "grad_norm": 0.17391139268875122,
+      "learning_rate": 0.001,
+      "loss": 1.7228,
+      "step": 16319
+    },
+    {
+      "epoch": 0.6904137405871901,
+      "grad_norm": 0.437977135181427,
+      "learning_rate": 0.001,
+      "loss": 2.7662,
+      "step": 16320
+    },
+    {
+      "epoch": 0.6904560453507065,
+      "grad_norm": 0.2368604987859726,
+      "learning_rate": 0.001,
+      "loss": 2.2516,
+      "step": 16321
+    },
+    {
+      "epoch": 0.6904983501142229,
+      "grad_norm": 0.1805589199066162,
+      "learning_rate": 0.001,
+      "loss": 2.1679,
+      "step": 16322
+    },
+    {
+      "epoch": 0.6905406548777392,
+      "grad_norm": 0.6197409629821777,
+      "learning_rate": 0.001,
+      "loss": 1.9621,
+      "step": 16323
+    },
+    {
+      "epoch": 0.6905829596412556,
+      "grad_norm": 0.16126328706741333,
+      "learning_rate": 0.001,
+      "loss": 2.1508,
+      "step": 16324
+    },
+    {
+      "epoch": 0.690625264404772,
+      "grad_norm": 0.17025581002235413,
+      "learning_rate": 0.001,
+      "loss": 2.3355,
+      "step": 16325
+    },
+    {
+      "epoch": 0.6906675691682883,
+      "grad_norm": 0.2806219756603241,
+      "learning_rate": 0.001,
+      "loss": 2.0408,
+      "step": 16326
+    },
+    {
+      "epoch": 0.6907098739318047,
+      "grad_norm": 0.4944941997528076,
+      "learning_rate": 0.001,
+      "loss": 3.1678,
+      "step": 16327
+    },
+    {
+      "epoch": 0.6907521786953211,
+      "grad_norm": 0.3849775195121765,
+      "learning_rate": 0.001,
+      "loss": 2.0565,
+      "step": 16328
+    },
+    {
+      "epoch": 0.6907944834588374,
+      "grad_norm": 0.17728754878044128,
+      "learning_rate": 0.001,
+      "loss": 1.5861,
+      "step": 16329
+    },
+    {
+      "epoch": 0.6908367882223538,
+      "grad_norm": 3.389814615249634,
+      "learning_rate": 0.001,
+      "loss": 2.1157,
+      "step": 16330
+    },
+    {
+      "epoch": 0.6908790929858702,
+      "grad_norm": 0.16745001077651978,
+      "learning_rate": 0.001,
+      "loss": 1.7655,
+      "step": 16331
+    },
+    {
+      "epoch": 0.6909213977493865,
+      "grad_norm": 0.2383284568786621,
+      "learning_rate": 0.001,
+      "loss": 2.2718,
+      "step": 16332
+    },
+    {
+      "epoch": 0.690963702512903,
+      "grad_norm": 1.5297582149505615,
+      "learning_rate": 0.001,
+      "loss": 2.8112,
+      "step": 16333
+    },
+    {
+      "epoch": 0.6910060072764194,
+      "grad_norm": 0.17354048788547516,
+      "learning_rate": 0.001,
+      "loss": 2.7349,
+      "step": 16334
+    },
+    {
+      "epoch": 0.6910483120399357,
+      "grad_norm": 0.1723877489566803,
+      "learning_rate": 0.001,
+      "loss": 2.1994,
+      "step": 16335
+    },
+    {
+      "epoch": 0.6910906168034521,
+      "grad_norm": 0.160502627491951,
+      "learning_rate": 0.001,
+      "loss": 2.5597,
+      "step": 16336
+    },
+    {
+      "epoch": 0.6911329215669685,
+      "grad_norm": 0.3956283628940582,
+      "learning_rate": 0.001,
+      "loss": 1.5582,
+      "step": 16337
+    },
+    {
+      "epoch": 0.6911752263304848,
+      "grad_norm": 0.27428823709487915,
+      "learning_rate": 0.001,
+      "loss": 2.2426,
+      "step": 16338
+    },
+    {
+      "epoch": 0.6912175310940012,
+      "grad_norm": 0.6840664148330688,
+      "learning_rate": 0.001,
+      "loss": 2.731,
+      "step": 16339
+    },
+    {
+      "epoch": 0.6912598358575176,
+      "grad_norm": 0.1819261908531189,
+      "learning_rate": 0.001,
+      "loss": 2.0739,
+      "step": 16340
+    },
+    {
+      "epoch": 0.6913021406210339,
+      "grad_norm": 0.2000657320022583,
+      "learning_rate": 0.001,
+      "loss": 3.0519,
+      "step": 16341
+    },
+    {
+      "epoch": 0.6913444453845503,
+      "grad_norm": 0.28833070397377014,
+      "learning_rate": 0.001,
+      "loss": 2.1887,
+      "step": 16342
+    },
+    {
+      "epoch": 0.6913867501480667,
+      "grad_norm": 0.24844685196876526,
+      "learning_rate": 0.001,
+      "loss": 2.7631,
+      "step": 16343
+    },
+    {
+      "epoch": 0.691429054911583,
+      "grad_norm": 0.16090954840183258,
+      "learning_rate": 0.001,
+      "loss": 2.0467,
+      "step": 16344
+    },
+    {
+      "epoch": 0.6914713596750994,
+      "grad_norm": 0.19617071747779846,
+      "learning_rate": 0.001,
+      "loss": 1.4285,
+      "step": 16345
+    },
+    {
+      "epoch": 0.6915136644386158,
+      "grad_norm": 0.16506654024124146,
+      "learning_rate": 0.001,
+      "loss": 1.3764,
+      "step": 16346
+    },
+    {
+      "epoch": 0.6915559692021321,
+      "grad_norm": 0.175397127866745,
+      "learning_rate": 0.001,
+      "loss": 1.7242,
+      "step": 16347
+    },
+    {
+      "epoch": 0.6915982739656485,
+      "grad_norm": 0.20223967730998993,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 16348
+    },
+    {
+      "epoch": 0.691640578729165,
+      "grad_norm": 0.17678618431091309,
+      "learning_rate": 0.001,
+      "loss": 2.9338,
+      "step": 16349
+    },
+    {
+      "epoch": 0.6916828834926813,
+      "grad_norm": 0.2138862907886505,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 16350
+    },
+    {
+      "epoch": 0.6917251882561977,
+      "grad_norm": 0.15194635093212128,
+      "learning_rate": 0.001,
+      "loss": 2.0574,
+      "step": 16351
+    },
+    {
+      "epoch": 0.6917674930197141,
+      "grad_norm": 0.2285081446170807,
+      "learning_rate": 0.001,
+      "loss": 3.0883,
+      "step": 16352
+    },
+    {
+      "epoch": 0.6918097977832304,
+      "grad_norm": 0.17893058061599731,
+      "learning_rate": 0.001,
+      "loss": 1.9451,
+      "step": 16353
+    },
+    {
+      "epoch": 0.6918521025467468,
+      "grad_norm": 0.1954171061515808,
+      "learning_rate": 0.001,
+      "loss": 1.7675,
+      "step": 16354
+    },
+    {
+      "epoch": 0.6918944073102631,
+      "grad_norm": 0.1901036947965622,
+      "learning_rate": 0.001,
+      "loss": 2.7754,
+      "step": 16355
+    },
+    {
+      "epoch": 0.6919367120737795,
+      "grad_norm": 0.2298295795917511,
+      "learning_rate": 0.001,
+      "loss": 1.8922,
+      "step": 16356
+    },
+    {
+      "epoch": 0.6919790168372959,
+      "grad_norm": 0.166233092546463,
+      "learning_rate": 0.001,
+      "loss": 1.6469,
+      "step": 16357
+    },
+    {
+      "epoch": 0.6920213216008122,
+      "grad_norm": 0.17486384510993958,
+      "learning_rate": 0.001,
+      "loss": 1.7152,
+      "step": 16358
+    },
+    {
+      "epoch": 0.6920636263643286,
+      "grad_norm": 0.2248300462961197,
+      "learning_rate": 0.001,
+      "loss": 1.6036,
+      "step": 16359
+    },
+    {
+      "epoch": 0.692105931127845,
+      "grad_norm": 0.36651915311813354,
+      "learning_rate": 0.001,
+      "loss": 3.0015,
+      "step": 16360
+    },
+    {
+      "epoch": 0.6921482358913613,
+      "grad_norm": 0.2662813663482666,
+      "learning_rate": 0.001,
+      "loss": 2.7571,
+      "step": 16361
+    },
+    {
+      "epoch": 0.6921905406548777,
+      "grad_norm": 0.540983259677887,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 16362
+    },
+    {
+      "epoch": 0.6922328454183941,
+      "grad_norm": 0.44526684284210205,
+      "learning_rate": 0.001,
+      "loss": 1.8432,
+      "step": 16363
+    },
+    {
+      "epoch": 0.6922751501819104,
+      "grad_norm": 4.076467990875244,
+      "learning_rate": 0.001,
+      "loss": 4.882,
+      "step": 16364
+    },
+    {
+      "epoch": 0.6923174549454268,
+      "grad_norm": 0.16594544053077698,
+      "learning_rate": 0.001,
+      "loss": 1.753,
+      "step": 16365
+    },
+    {
+      "epoch": 0.6923597597089433,
+      "grad_norm": 0.15641087293624878,
+      "learning_rate": 0.001,
+      "loss": 2.2164,
+      "step": 16366
+    },
+    {
+      "epoch": 0.6924020644724596,
+      "grad_norm": 0.14572037756443024,
+      "learning_rate": 0.001,
+      "loss": 3.1288,
+      "step": 16367
+    },
+    {
+      "epoch": 0.692444369235976,
+      "grad_norm": 0.18323473632335663,
+      "learning_rate": 0.001,
+      "loss": 1.6634,
+      "step": 16368
+    },
+    {
+      "epoch": 0.6924866739994924,
+      "grad_norm": 0.552975594997406,
+      "learning_rate": 0.001,
+      "loss": 2.5361,
+      "step": 16369
+    },
+    {
+      "epoch": 0.6925289787630087,
+      "grad_norm": 0.743918240070343,
+      "learning_rate": 0.001,
+      "loss": 3.3047,
+      "step": 16370
+    },
+    {
+      "epoch": 0.6925712835265251,
+      "grad_norm": 0.22077415883541107,
+      "learning_rate": 0.001,
+      "loss": 1.9185,
+      "step": 16371
+    },
+    {
+      "epoch": 0.6926135882900415,
+      "grad_norm": 4.2804365158081055,
+      "learning_rate": 0.001,
+      "loss": 1.6356,
+      "step": 16372
+    },
+    {
+      "epoch": 0.6926558930535578,
+      "grad_norm": 0.23495794832706451,
+      "learning_rate": 0.001,
+      "loss": 2.2379,
+      "step": 16373
+    },
+    {
+      "epoch": 0.6926981978170742,
+      "grad_norm": 0.14762520790100098,
+      "learning_rate": 0.001,
+      "loss": 1.7328,
+      "step": 16374
+    },
+    {
+      "epoch": 0.6927405025805906,
+      "grad_norm": 1.2095277309417725,
+      "learning_rate": 0.001,
+      "loss": 3.5081,
+      "step": 16375
+    },
+    {
+      "epoch": 0.6927828073441069,
+      "grad_norm": 0.16055262088775635,
+      "learning_rate": 0.001,
+      "loss": 2.0528,
+      "step": 16376
+    },
+    {
+      "epoch": 0.6928251121076233,
+      "grad_norm": 0.24955902993679047,
+      "learning_rate": 0.001,
+      "loss": 2.0347,
+      "step": 16377
+    },
+    {
+      "epoch": 0.6928674168711397,
+      "grad_norm": 0.8334779143333435,
+      "learning_rate": 0.001,
+      "loss": 2.8753,
+      "step": 16378
+    },
+    {
+      "epoch": 0.692909721634656,
+      "grad_norm": 0.18906332552433014,
+      "learning_rate": 0.001,
+      "loss": 1.6332,
+      "step": 16379
+    },
+    {
+      "epoch": 0.6929520263981724,
+      "grad_norm": 0.1773161143064499,
+      "learning_rate": 0.001,
+      "loss": 2.7939,
+      "step": 16380
+    },
+    {
+      "epoch": 0.6929943311616888,
+      "grad_norm": 0.18923841416835785,
+      "learning_rate": 0.001,
+      "loss": 2.376,
+      "step": 16381
+    },
+    {
+      "epoch": 0.6930366359252051,
+      "grad_norm": 0.1875029057264328,
+      "learning_rate": 0.001,
+      "loss": 2.5787,
+      "step": 16382
+    },
+    {
+      "epoch": 0.6930789406887216,
+      "grad_norm": 0.1562403440475464,
+      "learning_rate": 0.001,
+      "loss": 1.9452,
+      "step": 16383
+    },
+    {
+      "epoch": 0.693121245452238,
+      "grad_norm": 0.19524121284484863,
+      "learning_rate": 0.001,
+      "loss": 3.0866,
+      "step": 16384
+    },
+    {
+      "epoch": 0.6931635502157543,
+      "grad_norm": 0.1857270449399948,
+      "learning_rate": 0.001,
+      "loss": 1.9935,
+      "step": 16385
+    },
+    {
+      "epoch": 0.6932058549792707,
+      "grad_norm": 0.3167372941970825,
+      "learning_rate": 0.001,
+      "loss": 2.4217,
+      "step": 16386
+    },
+    {
+      "epoch": 0.6932481597427871,
+      "grad_norm": 0.24016937613487244,
+      "learning_rate": 0.001,
+      "loss": 2.8926,
+      "step": 16387
+    },
+    {
+      "epoch": 0.6932904645063034,
+      "grad_norm": 0.15906888246536255,
+      "learning_rate": 0.001,
+      "loss": 3.2683,
+      "step": 16388
+    },
+    {
+      "epoch": 0.6933327692698198,
+      "grad_norm": 0.2372499704360962,
+      "learning_rate": 0.001,
+      "loss": 2.9656,
+      "step": 16389
+    },
+    {
+      "epoch": 0.6933750740333362,
+      "grad_norm": 18.70325469970703,
+      "learning_rate": 0.001,
+      "loss": 1.8463,
+      "step": 16390
+    },
+    {
+      "epoch": 0.6934173787968525,
+      "grad_norm": 0.21969793736934662,
+      "learning_rate": 0.001,
+      "loss": 1.9638,
+      "step": 16391
+    },
+    {
+      "epoch": 0.6934596835603689,
+      "grad_norm": 1.0692784786224365,
+      "learning_rate": 0.001,
+      "loss": 2.1758,
+      "step": 16392
+    },
+    {
+      "epoch": 0.6935019883238853,
+      "grad_norm": 0.17494791746139526,
+      "learning_rate": 0.001,
+      "loss": 1.9434,
+      "step": 16393
+    },
+    {
+      "epoch": 0.6935442930874016,
+      "grad_norm": 10.130684852600098,
+      "learning_rate": 0.001,
+      "loss": 2.0469,
+      "step": 16394
+    },
+    {
+      "epoch": 0.693586597850918,
+      "grad_norm": 0.18543873727321625,
+      "learning_rate": 0.001,
+      "loss": 1.6253,
+      "step": 16395
+    },
+    {
+      "epoch": 0.6936289026144344,
+      "grad_norm": 0.176463320851326,
+      "learning_rate": 0.001,
+      "loss": 2.5161,
+      "step": 16396
+    },
+    {
+      "epoch": 0.6936712073779507,
+      "grad_norm": 0.21182700991630554,
+      "learning_rate": 0.001,
+      "loss": 1.9964,
+      "step": 16397
+    },
+    {
+      "epoch": 0.6937135121414671,
+      "grad_norm": 0.17434942722320557,
+      "learning_rate": 0.001,
+      "loss": 1.608,
+      "step": 16398
+    },
+    {
+      "epoch": 0.6937558169049834,
+      "grad_norm": 1.2661168575286865,
+      "learning_rate": 0.001,
+      "loss": 3.2148,
+      "step": 16399
+    },
+    {
+      "epoch": 0.6937981216684999,
+      "grad_norm": 16.555898666381836,
+      "learning_rate": 0.001,
+      "loss": 2.0895,
+      "step": 16400
+    },
+    {
+      "epoch": 0.6938404264320163,
+      "grad_norm": 0.1642748862504959,
+      "learning_rate": 0.001,
+      "loss": 1.9288,
+      "step": 16401
+    },
+    {
+      "epoch": 0.6938827311955326,
+      "grad_norm": 0.20912915468215942,
+      "learning_rate": 0.001,
+      "loss": 3.3075,
+      "step": 16402
+    },
+    {
+      "epoch": 0.693925035959049,
+      "grad_norm": 0.1803750842809677,
+      "learning_rate": 0.001,
+      "loss": 3.1573,
+      "step": 16403
+    },
+    {
+      "epoch": 0.6939673407225654,
+      "grad_norm": 0.21394596993923187,
+      "learning_rate": 0.001,
+      "loss": 2.6982,
+      "step": 16404
+    },
+    {
+      "epoch": 0.6940096454860817,
+      "grad_norm": 0.22637422382831573,
+      "learning_rate": 0.001,
+      "loss": 2.2632,
+      "step": 16405
+    },
+    {
+      "epoch": 0.6940519502495981,
+      "grad_norm": 0.3572201728820801,
+      "learning_rate": 0.001,
+      "loss": 2.7936,
+      "step": 16406
+    },
+    {
+      "epoch": 0.6940942550131145,
+      "grad_norm": 17.344480514526367,
+      "learning_rate": 0.001,
+      "loss": 1.5021,
+      "step": 16407
+    },
+    {
+      "epoch": 0.6941365597766308,
+      "grad_norm": 0.19448307156562805,
+      "learning_rate": 0.001,
+      "loss": 1.9725,
+      "step": 16408
+    },
+    {
+      "epoch": 0.6941788645401472,
+      "grad_norm": 0.18592120707035065,
+      "learning_rate": 0.001,
+      "loss": 2.0079,
+      "step": 16409
+    },
+    {
+      "epoch": 0.6942211693036636,
+      "grad_norm": 0.17572908103466034,
+      "learning_rate": 0.001,
+      "loss": 1.9132,
+      "step": 16410
+    },
+    {
+      "epoch": 0.6942634740671799,
+      "grad_norm": 0.31208327412605286,
+      "learning_rate": 0.001,
+      "loss": 1.6902,
+      "step": 16411
+    },
+    {
+      "epoch": 0.6943057788306963,
+      "grad_norm": 0.2312842458486557,
+      "learning_rate": 0.001,
+      "loss": 2.3156,
+      "step": 16412
+    },
+    {
+      "epoch": 0.6943480835942127,
+      "grad_norm": 0.25776514410972595,
+      "learning_rate": 0.001,
+      "loss": 1.3374,
+      "step": 16413
+    },
+    {
+      "epoch": 0.694390388357729,
+      "grad_norm": 0.7767313718795776,
+      "learning_rate": 0.001,
+      "loss": 2.6159,
+      "step": 16414
+    },
+    {
+      "epoch": 0.6944326931212454,
+      "grad_norm": 1.3222384452819824,
+      "learning_rate": 0.001,
+      "loss": 3.7811,
+      "step": 16415
+    },
+    {
+      "epoch": 0.6944749978847619,
+      "grad_norm": 0.22960853576660156,
+      "learning_rate": 0.001,
+      "loss": 1.8167,
+      "step": 16416
+    },
+    {
+      "epoch": 0.6945173026482782,
+      "grad_norm": 0.20826157927513123,
+      "learning_rate": 0.001,
+      "loss": 2.0062,
+      "step": 16417
+    },
+    {
+      "epoch": 0.6945596074117946,
+      "grad_norm": 2.312390089035034,
+      "learning_rate": 0.001,
+      "loss": 3.3735,
+      "step": 16418
+    },
+    {
+      "epoch": 0.694601912175311,
+      "grad_norm": 3.1580910682678223,
+      "learning_rate": 0.001,
+      "loss": 2.463,
+      "step": 16419
+    },
+    {
+      "epoch": 0.6946442169388273,
+      "grad_norm": 0.25120091438293457,
+      "learning_rate": 0.001,
+      "loss": 2.9196,
+      "step": 16420
+    },
+    {
+      "epoch": 0.6946865217023437,
+      "grad_norm": 0.15511690080165863,
+      "learning_rate": 0.001,
+      "loss": 1.6251,
+      "step": 16421
+    },
+    {
+      "epoch": 0.6947288264658601,
+      "grad_norm": 0.18598857522010803,
+      "learning_rate": 0.001,
+      "loss": 2.9597,
+      "step": 16422
+    },
+    {
+      "epoch": 0.6947711312293764,
+      "grad_norm": 0.16862650215625763,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 16423
+    },
+    {
+      "epoch": 0.6948134359928928,
+      "grad_norm": 0.18114817142486572,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 16424
+    },
+    {
+      "epoch": 0.6948557407564092,
+      "grad_norm": 0.19267553091049194,
+      "learning_rate": 0.001,
+      "loss": 2.3873,
+      "step": 16425
+    },
+    {
+      "epoch": 0.6948980455199255,
+      "grad_norm": 0.20283320546150208,
+      "learning_rate": 0.001,
+      "loss": 1.3301,
+      "step": 16426
+    },
+    {
+      "epoch": 0.6949403502834419,
+      "grad_norm": 2.084455728530884,
+      "learning_rate": 0.001,
+      "loss": 1.6759,
+      "step": 16427
+    },
+    {
+      "epoch": 0.6949826550469583,
+      "grad_norm": 2.8412082195281982,
+      "learning_rate": 0.001,
+      "loss": 3.4139,
+      "step": 16428
+    },
+    {
+      "epoch": 0.6950249598104746,
+      "grad_norm": 2.059316396713257,
+      "learning_rate": 0.001,
+      "loss": 1.9375,
+      "step": 16429
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 0.3789178431034088,
+      "learning_rate": 0.001,
+      "loss": 2.9269,
+      "step": 16430
+    },
+    {
+      "epoch": 0.6951095693375074,
+      "grad_norm": 0.31577736139297485,
+      "learning_rate": 0.001,
+      "loss": 2.9995,
+      "step": 16431
+    },
+    {
+      "epoch": 0.6951518741010237,
+      "grad_norm": 2.173356533050537,
+      "learning_rate": 0.001,
+      "loss": 2.4088,
+      "step": 16432
+    },
+    {
+      "epoch": 0.6951941788645402,
+      "grad_norm": 0.2261185199022293,
+      "learning_rate": 0.001,
+      "loss": 2.0083,
+      "step": 16433
+    },
+    {
+      "epoch": 0.6952364836280566,
+      "grad_norm": 0.23543325066566467,
+      "learning_rate": 0.001,
+      "loss": 1.9962,
+      "step": 16434
+    },
+    {
+      "epoch": 0.6952787883915729,
+      "grad_norm": 0.23661349713802338,
+      "learning_rate": 0.001,
+      "loss": 2.1091,
+      "step": 16435
+    },
+    {
+      "epoch": 0.6953210931550893,
+      "grad_norm": 0.20180752873420715,
+      "learning_rate": 0.001,
+      "loss": 2.3918,
+      "step": 16436
+    },
+    {
+      "epoch": 0.6953633979186057,
+      "grad_norm": 0.261025995016098,
+      "learning_rate": 0.001,
+      "loss": 2.3989,
+      "step": 16437
+    },
+    {
+      "epoch": 0.695405702682122,
+      "grad_norm": 0.24305380880832672,
+      "learning_rate": 0.001,
+      "loss": 2.3177,
+      "step": 16438
+    },
+    {
+      "epoch": 0.6954480074456384,
+      "grad_norm": 0.1875523179769516,
+      "learning_rate": 0.001,
+      "loss": 2.4939,
+      "step": 16439
+    },
+    {
+      "epoch": 0.6954903122091548,
+      "grad_norm": 0.20536570250988007,
+      "learning_rate": 0.001,
+      "loss": 2.3507,
+      "step": 16440
+    },
+    {
+      "epoch": 0.6955326169726711,
+      "grad_norm": 2.672262668609619,
+      "learning_rate": 0.001,
+      "loss": 2.8801,
+      "step": 16441
+    },
+    {
+      "epoch": 0.6955749217361875,
+      "grad_norm": 0.16370277106761932,
+      "learning_rate": 0.001,
+      "loss": 2.0448,
+      "step": 16442
+    },
+    {
+      "epoch": 0.6956172264997038,
+      "grad_norm": 0.3966878056526184,
+      "learning_rate": 0.001,
+      "loss": 1.4291,
+      "step": 16443
+    },
+    {
+      "epoch": 0.6956595312632202,
+      "grad_norm": 6.427881717681885,
+      "learning_rate": 0.001,
+      "loss": 2.0735,
+      "step": 16444
+    },
+    {
+      "epoch": 0.6957018360267366,
+      "grad_norm": 0.9864132404327393,
+      "learning_rate": 0.001,
+      "loss": 1.9811,
+      "step": 16445
+    },
+    {
+      "epoch": 0.6957441407902529,
+      "grad_norm": 0.19832006096839905,
+      "learning_rate": 0.001,
+      "loss": 2.0114,
+      "step": 16446
+    },
+    {
+      "epoch": 0.6957864455537693,
+      "grad_norm": 0.18039336800575256,
+      "learning_rate": 0.001,
+      "loss": 1.4557,
+      "step": 16447
+    },
+    {
+      "epoch": 0.6958287503172857,
+      "grad_norm": 0.18589988350868225,
+      "learning_rate": 0.001,
+      "loss": 1.7354,
+      "step": 16448
+    },
+    {
+      "epoch": 0.695871055080802,
+      "grad_norm": 0.21510185301303864,
+      "learning_rate": 0.001,
+      "loss": 2.3479,
+      "step": 16449
+    },
+    {
+      "epoch": 0.6959133598443185,
+      "grad_norm": 0.19630225002765656,
+      "learning_rate": 0.001,
+      "loss": 2.2254,
+      "step": 16450
+    },
+    {
+      "epoch": 0.6959556646078349,
+      "grad_norm": 0.18261785805225372,
+      "learning_rate": 0.001,
+      "loss": 2.2902,
+      "step": 16451
+    },
+    {
+      "epoch": 0.6959979693713512,
+      "grad_norm": 0.20006346702575684,
+      "learning_rate": 0.001,
+      "loss": 2.0152,
+      "step": 16452
+    },
+    {
+      "epoch": 0.6960402741348676,
+      "grad_norm": 0.18040160834789276,
+      "learning_rate": 0.001,
+      "loss": 1.8994,
+      "step": 16453
+    },
+    {
+      "epoch": 0.696082578898384,
+      "grad_norm": 1.6500322818756104,
+      "learning_rate": 0.001,
+      "loss": 1.7275,
+      "step": 16454
+    },
+    {
+      "epoch": 0.6961248836619003,
+      "grad_norm": 0.2118813842535019,
+      "learning_rate": 0.001,
+      "loss": 3.1393,
+      "step": 16455
+    },
+    {
+      "epoch": 0.6961671884254167,
+      "grad_norm": 0.18187031149864197,
+      "learning_rate": 0.001,
+      "loss": 2.1885,
+      "step": 16456
+    },
+    {
+      "epoch": 0.6962094931889331,
+      "grad_norm": 0.21792373061180115,
+      "learning_rate": 0.001,
+      "loss": 2.5011,
+      "step": 16457
+    },
+    {
+      "epoch": 0.6962517979524494,
+      "grad_norm": 0.2024918496608734,
+      "learning_rate": 0.001,
+      "loss": 2.1317,
+      "step": 16458
+    },
+    {
+      "epoch": 0.6962941027159658,
+      "grad_norm": 0.1720605045557022,
+      "learning_rate": 0.001,
+      "loss": 2.5661,
+      "step": 16459
+    },
+    {
+      "epoch": 0.6963364074794822,
+      "grad_norm": 0.20015959441661835,
+      "learning_rate": 0.001,
+      "loss": 2.6472,
+      "step": 16460
+    },
+    {
+      "epoch": 0.6963787122429985,
+      "grad_norm": 0.3034919500350952,
+      "learning_rate": 0.001,
+      "loss": 1.9821,
+      "step": 16461
+    },
+    {
+      "epoch": 0.6964210170065149,
+      "grad_norm": 0.1722230315208435,
+      "learning_rate": 0.001,
+      "loss": 1.7016,
+      "step": 16462
+    },
+    {
+      "epoch": 0.6964633217700313,
+      "grad_norm": 0.17962494492530823,
+      "learning_rate": 0.001,
+      "loss": 1.6498,
+      "step": 16463
+    },
+    {
+      "epoch": 0.6965056265335476,
+      "grad_norm": 0.6122218370437622,
+      "learning_rate": 0.001,
+      "loss": 1.7189,
+      "step": 16464
+    },
+    {
+      "epoch": 0.696547931297064,
+      "grad_norm": 0.17944207787513733,
+      "learning_rate": 0.001,
+      "loss": 1.5481,
+      "step": 16465
+    },
+    {
+      "epoch": 0.6965902360605805,
+      "grad_norm": 0.16219043731689453,
+      "learning_rate": 0.001,
+      "loss": 2.1806,
+      "step": 16466
+    },
+    {
+      "epoch": 0.6966325408240968,
+      "grad_norm": 0.2081691324710846,
+      "learning_rate": 0.001,
+      "loss": 3.2181,
+      "step": 16467
+    },
+    {
+      "epoch": 0.6966748455876132,
+      "grad_norm": 0.16732138395309448,
+      "learning_rate": 0.001,
+      "loss": 3.0854,
+      "step": 16468
+    },
+    {
+      "epoch": 0.6967171503511296,
+      "grad_norm": 0.1735231727361679,
+      "learning_rate": 0.001,
+      "loss": 2.2234,
+      "step": 16469
+    },
+    {
+      "epoch": 0.6967594551146459,
+      "grad_norm": 0.7906176447868347,
+      "learning_rate": 0.001,
+      "loss": 2.4975,
+      "step": 16470
+    },
+    {
+      "epoch": 0.6968017598781623,
+      "grad_norm": 0.20566585659980774,
+      "learning_rate": 0.001,
+      "loss": 3.4062,
+      "step": 16471
+    },
+    {
+      "epoch": 0.6968440646416787,
+      "grad_norm": 0.1833547055721283,
+      "learning_rate": 0.001,
+      "loss": 2.034,
+      "step": 16472
+    },
+    {
+      "epoch": 0.696886369405195,
+      "grad_norm": 0.18477390706539154,
+      "learning_rate": 0.001,
+      "loss": 2.5579,
+      "step": 16473
+    },
+    {
+      "epoch": 0.6969286741687114,
+      "grad_norm": 0.19416743516921997,
+      "learning_rate": 0.001,
+      "loss": 2.2212,
+      "step": 16474
+    },
+    {
+      "epoch": 0.6969709789322278,
+      "grad_norm": 0.2078043520450592,
+      "learning_rate": 0.001,
+      "loss": 2.0475,
+      "step": 16475
+    },
+    {
+      "epoch": 0.6970132836957441,
+      "grad_norm": 0.1452716588973999,
+      "learning_rate": 0.001,
+      "loss": 1.6866,
+      "step": 16476
+    },
+    {
+      "epoch": 0.6970555884592605,
+      "grad_norm": 0.1788160353899002,
+      "learning_rate": 0.001,
+      "loss": 2.1014,
+      "step": 16477
+    },
+    {
+      "epoch": 0.6970978932227769,
+      "grad_norm": 0.1941988617181778,
+      "learning_rate": 0.001,
+      "loss": 2.2228,
+      "step": 16478
+    },
+    {
+      "epoch": 0.6971401979862932,
+      "grad_norm": 0.5075099468231201,
+      "learning_rate": 0.001,
+      "loss": 2.046,
+      "step": 16479
+    },
+    {
+      "epoch": 0.6971825027498096,
+      "grad_norm": 0.16728682816028595,
+      "learning_rate": 0.001,
+      "loss": 2.5115,
+      "step": 16480
+    },
+    {
+      "epoch": 0.697224807513326,
+      "grad_norm": 0.17837141454219818,
+      "learning_rate": 0.001,
+      "loss": 1.8429,
+      "step": 16481
+    },
+    {
+      "epoch": 0.6972671122768423,
+      "grad_norm": 0.14722979068756104,
+      "learning_rate": 0.001,
+      "loss": 1.6878,
+      "step": 16482
+    },
+    {
+      "epoch": 0.6973094170403588,
+      "grad_norm": 0.17707088589668274,
+      "learning_rate": 0.001,
+      "loss": 2.0991,
+      "step": 16483
+    },
+    {
+      "epoch": 0.6973517218038752,
+      "grad_norm": 0.20449602603912354,
+      "learning_rate": 0.001,
+      "loss": 1.4112,
+      "step": 16484
+    },
+    {
+      "epoch": 0.6973940265673915,
+      "grad_norm": 0.18787558376789093,
+      "learning_rate": 0.001,
+      "loss": 1.9535,
+      "step": 16485
+    },
+    {
+      "epoch": 0.6974363313309079,
+      "grad_norm": 3.8804171085357666,
+      "learning_rate": 0.001,
+      "loss": 2.9737,
+      "step": 16486
+    },
+    {
+      "epoch": 0.6974786360944243,
+      "grad_norm": 0.18371431529521942,
+      "learning_rate": 0.001,
+      "loss": 2.5309,
+      "step": 16487
+    },
+    {
+      "epoch": 0.6975209408579406,
+      "grad_norm": 0.154661625623703,
+      "learning_rate": 0.001,
+      "loss": 1.7516,
+      "step": 16488
+    },
+    {
+      "epoch": 0.697563245621457,
+      "grad_norm": 0.15226295590400696,
+      "learning_rate": 0.001,
+      "loss": 2.5867,
+      "step": 16489
+    },
+    {
+      "epoch": 0.6976055503849733,
+      "grad_norm": 9.875614166259766,
+      "learning_rate": 0.001,
+      "loss": 2.3198,
+      "step": 16490
+    },
+    {
+      "epoch": 0.6976478551484897,
+      "grad_norm": 0.18575811386108398,
+      "learning_rate": 0.001,
+      "loss": 2.5731,
+      "step": 16491
+    },
+    {
+      "epoch": 0.6976901599120061,
+      "grad_norm": 0.18501834571361542,
+      "learning_rate": 0.001,
+      "loss": 1.9908,
+      "step": 16492
+    },
+    {
+      "epoch": 0.6977324646755224,
+      "grad_norm": 0.17492198944091797,
+      "learning_rate": 0.001,
+      "loss": 2.1035,
+      "step": 16493
+    },
+    {
+      "epoch": 0.6977747694390388,
+      "grad_norm": 0.1972741186618805,
+      "learning_rate": 0.001,
+      "loss": 2.0895,
+      "step": 16494
+    },
+    {
+      "epoch": 0.6978170742025552,
+      "grad_norm": 0.1590324491262436,
+      "learning_rate": 0.001,
+      "loss": 1.8644,
+      "step": 16495
+    },
+    {
+      "epoch": 0.6978593789660715,
+      "grad_norm": 0.15750128030776978,
+      "learning_rate": 0.001,
+      "loss": 2.3938,
+      "step": 16496
+    },
+    {
+      "epoch": 0.6979016837295879,
+      "grad_norm": 0.18353481590747833,
+      "learning_rate": 0.001,
+      "loss": 1.9701,
+      "step": 16497
+    },
+    {
+      "epoch": 0.6979439884931044,
+      "grad_norm": 0.4982626438140869,
+      "learning_rate": 0.001,
+      "loss": 2.8423,
+      "step": 16498
+    },
+    {
+      "epoch": 0.6979862932566206,
+      "grad_norm": 0.5572125911712646,
+      "learning_rate": 0.001,
+      "loss": 1.6008,
+      "step": 16499
+    },
+    {
+      "epoch": 0.6980285980201371,
+      "grad_norm": 0.18385641276836395,
+      "learning_rate": 0.001,
+      "loss": 1.7157,
+      "step": 16500
+    },
+    {
+      "epoch": 0.6980709027836535,
+      "grad_norm": 0.20316894352436066,
+      "learning_rate": 0.001,
+      "loss": 3.3949,
+      "step": 16501
+    },
+    {
+      "epoch": 0.6981132075471698,
+      "grad_norm": 0.21854710578918457,
+      "learning_rate": 0.001,
+      "loss": 2.4985,
+      "step": 16502
+    },
+    {
+      "epoch": 0.6981555123106862,
+      "grad_norm": 0.2654995620250702,
+      "learning_rate": 0.001,
+      "loss": 2.077,
+      "step": 16503
+    },
+    {
+      "epoch": 0.6981978170742026,
+      "grad_norm": 0.15483231842517853,
+      "learning_rate": 0.001,
+      "loss": 2.5364,
+      "step": 16504
+    },
+    {
+      "epoch": 0.6982401218377189,
+      "grad_norm": 0.1526549905538559,
+      "learning_rate": 0.001,
+      "loss": 2.8325,
+      "step": 16505
+    },
+    {
+      "epoch": 0.6982824266012353,
+      "grad_norm": 0.18169742822647095,
+      "learning_rate": 0.001,
+      "loss": 2.6684,
+      "step": 16506
+    },
+    {
+      "epoch": 0.6983247313647517,
+      "grad_norm": 0.19134920835494995,
+      "learning_rate": 0.001,
+      "loss": 2.8449,
+      "step": 16507
+    },
+    {
+      "epoch": 0.698367036128268,
+      "grad_norm": 3.5741662979125977,
+      "learning_rate": 0.001,
+      "loss": 2.137,
+      "step": 16508
+    },
+    {
+      "epoch": 0.6984093408917844,
+      "grad_norm": 0.42037051916122437,
+      "learning_rate": 0.001,
+      "loss": 1.8969,
+      "step": 16509
+    },
+    {
+      "epoch": 0.6984516456553008,
+      "grad_norm": 0.1691959798336029,
+      "learning_rate": 0.001,
+      "loss": 1.5428,
+      "step": 16510
+    },
+    {
+      "epoch": 0.6984939504188171,
+      "grad_norm": 0.19847933948040009,
+      "learning_rate": 0.001,
+      "loss": 1.9573,
+      "step": 16511
+    },
+    {
+      "epoch": 0.6985362551823335,
+      "grad_norm": 0.8859964609146118,
+      "learning_rate": 0.001,
+      "loss": 2.9135,
+      "step": 16512
+    },
+    {
+      "epoch": 0.6985785599458499,
+      "grad_norm": 0.17541298270225525,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 16513
+    },
+    {
+      "epoch": 0.6986208647093662,
+      "grad_norm": 0.1629028022289276,
+      "learning_rate": 0.001,
+      "loss": 2.2677,
+      "step": 16514
+    },
+    {
+      "epoch": 0.6986631694728827,
+      "grad_norm": 0.39890217781066895,
+      "learning_rate": 0.001,
+      "loss": 1.998,
+      "step": 16515
+    },
+    {
+      "epoch": 0.6987054742363991,
+      "grad_norm": 0.16391772031784058,
+      "learning_rate": 0.001,
+      "loss": 2.2344,
+      "step": 16516
+    },
+    {
+      "epoch": 0.6987477789999154,
+      "grad_norm": 0.20144188404083252,
+      "learning_rate": 0.001,
+      "loss": 2.3193,
+      "step": 16517
+    },
+    {
+      "epoch": 0.6987900837634318,
+      "grad_norm": 0.3551698923110962,
+      "learning_rate": 0.001,
+      "loss": 1.9678,
+      "step": 16518
+    },
+    {
+      "epoch": 0.6988323885269482,
+      "grad_norm": 3.031266450881958,
+      "learning_rate": 0.001,
+      "loss": 2.1602,
+      "step": 16519
+    },
+    {
+      "epoch": 0.6988746932904645,
+      "grad_norm": 0.21469104290008545,
+      "learning_rate": 0.001,
+      "loss": 2.2958,
+      "step": 16520
+    },
+    {
+      "epoch": 0.6989169980539809,
+      "grad_norm": 0.9255651831626892,
+      "learning_rate": 0.001,
+      "loss": 1.9495,
+      "step": 16521
+    },
+    {
+      "epoch": 0.6989593028174973,
+      "grad_norm": 0.1491638422012329,
+      "learning_rate": 0.001,
+      "loss": 1.8283,
+      "step": 16522
+    },
+    {
+      "epoch": 0.6990016075810136,
+      "grad_norm": 0.22015687823295593,
+      "learning_rate": 0.001,
+      "loss": 2.5978,
+      "step": 16523
+    },
+    {
+      "epoch": 0.69904391234453,
+      "grad_norm": 0.17289428412914276,
+      "learning_rate": 0.001,
+      "loss": 1.9345,
+      "step": 16524
+    },
+    {
+      "epoch": 0.6990862171080464,
+      "grad_norm": 0.1912022829055786,
+      "learning_rate": 0.001,
+      "loss": 2.4924,
+      "step": 16525
+    },
+    {
+      "epoch": 0.6991285218715627,
+      "grad_norm": 0.17688611149787903,
+      "learning_rate": 0.001,
+      "loss": 1.7264,
+      "step": 16526
+    },
+    {
+      "epoch": 0.6991708266350791,
+      "grad_norm": 0.1654282510280609,
+      "learning_rate": 0.001,
+      "loss": 2.0543,
+      "step": 16527
+    },
+    {
+      "epoch": 0.6992131313985955,
+      "grad_norm": 3.0969583988189697,
+      "learning_rate": 0.001,
+      "loss": 2.2927,
+      "step": 16528
+    },
+    {
+      "epoch": 0.6992554361621118,
+      "grad_norm": 0.2961462438106537,
+      "learning_rate": 0.001,
+      "loss": 1.3273,
+      "step": 16529
+    },
+    {
+      "epoch": 0.6992977409256282,
+      "grad_norm": 0.1830281913280487,
+      "learning_rate": 0.001,
+      "loss": 2.2964,
+      "step": 16530
+    },
+    {
+      "epoch": 0.6993400456891447,
+      "grad_norm": 0.19656279683113098,
+      "learning_rate": 0.001,
+      "loss": 2.521,
+      "step": 16531
+    },
+    {
+      "epoch": 0.699382350452661,
+      "grad_norm": 0.3053706884384155,
+      "learning_rate": 0.001,
+      "loss": 2.1476,
+      "step": 16532
+    },
+    {
+      "epoch": 0.6994246552161774,
+      "grad_norm": 0.3848680257797241,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 16533
+    },
+    {
+      "epoch": 0.6994669599796937,
+      "grad_norm": 0.20205725729465485,
+      "learning_rate": 0.001,
+      "loss": 2.2957,
+      "step": 16534
+    },
+    {
+      "epoch": 0.6995092647432101,
+      "grad_norm": 0.21945300698280334,
+      "learning_rate": 0.001,
+      "loss": 1.8217,
+      "step": 16535
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 0.3286508321762085,
+      "learning_rate": 0.001,
+      "loss": 2.1553,
+      "step": 16536
+    },
+    {
+      "epoch": 0.6995938742702428,
+      "grad_norm": 0.17390255630016327,
+      "learning_rate": 0.001,
+      "loss": 1.8902,
+      "step": 16537
+    },
+    {
+      "epoch": 0.6996361790337592,
+      "grad_norm": 0.2187155783176422,
+      "learning_rate": 0.001,
+      "loss": 2.5251,
+      "step": 16538
+    },
+    {
+      "epoch": 0.6996784837972756,
+      "grad_norm": 0.1886243224143982,
+      "learning_rate": 0.001,
+      "loss": 1.8132,
+      "step": 16539
+    },
+    {
+      "epoch": 0.6997207885607919,
+      "grad_norm": 0.14639221131801605,
+      "learning_rate": 0.001,
+      "loss": 2.2848,
+      "step": 16540
+    },
+    {
+      "epoch": 0.6997630933243083,
+      "grad_norm": 0.2016509771347046,
+      "learning_rate": 0.001,
+      "loss": 2.3703,
+      "step": 16541
+    },
+    {
+      "epoch": 0.6998053980878247,
+      "grad_norm": 0.2064632922410965,
+      "learning_rate": 0.001,
+      "loss": 1.8111,
+      "step": 16542
+    },
+    {
+      "epoch": 0.699847702851341,
+      "grad_norm": 0.304959237575531,
+      "learning_rate": 0.001,
+      "loss": 1.8825,
+      "step": 16543
+    },
+    {
+      "epoch": 0.6998900076148574,
+      "grad_norm": 0.23426057398319244,
+      "learning_rate": 0.001,
+      "loss": 1.7511,
+      "step": 16544
+    },
+    {
+      "epoch": 0.6999323123783738,
+      "grad_norm": 0.15036675333976746,
+      "learning_rate": 0.001,
+      "loss": 1.4636,
+      "step": 16545
+    },
+    {
+      "epoch": 0.6999746171418901,
+      "grad_norm": 0.18853148818016052,
+      "learning_rate": 0.001,
+      "loss": 3.3999,
+      "step": 16546
+    },
+    {
+      "epoch": 0.7000169219054065,
+      "grad_norm": 0.20281235873699188,
+      "learning_rate": 0.001,
+      "loss": 3.1192,
+      "step": 16547
+    },
+    {
+      "epoch": 0.700059226668923,
+      "grad_norm": 0.1816108375787735,
+      "learning_rate": 0.001,
+      "loss": 2.7407,
+      "step": 16548
+    },
+    {
+      "epoch": 0.7001015314324393,
+      "grad_norm": 0.28820428252220154,
+      "learning_rate": 0.001,
+      "loss": 2.1106,
+      "step": 16549
+    },
+    {
+      "epoch": 0.7001438361959557,
+      "grad_norm": 0.1509302407503128,
+      "learning_rate": 0.001,
+      "loss": 1.7362,
+      "step": 16550
+    },
+    {
+      "epoch": 0.7001861409594721,
+      "grad_norm": 0.14989890158176422,
+      "learning_rate": 0.001,
+      "loss": 1.9436,
+      "step": 16551
+    },
+    {
+      "epoch": 0.7002284457229884,
+      "grad_norm": 0.17388369143009186,
+      "learning_rate": 0.001,
+      "loss": 1.4622,
+      "step": 16552
+    },
+    {
+      "epoch": 0.7002707504865048,
+      "grad_norm": 0.28890854120254517,
+      "learning_rate": 0.001,
+      "loss": 2.0939,
+      "step": 16553
+    },
+    {
+      "epoch": 0.7003130552500212,
+      "grad_norm": 0.1795245260000229,
+      "learning_rate": 0.001,
+      "loss": 1.9428,
+      "step": 16554
+    },
+    {
+      "epoch": 0.7003553600135375,
+      "grad_norm": 0.178712397813797,
+      "learning_rate": 0.001,
+      "loss": 1.6748,
+      "step": 16555
+    },
+    {
+      "epoch": 0.7003976647770539,
+      "grad_norm": 0.16266508400440216,
+      "learning_rate": 0.001,
+      "loss": 1.7466,
+      "step": 16556
+    },
+    {
+      "epoch": 0.7004399695405703,
+      "grad_norm": 0.18018513917922974,
+      "learning_rate": 0.001,
+      "loss": 1.5014,
+      "step": 16557
+    },
+    {
+      "epoch": 0.7004822743040866,
+      "grad_norm": 0.20177042484283447,
+      "learning_rate": 0.001,
+      "loss": 1.8159,
+      "step": 16558
+    },
+    {
+      "epoch": 0.700524579067603,
+      "grad_norm": 0.4257431626319885,
+      "learning_rate": 0.001,
+      "loss": 2.8354,
+      "step": 16559
+    },
+    {
+      "epoch": 0.7005668838311194,
+      "grad_norm": 0.3945598006248474,
+      "learning_rate": 0.001,
+      "loss": 2.5141,
+      "step": 16560
+    },
+    {
+      "epoch": 0.7006091885946357,
+      "grad_norm": 0.1487802118062973,
+      "learning_rate": 0.001,
+      "loss": 1.9653,
+      "step": 16561
+    },
+    {
+      "epoch": 0.7006514933581521,
+      "grad_norm": 0.1524672657251358,
+      "learning_rate": 0.001,
+      "loss": 2.3534,
+      "step": 16562
+    },
+    {
+      "epoch": 0.7006937981216685,
+      "grad_norm": 0.1571507751941681,
+      "learning_rate": 0.001,
+      "loss": 3.0126,
+      "step": 16563
+    },
+    {
+      "epoch": 0.7007361028851848,
+      "grad_norm": 1.0442800521850586,
+      "learning_rate": 0.001,
+      "loss": 2.0008,
+      "step": 16564
+    },
+    {
+      "epoch": 0.7007784076487013,
+      "grad_norm": 0.13998109102249146,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 16565
+    },
+    {
+      "epoch": 0.7008207124122177,
+      "grad_norm": 0.1480410099029541,
+      "learning_rate": 0.001,
+      "loss": 1.6822,
+      "step": 16566
+    },
+    {
+      "epoch": 0.700863017175734,
+      "grad_norm": 0.2611854374408722,
+      "learning_rate": 0.001,
+      "loss": 1.9182,
+      "step": 16567
+    },
+    {
+      "epoch": 0.7009053219392504,
+      "grad_norm": 1.3647289276123047,
+      "learning_rate": 0.001,
+      "loss": 1.8359,
+      "step": 16568
+    },
+    {
+      "epoch": 0.7009476267027668,
+      "grad_norm": 0.15801845490932465,
+      "learning_rate": 0.001,
+      "loss": 1.8454,
+      "step": 16569
+    },
+    {
+      "epoch": 0.7009899314662831,
+      "grad_norm": 0.18349795043468475,
+      "learning_rate": 0.001,
+      "loss": 1.8095,
+      "step": 16570
+    },
+    {
+      "epoch": 0.7010322362297995,
+      "grad_norm": 0.16039691865444183,
+      "learning_rate": 0.001,
+      "loss": 2.1951,
+      "step": 16571
+    },
+    {
+      "epoch": 0.7010745409933159,
+      "grad_norm": 0.21815955638885498,
+      "learning_rate": 0.001,
+      "loss": 1.767,
+      "step": 16572
+    },
+    {
+      "epoch": 0.7011168457568322,
+      "grad_norm": 0.22475208342075348,
+      "learning_rate": 0.001,
+      "loss": 3.201,
+      "step": 16573
+    },
+    {
+      "epoch": 0.7011591505203486,
+      "grad_norm": 0.1533900797367096,
+      "learning_rate": 0.001,
+      "loss": 1.5355,
+      "step": 16574
+    },
+    {
+      "epoch": 0.701201455283865,
+      "grad_norm": 0.18003559112548828,
+      "learning_rate": 0.001,
+      "loss": 2.124,
+      "step": 16575
+    },
+    {
+      "epoch": 0.7012437600473813,
+      "grad_norm": 2.3861000537872314,
+      "learning_rate": 0.001,
+      "loss": 2.4133,
+      "step": 16576
+    },
+    {
+      "epoch": 0.7012860648108977,
+      "grad_norm": 0.18636606633663177,
+      "learning_rate": 0.001,
+      "loss": 1.7297,
+      "step": 16577
+    },
+    {
+      "epoch": 0.7013283695744141,
+      "grad_norm": 0.1709393411874771,
+      "learning_rate": 0.001,
+      "loss": 2.7423,
+      "step": 16578
+    },
+    {
+      "epoch": 0.7013706743379304,
+      "grad_norm": 0.1964663714170456,
+      "learning_rate": 0.001,
+      "loss": 2.2989,
+      "step": 16579
+    },
+    {
+      "epoch": 0.7014129791014468,
+      "grad_norm": 0.13408979773521423,
+      "learning_rate": 0.001,
+      "loss": 1.2903,
+      "step": 16580
+    },
+    {
+      "epoch": 0.7014552838649631,
+      "grad_norm": 0.17530257999897003,
+      "learning_rate": 0.001,
+      "loss": 2.0109,
+      "step": 16581
+    },
+    {
+      "epoch": 0.7014975886284796,
+      "grad_norm": 0.44883960485458374,
+      "learning_rate": 0.001,
+      "loss": 3.1007,
+      "step": 16582
+    },
+    {
+      "epoch": 0.701539893391996,
+      "grad_norm": 4.928041458129883,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 16583
+    },
+    {
+      "epoch": 0.7015821981555123,
+      "grad_norm": 0.1666925698518753,
+      "learning_rate": 0.001,
+      "loss": 2.2696,
+      "step": 16584
+    },
+    {
+      "epoch": 0.7016245029190287,
+      "grad_norm": 0.44181501865386963,
+      "learning_rate": 0.001,
+      "loss": 1.9347,
+      "step": 16585
+    },
+    {
+      "epoch": 0.7016668076825451,
+      "grad_norm": 0.17727473378181458,
+      "learning_rate": 0.001,
+      "loss": 2.1058,
+      "step": 16586
+    },
+    {
+      "epoch": 0.7017091124460614,
+      "grad_norm": 0.1318918913602829,
+      "learning_rate": 0.001,
+      "loss": 2.1093,
+      "step": 16587
+    },
+    {
+      "epoch": 0.7017514172095778,
+      "grad_norm": 0.1690230816602707,
+      "learning_rate": 0.001,
+      "loss": 2.2003,
+      "step": 16588
+    },
+    {
+      "epoch": 0.7017937219730942,
+      "grad_norm": 0.29973065853118896,
+      "learning_rate": 0.001,
+      "loss": 2.0176,
+      "step": 16589
+    },
+    {
+      "epoch": 0.7018360267366105,
+      "grad_norm": 0.5178530216217041,
+      "learning_rate": 0.001,
+      "loss": 3.1915,
+      "step": 16590
+    },
+    {
+      "epoch": 0.7018783315001269,
+      "grad_norm": 0.1572248637676239,
+      "learning_rate": 0.001,
+      "loss": 1.943,
+      "step": 16591
+    },
+    {
+      "epoch": 0.7019206362636433,
+      "grad_norm": 0.187464639544487,
+      "learning_rate": 0.001,
+      "loss": 2.0762,
+      "step": 16592
+    },
+    {
+      "epoch": 0.7019629410271596,
+      "grad_norm": 0.21802839636802673,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 16593
+    },
+    {
+      "epoch": 0.702005245790676,
+      "grad_norm": 0.20468945801258087,
+      "learning_rate": 0.001,
+      "loss": 3.3978,
+      "step": 16594
+    },
+    {
+      "epoch": 0.7020475505541924,
+      "grad_norm": 2.5467371940612793,
+      "learning_rate": 0.001,
+      "loss": 2.1057,
+      "step": 16595
+    },
+    {
+      "epoch": 0.7020898553177087,
+      "grad_norm": 0.198502317070961,
+      "learning_rate": 0.001,
+      "loss": 1.8921,
+      "step": 16596
+    },
+    {
+      "epoch": 0.7021321600812251,
+      "grad_norm": 0.18070580065250397,
+      "learning_rate": 0.001,
+      "loss": 2.1183,
+      "step": 16597
+    },
+    {
+      "epoch": 0.7021744648447416,
+      "grad_norm": 0.4479145109653473,
+      "learning_rate": 0.001,
+      "loss": 2.0895,
+      "step": 16598
+    },
+    {
+      "epoch": 0.7022167696082579,
+      "grad_norm": 0.16668649017810822,
+      "learning_rate": 0.001,
+      "loss": 1.7609,
+      "step": 16599
+    },
+    {
+      "epoch": 0.7022590743717743,
+      "grad_norm": 0.17309440672397614,
+      "learning_rate": 0.001,
+      "loss": 1.5591,
+      "step": 16600
+    },
+    {
+      "epoch": 0.7023013791352907,
+      "grad_norm": 1.5597689151763916,
+      "learning_rate": 0.001,
+      "loss": 2.1839,
+      "step": 16601
+    },
+    {
+      "epoch": 0.702343683898807,
+      "grad_norm": 0.2020406872034073,
+      "learning_rate": 0.001,
+      "loss": 2.4463,
+      "step": 16602
+    },
+    {
+      "epoch": 0.7023859886623234,
+      "grad_norm": 0.17011041939258575,
+      "learning_rate": 0.001,
+      "loss": 2.2795,
+      "step": 16603
+    },
+    {
+      "epoch": 0.7024282934258398,
+      "grad_norm": 0.1457245647907257,
+      "learning_rate": 0.001,
+      "loss": 2.8202,
+      "step": 16604
+    },
+    {
+      "epoch": 0.7024705981893561,
+      "grad_norm": 0.1977822333574295,
+      "learning_rate": 0.001,
+      "loss": 2.1196,
+      "step": 16605
+    },
+    {
+      "epoch": 0.7025129029528725,
+      "grad_norm": 0.14688114821910858,
+      "learning_rate": 0.001,
+      "loss": 1.9252,
+      "step": 16606
+    },
+    {
+      "epoch": 0.7025552077163889,
+      "grad_norm": 0.308075487613678,
+      "learning_rate": 0.001,
+      "loss": 2.4381,
+      "step": 16607
+    },
+    {
+      "epoch": 0.7025975124799052,
+      "grad_norm": 0.1714773029088974,
+      "learning_rate": 0.001,
+      "loss": 1.698,
+      "step": 16608
+    },
+    {
+      "epoch": 0.7026398172434216,
+      "grad_norm": 0.18509937822818756,
+      "learning_rate": 0.001,
+      "loss": 2.676,
+      "step": 16609
+    },
+    {
+      "epoch": 0.702682122006938,
+      "grad_norm": 0.37178945541381836,
+      "learning_rate": 0.001,
+      "loss": 3.245,
+      "step": 16610
+    },
+    {
+      "epoch": 0.7027244267704543,
+      "grad_norm": 0.15938927233219147,
+      "learning_rate": 0.001,
+      "loss": 1.8862,
+      "step": 16611
+    },
+    {
+      "epoch": 0.7027667315339707,
+      "grad_norm": 0.25779882073402405,
+      "learning_rate": 0.001,
+      "loss": 3.5429,
+      "step": 16612
+    },
+    {
+      "epoch": 0.7028090362974871,
+      "grad_norm": 0.1717577427625656,
+      "learning_rate": 0.001,
+      "loss": 1.811,
+      "step": 16613
+    },
+    {
+      "epoch": 0.7028513410610034,
+      "grad_norm": 0.25885605812072754,
+      "learning_rate": 0.001,
+      "loss": 2.4719,
+      "step": 16614
+    },
+    {
+      "epoch": 0.7028936458245199,
+      "grad_norm": 0.15932494401931763,
+      "learning_rate": 0.001,
+      "loss": 2.5687,
+      "step": 16615
+    },
+    {
+      "epoch": 0.7029359505880363,
+      "grad_norm": 0.15366816520690918,
+      "learning_rate": 0.001,
+      "loss": 2.3769,
+      "step": 16616
+    },
+    {
+      "epoch": 0.7029782553515526,
+      "grad_norm": 0.4392596483230591,
+      "learning_rate": 0.001,
+      "loss": 2.3932,
+      "step": 16617
+    },
+    {
+      "epoch": 0.703020560115069,
+      "grad_norm": 0.18615837395191193,
+      "learning_rate": 0.001,
+      "loss": 2.1175,
+      "step": 16618
+    },
+    {
+      "epoch": 0.7030628648785854,
+      "grad_norm": 0.20948411524295807,
+      "learning_rate": 0.001,
+      "loss": 2.4846,
+      "step": 16619
+    },
+    {
+      "epoch": 0.7031051696421017,
+      "grad_norm": 0.7131271362304688,
+      "learning_rate": 0.001,
+      "loss": 2.928,
+      "step": 16620
+    },
+    {
+      "epoch": 0.7031474744056181,
+      "grad_norm": 0.23972147703170776,
+      "learning_rate": 0.001,
+      "loss": 2.3654,
+      "step": 16621
+    },
+    {
+      "epoch": 0.7031897791691345,
+      "grad_norm": 0.23003694415092468,
+      "learning_rate": 0.001,
+      "loss": 2.846,
+      "step": 16622
+    },
+    {
+      "epoch": 0.7032320839326508,
+      "grad_norm": 0.14880388975143433,
+      "learning_rate": 0.001,
+      "loss": 1.8853,
+      "step": 16623
+    },
+    {
+      "epoch": 0.7032743886961672,
+      "grad_norm": 0.1396287977695465,
+      "learning_rate": 0.001,
+      "loss": 2.4475,
+      "step": 16624
+    },
+    {
+      "epoch": 0.7033166934596835,
+      "grad_norm": 0.146676704287529,
+      "learning_rate": 0.001,
+      "loss": 2.1553,
+      "step": 16625
+    },
+    {
+      "epoch": 0.7033589982231999,
+      "grad_norm": 0.17937001585960388,
+      "learning_rate": 0.001,
+      "loss": 2.787,
+      "step": 16626
+    },
+    {
+      "epoch": 0.7034013029867163,
+      "grad_norm": 0.18604117631912231,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 16627
+    },
+    {
+      "epoch": 0.7034436077502326,
+      "grad_norm": 3.127412796020508,
+      "learning_rate": 0.001,
+      "loss": 1.5014,
+      "step": 16628
+    },
+    {
+      "epoch": 0.703485912513749,
+      "grad_norm": 0.6706067323684692,
+      "learning_rate": 0.001,
+      "loss": 2.0265,
+      "step": 16629
+    },
+    {
+      "epoch": 0.7035282172772654,
+      "grad_norm": 0.13607746362686157,
+      "learning_rate": 0.001,
+      "loss": 1.5039,
+      "step": 16630
+    },
+    {
+      "epoch": 0.7035705220407817,
+      "grad_norm": 0.2080031931400299,
+      "learning_rate": 0.001,
+      "loss": 2.7118,
+      "step": 16631
+    },
+    {
+      "epoch": 0.7036128268042982,
+      "grad_norm": 0.2915956377983093,
+      "learning_rate": 0.001,
+      "loss": 1.9805,
+      "step": 16632
+    },
+    {
+      "epoch": 0.7036551315678146,
+      "grad_norm": 1.015211582183838,
+      "learning_rate": 0.001,
+      "loss": 2.4291,
+      "step": 16633
+    },
+    {
+      "epoch": 0.7036974363313309,
+      "grad_norm": 0.15686677396297455,
+      "learning_rate": 0.001,
+      "loss": 1.4743,
+      "step": 16634
+    },
+    {
+      "epoch": 0.7037397410948473,
+      "grad_norm": 0.3258562386035919,
+      "learning_rate": 0.001,
+      "loss": 2.1383,
+      "step": 16635
+    },
+    {
+      "epoch": 0.7037820458583637,
+      "grad_norm": 0.21752215921878815,
+      "learning_rate": 0.001,
+      "loss": 2.0518,
+      "step": 16636
+    },
+    {
+      "epoch": 0.70382435062188,
+      "grad_norm": 3.988706111907959,
+      "learning_rate": 0.001,
+      "loss": 3.0142,
+      "step": 16637
+    },
+    {
+      "epoch": 0.7038666553853964,
+      "grad_norm": 0.15791606903076172,
+      "learning_rate": 0.001,
+      "loss": 2.3269,
+      "step": 16638
+    },
+    {
+      "epoch": 0.7039089601489128,
+      "grad_norm": 0.16375815868377686,
+      "learning_rate": 0.001,
+      "loss": 1.5517,
+      "step": 16639
+    },
+    {
+      "epoch": 0.7039512649124291,
+      "grad_norm": 2.8950631618499756,
+      "learning_rate": 0.001,
+      "loss": 1.831,
+      "step": 16640
+    },
+    {
+      "epoch": 0.7039935696759455,
+      "grad_norm": 0.5850285887718201,
+      "learning_rate": 0.001,
+      "loss": 2.3522,
+      "step": 16641
+    },
+    {
+      "epoch": 0.7040358744394619,
+      "grad_norm": 0.25037726759910583,
+      "learning_rate": 0.001,
+      "loss": 2.2778,
+      "step": 16642
+    },
+    {
+      "epoch": 0.7040781792029782,
+      "grad_norm": 0.17441709339618683,
+      "learning_rate": 0.001,
+      "loss": 2.4771,
+      "step": 16643
+    },
+    {
+      "epoch": 0.7041204839664946,
+      "grad_norm": 0.15096983313560486,
+      "learning_rate": 0.001,
+      "loss": 1.4872,
+      "step": 16644
+    },
+    {
+      "epoch": 0.704162788730011,
+      "grad_norm": 1.1566590070724487,
+      "learning_rate": 0.001,
+      "loss": 1.9335,
+      "step": 16645
+    },
+    {
+      "epoch": 0.7042050934935273,
+      "grad_norm": 0.21050888299942017,
+      "learning_rate": 0.001,
+      "loss": 1.6326,
+      "step": 16646
+    },
+    {
+      "epoch": 0.7042473982570437,
+      "grad_norm": 0.24145705997943878,
+      "learning_rate": 0.001,
+      "loss": 3.1265,
+      "step": 16647
+    },
+    {
+      "epoch": 0.7042897030205602,
+      "grad_norm": 0.20239047706127167,
+      "learning_rate": 0.001,
+      "loss": 1.7814,
+      "step": 16648
+    },
+    {
+      "epoch": 0.7043320077840765,
+      "grad_norm": 0.5811108946800232,
+      "learning_rate": 0.001,
+      "loss": 2.4936,
+      "step": 16649
+    },
+    {
+      "epoch": 0.7043743125475929,
+      "grad_norm": 0.42122969031333923,
+      "learning_rate": 0.001,
+      "loss": 2.2731,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7044166173111093,
+      "grad_norm": 0.3062237501144409,
+      "learning_rate": 0.001,
+      "loss": 1.9344,
+      "step": 16651
+    },
+    {
+      "epoch": 0.7044589220746256,
+      "grad_norm": 1.1125757694244385,
+      "learning_rate": 0.001,
+      "loss": 2.0033,
+      "step": 16652
+    },
+    {
+      "epoch": 0.704501226838142,
+      "grad_norm": 0.20193973183631897,
+      "learning_rate": 0.001,
+      "loss": 2.0509,
+      "step": 16653
+    },
+    {
+      "epoch": 0.7045435316016584,
+      "grad_norm": 5.243770122528076,
+      "learning_rate": 0.001,
+      "loss": 1.7479,
+      "step": 16654
+    },
+    {
+      "epoch": 0.7045858363651747,
+      "grad_norm": 0.1896519958972931,
+      "learning_rate": 0.001,
+      "loss": 1.9381,
+      "step": 16655
+    },
+    {
+      "epoch": 0.7046281411286911,
+      "grad_norm": 0.21094900369644165,
+      "learning_rate": 0.001,
+      "loss": 1.6246,
+      "step": 16656
+    },
+    {
+      "epoch": 0.7046704458922075,
+      "grad_norm": 0.24396421015262604,
+      "learning_rate": 0.001,
+      "loss": 3.4794,
+      "step": 16657
+    },
+    {
+      "epoch": 0.7047127506557238,
+      "grad_norm": 0.30104437470436096,
+      "learning_rate": 0.001,
+      "loss": 3.0927,
+      "step": 16658
+    },
+    {
+      "epoch": 0.7047550554192402,
+      "grad_norm": 0.4394260346889496,
+      "learning_rate": 0.001,
+      "loss": 2.2789,
+      "step": 16659
+    },
+    {
+      "epoch": 0.7047973601827566,
+      "grad_norm": 0.23813438415527344,
+      "learning_rate": 0.001,
+      "loss": 1.8201,
+      "step": 16660
+    },
+    {
+      "epoch": 0.7048396649462729,
+      "grad_norm": 0.2260584682226181,
+      "learning_rate": 0.001,
+      "loss": 2.8637,
+      "step": 16661
+    },
+    {
+      "epoch": 0.7048819697097893,
+      "grad_norm": 0.17535214126110077,
+      "learning_rate": 0.001,
+      "loss": 1.7847,
+      "step": 16662
+    },
+    {
+      "epoch": 0.7049242744733057,
+      "grad_norm": 0.342427134513855,
+      "learning_rate": 0.001,
+      "loss": 2.5649,
+      "step": 16663
+    },
+    {
+      "epoch": 0.704966579236822,
+      "grad_norm": 0.2312489151954651,
+      "learning_rate": 0.001,
+      "loss": 2.3301,
+      "step": 16664
+    },
+    {
+      "epoch": 0.7050088840003385,
+      "grad_norm": 0.21483004093170166,
+      "learning_rate": 0.001,
+      "loss": 2.0729,
+      "step": 16665
+    },
+    {
+      "epoch": 0.7050511887638549,
+      "grad_norm": 2.1229746341705322,
+      "learning_rate": 0.001,
+      "loss": 2.058,
+      "step": 16666
+    },
+    {
+      "epoch": 0.7050934935273712,
+      "grad_norm": 0.20138874650001526,
+      "learning_rate": 0.001,
+      "loss": 2.5361,
+      "step": 16667
+    },
+    {
+      "epoch": 0.7051357982908876,
+      "grad_norm": 0.1767900288105011,
+      "learning_rate": 0.001,
+      "loss": 2.0497,
+      "step": 16668
+    },
+    {
+      "epoch": 0.7051781030544039,
+      "grad_norm": 0.23017890751361847,
+      "learning_rate": 0.001,
+      "loss": 1.6551,
+      "step": 16669
+    },
+    {
+      "epoch": 0.7052204078179203,
+      "grad_norm": 0.3998371362686157,
+      "learning_rate": 0.001,
+      "loss": 2.6177,
+      "step": 16670
+    },
+    {
+      "epoch": 0.7052627125814367,
+      "grad_norm": 0.1956208050251007,
+      "learning_rate": 0.001,
+      "loss": 2.4729,
+      "step": 16671
+    },
+    {
+      "epoch": 0.705305017344953,
+      "grad_norm": 0.13961687684059143,
+      "learning_rate": 0.001,
+      "loss": 2.5719,
+      "step": 16672
+    },
+    {
+      "epoch": 0.7053473221084694,
+      "grad_norm": 0.581203281879425,
+      "learning_rate": 0.001,
+      "loss": 3.0076,
+      "step": 16673
+    },
+    {
+      "epoch": 0.7053896268719858,
+      "grad_norm": 0.18303988873958588,
+      "learning_rate": 0.001,
+      "loss": 1.9536,
+      "step": 16674
+    },
+    {
+      "epoch": 0.7054319316355021,
+      "grad_norm": 0.18635371327400208,
+      "learning_rate": 0.001,
+      "loss": 1.5738,
+      "step": 16675
+    },
+    {
+      "epoch": 0.7054742363990185,
+      "grad_norm": 0.17525175213813782,
+      "learning_rate": 0.001,
+      "loss": 2.3192,
+      "step": 16676
+    },
+    {
+      "epoch": 0.7055165411625349,
+      "grad_norm": 1.5950613021850586,
+      "learning_rate": 0.001,
+      "loss": 2.6333,
+      "step": 16677
+    },
+    {
+      "epoch": 0.7055588459260512,
+      "grad_norm": 0.2035263627767563,
+      "learning_rate": 0.001,
+      "loss": 1.9577,
+      "step": 16678
+    },
+    {
+      "epoch": 0.7056011506895676,
+      "grad_norm": 0.16556823253631592,
+      "learning_rate": 0.001,
+      "loss": 2.3302,
+      "step": 16679
+    },
+    {
+      "epoch": 0.705643455453084,
+      "grad_norm": 6.488144397735596,
+      "learning_rate": 0.001,
+      "loss": 2.0294,
+      "step": 16680
+    },
+    {
+      "epoch": 0.7056857602166003,
+      "grad_norm": 0.21068356931209564,
+      "learning_rate": 0.001,
+      "loss": 3.265,
+      "step": 16681
+    },
+    {
+      "epoch": 0.7057280649801168,
+      "grad_norm": 0.1684761941432953,
+      "learning_rate": 0.001,
+      "loss": 2.9163,
+      "step": 16682
+    },
+    {
+      "epoch": 0.7057703697436332,
+      "grad_norm": 2.740279197692871,
+      "learning_rate": 0.001,
+      "loss": 1.8363,
+      "step": 16683
+    },
+    {
+      "epoch": 0.7058126745071495,
+      "grad_norm": 69.86604309082031,
+      "learning_rate": 0.001,
+      "loss": 2.5254,
+      "step": 16684
+    },
+    {
+      "epoch": 0.7058549792706659,
+      "grad_norm": 0.2736927568912506,
+      "learning_rate": 0.001,
+      "loss": 2.2184,
+      "step": 16685
+    },
+    {
+      "epoch": 0.7058972840341823,
+      "grad_norm": 0.21192248165607452,
+      "learning_rate": 0.001,
+      "loss": 2.7321,
+      "step": 16686
+    },
+    {
+      "epoch": 0.7059395887976986,
+      "grad_norm": 0.3148384988307953,
+      "learning_rate": 0.001,
+      "loss": 1.9477,
+      "step": 16687
+    },
+    {
+      "epoch": 0.705981893561215,
+      "grad_norm": 0.24907708168029785,
+      "learning_rate": 0.001,
+      "loss": 2.7906,
+      "step": 16688
+    },
+    {
+      "epoch": 0.7060241983247314,
+      "grad_norm": 0.3226729929447174,
+      "learning_rate": 0.001,
+      "loss": 2.8049,
+      "step": 16689
+    },
+    {
+      "epoch": 0.7060665030882477,
+      "grad_norm": 3.953677177429199,
+      "learning_rate": 0.001,
+      "loss": 2.0564,
+      "step": 16690
+    },
+    {
+      "epoch": 0.7061088078517641,
+      "grad_norm": 0.24923500418663025,
+      "learning_rate": 0.001,
+      "loss": 1.3582,
+      "step": 16691
+    },
+    {
+      "epoch": 0.7061511126152805,
+      "grad_norm": 0.9128309488296509,
+      "learning_rate": 0.001,
+      "loss": 3.0,
+      "step": 16692
+    },
+    {
+      "epoch": 0.7061934173787968,
+      "grad_norm": 0.30432406067848206,
+      "learning_rate": 0.001,
+      "loss": 2.2832,
+      "step": 16693
+    },
+    {
+      "epoch": 0.7062357221423132,
+      "grad_norm": 0.27826377749443054,
+      "learning_rate": 0.001,
+      "loss": 3.6967,
+      "step": 16694
+    },
+    {
+      "epoch": 0.7062780269058296,
+      "grad_norm": 1.1862016916275024,
+      "learning_rate": 0.001,
+      "loss": 2.8519,
+      "step": 16695
+    },
+    {
+      "epoch": 0.7063203316693459,
+      "grad_norm": 0.27995389699935913,
+      "learning_rate": 0.001,
+      "loss": 2.0341,
+      "step": 16696
+    },
+    {
+      "epoch": 0.7063626364328623,
+      "grad_norm": 0.20460890233516693,
+      "learning_rate": 0.001,
+      "loss": 2.0071,
+      "step": 16697
+    },
+    {
+      "epoch": 0.7064049411963788,
+      "grad_norm": 0.20291836559772491,
+      "learning_rate": 0.001,
+      "loss": 2.3107,
+      "step": 16698
+    },
+    {
+      "epoch": 0.706447245959895,
+      "grad_norm": 0.2692174017429352,
+      "learning_rate": 0.001,
+      "loss": 2.5738,
+      "step": 16699
+    },
+    {
+      "epoch": 0.7064895507234115,
+      "grad_norm": 0.1654096394777298,
+      "learning_rate": 0.001,
+      "loss": 2.286,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7065318554869279,
+      "grad_norm": 0.23862436413764954,
+      "learning_rate": 0.001,
+      "loss": 2.321,
+      "step": 16701
+    },
+    {
+      "epoch": 0.7065741602504442,
+      "grad_norm": 0.1421448439359665,
+      "learning_rate": 0.001,
+      "loss": 2.7367,
+      "step": 16702
+    },
+    {
+      "epoch": 0.7066164650139606,
+      "grad_norm": 0.5097910165786743,
+      "learning_rate": 0.001,
+      "loss": 2.5246,
+      "step": 16703
+    },
+    {
+      "epoch": 0.706658769777477,
+      "grad_norm": 0.19460159540176392,
+      "learning_rate": 0.001,
+      "loss": 2.2217,
+      "step": 16704
+    },
+    {
+      "epoch": 0.7067010745409933,
+      "grad_norm": 0.17561720311641693,
+      "learning_rate": 0.001,
+      "loss": 1.7901,
+      "step": 16705
+    },
+    {
+      "epoch": 0.7067433793045097,
+      "grad_norm": 0.16379234194755554,
+      "learning_rate": 0.001,
+      "loss": 2.0123,
+      "step": 16706
+    },
+    {
+      "epoch": 0.7067856840680261,
+      "grad_norm": 0.29534009099006653,
+      "learning_rate": 0.001,
+      "loss": 1.7291,
+      "step": 16707
+    },
+    {
+      "epoch": 0.7068279888315424,
+      "grad_norm": 0.16030588746070862,
+      "learning_rate": 0.001,
+      "loss": 2.4982,
+      "step": 16708
+    },
+    {
+      "epoch": 0.7068702935950588,
+      "grad_norm": 0.19317007064819336,
+      "learning_rate": 0.001,
+      "loss": 1.7916,
+      "step": 16709
+    },
+    {
+      "epoch": 0.7069125983585752,
+      "grad_norm": 16.575075149536133,
+      "learning_rate": 0.001,
+      "loss": 1.7214,
+      "step": 16710
+    },
+    {
+      "epoch": 0.7069549031220915,
+      "grad_norm": 0.6508944630622864,
+      "learning_rate": 0.001,
+      "loss": 3.3014,
+      "step": 16711
+    },
+    {
+      "epoch": 0.7069972078856079,
+      "grad_norm": 0.17529930174350739,
+      "learning_rate": 0.001,
+      "loss": 2.833,
+      "step": 16712
+    },
+    {
+      "epoch": 0.7070395126491243,
+      "grad_norm": 0.18948234617710114,
+      "learning_rate": 0.001,
+      "loss": 2.3365,
+      "step": 16713
+    },
+    {
+      "epoch": 0.7070818174126406,
+      "grad_norm": 0.19644437730312347,
+      "learning_rate": 0.001,
+      "loss": 3.7717,
+      "step": 16714
+    },
+    {
+      "epoch": 0.707124122176157,
+      "grad_norm": 0.36067622900009155,
+      "learning_rate": 0.001,
+      "loss": 2.016,
+      "step": 16715
+    },
+    {
+      "epoch": 0.7071664269396734,
+      "grad_norm": 0.1735169142484665,
+      "learning_rate": 0.001,
+      "loss": 2.1556,
+      "step": 16716
+    },
+    {
+      "epoch": 0.7072087317031898,
+      "grad_norm": 0.7500831484794617,
+      "learning_rate": 0.001,
+      "loss": 2.5418,
+      "step": 16717
+    },
+    {
+      "epoch": 0.7072510364667062,
+      "grad_norm": 0.1380908042192459,
+      "learning_rate": 0.001,
+      "loss": 2.0679,
+      "step": 16718
+    },
+    {
+      "epoch": 0.7072933412302225,
+      "grad_norm": 1.2244987487792969,
+      "learning_rate": 0.001,
+      "loss": 2.6043,
+      "step": 16719
+    },
+    {
+      "epoch": 0.7073356459937389,
+      "grad_norm": 0.14987333118915558,
+      "learning_rate": 0.001,
+      "loss": 2.4967,
+      "step": 16720
+    },
+    {
+      "epoch": 0.7073779507572553,
+      "grad_norm": 0.17000074684619904,
+      "learning_rate": 0.001,
+      "loss": 2.2771,
+      "step": 16721
+    },
+    {
+      "epoch": 0.7074202555207716,
+      "grad_norm": 2.4218010902404785,
+      "learning_rate": 0.001,
+      "loss": 2.0412,
+      "step": 16722
+    },
+    {
+      "epoch": 0.707462560284288,
+      "grad_norm": 0.24326300621032715,
+      "learning_rate": 0.001,
+      "loss": 2.3567,
+      "step": 16723
+    },
+    {
+      "epoch": 0.7075048650478044,
+      "grad_norm": 0.5091056227684021,
+      "learning_rate": 0.001,
+      "loss": 2.88,
+      "step": 16724
+    },
+    {
+      "epoch": 0.7075471698113207,
+      "grad_norm": 0.16985633969306946,
+      "learning_rate": 0.001,
+      "loss": 1.6534,
+      "step": 16725
+    },
+    {
+      "epoch": 0.7075894745748371,
+      "grad_norm": 0.20579244196414948,
+      "learning_rate": 0.001,
+      "loss": 1.6615,
+      "step": 16726
+    },
+    {
+      "epoch": 0.7076317793383535,
+      "grad_norm": 0.19666582345962524,
+      "learning_rate": 0.001,
+      "loss": 2.5436,
+      "step": 16727
+    },
+    {
+      "epoch": 0.7076740841018698,
+      "grad_norm": 0.1708710938692093,
+      "learning_rate": 0.001,
+      "loss": 1.5373,
+      "step": 16728
+    },
+    {
+      "epoch": 0.7077163888653862,
+      "grad_norm": 0.17626847326755524,
+      "learning_rate": 0.001,
+      "loss": 1.9095,
+      "step": 16729
+    },
+    {
+      "epoch": 0.7077586936289026,
+      "grad_norm": 0.19842320680618286,
+      "learning_rate": 0.001,
+      "loss": 2.8567,
+      "step": 16730
+    },
+    {
+      "epoch": 0.707800998392419,
+      "grad_norm": 0.17139360308647156,
+      "learning_rate": 0.001,
+      "loss": 2.3247,
+      "step": 16731
+    },
+    {
+      "epoch": 0.7078433031559354,
+      "grad_norm": 0.14886251091957092,
+      "learning_rate": 0.001,
+      "loss": 1.6924,
+      "step": 16732
+    },
+    {
+      "epoch": 0.7078856079194518,
+      "grad_norm": 0.182938814163208,
+      "learning_rate": 0.001,
+      "loss": 1.2942,
+      "step": 16733
+    },
+    {
+      "epoch": 0.7079279126829681,
+      "grad_norm": 0.21765394508838654,
+      "learning_rate": 0.001,
+      "loss": 2.992,
+      "step": 16734
+    },
+    {
+      "epoch": 0.7079702174464845,
+      "grad_norm": 49.08909606933594,
+      "learning_rate": 0.001,
+      "loss": 2.9176,
+      "step": 16735
+    },
+    {
+      "epoch": 0.7080125222100009,
+      "grad_norm": 0.15202230215072632,
+      "learning_rate": 0.001,
+      "loss": 1.8571,
+      "step": 16736
+    },
+    {
+      "epoch": 0.7080548269735172,
+      "grad_norm": 0.2932276129722595,
+      "learning_rate": 0.001,
+      "loss": 2.6166,
+      "step": 16737
+    },
+    {
+      "epoch": 0.7080971317370336,
+      "grad_norm": 0.15039752423763275,
+      "learning_rate": 0.001,
+      "loss": 1.5913,
+      "step": 16738
+    },
+    {
+      "epoch": 0.70813943650055,
+      "grad_norm": 0.1468665897846222,
+      "learning_rate": 0.001,
+      "loss": 2.4612,
+      "step": 16739
+    },
+    {
+      "epoch": 0.7081817412640663,
+      "grad_norm": 0.17689596116542816,
+      "learning_rate": 0.001,
+      "loss": 2.5566,
+      "step": 16740
+    },
+    {
+      "epoch": 0.7082240460275827,
+      "grad_norm": 0.22001603245735168,
+      "learning_rate": 0.001,
+      "loss": 2.2724,
+      "step": 16741
+    },
+    {
+      "epoch": 0.7082663507910991,
+      "grad_norm": 0.24464015662670135,
+      "learning_rate": 0.001,
+      "loss": 2.3597,
+      "step": 16742
+    },
+    {
+      "epoch": 0.7083086555546154,
+      "grad_norm": 0.1501077562570572,
+      "learning_rate": 0.001,
+      "loss": 2.4862,
+      "step": 16743
+    },
+    {
+      "epoch": 0.7083509603181318,
+      "grad_norm": 0.22457842528820038,
+      "learning_rate": 0.001,
+      "loss": 1.9533,
+      "step": 16744
+    },
+    {
+      "epoch": 0.7083932650816482,
+      "grad_norm": 0.48851919174194336,
+      "learning_rate": 0.001,
+      "loss": 1.8844,
+      "step": 16745
+    },
+    {
+      "epoch": 0.7084355698451645,
+      "grad_norm": 0.16567832231521606,
+      "learning_rate": 0.001,
+      "loss": 2.3328,
+      "step": 16746
+    },
+    {
+      "epoch": 0.708477874608681,
+      "grad_norm": 0.16666191816329956,
+      "learning_rate": 0.001,
+      "loss": 3.0872,
+      "step": 16747
+    },
+    {
+      "epoch": 0.7085201793721974,
+      "grad_norm": 0.1672544926404953,
+      "learning_rate": 0.001,
+      "loss": 2.2649,
+      "step": 16748
+    },
+    {
+      "epoch": 0.7085624841357137,
+      "grad_norm": 0.1558084934949875,
+      "learning_rate": 0.001,
+      "loss": 1.8319,
+      "step": 16749
+    },
+    {
+      "epoch": 0.7086047888992301,
+      "grad_norm": 0.14037245512008667,
+      "learning_rate": 0.001,
+      "loss": 2.3805,
+      "step": 16750
+    },
+    {
+      "epoch": 0.7086470936627465,
+      "grad_norm": 0.16425585746765137,
+      "learning_rate": 0.001,
+      "loss": 1.7784,
+      "step": 16751
+    },
+    {
+      "epoch": 0.7086893984262628,
+      "grad_norm": 0.7935085892677307,
+      "learning_rate": 0.001,
+      "loss": 2.1095,
+      "step": 16752
+    },
+    {
+      "epoch": 0.7087317031897792,
+      "grad_norm": 0.16057053208351135,
+      "learning_rate": 0.001,
+      "loss": 1.8058,
+      "step": 16753
+    },
+    {
+      "epoch": 0.7087740079532956,
+      "grad_norm": 0.16115844249725342,
+      "learning_rate": 0.001,
+      "loss": 1.6998,
+      "step": 16754
+    },
+    {
+      "epoch": 0.7088163127168119,
+      "grad_norm": 1.6587260961532593,
+      "learning_rate": 0.001,
+      "loss": 2.3739,
+      "step": 16755
+    },
+    {
+      "epoch": 0.7088586174803283,
+      "grad_norm": 0.11960441619157791,
+      "learning_rate": 0.001,
+      "loss": 1.2877,
+      "step": 16756
+    },
+    {
+      "epoch": 0.7089009222438447,
+      "grad_norm": 0.28413790464401245,
+      "learning_rate": 0.001,
+      "loss": 2.0965,
+      "step": 16757
+    },
+    {
+      "epoch": 0.708943227007361,
+      "grad_norm": 0.18682576715946198,
+      "learning_rate": 0.001,
+      "loss": 2.3906,
+      "step": 16758
+    },
+    {
+      "epoch": 0.7089855317708774,
+      "grad_norm": 0.15329422056674957,
+      "learning_rate": 0.001,
+      "loss": 2.0463,
+      "step": 16759
+    },
+    {
+      "epoch": 0.7090278365343937,
+      "grad_norm": 0.18216441571712494,
+      "learning_rate": 0.001,
+      "loss": 2.1109,
+      "step": 16760
+    },
+    {
+      "epoch": 0.7090701412979101,
+      "grad_norm": 0.38540518283843994,
+      "learning_rate": 0.001,
+      "loss": 1.7257,
+      "step": 16761
+    },
+    {
+      "epoch": 0.7091124460614265,
+      "grad_norm": 0.2819367051124573,
+      "learning_rate": 0.001,
+      "loss": 1.708,
+      "step": 16762
+    },
+    {
+      "epoch": 0.7091547508249428,
+      "grad_norm": 0.16718243062496185,
+      "learning_rate": 0.001,
+      "loss": 1.9525,
+      "step": 16763
+    },
+    {
+      "epoch": 0.7091970555884592,
+      "grad_norm": 0.2087067812681198,
+      "learning_rate": 0.001,
+      "loss": 2.1724,
+      "step": 16764
+    },
+    {
+      "epoch": 0.7092393603519757,
+      "grad_norm": 0.46061161160469055,
+      "learning_rate": 0.001,
+      "loss": 3.8056,
+      "step": 16765
+    },
+    {
+      "epoch": 0.709281665115492,
+      "grad_norm": 0.20006658136844635,
+      "learning_rate": 0.001,
+      "loss": 3.0642,
+      "step": 16766
+    },
+    {
+      "epoch": 0.7093239698790084,
+      "grad_norm": 0.15130923688411713,
+      "learning_rate": 0.001,
+      "loss": 2.0505,
+      "step": 16767
+    },
+    {
+      "epoch": 0.7093662746425248,
+      "grad_norm": 0.20808841288089752,
+      "learning_rate": 0.001,
+      "loss": 2.407,
+      "step": 16768
+    },
+    {
+      "epoch": 0.7094085794060411,
+      "grad_norm": 0.17218559980392456,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 16769
+    },
+    {
+      "epoch": 0.7094508841695575,
+      "grad_norm": 0.14409933984279633,
+      "learning_rate": 0.001,
+      "loss": 1.674,
+      "step": 16770
+    },
+    {
+      "epoch": 0.7094931889330739,
+      "grad_norm": 0.19326108694076538,
+      "learning_rate": 0.001,
+      "loss": 2.979,
+      "step": 16771
+    },
+    {
+      "epoch": 0.7095354936965902,
+      "grad_norm": 0.1743086278438568,
+      "learning_rate": 0.001,
+      "loss": 3.2637,
+      "step": 16772
+    },
+    {
+      "epoch": 0.7095777984601066,
+      "grad_norm": 0.17006252706050873,
+      "learning_rate": 0.001,
+      "loss": 1.9035,
+      "step": 16773
+    },
+    {
+      "epoch": 0.709620103223623,
+      "grad_norm": 0.1594936102628708,
+      "learning_rate": 0.001,
+      "loss": 1.8383,
+      "step": 16774
+    },
+    {
+      "epoch": 0.7096624079871393,
+      "grad_norm": 0.1599188596010208,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 16775
+    },
+    {
+      "epoch": 0.7097047127506557,
+      "grad_norm": 0.22632001340389252,
+      "learning_rate": 0.001,
+      "loss": 2.0174,
+      "step": 16776
+    },
+    {
+      "epoch": 0.7097470175141721,
+      "grad_norm": 0.151228666305542,
+      "learning_rate": 0.001,
+      "loss": 3.3681,
+      "step": 16777
+    },
+    {
+      "epoch": 0.7097893222776884,
+      "grad_norm": 0.154108464717865,
+      "learning_rate": 0.001,
+      "loss": 2.6175,
+      "step": 16778
+    },
+    {
+      "epoch": 0.7098316270412048,
+      "grad_norm": 0.23178794980049133,
+      "learning_rate": 0.001,
+      "loss": 2.6288,
+      "step": 16779
+    },
+    {
+      "epoch": 0.7098739318047212,
+      "grad_norm": 0.18430373072624207,
+      "learning_rate": 0.001,
+      "loss": 1.8585,
+      "step": 16780
+    },
+    {
+      "epoch": 0.7099162365682375,
+      "grad_norm": 1.811462640762329,
+      "learning_rate": 0.001,
+      "loss": 2.4217,
+      "step": 16781
+    },
+    {
+      "epoch": 0.709958541331754,
+      "grad_norm": 0.14598025381565094,
+      "learning_rate": 0.001,
+      "loss": 1.4804,
+      "step": 16782
+    },
+    {
+      "epoch": 0.7100008460952704,
+      "grad_norm": 0.162042036652565,
+      "learning_rate": 0.001,
+      "loss": 2.7988,
+      "step": 16783
+    },
+    {
+      "epoch": 0.7100431508587867,
+      "grad_norm": 0.24106714129447937,
+      "learning_rate": 0.001,
+      "loss": 2.4054,
+      "step": 16784
+    },
+    {
+      "epoch": 0.7100854556223031,
+      "grad_norm": 0.2140771746635437,
+      "learning_rate": 0.001,
+      "loss": 3.0376,
+      "step": 16785
+    },
+    {
+      "epoch": 0.7101277603858195,
+      "grad_norm": 0.24649621546268463,
+      "learning_rate": 0.001,
+      "loss": 1.7422,
+      "step": 16786
+    },
+    {
+      "epoch": 0.7101700651493358,
+      "grad_norm": 0.5776690244674683,
+      "learning_rate": 0.001,
+      "loss": 2.1208,
+      "step": 16787
+    },
+    {
+      "epoch": 0.7102123699128522,
+      "grad_norm": 0.21545056998729706,
+      "learning_rate": 0.001,
+      "loss": 2.2494,
+      "step": 16788
+    },
+    {
+      "epoch": 0.7102546746763686,
+      "grad_norm": 0.16463403403759003,
+      "learning_rate": 0.001,
+      "loss": 2.7351,
+      "step": 16789
+    },
+    {
+      "epoch": 0.7102969794398849,
+      "grad_norm": 0.20469972491264343,
+      "learning_rate": 0.001,
+      "loss": 2.3172,
+      "step": 16790
+    },
+    {
+      "epoch": 0.7103392842034013,
+      "grad_norm": 0.1745065301656723,
+      "learning_rate": 0.001,
+      "loss": 2.9764,
+      "step": 16791
+    },
+    {
+      "epoch": 0.7103815889669177,
+      "grad_norm": 0.15587376058101654,
+      "learning_rate": 0.001,
+      "loss": 2.395,
+      "step": 16792
+    },
+    {
+      "epoch": 0.710423893730434,
+      "grad_norm": 0.21353621780872345,
+      "learning_rate": 0.001,
+      "loss": 2.1667,
+      "step": 16793
+    },
+    {
+      "epoch": 0.7104661984939504,
+      "grad_norm": 1.4838371276855469,
+      "learning_rate": 0.001,
+      "loss": 2.7724,
+      "step": 16794
+    },
+    {
+      "epoch": 0.7105085032574668,
+      "grad_norm": 0.18408633768558502,
+      "learning_rate": 0.001,
+      "loss": 1.7745,
+      "step": 16795
+    },
+    {
+      "epoch": 0.7105508080209831,
+      "grad_norm": 0.20254147052764893,
+      "learning_rate": 0.001,
+      "loss": 3.6371,
+      "step": 16796
+    },
+    {
+      "epoch": 0.7105931127844995,
+      "grad_norm": 0.1831352263689041,
+      "learning_rate": 0.001,
+      "loss": 1.6486,
+      "step": 16797
+    },
+    {
+      "epoch": 0.710635417548016,
+      "grad_norm": 0.18303878605365753,
+      "learning_rate": 0.001,
+      "loss": 1.9915,
+      "step": 16798
+    },
+    {
+      "epoch": 0.7106777223115323,
+      "grad_norm": 8.696745872497559,
+      "learning_rate": 0.001,
+      "loss": 2.9612,
+      "step": 16799
+    },
+    {
+      "epoch": 0.7107200270750487,
+      "grad_norm": 0.16578525304794312,
+      "learning_rate": 0.001,
+      "loss": 2.5333,
+      "step": 16800
+    },
+    {
+      "epoch": 0.7107623318385651,
+      "grad_norm": 0.19835564494132996,
+      "learning_rate": 0.001,
+      "loss": 2.5376,
+      "step": 16801
+    },
+    {
+      "epoch": 0.7108046366020814,
+      "grad_norm": 0.18402199447155,
+      "learning_rate": 0.001,
+      "loss": 2.1194,
+      "step": 16802
+    },
+    {
+      "epoch": 0.7108469413655978,
+      "grad_norm": 0.39828217029571533,
+      "learning_rate": 0.001,
+      "loss": 2.7564,
+      "step": 16803
+    },
+    {
+      "epoch": 0.7108892461291141,
+      "grad_norm": 0.17361795902252197,
+      "learning_rate": 0.001,
+      "loss": 2.9527,
+      "step": 16804
+    },
+    {
+      "epoch": 0.7109315508926305,
+      "grad_norm": 0.2756977379322052,
+      "learning_rate": 0.001,
+      "loss": 2.1633,
+      "step": 16805
+    },
+    {
+      "epoch": 0.7109738556561469,
+      "grad_norm": 0.22067061066627502,
+      "learning_rate": 0.001,
+      "loss": 2.6841,
+      "step": 16806
+    },
+    {
+      "epoch": 0.7110161604196632,
+      "grad_norm": 0.1891237497329712,
+      "learning_rate": 0.001,
+      "loss": 1.4142,
+      "step": 16807
+    },
+    {
+      "epoch": 0.7110584651831796,
+      "grad_norm": 1.683262586593628,
+      "learning_rate": 0.001,
+      "loss": 2.0285,
+      "step": 16808
+    },
+    {
+      "epoch": 0.711100769946696,
+      "grad_norm": 0.16583381593227386,
+      "learning_rate": 0.001,
+      "loss": 1.7941,
+      "step": 16809
+    },
+    {
+      "epoch": 0.7111430747102123,
+      "grad_norm": 2.930053472518921,
+      "learning_rate": 0.001,
+      "loss": 1.8499,
+      "step": 16810
+    },
+    {
+      "epoch": 0.7111853794737287,
+      "grad_norm": 0.1657000333070755,
+      "learning_rate": 0.001,
+      "loss": 1.7437,
+      "step": 16811
+    },
+    {
+      "epoch": 0.7112276842372451,
+      "grad_norm": 0.16928398609161377,
+      "learning_rate": 0.001,
+      "loss": 1.7506,
+      "step": 16812
+    },
+    {
+      "epoch": 0.7112699890007614,
+      "grad_norm": 0.1839533895254135,
+      "learning_rate": 0.001,
+      "loss": 1.7639,
+      "step": 16813
+    },
+    {
+      "epoch": 0.7113122937642778,
+      "grad_norm": 0.15033124387264252,
+      "learning_rate": 0.001,
+      "loss": 2.3976,
+      "step": 16814
+    },
+    {
+      "epoch": 0.7113545985277943,
+      "grad_norm": 0.21158377826213837,
+      "learning_rate": 0.001,
+      "loss": 2.7739,
+      "step": 16815
+    },
+    {
+      "epoch": 0.7113969032913106,
+      "grad_norm": 0.14168700575828552,
+      "learning_rate": 0.001,
+      "loss": 1.7022,
+      "step": 16816
+    },
+    {
+      "epoch": 0.711439208054827,
+      "grad_norm": 0.17619457840919495,
+      "learning_rate": 0.001,
+      "loss": 3.0258,
+      "step": 16817
+    },
+    {
+      "epoch": 0.7114815128183434,
+      "grad_norm": 2.186882972717285,
+      "learning_rate": 0.001,
+      "loss": 1.669,
+      "step": 16818
+    },
+    {
+      "epoch": 0.7115238175818597,
+      "grad_norm": 0.37411370873451233,
+      "learning_rate": 0.001,
+      "loss": 1.7232,
+      "step": 16819
+    },
+    {
+      "epoch": 0.7115661223453761,
+      "grad_norm": 0.16331219673156738,
+      "learning_rate": 0.001,
+      "loss": 1.4706,
+      "step": 16820
+    },
+    {
+      "epoch": 0.7116084271088925,
+      "grad_norm": 0.9669393301010132,
+      "learning_rate": 0.001,
+      "loss": 2.8231,
+      "step": 16821
+    },
+    {
+      "epoch": 0.7116507318724088,
+      "grad_norm": 1.7849273681640625,
+      "learning_rate": 0.001,
+      "loss": 2.0226,
+      "step": 16822
+    },
+    {
+      "epoch": 0.7116930366359252,
+      "grad_norm": 5.401309013366699,
+      "learning_rate": 0.001,
+      "loss": 2.1295,
+      "step": 16823
+    },
+    {
+      "epoch": 0.7117353413994416,
+      "grad_norm": 4.94477653503418,
+      "learning_rate": 0.001,
+      "loss": 3.141,
+      "step": 16824
+    },
+    {
+      "epoch": 0.7117776461629579,
+      "grad_norm": 0.1646738350391388,
+      "learning_rate": 0.001,
+      "loss": 2.3503,
+      "step": 16825
+    },
+    {
+      "epoch": 0.7118199509264743,
+      "grad_norm": 0.2677449584007263,
+      "learning_rate": 0.001,
+      "loss": 1.6298,
+      "step": 16826
+    },
+    {
+      "epoch": 0.7118622556899907,
+      "grad_norm": 0.19275929033756256,
+      "learning_rate": 0.001,
+      "loss": 1.8281,
+      "step": 16827
+    },
+    {
+      "epoch": 0.711904560453507,
+      "grad_norm": 0.9130849242210388,
+      "learning_rate": 0.001,
+      "loss": 2.2454,
+      "step": 16828
+    },
+    {
+      "epoch": 0.7119468652170234,
+      "grad_norm": 0.16374343633651733,
+      "learning_rate": 0.001,
+      "loss": 1.355,
+      "step": 16829
+    },
+    {
+      "epoch": 0.7119891699805398,
+      "grad_norm": 0.19432541728019714,
+      "learning_rate": 0.001,
+      "loss": 3.3869,
+      "step": 16830
+    },
+    {
+      "epoch": 0.7120314747440561,
+      "grad_norm": 0.8363915681838989,
+      "learning_rate": 0.001,
+      "loss": 2.2074,
+      "step": 16831
+    },
+    {
+      "epoch": 0.7120737795075726,
+      "grad_norm": 0.18882228434085846,
+      "learning_rate": 0.001,
+      "loss": 2.0904,
+      "step": 16832
+    },
+    {
+      "epoch": 0.712116084271089,
+      "grad_norm": 0.16533978283405304,
+      "learning_rate": 0.001,
+      "loss": 1.8895,
+      "step": 16833
+    },
+    {
+      "epoch": 0.7121583890346053,
+      "grad_norm": 2.9509565830230713,
+      "learning_rate": 0.001,
+      "loss": 2.142,
+      "step": 16834
+    },
+    {
+      "epoch": 0.7122006937981217,
+      "grad_norm": 0.2698115408420563,
+      "learning_rate": 0.001,
+      "loss": 3.958,
+      "step": 16835
+    },
+    {
+      "epoch": 0.7122429985616381,
+      "grad_norm": 0.1884506493806839,
+      "learning_rate": 0.001,
+      "loss": 2.1897,
+      "step": 16836
+    },
+    {
+      "epoch": 0.7122853033251544,
+      "grad_norm": 59.53257751464844,
+      "learning_rate": 0.001,
+      "loss": 2.1418,
+      "step": 16837
+    },
+    {
+      "epoch": 0.7123276080886708,
+      "grad_norm": 4.942716121673584,
+      "learning_rate": 0.001,
+      "loss": 2.5548,
+      "step": 16838
+    },
+    {
+      "epoch": 0.7123699128521872,
+      "grad_norm": 0.3117887079715729,
+      "learning_rate": 0.001,
+      "loss": 1.6967,
+      "step": 16839
+    },
+    {
+      "epoch": 0.7124122176157035,
+      "grad_norm": 0.2594894766807556,
+      "learning_rate": 0.001,
+      "loss": 1.6713,
+      "step": 16840
+    },
+    {
+      "epoch": 0.7124545223792199,
+      "grad_norm": 0.37897738814353943,
+      "learning_rate": 0.001,
+      "loss": 1.9001,
+      "step": 16841
+    },
+    {
+      "epoch": 0.7124968271427363,
+      "grad_norm": 0.3171361982822418,
+      "learning_rate": 0.001,
+      "loss": 2.6247,
+      "step": 16842
+    },
+    {
+      "epoch": 0.7125391319062526,
+      "grad_norm": 0.15616539120674133,
+      "learning_rate": 0.001,
+      "loss": 2.1841,
+      "step": 16843
+    },
+    {
+      "epoch": 0.712581436669769,
+      "grad_norm": 0.363093763589859,
+      "learning_rate": 0.001,
+      "loss": 1.9907,
+      "step": 16844
+    },
+    {
+      "epoch": 0.7126237414332854,
+      "grad_norm": 0.20196084678173065,
+      "learning_rate": 0.001,
+      "loss": 2.3817,
+      "step": 16845
+    },
+    {
+      "epoch": 0.7126660461968017,
+      "grad_norm": 0.19358810782432556,
+      "learning_rate": 0.001,
+      "loss": 1.6368,
+      "step": 16846
+    },
+    {
+      "epoch": 0.7127083509603181,
+      "grad_norm": 0.24825577437877655,
+      "learning_rate": 0.001,
+      "loss": 2.6596,
+      "step": 16847
+    },
+    {
+      "epoch": 0.7127506557238346,
+      "grad_norm": 0.20185577869415283,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 16848
+    },
+    {
+      "epoch": 0.7127929604873509,
+      "grad_norm": 0.3401681184768677,
+      "learning_rate": 0.001,
+      "loss": 1.8744,
+      "step": 16849
+    },
+    {
+      "epoch": 0.7128352652508673,
+      "grad_norm": 0.2052794247865677,
+      "learning_rate": 0.001,
+      "loss": 1.8685,
+      "step": 16850
+    },
+    {
+      "epoch": 0.7128775700143836,
+      "grad_norm": 0.17829133570194244,
+      "learning_rate": 0.001,
+      "loss": 2.681,
+      "step": 16851
+    },
+    {
+      "epoch": 0.7129198747779,
+      "grad_norm": 0.2980566918849945,
+      "learning_rate": 0.001,
+      "loss": 2.1763,
+      "step": 16852
+    },
+    {
+      "epoch": 0.7129621795414164,
+      "grad_norm": 0.22543774545192719,
+      "learning_rate": 0.001,
+      "loss": 1.9288,
+      "step": 16853
+    },
+    {
+      "epoch": 0.7130044843049327,
+      "grad_norm": 1.7789720296859741,
+      "learning_rate": 0.001,
+      "loss": 2.1398,
+      "step": 16854
+    },
+    {
+      "epoch": 0.7130467890684491,
+      "grad_norm": 0.2461077868938446,
+      "learning_rate": 0.001,
+      "loss": 1.7258,
+      "step": 16855
+    },
+    {
+      "epoch": 0.7130890938319655,
+      "grad_norm": 0.1980772763490677,
+      "learning_rate": 0.001,
+      "loss": 2.8052,
+      "step": 16856
+    },
+    {
+      "epoch": 0.7131313985954818,
+      "grad_norm": 0.19356103241443634,
+      "learning_rate": 0.001,
+      "loss": 2.0402,
+      "step": 16857
+    },
+    {
+      "epoch": 0.7131737033589982,
+      "grad_norm": 0.21868008375167847,
+      "learning_rate": 0.001,
+      "loss": 2.9919,
+      "step": 16858
+    },
+    {
+      "epoch": 0.7132160081225146,
+      "grad_norm": 0.22265969216823578,
+      "learning_rate": 0.001,
+      "loss": 1.8722,
+      "step": 16859
+    },
+    {
+      "epoch": 0.7132583128860309,
+      "grad_norm": 0.9610694646835327,
+      "learning_rate": 0.001,
+      "loss": 1.7194,
+      "step": 16860
+    },
+    {
+      "epoch": 0.7133006176495473,
+      "grad_norm": 1.113793134689331,
+      "learning_rate": 0.001,
+      "loss": 2.0243,
+      "step": 16861
+    },
+    {
+      "epoch": 0.7133429224130637,
+      "grad_norm": 3.1779723167419434,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 16862
+    },
+    {
+      "epoch": 0.71338522717658,
+      "grad_norm": 0.1677677184343338,
+      "learning_rate": 0.001,
+      "loss": 2.1601,
+      "step": 16863
+    },
+    {
+      "epoch": 0.7134275319400964,
+      "grad_norm": 0.7339451313018799,
+      "learning_rate": 0.001,
+      "loss": 1.9814,
+      "step": 16864
+    },
+    {
+      "epoch": 0.7134698367036129,
+      "grad_norm": 0.4018705487251282,
+      "learning_rate": 0.001,
+      "loss": 2.3781,
+      "step": 16865
+    },
+    {
+      "epoch": 0.7135121414671292,
+      "grad_norm": 9.572480201721191,
+      "learning_rate": 0.001,
+      "loss": 1.9837,
+      "step": 16866
+    },
+    {
+      "epoch": 0.7135544462306456,
+      "grad_norm": 0.26376014947891235,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 16867
+    },
+    {
+      "epoch": 0.713596750994162,
+      "grad_norm": 1.0289099216461182,
+      "learning_rate": 0.001,
+      "loss": 2.5283,
+      "step": 16868
+    },
+    {
+      "epoch": 0.7136390557576783,
+      "grad_norm": 0.1844884753227234,
+      "learning_rate": 0.001,
+      "loss": 1.9274,
+      "step": 16869
+    },
+    {
+      "epoch": 0.7136813605211947,
+      "grad_norm": 0.19893215596675873,
+      "learning_rate": 0.001,
+      "loss": 2.1862,
+      "step": 16870
+    },
+    {
+      "epoch": 0.7137236652847111,
+      "grad_norm": 0.239218607544899,
+      "learning_rate": 0.001,
+      "loss": 1.6428,
+      "step": 16871
+    },
+    {
+      "epoch": 0.7137659700482274,
+      "grad_norm": 0.3456837832927704,
+      "learning_rate": 0.001,
+      "loss": 1.6866,
+      "step": 16872
+    },
+    {
+      "epoch": 0.7138082748117438,
+      "grad_norm": 0.20002485811710358,
+      "learning_rate": 0.001,
+      "loss": 2.0121,
+      "step": 16873
+    },
+    {
+      "epoch": 0.7138505795752602,
+      "grad_norm": 0.16409829258918762,
+      "learning_rate": 0.001,
+      "loss": 2.7127,
+      "step": 16874
+    },
+    {
+      "epoch": 0.7138928843387765,
+      "grad_norm": 0.2976275086402893,
+      "learning_rate": 0.001,
+      "loss": 2.683,
+      "step": 16875
+    },
+    {
+      "epoch": 0.7139351891022929,
+      "grad_norm": 0.1813412308692932,
+      "learning_rate": 0.001,
+      "loss": 1.4256,
+      "step": 16876
+    },
+    {
+      "epoch": 0.7139774938658093,
+      "grad_norm": 0.28270435333251953,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 16877
+    },
+    {
+      "epoch": 0.7140197986293256,
+      "grad_norm": 0.5542454719543457,
+      "learning_rate": 0.001,
+      "loss": 2.6197,
+      "step": 16878
+    },
+    {
+      "epoch": 0.714062103392842,
+      "grad_norm": 0.17437031865119934,
+      "learning_rate": 0.001,
+      "loss": 2.1137,
+      "step": 16879
+    },
+    {
+      "epoch": 0.7141044081563585,
+      "grad_norm": 0.3185507357120514,
+      "learning_rate": 0.001,
+      "loss": 2.0136,
+      "step": 16880
+    },
+    {
+      "epoch": 0.7141467129198747,
+      "grad_norm": 0.8088506460189819,
+      "learning_rate": 0.001,
+      "loss": 3.2072,
+      "step": 16881
+    },
+    {
+      "epoch": 0.7141890176833912,
+      "grad_norm": 0.15809717774391174,
+      "learning_rate": 0.001,
+      "loss": 1.428,
+      "step": 16882
+    },
+    {
+      "epoch": 0.7142313224469076,
+      "grad_norm": 0.16368800401687622,
+      "learning_rate": 0.001,
+      "loss": 2.1529,
+      "step": 16883
+    },
+    {
+      "epoch": 0.7142736272104239,
+      "grad_norm": 0.2843414843082428,
+      "learning_rate": 0.001,
+      "loss": 1.4441,
+      "step": 16884
+    },
+    {
+      "epoch": 0.7143159319739403,
+      "grad_norm": 0.20295199751853943,
+      "learning_rate": 0.001,
+      "loss": 2.4748,
+      "step": 16885
+    },
+    {
+      "epoch": 0.7143582367374567,
+      "grad_norm": 0.15212145447731018,
+      "learning_rate": 0.001,
+      "loss": 2.8044,
+      "step": 16886
+    },
+    {
+      "epoch": 0.714400541500973,
+      "grad_norm": 0.1592477262020111,
+      "learning_rate": 0.001,
+      "loss": 2.3869,
+      "step": 16887
+    },
+    {
+      "epoch": 0.7144428462644894,
+      "grad_norm": 0.18488934636116028,
+      "learning_rate": 0.001,
+      "loss": 2.8091,
+      "step": 16888
+    },
+    {
+      "epoch": 0.7144851510280058,
+      "grad_norm": 0.2524031698703766,
+      "learning_rate": 0.001,
+      "loss": 3.1258,
+      "step": 16889
+    },
+    {
+      "epoch": 0.7145274557915221,
+      "grad_norm": 0.19174166023731232,
+      "learning_rate": 0.001,
+      "loss": 2.3683,
+      "step": 16890
+    },
+    {
+      "epoch": 0.7145697605550385,
+      "grad_norm": 0.7119539380073547,
+      "learning_rate": 0.001,
+      "loss": 1.8931,
+      "step": 16891
+    },
+    {
+      "epoch": 0.7146120653185549,
+      "grad_norm": 0.14731153845787048,
+      "learning_rate": 0.001,
+      "loss": 3.5619,
+      "step": 16892
+    },
+    {
+      "epoch": 0.7146543700820712,
+      "grad_norm": 0.25515803694725037,
+      "learning_rate": 0.001,
+      "loss": 1.8279,
+      "step": 16893
+    },
+    {
+      "epoch": 0.7146966748455876,
+      "grad_norm": 0.1940464973449707,
+      "learning_rate": 0.001,
+      "loss": 2.2902,
+      "step": 16894
+    },
+    {
+      "epoch": 0.7147389796091039,
+      "grad_norm": 0.3229266405105591,
+      "learning_rate": 0.001,
+      "loss": 1.7167,
+      "step": 16895
+    },
+    {
+      "epoch": 0.7147812843726203,
+      "grad_norm": 0.166666179895401,
+      "learning_rate": 0.001,
+      "loss": 2.0199,
+      "step": 16896
+    },
+    {
+      "epoch": 0.7148235891361367,
+      "grad_norm": 0.14919742941856384,
+      "learning_rate": 0.001,
+      "loss": 1.6522,
+      "step": 16897
+    },
+    {
+      "epoch": 0.714865893899653,
+      "grad_norm": 0.1620715856552124,
+      "learning_rate": 0.001,
+      "loss": 2.5642,
+      "step": 16898
+    },
+    {
+      "epoch": 0.7149081986631695,
+      "grad_norm": 1.447795033454895,
+      "learning_rate": 0.001,
+      "loss": 2.2744,
+      "step": 16899
+    },
+    {
+      "epoch": 0.7149505034266859,
+      "grad_norm": 0.4487624764442444,
+      "learning_rate": 0.001,
+      "loss": 2.0419,
+      "step": 16900
+    },
+    {
+      "epoch": 0.7149928081902022,
+      "grad_norm": 0.21174852550029755,
+      "learning_rate": 0.001,
+      "loss": 3.2943,
+      "step": 16901
+    },
+    {
+      "epoch": 0.7150351129537186,
+      "grad_norm": 0.16627584397792816,
+      "learning_rate": 0.001,
+      "loss": 2.8095,
+      "step": 16902
+    },
+    {
+      "epoch": 0.715077417717235,
+      "grad_norm": 0.6215983629226685,
+      "learning_rate": 0.001,
+      "loss": 1.8961,
+      "step": 16903
+    },
+    {
+      "epoch": 0.7151197224807513,
+      "grad_norm": 0.3979867696762085,
+      "learning_rate": 0.001,
+      "loss": 2.1163,
+      "step": 16904
+    },
+    {
+      "epoch": 0.7151620272442677,
+      "grad_norm": 0.17654618620872498,
+      "learning_rate": 0.001,
+      "loss": 2.0656,
+      "step": 16905
+    },
+    {
+      "epoch": 0.7152043320077841,
+      "grad_norm": 0.15997301042079926,
+      "learning_rate": 0.001,
+      "loss": 1.4513,
+      "step": 16906
+    },
+    {
+      "epoch": 0.7152466367713004,
+      "grad_norm": 0.21495701372623444,
+      "learning_rate": 0.001,
+      "loss": 2.0596,
+      "step": 16907
+    },
+    {
+      "epoch": 0.7152889415348168,
+      "grad_norm": 0.3844386339187622,
+      "learning_rate": 0.001,
+      "loss": 1.82,
+      "step": 16908
+    },
+    {
+      "epoch": 0.7153312462983332,
+      "grad_norm": 1.2073756456375122,
+      "learning_rate": 0.001,
+      "loss": 1.7744,
+      "step": 16909
+    },
+    {
+      "epoch": 0.7153735510618495,
+      "grad_norm": 0.272298663854599,
+      "learning_rate": 0.001,
+      "loss": 3.1949,
+      "step": 16910
+    },
+    {
+      "epoch": 0.7154158558253659,
+      "grad_norm": 0.18787261843681335,
+      "learning_rate": 0.001,
+      "loss": 2.01,
+      "step": 16911
+    },
+    {
+      "epoch": 0.7154581605888823,
+      "grad_norm": 0.16972093284130096,
+      "learning_rate": 0.001,
+      "loss": 2.3674,
+      "step": 16912
+    },
+    {
+      "epoch": 0.7155004653523986,
+      "grad_norm": 0.20305274426937103,
+      "learning_rate": 0.001,
+      "loss": 2.1895,
+      "step": 16913
+    },
+    {
+      "epoch": 0.715542770115915,
+      "grad_norm": 0.16325430572032928,
+      "learning_rate": 0.001,
+      "loss": 1.6928,
+      "step": 16914
+    },
+    {
+      "epoch": 0.7155850748794315,
+      "grad_norm": 0.17961114645004272,
+      "learning_rate": 0.001,
+      "loss": 2.2898,
+      "step": 16915
+    },
+    {
+      "epoch": 0.7156273796429478,
+      "grad_norm": 0.15911836922168732,
+      "learning_rate": 0.001,
+      "loss": 2.2762,
+      "step": 16916
+    },
+    {
+      "epoch": 0.7156696844064642,
+      "grad_norm": 0.1571052074432373,
+      "learning_rate": 0.001,
+      "loss": 2.0235,
+      "step": 16917
+    },
+    {
+      "epoch": 0.7157119891699806,
+      "grad_norm": 0.17579875886440277,
+      "learning_rate": 0.001,
+      "loss": 2.8934,
+      "step": 16918
+    },
+    {
+      "epoch": 0.7157542939334969,
+      "grad_norm": 0.24277016520500183,
+      "learning_rate": 0.001,
+      "loss": 3.4171,
+      "step": 16919
+    },
+    {
+      "epoch": 0.7157965986970133,
+      "grad_norm": 1.3790236711502075,
+      "learning_rate": 0.001,
+      "loss": 3.0564,
+      "step": 16920
+    },
+    {
+      "epoch": 0.7158389034605297,
+      "grad_norm": 0.1847652643918991,
+      "learning_rate": 0.001,
+      "loss": 2.5091,
+      "step": 16921
+    },
+    {
+      "epoch": 0.715881208224046,
+      "grad_norm": 0.18262675404548645,
+      "learning_rate": 0.001,
+      "loss": 2.8952,
+      "step": 16922
+    },
+    {
+      "epoch": 0.7159235129875624,
+      "grad_norm": 0.5282437205314636,
+      "learning_rate": 0.001,
+      "loss": 2.3116,
+      "step": 16923
+    },
+    {
+      "epoch": 0.7159658177510788,
+      "grad_norm": 0.17553091049194336,
+      "learning_rate": 0.001,
+      "loss": 1.662,
+      "step": 16924
+    },
+    {
+      "epoch": 0.7160081225145951,
+      "grad_norm": 0.15575726330280304,
+      "learning_rate": 0.001,
+      "loss": 1.5341,
+      "step": 16925
+    },
+    {
+      "epoch": 0.7160504272781115,
+      "grad_norm": 0.1408965289592743,
+      "learning_rate": 0.001,
+      "loss": 1.2918,
+      "step": 16926
+    },
+    {
+      "epoch": 0.7160927320416279,
+      "grad_norm": 0.18358954787254333,
+      "learning_rate": 0.001,
+      "loss": 2.8717,
+      "step": 16927
+    },
+    {
+      "epoch": 0.7161350368051442,
+      "grad_norm": 0.15843123197555542,
+      "learning_rate": 0.001,
+      "loss": 2.5845,
+      "step": 16928
+    },
+    {
+      "epoch": 0.7161773415686606,
+      "grad_norm": 0.1516776978969574,
+      "learning_rate": 0.001,
+      "loss": 2.7124,
+      "step": 16929
+    },
+    {
+      "epoch": 0.716219646332177,
+      "grad_norm": 0.2799467146396637,
+      "learning_rate": 0.001,
+      "loss": 3.124,
+      "step": 16930
+    },
+    {
+      "epoch": 0.7162619510956933,
+      "grad_norm": 0.16782619059085846,
+      "learning_rate": 0.001,
+      "loss": 2.856,
+      "step": 16931
+    },
+    {
+      "epoch": 0.7163042558592098,
+      "grad_norm": 0.17226386070251465,
+      "learning_rate": 0.001,
+      "loss": 2.282,
+      "step": 16932
+    },
+    {
+      "epoch": 0.7163465606227262,
+      "grad_norm": 0.1673172563314438,
+      "learning_rate": 0.001,
+      "loss": 2.0755,
+      "step": 16933
+    },
+    {
+      "epoch": 0.7163888653862425,
+      "grad_norm": 0.20048165321350098,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 16934
+    },
+    {
+      "epoch": 0.7164311701497589,
+      "grad_norm": 0.3704041838645935,
+      "learning_rate": 0.001,
+      "loss": 1.8703,
+      "step": 16935
+    },
+    {
+      "epoch": 0.7164734749132753,
+      "grad_norm": 0.17238663136959076,
+      "learning_rate": 0.001,
+      "loss": 1.8328,
+      "step": 16936
+    },
+    {
+      "epoch": 0.7165157796767916,
+      "grad_norm": 0.5309063196182251,
+      "learning_rate": 0.001,
+      "loss": 2.7206,
+      "step": 16937
+    },
+    {
+      "epoch": 0.716558084440308,
+      "grad_norm": 0.28528085350990295,
+      "learning_rate": 0.001,
+      "loss": 1.8669,
+      "step": 16938
+    },
+    {
+      "epoch": 0.7166003892038244,
+      "grad_norm": 0.34017452597618103,
+      "learning_rate": 0.001,
+      "loss": 2.7997,
+      "step": 16939
+    },
+    {
+      "epoch": 0.7166426939673407,
+      "grad_norm": 0.19825142621994019,
+      "learning_rate": 0.001,
+      "loss": 3.2758,
+      "step": 16940
+    },
+    {
+      "epoch": 0.7166849987308571,
+      "grad_norm": 0.18345898389816284,
+      "learning_rate": 0.001,
+      "loss": 1.6755,
+      "step": 16941
+    },
+    {
+      "epoch": 0.7167273034943734,
+      "grad_norm": 0.15514928102493286,
+      "learning_rate": 0.001,
+      "loss": 1.7149,
+      "step": 16942
+    },
+    {
+      "epoch": 0.7167696082578898,
+      "grad_norm": 0.19425754249095917,
+      "learning_rate": 0.001,
+      "loss": 2.3292,
+      "step": 16943
+    },
+    {
+      "epoch": 0.7168119130214062,
+      "grad_norm": 0.21819639205932617,
+      "learning_rate": 0.001,
+      "loss": 2.918,
+      "step": 16944
+    },
+    {
+      "epoch": 0.7168542177849225,
+      "grad_norm": 0.16114571690559387,
+      "learning_rate": 0.001,
+      "loss": 1.7553,
+      "step": 16945
+    },
+    {
+      "epoch": 0.7168965225484389,
+      "grad_norm": 0.1726664900779724,
+      "learning_rate": 0.001,
+      "loss": 1.8447,
+      "step": 16946
+    },
+    {
+      "epoch": 0.7169388273119554,
+      "grad_norm": 0.17410875856876373,
+      "learning_rate": 0.001,
+      "loss": 1.649,
+      "step": 16947
+    },
+    {
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.17056071758270264,
+      "learning_rate": 0.001,
+      "loss": 2.7856,
+      "step": 16948
+    },
+    {
+      "epoch": 0.7170234368389881,
+      "grad_norm": 0.2786010801792145,
+      "learning_rate": 0.001,
+      "loss": 2.5549,
+      "step": 16949
+    },
+    {
+      "epoch": 0.7170657416025045,
+      "grad_norm": 3.383892774581909,
+      "learning_rate": 0.001,
+      "loss": 3.8689,
+      "step": 16950
+    },
+    {
+      "epoch": 0.7171080463660208,
+      "grad_norm": 0.8111663460731506,
+      "learning_rate": 0.001,
+      "loss": 1.7525,
+      "step": 16951
+    },
+    {
+      "epoch": 0.7171503511295372,
+      "grad_norm": 0.22883149981498718,
+      "learning_rate": 0.001,
+      "loss": 2.36,
+      "step": 16952
+    },
+    {
+      "epoch": 0.7171926558930536,
+      "grad_norm": 0.17972877621650696,
+      "learning_rate": 0.001,
+      "loss": 2.6088,
+      "step": 16953
+    },
+    {
+      "epoch": 0.7172349606565699,
+      "grad_norm": 0.2092818170785904,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 16954
+    },
+    {
+      "epoch": 0.7172772654200863,
+      "grad_norm": 0.16795043647289276,
+      "learning_rate": 0.001,
+      "loss": 2.887,
+      "step": 16955
+    },
+    {
+      "epoch": 0.7173195701836027,
+      "grad_norm": 0.16794079542160034,
+      "learning_rate": 0.001,
+      "loss": 2.2538,
+      "step": 16956
+    },
+    {
+      "epoch": 0.717361874947119,
+      "grad_norm": 0.1680205762386322,
+      "learning_rate": 0.001,
+      "loss": 1.7101,
+      "step": 16957
+    },
+    {
+      "epoch": 0.7174041797106354,
+      "grad_norm": 0.17401909828186035,
+      "learning_rate": 0.001,
+      "loss": 1.853,
+      "step": 16958
+    },
+    {
+      "epoch": 0.7174464844741518,
+      "grad_norm": 0.49490857124328613,
+      "learning_rate": 0.001,
+      "loss": 2.6701,
+      "step": 16959
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.17487838864326477,
+      "learning_rate": 0.001,
+      "loss": 1.2779,
+      "step": 16960
+    },
+    {
+      "epoch": 0.7175310940011845,
+      "grad_norm": 0.23202207684516907,
+      "learning_rate": 0.001,
+      "loss": 2.4539,
+      "step": 16961
+    },
+    {
+      "epoch": 0.7175733987647009,
+      "grad_norm": 0.4263302981853485,
+      "learning_rate": 0.001,
+      "loss": 1.7071,
+      "step": 16962
+    },
+    {
+      "epoch": 0.7176157035282172,
+      "grad_norm": 0.22260010242462158,
+      "learning_rate": 0.001,
+      "loss": 2.4392,
+      "step": 16963
+    },
+    {
+      "epoch": 0.7176580082917337,
+      "grad_norm": 0.15869098901748657,
+      "learning_rate": 0.001,
+      "loss": 1.7775,
+      "step": 16964
+    },
+    {
+      "epoch": 0.7177003130552501,
+      "grad_norm": 0.19737808406352997,
+      "learning_rate": 0.001,
+      "loss": 2.0257,
+      "step": 16965
+    },
+    {
+      "epoch": 0.7177426178187664,
+      "grad_norm": 0.15930262207984924,
+      "learning_rate": 0.001,
+      "loss": 3.0723,
+      "step": 16966
+    },
+    {
+      "epoch": 0.7177849225822828,
+      "grad_norm": 0.1594778150320053,
+      "learning_rate": 0.001,
+      "loss": 2.2068,
+      "step": 16967
+    },
+    {
+      "epoch": 0.7178272273457992,
+      "grad_norm": 0.17711618542671204,
+      "learning_rate": 0.001,
+      "loss": 1.7199,
+      "step": 16968
+    },
+    {
+      "epoch": 0.7178695321093155,
+      "grad_norm": 0.7712798118591309,
+      "learning_rate": 0.001,
+      "loss": 2.6924,
+      "step": 16969
+    },
+    {
+      "epoch": 0.7179118368728319,
+      "grad_norm": 0.16748692095279694,
+      "learning_rate": 0.001,
+      "loss": 2.1455,
+      "step": 16970
+    },
+    {
+      "epoch": 0.7179541416363483,
+      "grad_norm": 0.2054317146539688,
+      "learning_rate": 0.001,
+      "loss": 2.3641,
+      "step": 16971
+    },
+    {
+      "epoch": 0.7179964463998646,
+      "grad_norm": 1.4810477495193481,
+      "learning_rate": 0.001,
+      "loss": 2.5664,
+      "step": 16972
+    },
+    {
+      "epoch": 0.718038751163381,
+      "grad_norm": 0.3210020363330841,
+      "learning_rate": 0.001,
+      "loss": 2.4745,
+      "step": 16973
+    },
+    {
+      "epoch": 0.7180810559268974,
+      "grad_norm": 0.14637017250061035,
+      "learning_rate": 0.001,
+      "loss": 2.0775,
+      "step": 16974
+    },
+    {
+      "epoch": 0.7181233606904137,
+      "grad_norm": 0.1749821901321411,
+      "learning_rate": 0.001,
+      "loss": 2.4121,
+      "step": 16975
+    },
+    {
+      "epoch": 0.7181656654539301,
+      "grad_norm": 0.1919098049402237,
+      "learning_rate": 0.001,
+      "loss": 1.879,
+      "step": 16976
+    },
+    {
+      "epoch": 0.7182079702174465,
+      "grad_norm": 0.14838182926177979,
+      "learning_rate": 0.001,
+      "loss": 2.2794,
+      "step": 16977
+    },
+    {
+      "epoch": 0.7182502749809628,
+      "grad_norm": 0.3306111991405487,
+      "learning_rate": 0.001,
+      "loss": 1.7571,
+      "step": 16978
+    },
+    {
+      "epoch": 0.7182925797444792,
+      "grad_norm": 0.1290099322795868,
+      "learning_rate": 0.001,
+      "loss": 2.0743,
+      "step": 16979
+    },
+    {
+      "epoch": 0.7183348845079957,
+      "grad_norm": 0.2530765235424042,
+      "learning_rate": 0.001,
+      "loss": 1.92,
+      "step": 16980
+    },
+    {
+      "epoch": 0.718377189271512,
+      "grad_norm": 0.15586386620998383,
+      "learning_rate": 0.001,
+      "loss": 2.5149,
+      "step": 16981
+    },
+    {
+      "epoch": 0.7184194940350284,
+      "grad_norm": 0.4855019748210907,
+      "learning_rate": 0.001,
+      "loss": 1.9419,
+      "step": 16982
+    },
+    {
+      "epoch": 0.7184617987985448,
+      "grad_norm": 6.393121719360352,
+      "learning_rate": 0.001,
+      "loss": 1.7564,
+      "step": 16983
+    },
+    {
+      "epoch": 0.7185041035620611,
+      "grad_norm": 0.1833300143480301,
+      "learning_rate": 0.001,
+      "loss": 2.1363,
+      "step": 16984
+    },
+    {
+      "epoch": 0.7185464083255775,
+      "grad_norm": 0.1298752874135971,
+      "learning_rate": 0.001,
+      "loss": 2.2444,
+      "step": 16985
+    },
+    {
+      "epoch": 0.7185887130890938,
+      "grad_norm": 0.16792719066143036,
+      "learning_rate": 0.001,
+      "loss": 1.7757,
+      "step": 16986
+    },
+    {
+      "epoch": 0.7186310178526102,
+      "grad_norm": 1.4320098161697388,
+      "learning_rate": 0.001,
+      "loss": 2.647,
+      "step": 16987
+    },
+    {
+      "epoch": 0.7186733226161266,
+      "grad_norm": 0.13900575041770935,
+      "learning_rate": 0.001,
+      "loss": 1.6843,
+      "step": 16988
+    },
+    {
+      "epoch": 0.7187156273796429,
+      "grad_norm": 0.1738477498292923,
+      "learning_rate": 0.001,
+      "loss": 1.4185,
+      "step": 16989
+    },
+    {
+      "epoch": 0.7187579321431593,
+      "grad_norm": 0.5591545701026917,
+      "learning_rate": 0.001,
+      "loss": 2.4475,
+      "step": 16990
+    },
+    {
+      "epoch": 0.7188002369066757,
+      "grad_norm": 0.20126506686210632,
+      "learning_rate": 0.001,
+      "loss": 2.3001,
+      "step": 16991
+    },
+    {
+      "epoch": 0.718842541670192,
+      "grad_norm": 0.3284417986869812,
+      "learning_rate": 0.001,
+      "loss": 2.2221,
+      "step": 16992
+    },
+    {
+      "epoch": 0.7188848464337084,
+      "grad_norm": 0.1933998167514801,
+      "learning_rate": 0.001,
+      "loss": 1.9981,
+      "step": 16993
+    },
+    {
+      "epoch": 0.7189271511972248,
+      "grad_norm": 1.0560952425003052,
+      "learning_rate": 0.001,
+      "loss": 2.3462,
+      "step": 16994
+    },
+    {
+      "epoch": 0.7189694559607411,
+      "grad_norm": 0.16365966200828552,
+      "learning_rate": 0.001,
+      "loss": 2.6972,
+      "step": 16995
+    },
+    {
+      "epoch": 0.7190117607242575,
+      "grad_norm": 0.21969819068908691,
+      "learning_rate": 0.001,
+      "loss": 2.3659,
+      "step": 16996
+    },
+    {
+      "epoch": 0.719054065487774,
+      "grad_norm": 9.586769104003906,
+      "learning_rate": 0.001,
+      "loss": 1.9068,
+      "step": 16997
+    },
+    {
+      "epoch": 0.7190963702512903,
+      "grad_norm": 0.15527045726776123,
+      "learning_rate": 0.001,
+      "loss": 1.8663,
+      "step": 16998
+    },
+    {
+      "epoch": 0.7191386750148067,
+      "grad_norm": 0.25316473841667175,
+      "learning_rate": 0.001,
+      "loss": 2.7527,
+      "step": 16999
+    },
+    {
+      "epoch": 0.7191809797783231,
+      "grad_norm": 0.20590928196907043,
+      "learning_rate": 0.001,
+      "loss": 2.1561,
+      "step": 17000
+    },
+    {
+      "epoch": 0.7192232845418394,
+      "grad_norm": 0.23334155976772308,
+      "learning_rate": 0.001,
+      "loss": 1.6148,
+      "step": 17001
+    },
+    {
+      "epoch": 0.7192655893053558,
+      "grad_norm": 0.1720026433467865,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 17002
+    },
+    {
+      "epoch": 0.7193078940688722,
+      "grad_norm": 0.18993832170963287,
+      "learning_rate": 0.001,
+      "loss": 2.4176,
+      "step": 17003
+    },
+    {
+      "epoch": 0.7193501988323885,
+      "grad_norm": 1.2263435125350952,
+      "learning_rate": 0.001,
+      "loss": 1.695,
+      "step": 17004
+    },
+    {
+      "epoch": 0.7193925035959049,
+      "grad_norm": 0.20992887020111084,
+      "learning_rate": 0.001,
+      "loss": 3.1313,
+      "step": 17005
+    },
+    {
+      "epoch": 0.7194348083594213,
+      "grad_norm": 0.2623100280761719,
+      "learning_rate": 0.001,
+      "loss": 1.8112,
+      "step": 17006
+    },
+    {
+      "epoch": 0.7194771131229376,
+      "grad_norm": 0.2040976881980896,
+      "learning_rate": 0.001,
+      "loss": 2.8483,
+      "step": 17007
+    },
+    {
+      "epoch": 0.719519417886454,
+      "grad_norm": 0.1961153894662857,
+      "learning_rate": 0.001,
+      "loss": 3.994,
+      "step": 17008
+    },
+    {
+      "epoch": 0.7195617226499704,
+      "grad_norm": 0.1991744339466095,
+      "learning_rate": 0.001,
+      "loss": 1.9362,
+      "step": 17009
+    },
+    {
+      "epoch": 0.7196040274134867,
+      "grad_norm": 0.1773693561553955,
+      "learning_rate": 0.001,
+      "loss": 2.3754,
+      "step": 17010
+    },
+    {
+      "epoch": 0.7196463321770031,
+      "grad_norm": 0.322916716337204,
+      "learning_rate": 0.001,
+      "loss": 2.6735,
+      "step": 17011
+    },
+    {
+      "epoch": 0.7196886369405195,
+      "grad_norm": 0.1525036096572876,
+      "learning_rate": 0.001,
+      "loss": 2.265,
+      "step": 17012
+    },
+    {
+      "epoch": 0.7197309417040358,
+      "grad_norm": 0.15779347717761993,
+      "learning_rate": 0.001,
+      "loss": 1.6281,
+      "step": 17013
+    },
+    {
+      "epoch": 0.7197732464675523,
+      "grad_norm": 0.15959423780441284,
+      "learning_rate": 0.001,
+      "loss": 1.3413,
+      "step": 17014
+    },
+    {
+      "epoch": 0.7198155512310687,
+      "grad_norm": 0.3631104826927185,
+      "learning_rate": 0.001,
+      "loss": 1.6842,
+      "step": 17015
+    },
+    {
+      "epoch": 0.719857855994585,
+      "grad_norm": 0.474289208650589,
+      "learning_rate": 0.001,
+      "loss": 2.0346,
+      "step": 17016
+    },
+    {
+      "epoch": 0.7199001607581014,
+      "grad_norm": 0.19243498146533966,
+      "learning_rate": 0.001,
+      "loss": 2.3207,
+      "step": 17017
+    },
+    {
+      "epoch": 0.7199424655216178,
+      "grad_norm": 0.17244835197925568,
+      "learning_rate": 0.001,
+      "loss": 2.3529,
+      "step": 17018
+    },
+    {
+      "epoch": 0.7199847702851341,
+      "grad_norm": 0.16670437157154083,
+      "learning_rate": 0.001,
+      "loss": 2.4943,
+      "step": 17019
+    },
+    {
+      "epoch": 0.7200270750486505,
+      "grad_norm": 0.1707116663455963,
+      "learning_rate": 0.001,
+      "loss": 2.6172,
+      "step": 17020
+    },
+    {
+      "epoch": 0.7200693798121669,
+      "grad_norm": 0.17514576017856598,
+      "learning_rate": 0.001,
+      "loss": 1.9051,
+      "step": 17021
+    },
+    {
+      "epoch": 0.7201116845756832,
+      "grad_norm": 0.17995479702949524,
+      "learning_rate": 0.001,
+      "loss": 2.3423,
+      "step": 17022
+    },
+    {
+      "epoch": 0.7201539893391996,
+      "grad_norm": 0.1695273518562317,
+      "learning_rate": 0.001,
+      "loss": 1.9609,
+      "step": 17023
+    },
+    {
+      "epoch": 0.720196294102716,
+      "grad_norm": 0.18186168372631073,
+      "learning_rate": 0.001,
+      "loss": 1.8993,
+      "step": 17024
+    },
+    {
+      "epoch": 0.7202385988662323,
+      "grad_norm": 0.17733162641525269,
+      "learning_rate": 0.001,
+      "loss": 2.2507,
+      "step": 17025
+    },
+    {
+      "epoch": 0.7202809036297487,
+      "grad_norm": 0.17569288611412048,
+      "learning_rate": 0.001,
+      "loss": 2.4908,
+      "step": 17026
+    },
+    {
+      "epoch": 0.7203232083932651,
+      "grad_norm": 0.1461903601884842,
+      "learning_rate": 0.001,
+      "loss": 2.3501,
+      "step": 17027
+    },
+    {
+      "epoch": 0.7203655131567814,
+      "grad_norm": 0.13951033353805542,
+      "learning_rate": 0.001,
+      "loss": 1.5234,
+      "step": 17028
+    },
+    {
+      "epoch": 0.7204078179202978,
+      "grad_norm": 0.19020099937915802,
+      "learning_rate": 0.001,
+      "loss": 2.583,
+      "step": 17029
+    },
+    {
+      "epoch": 0.7204501226838141,
+      "grad_norm": 0.26185300946235657,
+      "learning_rate": 0.001,
+      "loss": 2.0379,
+      "step": 17030
+    },
+    {
+      "epoch": 0.7204924274473306,
+      "grad_norm": 0.15246087312698364,
+      "learning_rate": 0.001,
+      "loss": 1.6983,
+      "step": 17031
+    },
+    {
+      "epoch": 0.720534732210847,
+      "grad_norm": 0.21648335456848145,
+      "learning_rate": 0.001,
+      "loss": 2.4807,
+      "step": 17032
+    },
+    {
+      "epoch": 0.7205770369743633,
+      "grad_norm": 0.15605172514915466,
+      "learning_rate": 0.001,
+      "loss": 1.7671,
+      "step": 17033
+    },
+    {
+      "epoch": 0.7206193417378797,
+      "grad_norm": 0.1461627036333084,
+      "learning_rate": 0.001,
+      "loss": 1.6648,
+      "step": 17034
+    },
+    {
+      "epoch": 0.7206616465013961,
+      "grad_norm": 0.15967121720314026,
+      "learning_rate": 0.001,
+      "loss": 2.1308,
+      "step": 17035
+    },
+    {
+      "epoch": 0.7207039512649124,
+      "grad_norm": 0.17863650619983673,
+      "learning_rate": 0.001,
+      "loss": 2.646,
+      "step": 17036
+    },
+    {
+      "epoch": 0.7207462560284288,
+      "grad_norm": 0.15027737617492676,
+      "learning_rate": 0.001,
+      "loss": 2.302,
+      "step": 17037
+    },
+    {
+      "epoch": 0.7207885607919452,
+      "grad_norm": 0.30569708347320557,
+      "learning_rate": 0.001,
+      "loss": 3.3331,
+      "step": 17038
+    },
+    {
+      "epoch": 0.7208308655554615,
+      "grad_norm": 0.18736647069454193,
+      "learning_rate": 0.001,
+      "loss": 1.5433,
+      "step": 17039
+    },
+    {
+      "epoch": 0.7208731703189779,
+      "grad_norm": 0.16825266182422638,
+      "learning_rate": 0.001,
+      "loss": 1.511,
+      "step": 17040
+    },
+    {
+      "epoch": 0.7209154750824943,
+      "grad_norm": 0.1982458382844925,
+      "learning_rate": 0.001,
+      "loss": 2.066,
+      "step": 17041
+    },
+    {
+      "epoch": 0.7209577798460106,
+      "grad_norm": 0.1793079972267151,
+      "learning_rate": 0.001,
+      "loss": 1.6284,
+      "step": 17042
+    },
+    {
+      "epoch": 0.721000084609527,
+      "grad_norm": 0.1480327844619751,
+      "learning_rate": 0.001,
+      "loss": 1.4117,
+      "step": 17043
+    },
+    {
+      "epoch": 0.7210423893730434,
+      "grad_norm": 0.3210713267326355,
+      "learning_rate": 0.001,
+      "loss": 1.739,
+      "step": 17044
+    },
+    {
+      "epoch": 0.7210846941365597,
+      "grad_norm": 0.2352566421031952,
+      "learning_rate": 0.001,
+      "loss": 1.7573,
+      "step": 17045
+    },
+    {
+      "epoch": 0.7211269989000761,
+      "grad_norm": 0.1587694138288498,
+      "learning_rate": 0.001,
+      "loss": 2.3439,
+      "step": 17046
+    },
+    {
+      "epoch": 0.7211693036635926,
+      "grad_norm": 0.9405317306518555,
+      "learning_rate": 0.001,
+      "loss": 1.9967,
+      "step": 17047
+    },
+    {
+      "epoch": 0.7212116084271089,
+      "grad_norm": 0.1904739886522293,
+      "learning_rate": 0.001,
+      "loss": 1.9261,
+      "step": 17048
+    },
+    {
+      "epoch": 0.7212539131906253,
+      "grad_norm": 0.12879721820354462,
+      "learning_rate": 0.001,
+      "loss": 2.6484,
+      "step": 17049
+    },
+    {
+      "epoch": 0.7212962179541417,
+      "grad_norm": 0.17692124843597412,
+      "learning_rate": 0.001,
+      "loss": 2.0815,
+      "step": 17050
+    },
+    {
+      "epoch": 0.721338522717658,
+      "grad_norm": 0.16509714722633362,
+      "learning_rate": 0.001,
+      "loss": 2.9258,
+      "step": 17051
+    },
+    {
+      "epoch": 0.7213808274811744,
+      "grad_norm": 0.1631636917591095,
+      "learning_rate": 0.001,
+      "loss": 2.8743,
+      "step": 17052
+    },
+    {
+      "epoch": 0.7214231322446908,
+      "grad_norm": 0.174580916762352,
+      "learning_rate": 0.001,
+      "loss": 3.0192,
+      "step": 17053
+    },
+    {
+      "epoch": 0.7214654370082071,
+      "grad_norm": 0.8501453399658203,
+      "learning_rate": 0.001,
+      "loss": 1.8267,
+      "step": 17054
+    },
+    {
+      "epoch": 0.7215077417717235,
+      "grad_norm": 0.15166904032230377,
+      "learning_rate": 0.001,
+      "loss": 2.4201,
+      "step": 17055
+    },
+    {
+      "epoch": 0.7215500465352399,
+      "grad_norm": 0.16179314255714417,
+      "learning_rate": 0.001,
+      "loss": 1.9094,
+      "step": 17056
+    },
+    {
+      "epoch": 0.7215923512987562,
+      "grad_norm": 0.14122450351715088,
+      "learning_rate": 0.001,
+      "loss": 1.327,
+      "step": 17057
+    },
+    {
+      "epoch": 0.7216346560622726,
+      "grad_norm": 0.16468989849090576,
+      "learning_rate": 0.001,
+      "loss": 1.9848,
+      "step": 17058
+    },
+    {
+      "epoch": 0.721676960825789,
+      "grad_norm": 0.14296945929527283,
+      "learning_rate": 0.001,
+      "loss": 1.6847,
+      "step": 17059
+    },
+    {
+      "epoch": 0.7217192655893053,
+      "grad_norm": 0.1540268510580063,
+      "learning_rate": 0.001,
+      "loss": 1.4727,
+      "step": 17060
+    },
+    {
+      "epoch": 0.7217615703528217,
+      "grad_norm": 0.14178214967250824,
+      "learning_rate": 0.001,
+      "loss": 1.8472,
+      "step": 17061
+    },
+    {
+      "epoch": 0.7218038751163381,
+      "grad_norm": 0.1965574473142624,
+      "learning_rate": 0.001,
+      "loss": 2.9151,
+      "step": 17062
+    },
+    {
+      "epoch": 0.7218461798798544,
+      "grad_norm": 0.1626625508069992,
+      "learning_rate": 0.001,
+      "loss": 1.8456,
+      "step": 17063
+    },
+    {
+      "epoch": 0.7218884846433709,
+      "grad_norm": 1.8781062364578247,
+      "learning_rate": 0.001,
+      "loss": 2.1349,
+      "step": 17064
+    },
+    {
+      "epoch": 0.7219307894068873,
+      "grad_norm": 0.1811523288488388,
+      "learning_rate": 0.001,
+      "loss": 3.1633,
+      "step": 17065
+    },
+    {
+      "epoch": 0.7219730941704036,
+      "grad_norm": 0.1420523077249527,
+      "learning_rate": 0.001,
+      "loss": 2.1598,
+      "step": 17066
+    },
+    {
+      "epoch": 0.72201539893392,
+      "grad_norm": 0.23266121745109558,
+      "learning_rate": 0.001,
+      "loss": 2.5496,
+      "step": 17067
+    },
+    {
+      "epoch": 0.7220577036974364,
+      "grad_norm": 0.15366718173027039,
+      "learning_rate": 0.001,
+      "loss": 1.6903,
+      "step": 17068
+    },
+    {
+      "epoch": 0.7221000084609527,
+      "grad_norm": 0.1720408946275711,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 17069
+    },
+    {
+      "epoch": 0.7221423132244691,
+      "grad_norm": 0.13053718209266663,
+      "learning_rate": 0.001,
+      "loss": 2.9885,
+      "step": 17070
+    },
+    {
+      "epoch": 0.7221846179879855,
+      "grad_norm": 0.1682283878326416,
+      "learning_rate": 0.001,
+      "loss": 2.0249,
+      "step": 17071
+    },
+    {
+      "epoch": 0.7222269227515018,
+      "grad_norm": 0.1607704609632492,
+      "learning_rate": 0.001,
+      "loss": 2.3205,
+      "step": 17072
+    },
+    {
+      "epoch": 0.7222692275150182,
+      "grad_norm": 0.14773066341876984,
+      "learning_rate": 0.001,
+      "loss": 2.3242,
+      "step": 17073
+    },
+    {
+      "epoch": 0.7223115322785346,
+      "grad_norm": 2.609527349472046,
+      "learning_rate": 0.001,
+      "loss": 1.6714,
+      "step": 17074
+    },
+    {
+      "epoch": 0.7223538370420509,
+      "grad_norm": 0.22098509967327118,
+      "learning_rate": 0.001,
+      "loss": 2.6751,
+      "step": 17075
+    },
+    {
+      "epoch": 0.7223961418055673,
+      "grad_norm": 0.27444925904273987,
+      "learning_rate": 0.001,
+      "loss": 2.977,
+      "step": 17076
+    },
+    {
+      "epoch": 0.7224384465690836,
+      "grad_norm": 1.863974690437317,
+      "learning_rate": 0.001,
+      "loss": 1.914,
+      "step": 17077
+    },
+    {
+      "epoch": 0.7224807513326,
+      "grad_norm": 0.18942175805568695,
+      "learning_rate": 0.001,
+      "loss": 1.7231,
+      "step": 17078
+    },
+    {
+      "epoch": 0.7225230560961164,
+      "grad_norm": 20.305295944213867,
+      "learning_rate": 0.001,
+      "loss": 1.9652,
+      "step": 17079
+    },
+    {
+      "epoch": 0.7225653608596327,
+      "grad_norm": 1.116580843925476,
+      "learning_rate": 0.001,
+      "loss": 2.388,
+      "step": 17080
+    },
+    {
+      "epoch": 0.7226076656231492,
+      "grad_norm": 0.2017759084701538,
+      "learning_rate": 0.001,
+      "loss": 2.4368,
+      "step": 17081
+    },
+    {
+      "epoch": 0.7226499703866656,
+      "grad_norm": 0.20675040781497955,
+      "learning_rate": 0.001,
+      "loss": 1.8265,
+      "step": 17082
+    },
+    {
+      "epoch": 0.7226922751501819,
+      "grad_norm": 0.2491598129272461,
+      "learning_rate": 0.001,
+      "loss": 1.6281,
+      "step": 17083
+    },
+    {
+      "epoch": 0.7227345799136983,
+      "grad_norm": 0.7096131443977356,
+      "learning_rate": 0.001,
+      "loss": 2.6187,
+      "step": 17084
+    },
+    {
+      "epoch": 0.7227768846772147,
+      "grad_norm": 0.5838418006896973,
+      "learning_rate": 0.001,
+      "loss": 3.2109,
+      "step": 17085
+    },
+    {
+      "epoch": 0.722819189440731,
+      "grad_norm": 0.2073265165090561,
+      "learning_rate": 0.001,
+      "loss": 1.8221,
+      "step": 17086
+    },
+    {
+      "epoch": 0.7228614942042474,
+      "grad_norm": 0.13714540004730225,
+      "learning_rate": 0.001,
+      "loss": 2.1434,
+      "step": 17087
+    },
+    {
+      "epoch": 0.7229037989677638,
+      "grad_norm": 0.16703394055366516,
+      "learning_rate": 0.001,
+      "loss": 2.1473,
+      "step": 17088
+    },
+    {
+      "epoch": 0.7229461037312801,
+      "grad_norm": 1.2047275304794312,
+      "learning_rate": 0.001,
+      "loss": 2.9204,
+      "step": 17089
+    },
+    {
+      "epoch": 0.7229884084947965,
+      "grad_norm": 0.22018975019454956,
+      "learning_rate": 0.001,
+      "loss": 3.5448,
+      "step": 17090
+    },
+    {
+      "epoch": 0.7230307132583129,
+      "grad_norm": 0.4262462854385376,
+      "learning_rate": 0.001,
+      "loss": 2.1982,
+      "step": 17091
+    },
+    {
+      "epoch": 0.7230730180218292,
+      "grad_norm": 0.1653733253479004,
+      "learning_rate": 0.001,
+      "loss": 3.1182,
+      "step": 17092
+    },
+    {
+      "epoch": 0.7231153227853456,
+      "grad_norm": 0.20591098070144653,
+      "learning_rate": 0.001,
+      "loss": 3.3591,
+      "step": 17093
+    },
+    {
+      "epoch": 0.723157627548862,
+      "grad_norm": 0.177374005317688,
+      "learning_rate": 0.001,
+      "loss": 2.7171,
+      "step": 17094
+    },
+    {
+      "epoch": 0.7231999323123783,
+      "grad_norm": 0.17736870050430298,
+      "learning_rate": 0.001,
+      "loss": 1.6567,
+      "step": 17095
+    },
+    {
+      "epoch": 0.7232422370758947,
+      "grad_norm": 0.16085311770439148,
+      "learning_rate": 0.001,
+      "loss": 1.6649,
+      "step": 17096
+    },
+    {
+      "epoch": 0.7232845418394112,
+      "grad_norm": 0.1982753872871399,
+      "learning_rate": 0.001,
+      "loss": 1.6603,
+      "step": 17097
+    },
+    {
+      "epoch": 0.7233268466029275,
+      "grad_norm": 0.1861245036125183,
+      "learning_rate": 0.001,
+      "loss": 2.91,
+      "step": 17098
+    },
+    {
+      "epoch": 0.7233691513664439,
+      "grad_norm": 0.7297300100326538,
+      "learning_rate": 0.001,
+      "loss": 1.8701,
+      "step": 17099
+    },
+    {
+      "epoch": 0.7234114561299603,
+      "grad_norm": 0.16953283548355103,
+      "learning_rate": 0.001,
+      "loss": 3.1791,
+      "step": 17100
+    },
+    {
+      "epoch": 0.7234537608934766,
+      "grad_norm": 0.16749191284179688,
+      "learning_rate": 0.001,
+      "loss": 2.5591,
+      "step": 17101
+    },
+    {
+      "epoch": 0.723496065656993,
+      "grad_norm": 0.19384531676769257,
+      "learning_rate": 0.001,
+      "loss": 2.2332,
+      "step": 17102
+    },
+    {
+      "epoch": 0.7235383704205094,
+      "grad_norm": 2.1700196266174316,
+      "learning_rate": 0.001,
+      "loss": 3.0336,
+      "step": 17103
+    },
+    {
+      "epoch": 0.7235806751840257,
+      "grad_norm": 0.1668250858783722,
+      "learning_rate": 0.001,
+      "loss": 2.2885,
+      "step": 17104
+    },
+    {
+      "epoch": 0.7236229799475421,
+      "grad_norm": 0.17841757833957672,
+      "learning_rate": 0.001,
+      "loss": 2.3188,
+      "step": 17105
+    },
+    {
+      "epoch": 0.7236652847110585,
+      "grad_norm": 0.15527507662773132,
+      "learning_rate": 0.001,
+      "loss": 1.8841,
+      "step": 17106
+    },
+    {
+      "epoch": 0.7237075894745748,
+      "grad_norm": 0.5220174789428711,
+      "learning_rate": 0.001,
+      "loss": 1.6185,
+      "step": 17107
+    },
+    {
+      "epoch": 0.7237498942380912,
+      "grad_norm": 0.14695467054843903,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 17108
+    },
+    {
+      "epoch": 0.7237921990016076,
+      "grad_norm": 0.39175304770469666,
+      "learning_rate": 0.001,
+      "loss": 1.7712,
+      "step": 17109
+    },
+    {
+      "epoch": 0.7238345037651239,
+      "grad_norm": 0.15665383636951447,
+      "learning_rate": 0.001,
+      "loss": 2.5027,
+      "step": 17110
+    },
+    {
+      "epoch": 0.7238768085286403,
+      "grad_norm": 0.17397943139076233,
+      "learning_rate": 0.001,
+      "loss": 1.9146,
+      "step": 17111
+    },
+    {
+      "epoch": 0.7239191132921567,
+      "grad_norm": 0.18213899433612823,
+      "learning_rate": 0.001,
+      "loss": 2.6108,
+      "step": 17112
+    },
+    {
+      "epoch": 0.723961418055673,
+      "grad_norm": 0.1579909473657608,
+      "learning_rate": 0.001,
+      "loss": 1.4951,
+      "step": 17113
+    },
+    {
+      "epoch": 0.7240037228191895,
+      "grad_norm": 0.15795019268989563,
+      "learning_rate": 0.001,
+      "loss": 1.5673,
+      "step": 17114
+    },
+    {
+      "epoch": 0.7240460275827059,
+      "grad_norm": 0.24189157783985138,
+      "learning_rate": 0.001,
+      "loss": 3.2042,
+      "step": 17115
+    },
+    {
+      "epoch": 0.7240883323462222,
+      "grad_norm": 0.1958611160516739,
+      "learning_rate": 0.001,
+      "loss": 2.8201,
+      "step": 17116
+    },
+    {
+      "epoch": 0.7241306371097386,
+      "grad_norm": 0.18128181993961334,
+      "learning_rate": 0.001,
+      "loss": 2.0228,
+      "step": 17117
+    },
+    {
+      "epoch": 0.724172941873255,
+      "grad_norm": 0.1512492448091507,
+      "learning_rate": 0.001,
+      "loss": 2.3753,
+      "step": 17118
+    },
+    {
+      "epoch": 0.7242152466367713,
+      "grad_norm": 0.15748295187950134,
+      "learning_rate": 0.001,
+      "loss": 1.9492,
+      "step": 17119
+    },
+    {
+      "epoch": 0.7242575514002877,
+      "grad_norm": 15.251922607421875,
+      "learning_rate": 0.001,
+      "loss": 3.1576,
+      "step": 17120
+    },
+    {
+      "epoch": 0.724299856163804,
+      "grad_norm": 0.17477959394454956,
+      "learning_rate": 0.001,
+      "loss": 1.827,
+      "step": 17121
+    },
+    {
+      "epoch": 0.7243421609273204,
+      "grad_norm": 0.16635996103286743,
+      "learning_rate": 0.001,
+      "loss": 2.5203,
+      "step": 17122
+    },
+    {
+      "epoch": 0.7243844656908368,
+      "grad_norm": 0.12794765830039978,
+      "learning_rate": 0.001,
+      "loss": 2.4801,
+      "step": 17123
+    },
+    {
+      "epoch": 0.7244267704543531,
+      "grad_norm": 0.1665220856666565,
+      "learning_rate": 0.001,
+      "loss": 1.6915,
+      "step": 17124
+    },
+    {
+      "epoch": 0.7244690752178695,
+      "grad_norm": 0.20942756533622742,
+      "learning_rate": 0.001,
+      "loss": 2.1367,
+      "step": 17125
+    },
+    {
+      "epoch": 0.7245113799813859,
+      "grad_norm": 0.1444472074508667,
+      "learning_rate": 0.001,
+      "loss": 1.4025,
+      "step": 17126
+    },
+    {
+      "epoch": 0.7245536847449022,
+      "grad_norm": 0.5599656105041504,
+      "learning_rate": 0.001,
+      "loss": 1.6415,
+      "step": 17127
+    },
+    {
+      "epoch": 0.7245959895084186,
+      "grad_norm": 0.17753274738788605,
+      "learning_rate": 0.001,
+      "loss": 2.3513,
+      "step": 17128
+    },
+    {
+      "epoch": 0.724638294271935,
+      "grad_norm": 0.18549706041812897,
+      "learning_rate": 0.001,
+      "loss": 2.4151,
+      "step": 17129
+    },
+    {
+      "epoch": 0.7246805990354513,
+      "grad_norm": 0.17280645668506622,
+      "learning_rate": 0.001,
+      "loss": 2.5004,
+      "step": 17130
+    },
+    {
+      "epoch": 0.7247229037989678,
+      "grad_norm": 0.16694018244743347,
+      "learning_rate": 0.001,
+      "loss": 2.1255,
+      "step": 17131
+    },
+    {
+      "epoch": 0.7247652085624842,
+      "grad_norm": 0.1705455332994461,
+      "learning_rate": 0.001,
+      "loss": 1.967,
+      "step": 17132
+    },
+    {
+      "epoch": 0.7248075133260005,
+      "grad_norm": 0.1580754816532135,
+      "learning_rate": 0.001,
+      "loss": 1.9559,
+      "step": 17133
+    },
+    {
+      "epoch": 0.7248498180895169,
+      "grad_norm": 0.9335688352584839,
+      "learning_rate": 0.001,
+      "loss": 2.4301,
+      "step": 17134
+    },
+    {
+      "epoch": 0.7248921228530333,
+      "grad_norm": 0.1776943802833557,
+      "learning_rate": 0.001,
+      "loss": 1.4292,
+      "step": 17135
+    },
+    {
+      "epoch": 0.7249344276165496,
+      "grad_norm": 0.15362563729286194,
+      "learning_rate": 0.001,
+      "loss": 2.3276,
+      "step": 17136
+    },
+    {
+      "epoch": 0.724976732380066,
+      "grad_norm": 0.14412017166614532,
+      "learning_rate": 0.001,
+      "loss": 1.8388,
+      "step": 17137
+    },
+    {
+      "epoch": 0.7250190371435824,
+      "grad_norm": 0.2188391387462616,
+      "learning_rate": 0.001,
+      "loss": 2.7813,
+      "step": 17138
+    },
+    {
+      "epoch": 0.7250613419070987,
+      "grad_norm": 0.18921682238578796,
+      "learning_rate": 0.001,
+      "loss": 2.1525,
+      "step": 17139
+    },
+    {
+      "epoch": 0.7251036466706151,
+      "grad_norm": 0.1472938060760498,
+      "learning_rate": 0.001,
+      "loss": 2.2099,
+      "step": 17140
+    },
+    {
+      "epoch": 0.7251459514341315,
+      "grad_norm": 0.1493636518716812,
+      "learning_rate": 0.001,
+      "loss": 2.0206,
+      "step": 17141
+    },
+    {
+      "epoch": 0.7251882561976478,
+      "grad_norm": 0.16733522713184357,
+      "learning_rate": 0.001,
+      "loss": 2.0764,
+      "step": 17142
+    },
+    {
+      "epoch": 0.7252305609611642,
+      "grad_norm": 0.18534940481185913,
+      "learning_rate": 0.001,
+      "loss": 2.0909,
+      "step": 17143
+    },
+    {
+      "epoch": 0.7252728657246806,
+      "grad_norm": 0.25004249811172485,
+      "learning_rate": 0.001,
+      "loss": 1.9715,
+      "step": 17144
+    },
+    {
+      "epoch": 0.7253151704881969,
+      "grad_norm": 0.1925039291381836,
+      "learning_rate": 0.001,
+      "loss": 1.8828,
+      "step": 17145
+    },
+    {
+      "epoch": 0.7253574752517133,
+      "grad_norm": 0.19968828558921814,
+      "learning_rate": 0.001,
+      "loss": 1.8314,
+      "step": 17146
+    },
+    {
+      "epoch": 0.7253997800152298,
+      "grad_norm": 0.19976693391799927,
+      "learning_rate": 0.001,
+      "loss": 2.0041,
+      "step": 17147
+    },
+    {
+      "epoch": 0.725442084778746,
+      "grad_norm": 0.16204243898391724,
+      "learning_rate": 0.001,
+      "loss": 1.5418,
+      "step": 17148
+    },
+    {
+      "epoch": 0.7254843895422625,
+      "grad_norm": 0.6488047242164612,
+      "learning_rate": 0.001,
+      "loss": 1.5626,
+      "step": 17149
+    },
+    {
+      "epoch": 0.7255266943057789,
+      "grad_norm": 0.16644714772701263,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 17150
+    },
+    {
+      "epoch": 0.7255689990692952,
+      "grad_norm": 0.16991443932056427,
+      "learning_rate": 0.001,
+      "loss": 1.7933,
+      "step": 17151
+    },
+    {
+      "epoch": 0.7256113038328116,
+      "grad_norm": 0.23698163032531738,
+      "learning_rate": 0.001,
+      "loss": 2.3938,
+      "step": 17152
+    },
+    {
+      "epoch": 0.725653608596328,
+      "grad_norm": 1.1459918022155762,
+      "learning_rate": 0.001,
+      "loss": 2.0805,
+      "step": 17153
+    },
+    {
+      "epoch": 0.7256959133598443,
+      "grad_norm": 0.19081348180770874,
+      "learning_rate": 0.001,
+      "loss": 1.9892,
+      "step": 17154
+    },
+    {
+      "epoch": 0.7257382181233607,
+      "grad_norm": 0.20583724975585938,
+      "learning_rate": 0.001,
+      "loss": 1.9256,
+      "step": 17155
+    },
+    {
+      "epoch": 0.7257805228868771,
+      "grad_norm": 0.15794368088245392,
+      "learning_rate": 0.001,
+      "loss": 1.7678,
+      "step": 17156
+    },
+    {
+      "epoch": 0.7258228276503934,
+      "grad_norm": 0.22491112351417542,
+      "learning_rate": 0.001,
+      "loss": 2.1716,
+      "step": 17157
+    },
+    {
+      "epoch": 0.7258651324139098,
+      "grad_norm": 0.37008705735206604,
+      "learning_rate": 0.001,
+      "loss": 3.8031,
+      "step": 17158
+    },
+    {
+      "epoch": 0.7259074371774262,
+      "grad_norm": 0.2206166684627533,
+      "learning_rate": 0.001,
+      "loss": 2.6122,
+      "step": 17159
+    },
+    {
+      "epoch": 0.7259497419409425,
+      "grad_norm": 0.1652143895626068,
+      "learning_rate": 0.001,
+      "loss": 2.2458,
+      "step": 17160
+    },
+    {
+      "epoch": 0.7259920467044589,
+      "grad_norm": 0.16952086985111237,
+      "learning_rate": 0.001,
+      "loss": 1.978,
+      "step": 17161
+    },
+    {
+      "epoch": 0.7260343514679753,
+      "grad_norm": 0.15790002048015594,
+      "learning_rate": 0.001,
+      "loss": 1.9818,
+      "step": 17162
+    },
+    {
+      "epoch": 0.7260766562314916,
+      "grad_norm": 0.18033292889595032,
+      "learning_rate": 0.001,
+      "loss": 2.0203,
+      "step": 17163
+    },
+    {
+      "epoch": 0.7261189609950081,
+      "grad_norm": 0.19904761016368866,
+      "learning_rate": 0.001,
+      "loss": 3.5056,
+      "step": 17164
+    },
+    {
+      "epoch": 0.7261612657585244,
+      "grad_norm": 0.18638011813163757,
+      "learning_rate": 0.001,
+      "loss": 2.5683,
+      "step": 17165
+    },
+    {
+      "epoch": 0.7262035705220408,
+      "grad_norm": 0.1429300308227539,
+      "learning_rate": 0.001,
+      "loss": 1.5911,
+      "step": 17166
+    },
+    {
+      "epoch": 0.7262458752855572,
+      "grad_norm": 0.18004530668258667,
+      "learning_rate": 0.001,
+      "loss": 2.3988,
+      "step": 17167
+    },
+    {
+      "epoch": 0.7262881800490735,
+      "grad_norm": 0.16043299436569214,
+      "learning_rate": 0.001,
+      "loss": 1.6417,
+      "step": 17168
+    },
+    {
+      "epoch": 0.7263304848125899,
+      "grad_norm": 0.15215149521827698,
+      "learning_rate": 0.001,
+      "loss": 2.1354,
+      "step": 17169
+    },
+    {
+      "epoch": 0.7263727895761063,
+      "grad_norm": 0.2153785228729248,
+      "learning_rate": 0.001,
+      "loss": 3.3095,
+      "step": 17170
+    },
+    {
+      "epoch": 0.7264150943396226,
+      "grad_norm": 0.17379812896251678,
+      "learning_rate": 0.001,
+      "loss": 1.6931,
+      "step": 17171
+    },
+    {
+      "epoch": 0.726457399103139,
+      "grad_norm": 0.14409448206424713,
+      "learning_rate": 0.001,
+      "loss": 1.8867,
+      "step": 17172
+    },
+    {
+      "epoch": 0.7264997038666554,
+      "grad_norm": 0.16144266724586487,
+      "learning_rate": 0.001,
+      "loss": 1.6323,
+      "step": 17173
+    },
+    {
+      "epoch": 0.7265420086301717,
+      "grad_norm": 0.16007545590400696,
+      "learning_rate": 0.001,
+      "loss": 1.6542,
+      "step": 17174
+    },
+    {
+      "epoch": 0.7265843133936881,
+      "grad_norm": 0.12861347198486328,
+      "learning_rate": 0.001,
+      "loss": 1.838,
+      "step": 17175
+    },
+    {
+      "epoch": 0.7266266181572045,
+      "grad_norm": 0.1386764645576477,
+      "learning_rate": 0.001,
+      "loss": 1.7484,
+      "step": 17176
+    },
+    {
+      "epoch": 0.7266689229207208,
+      "grad_norm": 0.18406988680362701,
+      "learning_rate": 0.001,
+      "loss": 2.2947,
+      "step": 17177
+    },
+    {
+      "epoch": 0.7267112276842372,
+      "grad_norm": 0.1446983367204666,
+      "learning_rate": 0.001,
+      "loss": 2.2782,
+      "step": 17178
+    },
+    {
+      "epoch": 0.7267535324477536,
+      "grad_norm": 0.17378035187721252,
+      "learning_rate": 0.001,
+      "loss": 4.0799,
+      "step": 17179
+    },
+    {
+      "epoch": 0.72679583721127,
+      "grad_norm": 0.5141298174858093,
+      "learning_rate": 0.001,
+      "loss": 1.748,
+      "step": 17180
+    },
+    {
+      "epoch": 0.7268381419747864,
+      "grad_norm": 0.16872873902320862,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 17181
+    },
+    {
+      "epoch": 0.7268804467383028,
+      "grad_norm": 0.1536315381526947,
+      "learning_rate": 0.001,
+      "loss": 2.7305,
+      "step": 17182
+    },
+    {
+      "epoch": 0.7269227515018191,
+      "grad_norm": 17.74666404724121,
+      "learning_rate": 0.001,
+      "loss": 1.3806,
+      "step": 17183
+    },
+    {
+      "epoch": 0.7269650562653355,
+      "grad_norm": 0.17030657827854156,
+      "learning_rate": 0.001,
+      "loss": 2.2526,
+      "step": 17184
+    },
+    {
+      "epoch": 0.7270073610288519,
+      "grad_norm": 0.1639617383480072,
+      "learning_rate": 0.001,
+      "loss": 1.9684,
+      "step": 17185
+    },
+    {
+      "epoch": 0.7270496657923682,
+      "grad_norm": 0.18141882121562958,
+      "learning_rate": 0.001,
+      "loss": 1.8486,
+      "step": 17186
+    },
+    {
+      "epoch": 0.7270919705558846,
+      "grad_norm": 0.18478603661060333,
+      "learning_rate": 0.001,
+      "loss": 2.4985,
+      "step": 17187
+    },
+    {
+      "epoch": 0.727134275319401,
+      "grad_norm": 0.1748218983411789,
+      "learning_rate": 0.001,
+      "loss": 1.778,
+      "step": 17188
+    },
+    {
+      "epoch": 0.7271765800829173,
+      "grad_norm": 2.296082019805908,
+      "learning_rate": 0.001,
+      "loss": 2.8597,
+      "step": 17189
+    },
+    {
+      "epoch": 0.7272188848464337,
+      "grad_norm": 0.18864890933036804,
+      "learning_rate": 0.001,
+      "loss": 2.4064,
+      "step": 17190
+    },
+    {
+      "epoch": 0.7272611896099501,
+      "grad_norm": 0.17584924399852753,
+      "learning_rate": 0.001,
+      "loss": 2.0013,
+      "step": 17191
+    },
+    {
+      "epoch": 0.7273034943734664,
+      "grad_norm": 0.1358068883419037,
+      "learning_rate": 0.001,
+      "loss": 1.7993,
+      "step": 17192
+    },
+    {
+      "epoch": 0.7273457991369828,
+      "grad_norm": 0.17698509991168976,
+      "learning_rate": 0.001,
+      "loss": 2.1616,
+      "step": 17193
+    },
+    {
+      "epoch": 0.7273881039004992,
+      "grad_norm": 0.4649740159511566,
+      "learning_rate": 0.001,
+      "loss": 2.0258,
+      "step": 17194
+    },
+    {
+      "epoch": 0.7274304086640155,
+      "grad_norm": 0.15231867134571075,
+      "learning_rate": 0.001,
+      "loss": 2.1438,
+      "step": 17195
+    },
+    {
+      "epoch": 0.727472713427532,
+      "grad_norm": 0.19217731058597565,
+      "learning_rate": 0.001,
+      "loss": 3.0598,
+      "step": 17196
+    },
+    {
+      "epoch": 0.7275150181910484,
+      "grad_norm": 0.18122124671936035,
+      "learning_rate": 0.001,
+      "loss": 1.54,
+      "step": 17197
+    },
+    {
+      "epoch": 0.7275573229545647,
+      "grad_norm": 0.13571122288703918,
+      "learning_rate": 0.001,
+      "loss": 1.2435,
+      "step": 17198
+    },
+    {
+      "epoch": 0.7275996277180811,
+      "grad_norm": 0.14728273451328278,
+      "learning_rate": 0.001,
+      "loss": 1.6244,
+      "step": 17199
+    },
+    {
+      "epoch": 0.7276419324815975,
+      "grad_norm": 0.17009060084819794,
+      "learning_rate": 0.001,
+      "loss": 2.5318,
+      "step": 17200
+    },
+    {
+      "epoch": 0.7276842372451138,
+      "grad_norm": 0.17915566265583038,
+      "learning_rate": 0.001,
+      "loss": 2.2632,
+      "step": 17201
+    },
+    {
+      "epoch": 0.7277265420086302,
+      "grad_norm": 0.19644324481487274,
+      "learning_rate": 0.001,
+      "loss": 1.6086,
+      "step": 17202
+    },
+    {
+      "epoch": 0.7277688467721466,
+      "grad_norm": 0.16370825469493866,
+      "learning_rate": 0.001,
+      "loss": 1.992,
+      "step": 17203
+    },
+    {
+      "epoch": 0.7278111515356629,
+      "grad_norm": 0.219704270362854,
+      "learning_rate": 0.001,
+      "loss": 3.0066,
+      "step": 17204
+    },
+    {
+      "epoch": 0.7278534562991793,
+      "grad_norm": 0.2036728709936142,
+      "learning_rate": 0.001,
+      "loss": 2.1629,
+      "step": 17205
+    },
+    {
+      "epoch": 0.7278957610626957,
+      "grad_norm": 0.5341696739196777,
+      "learning_rate": 0.001,
+      "loss": 1.8417,
+      "step": 17206
+    },
+    {
+      "epoch": 0.727938065826212,
+      "grad_norm": 0.23344549536705017,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 17207
+    },
+    {
+      "epoch": 0.7279803705897284,
+      "grad_norm": 0.1574479639530182,
+      "learning_rate": 0.001,
+      "loss": 2.1478,
+      "step": 17208
+    },
+    {
+      "epoch": 0.7280226753532448,
+      "grad_norm": 0.14581382274627686,
+      "learning_rate": 0.001,
+      "loss": 3.0699,
+      "step": 17209
+    },
+    {
+      "epoch": 0.7280649801167611,
+      "grad_norm": 0.18210309743881226,
+      "learning_rate": 0.001,
+      "loss": 3.2913,
+      "step": 17210
+    },
+    {
+      "epoch": 0.7281072848802775,
+      "grad_norm": 0.2757953107357025,
+      "learning_rate": 0.001,
+      "loss": 2.9062,
+      "step": 17211
+    },
+    {
+      "epoch": 0.7281495896437938,
+      "grad_norm": 0.19155064225196838,
+      "learning_rate": 0.001,
+      "loss": 1.9776,
+      "step": 17212
+    },
+    {
+      "epoch": 0.7281918944073102,
+      "grad_norm": 0.17974048852920532,
+      "learning_rate": 0.001,
+      "loss": 2.187,
+      "step": 17213
+    },
+    {
+      "epoch": 0.7282341991708267,
+      "grad_norm": 0.16559724509716034,
+      "learning_rate": 0.001,
+      "loss": 1.7224,
+      "step": 17214
+    },
+    {
+      "epoch": 0.728276503934343,
+      "grad_norm": 0.1683637499809265,
+      "learning_rate": 0.001,
+      "loss": 3.5013,
+      "step": 17215
+    },
+    {
+      "epoch": 0.7283188086978594,
+      "grad_norm": 0.1559930294752121,
+      "learning_rate": 0.001,
+      "loss": 2.0094,
+      "step": 17216
+    },
+    {
+      "epoch": 0.7283611134613758,
+      "grad_norm": 0.17992942035198212,
+      "learning_rate": 0.001,
+      "loss": 2.7717,
+      "step": 17217
+    },
+    {
+      "epoch": 0.7284034182248921,
+      "grad_norm": 0.1760255992412567,
+      "learning_rate": 0.001,
+      "loss": 2.0994,
+      "step": 17218
+    },
+    {
+      "epoch": 0.7284457229884085,
+      "grad_norm": 0.15609480440616608,
+      "learning_rate": 0.001,
+      "loss": 2.3692,
+      "step": 17219
+    },
+    {
+      "epoch": 0.7284880277519249,
+      "grad_norm": 0.27221378684043884,
+      "learning_rate": 0.001,
+      "loss": 1.5019,
+      "step": 17220
+    },
+    {
+      "epoch": 0.7285303325154412,
+      "grad_norm": 0.1502402275800705,
+      "learning_rate": 0.001,
+      "loss": 2.3058,
+      "step": 17221
+    },
+    {
+      "epoch": 0.7285726372789576,
+      "grad_norm": 0.886763334274292,
+      "learning_rate": 0.001,
+      "loss": 2.2973,
+      "step": 17222
+    },
+    {
+      "epoch": 0.728614942042474,
+      "grad_norm": 0.1851658821105957,
+      "learning_rate": 0.001,
+      "loss": 1.7426,
+      "step": 17223
+    },
+    {
+      "epoch": 0.7286572468059903,
+      "grad_norm": 0.13750481605529785,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 17224
+    },
+    {
+      "epoch": 0.7286995515695067,
+      "grad_norm": 0.3507780134677887,
+      "learning_rate": 0.001,
+      "loss": 2.1812,
+      "step": 17225
+    },
+    {
+      "epoch": 0.7287418563330231,
+      "grad_norm": 0.2409433126449585,
+      "learning_rate": 0.001,
+      "loss": 1.9943,
+      "step": 17226
+    },
+    {
+      "epoch": 0.7287841610965394,
+      "grad_norm": 1.0777748823165894,
+      "learning_rate": 0.001,
+      "loss": 1.5415,
+      "step": 17227
+    },
+    {
+      "epoch": 0.7288264658600558,
+      "grad_norm": 0.5831173062324524,
+      "learning_rate": 0.001,
+      "loss": 2.9824,
+      "step": 17228
+    },
+    {
+      "epoch": 0.7288687706235722,
+      "grad_norm": 0.16224071383476257,
+      "learning_rate": 0.001,
+      "loss": 1.6938,
+      "step": 17229
+    },
+    {
+      "epoch": 0.7289110753870885,
+      "grad_norm": 0.3064275085926056,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 17230
+    },
+    {
+      "epoch": 0.728953380150605,
+      "grad_norm": 0.16514526307582855,
+      "learning_rate": 0.001,
+      "loss": 2.3135,
+      "step": 17231
+    },
+    {
+      "epoch": 0.7289956849141214,
+      "grad_norm": 0.23485006392002106,
+      "learning_rate": 0.001,
+      "loss": 2.1271,
+      "step": 17232
+    },
+    {
+      "epoch": 0.7290379896776377,
+      "grad_norm": 0.1845911592245102,
+      "learning_rate": 0.001,
+      "loss": 3.1486,
+      "step": 17233
+    },
+    {
+      "epoch": 0.7290802944411541,
+      "grad_norm": 0.23723876476287842,
+      "learning_rate": 0.001,
+      "loss": 3.0125,
+      "step": 17234
+    },
+    {
+      "epoch": 0.7291225992046705,
+      "grad_norm": 0.16877859830856323,
+      "learning_rate": 0.001,
+      "loss": 1.6634,
+      "step": 17235
+    },
+    {
+      "epoch": 0.7291649039681868,
+      "grad_norm": 0.1759745180606842,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 17236
+    },
+    {
+      "epoch": 0.7292072087317032,
+      "grad_norm": 0.14648011326789856,
+      "learning_rate": 0.001,
+      "loss": 1.6105,
+      "step": 17237
+    },
+    {
+      "epoch": 0.7292495134952196,
+      "grad_norm": 0.18712373077869415,
+      "learning_rate": 0.001,
+      "loss": 2.1293,
+      "step": 17238
+    },
+    {
+      "epoch": 0.7292918182587359,
+      "grad_norm": 0.1619027853012085,
+      "learning_rate": 0.001,
+      "loss": 1.3905,
+      "step": 17239
+    },
+    {
+      "epoch": 0.7293341230222523,
+      "grad_norm": 0.16455508768558502,
+      "learning_rate": 0.001,
+      "loss": 2.3091,
+      "step": 17240
+    },
+    {
+      "epoch": 0.7293764277857687,
+      "grad_norm": 0.173019140958786,
+      "learning_rate": 0.001,
+      "loss": 1.8305,
+      "step": 17241
+    },
+    {
+      "epoch": 0.729418732549285,
+      "grad_norm": 0.14237745106220245,
+      "learning_rate": 0.001,
+      "loss": 2.0488,
+      "step": 17242
+    },
+    {
+      "epoch": 0.7294610373128014,
+      "grad_norm": 0.1654769331216812,
+      "learning_rate": 0.001,
+      "loss": 1.6978,
+      "step": 17243
+    },
+    {
+      "epoch": 0.7295033420763178,
+      "grad_norm": 0.13907228410243988,
+      "learning_rate": 0.001,
+      "loss": 1.6468,
+      "step": 17244
+    },
+    {
+      "epoch": 0.7295456468398341,
+      "grad_norm": 0.15198984742164612,
+      "learning_rate": 0.001,
+      "loss": 1.644,
+      "step": 17245
+    },
+    {
+      "epoch": 0.7295879516033505,
+      "grad_norm": 0.3324761688709259,
+      "learning_rate": 0.001,
+      "loss": 1.8737,
+      "step": 17246
+    },
+    {
+      "epoch": 0.729630256366867,
+      "grad_norm": 0.16085344552993774,
+      "learning_rate": 0.001,
+      "loss": 1.588,
+      "step": 17247
+    },
+    {
+      "epoch": 0.7296725611303833,
+      "grad_norm": 0.21202977001667023,
+      "learning_rate": 0.001,
+      "loss": 2.3259,
+      "step": 17248
+    },
+    {
+      "epoch": 0.7297148658938997,
+      "grad_norm": 0.15043148398399353,
+      "learning_rate": 0.001,
+      "loss": 1.9283,
+      "step": 17249
+    },
+    {
+      "epoch": 0.7297571706574161,
+      "grad_norm": 0.1499723643064499,
+      "learning_rate": 0.001,
+      "loss": 1.2876,
+      "step": 17250
+    },
+    {
+      "epoch": 0.7297994754209324,
+      "grad_norm": 0.18416666984558105,
+      "learning_rate": 0.001,
+      "loss": 1.9966,
+      "step": 17251
+    },
+    {
+      "epoch": 0.7298417801844488,
+      "grad_norm": 0.19342102110385895,
+      "learning_rate": 0.001,
+      "loss": 2.1563,
+      "step": 17252
+    },
+    {
+      "epoch": 0.7298840849479652,
+      "grad_norm": 0.16368161141872406,
+      "learning_rate": 0.001,
+      "loss": 2.5488,
+      "step": 17253
+    },
+    {
+      "epoch": 0.7299263897114815,
+      "grad_norm": 0.19238623976707458,
+      "learning_rate": 0.001,
+      "loss": 2.4449,
+      "step": 17254
+    },
+    {
+      "epoch": 0.7299686944749979,
+      "grad_norm": 0.1476447880268097,
+      "learning_rate": 0.001,
+      "loss": 2.1851,
+      "step": 17255
+    },
+    {
+      "epoch": 0.7300109992385142,
+      "grad_norm": 2.3451356887817383,
+      "learning_rate": 0.001,
+      "loss": 1.8425,
+      "step": 17256
+    },
+    {
+      "epoch": 0.7300533040020306,
+      "grad_norm": 0.172628253698349,
+      "learning_rate": 0.001,
+      "loss": 2.2727,
+      "step": 17257
+    },
+    {
+      "epoch": 0.730095608765547,
+      "grad_norm": 0.23679950833320618,
+      "learning_rate": 0.001,
+      "loss": 2.0103,
+      "step": 17258
+    },
+    {
+      "epoch": 0.7301379135290633,
+      "grad_norm": 0.16247250139713287,
+      "learning_rate": 0.001,
+      "loss": 1.9596,
+      "step": 17259
+    },
+    {
+      "epoch": 0.7301802182925797,
+      "grad_norm": 0.9013334512710571,
+      "learning_rate": 0.001,
+      "loss": 3.134,
+      "step": 17260
+    },
+    {
+      "epoch": 0.7302225230560961,
+      "grad_norm": 0.15641777217388153,
+      "learning_rate": 0.001,
+      "loss": 2.2646,
+      "step": 17261
+    },
+    {
+      "epoch": 0.7302648278196124,
+      "grad_norm": 0.15012650191783905,
+      "learning_rate": 0.001,
+      "loss": 1.8738,
+      "step": 17262
+    },
+    {
+      "epoch": 0.7303071325831288,
+      "grad_norm": 0.16362197697162628,
+      "learning_rate": 0.001,
+      "loss": 2.5672,
+      "step": 17263
+    },
+    {
+      "epoch": 0.7303494373466453,
+      "grad_norm": 4.925112724304199,
+      "learning_rate": 0.001,
+      "loss": 3.0045,
+      "step": 17264
+    },
+    {
+      "epoch": 0.7303917421101616,
+      "grad_norm": 0.1557965725660324,
+      "learning_rate": 0.001,
+      "loss": 1.6906,
+      "step": 17265
+    },
+    {
+      "epoch": 0.730434046873678,
+      "grad_norm": 0.1717749536037445,
+      "learning_rate": 0.001,
+      "loss": 1.9104,
+      "step": 17266
+    },
+    {
+      "epoch": 0.7304763516371944,
+      "grad_norm": 0.17836451530456543,
+      "learning_rate": 0.001,
+      "loss": 2.1333,
+      "step": 17267
+    },
+    {
+      "epoch": 0.7305186564007107,
+      "grad_norm": 0.15282116830348969,
+      "learning_rate": 0.001,
+      "loss": 1.9864,
+      "step": 17268
+    },
+    {
+      "epoch": 0.7305609611642271,
+      "grad_norm": 0.16762694716453552,
+      "learning_rate": 0.001,
+      "loss": 2.6978,
+      "step": 17269
+    },
+    {
+      "epoch": 0.7306032659277435,
+      "grad_norm": 0.3359089493751526,
+      "learning_rate": 0.001,
+      "loss": 2.244,
+      "step": 17270
+    },
+    {
+      "epoch": 0.7306455706912598,
+      "grad_norm": 0.6835538148880005,
+      "learning_rate": 0.001,
+      "loss": 3.3602,
+      "step": 17271
+    },
+    {
+      "epoch": 0.7306878754547762,
+      "grad_norm": 0.16285963356494904,
+      "learning_rate": 0.001,
+      "loss": 1.4811,
+      "step": 17272
+    },
+    {
+      "epoch": 0.7307301802182926,
+      "grad_norm": 1.3391896486282349,
+      "learning_rate": 0.001,
+      "loss": 2.6051,
+      "step": 17273
+    },
+    {
+      "epoch": 0.7307724849818089,
+      "grad_norm": 0.17385242879390717,
+      "learning_rate": 0.001,
+      "loss": 3.2361,
+      "step": 17274
+    },
+    {
+      "epoch": 0.7308147897453253,
+      "grad_norm": 5.107343673706055,
+      "learning_rate": 0.001,
+      "loss": 2.1832,
+      "step": 17275
+    },
+    {
+      "epoch": 0.7308570945088417,
+      "grad_norm": 0.14723435044288635,
+      "learning_rate": 0.001,
+      "loss": 1.8271,
+      "step": 17276
+    },
+    {
+      "epoch": 0.730899399272358,
+      "grad_norm": 0.20319636166095734,
+      "learning_rate": 0.001,
+      "loss": 2.0419,
+      "step": 17277
+    },
+    {
+      "epoch": 0.7309417040358744,
+      "grad_norm": 0.21243970096111298,
+      "learning_rate": 0.001,
+      "loss": 2.0175,
+      "step": 17278
+    },
+    {
+      "epoch": 0.7309840087993908,
+      "grad_norm": 0.16713847219944,
+      "learning_rate": 0.001,
+      "loss": 1.5741,
+      "step": 17279
+    },
+    {
+      "epoch": 0.7310263135629071,
+      "grad_norm": 0.16459450125694275,
+      "learning_rate": 0.001,
+      "loss": 2.6241,
+      "step": 17280
+    },
+    {
+      "epoch": 0.7310686183264236,
+      "grad_norm": 0.13629010319709778,
+      "learning_rate": 0.001,
+      "loss": 2.9519,
+      "step": 17281
+    },
+    {
+      "epoch": 0.73111092308994,
+      "grad_norm": 0.1602662056684494,
+      "learning_rate": 0.001,
+      "loss": 1.6162,
+      "step": 17282
+    },
+    {
+      "epoch": 0.7311532278534563,
+      "grad_norm": 0.36450186371803284,
+      "learning_rate": 0.001,
+      "loss": 1.7024,
+      "step": 17283
+    },
+    {
+      "epoch": 0.7311955326169727,
+      "grad_norm": 0.1558094173669815,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 17284
+    },
+    {
+      "epoch": 0.7312378373804891,
+      "grad_norm": 0.161569282412529,
+      "learning_rate": 0.001,
+      "loss": 1.6652,
+      "step": 17285
+    },
+    {
+      "epoch": 0.7312801421440054,
+      "grad_norm": 0.394671767950058,
+      "learning_rate": 0.001,
+      "loss": 2.9097,
+      "step": 17286
+    },
+    {
+      "epoch": 0.7313224469075218,
+      "grad_norm": 0.1987708956003189,
+      "learning_rate": 0.001,
+      "loss": 2.3612,
+      "step": 17287
+    },
+    {
+      "epoch": 0.7313647516710382,
+      "grad_norm": 0.19308647513389587,
+      "learning_rate": 0.001,
+      "loss": 2.5338,
+      "step": 17288
+    },
+    {
+      "epoch": 0.7314070564345545,
+      "grad_norm": 0.15085269510746002,
+      "learning_rate": 0.001,
+      "loss": 2.0816,
+      "step": 17289
+    },
+    {
+      "epoch": 0.7314493611980709,
+      "grad_norm": 0.15625354647636414,
+      "learning_rate": 0.001,
+      "loss": 1.6518,
+      "step": 17290
+    },
+    {
+      "epoch": 0.7314916659615873,
+      "grad_norm": 6.045670032501221,
+      "learning_rate": 0.001,
+      "loss": 2.2589,
+      "step": 17291
+    },
+    {
+      "epoch": 0.7315339707251036,
+      "grad_norm": 1.6492812633514404,
+      "learning_rate": 0.001,
+      "loss": 2.7467,
+      "step": 17292
+    },
+    {
+      "epoch": 0.73157627548862,
+      "grad_norm": 0.18798352777957916,
+      "learning_rate": 0.001,
+      "loss": 1.8589,
+      "step": 17293
+    },
+    {
+      "epoch": 0.7316185802521364,
+      "grad_norm": 0.1604667603969574,
+      "learning_rate": 0.001,
+      "loss": 2.51,
+      "step": 17294
+    },
+    {
+      "epoch": 0.7316608850156527,
+      "grad_norm": 0.15313079953193665,
+      "learning_rate": 0.001,
+      "loss": 2.0193,
+      "step": 17295
+    },
+    {
+      "epoch": 0.7317031897791691,
+      "grad_norm": 0.28037911653518677,
+      "learning_rate": 0.001,
+      "loss": 1.8532,
+      "step": 17296
+    },
+    {
+      "epoch": 0.7317454945426856,
+      "grad_norm": 0.21967746317386627,
+      "learning_rate": 0.001,
+      "loss": 2.5508,
+      "step": 17297
+    },
+    {
+      "epoch": 0.7317877993062019,
+      "grad_norm": 0.16519500315189362,
+      "learning_rate": 0.001,
+      "loss": 1.861,
+      "step": 17298
+    },
+    {
+      "epoch": 0.7318301040697183,
+      "grad_norm": 0.22370240092277527,
+      "learning_rate": 0.001,
+      "loss": 3.3633,
+      "step": 17299
+    },
+    {
+      "epoch": 0.7318724088332346,
+      "grad_norm": 0.18574631214141846,
+      "learning_rate": 0.001,
+      "loss": 2.219,
+      "step": 17300
+    },
+    {
+      "epoch": 0.731914713596751,
+      "grad_norm": 6.258697986602783,
+      "learning_rate": 0.001,
+      "loss": 2.1701,
+      "step": 17301
+    },
+    {
+      "epoch": 0.7319570183602674,
+      "grad_norm": 0.16539275646209717,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 17302
+    },
+    {
+      "epoch": 0.7319993231237837,
+      "grad_norm": 0.47777655720710754,
+      "learning_rate": 0.001,
+      "loss": 1.445,
+      "step": 17303
+    },
+    {
+      "epoch": 0.7320416278873001,
+      "grad_norm": 0.16214773058891296,
+      "learning_rate": 0.001,
+      "loss": 2.2202,
+      "step": 17304
+    },
+    {
+      "epoch": 0.7320839326508165,
+      "grad_norm": 0.16900905966758728,
+      "learning_rate": 0.001,
+      "loss": 2.3208,
+      "step": 17305
+    },
+    {
+      "epoch": 0.7321262374143328,
+      "grad_norm": 0.2838803231716156,
+      "learning_rate": 0.001,
+      "loss": 3.2839,
+      "step": 17306
+    },
+    {
+      "epoch": 0.7321685421778492,
+      "grad_norm": 0.19927377998828888,
+      "learning_rate": 0.001,
+      "loss": 1.9397,
+      "step": 17307
+    },
+    {
+      "epoch": 0.7322108469413656,
+      "grad_norm": 0.2335241436958313,
+      "learning_rate": 0.001,
+      "loss": 2.9701,
+      "step": 17308
+    },
+    {
+      "epoch": 0.7322531517048819,
+      "grad_norm": 0.1455473154783249,
+      "learning_rate": 0.001,
+      "loss": 2.0696,
+      "step": 17309
+    },
+    {
+      "epoch": 0.7322954564683983,
+      "grad_norm": 0.17584875226020813,
+      "learning_rate": 0.001,
+      "loss": 1.9355,
+      "step": 17310
+    },
+    {
+      "epoch": 0.7323377612319147,
+      "grad_norm": 0.1485612541437149,
+      "learning_rate": 0.001,
+      "loss": 1.7784,
+      "step": 17311
+    },
+    {
+      "epoch": 0.732380065995431,
+      "grad_norm": 0.17357254028320312,
+      "learning_rate": 0.001,
+      "loss": 2.162,
+      "step": 17312
+    },
+    {
+      "epoch": 0.7324223707589474,
+      "grad_norm": 0.322561651468277,
+      "learning_rate": 0.001,
+      "loss": 1.7913,
+      "step": 17313
+    },
+    {
+      "epoch": 0.7324646755224639,
+      "grad_norm": 0.17470252513885498,
+      "learning_rate": 0.001,
+      "loss": 1.835,
+      "step": 17314
+    },
+    {
+      "epoch": 0.7325069802859802,
+      "grad_norm": 0.1539514809846878,
+      "learning_rate": 0.001,
+      "loss": 1.8439,
+      "step": 17315
+    },
+    {
+      "epoch": 0.7325492850494966,
+      "grad_norm": 0.13950100541114807,
+      "learning_rate": 0.001,
+      "loss": 1.9242,
+      "step": 17316
+    },
+    {
+      "epoch": 0.732591589813013,
+      "grad_norm": 1.2174321413040161,
+      "learning_rate": 0.001,
+      "loss": 2.0929,
+      "step": 17317
+    },
+    {
+      "epoch": 0.7326338945765293,
+      "grad_norm": 0.6627561450004578,
+      "learning_rate": 0.001,
+      "loss": 2.2273,
+      "step": 17318
+    },
+    {
+      "epoch": 0.7326761993400457,
+      "grad_norm": 0.17448411881923676,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 17319
+    },
+    {
+      "epoch": 0.7327185041035621,
+      "grad_norm": 0.20414674282073975,
+      "learning_rate": 0.001,
+      "loss": 2.4904,
+      "step": 17320
+    },
+    {
+      "epoch": 0.7327608088670784,
+      "grad_norm": 3.205564022064209,
+      "learning_rate": 0.001,
+      "loss": 2.7201,
+      "step": 17321
+    },
+    {
+      "epoch": 0.7328031136305948,
+      "grad_norm": 0.17250360548496246,
+      "learning_rate": 0.001,
+      "loss": 2.3204,
+      "step": 17322
+    },
+    {
+      "epoch": 0.7328454183941112,
+      "grad_norm": 0.1782556027173996,
+      "learning_rate": 0.001,
+      "loss": 1.8324,
+      "step": 17323
+    },
+    {
+      "epoch": 0.7328877231576275,
+      "grad_norm": 0.15511076152324677,
+      "learning_rate": 0.001,
+      "loss": 1.9694,
+      "step": 17324
+    },
+    {
+      "epoch": 0.7329300279211439,
+      "grad_norm": 0.49949324131011963,
+      "learning_rate": 0.001,
+      "loss": 3.1143,
+      "step": 17325
+    },
+    {
+      "epoch": 0.7329723326846603,
+      "grad_norm": 0.1534297913312912,
+      "learning_rate": 0.001,
+      "loss": 1.7866,
+      "step": 17326
+    },
+    {
+      "epoch": 0.7330146374481766,
+      "grad_norm": 0.15869884192943573,
+      "learning_rate": 0.001,
+      "loss": 1.9116,
+      "step": 17327
+    },
+    {
+      "epoch": 0.733056942211693,
+      "grad_norm": 0.14846599102020264,
+      "learning_rate": 0.001,
+      "loss": 2.6805,
+      "step": 17328
+    },
+    {
+      "epoch": 0.7330992469752095,
+      "grad_norm": 0.15433016419410706,
+      "learning_rate": 0.001,
+      "loss": 2.7788,
+      "step": 17329
+    },
+    {
+      "epoch": 0.7331415517387257,
+      "grad_norm": 0.17470060288906097,
+      "learning_rate": 0.001,
+      "loss": 1.5982,
+      "step": 17330
+    },
+    {
+      "epoch": 0.7331838565022422,
+      "grad_norm": 0.1676022708415985,
+      "learning_rate": 0.001,
+      "loss": 3.3472,
+      "step": 17331
+    },
+    {
+      "epoch": 0.7332261612657586,
+      "grad_norm": 0.16653110086917877,
+      "learning_rate": 0.001,
+      "loss": 1.7236,
+      "step": 17332
+    },
+    {
+      "epoch": 0.7332684660292749,
+      "grad_norm": 0.1555767059326172,
+      "learning_rate": 0.001,
+      "loss": 2.201,
+      "step": 17333
+    },
+    {
+      "epoch": 0.7333107707927913,
+      "grad_norm": 3.2302515506744385,
+      "learning_rate": 0.001,
+      "loss": 2.3596,
+      "step": 17334
+    },
+    {
+      "epoch": 0.7333530755563077,
+      "grad_norm": 0.13147684931755066,
+      "learning_rate": 0.001,
+      "loss": 1.4347,
+      "step": 17335
+    },
+    {
+      "epoch": 0.733395380319824,
+      "grad_norm": 0.1869000345468521,
+      "learning_rate": 0.001,
+      "loss": 1.6992,
+      "step": 17336
+    },
+    {
+      "epoch": 0.7334376850833404,
+      "grad_norm": 1.2183090448379517,
+      "learning_rate": 0.001,
+      "loss": 2.0567,
+      "step": 17337
+    },
+    {
+      "epoch": 0.7334799898468568,
+      "grad_norm": 0.1941007673740387,
+      "learning_rate": 0.001,
+      "loss": 1.8982,
+      "step": 17338
+    },
+    {
+      "epoch": 0.7335222946103731,
+      "grad_norm": 0.16951008141040802,
+      "learning_rate": 0.001,
+      "loss": 2.2301,
+      "step": 17339
+    },
+    {
+      "epoch": 0.7335645993738895,
+      "grad_norm": 0.19473901391029358,
+      "learning_rate": 0.001,
+      "loss": 2.7647,
+      "step": 17340
+    },
+    {
+      "epoch": 0.7336069041374059,
+      "grad_norm": 0.22168588638305664,
+      "learning_rate": 0.001,
+      "loss": 1.8121,
+      "step": 17341
+    },
+    {
+      "epoch": 0.7336492089009222,
+      "grad_norm": 0.9966937303543091,
+      "learning_rate": 0.001,
+      "loss": 1.607,
+      "step": 17342
+    },
+    {
+      "epoch": 0.7336915136644386,
+      "grad_norm": 0.17989544570446014,
+      "learning_rate": 0.001,
+      "loss": 1.695,
+      "step": 17343
+    },
+    {
+      "epoch": 0.733733818427955,
+      "grad_norm": 0.2009655088186264,
+      "learning_rate": 0.001,
+      "loss": 2.5365,
+      "step": 17344
+    },
+    {
+      "epoch": 0.7337761231914713,
+      "grad_norm": 0.17318382859230042,
+      "learning_rate": 0.001,
+      "loss": 2.3375,
+      "step": 17345
+    },
+    {
+      "epoch": 0.7338184279549878,
+      "grad_norm": 0.5182844996452332,
+      "learning_rate": 0.001,
+      "loss": 1.7107,
+      "step": 17346
+    },
+    {
+      "epoch": 0.733860732718504,
+      "grad_norm": 0.15933315455913544,
+      "learning_rate": 0.001,
+      "loss": 2.7681,
+      "step": 17347
+    },
+    {
+      "epoch": 0.7339030374820205,
+      "grad_norm": 0.17201949656009674,
+      "learning_rate": 0.001,
+      "loss": 2.4166,
+      "step": 17348
+    },
+    {
+      "epoch": 0.7339453422455369,
+      "grad_norm": 0.15629622340202332,
+      "learning_rate": 0.001,
+      "loss": 1.4808,
+      "step": 17349
+    },
+    {
+      "epoch": 0.7339876470090532,
+      "grad_norm": 0.17875443398952484,
+      "learning_rate": 0.001,
+      "loss": 1.7621,
+      "step": 17350
+    },
+    {
+      "epoch": 0.7340299517725696,
+      "grad_norm": 0.18575279414653778,
+      "learning_rate": 0.001,
+      "loss": 2.5072,
+      "step": 17351
+    },
+    {
+      "epoch": 0.734072256536086,
+      "grad_norm": 0.17435061931610107,
+      "learning_rate": 0.001,
+      "loss": 1.8393,
+      "step": 17352
+    },
+    {
+      "epoch": 0.7341145612996023,
+      "grad_norm": 0.18890979886054993,
+      "learning_rate": 0.001,
+      "loss": 2.0328,
+      "step": 17353
+    },
+    {
+      "epoch": 0.7341568660631187,
+      "grad_norm": 0.41785550117492676,
+      "learning_rate": 0.001,
+      "loss": 2.1083,
+      "step": 17354
+    },
+    {
+      "epoch": 0.7341991708266351,
+      "grad_norm": 0.16583463549613953,
+      "learning_rate": 0.001,
+      "loss": 1.7583,
+      "step": 17355
+    },
+    {
+      "epoch": 0.7342414755901514,
+      "grad_norm": 0.19213053584098816,
+      "learning_rate": 0.001,
+      "loss": 1.8698,
+      "step": 17356
+    },
+    {
+      "epoch": 0.7342837803536678,
+      "grad_norm": 0.7036476731300354,
+      "learning_rate": 0.001,
+      "loss": 2.8578,
+      "step": 17357
+    },
+    {
+      "epoch": 0.7343260851171842,
+      "grad_norm": 0.48419860005378723,
+      "learning_rate": 0.001,
+      "loss": 2.4056,
+      "step": 17358
+    },
+    {
+      "epoch": 0.7343683898807005,
+      "grad_norm": 0.16891148686408997,
+      "learning_rate": 0.001,
+      "loss": 2.6222,
+      "step": 17359
+    },
+    {
+      "epoch": 0.7344106946442169,
+      "grad_norm": 0.19437286257743835,
+      "learning_rate": 0.001,
+      "loss": 1.9693,
+      "step": 17360
+    },
+    {
+      "epoch": 0.7344529994077333,
+      "grad_norm": 0.1752423644065857,
+      "learning_rate": 0.001,
+      "loss": 2.8185,
+      "step": 17361
+    },
+    {
+      "epoch": 0.7344953041712496,
+      "grad_norm": 0.19541260600090027,
+      "learning_rate": 0.001,
+      "loss": 2.6349,
+      "step": 17362
+    },
+    {
+      "epoch": 0.734537608934766,
+      "grad_norm": 0.19351324439048767,
+      "learning_rate": 0.001,
+      "loss": 2.8542,
+      "step": 17363
+    },
+    {
+      "epoch": 0.7345799136982825,
+      "grad_norm": 0.17461591958999634,
+      "learning_rate": 0.001,
+      "loss": 2.2905,
+      "step": 17364
+    },
+    {
+      "epoch": 0.7346222184617988,
+      "grad_norm": 0.16678562760353088,
+      "learning_rate": 0.001,
+      "loss": 1.2962,
+      "step": 17365
+    },
+    {
+      "epoch": 0.7346645232253152,
+      "grad_norm": 0.16038046777248383,
+      "learning_rate": 0.001,
+      "loss": 1.9527,
+      "step": 17366
+    },
+    {
+      "epoch": 0.7347068279888316,
+      "grad_norm": 0.1805296242237091,
+      "learning_rate": 0.001,
+      "loss": 3.3531,
+      "step": 17367
+    },
+    {
+      "epoch": 0.7347491327523479,
+      "grad_norm": 0.8957823514938354,
+      "learning_rate": 0.001,
+      "loss": 1.7126,
+      "step": 17368
+    },
+    {
+      "epoch": 0.7347914375158643,
+      "grad_norm": 0.1552397757768631,
+      "learning_rate": 0.001,
+      "loss": 1.6727,
+      "step": 17369
+    },
+    {
+      "epoch": 0.7348337422793807,
+      "grad_norm": 0.2678411900997162,
+      "learning_rate": 0.001,
+      "loss": 1.7442,
+      "step": 17370
+    },
+    {
+      "epoch": 0.734876047042897,
+      "grad_norm": 2.7068116664886475,
+      "learning_rate": 0.001,
+      "loss": 2.5148,
+      "step": 17371
+    },
+    {
+      "epoch": 0.7349183518064134,
+      "grad_norm": 0.18146149814128876,
+      "learning_rate": 0.001,
+      "loss": 2.6581,
+      "step": 17372
+    },
+    {
+      "epoch": 0.7349606565699298,
+      "grad_norm": 0.18075819313526154,
+      "learning_rate": 0.001,
+      "loss": 1.3142,
+      "step": 17373
+    },
+    {
+      "epoch": 0.7350029613334461,
+      "grad_norm": 0.2121601551771164,
+      "learning_rate": 0.001,
+      "loss": 1.9923,
+      "step": 17374
+    },
+    {
+      "epoch": 0.7350452660969625,
+      "grad_norm": 0.15848374366760254,
+      "learning_rate": 0.001,
+      "loss": 1.6758,
+      "step": 17375
+    },
+    {
+      "epoch": 0.7350875708604789,
+      "grad_norm": 0.17913897335529327,
+      "learning_rate": 0.001,
+      "loss": 2.6544,
+      "step": 17376
+    },
+    {
+      "epoch": 0.7351298756239952,
+      "grad_norm": 0.6060518026351929,
+      "learning_rate": 0.001,
+      "loss": 2.1991,
+      "step": 17377
+    },
+    {
+      "epoch": 0.7351721803875116,
+      "grad_norm": 0.16152165830135345,
+      "learning_rate": 0.001,
+      "loss": 1.9824,
+      "step": 17378
+    },
+    {
+      "epoch": 0.735214485151028,
+      "grad_norm": 0.21107393503189087,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 17379
+    },
+    {
+      "epoch": 0.7352567899145444,
+      "grad_norm": 0.26977473497390747,
+      "learning_rate": 0.001,
+      "loss": 2.0408,
+      "step": 17380
+    },
+    {
+      "epoch": 0.7352990946780608,
+      "grad_norm": 0.18134251236915588,
+      "learning_rate": 0.001,
+      "loss": 2.6027,
+      "step": 17381
+    },
+    {
+      "epoch": 0.7353413994415772,
+      "grad_norm": 0.1814247965812683,
+      "learning_rate": 0.001,
+      "loss": 2.2769,
+      "step": 17382
+    },
+    {
+      "epoch": 0.7353837042050935,
+      "grad_norm": 0.1722179353237152,
+      "learning_rate": 0.001,
+      "loss": 2.6387,
+      "step": 17383
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 0.6016577482223511,
+      "learning_rate": 0.001,
+      "loss": 3.3811,
+      "step": 17384
+    },
+    {
+      "epoch": 0.7354683137321263,
+      "grad_norm": 0.16259095072746277,
+      "learning_rate": 0.001,
+      "loss": 2.2258,
+      "step": 17385
+    },
+    {
+      "epoch": 0.7355106184956426,
+      "grad_norm": 74.26570892333984,
+      "learning_rate": 0.001,
+      "loss": 1.5017,
+      "step": 17386
+    },
+    {
+      "epoch": 0.735552923259159,
+      "grad_norm": 0.2374870926141739,
+      "learning_rate": 0.001,
+      "loss": 2.5826,
+      "step": 17387
+    },
+    {
+      "epoch": 0.7355952280226754,
+      "grad_norm": 0.15083283185958862,
+      "learning_rate": 0.001,
+      "loss": 1.9046,
+      "step": 17388
+    },
+    {
+      "epoch": 0.7356375327861917,
+      "grad_norm": 0.2028711885213852,
+      "learning_rate": 0.001,
+      "loss": 2.3847,
+      "step": 17389
+    },
+    {
+      "epoch": 0.7356798375497081,
+      "grad_norm": 0.20516479015350342,
+      "learning_rate": 0.001,
+      "loss": 2.4515,
+      "step": 17390
+    },
+    {
+      "epoch": 0.7357221423132244,
+      "grad_norm": 0.2230457216501236,
+      "learning_rate": 0.001,
+      "loss": 2.0254,
+      "step": 17391
+    },
+    {
+      "epoch": 0.7357644470767408,
+      "grad_norm": 0.18468298017978668,
+      "learning_rate": 0.001,
+      "loss": 3.509,
+      "step": 17392
+    },
+    {
+      "epoch": 0.7358067518402572,
+      "grad_norm": 0.4097122848033905,
+      "learning_rate": 0.001,
+      "loss": 2.2616,
+      "step": 17393
+    },
+    {
+      "epoch": 0.7358490566037735,
+      "grad_norm": 0.17995667457580566,
+      "learning_rate": 0.001,
+      "loss": 2.5077,
+      "step": 17394
+    },
+    {
+      "epoch": 0.7358913613672899,
+      "grad_norm": 0.5382487177848816,
+      "learning_rate": 0.001,
+      "loss": 2.278,
+      "step": 17395
+    },
+    {
+      "epoch": 0.7359336661308064,
+      "grad_norm": 7.614555358886719,
+      "learning_rate": 0.001,
+      "loss": 2.148,
+      "step": 17396
+    },
+    {
+      "epoch": 0.7359759708943227,
+      "grad_norm": 0.22902828454971313,
+      "learning_rate": 0.001,
+      "loss": 3.9575,
+      "step": 17397
+    },
+    {
+      "epoch": 0.7360182756578391,
+      "grad_norm": 0.193025603890419,
+      "learning_rate": 0.001,
+      "loss": 2.3315,
+      "step": 17398
+    },
+    {
+      "epoch": 0.7360605804213555,
+      "grad_norm": 0.164814755320549,
+      "learning_rate": 0.001,
+      "loss": 3.175,
+      "step": 17399
+    },
+    {
+      "epoch": 0.7361028851848718,
+      "grad_norm": 0.16597630083560944,
+      "learning_rate": 0.001,
+      "loss": 2.1436,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7361451899483882,
+      "grad_norm": 0.264950692653656,
+      "learning_rate": 0.001,
+      "loss": 2.5343,
+      "step": 17401
+    },
+    {
+      "epoch": 0.7361874947119046,
+      "grad_norm": 0.23448620736598969,
+      "learning_rate": 0.001,
+      "loss": 2.1109,
+      "step": 17402
+    },
+    {
+      "epoch": 0.7362297994754209,
+      "grad_norm": 1.8346327543258667,
+      "learning_rate": 0.001,
+      "loss": 2.1264,
+      "step": 17403
+    },
+    {
+      "epoch": 0.7362721042389373,
+      "grad_norm": 0.3137005865573883,
+      "learning_rate": 0.001,
+      "loss": 3.8748,
+      "step": 17404
+    },
+    {
+      "epoch": 0.7363144090024537,
+      "grad_norm": 0.15299704670906067,
+      "learning_rate": 0.001,
+      "loss": 1.8221,
+      "step": 17405
+    },
+    {
+      "epoch": 0.73635671376597,
+      "grad_norm": 0.3703068494796753,
+      "learning_rate": 0.001,
+      "loss": 2.0435,
+      "step": 17406
+    },
+    {
+      "epoch": 0.7363990185294864,
+      "grad_norm": 0.15267716348171234,
+      "learning_rate": 0.001,
+      "loss": 2.9499,
+      "step": 17407
+    },
+    {
+      "epoch": 0.7364413232930028,
+      "grad_norm": 0.2238219976425171,
+      "learning_rate": 0.001,
+      "loss": 1.8026,
+      "step": 17408
+    },
+    {
+      "epoch": 0.7364836280565191,
+      "grad_norm": 0.22296573221683502,
+      "learning_rate": 0.001,
+      "loss": 2.8872,
+      "step": 17409
+    },
+    {
+      "epoch": 0.7365259328200355,
+      "grad_norm": 0.3025241494178772,
+      "learning_rate": 0.001,
+      "loss": 3.631,
+      "step": 17410
+    },
+    {
+      "epoch": 0.7365682375835519,
+      "grad_norm": 0.19598890841007233,
+      "learning_rate": 0.001,
+      "loss": 3.2347,
+      "step": 17411
+    },
+    {
+      "epoch": 0.7366105423470682,
+      "grad_norm": 0.15584322810173035,
+      "learning_rate": 0.001,
+      "loss": 2.8041,
+      "step": 17412
+    },
+    {
+      "epoch": 0.7366528471105847,
+      "grad_norm": 10.525110244750977,
+      "learning_rate": 0.001,
+      "loss": 2.9718,
+      "step": 17413
+    },
+    {
+      "epoch": 0.7366951518741011,
+      "grad_norm": 0.19018413126468658,
+      "learning_rate": 0.001,
+      "loss": 1.9068,
+      "step": 17414
+    },
+    {
+      "epoch": 0.7367374566376174,
+      "grad_norm": 0.1911727637052536,
+      "learning_rate": 0.001,
+      "loss": 2.2898,
+      "step": 17415
+    },
+    {
+      "epoch": 0.7367797614011338,
+      "grad_norm": 0.2422974407672882,
+      "learning_rate": 0.001,
+      "loss": 1.8923,
+      "step": 17416
+    },
+    {
+      "epoch": 0.7368220661646502,
+      "grad_norm": 0.18192513287067413,
+      "learning_rate": 0.001,
+      "loss": 3.3953,
+      "step": 17417
+    },
+    {
+      "epoch": 0.7368643709281665,
+      "grad_norm": 1.0662686824798584,
+      "learning_rate": 0.001,
+      "loss": 3.1413,
+      "step": 17418
+    },
+    {
+      "epoch": 0.7369066756916829,
+      "grad_norm": 0.20433665812015533,
+      "learning_rate": 0.001,
+      "loss": 2.1175,
+      "step": 17419
+    },
+    {
+      "epoch": 0.7369489804551993,
+      "grad_norm": 0.2949641942977905,
+      "learning_rate": 0.001,
+      "loss": 1.5695,
+      "step": 17420
+    },
+    {
+      "epoch": 0.7369912852187156,
+      "grad_norm": 0.21626880764961243,
+      "learning_rate": 0.001,
+      "loss": 2.4829,
+      "step": 17421
+    },
+    {
+      "epoch": 0.737033589982232,
+      "grad_norm": 0.18955782055854797,
+      "learning_rate": 0.001,
+      "loss": 2.3706,
+      "step": 17422
+    },
+    {
+      "epoch": 0.7370758947457484,
+      "grad_norm": 0.19381113350391388,
+      "learning_rate": 0.001,
+      "loss": 1.5407,
+      "step": 17423
+    },
+    {
+      "epoch": 0.7371181995092647,
+      "grad_norm": 0.5135764479637146,
+      "learning_rate": 0.001,
+      "loss": 2.4144,
+      "step": 17424
+    },
+    {
+      "epoch": 0.7371605042727811,
+      "grad_norm": 0.15292295813560486,
+      "learning_rate": 0.001,
+      "loss": 2.3463,
+      "step": 17425
+    },
+    {
+      "epoch": 0.7372028090362975,
+      "grad_norm": 5.058320045471191,
+      "learning_rate": 0.001,
+      "loss": 1.8338,
+      "step": 17426
+    },
+    {
+      "epoch": 0.7372451137998138,
+      "grad_norm": 0.22380977869033813,
+      "learning_rate": 0.001,
+      "loss": 2.2025,
+      "step": 17427
+    },
+    {
+      "epoch": 0.7372874185633302,
+      "grad_norm": 0.17026227712631226,
+      "learning_rate": 0.001,
+      "loss": 2.0961,
+      "step": 17428
+    },
+    {
+      "epoch": 0.7373297233268467,
+      "grad_norm": 0.1634519249200821,
+      "learning_rate": 0.001,
+      "loss": 3.1614,
+      "step": 17429
+    },
+    {
+      "epoch": 0.737372028090363,
+      "grad_norm": 0.17133593559265137,
+      "learning_rate": 0.001,
+      "loss": 2.1654,
+      "step": 17430
+    },
+    {
+      "epoch": 0.7374143328538794,
+      "grad_norm": 2.215588092803955,
+      "learning_rate": 0.001,
+      "loss": 1.7608,
+      "step": 17431
+    },
+    {
+      "epoch": 0.7374566376173958,
+      "grad_norm": 3.688143730163574,
+      "learning_rate": 0.001,
+      "loss": 2.6195,
+      "step": 17432
+    },
+    {
+      "epoch": 0.7374989423809121,
+      "grad_norm": 0.18635474145412445,
+      "learning_rate": 0.001,
+      "loss": 2.0476,
+      "step": 17433
+    },
+    {
+      "epoch": 0.7375412471444285,
+      "grad_norm": 0.17662550508975983,
+      "learning_rate": 0.001,
+      "loss": 2.4137,
+      "step": 17434
+    },
+    {
+      "epoch": 0.7375835519079449,
+      "grad_norm": 0.20594938099384308,
+      "learning_rate": 0.001,
+      "loss": 2.4498,
+      "step": 17435
+    },
+    {
+      "epoch": 0.7376258566714612,
+      "grad_norm": 0.33524230122566223,
+      "learning_rate": 0.001,
+      "loss": 2.8723,
+      "step": 17436
+    },
+    {
+      "epoch": 0.7376681614349776,
+      "grad_norm": 0.2059050351381302,
+      "learning_rate": 0.001,
+      "loss": 2.2312,
+      "step": 17437
+    },
+    {
+      "epoch": 0.7377104661984939,
+      "grad_norm": 0.2377101182937622,
+      "learning_rate": 0.001,
+      "loss": 2.8656,
+      "step": 17438
+    },
+    {
+      "epoch": 0.7377527709620103,
+      "grad_norm": 0.18826983869075775,
+      "learning_rate": 0.001,
+      "loss": 1.5286,
+      "step": 17439
+    },
+    {
+      "epoch": 0.7377950757255267,
+      "grad_norm": 0.16893354058265686,
+      "learning_rate": 0.001,
+      "loss": 1.7591,
+      "step": 17440
+    },
+    {
+      "epoch": 0.737837380489043,
+      "grad_norm": 0.20331987738609314,
+      "learning_rate": 0.001,
+      "loss": 1.9848,
+      "step": 17441
+    },
+    {
+      "epoch": 0.7378796852525594,
+      "grad_norm": 0.291287362575531,
+      "learning_rate": 0.001,
+      "loss": 2.1354,
+      "step": 17442
+    },
+    {
+      "epoch": 0.7379219900160758,
+      "grad_norm": 0.17716658115386963,
+      "learning_rate": 0.001,
+      "loss": 2.696,
+      "step": 17443
+    },
+    {
+      "epoch": 0.7379642947795921,
+      "grad_norm": 0.18112552165985107,
+      "learning_rate": 0.001,
+      "loss": 2.3412,
+      "step": 17444
+    },
+    {
+      "epoch": 0.7380065995431085,
+      "grad_norm": 0.2917419672012329,
+      "learning_rate": 0.001,
+      "loss": 1.7166,
+      "step": 17445
+    },
+    {
+      "epoch": 0.738048904306625,
+      "grad_norm": 0.17572566866874695,
+      "learning_rate": 0.001,
+      "loss": 1.9647,
+      "step": 17446
+    },
+    {
+      "epoch": 0.7380912090701413,
+      "grad_norm": 0.27557796239852905,
+      "learning_rate": 0.001,
+      "loss": 2.2187,
+      "step": 17447
+    },
+    {
+      "epoch": 0.7381335138336577,
+      "grad_norm": 10.137809753417969,
+      "learning_rate": 0.001,
+      "loss": 1.7207,
+      "step": 17448
+    },
+    {
+      "epoch": 0.7381758185971741,
+      "grad_norm": 0.20233531296253204,
+      "learning_rate": 0.001,
+      "loss": 1.4421,
+      "step": 17449
+    },
+    {
+      "epoch": 0.7382181233606904,
+      "grad_norm": 0.2746541202068329,
+      "learning_rate": 0.001,
+      "loss": 1.8292,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7382604281242068,
+      "grad_norm": 0.2038559764623642,
+      "learning_rate": 0.001,
+      "loss": 2.6724,
+      "step": 17451
+    },
+    {
+      "epoch": 0.7383027328877232,
+      "grad_norm": 0.22853845357894897,
+      "learning_rate": 0.001,
+      "loss": 2.7177,
+      "step": 17452
+    },
+    {
+      "epoch": 0.7383450376512395,
+      "grad_norm": 0.24598032236099243,
+      "learning_rate": 0.001,
+      "loss": 2.0161,
+      "step": 17453
+    },
+    {
+      "epoch": 0.7383873424147559,
+      "grad_norm": 0.24328875541687012,
+      "learning_rate": 0.001,
+      "loss": 3.1901,
+      "step": 17454
+    },
+    {
+      "epoch": 0.7384296471782723,
+      "grad_norm": 0.18322333693504333,
+      "learning_rate": 0.001,
+      "loss": 1.7166,
+      "step": 17455
+    },
+    {
+      "epoch": 0.7384719519417886,
+      "grad_norm": 0.2122056633234024,
+      "learning_rate": 0.001,
+      "loss": 2.1863,
+      "step": 17456
+    },
+    {
+      "epoch": 0.738514256705305,
+      "grad_norm": 0.22805139422416687,
+      "learning_rate": 0.001,
+      "loss": 2.4584,
+      "step": 17457
+    },
+    {
+      "epoch": 0.7385565614688214,
+      "grad_norm": 2.5409958362579346,
+      "learning_rate": 0.001,
+      "loss": 2.2642,
+      "step": 17458
+    },
+    {
+      "epoch": 0.7385988662323377,
+      "grad_norm": 0.2093057632446289,
+      "learning_rate": 0.001,
+      "loss": 1.5797,
+      "step": 17459
+    },
+    {
+      "epoch": 0.7386411709958541,
+      "grad_norm": 0.1915285587310791,
+      "learning_rate": 0.001,
+      "loss": 2.1698,
+      "step": 17460
+    },
+    {
+      "epoch": 0.7386834757593705,
+      "grad_norm": 1.6814018487930298,
+      "learning_rate": 0.001,
+      "loss": 1.5954,
+      "step": 17461
+    },
+    {
+      "epoch": 0.7387257805228868,
+      "grad_norm": 0.19684362411499023,
+      "learning_rate": 0.001,
+      "loss": 1.8076,
+      "step": 17462
+    },
+    {
+      "epoch": 0.7387680852864033,
+      "grad_norm": 0.18094508349895477,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 17463
+    },
+    {
+      "epoch": 0.7388103900499197,
+      "grad_norm": 1.161989450454712,
+      "learning_rate": 0.001,
+      "loss": 2.7361,
+      "step": 17464
+    },
+    {
+      "epoch": 0.738852694813436,
+      "grad_norm": 0.21886514127254486,
+      "learning_rate": 0.001,
+      "loss": 2.6973,
+      "step": 17465
+    },
+    {
+      "epoch": 0.7388949995769524,
+      "grad_norm": 0.201472669839859,
+      "learning_rate": 0.001,
+      "loss": 2.3263,
+      "step": 17466
+    },
+    {
+      "epoch": 0.7389373043404688,
+      "grad_norm": 0.17126241326332092,
+      "learning_rate": 0.001,
+      "loss": 2.0101,
+      "step": 17467
+    },
+    {
+      "epoch": 0.7389796091039851,
+      "grad_norm": 0.18553845584392548,
+      "learning_rate": 0.001,
+      "loss": 1.9736,
+      "step": 17468
+    },
+    {
+      "epoch": 0.7390219138675015,
+      "grad_norm": 0.17792284488677979,
+      "learning_rate": 0.001,
+      "loss": 2.87,
+      "step": 17469
+    },
+    {
+      "epoch": 0.7390642186310179,
+      "grad_norm": 0.14445818960666656,
+      "learning_rate": 0.001,
+      "loss": 1.7331,
+      "step": 17470
+    },
+    {
+      "epoch": 0.7391065233945342,
+      "grad_norm": 0.15313506126403809,
+      "learning_rate": 0.001,
+      "loss": 1.9839,
+      "step": 17471
+    },
+    {
+      "epoch": 0.7391488281580506,
+      "grad_norm": 0.17701950669288635,
+      "learning_rate": 0.001,
+      "loss": 2.0847,
+      "step": 17472
+    },
+    {
+      "epoch": 0.739191132921567,
+      "grad_norm": 0.1592811495065689,
+      "learning_rate": 0.001,
+      "loss": 2.9718,
+      "step": 17473
+    },
+    {
+      "epoch": 0.7392334376850833,
+      "grad_norm": 0.1532157063484192,
+      "learning_rate": 0.001,
+      "loss": 2.1228,
+      "step": 17474
+    },
+    {
+      "epoch": 0.7392757424485997,
+      "grad_norm": 0.1617754101753235,
+      "learning_rate": 0.001,
+      "loss": 2.5941,
+      "step": 17475
+    },
+    {
+      "epoch": 0.7393180472121161,
+      "grad_norm": 0.1594044715166092,
+      "learning_rate": 0.001,
+      "loss": 2.0657,
+      "step": 17476
+    },
+    {
+      "epoch": 0.7393603519756324,
+      "grad_norm": 0.18134772777557373,
+      "learning_rate": 0.001,
+      "loss": 1.7252,
+      "step": 17477
+    },
+    {
+      "epoch": 0.7394026567391488,
+      "grad_norm": 0.14515188336372375,
+      "learning_rate": 0.001,
+      "loss": 1.3929,
+      "step": 17478
+    },
+    {
+      "epoch": 0.7394449615026653,
+      "grad_norm": 0.2504208981990814,
+      "learning_rate": 0.001,
+      "loss": 2.7328,
+      "step": 17479
+    },
+    {
+      "epoch": 0.7394872662661816,
+      "grad_norm": 0.24100682139396667,
+      "learning_rate": 0.001,
+      "loss": 1.7722,
+      "step": 17480
+    },
+    {
+      "epoch": 0.739529571029698,
+      "grad_norm": 0.19070883095264435,
+      "learning_rate": 0.001,
+      "loss": 1.6493,
+      "step": 17481
+    },
+    {
+      "epoch": 0.7395718757932143,
+      "grad_norm": 0.2634289264678955,
+      "learning_rate": 0.001,
+      "loss": 2.8336,
+      "step": 17482
+    },
+    {
+      "epoch": 0.7396141805567307,
+      "grad_norm": 0.1580793410539627,
+      "learning_rate": 0.001,
+      "loss": 2.444,
+      "step": 17483
+    },
+    {
+      "epoch": 0.7396564853202471,
+      "grad_norm": 0.17331601679325104,
+      "learning_rate": 0.001,
+      "loss": 3.8579,
+      "step": 17484
+    },
+    {
+      "epoch": 0.7396987900837634,
+      "grad_norm": 0.13026723265647888,
+      "learning_rate": 0.001,
+      "loss": 2.681,
+      "step": 17485
+    },
+    {
+      "epoch": 0.7397410948472798,
+      "grad_norm": 0.16975976526737213,
+      "learning_rate": 0.001,
+      "loss": 1.8355,
+      "step": 17486
+    },
+    {
+      "epoch": 0.7397833996107962,
+      "grad_norm": 0.1425335556268692,
+      "learning_rate": 0.001,
+      "loss": 2.1688,
+      "step": 17487
+    },
+    {
+      "epoch": 0.7398257043743125,
+      "grad_norm": 0.16109095513820648,
+      "learning_rate": 0.001,
+      "loss": 2.1205,
+      "step": 17488
+    },
+    {
+      "epoch": 0.7398680091378289,
+      "grad_norm": 0.14170965552330017,
+      "learning_rate": 0.001,
+      "loss": 1.7205,
+      "step": 17489
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 0.1703367382287979,
+      "learning_rate": 0.001,
+      "loss": 2.2542,
+      "step": 17490
+    },
+    {
+      "epoch": 0.7399526186648616,
+      "grad_norm": 0.18292082846164703,
+      "learning_rate": 0.001,
+      "loss": 2.0744,
+      "step": 17491
+    },
+    {
+      "epoch": 0.739994923428378,
+      "grad_norm": 0.15937143564224243,
+      "learning_rate": 0.001,
+      "loss": 2.5822,
+      "step": 17492
+    },
+    {
+      "epoch": 0.7400372281918944,
+      "grad_norm": 0.14507554471492767,
+      "learning_rate": 0.001,
+      "loss": 1.8052,
+      "step": 17493
+    },
+    {
+      "epoch": 0.7400795329554107,
+      "grad_norm": 0.4141727387905121,
+      "learning_rate": 0.001,
+      "loss": 2.2241,
+      "step": 17494
+    },
+    {
+      "epoch": 0.7401218377189271,
+      "grad_norm": 0.14581911265850067,
+      "learning_rate": 0.001,
+      "loss": 1.9695,
+      "step": 17495
+    },
+    {
+      "epoch": 0.7401641424824436,
+      "grad_norm": 0.1397586613893509,
+      "learning_rate": 0.001,
+      "loss": 2.4452,
+      "step": 17496
+    },
+    {
+      "epoch": 0.7402064472459599,
+      "grad_norm": 0.21478283405303955,
+      "learning_rate": 0.001,
+      "loss": 2.4327,
+      "step": 17497
+    },
+    {
+      "epoch": 0.7402487520094763,
+      "grad_norm": 0.16876371204853058,
+      "learning_rate": 0.001,
+      "loss": 2.7402,
+      "step": 17498
+    },
+    {
+      "epoch": 0.7402910567729927,
+      "grad_norm": 0.3392346203327179,
+      "learning_rate": 0.001,
+      "loss": 2.3049,
+      "step": 17499
+    },
+    {
+      "epoch": 0.740333361536509,
+      "grad_norm": 0.1360718309879303,
+      "learning_rate": 0.001,
+      "loss": 2.0013,
+      "step": 17500
+    },
+    {
+      "epoch": 0.7403756663000254,
+      "grad_norm": 0.14087393879890442,
+      "learning_rate": 0.001,
+      "loss": 1.7843,
+      "step": 17501
+    },
+    {
+      "epoch": 0.7404179710635418,
+      "grad_norm": 0.15217366814613342,
+      "learning_rate": 0.001,
+      "loss": 3.6722,
+      "step": 17502
+    },
+    {
+      "epoch": 0.7404602758270581,
+      "grad_norm": 0.17477324604988098,
+      "learning_rate": 0.001,
+      "loss": 2.2989,
+      "step": 17503
+    },
+    {
+      "epoch": 0.7405025805905745,
+      "grad_norm": 16.218931198120117,
+      "learning_rate": 0.001,
+      "loss": 1.676,
+      "step": 17504
+    },
+    {
+      "epoch": 0.7405448853540909,
+      "grad_norm": 0.1638619303703308,
+      "learning_rate": 0.001,
+      "loss": 2.3842,
+      "step": 17505
+    },
+    {
+      "epoch": 0.7405871901176072,
+      "grad_norm": 31.428306579589844,
+      "learning_rate": 0.001,
+      "loss": 3.2131,
+      "step": 17506
+    },
+    {
+      "epoch": 0.7406294948811236,
+      "grad_norm": 0.1608552634716034,
+      "learning_rate": 0.001,
+      "loss": 1.9849,
+      "step": 17507
+    },
+    {
+      "epoch": 0.74067179964464,
+      "grad_norm": 0.20988480746746063,
+      "learning_rate": 0.001,
+      "loss": 2.2174,
+      "step": 17508
+    },
+    {
+      "epoch": 0.7407141044081563,
+      "grad_norm": 3.444148302078247,
+      "learning_rate": 0.001,
+      "loss": 2.4545,
+      "step": 17509
+    },
+    {
+      "epoch": 0.7407564091716727,
+      "grad_norm": 0.17780986428260803,
+      "learning_rate": 0.001,
+      "loss": 2.6699,
+      "step": 17510
+    },
+    {
+      "epoch": 0.7407987139351891,
+      "grad_norm": 0.19687338173389435,
+      "learning_rate": 0.001,
+      "loss": 2.1274,
+      "step": 17511
+    },
+    {
+      "epoch": 0.7408410186987054,
+      "grad_norm": 0.3244110643863678,
+      "learning_rate": 0.001,
+      "loss": 1.981,
+      "step": 17512
+    },
+    {
+      "epoch": 0.7408833234622219,
+      "grad_norm": 0.19059693813323975,
+      "learning_rate": 0.001,
+      "loss": 2.4835,
+      "step": 17513
+    },
+    {
+      "epoch": 0.7409256282257383,
+      "grad_norm": 0.2618594169616699,
+      "learning_rate": 0.001,
+      "loss": 3.4605,
+      "step": 17514
+    },
+    {
+      "epoch": 0.7409679329892546,
+      "grad_norm": 0.4329042136669159,
+      "learning_rate": 0.001,
+      "loss": 2.3207,
+      "step": 17515
+    },
+    {
+      "epoch": 0.741010237752771,
+      "grad_norm": 0.45828118920326233,
+      "learning_rate": 0.001,
+      "loss": 2.6254,
+      "step": 17516
+    },
+    {
+      "epoch": 0.7410525425162874,
+      "grad_norm": 0.17340798676013947,
+      "learning_rate": 0.001,
+      "loss": 3.2398,
+      "step": 17517
+    },
+    {
+      "epoch": 0.7410948472798037,
+      "grad_norm": 0.1793540120124817,
+      "learning_rate": 0.001,
+      "loss": 2.3224,
+      "step": 17518
+    },
+    {
+      "epoch": 0.7411371520433201,
+      "grad_norm": 0.1443507820367813,
+      "learning_rate": 0.001,
+      "loss": 1.4805,
+      "step": 17519
+    },
+    {
+      "epoch": 0.7411794568068365,
+      "grad_norm": 0.1604100912809372,
+      "learning_rate": 0.001,
+      "loss": 1.6807,
+      "step": 17520
+    },
+    {
+      "epoch": 0.7412217615703528,
+      "grad_norm": 0.17397022247314453,
+      "learning_rate": 0.001,
+      "loss": 1.6518,
+      "step": 17521
+    },
+    {
+      "epoch": 0.7412640663338692,
+      "grad_norm": 0.49950161576271057,
+      "learning_rate": 0.001,
+      "loss": 2.6769,
+      "step": 17522
+    },
+    {
+      "epoch": 0.7413063710973856,
+      "grad_norm": 0.15705753862857819,
+      "learning_rate": 0.001,
+      "loss": 1.4069,
+      "step": 17523
+    },
+    {
+      "epoch": 0.7413486758609019,
+      "grad_norm": 0.14743894338607788,
+      "learning_rate": 0.001,
+      "loss": 2.8272,
+      "step": 17524
+    },
+    {
+      "epoch": 0.7413909806244183,
+      "grad_norm": 0.1543998420238495,
+      "learning_rate": 0.001,
+      "loss": 3.0686,
+      "step": 17525
+    },
+    {
+      "epoch": 0.7414332853879346,
+      "grad_norm": 0.16085979342460632,
+      "learning_rate": 0.001,
+      "loss": 1.5601,
+      "step": 17526
+    },
+    {
+      "epoch": 0.741475590151451,
+      "grad_norm": 0.22067369520664215,
+      "learning_rate": 0.001,
+      "loss": 3.7326,
+      "step": 17527
+    },
+    {
+      "epoch": 0.7415178949149674,
+      "grad_norm": 0.17484331130981445,
+      "learning_rate": 0.001,
+      "loss": 2.4677,
+      "step": 17528
+    },
+    {
+      "epoch": 0.7415601996784837,
+      "grad_norm": 0.3703678548336029,
+      "learning_rate": 0.001,
+      "loss": 2.2413,
+      "step": 17529
+    },
+    {
+      "epoch": 0.7416025044420002,
+      "grad_norm": 0.2962591350078583,
+      "learning_rate": 0.001,
+      "loss": 1.7069,
+      "step": 17530
+    },
+    {
+      "epoch": 0.7416448092055166,
+      "grad_norm": 0.656477689743042,
+      "learning_rate": 0.001,
+      "loss": 2.4666,
+      "step": 17531
+    },
+    {
+      "epoch": 0.7416871139690329,
+      "grad_norm": 0.36826154589653015,
+      "learning_rate": 0.001,
+      "loss": 1.8836,
+      "step": 17532
+    },
+    {
+      "epoch": 0.7417294187325493,
+      "grad_norm": 0.17669114470481873,
+      "learning_rate": 0.001,
+      "loss": 1.9258,
+      "step": 17533
+    },
+    {
+      "epoch": 0.7417717234960657,
+      "grad_norm": 0.14634403586387634,
+      "learning_rate": 0.001,
+      "loss": 1.8056,
+      "step": 17534
+    },
+    {
+      "epoch": 0.741814028259582,
+      "grad_norm": 0.15235856175422668,
+      "learning_rate": 0.001,
+      "loss": 3.3709,
+      "step": 17535
+    },
+    {
+      "epoch": 0.7418563330230984,
+      "grad_norm": 0.1723131686449051,
+      "learning_rate": 0.001,
+      "loss": 2.7196,
+      "step": 17536
+    },
+    {
+      "epoch": 0.7418986377866148,
+      "grad_norm": 0.1826736032962799,
+      "learning_rate": 0.001,
+      "loss": 2.494,
+      "step": 17537
+    },
+    {
+      "epoch": 0.7419409425501311,
+      "grad_norm": 0.16118574142456055,
+      "learning_rate": 0.001,
+      "loss": 2.9765,
+      "step": 17538
+    },
+    {
+      "epoch": 0.7419832473136475,
+      "grad_norm": 0.4343286454677582,
+      "learning_rate": 0.001,
+      "loss": 1.7836,
+      "step": 17539
+    },
+    {
+      "epoch": 0.7420255520771639,
+      "grad_norm": 1.067611575126648,
+      "learning_rate": 0.001,
+      "loss": 2.923,
+      "step": 17540
+    },
+    {
+      "epoch": 0.7420678568406802,
+      "grad_norm": 0.178229421377182,
+      "learning_rate": 0.001,
+      "loss": 2.2143,
+      "step": 17541
+    },
+    {
+      "epoch": 0.7421101616041966,
+      "grad_norm": 0.15373486280441284,
+      "learning_rate": 0.001,
+      "loss": 1.7178,
+      "step": 17542
+    },
+    {
+      "epoch": 0.742152466367713,
+      "grad_norm": 0.16188517212867737,
+      "learning_rate": 0.001,
+      "loss": 2.612,
+      "step": 17543
+    },
+    {
+      "epoch": 0.7421947711312293,
+      "grad_norm": 0.1258399486541748,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 17544
+    },
+    {
+      "epoch": 0.7422370758947457,
+      "grad_norm": 0.14767125248908997,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 17545
+    },
+    {
+      "epoch": 0.7422793806582622,
+      "grad_norm": 0.8984226584434509,
+      "learning_rate": 0.001,
+      "loss": 2.3823,
+      "step": 17546
+    },
+    {
+      "epoch": 0.7423216854217785,
+      "grad_norm": 0.12772268056869507,
+      "learning_rate": 0.001,
+      "loss": 2.3786,
+      "step": 17547
+    },
+    {
+      "epoch": 0.7423639901852949,
+      "grad_norm": 0.14280395209789276,
+      "learning_rate": 0.001,
+      "loss": 2.6982,
+      "step": 17548
+    },
+    {
+      "epoch": 0.7424062949488113,
+      "grad_norm": 0.15735042095184326,
+      "learning_rate": 0.001,
+      "loss": 2.907,
+      "step": 17549
+    },
+    {
+      "epoch": 0.7424485997123276,
+      "grad_norm": 0.22708289325237274,
+      "learning_rate": 0.001,
+      "loss": 2.9517,
+      "step": 17550
+    },
+    {
+      "epoch": 0.742490904475844,
+      "grad_norm": 0.5488585233688354,
+      "learning_rate": 0.001,
+      "loss": 3.046,
+      "step": 17551
+    },
+    {
+      "epoch": 0.7425332092393604,
+      "grad_norm": 0.15782161056995392,
+      "learning_rate": 0.001,
+      "loss": 2.2142,
+      "step": 17552
+    },
+    {
+      "epoch": 0.7425755140028767,
+      "grad_norm": 0.147671177983284,
+      "learning_rate": 0.001,
+      "loss": 2.2613,
+      "step": 17553
+    },
+    {
+      "epoch": 0.7426178187663931,
+      "grad_norm": 0.16425222158432007,
+      "learning_rate": 0.001,
+      "loss": 1.8327,
+      "step": 17554
+    },
+    {
+      "epoch": 0.7426601235299095,
+      "grad_norm": 0.15929697453975677,
+      "learning_rate": 0.001,
+      "loss": 2.6455,
+      "step": 17555
+    },
+    {
+      "epoch": 0.7427024282934258,
+      "grad_norm": 0.16624753177165985,
+      "learning_rate": 0.001,
+      "loss": 1.5187,
+      "step": 17556
+    },
+    {
+      "epoch": 0.7427447330569422,
+      "grad_norm": 0.36340129375457764,
+      "learning_rate": 0.001,
+      "loss": 2.103,
+      "step": 17557
+    },
+    {
+      "epoch": 0.7427870378204586,
+      "grad_norm": 0.20203012228012085,
+      "learning_rate": 0.001,
+      "loss": 3.6063,
+      "step": 17558
+    },
+    {
+      "epoch": 0.7428293425839749,
+      "grad_norm": 0.2196560502052307,
+      "learning_rate": 0.001,
+      "loss": 2.2121,
+      "step": 17559
+    },
+    {
+      "epoch": 0.7428716473474913,
+      "grad_norm": 0.16001419723033905,
+      "learning_rate": 0.001,
+      "loss": 2.8535,
+      "step": 17560
+    },
+    {
+      "epoch": 0.7429139521110077,
+      "grad_norm": 0.2432798594236374,
+      "learning_rate": 0.001,
+      "loss": 2.1841,
+      "step": 17561
+    },
+    {
+      "epoch": 0.742956256874524,
+      "grad_norm": 0.1755453497171402,
+      "learning_rate": 0.001,
+      "loss": 2.4195,
+      "step": 17562
+    },
+    {
+      "epoch": 0.7429985616380405,
+      "grad_norm": 0.16746820509433746,
+      "learning_rate": 0.001,
+      "loss": 1.645,
+      "step": 17563
+    },
+    {
+      "epoch": 0.7430408664015569,
+      "grad_norm": 0.1702737808227539,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 17564
+    },
+    {
+      "epoch": 0.7430831711650732,
+      "grad_norm": 0.15115857124328613,
+      "learning_rate": 0.001,
+      "loss": 1.7634,
+      "step": 17565
+    },
+    {
+      "epoch": 0.7431254759285896,
+      "grad_norm": 0.192986860871315,
+      "learning_rate": 0.001,
+      "loss": 2.5294,
+      "step": 17566
+    },
+    {
+      "epoch": 0.743167780692106,
+      "grad_norm": 0.1629578322172165,
+      "learning_rate": 0.001,
+      "loss": 1.551,
+      "step": 17567
+    },
+    {
+      "epoch": 0.7432100854556223,
+      "grad_norm": 2.344882011413574,
+      "learning_rate": 0.001,
+      "loss": 3.6963,
+      "step": 17568
+    },
+    {
+      "epoch": 0.7432523902191387,
+      "grad_norm": 0.19861623644828796,
+      "learning_rate": 0.001,
+      "loss": 2.5566,
+      "step": 17569
+    },
+    {
+      "epoch": 0.7432946949826551,
+      "grad_norm": 0.1944287270307541,
+      "learning_rate": 0.001,
+      "loss": 1.4486,
+      "step": 17570
+    },
+    {
+      "epoch": 0.7433369997461714,
+      "grad_norm": 0.2017359435558319,
+      "learning_rate": 0.001,
+      "loss": 2.2659,
+      "step": 17571
+    },
+    {
+      "epoch": 0.7433793045096878,
+      "grad_norm": 0.19808664917945862,
+      "learning_rate": 0.001,
+      "loss": 2.0332,
+      "step": 17572
+    },
+    {
+      "epoch": 0.7434216092732041,
+      "grad_norm": 0.1653938591480255,
+      "learning_rate": 0.001,
+      "loss": 2.2198,
+      "step": 17573
+    },
+    {
+      "epoch": 0.7434639140367205,
+      "grad_norm": 0.20208589732646942,
+      "learning_rate": 0.001,
+      "loss": 2.2754,
+      "step": 17574
+    },
+    {
+      "epoch": 0.7435062188002369,
+      "grad_norm": 0.1551780104637146,
+      "learning_rate": 0.001,
+      "loss": 1.8926,
+      "step": 17575
+    },
+    {
+      "epoch": 0.7435485235637532,
+      "grad_norm": 0.20212167501449585,
+      "learning_rate": 0.001,
+      "loss": 2.2995,
+      "step": 17576
+    },
+    {
+      "epoch": 0.7435908283272696,
+      "grad_norm": 0.2041611671447754,
+      "learning_rate": 0.001,
+      "loss": 2.7081,
+      "step": 17577
+    },
+    {
+      "epoch": 0.743633133090786,
+      "grad_norm": 0.18079808354377747,
+      "learning_rate": 0.001,
+      "loss": 3.0791,
+      "step": 17578
+    },
+    {
+      "epoch": 0.7436754378543023,
+      "grad_norm": 0.15058568120002747,
+      "learning_rate": 0.001,
+      "loss": 2.3326,
+      "step": 17579
+    },
+    {
+      "epoch": 0.7437177426178188,
+      "grad_norm": 0.18063566088676453,
+      "learning_rate": 0.001,
+      "loss": 2.0293,
+      "step": 17580
+    },
+    {
+      "epoch": 0.7437600473813352,
+      "grad_norm": 0.16308987140655518,
+      "learning_rate": 0.001,
+      "loss": 2.3896,
+      "step": 17581
+    },
+    {
+      "epoch": 0.7438023521448515,
+      "grad_norm": 0.14598752558231354,
+      "learning_rate": 0.001,
+      "loss": 1.7756,
+      "step": 17582
+    },
+    {
+      "epoch": 0.7438446569083679,
+      "grad_norm": 0.17939159274101257,
+      "learning_rate": 0.001,
+      "loss": 2.0577,
+      "step": 17583
+    },
+    {
+      "epoch": 0.7438869616718843,
+      "grad_norm": 0.2707332670688629,
+      "learning_rate": 0.001,
+      "loss": 2.4434,
+      "step": 17584
+    },
+    {
+      "epoch": 0.7439292664354006,
+      "grad_norm": 0.15140265226364136,
+      "learning_rate": 0.001,
+      "loss": 1.9017,
+      "step": 17585
+    },
+    {
+      "epoch": 0.743971571198917,
+      "grad_norm": 0.18124867975711823,
+      "learning_rate": 0.001,
+      "loss": 2.7051,
+      "step": 17586
+    },
+    {
+      "epoch": 0.7440138759624334,
+      "grad_norm": 0.20102323591709137,
+      "learning_rate": 0.001,
+      "loss": 2.3188,
+      "step": 17587
+    },
+    {
+      "epoch": 0.7440561807259497,
+      "grad_norm": 1.4687132835388184,
+      "learning_rate": 0.001,
+      "loss": 2.4702,
+      "step": 17588
+    },
+    {
+      "epoch": 0.7440984854894661,
+      "grad_norm": 0.4242343604564667,
+      "learning_rate": 0.001,
+      "loss": 1.6438,
+      "step": 17589
+    },
+    {
+      "epoch": 0.7441407902529825,
+      "grad_norm": 0.1698838770389557,
+      "learning_rate": 0.001,
+      "loss": 1.8073,
+      "step": 17590
+    },
+    {
+      "epoch": 0.7441830950164988,
+      "grad_norm": 0.3089980483055115,
+      "learning_rate": 0.001,
+      "loss": 2.1405,
+      "step": 17591
+    },
+    {
+      "epoch": 0.7442253997800152,
+      "grad_norm": 0.8378577828407288,
+      "learning_rate": 0.001,
+      "loss": 1.7228,
+      "step": 17592
+    },
+    {
+      "epoch": 0.7442677045435316,
+      "grad_norm": 0.14282885193824768,
+      "learning_rate": 0.001,
+      "loss": 2.7357,
+      "step": 17593
+    },
+    {
+      "epoch": 0.7443100093070479,
+      "grad_norm": 0.6160475015640259,
+      "learning_rate": 0.001,
+      "loss": 1.9501,
+      "step": 17594
+    },
+    {
+      "epoch": 0.7443523140705643,
+      "grad_norm": 0.17935483157634735,
+      "learning_rate": 0.001,
+      "loss": 2.9295,
+      "step": 17595
+    },
+    {
+      "epoch": 0.7443946188340808,
+      "grad_norm": 0.15372052788734436,
+      "learning_rate": 0.001,
+      "loss": 1.741,
+      "step": 17596
+    },
+    {
+      "epoch": 0.7444369235975971,
+      "grad_norm": 0.13836802542209625,
+      "learning_rate": 0.001,
+      "loss": 2.3167,
+      "step": 17597
+    },
+    {
+      "epoch": 0.7444792283611135,
+      "grad_norm": 0.4473693370819092,
+      "learning_rate": 0.001,
+      "loss": 1.2818,
+      "step": 17598
+    },
+    {
+      "epoch": 0.7445215331246299,
+      "grad_norm": 0.1744404435157776,
+      "learning_rate": 0.001,
+      "loss": 1.7035,
+      "step": 17599
+    },
+    {
+      "epoch": 0.7445638378881462,
+      "grad_norm": 0.1443067491054535,
+      "learning_rate": 0.001,
+      "loss": 2.1968,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7446061426516626,
+      "grad_norm": 0.14389725029468536,
+      "learning_rate": 0.001,
+      "loss": 1.8681,
+      "step": 17601
+    },
+    {
+      "epoch": 0.744648447415179,
+      "grad_norm": 0.1405625343322754,
+      "learning_rate": 0.001,
+      "loss": 2.3588,
+      "step": 17602
+    },
+    {
+      "epoch": 0.7446907521786953,
+      "grad_norm": 0.16592542827129364,
+      "learning_rate": 0.001,
+      "loss": 1.8942,
+      "step": 17603
+    },
+    {
+      "epoch": 0.7447330569422117,
+      "grad_norm": 0.23292699456214905,
+      "learning_rate": 0.001,
+      "loss": 3.1274,
+      "step": 17604
+    },
+    {
+      "epoch": 0.7447753617057281,
+      "grad_norm": 0.23173336684703827,
+      "learning_rate": 0.001,
+      "loss": 2.626,
+      "step": 17605
+    },
+    {
+      "epoch": 0.7448176664692444,
+      "grad_norm": 0.1321907639503479,
+      "learning_rate": 0.001,
+      "loss": 1.5545,
+      "step": 17606
+    },
+    {
+      "epoch": 0.7448599712327608,
+      "grad_norm": 0.15251246094703674,
+      "learning_rate": 0.001,
+      "loss": 1.5877,
+      "step": 17607
+    },
+    {
+      "epoch": 0.7449022759962772,
+      "grad_norm": 0.16497798264026642,
+      "learning_rate": 0.001,
+      "loss": 2.6468,
+      "step": 17608
+    },
+    {
+      "epoch": 0.7449445807597935,
+      "grad_norm": 0.15258750319480896,
+      "learning_rate": 0.001,
+      "loss": 2.7736,
+      "step": 17609
+    },
+    {
+      "epoch": 0.7449868855233099,
+      "grad_norm": 0.1986195147037506,
+      "learning_rate": 0.001,
+      "loss": 2.0389,
+      "step": 17610
+    },
+    {
+      "epoch": 0.7450291902868263,
+      "grad_norm": 0.13651925325393677,
+      "learning_rate": 0.001,
+      "loss": 1.9696,
+      "step": 17611
+    },
+    {
+      "epoch": 0.7450714950503426,
+      "grad_norm": 0.17349354922771454,
+      "learning_rate": 0.001,
+      "loss": 1.5452,
+      "step": 17612
+    },
+    {
+      "epoch": 0.7451137998138591,
+      "grad_norm": 0.18067964911460876,
+      "learning_rate": 0.001,
+      "loss": 1.5918,
+      "step": 17613
+    },
+    {
+      "epoch": 0.7451561045773755,
+      "grad_norm": 91.02615356445312,
+      "learning_rate": 0.001,
+      "loss": 1.9013,
+      "step": 17614
+    },
+    {
+      "epoch": 0.7451984093408918,
+      "grad_norm": 0.17014148831367493,
+      "learning_rate": 0.001,
+      "loss": 2.085,
+      "step": 17615
+    },
+    {
+      "epoch": 0.7452407141044082,
+      "grad_norm": 0.6066452860832214,
+      "learning_rate": 0.001,
+      "loss": 2.2974,
+      "step": 17616
+    },
+    {
+      "epoch": 0.7452830188679245,
+      "grad_norm": 0.1638554334640503,
+      "learning_rate": 0.001,
+      "loss": 1.9358,
+      "step": 17617
+    },
+    {
+      "epoch": 0.7453253236314409,
+      "grad_norm": 0.17447148263454437,
+      "learning_rate": 0.001,
+      "loss": 2.4924,
+      "step": 17618
+    },
+    {
+      "epoch": 0.7453676283949573,
+      "grad_norm": 0.21377401053905487,
+      "learning_rate": 0.001,
+      "loss": 1.7172,
+      "step": 17619
+    },
+    {
+      "epoch": 0.7454099331584736,
+      "grad_norm": 0.21663278341293335,
+      "learning_rate": 0.001,
+      "loss": 1.6387,
+      "step": 17620
+    },
+    {
+      "epoch": 0.74545223792199,
+      "grad_norm": 0.1757373809814453,
+      "learning_rate": 0.001,
+      "loss": 2.1301,
+      "step": 17621
+    },
+    {
+      "epoch": 0.7454945426855064,
+      "grad_norm": 0.2108677327632904,
+      "learning_rate": 0.001,
+      "loss": 1.9894,
+      "step": 17622
+    },
+    {
+      "epoch": 0.7455368474490227,
+      "grad_norm": 0.23199456930160522,
+      "learning_rate": 0.001,
+      "loss": 1.837,
+      "step": 17623
+    },
+    {
+      "epoch": 0.7455791522125391,
+      "grad_norm": 0.2354292869567871,
+      "learning_rate": 0.001,
+      "loss": 1.8365,
+      "step": 17624
+    },
+    {
+      "epoch": 0.7456214569760555,
+      "grad_norm": 0.2139493077993393,
+      "learning_rate": 0.001,
+      "loss": 2.3974,
+      "step": 17625
+    },
+    {
+      "epoch": 0.7456637617395718,
+      "grad_norm": 0.21422770619392395,
+      "learning_rate": 0.001,
+      "loss": 3.0719,
+      "step": 17626
+    },
+    {
+      "epoch": 0.7457060665030882,
+      "grad_norm": 0.28964513540267944,
+      "learning_rate": 0.001,
+      "loss": 1.8213,
+      "step": 17627
+    },
+    {
+      "epoch": 0.7457483712666046,
+      "grad_norm": 0.148183211684227,
+      "learning_rate": 0.001,
+      "loss": 1.9446,
+      "step": 17628
+    },
+    {
+      "epoch": 0.745790676030121,
+      "grad_norm": 0.18475840985774994,
+      "learning_rate": 0.001,
+      "loss": 2.2334,
+      "step": 17629
+    },
+    {
+      "epoch": 0.7458329807936374,
+      "grad_norm": 0.15556128323078156,
+      "learning_rate": 0.001,
+      "loss": 2.723,
+      "step": 17630
+    },
+    {
+      "epoch": 0.7458752855571538,
+      "grad_norm": 0.23832905292510986,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 17631
+    },
+    {
+      "epoch": 0.7459175903206701,
+      "grad_norm": 0.4453236162662506,
+      "learning_rate": 0.001,
+      "loss": 2.9713,
+      "step": 17632
+    },
+    {
+      "epoch": 0.7459598950841865,
+      "grad_norm": 0.17969189584255219,
+      "learning_rate": 0.001,
+      "loss": 2.0573,
+      "step": 17633
+    },
+    {
+      "epoch": 0.7460021998477029,
+      "grad_norm": 0.32808932662010193,
+      "learning_rate": 0.001,
+      "loss": 1.9398,
+      "step": 17634
+    },
+    {
+      "epoch": 0.7460445046112192,
+      "grad_norm": 0.1659063994884491,
+      "learning_rate": 0.001,
+      "loss": 1.8874,
+      "step": 17635
+    },
+    {
+      "epoch": 0.7460868093747356,
+      "grad_norm": 0.14386507868766785,
+      "learning_rate": 0.001,
+      "loss": 2.0049,
+      "step": 17636
+    },
+    {
+      "epoch": 0.746129114138252,
+      "grad_norm": 0.174132838845253,
+      "learning_rate": 0.001,
+      "loss": 2.1084,
+      "step": 17637
+    },
+    {
+      "epoch": 0.7461714189017683,
+      "grad_norm": 15.538298606872559,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 17638
+    },
+    {
+      "epoch": 0.7462137236652847,
+      "grad_norm": 2.191509962081909,
+      "learning_rate": 0.001,
+      "loss": 2.0302,
+      "step": 17639
+    },
+    {
+      "epoch": 0.7462560284288011,
+      "grad_norm": 0.2102295458316803,
+      "learning_rate": 0.001,
+      "loss": 2.0219,
+      "step": 17640
+    },
+    {
+      "epoch": 0.7462983331923174,
+      "grad_norm": 0.4853608310222626,
+      "learning_rate": 0.001,
+      "loss": 1.6501,
+      "step": 17641
+    },
+    {
+      "epoch": 0.7463406379558338,
+      "grad_norm": 0.36355969309806824,
+      "learning_rate": 0.001,
+      "loss": 1.8691,
+      "step": 17642
+    },
+    {
+      "epoch": 0.7463829427193502,
+      "grad_norm": 0.1756628006696701,
+      "learning_rate": 0.001,
+      "loss": 2.27,
+      "step": 17643
+    },
+    {
+      "epoch": 0.7464252474828665,
+      "grad_norm": 0.1729697287082672,
+      "learning_rate": 0.001,
+      "loss": 3.3898,
+      "step": 17644
+    },
+    {
+      "epoch": 0.746467552246383,
+      "grad_norm": 0.17120568454265594,
+      "learning_rate": 0.001,
+      "loss": 2.0535,
+      "step": 17645
+    },
+    {
+      "epoch": 0.7465098570098994,
+      "grad_norm": 0.18185213208198547,
+      "learning_rate": 0.001,
+      "loss": 1.8239,
+      "step": 17646
+    },
+    {
+      "epoch": 0.7465521617734157,
+      "grad_norm": 0.13639383018016815,
+      "learning_rate": 0.001,
+      "loss": 1.7961,
+      "step": 17647
+    },
+    {
+      "epoch": 0.7465944665369321,
+      "grad_norm": 2.5754916667938232,
+      "learning_rate": 0.001,
+      "loss": 1.9791,
+      "step": 17648
+    },
+    {
+      "epoch": 0.7466367713004485,
+      "grad_norm": 0.13714253902435303,
+      "learning_rate": 0.001,
+      "loss": 1.497,
+      "step": 17649
+    },
+    {
+      "epoch": 0.7466790760639648,
+      "grad_norm": 1.1888046264648438,
+      "learning_rate": 0.001,
+      "loss": 1.7414,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7467213808274812,
+      "grad_norm": 0.1794203370809555,
+      "learning_rate": 0.001,
+      "loss": 1.3829,
+      "step": 17651
+    },
+    {
+      "epoch": 0.7467636855909976,
+      "grad_norm": 2.2134206295013428,
+      "learning_rate": 0.001,
+      "loss": 1.8633,
+      "step": 17652
+    },
+    {
+      "epoch": 0.7468059903545139,
+      "grad_norm": 0.1946529895067215,
+      "learning_rate": 0.001,
+      "loss": 2.1527,
+      "step": 17653
+    },
+    {
+      "epoch": 0.7468482951180303,
+      "grad_norm": 0.18843619525432587,
+      "learning_rate": 0.001,
+      "loss": 1.8838,
+      "step": 17654
+    },
+    {
+      "epoch": 0.7468905998815467,
+      "grad_norm": 0.30551382899284363,
+      "learning_rate": 0.001,
+      "loss": 2.2091,
+      "step": 17655
+    },
+    {
+      "epoch": 0.746932904645063,
+      "grad_norm": 0.18397283554077148,
+      "learning_rate": 0.001,
+      "loss": 1.8819,
+      "step": 17656
+    },
+    {
+      "epoch": 0.7469752094085794,
+      "grad_norm": 0.4012526273727417,
+      "learning_rate": 0.001,
+      "loss": 2.8932,
+      "step": 17657
+    },
+    {
+      "epoch": 0.7470175141720958,
+      "grad_norm": 0.1578911393880844,
+      "learning_rate": 0.001,
+      "loss": 2.5474,
+      "step": 17658
+    },
+    {
+      "epoch": 0.7470598189356121,
+      "grad_norm": 0.16640107333660126,
+      "learning_rate": 0.001,
+      "loss": 1.9356,
+      "step": 17659
+    },
+    {
+      "epoch": 0.7471021236991285,
+      "grad_norm": 9.117905616760254,
+      "learning_rate": 0.001,
+      "loss": 2.1185,
+      "step": 17660
+    },
+    {
+      "epoch": 0.7471444284626448,
+      "grad_norm": 0.7601824998855591,
+      "learning_rate": 0.001,
+      "loss": 2.4204,
+      "step": 17661
+    },
+    {
+      "epoch": 0.7471867332261612,
+      "grad_norm": 0.2028346061706543,
+      "learning_rate": 0.001,
+      "loss": 2.3328,
+      "step": 17662
+    },
+    {
+      "epoch": 0.7472290379896777,
+      "grad_norm": 0.4436508119106293,
+      "learning_rate": 0.001,
+      "loss": 2.5131,
+      "step": 17663
+    },
+    {
+      "epoch": 0.747271342753194,
+      "grad_norm": 0.19675542414188385,
+      "learning_rate": 0.001,
+      "loss": 2.5793,
+      "step": 17664
+    },
+    {
+      "epoch": 0.7473136475167104,
+      "grad_norm": 0.22706976532936096,
+      "learning_rate": 0.001,
+      "loss": 2.364,
+      "step": 17665
+    },
+    {
+      "epoch": 0.7473559522802268,
+      "grad_norm": 0.19858993589878082,
+      "learning_rate": 0.001,
+      "loss": 2.0174,
+      "step": 17666
+    },
+    {
+      "epoch": 0.7473982570437431,
+      "grad_norm": 0.20804426074028015,
+      "learning_rate": 0.001,
+      "loss": 1.8808,
+      "step": 17667
+    },
+    {
+      "epoch": 0.7474405618072595,
+      "grad_norm": 0.21756567060947418,
+      "learning_rate": 0.001,
+      "loss": 1.7364,
+      "step": 17668
+    },
+    {
+      "epoch": 0.7474828665707759,
+      "grad_norm": 17.11835479736328,
+      "learning_rate": 0.001,
+      "loss": 2.8228,
+      "step": 17669
+    },
+    {
+      "epoch": 0.7475251713342922,
+      "grad_norm": 0.1679975390434265,
+      "learning_rate": 0.001,
+      "loss": 2.387,
+      "step": 17670
+    },
+    {
+      "epoch": 0.7475674760978086,
+      "grad_norm": 0.16715805232524872,
+      "learning_rate": 0.001,
+      "loss": 2.7919,
+      "step": 17671
+    },
+    {
+      "epoch": 0.747609780861325,
+      "grad_norm": 0.17563515901565552,
+      "learning_rate": 0.001,
+      "loss": 2.2313,
+      "step": 17672
+    },
+    {
+      "epoch": 0.7476520856248413,
+      "grad_norm": 0.16387246549129486,
+      "learning_rate": 0.001,
+      "loss": 1.6362,
+      "step": 17673
+    },
+    {
+      "epoch": 0.7476943903883577,
+      "grad_norm": 0.1781141310930252,
+      "learning_rate": 0.001,
+      "loss": 4.1604,
+      "step": 17674
+    },
+    {
+      "epoch": 0.7477366951518741,
+      "grad_norm": 0.17480358481407166,
+      "learning_rate": 0.001,
+      "loss": 1.872,
+      "step": 17675
+    },
+    {
+      "epoch": 0.7477789999153904,
+      "grad_norm": 0.192658469080925,
+      "learning_rate": 0.001,
+      "loss": 1.8116,
+      "step": 17676
+    },
+    {
+      "epoch": 0.7478213046789068,
+      "grad_norm": 0.2537101209163666,
+      "learning_rate": 0.001,
+      "loss": 3.9484,
+      "step": 17677
+    },
+    {
+      "epoch": 0.7478636094424232,
+      "grad_norm": 1.3975545167922974,
+      "learning_rate": 0.001,
+      "loss": 2.4941,
+      "step": 17678
+    },
+    {
+      "epoch": 0.7479059142059395,
+      "grad_norm": 0.21007265150547028,
+      "learning_rate": 0.001,
+      "loss": 2.1936,
+      "step": 17679
+    },
+    {
+      "epoch": 0.747948218969456,
+      "grad_norm": 0.20976990461349487,
+      "learning_rate": 0.001,
+      "loss": 2.3444,
+      "step": 17680
+    },
+    {
+      "epoch": 0.7479905237329724,
+      "grad_norm": 0.28432410955429077,
+      "learning_rate": 0.001,
+      "loss": 2.4682,
+      "step": 17681
+    },
+    {
+      "epoch": 0.7480328284964887,
+      "grad_norm": 0.17756503820419312,
+      "learning_rate": 0.001,
+      "loss": 3.1183,
+      "step": 17682
+    },
+    {
+      "epoch": 0.7480751332600051,
+      "grad_norm": 0.148085817694664,
+      "learning_rate": 0.001,
+      "loss": 1.5892,
+      "step": 17683
+    },
+    {
+      "epoch": 0.7481174380235215,
+      "grad_norm": 0.15588833391666412,
+      "learning_rate": 0.001,
+      "loss": 2.1042,
+      "step": 17684
+    },
+    {
+      "epoch": 0.7481597427870378,
+      "grad_norm": 0.16696330904960632,
+      "learning_rate": 0.001,
+      "loss": 1.4378,
+      "step": 17685
+    },
+    {
+      "epoch": 0.7482020475505542,
+      "grad_norm": 0.14883705973625183,
+      "learning_rate": 0.001,
+      "loss": 1.5751,
+      "step": 17686
+    },
+    {
+      "epoch": 0.7482443523140706,
+      "grad_norm": 0.8387157320976257,
+      "learning_rate": 0.001,
+      "loss": 4.0032,
+      "step": 17687
+    },
+    {
+      "epoch": 0.7482866570775869,
+      "grad_norm": 0.16816678643226624,
+      "learning_rate": 0.001,
+      "loss": 2.0232,
+      "step": 17688
+    },
+    {
+      "epoch": 0.7483289618411033,
+      "grad_norm": 0.17581240832805634,
+      "learning_rate": 0.001,
+      "loss": 2.221,
+      "step": 17689
+    },
+    {
+      "epoch": 0.7483712666046197,
+      "grad_norm": 0.20603400468826294,
+      "learning_rate": 0.001,
+      "loss": 1.6247,
+      "step": 17690
+    },
+    {
+      "epoch": 0.748413571368136,
+      "grad_norm": 1.0062761306762695,
+      "learning_rate": 0.001,
+      "loss": 2.0145,
+      "step": 17691
+    },
+    {
+      "epoch": 0.7484558761316524,
+      "grad_norm": 0.1723836064338684,
+      "learning_rate": 0.001,
+      "loss": 2.6105,
+      "step": 17692
+    },
+    {
+      "epoch": 0.7484981808951688,
+      "grad_norm": 0.1439078003168106,
+      "learning_rate": 0.001,
+      "loss": 2.2332,
+      "step": 17693
+    },
+    {
+      "epoch": 0.7485404856586851,
+      "grad_norm": 0.19162747263908386,
+      "learning_rate": 0.001,
+      "loss": 2.2636,
+      "step": 17694
+    },
+    {
+      "epoch": 0.7485827904222015,
+      "grad_norm": 0.2729785740375519,
+      "learning_rate": 0.001,
+      "loss": 2.4435,
+      "step": 17695
+    },
+    {
+      "epoch": 0.748625095185718,
+      "grad_norm": 0.9558886885643005,
+      "learning_rate": 0.001,
+      "loss": 2.0409,
+      "step": 17696
+    },
+    {
+      "epoch": 0.7486673999492343,
+      "grad_norm": 0.1588534116744995,
+      "learning_rate": 0.001,
+      "loss": 2.9943,
+      "step": 17697
+    },
+    {
+      "epoch": 0.7487097047127507,
+      "grad_norm": 0.21727414429187775,
+      "learning_rate": 0.001,
+      "loss": 1.9491,
+      "step": 17698
+    },
+    {
+      "epoch": 0.7487520094762671,
+      "grad_norm": 0.18474645912647247,
+      "learning_rate": 0.001,
+      "loss": 2.6354,
+      "step": 17699
+    },
+    {
+      "epoch": 0.7487943142397834,
+      "grad_norm": 0.23161320388317108,
+      "learning_rate": 0.001,
+      "loss": 2.2559,
+      "step": 17700
+    },
+    {
+      "epoch": 0.7488366190032998,
+      "grad_norm": 0.16369740664958954,
+      "learning_rate": 0.001,
+      "loss": 1.8356,
+      "step": 17701
+    },
+    {
+      "epoch": 0.7488789237668162,
+      "grad_norm": 0.17817635834217072,
+      "learning_rate": 0.001,
+      "loss": 1.8854,
+      "step": 17702
+    },
+    {
+      "epoch": 0.7489212285303325,
+      "grad_norm": 0.18831977248191833,
+      "learning_rate": 0.001,
+      "loss": 1.8879,
+      "step": 17703
+    },
+    {
+      "epoch": 0.7489635332938489,
+      "grad_norm": 0.4837913513183594,
+      "learning_rate": 0.001,
+      "loss": 1.9171,
+      "step": 17704
+    },
+    {
+      "epoch": 0.7490058380573653,
+      "grad_norm": 0.5666935443878174,
+      "learning_rate": 0.001,
+      "loss": 3.624,
+      "step": 17705
+    },
+    {
+      "epoch": 0.7490481428208816,
+      "grad_norm": 0.16191402077674866,
+      "learning_rate": 0.001,
+      "loss": 3.664,
+      "step": 17706
+    },
+    {
+      "epoch": 0.749090447584398,
+      "grad_norm": 0.16512976586818695,
+      "learning_rate": 0.001,
+      "loss": 1.9419,
+      "step": 17707
+    },
+    {
+      "epoch": 0.7491327523479143,
+      "grad_norm": 0.15622422099113464,
+      "learning_rate": 0.001,
+      "loss": 2.814,
+      "step": 17708
+    },
+    {
+      "epoch": 0.7491750571114307,
+      "grad_norm": 0.17258132994174957,
+      "learning_rate": 0.001,
+      "loss": 2.0253,
+      "step": 17709
+    },
+    {
+      "epoch": 0.7492173618749471,
+      "grad_norm": 0.19456447660923004,
+      "learning_rate": 0.001,
+      "loss": 2.9622,
+      "step": 17710
+    },
+    {
+      "epoch": 0.7492596666384634,
+      "grad_norm": 2.774994134902954,
+      "learning_rate": 0.001,
+      "loss": 2.2187,
+      "step": 17711
+    },
+    {
+      "epoch": 0.7493019714019798,
+      "grad_norm": 0.17640100419521332,
+      "learning_rate": 0.001,
+      "loss": 1.976,
+      "step": 17712
+    },
+    {
+      "epoch": 0.7493442761654963,
+      "grad_norm": 0.17278911173343658,
+      "learning_rate": 0.001,
+      "loss": 2.401,
+      "step": 17713
+    },
+    {
+      "epoch": 0.7493865809290126,
+      "grad_norm": 0.20537172257900238,
+      "learning_rate": 0.001,
+      "loss": 3.0728,
+      "step": 17714
+    },
+    {
+      "epoch": 0.749428885692529,
+      "grad_norm": 0.35584142804145813,
+      "learning_rate": 0.001,
+      "loss": 3.3195,
+      "step": 17715
+    },
+    {
+      "epoch": 0.7494711904560454,
+      "grad_norm": 0.20729830861091614,
+      "learning_rate": 0.001,
+      "loss": 2.5783,
+      "step": 17716
+    },
+    {
+      "epoch": 0.7495134952195617,
+      "grad_norm": 0.1861492097377777,
+      "learning_rate": 0.001,
+      "loss": 1.9174,
+      "step": 17717
+    },
+    {
+      "epoch": 0.7495557999830781,
+      "grad_norm": 0.18184538185596466,
+      "learning_rate": 0.001,
+      "loss": 2.2683,
+      "step": 17718
+    },
+    {
+      "epoch": 0.7495981047465945,
+      "grad_norm": 0.15846717357635498,
+      "learning_rate": 0.001,
+      "loss": 1.5975,
+      "step": 17719
+    },
+    {
+      "epoch": 0.7496404095101108,
+      "grad_norm": 0.15116605162620544,
+      "learning_rate": 0.001,
+      "loss": 1.2862,
+      "step": 17720
+    },
+    {
+      "epoch": 0.7496827142736272,
+      "grad_norm": 0.15482908487319946,
+      "learning_rate": 0.001,
+      "loss": 2.0605,
+      "step": 17721
+    },
+    {
+      "epoch": 0.7497250190371436,
+      "grad_norm": 0.1819373369216919,
+      "learning_rate": 0.001,
+      "loss": 2.8881,
+      "step": 17722
+    },
+    {
+      "epoch": 0.7497673238006599,
+      "grad_norm": 0.16253606975078583,
+      "learning_rate": 0.001,
+      "loss": 1.448,
+      "step": 17723
+    },
+    {
+      "epoch": 0.7498096285641763,
+      "grad_norm": 0.15486584603786469,
+      "learning_rate": 0.001,
+      "loss": 2.7328,
+      "step": 17724
+    },
+    {
+      "epoch": 0.7498519333276927,
+      "grad_norm": 0.14722788333892822,
+      "learning_rate": 0.001,
+      "loss": 2.3268,
+      "step": 17725
+    },
+    {
+      "epoch": 0.749894238091209,
+      "grad_norm": 0.928293764591217,
+      "learning_rate": 0.001,
+      "loss": 2.1948,
+      "step": 17726
+    },
+    {
+      "epoch": 0.7499365428547254,
+      "grad_norm": 0.14269466698169708,
+      "learning_rate": 0.001,
+      "loss": 1.9547,
+      "step": 17727
+    },
+    {
+      "epoch": 0.7499788476182419,
+      "grad_norm": 0.7089377641677856,
+      "learning_rate": 0.001,
+      "loss": 2.996,
+      "step": 17728
+    },
+    {
+      "epoch": 0.7500211523817581,
+      "grad_norm": 2.6846911907196045,
+      "learning_rate": 0.001,
+      "loss": 1.8741,
+      "step": 17729
+    },
+    {
+      "epoch": 0.7500634571452746,
+      "grad_norm": 0.13764113187789917,
+      "learning_rate": 0.001,
+      "loss": 2.2695,
+      "step": 17730
+    },
+    {
+      "epoch": 0.750105761908791,
+      "grad_norm": 0.16455744206905365,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 17731
+    },
+    {
+      "epoch": 0.7501480666723073,
+      "grad_norm": 1.5582802295684814,
+      "learning_rate": 0.001,
+      "loss": 2.6966,
+      "step": 17732
+    },
+    {
+      "epoch": 0.7501903714358237,
+      "grad_norm": 0.15412558615207672,
+      "learning_rate": 0.001,
+      "loss": 1.8025,
+      "step": 17733
+    },
+    {
+      "epoch": 0.7502326761993401,
+      "grad_norm": 0.17679987847805023,
+      "learning_rate": 0.001,
+      "loss": 3.0607,
+      "step": 17734
+    },
+    {
+      "epoch": 0.7502749809628564,
+      "grad_norm": 0.2634994685649872,
+      "learning_rate": 0.001,
+      "loss": 2.326,
+      "step": 17735
+    },
+    {
+      "epoch": 0.7503172857263728,
+      "grad_norm": 0.18038220703601837,
+      "learning_rate": 0.001,
+      "loss": 1.9938,
+      "step": 17736
+    },
+    {
+      "epoch": 0.7503595904898892,
+      "grad_norm": 0.1706598848104477,
+      "learning_rate": 0.001,
+      "loss": 2.8744,
+      "step": 17737
+    },
+    {
+      "epoch": 0.7504018952534055,
+      "grad_norm": 0.16422058641910553,
+      "learning_rate": 0.001,
+      "loss": 2.5134,
+      "step": 17738
+    },
+    {
+      "epoch": 0.7504442000169219,
+      "grad_norm": 0.1558268815279007,
+      "learning_rate": 0.001,
+      "loss": 2.8615,
+      "step": 17739
+    },
+    {
+      "epoch": 0.7504865047804383,
+      "grad_norm": 0.15481281280517578,
+      "learning_rate": 0.001,
+      "loss": 2.537,
+      "step": 17740
+    },
+    {
+      "epoch": 0.7505288095439546,
+      "grad_norm": 0.1353859305381775,
+      "learning_rate": 0.001,
+      "loss": 1.9273,
+      "step": 17741
+    },
+    {
+      "epoch": 0.750571114307471,
+      "grad_norm": 0.1670721471309662,
+      "learning_rate": 0.001,
+      "loss": 2.0898,
+      "step": 17742
+    },
+    {
+      "epoch": 0.7506134190709874,
+      "grad_norm": 0.17640924453735352,
+      "learning_rate": 0.001,
+      "loss": 1.8176,
+      "step": 17743
+    },
+    {
+      "epoch": 0.7506557238345037,
+      "grad_norm": 0.18629953265190125,
+      "learning_rate": 0.001,
+      "loss": 2.1797,
+      "step": 17744
+    },
+    {
+      "epoch": 0.7506980285980202,
+      "grad_norm": 0.2514755129814148,
+      "learning_rate": 0.001,
+      "loss": 3.1144,
+      "step": 17745
+    },
+    {
+      "epoch": 0.7507403333615366,
+      "grad_norm": 2.408249855041504,
+      "learning_rate": 0.001,
+      "loss": 1.7701,
+      "step": 17746
+    },
+    {
+      "epoch": 0.7507826381250529,
+      "grad_norm": 1.3865883350372314,
+      "learning_rate": 0.001,
+      "loss": 2.3281,
+      "step": 17747
+    },
+    {
+      "epoch": 0.7508249428885693,
+      "grad_norm": 0.16517174243927002,
+      "learning_rate": 0.001,
+      "loss": 2.4022,
+      "step": 17748
+    },
+    {
+      "epoch": 0.7508672476520857,
+      "grad_norm": 7.395416736602783,
+      "learning_rate": 0.001,
+      "loss": 2.9588,
+      "step": 17749
+    },
+    {
+      "epoch": 0.750909552415602,
+      "grad_norm": 0.24712392687797546,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7509518571791184,
+      "grad_norm": 0.14635969698429108,
+      "learning_rate": 0.001,
+      "loss": 1.5613,
+      "step": 17751
+    },
+    {
+      "epoch": 0.7509941619426347,
+      "grad_norm": 0.16035181283950806,
+      "learning_rate": 0.001,
+      "loss": 2.7492,
+      "step": 17752
+    },
+    {
+      "epoch": 0.7510364667061511,
+      "grad_norm": 0.17972306907176971,
+      "learning_rate": 0.001,
+      "loss": 2.2771,
+      "step": 17753
+    },
+    {
+      "epoch": 0.7510787714696675,
+      "grad_norm": 0.2305939793586731,
+      "learning_rate": 0.001,
+      "loss": 1.5971,
+      "step": 17754
+    },
+    {
+      "epoch": 0.7511210762331838,
+      "grad_norm": 0.21389088034629822,
+      "learning_rate": 0.001,
+      "loss": 2.2783,
+      "step": 17755
+    },
+    {
+      "epoch": 0.7511633809967002,
+      "grad_norm": 0.17007265985012054,
+      "learning_rate": 0.001,
+      "loss": 2.0106,
+      "step": 17756
+    },
+    {
+      "epoch": 0.7512056857602166,
+      "grad_norm": 0.18640732765197754,
+      "learning_rate": 0.001,
+      "loss": 1.8577,
+      "step": 17757
+    },
+    {
+      "epoch": 0.7512479905237329,
+      "grad_norm": 15.663119316101074,
+      "learning_rate": 0.001,
+      "loss": 2.9581,
+      "step": 17758
+    },
+    {
+      "epoch": 0.7512902952872493,
+      "grad_norm": 4.945131778717041,
+      "learning_rate": 0.001,
+      "loss": 2.8615,
+      "step": 17759
+    },
+    {
+      "epoch": 0.7513326000507657,
+      "grad_norm": 0.16905252635478973,
+      "learning_rate": 0.001,
+      "loss": 2.2082,
+      "step": 17760
+    },
+    {
+      "epoch": 0.751374904814282,
+      "grad_norm": 0.18479637801647186,
+      "learning_rate": 0.001,
+      "loss": 1.8339,
+      "step": 17761
+    },
+    {
+      "epoch": 0.7514172095777985,
+      "grad_norm": 0.16893810033798218,
+      "learning_rate": 0.001,
+      "loss": 3.0538,
+      "step": 17762
+    },
+    {
+      "epoch": 0.7514595143413149,
+      "grad_norm": 0.4891713559627533,
+      "learning_rate": 0.001,
+      "loss": 2.1883,
+      "step": 17763
+    },
+    {
+      "epoch": 0.7515018191048312,
+      "grad_norm": 2.8719229698181152,
+      "learning_rate": 0.001,
+      "loss": 2.5028,
+      "step": 17764
+    },
+    {
+      "epoch": 0.7515441238683476,
+      "grad_norm": 0.335845947265625,
+      "learning_rate": 0.001,
+      "loss": 1.8828,
+      "step": 17765
+    },
+    {
+      "epoch": 0.751586428631864,
+      "grad_norm": 0.17647729814052582,
+      "learning_rate": 0.001,
+      "loss": 1.4358,
+      "step": 17766
+    },
+    {
+      "epoch": 0.7516287333953803,
+      "grad_norm": 38.300811767578125,
+      "learning_rate": 0.001,
+      "loss": 2.461,
+      "step": 17767
+    },
+    {
+      "epoch": 0.7516710381588967,
+      "grad_norm": 0.25352421402931213,
+      "learning_rate": 0.001,
+      "loss": 2.5402,
+      "step": 17768
+    },
+    {
+      "epoch": 0.7517133429224131,
+      "grad_norm": 0.1729395091533661,
+      "learning_rate": 0.001,
+      "loss": 1.8336,
+      "step": 17769
+    },
+    {
+      "epoch": 0.7517556476859294,
+      "grad_norm": 0.2317625731229782,
+      "learning_rate": 0.001,
+      "loss": 2.022,
+      "step": 17770
+    },
+    {
+      "epoch": 0.7517979524494458,
+      "grad_norm": 23.282867431640625,
+      "learning_rate": 0.001,
+      "loss": 3.9744,
+      "step": 17771
+    },
+    {
+      "epoch": 0.7518402572129622,
+      "grad_norm": 0.25208935141563416,
+      "learning_rate": 0.001,
+      "loss": 2.7925,
+      "step": 17772
+    },
+    {
+      "epoch": 0.7518825619764785,
+      "grad_norm": 0.16493536531925201,
+      "learning_rate": 0.001,
+      "loss": 1.7565,
+      "step": 17773
+    },
+    {
+      "epoch": 0.7519248667399949,
+      "grad_norm": 0.8458042144775391,
+      "learning_rate": 0.001,
+      "loss": 2.9931,
+      "step": 17774
+    },
+    {
+      "epoch": 0.7519671715035113,
+      "grad_norm": 0.22725796699523926,
+      "learning_rate": 0.001,
+      "loss": 2.1438,
+      "step": 17775
+    },
+    {
+      "epoch": 0.7520094762670276,
+      "grad_norm": 0.15165364742279053,
+      "learning_rate": 0.001,
+      "loss": 1.9869,
+      "step": 17776
+    },
+    {
+      "epoch": 0.752051781030544,
+      "grad_norm": 0.1665864735841751,
+      "learning_rate": 0.001,
+      "loss": 1.4288,
+      "step": 17777
+    },
+    {
+      "epoch": 0.7520940857940605,
+      "grad_norm": 0.20887833833694458,
+      "learning_rate": 0.001,
+      "loss": 2.0971,
+      "step": 17778
+    },
+    {
+      "epoch": 0.7521363905575768,
+      "grad_norm": 0.6601106524467468,
+      "learning_rate": 0.001,
+      "loss": 1.7179,
+      "step": 17779
+    },
+    {
+      "epoch": 0.7521786953210932,
+      "grad_norm": 1.797553300857544,
+      "learning_rate": 0.001,
+      "loss": 1.86,
+      "step": 17780
+    },
+    {
+      "epoch": 0.7522210000846096,
+      "grad_norm": 0.6821169853210449,
+      "learning_rate": 0.001,
+      "loss": 1.9908,
+      "step": 17781
+    },
+    {
+      "epoch": 0.7522633048481259,
+      "grad_norm": 0.24313047528266907,
+      "learning_rate": 0.001,
+      "loss": 1.7132,
+      "step": 17782
+    },
+    {
+      "epoch": 0.7523056096116423,
+      "grad_norm": 0.16890186071395874,
+      "learning_rate": 0.001,
+      "loss": 2.7889,
+      "step": 17783
+    },
+    {
+      "epoch": 0.7523479143751587,
+      "grad_norm": 0.3041240870952606,
+      "learning_rate": 0.001,
+      "loss": 1.5615,
+      "step": 17784
+    },
+    {
+      "epoch": 0.752390219138675,
+      "grad_norm": 0.17701970040798187,
+      "learning_rate": 0.001,
+      "loss": 1.9374,
+      "step": 17785
+    },
+    {
+      "epoch": 0.7524325239021914,
+      "grad_norm": 0.16033196449279785,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 17786
+    },
+    {
+      "epoch": 0.7524748286657078,
+      "grad_norm": 0.18691828846931458,
+      "learning_rate": 0.001,
+      "loss": 2.0472,
+      "step": 17787
+    },
+    {
+      "epoch": 0.7525171334292241,
+      "grad_norm": 0.20954419672489166,
+      "learning_rate": 0.001,
+      "loss": 1.2313,
+      "step": 17788
+    },
+    {
+      "epoch": 0.7525594381927405,
+      "grad_norm": 0.17444777488708496,
+      "learning_rate": 0.001,
+      "loss": 2.1319,
+      "step": 17789
+    },
+    {
+      "epoch": 0.7526017429562569,
+      "grad_norm": 0.1553867757320404,
+      "learning_rate": 0.001,
+      "loss": 2.1579,
+      "step": 17790
+    },
+    {
+      "epoch": 0.7526440477197732,
+      "grad_norm": 0.16180811822414398,
+      "learning_rate": 0.001,
+      "loss": 1.8497,
+      "step": 17791
+    },
+    {
+      "epoch": 0.7526863524832896,
+      "grad_norm": 0.16590484976768494,
+      "learning_rate": 0.001,
+      "loss": 1.7005,
+      "step": 17792
+    },
+    {
+      "epoch": 0.752728657246806,
+      "grad_norm": 0.161569282412529,
+      "learning_rate": 0.001,
+      "loss": 1.9812,
+      "step": 17793
+    },
+    {
+      "epoch": 0.7527709620103223,
+      "grad_norm": 0.1726117730140686,
+      "learning_rate": 0.001,
+      "loss": 1.6706,
+      "step": 17794
+    },
+    {
+      "epoch": 0.7528132667738388,
+      "grad_norm": 1.3740065097808838,
+      "learning_rate": 0.001,
+      "loss": 3.4836,
+      "step": 17795
+    },
+    {
+      "epoch": 0.7528555715373552,
+      "grad_norm": 0.20311777293682098,
+      "learning_rate": 0.001,
+      "loss": 3.3955,
+      "step": 17796
+    },
+    {
+      "epoch": 0.7528978763008715,
+      "grad_norm": 0.1724340319633484,
+      "learning_rate": 0.001,
+      "loss": 2.1475,
+      "step": 17797
+    },
+    {
+      "epoch": 0.7529401810643879,
+      "grad_norm": 0.15523028373718262,
+      "learning_rate": 0.001,
+      "loss": 2.4205,
+      "step": 17798
+    },
+    {
+      "epoch": 0.7529824858279042,
+      "grad_norm": 0.16621935367584229,
+      "learning_rate": 0.001,
+      "loss": 2.1048,
+      "step": 17799
+    },
+    {
+      "epoch": 0.7530247905914206,
+      "grad_norm": 0.1458512395620346,
+      "learning_rate": 0.001,
+      "loss": 2.0226,
+      "step": 17800
+    },
+    {
+      "epoch": 0.753067095354937,
+      "grad_norm": 1.5626475811004639,
+      "learning_rate": 0.001,
+      "loss": 1.4764,
+      "step": 17801
+    },
+    {
+      "epoch": 0.7531094001184533,
+      "grad_norm": 0.17712949216365814,
+      "learning_rate": 0.001,
+      "loss": 2.0474,
+      "step": 17802
+    },
+    {
+      "epoch": 0.7531517048819697,
+      "grad_norm": 0.184153750538826,
+      "learning_rate": 0.001,
+      "loss": 1.932,
+      "step": 17803
+    },
+    {
+      "epoch": 0.7531940096454861,
+      "grad_norm": 0.17124585807323456,
+      "learning_rate": 0.001,
+      "loss": 1.8677,
+      "step": 17804
+    },
+    {
+      "epoch": 0.7532363144090024,
+      "grad_norm": 0.18082095682621002,
+      "learning_rate": 0.001,
+      "loss": 2.0149,
+      "step": 17805
+    },
+    {
+      "epoch": 0.7532786191725188,
+      "grad_norm": 0.4178064465522766,
+      "learning_rate": 0.001,
+      "loss": 2.1771,
+      "step": 17806
+    },
+    {
+      "epoch": 0.7533209239360352,
+      "grad_norm": 0.16732126474380493,
+      "learning_rate": 0.001,
+      "loss": 2.033,
+      "step": 17807
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.6002020239830017,
+      "learning_rate": 0.001,
+      "loss": 2.7392,
+      "step": 17808
+    },
+    {
+      "epoch": 0.7534055334630679,
+      "grad_norm": 0.1773737519979477,
+      "learning_rate": 0.001,
+      "loss": 1.8149,
+      "step": 17809
+    },
+    {
+      "epoch": 0.7534478382265843,
+      "grad_norm": 1.7740546464920044,
+      "learning_rate": 0.001,
+      "loss": 2.1816,
+      "step": 17810
+    },
+    {
+      "epoch": 0.7534901429901006,
+      "grad_norm": 0.6229245066642761,
+      "learning_rate": 0.001,
+      "loss": 1.4621,
+      "step": 17811
+    },
+    {
+      "epoch": 0.753532447753617,
+      "grad_norm": 1.3649619817733765,
+      "learning_rate": 0.001,
+      "loss": 1.8576,
+      "step": 17812
+    },
+    {
+      "epoch": 0.7535747525171335,
+      "grad_norm": 0.14693179726600647,
+      "learning_rate": 0.001,
+      "loss": 1.8112,
+      "step": 17813
+    },
+    {
+      "epoch": 0.7536170572806498,
+      "grad_norm": 0.5923940539360046,
+      "learning_rate": 0.001,
+      "loss": 2.9757,
+      "step": 17814
+    },
+    {
+      "epoch": 0.7536593620441662,
+      "grad_norm": 0.2543085217475891,
+      "learning_rate": 0.001,
+      "loss": 3.2552,
+      "step": 17815
+    },
+    {
+      "epoch": 0.7537016668076826,
+      "grad_norm": 0.1711706966161728,
+      "learning_rate": 0.001,
+      "loss": 1.7678,
+      "step": 17816
+    },
+    {
+      "epoch": 0.7537439715711989,
+      "grad_norm": 0.1597229242324829,
+      "learning_rate": 0.001,
+      "loss": 2.3738,
+      "step": 17817
+    },
+    {
+      "epoch": 0.7537862763347153,
+      "grad_norm": 1.7806466817855835,
+      "learning_rate": 0.001,
+      "loss": 3.6486,
+      "step": 17818
+    },
+    {
+      "epoch": 0.7538285810982317,
+      "grad_norm": 0.1596599519252777,
+      "learning_rate": 0.001,
+      "loss": 3.2813,
+      "step": 17819
+    },
+    {
+      "epoch": 0.753870885861748,
+      "grad_norm": 0.4388027787208557,
+      "learning_rate": 0.001,
+      "loss": 2.2502,
+      "step": 17820
+    },
+    {
+      "epoch": 0.7539131906252644,
+      "grad_norm": 0.15271490812301636,
+      "learning_rate": 0.001,
+      "loss": 3.0926,
+      "step": 17821
+    },
+    {
+      "epoch": 0.7539554953887808,
+      "grad_norm": 0.19142945110797882,
+      "learning_rate": 0.001,
+      "loss": 1.8214,
+      "step": 17822
+    },
+    {
+      "epoch": 0.7539978001522971,
+      "grad_norm": 0.18499502539634705,
+      "learning_rate": 0.001,
+      "loss": 1.8132,
+      "step": 17823
+    },
+    {
+      "epoch": 0.7540401049158135,
+      "grad_norm": 0.21583931148052216,
+      "learning_rate": 0.001,
+      "loss": 2.8415,
+      "step": 17824
+    },
+    {
+      "epoch": 0.7540824096793299,
+      "grad_norm": 0.12202106416225433,
+      "learning_rate": 0.001,
+      "loss": 1.5671,
+      "step": 17825
+    },
+    {
+      "epoch": 0.7541247144428462,
+      "grad_norm": 0.17914700508117676,
+      "learning_rate": 0.001,
+      "loss": 2.0909,
+      "step": 17826
+    },
+    {
+      "epoch": 0.7541670192063626,
+      "grad_norm": 0.17888757586479187,
+      "learning_rate": 0.001,
+      "loss": 1.953,
+      "step": 17827
+    },
+    {
+      "epoch": 0.754209323969879,
+      "grad_norm": 0.16106991469860077,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 17828
+    },
+    {
+      "epoch": 0.7542516287333954,
+      "grad_norm": 0.13767673075199127,
+      "learning_rate": 0.001,
+      "loss": 2.0734,
+      "step": 17829
+    },
+    {
+      "epoch": 0.7542939334969118,
+      "grad_norm": 0.20107002556324005,
+      "learning_rate": 0.001,
+      "loss": 2.2336,
+      "step": 17830
+    },
+    {
+      "epoch": 0.7543362382604282,
+      "grad_norm": 0.1628139615058899,
+      "learning_rate": 0.001,
+      "loss": 2.8548,
+      "step": 17831
+    },
+    {
+      "epoch": 0.7543785430239445,
+      "grad_norm": 0.20057493448257446,
+      "learning_rate": 0.001,
+      "loss": 1.914,
+      "step": 17832
+    },
+    {
+      "epoch": 0.7544208477874609,
+      "grad_norm": 0.1650390923023224,
+      "learning_rate": 0.001,
+      "loss": 2.3508,
+      "step": 17833
+    },
+    {
+      "epoch": 0.7544631525509773,
+      "grad_norm": 0.14668241143226624,
+      "learning_rate": 0.001,
+      "loss": 2.1428,
+      "step": 17834
+    },
+    {
+      "epoch": 0.7545054573144936,
+      "grad_norm": 0.24032439291477203,
+      "learning_rate": 0.001,
+      "loss": 2.4814,
+      "step": 17835
+    },
+    {
+      "epoch": 0.75454776207801,
+      "grad_norm": 0.1620781570672989,
+      "learning_rate": 0.001,
+      "loss": 2.1885,
+      "step": 17836
+    },
+    {
+      "epoch": 0.7545900668415264,
+      "grad_norm": 0.5074538588523865,
+      "learning_rate": 0.001,
+      "loss": 2.2945,
+      "step": 17837
+    },
+    {
+      "epoch": 0.7546323716050427,
+      "grad_norm": 0.1588655561208725,
+      "learning_rate": 0.001,
+      "loss": 1.9829,
+      "step": 17838
+    },
+    {
+      "epoch": 0.7546746763685591,
+      "grad_norm": 0.13773521780967712,
+      "learning_rate": 0.001,
+      "loss": 2.0773,
+      "step": 17839
+    },
+    {
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.16354626417160034,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 17840
+    },
+    {
+      "epoch": 0.7547592858955918,
+      "grad_norm": 0.15736053884029388,
+      "learning_rate": 0.001,
+      "loss": 1.4906,
+      "step": 17841
+    },
+    {
+      "epoch": 0.7548015906591082,
+      "grad_norm": 0.13895869255065918,
+      "learning_rate": 0.001,
+      "loss": 1.3762,
+      "step": 17842
+    },
+    {
+      "epoch": 0.7548438954226245,
+      "grad_norm": 0.16241589188575745,
+      "learning_rate": 0.001,
+      "loss": 2.8513,
+      "step": 17843
+    },
+    {
+      "epoch": 0.7548862001861409,
+      "grad_norm": 0.1497146338224411,
+      "learning_rate": 0.001,
+      "loss": 1.982,
+      "step": 17844
+    },
+    {
+      "epoch": 0.7549285049496574,
+      "grad_norm": 0.6093889474868774,
+      "learning_rate": 0.001,
+      "loss": 2.6218,
+      "step": 17845
+    },
+    {
+      "epoch": 0.7549708097131737,
+      "grad_norm": 0.2828693091869354,
+      "learning_rate": 0.001,
+      "loss": 2.0783,
+      "step": 17846
+    },
+    {
+      "epoch": 0.7550131144766901,
+      "grad_norm": 0.17641054093837738,
+      "learning_rate": 0.001,
+      "loss": 2.1691,
+      "step": 17847
+    },
+    {
+      "epoch": 0.7550554192402065,
+      "grad_norm": 0.4624340534210205,
+      "learning_rate": 0.001,
+      "loss": 1.5838,
+      "step": 17848
+    },
+    {
+      "epoch": 0.7550977240037228,
+      "grad_norm": 0.18794561922550201,
+      "learning_rate": 0.001,
+      "loss": 2.8615,
+      "step": 17849
+    },
+    {
+      "epoch": 0.7551400287672392,
+      "grad_norm": 0.1454625278711319,
+      "learning_rate": 0.001,
+      "loss": 2.3494,
+      "step": 17850
+    },
+    {
+      "epoch": 0.7551823335307556,
+      "grad_norm": 0.16596491634845734,
+      "learning_rate": 0.001,
+      "loss": 2.1871,
+      "step": 17851
+    },
+    {
+      "epoch": 0.7552246382942719,
+      "grad_norm": 0.1360815316438675,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 17852
+    },
+    {
+      "epoch": 0.7552669430577883,
+      "grad_norm": 0.12390490621328354,
+      "learning_rate": 0.001,
+      "loss": 1.8553,
+      "step": 17853
+    },
+    {
+      "epoch": 0.7553092478213047,
+      "grad_norm": 0.14053763449192047,
+      "learning_rate": 0.001,
+      "loss": 1.8414,
+      "step": 17854
+    },
+    {
+      "epoch": 0.755351552584821,
+      "grad_norm": 1.1963753700256348,
+      "learning_rate": 0.001,
+      "loss": 2.295,
+      "step": 17855
+    },
+    {
+      "epoch": 0.7553938573483374,
+      "grad_norm": 0.17305585741996765,
+      "learning_rate": 0.001,
+      "loss": 2.9799,
+      "step": 17856
+    },
+    {
+      "epoch": 0.7554361621118538,
+      "grad_norm": 0.9570670127868652,
+      "learning_rate": 0.001,
+      "loss": 3.0922,
+      "step": 17857
+    },
+    {
+      "epoch": 0.7554784668753701,
+      "grad_norm": 0.1583283543586731,
+      "learning_rate": 0.001,
+      "loss": 2.4881,
+      "step": 17858
+    },
+    {
+      "epoch": 0.7555207716388865,
+      "grad_norm": 0.18101347982883453,
+      "learning_rate": 0.001,
+      "loss": 2.2682,
+      "step": 17859
+    },
+    {
+      "epoch": 0.7555630764024029,
+      "grad_norm": 0.17083586752414703,
+      "learning_rate": 0.001,
+      "loss": 2.5757,
+      "step": 17860
+    },
+    {
+      "epoch": 0.7556053811659192,
+      "grad_norm": 0.2182934582233429,
+      "learning_rate": 0.001,
+      "loss": 2.3445,
+      "step": 17861
+    },
+    {
+      "epoch": 0.7556476859294357,
+      "grad_norm": 0.13147765398025513,
+      "learning_rate": 0.001,
+      "loss": 1.6128,
+      "step": 17862
+    },
+    {
+      "epoch": 0.7556899906929521,
+      "grad_norm": 0.21211394667625427,
+      "learning_rate": 0.001,
+      "loss": 2.4639,
+      "step": 17863
+    },
+    {
+      "epoch": 0.7557322954564684,
+      "grad_norm": 0.29777416586875916,
+      "learning_rate": 0.001,
+      "loss": 2.3622,
+      "step": 17864
+    },
+    {
+      "epoch": 0.7557746002199848,
+      "grad_norm": 0.20672816038131714,
+      "learning_rate": 0.001,
+      "loss": 2.1961,
+      "step": 17865
+    },
+    {
+      "epoch": 0.7558169049835012,
+      "grad_norm": 2.0547661781311035,
+      "learning_rate": 0.001,
+      "loss": 4.3223,
+      "step": 17866
+    },
+    {
+      "epoch": 0.7558592097470175,
+      "grad_norm": 0.40489253401756287,
+      "learning_rate": 0.001,
+      "loss": 2.754,
+      "step": 17867
+    },
+    {
+      "epoch": 0.7559015145105339,
+      "grad_norm": 0.18186049163341522,
+      "learning_rate": 0.001,
+      "loss": 1.9801,
+      "step": 17868
+    },
+    {
+      "epoch": 0.7559438192740503,
+      "grad_norm": 0.1969180703163147,
+      "learning_rate": 0.001,
+      "loss": 1.7136,
+      "step": 17869
+    },
+    {
+      "epoch": 0.7559861240375666,
+      "grad_norm": 0.16497890651226044,
+      "learning_rate": 0.001,
+      "loss": 1.6463,
+      "step": 17870
+    },
+    {
+      "epoch": 0.756028428801083,
+      "grad_norm": 0.16345803439617157,
+      "learning_rate": 0.001,
+      "loss": 2.0088,
+      "step": 17871
+    },
+    {
+      "epoch": 0.7560707335645994,
+      "grad_norm": 0.2046748548746109,
+      "learning_rate": 0.001,
+      "loss": 2.7136,
+      "step": 17872
+    },
+    {
+      "epoch": 0.7561130383281157,
+      "grad_norm": 0.3788432776927948,
+      "learning_rate": 0.001,
+      "loss": 3.172,
+      "step": 17873
+    },
+    {
+      "epoch": 0.7561553430916321,
+      "grad_norm": 5.223359107971191,
+      "learning_rate": 0.001,
+      "loss": 1.9627,
+      "step": 17874
+    },
+    {
+      "epoch": 0.7561976478551485,
+      "grad_norm": 0.1771371066570282,
+      "learning_rate": 0.001,
+      "loss": 2.7354,
+      "step": 17875
+    },
+    {
+      "epoch": 0.7562399526186648,
+      "grad_norm": 0.19962891936302185,
+      "learning_rate": 0.001,
+      "loss": 2.2917,
+      "step": 17876
+    },
+    {
+      "epoch": 0.7562822573821812,
+      "grad_norm": 0.1908496767282486,
+      "learning_rate": 0.001,
+      "loss": 3.2174,
+      "step": 17877
+    },
+    {
+      "epoch": 0.7563245621456977,
+      "grad_norm": 0.15644274652004242,
+      "learning_rate": 0.001,
+      "loss": 2.0214,
+      "step": 17878
+    },
+    {
+      "epoch": 0.756366866909214,
+      "grad_norm": 0.3016762435436249,
+      "learning_rate": 0.001,
+      "loss": 1.9582,
+      "step": 17879
+    },
+    {
+      "epoch": 0.7564091716727304,
+      "grad_norm": 0.14754663407802582,
+      "learning_rate": 0.001,
+      "loss": 2.2023,
+      "step": 17880
+    },
+    {
+      "epoch": 0.7564514764362468,
+      "grad_norm": 0.14365266263484955,
+      "learning_rate": 0.001,
+      "loss": 2.4457,
+      "step": 17881
+    },
+    {
+      "epoch": 0.7564937811997631,
+      "grad_norm": 0.15732894837856293,
+      "learning_rate": 0.001,
+      "loss": 2.6365,
+      "step": 17882
+    },
+    {
+      "epoch": 0.7565360859632795,
+      "grad_norm": 0.17526346445083618,
+      "learning_rate": 0.001,
+      "loss": 2.0218,
+      "step": 17883
+    },
+    {
+      "epoch": 0.7565783907267959,
+      "grad_norm": 0.15322791039943695,
+      "learning_rate": 0.001,
+      "loss": 1.6749,
+      "step": 17884
+    },
+    {
+      "epoch": 0.7566206954903122,
+      "grad_norm": 0.24172990024089813,
+      "learning_rate": 0.001,
+      "loss": 2.5218,
+      "step": 17885
+    },
+    {
+      "epoch": 0.7566630002538286,
+      "grad_norm": 0.15460051596164703,
+      "learning_rate": 0.001,
+      "loss": 2.4132,
+      "step": 17886
+    },
+    {
+      "epoch": 0.7567053050173449,
+      "grad_norm": 0.13354411721229553,
+      "learning_rate": 0.001,
+      "loss": 2.9439,
+      "step": 17887
+    },
+    {
+      "epoch": 0.7567476097808613,
+      "grad_norm": 0.13748109340667725,
+      "learning_rate": 0.001,
+      "loss": 1.4254,
+      "step": 17888
+    },
+    {
+      "epoch": 0.7567899145443777,
+      "grad_norm": 0.24838584661483765,
+      "learning_rate": 0.001,
+      "loss": 2.0226,
+      "step": 17889
+    },
+    {
+      "epoch": 0.756832219307894,
+      "grad_norm": 0.15621478855609894,
+      "learning_rate": 0.001,
+      "loss": 3.084,
+      "step": 17890
+    },
+    {
+      "epoch": 0.7568745240714104,
+      "grad_norm": 0.543929934501648,
+      "learning_rate": 0.001,
+      "loss": 1.88,
+      "step": 17891
+    },
+    {
+      "epoch": 0.7569168288349268,
+      "grad_norm": 0.4589284062385559,
+      "learning_rate": 0.001,
+      "loss": 2.0766,
+      "step": 17892
+    },
+    {
+      "epoch": 0.7569591335984431,
+      "grad_norm": 0.4541616141796112,
+      "learning_rate": 0.001,
+      "loss": 2.6497,
+      "step": 17893
+    },
+    {
+      "epoch": 0.7570014383619595,
+      "grad_norm": 0.1465660035610199,
+      "learning_rate": 0.001,
+      "loss": 2.4708,
+      "step": 17894
+    },
+    {
+      "epoch": 0.757043743125476,
+      "grad_norm": 0.2618861198425293,
+      "learning_rate": 0.001,
+      "loss": 2.753,
+      "step": 17895
+    },
+    {
+      "epoch": 0.7570860478889923,
+      "grad_norm": 0.15626366436481476,
+      "learning_rate": 0.001,
+      "loss": 2.0046,
+      "step": 17896
+    },
+    {
+      "epoch": 0.7571283526525087,
+      "grad_norm": 0.15386773645877838,
+      "learning_rate": 0.001,
+      "loss": 2.168,
+      "step": 17897
+    },
+    {
+      "epoch": 0.7571706574160251,
+      "grad_norm": 0.15330027043819427,
+      "learning_rate": 0.001,
+      "loss": 2.3295,
+      "step": 17898
+    },
+    {
+      "epoch": 0.7572129621795414,
+      "grad_norm": 0.15845079720020294,
+      "learning_rate": 0.001,
+      "loss": 2.0764,
+      "step": 17899
+    },
+    {
+      "epoch": 0.7572552669430578,
+      "grad_norm": 0.5372272729873657,
+      "learning_rate": 0.001,
+      "loss": 2.3282,
+      "step": 17900
+    },
+    {
+      "epoch": 0.7572975717065742,
+      "grad_norm": 0.1655784249305725,
+      "learning_rate": 0.001,
+      "loss": 1.5967,
+      "step": 17901
+    },
+    {
+      "epoch": 0.7573398764700905,
+      "grad_norm": 0.17060010135173798,
+      "learning_rate": 0.001,
+      "loss": 3.1258,
+      "step": 17902
+    },
+    {
+      "epoch": 0.7573821812336069,
+      "grad_norm": 0.16431990265846252,
+      "learning_rate": 0.001,
+      "loss": 1.9269,
+      "step": 17903
+    },
+    {
+      "epoch": 0.7574244859971233,
+      "grad_norm": 0.17108014225959778,
+      "learning_rate": 0.001,
+      "loss": 2.9287,
+      "step": 17904
+    },
+    {
+      "epoch": 0.7574667907606396,
+      "grad_norm": 0.14799624681472778,
+      "learning_rate": 0.001,
+      "loss": 1.8449,
+      "step": 17905
+    },
+    {
+      "epoch": 0.757509095524156,
+      "grad_norm": 0.3792171776294708,
+      "learning_rate": 0.001,
+      "loss": 2.2688,
+      "step": 17906
+    },
+    {
+      "epoch": 0.7575514002876724,
+      "grad_norm": 0.2001536786556244,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 17907
+    },
+    {
+      "epoch": 0.7575937050511887,
+      "grad_norm": 0.1881311982870102,
+      "learning_rate": 0.001,
+      "loss": 2.8072,
+      "step": 17908
+    },
+    {
+      "epoch": 0.7576360098147051,
+      "grad_norm": 0.16043172776699066,
+      "learning_rate": 0.001,
+      "loss": 2.4464,
+      "step": 17909
+    },
+    {
+      "epoch": 0.7576783145782215,
+      "grad_norm": 0.16389212012290955,
+      "learning_rate": 0.001,
+      "loss": 1.8501,
+      "step": 17910
+    },
+    {
+      "epoch": 0.7577206193417378,
+      "grad_norm": 0.14742010831832886,
+      "learning_rate": 0.001,
+      "loss": 1.5688,
+      "step": 17911
+    },
+    {
+      "epoch": 0.7577629241052543,
+      "grad_norm": 0.3455764949321747,
+      "learning_rate": 0.001,
+      "loss": 1.6951,
+      "step": 17912
+    },
+    {
+      "epoch": 0.7578052288687707,
+      "grad_norm": 0.15572752058506012,
+      "learning_rate": 0.001,
+      "loss": 2.8535,
+      "step": 17913
+    },
+    {
+      "epoch": 0.757847533632287,
+      "grad_norm": 0.14037249982357025,
+      "learning_rate": 0.001,
+      "loss": 2.1352,
+      "step": 17914
+    },
+    {
+      "epoch": 0.7578898383958034,
+      "grad_norm": 0.14692841470241547,
+      "learning_rate": 0.001,
+      "loss": 2.1657,
+      "step": 17915
+    },
+    {
+      "epoch": 0.7579321431593198,
+      "grad_norm": 0.15857268869876862,
+      "learning_rate": 0.001,
+      "loss": 2.1919,
+      "step": 17916
+    },
+    {
+      "epoch": 0.7579744479228361,
+      "grad_norm": 0.20824623107910156,
+      "learning_rate": 0.001,
+      "loss": 2.8121,
+      "step": 17917
+    },
+    {
+      "epoch": 0.7580167526863525,
+      "grad_norm": 0.1485569328069687,
+      "learning_rate": 0.001,
+      "loss": 3.027,
+      "step": 17918
+    },
+    {
+      "epoch": 0.7580590574498689,
+      "grad_norm": 0.3441106081008911,
+      "learning_rate": 0.001,
+      "loss": 1.4367,
+      "step": 17919
+    },
+    {
+      "epoch": 0.7581013622133852,
+      "grad_norm": 0.9153411984443665,
+      "learning_rate": 0.001,
+      "loss": 1.6852,
+      "step": 17920
+    },
+    {
+      "epoch": 0.7581436669769016,
+      "grad_norm": 0.15342696011066437,
+      "learning_rate": 0.001,
+      "loss": 2.0896,
+      "step": 17921
+    },
+    {
+      "epoch": 0.758185971740418,
+      "grad_norm": 0.1278454065322876,
+      "learning_rate": 0.001,
+      "loss": 1.6947,
+      "step": 17922
+    },
+    {
+      "epoch": 0.7582282765039343,
+      "grad_norm": 0.22508245706558228,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 17923
+    },
+    {
+      "epoch": 0.7582705812674507,
+      "grad_norm": 0.1679566353559494,
+      "learning_rate": 0.001,
+      "loss": 2.0825,
+      "step": 17924
+    },
+    {
+      "epoch": 0.7583128860309671,
+      "grad_norm": 0.18976467847824097,
+      "learning_rate": 0.001,
+      "loss": 2.0653,
+      "step": 17925
+    },
+    {
+      "epoch": 0.7583551907944834,
+      "grad_norm": 0.13648661971092224,
+      "learning_rate": 0.001,
+      "loss": 1.651,
+      "step": 17926
+    },
+    {
+      "epoch": 0.7583974955579998,
+      "grad_norm": 0.14440205693244934,
+      "learning_rate": 0.001,
+      "loss": 1.8385,
+      "step": 17927
+    },
+    {
+      "epoch": 0.7584398003215163,
+      "grad_norm": 0.14819002151489258,
+      "learning_rate": 0.001,
+      "loss": 1.4996,
+      "step": 17928
+    },
+    {
+      "epoch": 0.7584821050850326,
+      "grad_norm": 0.17857134342193604,
+      "learning_rate": 0.001,
+      "loss": 1.8688,
+      "step": 17929
+    },
+    {
+      "epoch": 0.758524409848549,
+      "grad_norm": 0.17260822653770447,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 17930
+    },
+    {
+      "epoch": 0.7585667146120654,
+      "grad_norm": 0.15363721549510956,
+      "learning_rate": 0.001,
+      "loss": 2.2281,
+      "step": 17931
+    },
+    {
+      "epoch": 0.7586090193755817,
+      "grad_norm": 0.2327968031167984,
+      "learning_rate": 0.001,
+      "loss": 2.4439,
+      "step": 17932
+    },
+    {
+      "epoch": 0.7586513241390981,
+      "grad_norm": 0.1449955552816391,
+      "learning_rate": 0.001,
+      "loss": 1.5978,
+      "step": 17933
+    },
+    {
+      "epoch": 0.7586936289026144,
+      "grad_norm": 0.14919523894786835,
+      "learning_rate": 0.001,
+      "loss": 1.8171,
+      "step": 17934
+    },
+    {
+      "epoch": 0.7587359336661308,
+      "grad_norm": 0.12412866950035095,
+      "learning_rate": 0.001,
+      "loss": 2.6799,
+      "step": 17935
+    },
+    {
+      "epoch": 0.7587782384296472,
+      "grad_norm": 0.16862471401691437,
+      "learning_rate": 0.001,
+      "loss": 2.93,
+      "step": 17936
+    },
+    {
+      "epoch": 0.7588205431931635,
+      "grad_norm": 0.16229136288166046,
+      "learning_rate": 0.001,
+      "loss": 1.2241,
+      "step": 17937
+    },
+    {
+      "epoch": 0.7588628479566799,
+      "grad_norm": 0.12667950987815857,
+      "learning_rate": 0.001,
+      "loss": 1.3335,
+      "step": 17938
+    },
+    {
+      "epoch": 0.7589051527201963,
+      "grad_norm": 0.1758270263671875,
+      "learning_rate": 0.001,
+      "loss": 2.1729,
+      "step": 17939
+    },
+    {
+      "epoch": 0.7589474574837126,
+      "grad_norm": 0.17729204893112183,
+      "learning_rate": 0.001,
+      "loss": 3.5026,
+      "step": 17940
+    },
+    {
+      "epoch": 0.758989762247229,
+      "grad_norm": 0.16940808296203613,
+      "learning_rate": 0.001,
+      "loss": 2.3168,
+      "step": 17941
+    },
+    {
+      "epoch": 0.7590320670107454,
+      "grad_norm": 0.14315323531627655,
+      "learning_rate": 0.001,
+      "loss": 1.6935,
+      "step": 17942
+    },
+    {
+      "epoch": 0.7590743717742617,
+      "grad_norm": 0.14111851155757904,
+      "learning_rate": 0.001,
+      "loss": 1.6618,
+      "step": 17943
+    },
+    {
+      "epoch": 0.7591166765377781,
+      "grad_norm": 0.28591257333755493,
+      "learning_rate": 0.001,
+      "loss": 3.0757,
+      "step": 17944
+    },
+    {
+      "epoch": 0.7591589813012946,
+      "grad_norm": 0.15771640837192535,
+      "learning_rate": 0.001,
+      "loss": 2.0072,
+      "step": 17945
+    },
+    {
+      "epoch": 0.7592012860648109,
+      "grad_norm": 0.15648822486400604,
+      "learning_rate": 0.001,
+      "loss": 2.8095,
+      "step": 17946
+    },
+    {
+      "epoch": 0.7592435908283273,
+      "grad_norm": 0.20096895098686218,
+      "learning_rate": 0.001,
+      "loss": 2.3453,
+      "step": 17947
+    },
+    {
+      "epoch": 0.7592858955918437,
+      "grad_norm": 0.17997436225414276,
+      "learning_rate": 0.001,
+      "loss": 1.9256,
+      "step": 17948
+    },
+    {
+      "epoch": 0.75932820035536,
+      "grad_norm": 0.1645924299955368,
+      "learning_rate": 0.001,
+      "loss": 1.9429,
+      "step": 17949
+    },
+    {
+      "epoch": 0.7593705051188764,
+      "grad_norm": 0.14926403760910034,
+      "learning_rate": 0.001,
+      "loss": 2.39,
+      "step": 17950
+    },
+    {
+      "epoch": 0.7594128098823928,
+      "grad_norm": 0.17297221720218658,
+      "learning_rate": 0.001,
+      "loss": 1.5502,
+      "step": 17951
+    },
+    {
+      "epoch": 0.7594551146459091,
+      "grad_norm": 0.1334647685289383,
+      "learning_rate": 0.001,
+      "loss": 1.7092,
+      "step": 17952
+    },
+    {
+      "epoch": 0.7594974194094255,
+      "grad_norm": 0.7537721395492554,
+      "learning_rate": 0.001,
+      "loss": 1.3184,
+      "step": 17953
+    },
+    {
+      "epoch": 0.7595397241729419,
+      "grad_norm": 0.2796461880207062,
+      "learning_rate": 0.001,
+      "loss": 2.0563,
+      "step": 17954
+    },
+    {
+      "epoch": 0.7595820289364582,
+      "grad_norm": 0.2787031829357147,
+      "learning_rate": 0.001,
+      "loss": 2.0524,
+      "step": 17955
+    },
+    {
+      "epoch": 0.7596243336999746,
+      "grad_norm": 0.172209694981575,
+      "learning_rate": 0.001,
+      "loss": 1.8967,
+      "step": 17956
+    },
+    {
+      "epoch": 0.759666638463491,
+      "grad_norm": 0.15456169843673706,
+      "learning_rate": 0.001,
+      "loss": 2.1134,
+      "step": 17957
+    },
+    {
+      "epoch": 0.7597089432270073,
+      "grad_norm": 0.16440053284168243,
+      "learning_rate": 0.001,
+      "loss": 1.9277,
+      "step": 17958
+    },
+    {
+      "epoch": 0.7597512479905237,
+      "grad_norm": 0.2558080554008484,
+      "learning_rate": 0.001,
+      "loss": 2.6084,
+      "step": 17959
+    },
+    {
+      "epoch": 0.7597935527540401,
+      "grad_norm": 0.1783752590417862,
+      "learning_rate": 0.001,
+      "loss": 2.1082,
+      "step": 17960
+    },
+    {
+      "epoch": 0.7598358575175564,
+      "grad_norm": 0.14168411493301392,
+      "learning_rate": 0.001,
+      "loss": 1.8182,
+      "step": 17961
+    },
+    {
+      "epoch": 0.7598781622810729,
+      "grad_norm": 0.1725158542394638,
+      "learning_rate": 0.001,
+      "loss": 2.5628,
+      "step": 17962
+    },
+    {
+      "epoch": 0.7599204670445893,
+      "grad_norm": 0.13498811423778534,
+      "learning_rate": 0.001,
+      "loss": 2.1323,
+      "step": 17963
+    },
+    {
+      "epoch": 0.7599627718081056,
+      "grad_norm": 0.15643754601478577,
+      "learning_rate": 0.001,
+      "loss": 2.3276,
+      "step": 17964
+    },
+    {
+      "epoch": 0.760005076571622,
+      "grad_norm": 0.16019053757190704,
+      "learning_rate": 0.001,
+      "loss": 2.2398,
+      "step": 17965
+    },
+    {
+      "epoch": 0.7600473813351384,
+      "grad_norm": 0.1496978998184204,
+      "learning_rate": 0.001,
+      "loss": 3.8534,
+      "step": 17966
+    },
+    {
+      "epoch": 0.7600896860986547,
+      "grad_norm": 0.18137343227863312,
+      "learning_rate": 0.001,
+      "loss": 2.4502,
+      "step": 17967
+    },
+    {
+      "epoch": 0.7601319908621711,
+      "grad_norm": 0.14350903034210205,
+      "learning_rate": 0.001,
+      "loss": 2.6869,
+      "step": 17968
+    },
+    {
+      "epoch": 0.7601742956256875,
+      "grad_norm": 0.17714965343475342,
+      "learning_rate": 0.001,
+      "loss": 1.9266,
+      "step": 17969
+    },
+    {
+      "epoch": 0.7602166003892038,
+      "grad_norm": 0.13317923247814178,
+      "learning_rate": 0.001,
+      "loss": 1.5833,
+      "step": 17970
+    },
+    {
+      "epoch": 0.7602589051527202,
+      "grad_norm": 0.13361962139606476,
+      "learning_rate": 0.001,
+      "loss": 1.6225,
+      "step": 17971
+    },
+    {
+      "epoch": 0.7603012099162366,
+      "grad_norm": 0.1693313717842102,
+      "learning_rate": 0.001,
+      "loss": 1.9779,
+      "step": 17972
+    },
+    {
+      "epoch": 0.7603435146797529,
+      "grad_norm": 1.043839454650879,
+      "learning_rate": 0.001,
+      "loss": 1.9982,
+      "step": 17973
+    },
+    {
+      "epoch": 0.7603858194432693,
+      "grad_norm": 0.19669762253761292,
+      "learning_rate": 0.001,
+      "loss": 2.6891,
+      "step": 17974
+    },
+    {
+      "epoch": 0.7604281242067857,
+      "grad_norm": 0.1441902369260788,
+      "learning_rate": 0.001,
+      "loss": 2.0312,
+      "step": 17975
+    },
+    {
+      "epoch": 0.760470428970302,
+      "grad_norm": 0.13222302496433258,
+      "learning_rate": 0.001,
+      "loss": 2.0124,
+      "step": 17976
+    },
+    {
+      "epoch": 0.7605127337338184,
+      "grad_norm": 0.16443274915218353,
+      "learning_rate": 0.001,
+      "loss": 3.1681,
+      "step": 17977
+    },
+    {
+      "epoch": 0.7605550384973347,
+      "grad_norm": 0.1863120049238205,
+      "learning_rate": 0.001,
+      "loss": 2.4867,
+      "step": 17978
+    },
+    {
+      "epoch": 0.7605973432608512,
+      "grad_norm": 0.15816029906272888,
+      "learning_rate": 0.001,
+      "loss": 3.2455,
+      "step": 17979
+    },
+    {
+      "epoch": 0.7606396480243676,
+      "grad_norm": 0.20944538712501526,
+      "learning_rate": 0.001,
+      "loss": 2.5672,
+      "step": 17980
+    },
+    {
+      "epoch": 0.7606819527878839,
+      "grad_norm": 0.13495473563671112,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 17981
+    },
+    {
+      "epoch": 0.7607242575514003,
+      "grad_norm": 0.160276859998703,
+      "learning_rate": 0.001,
+      "loss": 2.5909,
+      "step": 17982
+    },
+    {
+      "epoch": 0.7607665623149167,
+      "grad_norm": 0.14563147723674774,
+      "learning_rate": 0.001,
+      "loss": 1.6701,
+      "step": 17983
+    },
+    {
+      "epoch": 0.760808867078433,
+      "grad_norm": 0.16634686291217804,
+      "learning_rate": 0.001,
+      "loss": 2.6882,
+      "step": 17984
+    },
+    {
+      "epoch": 0.7608511718419494,
+      "grad_norm": 0.17471922934055328,
+      "learning_rate": 0.001,
+      "loss": 3.1525,
+      "step": 17985
+    },
+    {
+      "epoch": 0.7608934766054658,
+      "grad_norm": 0.18728522956371307,
+      "learning_rate": 0.001,
+      "loss": 3.5754,
+      "step": 17986
+    },
+    {
+      "epoch": 0.7609357813689821,
+      "grad_norm": 0.13889671862125397,
+      "learning_rate": 0.001,
+      "loss": 2.6335,
+      "step": 17987
+    },
+    {
+      "epoch": 0.7609780861324985,
+      "grad_norm": 0.1883813589811325,
+      "learning_rate": 0.001,
+      "loss": 2.5539,
+      "step": 17988
+    },
+    {
+      "epoch": 0.7610203908960149,
+      "grad_norm": 0.16442306339740753,
+      "learning_rate": 0.001,
+      "loss": 2.6422,
+      "step": 17989
+    },
+    {
+      "epoch": 0.7610626956595312,
+      "grad_norm": 0.154715433716774,
+      "learning_rate": 0.001,
+      "loss": 2.8743,
+      "step": 17990
+    },
+    {
+      "epoch": 0.7611050004230476,
+      "grad_norm": 1.595334529876709,
+      "learning_rate": 0.001,
+      "loss": 2.4011,
+      "step": 17991
+    },
+    {
+      "epoch": 0.761147305186564,
+      "grad_norm": 0.14046059548854828,
+      "learning_rate": 0.001,
+      "loss": 2.2203,
+      "step": 17992
+    },
+    {
+      "epoch": 0.7611896099500803,
+      "grad_norm": 0.16517886519432068,
+      "learning_rate": 0.001,
+      "loss": 1.8345,
+      "step": 17993
+    },
+    {
+      "epoch": 0.7612319147135967,
+      "grad_norm": 0.15693698823451996,
+      "learning_rate": 0.001,
+      "loss": 2.506,
+      "step": 17994
+    },
+    {
+      "epoch": 0.7612742194771132,
+      "grad_norm": 0.2535119950771332,
+      "learning_rate": 0.001,
+      "loss": 2.2293,
+      "step": 17995
+    },
+    {
+      "epoch": 0.7613165242406295,
+      "grad_norm": 0.30919238924980164,
+      "learning_rate": 0.001,
+      "loss": 2.0888,
+      "step": 17996
+    },
+    {
+      "epoch": 0.7613588290041459,
+      "grad_norm": 0.21086013317108154,
+      "learning_rate": 0.001,
+      "loss": 1.4487,
+      "step": 17997
+    },
+    {
+      "epoch": 0.7614011337676623,
+      "grad_norm": 0.14983509480953217,
+      "learning_rate": 0.001,
+      "loss": 1.9109,
+      "step": 17998
+    },
+    {
+      "epoch": 0.7614434385311786,
+      "grad_norm": 0.16807793080806732,
+      "learning_rate": 0.001,
+      "loss": 4.2247,
+      "step": 17999
+    },
+    {
+      "epoch": 0.761485743294695,
+      "grad_norm": 0.15390969812870026,
+      "learning_rate": 0.001,
+      "loss": 1.5664,
+      "step": 18000
+    },
+    {
+      "epoch": 0.7615280480582114,
+      "grad_norm": 0.16306841373443604,
+      "learning_rate": 0.001,
+      "loss": 1.6577,
+      "step": 18001
+    },
+    {
+      "epoch": 0.7615703528217277,
+      "grad_norm": 0.16728578507900238,
+      "learning_rate": 0.001,
+      "loss": 1.9575,
+      "step": 18002
+    },
+    {
+      "epoch": 0.7616126575852441,
+      "grad_norm": 0.16129907965660095,
+      "learning_rate": 0.001,
+      "loss": 1.9566,
+      "step": 18003
+    },
+    {
+      "epoch": 0.7616549623487605,
+      "grad_norm": 0.1789942979812622,
+      "learning_rate": 0.001,
+      "loss": 1.806,
+      "step": 18004
+    },
+    {
+      "epoch": 0.7616972671122768,
+      "grad_norm": 0.15799731016159058,
+      "learning_rate": 0.001,
+      "loss": 2.8574,
+      "step": 18005
+    },
+    {
+      "epoch": 0.7617395718757932,
+      "grad_norm": 0.13212554156780243,
+      "learning_rate": 0.001,
+      "loss": 1.6819,
+      "step": 18006
+    },
+    {
+      "epoch": 0.7617818766393096,
+      "grad_norm": 0.13850046694278717,
+      "learning_rate": 0.001,
+      "loss": 2.0633,
+      "step": 18007
+    },
+    {
+      "epoch": 0.7618241814028259,
+      "grad_norm": 0.22076542675495148,
+      "learning_rate": 0.001,
+      "loss": 2.0944,
+      "step": 18008
+    },
+    {
+      "epoch": 0.7618664861663423,
+      "grad_norm": 0.18561388552188873,
+      "learning_rate": 0.001,
+      "loss": 1.7448,
+      "step": 18009
+    },
+    {
+      "epoch": 0.7619087909298587,
+      "grad_norm": 0.15947283804416656,
+      "learning_rate": 0.001,
+      "loss": 1.9042,
+      "step": 18010
+    },
+    {
+      "epoch": 0.761951095693375,
+      "grad_norm": 0.49783751368522644,
+      "learning_rate": 0.001,
+      "loss": 2.8313,
+      "step": 18011
+    },
+    {
+      "epoch": 0.7619934004568915,
+      "grad_norm": 0.16959191858768463,
+      "learning_rate": 0.001,
+      "loss": 2.4751,
+      "step": 18012
+    },
+    {
+      "epoch": 0.7620357052204079,
+      "grad_norm": 0.1717720329761505,
+      "learning_rate": 0.001,
+      "loss": 3.4214,
+      "step": 18013
+    },
+    {
+      "epoch": 0.7620780099839242,
+      "grad_norm": 0.1780117154121399,
+      "learning_rate": 0.001,
+      "loss": 2.4375,
+      "step": 18014
+    },
+    {
+      "epoch": 0.7621203147474406,
+      "grad_norm": 0.1493871957063675,
+      "learning_rate": 0.001,
+      "loss": 1.8541,
+      "step": 18015
+    },
+    {
+      "epoch": 0.762162619510957,
+      "grad_norm": 0.1525392383337021,
+      "learning_rate": 0.001,
+      "loss": 1.7101,
+      "step": 18016
+    },
+    {
+      "epoch": 0.7622049242744733,
+      "grad_norm": 0.14318202435970306,
+      "learning_rate": 0.001,
+      "loss": 3.5063,
+      "step": 18017
+    },
+    {
+      "epoch": 0.7622472290379897,
+      "grad_norm": 0.5359983444213867,
+      "learning_rate": 0.001,
+      "loss": 1.6701,
+      "step": 18018
+    },
+    {
+      "epoch": 0.7622895338015061,
+      "grad_norm": 0.19831795990467072,
+      "learning_rate": 0.001,
+      "loss": 2.242,
+      "step": 18019
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 0.4935970604419708,
+      "learning_rate": 0.001,
+      "loss": 2.697,
+      "step": 18020
+    },
+    {
+      "epoch": 0.7623741433285388,
+      "grad_norm": 0.19180017709732056,
+      "learning_rate": 0.001,
+      "loss": 2.217,
+      "step": 18021
+    },
+    {
+      "epoch": 0.7624164480920551,
+      "grad_norm": 0.25960850715637207,
+      "learning_rate": 0.001,
+      "loss": 1.561,
+      "step": 18022
+    },
+    {
+      "epoch": 0.7624587528555715,
+      "grad_norm": 0.25631117820739746,
+      "learning_rate": 0.001,
+      "loss": 2.6779,
+      "step": 18023
+    },
+    {
+      "epoch": 0.7625010576190879,
+      "grad_norm": 2.3049187660217285,
+      "learning_rate": 0.001,
+      "loss": 2.1273,
+      "step": 18024
+    },
+    {
+      "epoch": 0.7625433623826042,
+      "grad_norm": 0.1636704057455063,
+      "learning_rate": 0.001,
+      "loss": 1.6586,
+      "step": 18025
+    },
+    {
+      "epoch": 0.7625856671461206,
+      "grad_norm": 0.17722514271736145,
+      "learning_rate": 0.001,
+      "loss": 1.6602,
+      "step": 18026
+    },
+    {
+      "epoch": 0.762627971909637,
+      "grad_norm": 0.22410543262958527,
+      "learning_rate": 0.001,
+      "loss": 2.5346,
+      "step": 18027
+    },
+    {
+      "epoch": 0.7626702766731533,
+      "grad_norm": 0.15916574001312256,
+      "learning_rate": 0.001,
+      "loss": 2.2563,
+      "step": 18028
+    },
+    {
+      "epoch": 0.7627125814366698,
+      "grad_norm": 0.17527827620506287,
+      "learning_rate": 0.001,
+      "loss": 2.4785,
+      "step": 18029
+    },
+    {
+      "epoch": 0.7627548862001862,
+      "grad_norm": 0.2694207727909088,
+      "learning_rate": 0.001,
+      "loss": 2.3781,
+      "step": 18030
+    },
+    {
+      "epoch": 0.7627971909637025,
+      "grad_norm": 0.2880823314189911,
+      "learning_rate": 0.001,
+      "loss": 2.6001,
+      "step": 18031
+    },
+    {
+      "epoch": 0.7628394957272189,
+      "grad_norm": 0.165915384888649,
+      "learning_rate": 0.001,
+      "loss": 2.18,
+      "step": 18032
+    },
+    {
+      "epoch": 0.7628818004907353,
+      "grad_norm": 0.19809134304523468,
+      "learning_rate": 0.001,
+      "loss": 2.8755,
+      "step": 18033
+    },
+    {
+      "epoch": 0.7629241052542516,
+      "grad_norm": 0.18514703214168549,
+      "learning_rate": 0.001,
+      "loss": 1.9967,
+      "step": 18034
+    },
+    {
+      "epoch": 0.762966410017768,
+      "grad_norm": 0.17875197529792786,
+      "learning_rate": 0.001,
+      "loss": 2.2859,
+      "step": 18035
+    },
+    {
+      "epoch": 0.7630087147812844,
+      "grad_norm": 0.1806037873029709,
+      "learning_rate": 0.001,
+      "loss": 2.6236,
+      "step": 18036
+    },
+    {
+      "epoch": 0.7630510195448007,
+      "grad_norm": 0.15905499458312988,
+      "learning_rate": 0.001,
+      "loss": 2.1554,
+      "step": 18037
+    },
+    {
+      "epoch": 0.7630933243083171,
+      "grad_norm": 0.1544172763824463,
+      "learning_rate": 0.001,
+      "loss": 1.8756,
+      "step": 18038
+    },
+    {
+      "epoch": 0.7631356290718335,
+      "grad_norm": 0.20902614295482635,
+      "learning_rate": 0.001,
+      "loss": 2.6106,
+      "step": 18039
+    },
+    {
+      "epoch": 0.7631779338353498,
+      "grad_norm": 0.15740123391151428,
+      "learning_rate": 0.001,
+      "loss": 2.3193,
+      "step": 18040
+    },
+    {
+      "epoch": 0.7632202385988662,
+      "grad_norm": 0.15449416637420654,
+      "learning_rate": 0.001,
+      "loss": 1.6584,
+      "step": 18041
+    },
+    {
+      "epoch": 0.7632625433623826,
+      "grad_norm": 0.17526280879974365,
+      "learning_rate": 0.001,
+      "loss": 1.9062,
+      "step": 18042
+    },
+    {
+      "epoch": 0.7633048481258989,
+      "grad_norm": 0.19311073422431946,
+      "learning_rate": 0.001,
+      "loss": 2.3969,
+      "step": 18043
+    },
+    {
+      "epoch": 0.7633471528894153,
+      "grad_norm": 0.17864413559436798,
+      "learning_rate": 0.001,
+      "loss": 1.7516,
+      "step": 18044
+    },
+    {
+      "epoch": 0.7633894576529318,
+      "grad_norm": 0.16350750625133514,
+      "learning_rate": 0.001,
+      "loss": 1.8934,
+      "step": 18045
+    },
+    {
+      "epoch": 0.7634317624164481,
+      "grad_norm": 0.29366812109947205,
+      "learning_rate": 0.001,
+      "loss": 3.6984,
+      "step": 18046
+    },
+    {
+      "epoch": 0.7634740671799645,
+      "grad_norm": 0.8536985516548157,
+      "learning_rate": 0.001,
+      "loss": 1.4539,
+      "step": 18047
+    },
+    {
+      "epoch": 0.7635163719434809,
+      "grad_norm": 0.1530151069164276,
+      "learning_rate": 0.001,
+      "loss": 1.9095,
+      "step": 18048
+    },
+    {
+      "epoch": 0.7635586767069972,
+      "grad_norm": 0.1519416868686676,
+      "learning_rate": 0.001,
+      "loss": 1.5347,
+      "step": 18049
+    },
+    {
+      "epoch": 0.7636009814705136,
+      "grad_norm": 0.31431666016578674,
+      "learning_rate": 0.001,
+      "loss": 2.056,
+      "step": 18050
+    },
+    {
+      "epoch": 0.76364328623403,
+      "grad_norm": 0.9129075407981873,
+      "learning_rate": 0.001,
+      "loss": 1.9324,
+      "step": 18051
+    },
+    {
+      "epoch": 0.7636855909975463,
+      "grad_norm": 0.1397624909877777,
+      "learning_rate": 0.001,
+      "loss": 1.7459,
+      "step": 18052
+    },
+    {
+      "epoch": 0.7637278957610627,
+      "grad_norm": 0.1634640395641327,
+      "learning_rate": 0.001,
+      "loss": 2.0907,
+      "step": 18053
+    },
+    {
+      "epoch": 0.7637702005245791,
+      "grad_norm": 0.18420040607452393,
+      "learning_rate": 0.001,
+      "loss": 2.1298,
+      "step": 18054
+    },
+    {
+      "epoch": 0.7638125052880954,
+      "grad_norm": 1.812959909439087,
+      "learning_rate": 0.001,
+      "loss": 3.1896,
+      "step": 18055
+    },
+    {
+      "epoch": 0.7638548100516118,
+      "grad_norm": 0.40835338830947876,
+      "learning_rate": 0.001,
+      "loss": 2.3623,
+      "step": 18056
+    },
+    {
+      "epoch": 0.7638971148151282,
+      "grad_norm": 5.703685283660889,
+      "learning_rate": 0.001,
+      "loss": 2.2481,
+      "step": 18057
+    },
+    {
+      "epoch": 0.7639394195786445,
+      "grad_norm": 0.18355365097522736,
+      "learning_rate": 0.001,
+      "loss": 2.5109,
+      "step": 18058
+    },
+    {
+      "epoch": 0.7639817243421609,
+      "grad_norm": 38.65842056274414,
+      "learning_rate": 0.001,
+      "loss": 2.5208,
+      "step": 18059
+    },
+    {
+      "epoch": 0.7640240291056773,
+      "grad_norm": 0.1402842104434967,
+      "learning_rate": 0.001,
+      "loss": 2.0796,
+      "step": 18060
+    },
+    {
+      "epoch": 0.7640663338691936,
+      "grad_norm": 0.28925231099128723,
+      "learning_rate": 0.001,
+      "loss": 2.1025,
+      "step": 18061
+    },
+    {
+      "epoch": 0.7641086386327101,
+      "grad_norm": 0.1598457396030426,
+      "learning_rate": 0.001,
+      "loss": 1.9011,
+      "step": 18062
+    },
+    {
+      "epoch": 0.7641509433962265,
+      "grad_norm": 0.21429701149463654,
+      "learning_rate": 0.001,
+      "loss": 1.6832,
+      "step": 18063
+    },
+    {
+      "epoch": 0.7641932481597428,
+      "grad_norm": 0.1933804154396057,
+      "learning_rate": 0.001,
+      "loss": 2.435,
+      "step": 18064
+    },
+    {
+      "epoch": 0.7642355529232592,
+      "grad_norm": 0.21492686867713928,
+      "learning_rate": 0.001,
+      "loss": 1.7783,
+      "step": 18065
+    },
+    {
+      "epoch": 0.7642778576867756,
+      "grad_norm": 0.19295653700828552,
+      "learning_rate": 0.001,
+      "loss": 1.6563,
+      "step": 18066
+    },
+    {
+      "epoch": 0.7643201624502919,
+      "grad_norm": 0.22416500747203827,
+      "learning_rate": 0.001,
+      "loss": 1.7468,
+      "step": 18067
+    },
+    {
+      "epoch": 0.7643624672138083,
+      "grad_norm": 0.1974020004272461,
+      "learning_rate": 0.001,
+      "loss": 2.6946,
+      "step": 18068
+    },
+    {
+      "epoch": 0.7644047719773246,
+      "grad_norm": 0.1901572346687317,
+      "learning_rate": 0.001,
+      "loss": 2.1431,
+      "step": 18069
+    },
+    {
+      "epoch": 0.764447076740841,
+      "grad_norm": 0.1712511032819748,
+      "learning_rate": 0.001,
+      "loss": 2.1096,
+      "step": 18070
+    },
+    {
+      "epoch": 0.7644893815043574,
+      "grad_norm": 0.1516590565443039,
+      "learning_rate": 0.001,
+      "loss": 1.7473,
+      "step": 18071
+    },
+    {
+      "epoch": 0.7645316862678737,
+      "grad_norm": 0.7932908535003662,
+      "learning_rate": 0.001,
+      "loss": 1.8818,
+      "step": 18072
+    },
+    {
+      "epoch": 0.7645739910313901,
+      "grad_norm": 0.12099297344684601,
+      "learning_rate": 0.001,
+      "loss": 1.6472,
+      "step": 18073
+    },
+    {
+      "epoch": 0.7646162957949065,
+      "grad_norm": 0.1378883421421051,
+      "learning_rate": 0.001,
+      "loss": 1.8047,
+      "step": 18074
+    },
+    {
+      "epoch": 0.7646586005584228,
+      "grad_norm": 0.1538868397474289,
+      "learning_rate": 0.001,
+      "loss": 2.1053,
+      "step": 18075
+    },
+    {
+      "epoch": 0.7647009053219392,
+      "grad_norm": 2.756589889526367,
+      "learning_rate": 0.001,
+      "loss": 2.4575,
+      "step": 18076
+    },
+    {
+      "epoch": 0.7647432100854556,
+      "grad_norm": 0.16562233865261078,
+      "learning_rate": 0.001,
+      "loss": 1.8964,
+      "step": 18077
+    },
+    {
+      "epoch": 0.764785514848972,
+      "grad_norm": 0.17701025307178497,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 18078
+    },
+    {
+      "epoch": 0.7648278196124884,
+      "grad_norm": 0.2576713562011719,
+      "learning_rate": 0.001,
+      "loss": 1.7816,
+      "step": 18079
+    },
+    {
+      "epoch": 0.7648701243760048,
+      "grad_norm": 0.20448629558086395,
+      "learning_rate": 0.001,
+      "loss": 2.3767,
+      "step": 18080
+    },
+    {
+      "epoch": 0.7649124291395211,
+      "grad_norm": 0.18108433485031128,
+      "learning_rate": 0.001,
+      "loss": 1.4452,
+      "step": 18081
+    },
+    {
+      "epoch": 0.7649547339030375,
+      "grad_norm": 0.17302542924880981,
+      "learning_rate": 0.001,
+      "loss": 2.3699,
+      "step": 18082
+    },
+    {
+      "epoch": 0.7649970386665539,
+      "grad_norm": 0.16477636992931366,
+      "learning_rate": 0.001,
+      "loss": 2.436,
+      "step": 18083
+    },
+    {
+      "epoch": 0.7650393434300702,
+      "grad_norm": 0.1961863785982132,
+      "learning_rate": 0.001,
+      "loss": 2.1719,
+      "step": 18084
+    },
+    {
+      "epoch": 0.7650816481935866,
+      "grad_norm": 0.14665809273719788,
+      "learning_rate": 0.001,
+      "loss": 1.5338,
+      "step": 18085
+    },
+    {
+      "epoch": 0.765123952957103,
+      "grad_norm": 0.16945761442184448,
+      "learning_rate": 0.001,
+      "loss": 3.4643,
+      "step": 18086
+    },
+    {
+      "epoch": 0.7651662577206193,
+      "grad_norm": 0.16046762466430664,
+      "learning_rate": 0.001,
+      "loss": 1.7755,
+      "step": 18087
+    },
+    {
+      "epoch": 0.7652085624841357,
+      "grad_norm": 0.15023651719093323,
+      "learning_rate": 0.001,
+      "loss": 2.3674,
+      "step": 18088
+    },
+    {
+      "epoch": 0.7652508672476521,
+      "grad_norm": 0.18328256905078888,
+      "learning_rate": 0.001,
+      "loss": 3.5899,
+      "step": 18089
+    },
+    {
+      "epoch": 0.7652931720111684,
+      "grad_norm": 0.164375901222229,
+      "learning_rate": 0.001,
+      "loss": 2.2567,
+      "step": 18090
+    },
+    {
+      "epoch": 0.7653354767746848,
+      "grad_norm": 0.15443533658981323,
+      "learning_rate": 0.001,
+      "loss": 1.6702,
+      "step": 18091
+    },
+    {
+      "epoch": 0.7653777815382012,
+      "grad_norm": 0.1532910019159317,
+      "learning_rate": 0.001,
+      "loss": 1.7926,
+      "step": 18092
+    },
+    {
+      "epoch": 0.7654200863017175,
+      "grad_norm": 0.17233017086982727,
+      "learning_rate": 0.001,
+      "loss": 1.9303,
+      "step": 18093
+    },
+    {
+      "epoch": 0.765462391065234,
+      "grad_norm": 0.3771326243877411,
+      "learning_rate": 0.001,
+      "loss": 2.6317,
+      "step": 18094
+    },
+    {
+      "epoch": 0.7655046958287504,
+      "grad_norm": 0.15392722189426422,
+      "learning_rate": 0.001,
+      "loss": 2.2798,
+      "step": 18095
+    },
+    {
+      "epoch": 0.7655470005922667,
+      "grad_norm": 0.15488563477993011,
+      "learning_rate": 0.001,
+      "loss": 2.1111,
+      "step": 18096
+    },
+    {
+      "epoch": 0.7655893053557831,
+      "grad_norm": 0.17814667522907257,
+      "learning_rate": 0.001,
+      "loss": 3.3561,
+      "step": 18097
+    },
+    {
+      "epoch": 0.7656316101192995,
+      "grad_norm": 0.26778027415275574,
+      "learning_rate": 0.001,
+      "loss": 1.7169,
+      "step": 18098
+    },
+    {
+      "epoch": 0.7656739148828158,
+      "grad_norm": 0.18762215971946716,
+      "learning_rate": 0.001,
+      "loss": 2.769,
+      "step": 18099
+    },
+    {
+      "epoch": 0.7657162196463322,
+      "grad_norm": 20.262495040893555,
+      "learning_rate": 0.001,
+      "loss": 2.2304,
+      "step": 18100
+    },
+    {
+      "epoch": 0.7657585244098486,
+      "grad_norm": 0.15660881996154785,
+      "learning_rate": 0.001,
+      "loss": 2.1479,
+      "step": 18101
+    },
+    {
+      "epoch": 0.7658008291733649,
+      "grad_norm": 0.1680443435907364,
+      "learning_rate": 0.001,
+      "loss": 1.9905,
+      "step": 18102
+    },
+    {
+      "epoch": 0.7658431339368813,
+      "grad_norm": 0.16119442880153656,
+      "learning_rate": 0.001,
+      "loss": 2.5656,
+      "step": 18103
+    },
+    {
+      "epoch": 0.7658854387003977,
+      "grad_norm": 0.1531805545091629,
+      "learning_rate": 0.001,
+      "loss": 1.4154,
+      "step": 18104
+    },
+    {
+      "epoch": 0.765927743463914,
+      "grad_norm": 1.081006407737732,
+      "learning_rate": 0.001,
+      "loss": 2.4958,
+      "step": 18105
+    },
+    {
+      "epoch": 0.7659700482274304,
+      "grad_norm": 0.24466001987457275,
+      "learning_rate": 0.001,
+      "loss": 2.9361,
+      "step": 18106
+    },
+    {
+      "epoch": 0.7660123529909468,
+      "grad_norm": 0.17916032671928406,
+      "learning_rate": 0.001,
+      "loss": 1.9019,
+      "step": 18107
+    },
+    {
+      "epoch": 0.7660546577544631,
+      "grad_norm": 0.19994664192199707,
+      "learning_rate": 0.001,
+      "loss": 1.9385,
+      "step": 18108
+    },
+    {
+      "epoch": 0.7660969625179795,
+      "grad_norm": 0.2043987214565277,
+      "learning_rate": 0.001,
+      "loss": 2.122,
+      "step": 18109
+    },
+    {
+      "epoch": 0.766139267281496,
+      "grad_norm": 0.18687529861927032,
+      "learning_rate": 0.001,
+      "loss": 1.7168,
+      "step": 18110
+    },
+    {
+      "epoch": 0.7661815720450122,
+      "grad_norm": 0.32144445180892944,
+      "learning_rate": 0.001,
+      "loss": 2.2964,
+      "step": 18111
+    },
+    {
+      "epoch": 0.7662238768085287,
+      "grad_norm": 0.1715632677078247,
+      "learning_rate": 0.001,
+      "loss": 1.7487,
+      "step": 18112
+    },
+    {
+      "epoch": 0.766266181572045,
+      "grad_norm": 0.22985003888607025,
+      "learning_rate": 0.001,
+      "loss": 2.0248,
+      "step": 18113
+    },
+    {
+      "epoch": 0.7663084863355614,
+      "grad_norm": 1.1345404386520386,
+      "learning_rate": 0.001,
+      "loss": 2.2758,
+      "step": 18114
+    },
+    {
+      "epoch": 0.7663507910990778,
+      "grad_norm": 0.18999895453453064,
+      "learning_rate": 0.001,
+      "loss": 1.8452,
+      "step": 18115
+    },
+    {
+      "epoch": 0.7663930958625941,
+      "grad_norm": 0.8503357768058777,
+      "learning_rate": 0.001,
+      "loss": 2.1866,
+      "step": 18116
+    },
+    {
+      "epoch": 0.7664354006261105,
+      "grad_norm": 0.19124074280261993,
+      "learning_rate": 0.001,
+      "loss": 2.4042,
+      "step": 18117
+    },
+    {
+      "epoch": 0.7664777053896269,
+      "grad_norm": 0.20431570708751678,
+      "learning_rate": 0.001,
+      "loss": 1.7436,
+      "step": 18118
+    },
+    {
+      "epoch": 0.7665200101531432,
+      "grad_norm": 0.2346670776605606,
+      "learning_rate": 0.001,
+      "loss": 2.3448,
+      "step": 18119
+    },
+    {
+      "epoch": 0.7665623149166596,
+      "grad_norm": 0.19955824315547943,
+      "learning_rate": 0.001,
+      "loss": 3.1449,
+      "step": 18120
+    },
+    {
+      "epoch": 0.766604619680176,
+      "grad_norm": 0.1841670423746109,
+      "learning_rate": 0.001,
+      "loss": 1.9185,
+      "step": 18121
+    },
+    {
+      "epoch": 0.7666469244436923,
+      "grad_norm": 0.44874322414398193,
+      "learning_rate": 0.001,
+      "loss": 2.127,
+      "step": 18122
+    },
+    {
+      "epoch": 0.7666892292072087,
+      "grad_norm": 0.1544332355260849,
+      "learning_rate": 0.001,
+      "loss": 2.2697,
+      "step": 18123
+    },
+    {
+      "epoch": 0.7667315339707251,
+      "grad_norm": 0.1674651801586151,
+      "learning_rate": 0.001,
+      "loss": 1.9383,
+      "step": 18124
+    },
+    {
+      "epoch": 0.7667738387342414,
+      "grad_norm": 3.113783359527588,
+      "learning_rate": 0.001,
+      "loss": 1.4092,
+      "step": 18125
+    },
+    {
+      "epoch": 0.7668161434977578,
+      "grad_norm": 0.1927240639925003,
+      "learning_rate": 0.001,
+      "loss": 3.3805,
+      "step": 18126
+    },
+    {
+      "epoch": 0.7668584482612743,
+      "grad_norm": 0.21248434484004974,
+      "learning_rate": 0.001,
+      "loss": 2.3386,
+      "step": 18127
+    },
+    {
+      "epoch": 0.7669007530247905,
+      "grad_norm": 0.2704300582408905,
+      "learning_rate": 0.001,
+      "loss": 2.011,
+      "step": 18128
+    },
+    {
+      "epoch": 0.766943057788307,
+      "grad_norm": 0.33151501417160034,
+      "learning_rate": 0.001,
+      "loss": 1.9698,
+      "step": 18129
+    },
+    {
+      "epoch": 0.7669853625518234,
+      "grad_norm": 0.2982079088687897,
+      "learning_rate": 0.001,
+      "loss": 2.9232,
+      "step": 18130
+    },
+    {
+      "epoch": 0.7670276673153397,
+      "grad_norm": 0.185440793633461,
+      "learning_rate": 0.001,
+      "loss": 2.1382,
+      "step": 18131
+    },
+    {
+      "epoch": 0.7670699720788561,
+      "grad_norm": 0.1537931114435196,
+      "learning_rate": 0.001,
+      "loss": 3.2961,
+      "step": 18132
+    },
+    {
+      "epoch": 0.7671122768423725,
+      "grad_norm": 0.19790029525756836,
+      "learning_rate": 0.001,
+      "loss": 1.857,
+      "step": 18133
+    },
+    {
+      "epoch": 0.7671545816058888,
+      "grad_norm": 0.16664250195026398,
+      "learning_rate": 0.001,
+      "loss": 2.3612,
+      "step": 18134
+    },
+    {
+      "epoch": 0.7671968863694052,
+      "grad_norm": 0.19958829879760742,
+      "learning_rate": 0.001,
+      "loss": 1.444,
+      "step": 18135
+    },
+    {
+      "epoch": 0.7672391911329216,
+      "grad_norm": 0.25182950496673584,
+      "learning_rate": 0.001,
+      "loss": 2.2047,
+      "step": 18136
+    },
+    {
+      "epoch": 0.7672814958964379,
+      "grad_norm": 0.19357016682624817,
+      "learning_rate": 0.001,
+      "loss": 2.5632,
+      "step": 18137
+    },
+    {
+      "epoch": 0.7673238006599543,
+      "grad_norm": 0.20123781263828278,
+      "learning_rate": 0.001,
+      "loss": 3.6875,
+      "step": 18138
+    },
+    {
+      "epoch": 0.7673661054234707,
+      "grad_norm": 0.21408236026763916,
+      "learning_rate": 0.001,
+      "loss": 2.7204,
+      "step": 18139
+    },
+    {
+      "epoch": 0.767408410186987,
+      "grad_norm": 0.1696946918964386,
+      "learning_rate": 0.001,
+      "loss": 2.0886,
+      "step": 18140
+    },
+    {
+      "epoch": 0.7674507149505034,
+      "grad_norm": 0.8207197189331055,
+      "learning_rate": 0.001,
+      "loss": 2.7853,
+      "step": 18141
+    },
+    {
+      "epoch": 0.7674930197140198,
+      "grad_norm": 2.7755489349365234,
+      "learning_rate": 0.001,
+      "loss": 4.5532,
+      "step": 18142
+    },
+    {
+      "epoch": 0.7675353244775361,
+      "grad_norm": 0.20465411245822906,
+      "learning_rate": 0.001,
+      "loss": 1.803,
+      "step": 18143
+    },
+    {
+      "epoch": 0.7675776292410526,
+      "grad_norm": 0.16455033421516418,
+      "learning_rate": 0.001,
+      "loss": 2.0663,
+      "step": 18144
+    },
+    {
+      "epoch": 0.767619934004569,
+      "grad_norm": 3.680520534515381,
+      "learning_rate": 0.001,
+      "loss": 1.9112,
+      "step": 18145
+    },
+    {
+      "epoch": 0.7676622387680853,
+      "grad_norm": 0.20432768762111664,
+      "learning_rate": 0.001,
+      "loss": 1.8184,
+      "step": 18146
+    },
+    {
+      "epoch": 0.7677045435316017,
+      "grad_norm": 0.193999245762825,
+      "learning_rate": 0.001,
+      "loss": 2.0712,
+      "step": 18147
+    },
+    {
+      "epoch": 0.7677468482951181,
+      "grad_norm": 0.18594130873680115,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 18148
+    },
+    {
+      "epoch": 0.7677891530586344,
+      "grad_norm": 5.5671491622924805,
+      "learning_rate": 0.001,
+      "loss": 2.2832,
+      "step": 18149
+    },
+    {
+      "epoch": 0.7678314578221508,
+      "grad_norm": 0.22034242749214172,
+      "learning_rate": 0.001,
+      "loss": 3.3777,
+      "step": 18150
+    },
+    {
+      "epoch": 0.7678737625856672,
+      "grad_norm": 0.16389696300029755,
+      "learning_rate": 0.001,
+      "loss": 2.6919,
+      "step": 18151
+    },
+    {
+      "epoch": 0.7679160673491835,
+      "grad_norm": 4.396114349365234,
+      "learning_rate": 0.001,
+      "loss": 2.7745,
+      "step": 18152
+    },
+    {
+      "epoch": 0.7679583721126999,
+      "grad_norm": 0.18078528344631195,
+      "learning_rate": 0.001,
+      "loss": 2.5654,
+      "step": 18153
+    },
+    {
+      "epoch": 0.7680006768762163,
+      "grad_norm": 0.2645280361175537,
+      "learning_rate": 0.001,
+      "loss": 1.831,
+      "step": 18154
+    },
+    {
+      "epoch": 0.7680429816397326,
+      "grad_norm": 0.2649464011192322,
+      "learning_rate": 0.001,
+      "loss": 2.6965,
+      "step": 18155
+    },
+    {
+      "epoch": 0.768085286403249,
+      "grad_norm": 0.22064730525016785,
+      "learning_rate": 0.001,
+      "loss": 2.3275,
+      "step": 18156
+    },
+    {
+      "epoch": 0.7681275911667654,
+      "grad_norm": 0.22132393717765808,
+      "learning_rate": 0.001,
+      "loss": 1.9728,
+      "step": 18157
+    },
+    {
+      "epoch": 0.7681698959302817,
+      "grad_norm": 0.20887607336044312,
+      "learning_rate": 0.001,
+      "loss": 2.3195,
+      "step": 18158
+    },
+    {
+      "epoch": 0.7682122006937981,
+      "grad_norm": 1.3075305223464966,
+      "learning_rate": 0.001,
+      "loss": 2.5001,
+      "step": 18159
+    },
+    {
+      "epoch": 0.7682545054573144,
+      "grad_norm": 0.395717978477478,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 18160
+    },
+    {
+      "epoch": 0.7682968102208309,
+      "grad_norm": 0.18617072701454163,
+      "learning_rate": 0.001,
+      "loss": 2.0584,
+      "step": 18161
+    },
+    {
+      "epoch": 0.7683391149843473,
+      "grad_norm": 2.8809750080108643,
+      "learning_rate": 0.001,
+      "loss": 3.424,
+      "step": 18162
+    },
+    {
+      "epoch": 0.7683814197478636,
+      "grad_norm": 0.1813565492630005,
+      "learning_rate": 0.001,
+      "loss": 2.0456,
+      "step": 18163
+    },
+    {
+      "epoch": 0.76842372451138,
+      "grad_norm": 0.15641596913337708,
+      "learning_rate": 0.001,
+      "loss": 2.2862,
+      "step": 18164
+    },
+    {
+      "epoch": 0.7684660292748964,
+      "grad_norm": 0.1734810769557953,
+      "learning_rate": 0.001,
+      "loss": 3.9354,
+      "step": 18165
+    },
+    {
+      "epoch": 0.7685083340384127,
+      "grad_norm": 0.17491766810417175,
+      "learning_rate": 0.001,
+      "loss": 2.1404,
+      "step": 18166
+    },
+    {
+      "epoch": 0.7685506388019291,
+      "grad_norm": 0.1915895938873291,
+      "learning_rate": 0.001,
+      "loss": 2.9,
+      "step": 18167
+    },
+    {
+      "epoch": 0.7685929435654455,
+      "grad_norm": 0.1676090508699417,
+      "learning_rate": 0.001,
+      "loss": 2.2399,
+      "step": 18168
+    },
+    {
+      "epoch": 0.7686352483289618,
+      "grad_norm": 0.19483682513237,
+      "learning_rate": 0.001,
+      "loss": 2.1038,
+      "step": 18169
+    },
+    {
+      "epoch": 0.7686775530924782,
+      "grad_norm": 0.19385400414466858,
+      "learning_rate": 0.001,
+      "loss": 1.5603,
+      "step": 18170
+    },
+    {
+      "epoch": 0.7687198578559946,
+      "grad_norm": 0.15440934896469116,
+      "learning_rate": 0.001,
+      "loss": 1.5675,
+      "step": 18171
+    },
+    {
+      "epoch": 0.7687621626195109,
+      "grad_norm": 0.1513241082429886,
+      "learning_rate": 0.001,
+      "loss": 1.92,
+      "step": 18172
+    },
+    {
+      "epoch": 0.7688044673830273,
+      "grad_norm": 0.1512671858072281,
+      "learning_rate": 0.001,
+      "loss": 2.1714,
+      "step": 18173
+    },
+    {
+      "epoch": 0.7688467721465437,
+      "grad_norm": 0.31254085898399353,
+      "learning_rate": 0.001,
+      "loss": 4.2346,
+      "step": 18174
+    },
+    {
+      "epoch": 0.76888907691006,
+      "grad_norm": 0.14157380163669586,
+      "learning_rate": 0.001,
+      "loss": 1.5382,
+      "step": 18175
+    },
+    {
+      "epoch": 0.7689313816735764,
+      "grad_norm": 0.1927814781665802,
+      "learning_rate": 0.001,
+      "loss": 2.2692,
+      "step": 18176
+    },
+    {
+      "epoch": 0.7689736864370929,
+      "grad_norm": 0.6695175766944885,
+      "learning_rate": 0.001,
+      "loss": 2.4544,
+      "step": 18177
+    },
+    {
+      "epoch": 0.7690159912006092,
+      "grad_norm": 0.15355220437049866,
+      "learning_rate": 0.001,
+      "loss": 1.2551,
+      "step": 18178
+    },
+    {
+      "epoch": 0.7690582959641256,
+      "grad_norm": 1.621734380722046,
+      "learning_rate": 0.001,
+      "loss": 2.1838,
+      "step": 18179
+    },
+    {
+      "epoch": 0.769100600727642,
+      "grad_norm": 0.13713295757770538,
+      "learning_rate": 0.001,
+      "loss": 1.9986,
+      "step": 18180
+    },
+    {
+      "epoch": 0.7691429054911583,
+      "grad_norm": 0.3079672157764435,
+      "learning_rate": 0.001,
+      "loss": 2.9259,
+      "step": 18181
+    },
+    {
+      "epoch": 0.7691852102546747,
+      "grad_norm": 0.7089863419532776,
+      "learning_rate": 0.001,
+      "loss": 2.9016,
+      "step": 18182
+    },
+    {
+      "epoch": 0.7692275150181911,
+      "grad_norm": 1.244018316268921,
+      "learning_rate": 0.001,
+      "loss": 3.3858,
+      "step": 18183
+    },
+    {
+      "epoch": 0.7692698197817074,
+      "grad_norm": 0.19387851655483246,
+      "learning_rate": 0.001,
+      "loss": 3.4741,
+      "step": 18184
+    },
+    {
+      "epoch": 0.7693121245452238,
+      "grad_norm": 0.20450355112552643,
+      "learning_rate": 0.001,
+      "loss": 1.7754,
+      "step": 18185
+    },
+    {
+      "epoch": 0.7693544293087402,
+      "grad_norm": 0.18945342302322388,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 18186
+    },
+    {
+      "epoch": 0.7693967340722565,
+      "grad_norm": 0.191238135099411,
+      "learning_rate": 0.001,
+      "loss": 2.1089,
+      "step": 18187
+    },
+    {
+      "epoch": 0.7694390388357729,
+      "grad_norm": 0.2367229014635086,
+      "learning_rate": 0.001,
+      "loss": 2.7953,
+      "step": 18188
+    },
+    {
+      "epoch": 0.7694813435992893,
+      "grad_norm": 2.5353989601135254,
+      "learning_rate": 0.001,
+      "loss": 1.9474,
+      "step": 18189
+    },
+    {
+      "epoch": 0.7695236483628056,
+      "grad_norm": 2.5594029426574707,
+      "learning_rate": 0.001,
+      "loss": 1.7098,
+      "step": 18190
+    },
+    {
+      "epoch": 0.769565953126322,
+      "grad_norm": 0.21357464790344238,
+      "learning_rate": 0.001,
+      "loss": 3.2049,
+      "step": 18191
+    },
+    {
+      "epoch": 0.7696082578898384,
+      "grad_norm": 0.1735057681798935,
+      "learning_rate": 0.001,
+      "loss": 3.6089,
+      "step": 18192
+    },
+    {
+      "epoch": 0.7696505626533547,
+      "grad_norm": 0.18362459540367126,
+      "learning_rate": 0.001,
+      "loss": 1.9546,
+      "step": 18193
+    },
+    {
+      "epoch": 0.7696928674168712,
+      "grad_norm": 2.1002511978149414,
+      "learning_rate": 0.001,
+      "loss": 2.7012,
+      "step": 18194
+    },
+    {
+      "epoch": 0.7697351721803876,
+      "grad_norm": 0.1482006460428238,
+      "learning_rate": 0.001,
+      "loss": 1.4662,
+      "step": 18195
+    },
+    {
+      "epoch": 0.7697774769439039,
+      "grad_norm": 0.6873133778572083,
+      "learning_rate": 0.001,
+      "loss": 4.3751,
+      "step": 18196
+    },
+    {
+      "epoch": 0.7698197817074203,
+      "grad_norm": 0.19415785372257233,
+      "learning_rate": 0.001,
+      "loss": 2.9213,
+      "step": 18197
+    },
+    {
+      "epoch": 0.7698620864709367,
+      "grad_norm": 0.18056610226631165,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 18198
+    },
+    {
+      "epoch": 0.769904391234453,
+      "grad_norm": 0.15959571301937103,
+      "learning_rate": 0.001,
+      "loss": 2.5785,
+      "step": 18199
+    },
+    {
+      "epoch": 0.7699466959979694,
+      "grad_norm": 1.567895770072937,
+      "learning_rate": 0.001,
+      "loss": 1.4776,
+      "step": 18200
+    },
+    {
+      "epoch": 0.7699890007614858,
+      "grad_norm": 0.17681176960468292,
+      "learning_rate": 0.001,
+      "loss": 1.8845,
+      "step": 18201
+    },
+    {
+      "epoch": 0.7700313055250021,
+      "grad_norm": 0.16770051419734955,
+      "learning_rate": 0.001,
+      "loss": 1.8349,
+      "step": 18202
+    },
+    {
+      "epoch": 0.7700736102885185,
+      "grad_norm": 2.6900062561035156,
+      "learning_rate": 0.001,
+      "loss": 2.6632,
+      "step": 18203
+    },
+    {
+      "epoch": 0.7701159150520348,
+      "grad_norm": 0.18311475217342377,
+      "learning_rate": 0.001,
+      "loss": 1.4378,
+      "step": 18204
+    },
+    {
+      "epoch": 0.7701582198155512,
+      "grad_norm": 0.19490399956703186,
+      "learning_rate": 0.001,
+      "loss": 2.3233,
+      "step": 18205
+    },
+    {
+      "epoch": 0.7702005245790676,
+      "grad_norm": 0.17303673923015594,
+      "learning_rate": 0.001,
+      "loss": 1.9792,
+      "step": 18206
+    },
+    {
+      "epoch": 0.7702428293425839,
+      "grad_norm": 0.18099454045295715,
+      "learning_rate": 0.001,
+      "loss": 2.1376,
+      "step": 18207
+    },
+    {
+      "epoch": 0.7702851341061003,
+      "grad_norm": 0.1846335083246231,
+      "learning_rate": 0.001,
+      "loss": 2.6739,
+      "step": 18208
+    },
+    {
+      "epoch": 0.7703274388696167,
+      "grad_norm": 0.9617498517036438,
+      "learning_rate": 0.001,
+      "loss": 2.0516,
+      "step": 18209
+    },
+    {
+      "epoch": 0.770369743633133,
+      "grad_norm": 0.1959528774023056,
+      "learning_rate": 0.001,
+      "loss": 1.6241,
+      "step": 18210
+    },
+    {
+      "epoch": 0.7704120483966495,
+      "grad_norm": 0.18654143810272217,
+      "learning_rate": 0.001,
+      "loss": 2.8616,
+      "step": 18211
+    },
+    {
+      "epoch": 0.7704543531601659,
+      "grad_norm": 0.19656072556972504,
+      "learning_rate": 0.001,
+      "loss": 1.9619,
+      "step": 18212
+    },
+    {
+      "epoch": 0.7704966579236822,
+      "grad_norm": 0.42158690094947815,
+      "learning_rate": 0.001,
+      "loss": 2.3077,
+      "step": 18213
+    },
+    {
+      "epoch": 0.7705389626871986,
+      "grad_norm": 0.2899084687232971,
+      "learning_rate": 0.001,
+      "loss": 1.9099,
+      "step": 18214
+    },
+    {
+      "epoch": 0.770581267450715,
+      "grad_norm": 0.19894026219844818,
+      "learning_rate": 0.001,
+      "loss": 2.0858,
+      "step": 18215
+    },
+    {
+      "epoch": 0.7706235722142313,
+      "grad_norm": 0.16096381843090057,
+      "learning_rate": 0.001,
+      "loss": 1.724,
+      "step": 18216
+    },
+    {
+      "epoch": 0.7706658769777477,
+      "grad_norm": 0.19218985736370087,
+      "learning_rate": 0.001,
+      "loss": 2.7003,
+      "step": 18217
+    },
+    {
+      "epoch": 0.7707081817412641,
+      "grad_norm": 0.24575604498386383,
+      "learning_rate": 0.001,
+      "loss": 2.7877,
+      "step": 18218
+    },
+    {
+      "epoch": 0.7707504865047804,
+      "grad_norm": 0.17563903331756592,
+      "learning_rate": 0.001,
+      "loss": 2.192,
+      "step": 18219
+    },
+    {
+      "epoch": 0.7707927912682968,
+      "grad_norm": 0.18912796676158905,
+      "learning_rate": 0.001,
+      "loss": 2.3836,
+      "step": 18220
+    },
+    {
+      "epoch": 0.7708350960318132,
+      "grad_norm": 0.16776569187641144,
+      "learning_rate": 0.001,
+      "loss": 1.9411,
+      "step": 18221
+    },
+    {
+      "epoch": 0.7708774007953295,
+      "grad_norm": 0.18997344374656677,
+      "learning_rate": 0.001,
+      "loss": 2.5683,
+      "step": 18222
+    },
+    {
+      "epoch": 0.7709197055588459,
+      "grad_norm": 3.814743757247925,
+      "learning_rate": 0.001,
+      "loss": 2.4773,
+      "step": 18223
+    },
+    {
+      "epoch": 0.7709620103223623,
+      "grad_norm": 0.17335738241672516,
+      "learning_rate": 0.001,
+      "loss": 1.8626,
+      "step": 18224
+    },
+    {
+      "epoch": 0.7710043150858786,
+      "grad_norm": 1.7937777042388916,
+      "learning_rate": 0.001,
+      "loss": 2.8774,
+      "step": 18225
+    },
+    {
+      "epoch": 0.771046619849395,
+      "grad_norm": 0.16799603402614594,
+      "learning_rate": 0.001,
+      "loss": 1.7608,
+      "step": 18226
+    },
+    {
+      "epoch": 0.7710889246129115,
+      "grad_norm": 0.17312444746494293,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 18227
+    },
+    {
+      "epoch": 0.7711312293764278,
+      "grad_norm": 1.968964695930481,
+      "learning_rate": 0.001,
+      "loss": 1.916,
+      "step": 18228
+    },
+    {
+      "epoch": 0.7711735341399442,
+      "grad_norm": 0.1496216207742691,
+      "learning_rate": 0.001,
+      "loss": 1.789,
+      "step": 18229
+    },
+    {
+      "epoch": 0.7712158389034606,
+      "grad_norm": 0.15248148143291473,
+      "learning_rate": 0.001,
+      "loss": 2.2559,
+      "step": 18230
+    },
+    {
+      "epoch": 0.7712581436669769,
+      "grad_norm": 0.2088804543018341,
+      "learning_rate": 0.001,
+      "loss": 2.0968,
+      "step": 18231
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.20815947651863098,
+      "learning_rate": 0.001,
+      "loss": 2.6797,
+      "step": 18232
+    },
+    {
+      "epoch": 0.7713427531940097,
+      "grad_norm": 0.5374844670295715,
+      "learning_rate": 0.001,
+      "loss": 3.3856,
+      "step": 18233
+    },
+    {
+      "epoch": 0.771385057957526,
+      "grad_norm": 0.26020729541778564,
+      "learning_rate": 0.001,
+      "loss": 1.9107,
+      "step": 18234
+    },
+    {
+      "epoch": 0.7714273627210424,
+      "grad_norm": 0.27381160855293274,
+      "learning_rate": 0.001,
+      "loss": 2.0782,
+      "step": 18235
+    },
+    {
+      "epoch": 0.7714696674845588,
+      "grad_norm": 0.23972102999687195,
+      "learning_rate": 0.001,
+      "loss": 2.4307,
+      "step": 18236
+    },
+    {
+      "epoch": 0.7715119722480751,
+      "grad_norm": 0.1825583428144455,
+      "learning_rate": 0.001,
+      "loss": 1.7901,
+      "step": 18237
+    },
+    {
+      "epoch": 0.7715542770115915,
+      "grad_norm": 0.28334465622901917,
+      "learning_rate": 0.001,
+      "loss": 1.9967,
+      "step": 18238
+    },
+    {
+      "epoch": 0.7715965817751079,
+      "grad_norm": 1.4033657312393188,
+      "learning_rate": 0.001,
+      "loss": 3.312,
+      "step": 18239
+    },
+    {
+      "epoch": 0.7716388865386242,
+      "grad_norm": 0.20028841495513916,
+      "learning_rate": 0.001,
+      "loss": 1.7072,
+      "step": 18240
+    },
+    {
+      "epoch": 0.7716811913021406,
+      "grad_norm": 0.1848575472831726,
+      "learning_rate": 0.001,
+      "loss": 1.6997,
+      "step": 18241
+    },
+    {
+      "epoch": 0.771723496065657,
+      "grad_norm": 0.1979037970304489,
+      "learning_rate": 0.001,
+      "loss": 2.4092,
+      "step": 18242
+    },
+    {
+      "epoch": 0.7717658008291733,
+      "grad_norm": 0.2577193081378937,
+      "learning_rate": 0.001,
+      "loss": 2.3364,
+      "step": 18243
+    },
+    {
+      "epoch": 0.7718081055926898,
+      "grad_norm": 0.21102125942707062,
+      "learning_rate": 0.001,
+      "loss": 2.8763,
+      "step": 18244
+    },
+    {
+      "epoch": 0.7718504103562062,
+      "grad_norm": 0.17282043397426605,
+      "learning_rate": 0.001,
+      "loss": 1.6663,
+      "step": 18245
+    },
+    {
+      "epoch": 0.7718927151197225,
+      "grad_norm": 0.2045888453722,
+      "learning_rate": 0.001,
+      "loss": 2.0012,
+      "step": 18246
+    },
+    {
+      "epoch": 0.7719350198832389,
+      "grad_norm": 0.22302895784378052,
+      "learning_rate": 0.001,
+      "loss": 2.485,
+      "step": 18247
+    },
+    {
+      "epoch": 0.7719773246467552,
+      "grad_norm": 0.19768020510673523,
+      "learning_rate": 0.001,
+      "loss": 2.3736,
+      "step": 18248
+    },
+    {
+      "epoch": 0.7720196294102716,
+      "grad_norm": 0.15795622766017914,
+      "learning_rate": 0.001,
+      "loss": 1.9866,
+      "step": 18249
+    },
+    {
+      "epoch": 0.772061934173788,
+      "grad_norm": 0.1597718894481659,
+      "learning_rate": 0.001,
+      "loss": 2.3061,
+      "step": 18250
+    },
+    {
+      "epoch": 0.7721042389373043,
+      "grad_norm": 0.14455056190490723,
+      "learning_rate": 0.001,
+      "loss": 2.7711,
+      "step": 18251
+    },
+    {
+      "epoch": 0.7721465437008207,
+      "grad_norm": 0.16666211187839508,
+      "learning_rate": 0.001,
+      "loss": 2.4853,
+      "step": 18252
+    },
+    {
+      "epoch": 0.7721888484643371,
+      "grad_norm": 0.15804295241832733,
+      "learning_rate": 0.001,
+      "loss": 2.7146,
+      "step": 18253
+    },
+    {
+      "epoch": 0.7722311532278534,
+      "grad_norm": 0.14087384939193726,
+      "learning_rate": 0.001,
+      "loss": 1.6901,
+      "step": 18254
+    },
+    {
+      "epoch": 0.7722734579913698,
+      "grad_norm": 0.24747240543365479,
+      "learning_rate": 0.001,
+      "loss": 2.3874,
+      "step": 18255
+    },
+    {
+      "epoch": 0.7723157627548862,
+      "grad_norm": 0.20662999153137207,
+      "learning_rate": 0.001,
+      "loss": 1.7758,
+      "step": 18256
+    },
+    {
+      "epoch": 0.7723580675184025,
+      "grad_norm": 0.15056248009204865,
+      "learning_rate": 0.001,
+      "loss": 2.264,
+      "step": 18257
+    },
+    {
+      "epoch": 0.7724003722819189,
+      "grad_norm": 0.1590527892112732,
+      "learning_rate": 0.001,
+      "loss": 2.5744,
+      "step": 18258
+    },
+    {
+      "epoch": 0.7724426770454353,
+      "grad_norm": 0.18478628993034363,
+      "learning_rate": 0.001,
+      "loss": 2.4574,
+      "step": 18259
+    },
+    {
+      "epoch": 0.7724849818089516,
+      "grad_norm": 0.47995075583457947,
+      "learning_rate": 0.001,
+      "loss": 2.2169,
+      "step": 18260
+    },
+    {
+      "epoch": 0.772527286572468,
+      "grad_norm": 0.18007495999336243,
+      "learning_rate": 0.001,
+      "loss": 2.0898,
+      "step": 18261
+    },
+    {
+      "epoch": 0.7725695913359845,
+      "grad_norm": 0.15627458691596985,
+      "learning_rate": 0.001,
+      "loss": 2.0773,
+      "step": 18262
+    },
+    {
+      "epoch": 0.7726118960995008,
+      "grad_norm": 0.18800586462020874,
+      "learning_rate": 0.001,
+      "loss": 3.2193,
+      "step": 18263
+    },
+    {
+      "epoch": 0.7726542008630172,
+      "grad_norm": 0.3073885440826416,
+      "learning_rate": 0.001,
+      "loss": 2.0712,
+      "step": 18264
+    },
+    {
+      "epoch": 0.7726965056265336,
+      "grad_norm": 0.1837342083454132,
+      "learning_rate": 0.001,
+      "loss": 3.1809,
+      "step": 18265
+    },
+    {
+      "epoch": 0.7727388103900499,
+      "grad_norm": 0.24296334385871887,
+      "learning_rate": 0.001,
+      "loss": 2.2377,
+      "step": 18266
+    },
+    {
+      "epoch": 0.7727811151535663,
+      "grad_norm": 0.17990991473197937,
+      "learning_rate": 0.001,
+      "loss": 2.3957,
+      "step": 18267
+    },
+    {
+      "epoch": 0.7728234199170827,
+      "grad_norm": 0.1581333428621292,
+      "learning_rate": 0.001,
+      "loss": 2.1276,
+      "step": 18268
+    },
+    {
+      "epoch": 0.772865724680599,
+      "grad_norm": 0.1533811092376709,
+      "learning_rate": 0.001,
+      "loss": 1.9765,
+      "step": 18269
+    },
+    {
+      "epoch": 0.7729080294441154,
+      "grad_norm": 0.3548398017883301,
+      "learning_rate": 0.001,
+      "loss": 3.269,
+      "step": 18270
+    },
+    {
+      "epoch": 0.7729503342076318,
+      "grad_norm": 0.16134926676750183,
+      "learning_rate": 0.001,
+      "loss": 2.0527,
+      "step": 18271
+    },
+    {
+      "epoch": 0.7729926389711481,
+      "grad_norm": 0.713190495967865,
+      "learning_rate": 0.001,
+      "loss": 1.8366,
+      "step": 18272
+    },
+    {
+      "epoch": 0.7730349437346645,
+      "grad_norm": 0.8022021055221558,
+      "learning_rate": 0.001,
+      "loss": 1.7423,
+      "step": 18273
+    },
+    {
+      "epoch": 0.7730772484981809,
+      "grad_norm": 0.38094237446784973,
+      "learning_rate": 0.001,
+      "loss": 2.4639,
+      "step": 18274
+    },
+    {
+      "epoch": 0.7731195532616972,
+      "grad_norm": 0.1615244299173355,
+      "learning_rate": 0.001,
+      "loss": 2.2037,
+      "step": 18275
+    },
+    {
+      "epoch": 0.7731618580252136,
+      "grad_norm": 0.23506537079811096,
+      "learning_rate": 0.001,
+      "loss": 2.9041,
+      "step": 18276
+    },
+    {
+      "epoch": 0.77320416278873,
+      "grad_norm": 0.26342645287513733,
+      "learning_rate": 0.001,
+      "loss": 2.0227,
+      "step": 18277
+    },
+    {
+      "epoch": 0.7732464675522464,
+      "grad_norm": 0.16665269434452057,
+      "learning_rate": 0.001,
+      "loss": 1.5072,
+      "step": 18278
+    },
+    {
+      "epoch": 0.7732887723157628,
+      "grad_norm": 0.15748518705368042,
+      "learning_rate": 0.001,
+      "loss": 1.9506,
+      "step": 18279
+    },
+    {
+      "epoch": 0.7733310770792792,
+      "grad_norm": 0.20437052845954895,
+      "learning_rate": 0.001,
+      "loss": 1.7317,
+      "step": 18280
+    },
+    {
+      "epoch": 0.7733733818427955,
+      "grad_norm": 0.1785205602645874,
+      "learning_rate": 0.001,
+      "loss": 2.6008,
+      "step": 18281
+    },
+    {
+      "epoch": 0.7734156866063119,
+      "grad_norm": 0.15185962617397308,
+      "learning_rate": 0.001,
+      "loss": 2.136,
+      "step": 18282
+    },
+    {
+      "epoch": 0.7734579913698283,
+      "grad_norm": 0.1808384209871292,
+      "learning_rate": 0.001,
+      "loss": 1.8633,
+      "step": 18283
+    },
+    {
+      "epoch": 0.7735002961333446,
+      "grad_norm": 0.2651410698890686,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 18284
+    },
+    {
+      "epoch": 0.773542600896861,
+      "grad_norm": 0.15414567291736603,
+      "learning_rate": 0.001,
+      "loss": 2.2774,
+      "step": 18285
+    },
+    {
+      "epoch": 0.7735849056603774,
+      "grad_norm": 0.16589045524597168,
+      "learning_rate": 0.001,
+      "loss": 1.5533,
+      "step": 18286
+    },
+    {
+      "epoch": 0.7736272104238937,
+      "grad_norm": 0.15032519400119781,
+      "learning_rate": 0.001,
+      "loss": 1.6888,
+      "step": 18287
+    },
+    {
+      "epoch": 0.7736695151874101,
+      "grad_norm": 0.7096246480941772,
+      "learning_rate": 0.001,
+      "loss": 1.9838,
+      "step": 18288
+    },
+    {
+      "epoch": 0.7737118199509265,
+      "grad_norm": 0.16361872851848602,
+      "learning_rate": 0.001,
+      "loss": 1.6044,
+      "step": 18289
+    },
+    {
+      "epoch": 0.7737541247144428,
+      "grad_norm": 0.3765888810157776,
+      "learning_rate": 0.001,
+      "loss": 1.7739,
+      "step": 18290
+    },
+    {
+      "epoch": 0.7737964294779592,
+      "grad_norm": 0.3084273934364319,
+      "learning_rate": 0.001,
+      "loss": 1.2695,
+      "step": 18291
+    },
+    {
+      "epoch": 0.7738387342414756,
+      "grad_norm": 0.17378173768520355,
+      "learning_rate": 0.001,
+      "loss": 1.9922,
+      "step": 18292
+    },
+    {
+      "epoch": 0.7738810390049919,
+      "grad_norm": 0.671561598777771,
+      "learning_rate": 0.001,
+      "loss": 2.7077,
+      "step": 18293
+    },
+    {
+      "epoch": 0.7739233437685084,
+      "grad_norm": 0.17134688794612885,
+      "learning_rate": 0.001,
+      "loss": 2.1455,
+      "step": 18294
+    },
+    {
+      "epoch": 0.7739656485320247,
+      "grad_norm": 0.15048515796661377,
+      "learning_rate": 0.001,
+      "loss": 2.2337,
+      "step": 18295
+    },
+    {
+      "epoch": 0.7740079532955411,
+      "grad_norm": 0.14688226580619812,
+      "learning_rate": 0.001,
+      "loss": 2.2607,
+      "step": 18296
+    },
+    {
+      "epoch": 0.7740502580590575,
+      "grad_norm": 0.15102264285087585,
+      "learning_rate": 0.001,
+      "loss": 2.3796,
+      "step": 18297
+    },
+    {
+      "epoch": 0.7740925628225738,
+      "grad_norm": 0.15952768921852112,
+      "learning_rate": 0.001,
+      "loss": 2.2732,
+      "step": 18298
+    },
+    {
+      "epoch": 0.7741348675860902,
+      "grad_norm": 0.15502522885799408,
+      "learning_rate": 0.001,
+      "loss": 2.3354,
+      "step": 18299
+    },
+    {
+      "epoch": 0.7741771723496066,
+      "grad_norm": 0.17360159754753113,
+      "learning_rate": 0.001,
+      "loss": 1.4465,
+      "step": 18300
+    },
+    {
+      "epoch": 0.7742194771131229,
+      "grad_norm": 0.14462324976921082,
+      "learning_rate": 0.001,
+      "loss": 1.5657,
+      "step": 18301
+    },
+    {
+      "epoch": 0.7742617818766393,
+      "grad_norm": 0.22618313133716583,
+      "learning_rate": 0.001,
+      "loss": 1.497,
+      "step": 18302
+    },
+    {
+      "epoch": 0.7743040866401557,
+      "grad_norm": 0.14733265340328217,
+      "learning_rate": 0.001,
+      "loss": 2.7168,
+      "step": 18303
+    },
+    {
+      "epoch": 0.774346391403672,
+      "grad_norm": 0.16763785481452942,
+      "learning_rate": 0.001,
+      "loss": 2.0218,
+      "step": 18304
+    },
+    {
+      "epoch": 0.7743886961671884,
+      "grad_norm": 0.14719419181346893,
+      "learning_rate": 0.001,
+      "loss": 1.8908,
+      "step": 18305
+    },
+    {
+      "epoch": 0.7744310009307048,
+      "grad_norm": 0.14694784581661224,
+      "learning_rate": 0.001,
+      "loss": 2.3633,
+      "step": 18306
+    },
+    {
+      "epoch": 0.7744733056942211,
+      "grad_norm": 0.1397048681974411,
+      "learning_rate": 0.001,
+      "loss": 1.574,
+      "step": 18307
+    },
+    {
+      "epoch": 0.7745156104577375,
+      "grad_norm": 0.138363778591156,
+      "learning_rate": 0.001,
+      "loss": 2.1553,
+      "step": 18308
+    },
+    {
+      "epoch": 0.774557915221254,
+      "grad_norm": 0.17978732287883759,
+      "learning_rate": 0.001,
+      "loss": 2.1659,
+      "step": 18309
+    },
+    {
+      "epoch": 0.7746002199847702,
+      "grad_norm": 2.1564154624938965,
+      "learning_rate": 0.001,
+      "loss": 2.5153,
+      "step": 18310
+    },
+    {
+      "epoch": 0.7746425247482867,
+      "grad_norm": 0.1512441188097,
+      "learning_rate": 0.001,
+      "loss": 1.7129,
+      "step": 18311
+    },
+    {
+      "epoch": 0.7746848295118031,
+      "grad_norm": 0.21785925328731537,
+      "learning_rate": 0.001,
+      "loss": 2.0366,
+      "step": 18312
+    },
+    {
+      "epoch": 0.7747271342753194,
+      "grad_norm": 0.5639128684997559,
+      "learning_rate": 0.001,
+      "loss": 1.4766,
+      "step": 18313
+    },
+    {
+      "epoch": 0.7747694390388358,
+      "grad_norm": 0.2070891112089157,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 18314
+    },
+    {
+      "epoch": 0.7748117438023522,
+      "grad_norm": 0.16625727713108063,
+      "learning_rate": 0.001,
+      "loss": 1.5067,
+      "step": 18315
+    },
+    {
+      "epoch": 0.7748540485658685,
+      "grad_norm": 0.5115146040916443,
+      "learning_rate": 0.001,
+      "loss": 2.7549,
+      "step": 18316
+    },
+    {
+      "epoch": 0.7748963533293849,
+      "grad_norm": 0.17250525951385498,
+      "learning_rate": 0.001,
+      "loss": 1.7243,
+      "step": 18317
+    },
+    {
+      "epoch": 0.7749386580929013,
+      "grad_norm": 1.3713905811309814,
+      "learning_rate": 0.001,
+      "loss": 2.225,
+      "step": 18318
+    },
+    {
+      "epoch": 0.7749809628564176,
+      "grad_norm": 0.21677060425281525,
+      "learning_rate": 0.001,
+      "loss": 3.3116,
+      "step": 18319
+    },
+    {
+      "epoch": 0.775023267619934,
+      "grad_norm": 0.21051226556301117,
+      "learning_rate": 0.001,
+      "loss": 2.2168,
+      "step": 18320
+    },
+    {
+      "epoch": 0.7750655723834504,
+      "grad_norm": 0.15511994063854218,
+      "learning_rate": 0.001,
+      "loss": 1.6332,
+      "step": 18321
+    },
+    {
+      "epoch": 0.7751078771469667,
+      "grad_norm": 0.19078783690929413,
+      "learning_rate": 0.001,
+      "loss": 2.149,
+      "step": 18322
+    },
+    {
+      "epoch": 0.7751501819104831,
+      "grad_norm": 0.186100572347641,
+      "learning_rate": 0.001,
+      "loss": 2.3936,
+      "step": 18323
+    },
+    {
+      "epoch": 0.7751924866739995,
+      "grad_norm": 0.19699545204639435,
+      "learning_rate": 0.001,
+      "loss": 1.6859,
+      "step": 18324
+    },
+    {
+      "epoch": 0.7752347914375158,
+      "grad_norm": 0.1534021496772766,
+      "learning_rate": 0.001,
+      "loss": 1.7206,
+      "step": 18325
+    },
+    {
+      "epoch": 0.7752770962010322,
+      "grad_norm": 0.1593978852033615,
+      "learning_rate": 0.001,
+      "loss": 2.3795,
+      "step": 18326
+    },
+    {
+      "epoch": 0.7753194009645487,
+      "grad_norm": 0.9152129292488098,
+      "learning_rate": 0.001,
+      "loss": 2.5192,
+      "step": 18327
+    },
+    {
+      "epoch": 0.775361705728065,
+      "grad_norm": 3.8987066745758057,
+      "learning_rate": 0.001,
+      "loss": 1.7634,
+      "step": 18328
+    },
+    {
+      "epoch": 0.7754040104915814,
+      "grad_norm": 0.16911078989505768,
+      "learning_rate": 0.001,
+      "loss": 1.7858,
+      "step": 18329
+    },
+    {
+      "epoch": 0.7754463152550978,
+      "grad_norm": 0.1814729869365692,
+      "learning_rate": 0.001,
+      "loss": 2.2558,
+      "step": 18330
+    },
+    {
+      "epoch": 0.7754886200186141,
+      "grad_norm": 0.15336821973323822,
+      "learning_rate": 0.001,
+      "loss": 2.127,
+      "step": 18331
+    },
+    {
+      "epoch": 0.7755309247821305,
+      "grad_norm": 0.7835121750831604,
+      "learning_rate": 0.001,
+      "loss": 2.1583,
+      "step": 18332
+    },
+    {
+      "epoch": 0.7755732295456469,
+      "grad_norm": 0.19294503331184387,
+      "learning_rate": 0.001,
+      "loss": 2.2585,
+      "step": 18333
+    },
+    {
+      "epoch": 0.7756155343091632,
+      "grad_norm": 6.45863151550293,
+      "learning_rate": 0.001,
+      "loss": 2.1715,
+      "step": 18334
+    },
+    {
+      "epoch": 0.7756578390726796,
+      "grad_norm": 0.17118413746356964,
+      "learning_rate": 0.001,
+      "loss": 1.3514,
+      "step": 18335
+    },
+    {
+      "epoch": 0.775700143836196,
+      "grad_norm": 0.24696171283721924,
+      "learning_rate": 0.001,
+      "loss": 2.2035,
+      "step": 18336
+    },
+    {
+      "epoch": 0.7757424485997123,
+      "grad_norm": 0.33256638050079346,
+      "learning_rate": 0.001,
+      "loss": 2.5719,
+      "step": 18337
+    },
+    {
+      "epoch": 0.7757847533632287,
+      "grad_norm": 0.19757241010665894,
+      "learning_rate": 0.001,
+      "loss": 2.4873,
+      "step": 18338
+    },
+    {
+      "epoch": 0.775827058126745,
+      "grad_norm": 0.17525336146354675,
+      "learning_rate": 0.001,
+      "loss": 2.2333,
+      "step": 18339
+    },
+    {
+      "epoch": 0.7758693628902614,
+      "grad_norm": 0.15870629251003265,
+      "learning_rate": 0.001,
+      "loss": 1.8791,
+      "step": 18340
+    },
+    {
+      "epoch": 0.7759116676537778,
+      "grad_norm": 0.13413628935813904,
+      "learning_rate": 0.001,
+      "loss": 2.8583,
+      "step": 18341
+    },
+    {
+      "epoch": 0.7759539724172941,
+      "grad_norm": 0.15136483311653137,
+      "learning_rate": 0.001,
+      "loss": 2.368,
+      "step": 18342
+    },
+    {
+      "epoch": 0.7759962771808105,
+      "grad_norm": 0.15002518892288208,
+      "learning_rate": 0.001,
+      "loss": 1.6174,
+      "step": 18343
+    },
+    {
+      "epoch": 0.776038581944327,
+      "grad_norm": 0.15149597823619843,
+      "learning_rate": 0.001,
+      "loss": 2.7551,
+      "step": 18344
+    },
+    {
+      "epoch": 0.7760808867078433,
+      "grad_norm": 0.548881471157074,
+      "learning_rate": 0.001,
+      "loss": 1.8676,
+      "step": 18345
+    },
+    {
+      "epoch": 0.7761231914713597,
+      "grad_norm": 0.1294621080160141,
+      "learning_rate": 0.001,
+      "loss": 1.3881,
+      "step": 18346
+    },
+    {
+      "epoch": 0.7761654962348761,
+      "grad_norm": 0.1591734141111374,
+      "learning_rate": 0.001,
+      "loss": 2.6516,
+      "step": 18347
+    },
+    {
+      "epoch": 0.7762078009983924,
+      "grad_norm": 0.7913060784339905,
+      "learning_rate": 0.001,
+      "loss": 2.4505,
+      "step": 18348
+    },
+    {
+      "epoch": 0.7762501057619088,
+      "grad_norm": 8.328814506530762,
+      "learning_rate": 0.001,
+      "loss": 2.4966,
+      "step": 18349
+    },
+    {
+      "epoch": 0.7762924105254252,
+      "grad_norm": 0.15105003118515015,
+      "learning_rate": 0.001,
+      "loss": 2.0033,
+      "step": 18350
+    },
+    {
+      "epoch": 0.7763347152889415,
+      "grad_norm": 0.1463308483362198,
+      "learning_rate": 0.001,
+      "loss": 2.31,
+      "step": 18351
+    },
+    {
+      "epoch": 0.7763770200524579,
+      "grad_norm": 1.781333327293396,
+      "learning_rate": 0.001,
+      "loss": 3.0765,
+      "step": 18352
+    },
+    {
+      "epoch": 0.7764193248159743,
+      "grad_norm": 0.17749984562397003,
+      "learning_rate": 0.001,
+      "loss": 2.4567,
+      "step": 18353
+    },
+    {
+      "epoch": 0.7764616295794906,
+      "grad_norm": 0.16004282236099243,
+      "learning_rate": 0.001,
+      "loss": 2.1417,
+      "step": 18354
+    },
+    {
+      "epoch": 0.776503934343007,
+      "grad_norm": 0.18965484201908112,
+      "learning_rate": 0.001,
+      "loss": 2.5025,
+      "step": 18355
+    },
+    {
+      "epoch": 0.7765462391065234,
+      "grad_norm": 0.4342174530029297,
+      "learning_rate": 0.001,
+      "loss": 1.8881,
+      "step": 18356
+    },
+    {
+      "epoch": 0.7765885438700397,
+      "grad_norm": 1.2400356531143188,
+      "learning_rate": 0.001,
+      "loss": 3.1552,
+      "step": 18357
+    },
+    {
+      "epoch": 0.7766308486335561,
+      "grad_norm": 0.33312109112739563,
+      "learning_rate": 0.001,
+      "loss": 2.3914,
+      "step": 18358
+    },
+    {
+      "epoch": 0.7766731533970725,
+      "grad_norm": 0.1646675318479538,
+      "learning_rate": 0.001,
+      "loss": 2.3323,
+      "step": 18359
+    },
+    {
+      "epoch": 0.7767154581605888,
+      "grad_norm": 0.16873008012771606,
+      "learning_rate": 0.001,
+      "loss": 2.777,
+      "step": 18360
+    },
+    {
+      "epoch": 0.7767577629241053,
+      "grad_norm": 0.2983357906341553,
+      "learning_rate": 0.001,
+      "loss": 2.6237,
+      "step": 18361
+    },
+    {
+      "epoch": 0.7768000676876217,
+      "grad_norm": 10.579153060913086,
+      "learning_rate": 0.001,
+      "loss": 1.5843,
+      "step": 18362
+    },
+    {
+      "epoch": 0.776842372451138,
+      "grad_norm": 0.1804652363061905,
+      "learning_rate": 0.001,
+      "loss": 2.4111,
+      "step": 18363
+    },
+    {
+      "epoch": 0.7768846772146544,
+      "grad_norm": 0.24087312817573547,
+      "learning_rate": 0.001,
+      "loss": 2.2012,
+      "step": 18364
+    },
+    {
+      "epoch": 0.7769269819781708,
+      "grad_norm": 0.19113610684871674,
+      "learning_rate": 0.001,
+      "loss": 2.2822,
+      "step": 18365
+    },
+    {
+      "epoch": 0.7769692867416871,
+      "grad_norm": 0.20484258234500885,
+      "learning_rate": 0.001,
+      "loss": 2.0493,
+      "step": 18366
+    },
+    {
+      "epoch": 0.7770115915052035,
+      "grad_norm": 0.19894827902317047,
+      "learning_rate": 0.001,
+      "loss": 3.3502,
+      "step": 18367
+    },
+    {
+      "epoch": 0.7770538962687199,
+      "grad_norm": 0.20970451831817627,
+      "learning_rate": 0.001,
+      "loss": 1.6449,
+      "step": 18368
+    },
+    {
+      "epoch": 0.7770962010322362,
+      "grad_norm": 0.1429346650838852,
+      "learning_rate": 0.001,
+      "loss": 1.8039,
+      "step": 18369
+    },
+    {
+      "epoch": 0.7771385057957526,
+      "grad_norm": 0.16938546299934387,
+      "learning_rate": 0.001,
+      "loss": 1.9748,
+      "step": 18370
+    },
+    {
+      "epoch": 0.777180810559269,
+      "grad_norm": 0.9999366998672485,
+      "learning_rate": 0.001,
+      "loss": 1.5648,
+      "step": 18371
+    },
+    {
+      "epoch": 0.7772231153227853,
+      "grad_norm": 0.16697566211223602,
+      "learning_rate": 0.001,
+      "loss": 1.9137,
+      "step": 18372
+    },
+    {
+      "epoch": 0.7772654200863017,
+      "grad_norm": 0.1452540159225464,
+      "learning_rate": 0.001,
+      "loss": 2.1988,
+      "step": 18373
+    },
+    {
+      "epoch": 0.7773077248498181,
+      "grad_norm": 0.19160769879817963,
+      "learning_rate": 0.001,
+      "loss": 3.139,
+      "step": 18374
+    },
+    {
+      "epoch": 0.7773500296133344,
+      "grad_norm": 0.1582278609275818,
+      "learning_rate": 0.001,
+      "loss": 2.0894,
+      "step": 18375
+    },
+    {
+      "epoch": 0.7773923343768508,
+      "grad_norm": 0.27617523074150085,
+      "learning_rate": 0.001,
+      "loss": 2.507,
+      "step": 18376
+    },
+    {
+      "epoch": 0.7774346391403673,
+      "grad_norm": 0.15912631154060364,
+      "learning_rate": 0.001,
+      "loss": 1.2789,
+      "step": 18377
+    },
+    {
+      "epoch": 0.7774769439038836,
+      "grad_norm": 0.13841427862644196,
+      "learning_rate": 0.001,
+      "loss": 1.6491,
+      "step": 18378
+    },
+    {
+      "epoch": 0.7775192486674,
+      "grad_norm": 0.1486634612083435,
+      "learning_rate": 0.001,
+      "loss": 1.6654,
+      "step": 18379
+    },
+    {
+      "epoch": 0.7775615534309164,
+      "grad_norm": 0.16134053468704224,
+      "learning_rate": 0.001,
+      "loss": 1.6808,
+      "step": 18380
+    },
+    {
+      "epoch": 0.7776038581944327,
+      "grad_norm": 0.5859285593032837,
+      "learning_rate": 0.001,
+      "loss": 1.9378,
+      "step": 18381
+    },
+    {
+      "epoch": 0.7776461629579491,
+      "grad_norm": 0.1989150196313858,
+      "learning_rate": 0.001,
+      "loss": 2.8664,
+      "step": 18382
+    },
+    {
+      "epoch": 0.7776884677214654,
+      "grad_norm": 0.21168352663516998,
+      "learning_rate": 0.001,
+      "loss": 2.7791,
+      "step": 18383
+    },
+    {
+      "epoch": 0.7777307724849818,
+      "grad_norm": 0.16079917550086975,
+      "learning_rate": 0.001,
+      "loss": 1.5344,
+      "step": 18384
+    },
+    {
+      "epoch": 0.7777730772484982,
+      "grad_norm": 0.17647407948970795,
+      "learning_rate": 0.001,
+      "loss": 2.9867,
+      "step": 18385
+    },
+    {
+      "epoch": 0.7778153820120145,
+      "grad_norm": 0.15070800483226776,
+      "learning_rate": 0.001,
+      "loss": 2.0682,
+      "step": 18386
+    },
+    {
+      "epoch": 0.7778576867755309,
+      "grad_norm": 0.1633099466562271,
+      "learning_rate": 0.001,
+      "loss": 1.5573,
+      "step": 18387
+    },
+    {
+      "epoch": 0.7778999915390473,
+      "grad_norm": 0.1457083374261856,
+      "learning_rate": 0.001,
+      "loss": 1.8068,
+      "step": 18388
+    },
+    {
+      "epoch": 0.7779422963025636,
+      "grad_norm": 0.1393345296382904,
+      "learning_rate": 0.001,
+      "loss": 2.3662,
+      "step": 18389
+    },
+    {
+      "epoch": 0.77798460106608,
+      "grad_norm": 0.17682814598083496,
+      "learning_rate": 0.001,
+      "loss": 2.6571,
+      "step": 18390
+    },
+    {
+      "epoch": 0.7780269058295964,
+      "grad_norm": 0.1801498383283615,
+      "learning_rate": 0.001,
+      "loss": 2.6531,
+      "step": 18391
+    },
+    {
+      "epoch": 0.7780692105931127,
+      "grad_norm": 0.14461541175842285,
+      "learning_rate": 0.001,
+      "loss": 1.6549,
+      "step": 18392
+    },
+    {
+      "epoch": 0.7781115153566291,
+      "grad_norm": 0.16726653277873993,
+      "learning_rate": 0.001,
+      "loss": 2.2859,
+      "step": 18393
+    },
+    {
+      "epoch": 0.7781538201201456,
+      "grad_norm": 0.1532205492258072,
+      "learning_rate": 0.001,
+      "loss": 1.4662,
+      "step": 18394
+    },
+    {
+      "epoch": 0.7781961248836619,
+      "grad_norm": 0.13347402215003967,
+      "learning_rate": 0.001,
+      "loss": 2.1848,
+      "step": 18395
+    },
+    {
+      "epoch": 0.7782384296471783,
+      "grad_norm": 0.20856840908527374,
+      "learning_rate": 0.001,
+      "loss": 2.0527,
+      "step": 18396
+    },
+    {
+      "epoch": 0.7782807344106947,
+      "grad_norm": 0.15012764930725098,
+      "learning_rate": 0.001,
+      "loss": 2.0695,
+      "step": 18397
+    },
+    {
+      "epoch": 0.778323039174211,
+      "grad_norm": 2.177178144454956,
+      "learning_rate": 0.001,
+      "loss": 1.4426,
+      "step": 18398
+    },
+    {
+      "epoch": 0.7783653439377274,
+      "grad_norm": 0.22611618041992188,
+      "learning_rate": 0.001,
+      "loss": 2.052,
+      "step": 18399
+    },
+    {
+      "epoch": 0.7784076487012438,
+      "grad_norm": 0.1328599900007248,
+      "learning_rate": 0.001,
+      "loss": 1.7561,
+      "step": 18400
+    },
+    {
+      "epoch": 0.7784499534647601,
+      "grad_norm": 0.14337456226348877,
+      "learning_rate": 0.001,
+      "loss": 1.1991,
+      "step": 18401
+    },
+    {
+      "epoch": 0.7784922582282765,
+      "grad_norm": 0.15648649632930756,
+      "learning_rate": 0.001,
+      "loss": 2.0966,
+      "step": 18402
+    },
+    {
+      "epoch": 0.7785345629917929,
+      "grad_norm": 0.15187643468379974,
+      "learning_rate": 0.001,
+      "loss": 1.8629,
+      "step": 18403
+    },
+    {
+      "epoch": 0.7785768677553092,
+      "grad_norm": 0.12914754450321198,
+      "learning_rate": 0.001,
+      "loss": 1.6909,
+      "step": 18404
+    },
+    {
+      "epoch": 0.7786191725188256,
+      "grad_norm": 0.14733096957206726,
+      "learning_rate": 0.001,
+      "loss": 1.9314,
+      "step": 18405
+    },
+    {
+      "epoch": 0.778661477282342,
+      "grad_norm": 0.19098341464996338,
+      "learning_rate": 0.001,
+      "loss": 2.2461,
+      "step": 18406
+    },
+    {
+      "epoch": 0.7787037820458583,
+      "grad_norm": 0.14427435398101807,
+      "learning_rate": 0.001,
+      "loss": 1.5976,
+      "step": 18407
+    },
+    {
+      "epoch": 0.7787460868093747,
+      "grad_norm": 0.14376568794250488,
+      "learning_rate": 0.001,
+      "loss": 1.7142,
+      "step": 18408
+    },
+    {
+      "epoch": 0.7787883915728911,
+      "grad_norm": 0.1582573801279068,
+      "learning_rate": 0.001,
+      "loss": 3.5712,
+      "step": 18409
+    },
+    {
+      "epoch": 0.7788306963364074,
+      "grad_norm": 0.16873016953468323,
+      "learning_rate": 0.001,
+      "loss": 2.0883,
+      "step": 18410
+    },
+    {
+      "epoch": 0.7788730010999239,
+      "grad_norm": 0.14867006242275238,
+      "learning_rate": 0.001,
+      "loss": 1.5226,
+      "step": 18411
+    },
+    {
+      "epoch": 0.7789153058634403,
+      "grad_norm": 0.1830413043498993,
+      "learning_rate": 0.001,
+      "loss": 1.5773,
+      "step": 18412
+    },
+    {
+      "epoch": 0.7789576106269566,
+      "grad_norm": 0.21609345078468323,
+      "learning_rate": 0.001,
+      "loss": 2.2652,
+      "step": 18413
+    },
+    {
+      "epoch": 0.778999915390473,
+      "grad_norm": 0.36571288108825684,
+      "learning_rate": 0.001,
+      "loss": 1.8297,
+      "step": 18414
+    },
+    {
+      "epoch": 0.7790422201539894,
+      "grad_norm": 0.15142600238323212,
+      "learning_rate": 0.001,
+      "loss": 1.5149,
+      "step": 18415
+    },
+    {
+      "epoch": 0.7790845249175057,
+      "grad_norm": 0.7442256212234497,
+      "learning_rate": 0.001,
+      "loss": 4.0185,
+      "step": 18416
+    },
+    {
+      "epoch": 0.7791268296810221,
+      "grad_norm": 0.28748589754104614,
+      "learning_rate": 0.001,
+      "loss": 3.1263,
+      "step": 18417
+    },
+    {
+      "epoch": 0.7791691344445385,
+      "grad_norm": 0.16540317237377167,
+      "learning_rate": 0.001,
+      "loss": 3.5016,
+      "step": 18418
+    },
+    {
+      "epoch": 0.7792114392080548,
+      "grad_norm": 0.15089643001556396,
+      "learning_rate": 0.001,
+      "loss": 2.1887,
+      "step": 18419
+    },
+    {
+      "epoch": 0.7792537439715712,
+      "grad_norm": 0.18112583458423615,
+      "learning_rate": 0.001,
+      "loss": 1.6532,
+      "step": 18420
+    },
+    {
+      "epoch": 0.7792960487350876,
+      "grad_norm": 0.14614035189151764,
+      "learning_rate": 0.001,
+      "loss": 1.8557,
+      "step": 18421
+    },
+    {
+      "epoch": 0.7793383534986039,
+      "grad_norm": 0.1359279751777649,
+      "learning_rate": 0.001,
+      "loss": 1.6901,
+      "step": 18422
+    },
+    {
+      "epoch": 0.7793806582621203,
+      "grad_norm": 1.8162649869918823,
+      "learning_rate": 0.001,
+      "loss": 2.0045,
+      "step": 18423
+    },
+    {
+      "epoch": 0.7794229630256367,
+      "grad_norm": 0.33419543504714966,
+      "learning_rate": 0.001,
+      "loss": 2.7161,
+      "step": 18424
+    },
+    {
+      "epoch": 0.779465267789153,
+      "grad_norm": 0.14467467367649078,
+      "learning_rate": 0.001,
+      "loss": 1.7674,
+      "step": 18425
+    },
+    {
+      "epoch": 0.7795075725526694,
+      "grad_norm": 0.17725689709186554,
+      "learning_rate": 0.001,
+      "loss": 1.6678,
+      "step": 18426
+    },
+    {
+      "epoch": 0.7795498773161859,
+      "grad_norm": 0.1722497195005417,
+      "learning_rate": 0.001,
+      "loss": 1.9924,
+      "step": 18427
+    },
+    {
+      "epoch": 0.7795921820797022,
+      "grad_norm": 0.17150290310382843,
+      "learning_rate": 0.001,
+      "loss": 1.5635,
+      "step": 18428
+    },
+    {
+      "epoch": 0.7796344868432186,
+      "grad_norm": 0.14820575714111328,
+      "learning_rate": 0.001,
+      "loss": 2.9991,
+      "step": 18429
+    },
+    {
+      "epoch": 0.7796767916067349,
+      "grad_norm": 0.18718744814395905,
+      "learning_rate": 0.001,
+      "loss": 2.1338,
+      "step": 18430
+    },
+    {
+      "epoch": 0.7797190963702513,
+      "grad_norm": 0.17354276776313782,
+      "learning_rate": 0.001,
+      "loss": 2.0303,
+      "step": 18431
+    },
+    {
+      "epoch": 0.7797614011337677,
+      "grad_norm": 0.17366386950016022,
+      "learning_rate": 0.001,
+      "loss": 1.5947,
+      "step": 18432
+    },
+    {
+      "epoch": 0.779803705897284,
+      "grad_norm": 0.1608889102935791,
+      "learning_rate": 0.001,
+      "loss": 3.5028,
+      "step": 18433
+    },
+    {
+      "epoch": 0.7798460106608004,
+      "grad_norm": 0.16302143037319183,
+      "learning_rate": 0.001,
+      "loss": 2.1618,
+      "step": 18434
+    },
+    {
+      "epoch": 0.7798883154243168,
+      "grad_norm": 0.19998212158679962,
+      "learning_rate": 0.001,
+      "loss": 1.3854,
+      "step": 18435
+    },
+    {
+      "epoch": 0.7799306201878331,
+      "grad_norm": 0.20762750506401062,
+      "learning_rate": 0.001,
+      "loss": 3.0683,
+      "step": 18436
+    },
+    {
+      "epoch": 0.7799729249513495,
+      "grad_norm": 0.1632770299911499,
+      "learning_rate": 0.001,
+      "loss": 1.3391,
+      "step": 18437
+    },
+    {
+      "epoch": 0.7800152297148659,
+      "grad_norm": 0.4506990909576416,
+      "learning_rate": 0.001,
+      "loss": 3.0315,
+      "step": 18438
+    },
+    {
+      "epoch": 0.7800575344783822,
+      "grad_norm": 0.1689714938402176,
+      "learning_rate": 0.001,
+      "loss": 2.7837,
+      "step": 18439
+    },
+    {
+      "epoch": 0.7800998392418986,
+      "grad_norm": 0.16053323447704315,
+      "learning_rate": 0.001,
+      "loss": 1.4578,
+      "step": 18440
+    },
+    {
+      "epoch": 0.780142144005415,
+      "grad_norm": 0.14367254078388214,
+      "learning_rate": 0.001,
+      "loss": 2.418,
+      "step": 18441
+    },
+    {
+      "epoch": 0.7801844487689313,
+      "grad_norm": 0.6091342568397522,
+      "learning_rate": 0.001,
+      "loss": 2.745,
+      "step": 18442
+    },
+    {
+      "epoch": 0.7802267535324477,
+      "grad_norm": 0.2298995554447174,
+      "learning_rate": 0.001,
+      "loss": 3.9425,
+      "step": 18443
+    },
+    {
+      "epoch": 0.7802690582959642,
+      "grad_norm": 0.15140703320503235,
+      "learning_rate": 0.001,
+      "loss": 2.7755,
+      "step": 18444
+    },
+    {
+      "epoch": 0.7803113630594805,
+      "grad_norm": 0.29984918236732483,
+      "learning_rate": 0.001,
+      "loss": 2.1757,
+      "step": 18445
+    },
+    {
+      "epoch": 0.7803536678229969,
+      "grad_norm": 0.18075025081634521,
+      "learning_rate": 0.001,
+      "loss": 1.7825,
+      "step": 18446
+    },
+    {
+      "epoch": 0.7803959725865133,
+      "grad_norm": 0.23606182634830475,
+      "learning_rate": 0.001,
+      "loss": 1.9542,
+      "step": 18447
+    },
+    {
+      "epoch": 0.7804382773500296,
+      "grad_norm": 0.17639142274856567,
+      "learning_rate": 0.001,
+      "loss": 2.5699,
+      "step": 18448
+    },
+    {
+      "epoch": 0.780480582113546,
+      "grad_norm": 0.3127742111682892,
+      "learning_rate": 0.001,
+      "loss": 1.3408,
+      "step": 18449
+    },
+    {
+      "epoch": 0.7805228868770624,
+      "grad_norm": 0.2586624026298523,
+      "learning_rate": 0.001,
+      "loss": 1.4728,
+      "step": 18450
+    },
+    {
+      "epoch": 0.7805651916405787,
+      "grad_norm": 0.14627474546432495,
+      "learning_rate": 0.001,
+      "loss": 1.5882,
+      "step": 18451
+    },
+    {
+      "epoch": 0.7806074964040951,
+      "grad_norm": 0.22566846013069153,
+      "learning_rate": 0.001,
+      "loss": 1.6143,
+      "step": 18452
+    },
+    {
+      "epoch": 0.7806498011676115,
+      "grad_norm": 0.5345284342765808,
+      "learning_rate": 0.001,
+      "loss": 2.4135,
+      "step": 18453
+    },
+    {
+      "epoch": 0.7806921059311278,
+      "grad_norm": 0.16989070177078247,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 18454
+    },
+    {
+      "epoch": 0.7807344106946442,
+      "grad_norm": 0.7486119866371155,
+      "learning_rate": 0.001,
+      "loss": 2.1102,
+      "step": 18455
+    },
+    {
+      "epoch": 0.7807767154581606,
+      "grad_norm": 0.1473877727985382,
+      "learning_rate": 0.001,
+      "loss": 2.0649,
+      "step": 18456
+    },
+    {
+      "epoch": 0.7808190202216769,
+      "grad_norm": 0.1684478372335434,
+      "learning_rate": 0.001,
+      "loss": 1.9911,
+      "step": 18457
+    },
+    {
+      "epoch": 0.7808613249851933,
+      "grad_norm": 2.9781100749969482,
+      "learning_rate": 0.001,
+      "loss": 1.858,
+      "step": 18458
+    },
+    {
+      "epoch": 0.7809036297487097,
+      "grad_norm": 0.13891150057315826,
+      "learning_rate": 0.001,
+      "loss": 1.716,
+      "step": 18459
+    },
+    {
+      "epoch": 0.780945934512226,
+      "grad_norm": 0.16608332097530365,
+      "learning_rate": 0.001,
+      "loss": 1.6527,
+      "step": 18460
+    },
+    {
+      "epoch": 0.7809882392757425,
+      "grad_norm": 0.161564901471138,
+      "learning_rate": 0.001,
+      "loss": 2.0179,
+      "step": 18461
+    },
+    {
+      "epoch": 0.7810305440392589,
+      "grad_norm": 0.17590850591659546,
+      "learning_rate": 0.001,
+      "loss": 1.8993,
+      "step": 18462
+    },
+    {
+      "epoch": 0.7810728488027752,
+      "grad_norm": 0.15447676181793213,
+      "learning_rate": 0.001,
+      "loss": 2.4161,
+      "step": 18463
+    },
+    {
+      "epoch": 0.7811151535662916,
+      "grad_norm": 0.13972973823547363,
+      "learning_rate": 0.001,
+      "loss": 2.4988,
+      "step": 18464
+    },
+    {
+      "epoch": 0.781157458329808,
+      "grad_norm": 0.1585020124912262,
+      "learning_rate": 0.001,
+      "loss": 2.4953,
+      "step": 18465
+    },
+    {
+      "epoch": 0.7811997630933243,
+      "grad_norm": 0.7317923307418823,
+      "learning_rate": 0.001,
+      "loss": 2.9961,
+      "step": 18466
+    },
+    {
+      "epoch": 0.7812420678568407,
+      "grad_norm": 0.1768050640821457,
+      "learning_rate": 0.001,
+      "loss": 3.3468,
+      "step": 18467
+    },
+    {
+      "epoch": 0.7812843726203571,
+      "grad_norm": 0.22169825434684753,
+      "learning_rate": 0.001,
+      "loss": 2.0874,
+      "step": 18468
+    },
+    {
+      "epoch": 0.7813266773838734,
+      "grad_norm": 0.16282053291797638,
+      "learning_rate": 0.001,
+      "loss": 2.3298,
+      "step": 18469
+    },
+    {
+      "epoch": 0.7813689821473898,
+      "grad_norm": 0.13636824488639832,
+      "learning_rate": 0.001,
+      "loss": 1.5925,
+      "step": 18470
+    },
+    {
+      "epoch": 0.7814112869109062,
+      "grad_norm": 0.1489405781030655,
+      "learning_rate": 0.001,
+      "loss": 2.6353,
+      "step": 18471
+    },
+    {
+      "epoch": 0.7814535916744225,
+      "grad_norm": 0.19491471350193024,
+      "learning_rate": 0.001,
+      "loss": 3.0984,
+      "step": 18472
+    },
+    {
+      "epoch": 0.7814958964379389,
+      "grad_norm": 0.22249197959899902,
+      "learning_rate": 0.001,
+      "loss": 2.7739,
+      "step": 18473
+    },
+    {
+      "epoch": 0.7815382012014552,
+      "grad_norm": 0.21340025961399078,
+      "learning_rate": 0.001,
+      "loss": 2.2505,
+      "step": 18474
+    },
+    {
+      "epoch": 0.7815805059649716,
+      "grad_norm": 0.16923391819000244,
+      "learning_rate": 0.001,
+      "loss": 2.5597,
+      "step": 18475
+    },
+    {
+      "epoch": 0.781622810728488,
+      "grad_norm": 0.1564716249704361,
+      "learning_rate": 0.001,
+      "loss": 2.6755,
+      "step": 18476
+    },
+    {
+      "epoch": 0.7816651154920043,
+      "grad_norm": 0.18122835457324982,
+      "learning_rate": 0.001,
+      "loss": 1.8659,
+      "step": 18477
+    },
+    {
+      "epoch": 0.7817074202555208,
+      "grad_norm": 0.5826538801193237,
+      "learning_rate": 0.001,
+      "loss": 2.1115,
+      "step": 18478
+    },
+    {
+      "epoch": 0.7817497250190372,
+      "grad_norm": 0.15686936676502228,
+      "learning_rate": 0.001,
+      "loss": 2.1577,
+      "step": 18479
+    },
+    {
+      "epoch": 0.7817920297825535,
+      "grad_norm": 0.1563858985900879,
+      "learning_rate": 0.001,
+      "loss": 1.9294,
+      "step": 18480
+    },
+    {
+      "epoch": 0.7818343345460699,
+      "grad_norm": 0.6455838680267334,
+      "learning_rate": 0.001,
+      "loss": 2.0729,
+      "step": 18481
+    },
+    {
+      "epoch": 0.7818766393095863,
+      "grad_norm": 0.16011108458042145,
+      "learning_rate": 0.001,
+      "loss": 2.1436,
+      "step": 18482
+    },
+    {
+      "epoch": 0.7819189440731026,
+      "grad_norm": 0.1725652515888214,
+      "learning_rate": 0.001,
+      "loss": 2.3672,
+      "step": 18483
+    },
+    {
+      "epoch": 0.781961248836619,
+      "grad_norm": 0.17138071358203888,
+      "learning_rate": 0.001,
+      "loss": 2.05,
+      "step": 18484
+    },
+    {
+      "epoch": 0.7820035536001354,
+      "grad_norm": 0.16126024723052979,
+      "learning_rate": 0.001,
+      "loss": 1.8446,
+      "step": 18485
+    },
+    {
+      "epoch": 0.7820458583636517,
+      "grad_norm": 0.20708617568016052,
+      "learning_rate": 0.001,
+      "loss": 2.3597,
+      "step": 18486
+    },
+    {
+      "epoch": 0.7820881631271681,
+      "grad_norm": 0.3041653037071228,
+      "learning_rate": 0.001,
+      "loss": 2.9242,
+      "step": 18487
+    },
+    {
+      "epoch": 0.7821304678906845,
+      "grad_norm": 0.2294047325849533,
+      "learning_rate": 0.001,
+      "loss": 2.7331,
+      "step": 18488
+    },
+    {
+      "epoch": 0.7821727726542008,
+      "grad_norm": 0.16051490604877472,
+      "learning_rate": 0.001,
+      "loss": 1.3815,
+      "step": 18489
+    },
+    {
+      "epoch": 0.7822150774177172,
+      "grad_norm": 0.1533360332250595,
+      "learning_rate": 0.001,
+      "loss": 1.7686,
+      "step": 18490
+    },
+    {
+      "epoch": 0.7822573821812336,
+      "grad_norm": 0.3606812655925751,
+      "learning_rate": 0.001,
+      "loss": 2.5197,
+      "step": 18491
+    },
+    {
+      "epoch": 0.7822996869447499,
+      "grad_norm": 0.15387903153896332,
+      "learning_rate": 0.001,
+      "loss": 1.8732,
+      "step": 18492
+    },
+    {
+      "epoch": 0.7823419917082663,
+      "grad_norm": 0.1506211757659912,
+      "learning_rate": 0.001,
+      "loss": 1.9354,
+      "step": 18493
+    },
+    {
+      "epoch": 0.7823842964717828,
+      "grad_norm": 0.1458059847354889,
+      "learning_rate": 0.001,
+      "loss": 1.6647,
+      "step": 18494
+    },
+    {
+      "epoch": 0.7824266012352991,
+      "grad_norm": 0.1346750259399414,
+      "learning_rate": 0.001,
+      "loss": 1.8682,
+      "step": 18495
+    },
+    {
+      "epoch": 0.7824689059988155,
+      "grad_norm": 0.1503671556711197,
+      "learning_rate": 0.001,
+      "loss": 1.8601,
+      "step": 18496
+    },
+    {
+      "epoch": 0.7825112107623319,
+      "grad_norm": 0.15590086579322815,
+      "learning_rate": 0.001,
+      "loss": 3.2885,
+      "step": 18497
+    },
+    {
+      "epoch": 0.7825535155258482,
+      "grad_norm": 0.16297362744808197,
+      "learning_rate": 0.001,
+      "loss": 2.2588,
+      "step": 18498
+    },
+    {
+      "epoch": 0.7825958202893646,
+      "grad_norm": 0.15889789164066315,
+      "learning_rate": 0.001,
+      "loss": 1.5667,
+      "step": 18499
+    },
+    {
+      "epoch": 0.782638125052881,
+      "grad_norm": 0.16023693978786469,
+      "learning_rate": 0.001,
+      "loss": 2.3448,
+      "step": 18500
+    },
+    {
+      "epoch": 0.7826804298163973,
+      "grad_norm": 0.19175249338150024,
+      "learning_rate": 0.001,
+      "loss": 2.4261,
+      "step": 18501
+    },
+    {
+      "epoch": 0.7827227345799137,
+      "grad_norm": 0.14406760036945343,
+      "learning_rate": 0.001,
+      "loss": 1.5323,
+      "step": 18502
+    },
+    {
+      "epoch": 0.7827650393434301,
+      "grad_norm": 0.14279572665691376,
+      "learning_rate": 0.001,
+      "loss": 1.4731,
+      "step": 18503
+    },
+    {
+      "epoch": 0.7828073441069464,
+      "grad_norm": 7.251346111297607,
+      "learning_rate": 0.001,
+      "loss": 2.6161,
+      "step": 18504
+    },
+    {
+      "epoch": 0.7828496488704628,
+      "grad_norm": 0.580826997756958,
+      "learning_rate": 0.001,
+      "loss": 2.6049,
+      "step": 18505
+    },
+    {
+      "epoch": 0.7828919536339792,
+      "grad_norm": 2.745222330093384,
+      "learning_rate": 0.001,
+      "loss": 2.3253,
+      "step": 18506
+    },
+    {
+      "epoch": 0.7829342583974955,
+      "grad_norm": 0.14414088428020477,
+      "learning_rate": 0.001,
+      "loss": 1.7045,
+      "step": 18507
+    },
+    {
+      "epoch": 0.7829765631610119,
+      "grad_norm": 0.1275712102651596,
+      "learning_rate": 0.001,
+      "loss": 2.0115,
+      "step": 18508
+    },
+    {
+      "epoch": 0.7830188679245284,
+      "grad_norm": 0.1394435614347458,
+      "learning_rate": 0.001,
+      "loss": 1.4622,
+      "step": 18509
+    },
+    {
+      "epoch": 0.7830611726880446,
+      "grad_norm": 0.1480412781238556,
+      "learning_rate": 0.001,
+      "loss": 1.8277,
+      "step": 18510
+    },
+    {
+      "epoch": 0.7831034774515611,
+      "grad_norm": 0.14259982109069824,
+      "learning_rate": 0.001,
+      "loss": 2.2549,
+      "step": 18511
+    },
+    {
+      "epoch": 0.7831457822150775,
+      "grad_norm": 2.498133897781372,
+      "learning_rate": 0.001,
+      "loss": 3.1715,
+      "step": 18512
+    },
+    {
+      "epoch": 0.7831880869785938,
+      "grad_norm": 0.19657468795776367,
+      "learning_rate": 0.001,
+      "loss": 2.4656,
+      "step": 18513
+    },
+    {
+      "epoch": 0.7832303917421102,
+      "grad_norm": 0.1902858018875122,
+      "learning_rate": 0.001,
+      "loss": 2.3779,
+      "step": 18514
+    },
+    {
+      "epoch": 0.7832726965056266,
+      "grad_norm": 2.73234486579895,
+      "learning_rate": 0.001,
+      "loss": 2.9337,
+      "step": 18515
+    },
+    {
+      "epoch": 0.7833150012691429,
+      "grad_norm": 0.16122861206531525,
+      "learning_rate": 0.001,
+      "loss": 3.7783,
+      "step": 18516
+    },
+    {
+      "epoch": 0.7833573060326593,
+      "grad_norm": 0.146601602435112,
+      "learning_rate": 0.001,
+      "loss": 1.7877,
+      "step": 18517
+    },
+    {
+      "epoch": 0.7833996107961756,
+      "grad_norm": 0.1656588464975357,
+      "learning_rate": 0.001,
+      "loss": 2.4555,
+      "step": 18518
+    },
+    {
+      "epoch": 0.783441915559692,
+      "grad_norm": 0.1633993536233902,
+      "learning_rate": 0.001,
+      "loss": 2.131,
+      "step": 18519
+    },
+    {
+      "epoch": 0.7834842203232084,
+      "grad_norm": 0.13799798488616943,
+      "learning_rate": 0.001,
+      "loss": 1.6201,
+      "step": 18520
+    },
+    {
+      "epoch": 0.7835265250867247,
+      "grad_norm": 0.1774521917104721,
+      "learning_rate": 0.001,
+      "loss": 2.4657,
+      "step": 18521
+    },
+    {
+      "epoch": 0.7835688298502411,
+      "grad_norm": 0.17615440487861633,
+      "learning_rate": 0.001,
+      "loss": 2.1959,
+      "step": 18522
+    },
+    {
+      "epoch": 0.7836111346137575,
+      "grad_norm": 0.18356232345104218,
+      "learning_rate": 0.001,
+      "loss": 2.6112,
+      "step": 18523
+    },
+    {
+      "epoch": 0.7836534393772738,
+      "grad_norm": 0.1374877244234085,
+      "learning_rate": 0.001,
+      "loss": 1.8857,
+      "step": 18524
+    },
+    {
+      "epoch": 0.7836957441407902,
+      "grad_norm": 0.2870613634586334,
+      "learning_rate": 0.001,
+      "loss": 1.7536,
+      "step": 18525
+    },
+    {
+      "epoch": 0.7837380489043067,
+      "grad_norm": 0.16457685828208923,
+      "learning_rate": 0.001,
+      "loss": 1.7759,
+      "step": 18526
+    },
+    {
+      "epoch": 0.783780353667823,
+      "grad_norm": 0.12657257914543152,
+      "learning_rate": 0.001,
+      "loss": 2.1712,
+      "step": 18527
+    },
+    {
+      "epoch": 0.7838226584313394,
+      "grad_norm": 1.0247093439102173,
+      "learning_rate": 0.001,
+      "loss": 1.3212,
+      "step": 18528
+    },
+    {
+      "epoch": 0.7838649631948558,
+      "grad_norm": 2.865370273590088,
+      "learning_rate": 0.001,
+      "loss": 2.1017,
+      "step": 18529
+    },
+    {
+      "epoch": 0.7839072679583721,
+      "grad_norm": 0.2102050930261612,
+      "learning_rate": 0.001,
+      "loss": 1.917,
+      "step": 18530
+    },
+    {
+      "epoch": 0.7839495727218885,
+      "grad_norm": 0.23753419518470764,
+      "learning_rate": 0.001,
+      "loss": 2.7787,
+      "step": 18531
+    },
+    {
+      "epoch": 0.7839918774854049,
+      "grad_norm": 0.15850041806697845,
+      "learning_rate": 0.001,
+      "loss": 1.5583,
+      "step": 18532
+    },
+    {
+      "epoch": 0.7840341822489212,
+      "grad_norm": 0.1486329436302185,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 18533
+    },
+    {
+      "epoch": 0.7840764870124376,
+      "grad_norm": 0.13608387112617493,
+      "learning_rate": 0.001,
+      "loss": 2.6181,
+      "step": 18534
+    },
+    {
+      "epoch": 0.784118791775954,
+      "grad_norm": 0.1750655174255371,
+      "learning_rate": 0.001,
+      "loss": 3.3077,
+      "step": 18535
+    },
+    {
+      "epoch": 0.7841610965394703,
+      "grad_norm": 0.15577921271324158,
+      "learning_rate": 0.001,
+      "loss": 2.3589,
+      "step": 18536
+    },
+    {
+      "epoch": 0.7842034013029867,
+      "grad_norm": 0.16472339630126953,
+      "learning_rate": 0.001,
+      "loss": 2.7532,
+      "step": 18537
+    },
+    {
+      "epoch": 0.7842457060665031,
+      "grad_norm": 0.8173422813415527,
+      "learning_rate": 0.001,
+      "loss": 2.1228,
+      "step": 18538
+    },
+    {
+      "epoch": 0.7842880108300194,
+      "grad_norm": 0.16363029181957245,
+      "learning_rate": 0.001,
+      "loss": 3.242,
+      "step": 18539
+    },
+    {
+      "epoch": 0.7843303155935358,
+      "grad_norm": 1.0863935947418213,
+      "learning_rate": 0.001,
+      "loss": 2.4159,
+      "step": 18540
+    },
+    {
+      "epoch": 0.7843726203570522,
+      "grad_norm": 0.17449085414409637,
+      "learning_rate": 0.001,
+      "loss": 3.3047,
+      "step": 18541
+    },
+    {
+      "epoch": 0.7844149251205685,
+      "grad_norm": 0.2527814507484436,
+      "learning_rate": 0.001,
+      "loss": 2.5967,
+      "step": 18542
+    },
+    {
+      "epoch": 0.784457229884085,
+      "grad_norm": 0.13562943041324615,
+      "learning_rate": 0.001,
+      "loss": 2.5034,
+      "step": 18543
+    },
+    {
+      "epoch": 0.7844995346476014,
+      "grad_norm": 0.22819682955741882,
+      "learning_rate": 0.001,
+      "loss": 1.9655,
+      "step": 18544
+    },
+    {
+      "epoch": 0.7845418394111177,
+      "grad_norm": 0.7350271940231323,
+      "learning_rate": 0.001,
+      "loss": 2.8429,
+      "step": 18545
+    },
+    {
+      "epoch": 0.7845841441746341,
+      "grad_norm": 0.16228371858596802,
+      "learning_rate": 0.001,
+      "loss": 1.4537,
+      "step": 18546
+    },
+    {
+      "epoch": 0.7846264489381505,
+      "grad_norm": 0.5120057463645935,
+      "learning_rate": 0.001,
+      "loss": 1.5295,
+      "step": 18547
+    },
+    {
+      "epoch": 0.7846687537016668,
+      "grad_norm": 0.17189784348011017,
+      "learning_rate": 0.001,
+      "loss": 2.5171,
+      "step": 18548
+    },
+    {
+      "epoch": 0.7847110584651832,
+      "grad_norm": 0.2788395285606384,
+      "learning_rate": 0.001,
+      "loss": 1.7338,
+      "step": 18549
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 0.17548993229866028,
+      "learning_rate": 0.001,
+      "loss": 2.0646,
+      "step": 18550
+    },
+    {
+      "epoch": 0.7847956679922159,
+      "grad_norm": 0.15027286112308502,
+      "learning_rate": 0.001,
+      "loss": 1.6634,
+      "step": 18551
+    },
+    {
+      "epoch": 0.7848379727557323,
+      "grad_norm": 0.14293362200260162,
+      "learning_rate": 0.001,
+      "loss": 1.7618,
+      "step": 18552
+    },
+    {
+      "epoch": 0.7848802775192487,
+      "grad_norm": 0.17201675474643707,
+      "learning_rate": 0.001,
+      "loss": 2.2748,
+      "step": 18553
+    },
+    {
+      "epoch": 0.784922582282765,
+      "grad_norm": 0.22527389228343964,
+      "learning_rate": 0.001,
+      "loss": 1.3851,
+      "step": 18554
+    },
+    {
+      "epoch": 0.7849648870462814,
+      "grad_norm": 0.16594509780406952,
+      "learning_rate": 0.001,
+      "loss": 2.1222,
+      "step": 18555
+    },
+    {
+      "epoch": 0.7850071918097978,
+      "grad_norm": 0.1816147267818451,
+      "learning_rate": 0.001,
+      "loss": 2.4502,
+      "step": 18556
+    },
+    {
+      "epoch": 0.7850494965733141,
+      "grad_norm": 0.14666712284088135,
+      "learning_rate": 0.001,
+      "loss": 1.4992,
+      "step": 18557
+    },
+    {
+      "epoch": 0.7850918013368305,
+      "grad_norm": 0.14625667035579681,
+      "learning_rate": 0.001,
+      "loss": 2.2672,
+      "step": 18558
+    },
+    {
+      "epoch": 0.785134106100347,
+      "grad_norm": 0.5537320375442505,
+      "learning_rate": 0.001,
+      "loss": 1.6027,
+      "step": 18559
+    },
+    {
+      "epoch": 0.7851764108638633,
+      "grad_norm": 0.17199261486530304,
+      "learning_rate": 0.001,
+      "loss": 2.5807,
+      "step": 18560
+    },
+    {
+      "epoch": 0.7852187156273797,
+      "grad_norm": 0.14282803237438202,
+      "learning_rate": 0.001,
+      "loss": 2.0102,
+      "step": 18561
+    },
+    {
+      "epoch": 0.7852610203908961,
+      "grad_norm": 0.7231976985931396,
+      "learning_rate": 0.001,
+      "loss": 2.3132,
+      "step": 18562
+    },
+    {
+      "epoch": 0.7853033251544124,
+      "grad_norm": 0.15639185905456543,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 18563
+    },
+    {
+      "epoch": 0.7853456299179288,
+      "grad_norm": 0.1664103865623474,
+      "learning_rate": 0.001,
+      "loss": 2.3088,
+      "step": 18564
+    },
+    {
+      "epoch": 0.7853879346814451,
+      "grad_norm": 0.18371713161468506,
+      "learning_rate": 0.001,
+      "loss": 2.4793,
+      "step": 18565
+    },
+    {
+      "epoch": 0.7854302394449615,
+      "grad_norm": 1.8584163188934326,
+      "learning_rate": 0.001,
+      "loss": 1.5245,
+      "step": 18566
+    },
+    {
+      "epoch": 0.7854725442084779,
+      "grad_norm": 0.44644030928611755,
+      "learning_rate": 0.001,
+      "loss": 2.4425,
+      "step": 18567
+    },
+    {
+      "epoch": 0.7855148489719942,
+      "grad_norm": 0.15669460594654083,
+      "learning_rate": 0.001,
+      "loss": 1.7045,
+      "step": 18568
+    },
+    {
+      "epoch": 0.7855571537355106,
+      "grad_norm": 0.24448037147521973,
+      "learning_rate": 0.001,
+      "loss": 1.8752,
+      "step": 18569
+    },
+    {
+      "epoch": 0.785599458499027,
+      "grad_norm": 0.1537441611289978,
+      "learning_rate": 0.001,
+      "loss": 2.3786,
+      "step": 18570
+    },
+    {
+      "epoch": 0.7856417632625433,
+      "grad_norm": 0.14563751220703125,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 18571
+    },
+    {
+      "epoch": 0.7856840680260597,
+      "grad_norm": 0.17651155591011047,
+      "learning_rate": 0.001,
+      "loss": 2.319,
+      "step": 18572
+    },
+    {
+      "epoch": 0.7857263727895761,
+      "grad_norm": 0.17922982573509216,
+      "learning_rate": 0.001,
+      "loss": 2.112,
+      "step": 18573
+    },
+    {
+      "epoch": 0.7857686775530924,
+      "grad_norm": 0.1861950010061264,
+      "learning_rate": 0.001,
+      "loss": 2.7733,
+      "step": 18574
+    },
+    {
+      "epoch": 0.7858109823166088,
+      "grad_norm": 0.156846284866333,
+      "learning_rate": 0.001,
+      "loss": 2.2694,
+      "step": 18575
+    },
+    {
+      "epoch": 0.7858532870801253,
+      "grad_norm": 0.4748559892177582,
+      "learning_rate": 0.001,
+      "loss": 2.1577,
+      "step": 18576
+    },
+    {
+      "epoch": 0.7858955918436415,
+      "grad_norm": 0.14767950773239136,
+      "learning_rate": 0.001,
+      "loss": 2.1209,
+      "step": 18577
+    },
+    {
+      "epoch": 0.785937896607158,
+      "grad_norm": 0.5291089415550232,
+      "learning_rate": 0.001,
+      "loss": 2.5013,
+      "step": 18578
+    },
+    {
+      "epoch": 0.7859802013706744,
+      "grad_norm": 0.1678914725780487,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 18579
+    },
+    {
+      "epoch": 0.7860225061341907,
+      "grad_norm": 0.1400989443063736,
+      "learning_rate": 0.001,
+      "loss": 1.9519,
+      "step": 18580
+    },
+    {
+      "epoch": 0.7860648108977071,
+      "grad_norm": 0.20776335895061493,
+      "learning_rate": 0.001,
+      "loss": 2.9314,
+      "step": 18581
+    },
+    {
+      "epoch": 0.7861071156612235,
+      "grad_norm": 0.16925200819969177,
+      "learning_rate": 0.001,
+      "loss": 3.7282,
+      "step": 18582
+    },
+    {
+      "epoch": 0.7861494204247398,
+      "grad_norm": 0.14598102867603302,
+      "learning_rate": 0.001,
+      "loss": 2.3061,
+      "step": 18583
+    },
+    {
+      "epoch": 0.7861917251882562,
+      "grad_norm": 0.6664958000183105,
+      "learning_rate": 0.001,
+      "loss": 2.0033,
+      "step": 18584
+    },
+    {
+      "epoch": 0.7862340299517726,
+      "grad_norm": 0.14569751918315887,
+      "learning_rate": 0.001,
+      "loss": 2.4761,
+      "step": 18585
+    },
+    {
+      "epoch": 0.7862763347152889,
+      "grad_norm": 0.2870123088359833,
+      "learning_rate": 0.001,
+      "loss": 2.782,
+      "step": 18586
+    },
+    {
+      "epoch": 0.7863186394788053,
+      "grad_norm": 0.16477970778942108,
+      "learning_rate": 0.001,
+      "loss": 1.6753,
+      "step": 18587
+    },
+    {
+      "epoch": 0.7863609442423217,
+      "grad_norm": 0.18875035643577576,
+      "learning_rate": 0.001,
+      "loss": 1.7905,
+      "step": 18588
+    },
+    {
+      "epoch": 0.786403249005838,
+      "grad_norm": 0.1353473663330078,
+      "learning_rate": 0.001,
+      "loss": 3.2058,
+      "step": 18589
+    },
+    {
+      "epoch": 0.7864455537693544,
+      "grad_norm": 0.2390325963497162,
+      "learning_rate": 0.001,
+      "loss": 2.155,
+      "step": 18590
+    },
+    {
+      "epoch": 0.7864878585328708,
+      "grad_norm": 0.1434008628129959,
+      "learning_rate": 0.001,
+      "loss": 2.0458,
+      "step": 18591
+    },
+    {
+      "epoch": 0.7865301632963871,
+      "grad_norm": 0.1311628371477127,
+      "learning_rate": 0.001,
+      "loss": 1.5251,
+      "step": 18592
+    },
+    {
+      "epoch": 0.7865724680599036,
+      "grad_norm": 0.20478565990924835,
+      "learning_rate": 0.001,
+      "loss": 2.3965,
+      "step": 18593
+    },
+    {
+      "epoch": 0.78661477282342,
+      "grad_norm": 0.157118022441864,
+      "learning_rate": 0.001,
+      "loss": 2.0589,
+      "step": 18594
+    },
+    {
+      "epoch": 0.7866570775869363,
+      "grad_norm": 0.17026092112064362,
+      "learning_rate": 0.001,
+      "loss": 2.3213,
+      "step": 18595
+    },
+    {
+      "epoch": 0.7866993823504527,
+      "grad_norm": 0.12616808712482452,
+      "learning_rate": 0.001,
+      "loss": 1.5168,
+      "step": 18596
+    },
+    {
+      "epoch": 0.7867416871139691,
+      "grad_norm": 0.12917254865169525,
+      "learning_rate": 0.001,
+      "loss": 1.4579,
+      "step": 18597
+    },
+    {
+      "epoch": 0.7867839918774854,
+      "grad_norm": 0.18053403496742249,
+      "learning_rate": 0.001,
+      "loss": 2.1042,
+      "step": 18598
+    },
+    {
+      "epoch": 0.7868262966410018,
+      "grad_norm": 0.17954039573669434,
+      "learning_rate": 0.001,
+      "loss": 2.2459,
+      "step": 18599
+    },
+    {
+      "epoch": 0.7868686014045182,
+      "grad_norm": 0.23645149171352386,
+      "learning_rate": 0.001,
+      "loss": 2.0196,
+      "step": 18600
+    },
+    {
+      "epoch": 0.7869109061680345,
+      "grad_norm": 0.1373950093984604,
+      "learning_rate": 0.001,
+      "loss": 1.6001,
+      "step": 18601
+    },
+    {
+      "epoch": 0.7869532109315509,
+      "grad_norm": 0.17674638330936432,
+      "learning_rate": 0.001,
+      "loss": 3.6847,
+      "step": 18602
+    },
+    {
+      "epoch": 0.7869955156950673,
+      "grad_norm": 0.1795940101146698,
+      "learning_rate": 0.001,
+      "loss": 2.7799,
+      "step": 18603
+    },
+    {
+      "epoch": 0.7870378204585836,
+      "grad_norm": 0.16243141889572144,
+      "learning_rate": 0.001,
+      "loss": 1.7403,
+      "step": 18604
+    },
+    {
+      "epoch": 0.7870801252221,
+      "grad_norm": 13.920958518981934,
+      "learning_rate": 0.001,
+      "loss": 1.9834,
+      "step": 18605
+    },
+    {
+      "epoch": 0.7871224299856164,
+      "grad_norm": 0.16762857139110565,
+      "learning_rate": 0.001,
+      "loss": 1.9104,
+      "step": 18606
+    },
+    {
+      "epoch": 0.7871647347491327,
+      "grad_norm": 0.15232449769973755,
+      "learning_rate": 0.001,
+      "loss": 1.9612,
+      "step": 18607
+    },
+    {
+      "epoch": 0.7872070395126491,
+      "grad_norm": 0.1300850510597229,
+      "learning_rate": 0.001,
+      "loss": 1.663,
+      "step": 18608
+    },
+    {
+      "epoch": 0.7872493442761654,
+      "grad_norm": 0.12953154742717743,
+      "learning_rate": 0.001,
+      "loss": 2.6655,
+      "step": 18609
+    },
+    {
+      "epoch": 0.7872916490396819,
+      "grad_norm": 0.1451409012079239,
+      "learning_rate": 0.001,
+      "loss": 2.2023,
+      "step": 18610
+    },
+    {
+      "epoch": 0.7873339538031983,
+      "grad_norm": 0.370164692401886,
+      "learning_rate": 0.001,
+      "loss": 2.1399,
+      "step": 18611
+    },
+    {
+      "epoch": 0.7873762585667146,
+      "grad_norm": 0.3032169044017792,
+      "learning_rate": 0.001,
+      "loss": 1.439,
+      "step": 18612
+    },
+    {
+      "epoch": 0.787418563330231,
+      "grad_norm": 0.15478359162807465,
+      "learning_rate": 0.001,
+      "loss": 1.7436,
+      "step": 18613
+    },
+    {
+      "epoch": 0.7874608680937474,
+      "grad_norm": 0.1523745208978653,
+      "learning_rate": 0.001,
+      "loss": 1.7031,
+      "step": 18614
+    },
+    {
+      "epoch": 0.7875031728572637,
+      "grad_norm": 0.1674836277961731,
+      "learning_rate": 0.001,
+      "loss": 1.8924,
+      "step": 18615
+    },
+    {
+      "epoch": 0.7875454776207801,
+      "grad_norm": 0.16084153950214386,
+      "learning_rate": 0.001,
+      "loss": 3.0371,
+      "step": 18616
+    },
+    {
+      "epoch": 0.7875877823842965,
+      "grad_norm": 0.15043555200099945,
+      "learning_rate": 0.001,
+      "loss": 1.669,
+      "step": 18617
+    },
+    {
+      "epoch": 0.7876300871478128,
+      "grad_norm": 0.15042801201343536,
+      "learning_rate": 0.001,
+      "loss": 1.4761,
+      "step": 18618
+    },
+    {
+      "epoch": 0.7876723919113292,
+      "grad_norm": 0.20407025516033173,
+      "learning_rate": 0.001,
+      "loss": 2.5795,
+      "step": 18619
+    },
+    {
+      "epoch": 0.7877146966748456,
+      "grad_norm": 0.37123432755470276,
+      "learning_rate": 0.001,
+      "loss": 1.9516,
+      "step": 18620
+    },
+    {
+      "epoch": 0.7877570014383619,
+      "grad_norm": 0.14106714725494385,
+      "learning_rate": 0.001,
+      "loss": 3.1985,
+      "step": 18621
+    },
+    {
+      "epoch": 0.7877993062018783,
+      "grad_norm": 0.16978901624679565,
+      "learning_rate": 0.001,
+      "loss": 2.6184,
+      "step": 18622
+    },
+    {
+      "epoch": 0.7878416109653947,
+      "grad_norm": 0.32933205366134644,
+      "learning_rate": 0.001,
+      "loss": 3.5388,
+      "step": 18623
+    },
+    {
+      "epoch": 0.787883915728911,
+      "grad_norm": 0.16782082617282867,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 18624
+    },
+    {
+      "epoch": 0.7879262204924274,
+      "grad_norm": 0.16868159174919128,
+      "learning_rate": 0.001,
+      "loss": 2.1474,
+      "step": 18625
+    },
+    {
+      "epoch": 0.7879685252559439,
+      "grad_norm": 0.20239630341529846,
+      "learning_rate": 0.001,
+      "loss": 2.5787,
+      "step": 18626
+    },
+    {
+      "epoch": 0.7880108300194602,
+      "grad_norm": 0.49413931369781494,
+      "learning_rate": 0.001,
+      "loss": 2.7789,
+      "step": 18627
+    },
+    {
+      "epoch": 0.7880531347829766,
+      "grad_norm": 0.19046078622341156,
+      "learning_rate": 0.001,
+      "loss": 2.8731,
+      "step": 18628
+    },
+    {
+      "epoch": 0.788095439546493,
+      "grad_norm": 0.6242032647132874,
+      "learning_rate": 0.001,
+      "loss": 3.4162,
+      "step": 18629
+    },
+    {
+      "epoch": 0.7881377443100093,
+      "grad_norm": 0.29068413376808167,
+      "learning_rate": 0.001,
+      "loss": 2.5984,
+      "step": 18630
+    },
+    {
+      "epoch": 0.7881800490735257,
+      "grad_norm": 0.159669429063797,
+      "learning_rate": 0.001,
+      "loss": 2.885,
+      "step": 18631
+    },
+    {
+      "epoch": 0.7882223538370421,
+      "grad_norm": 0.14053291082382202,
+      "learning_rate": 0.001,
+      "loss": 1.7608,
+      "step": 18632
+    },
+    {
+      "epoch": 0.7882646586005584,
+      "grad_norm": 0.34683123230934143,
+      "learning_rate": 0.001,
+      "loss": 1.8457,
+      "step": 18633
+    },
+    {
+      "epoch": 0.7883069633640748,
+      "grad_norm": 0.1669548749923706,
+      "learning_rate": 0.001,
+      "loss": 1.6075,
+      "step": 18634
+    },
+    {
+      "epoch": 0.7883492681275912,
+      "grad_norm": 0.19117628037929535,
+      "learning_rate": 0.001,
+      "loss": 2.7102,
+      "step": 18635
+    },
+    {
+      "epoch": 0.7883915728911075,
+      "grad_norm": 0.13999809324741364,
+      "learning_rate": 0.001,
+      "loss": 1.5316,
+      "step": 18636
+    },
+    {
+      "epoch": 0.7884338776546239,
+      "grad_norm": 0.3912406265735626,
+      "learning_rate": 0.001,
+      "loss": 2.5317,
+      "step": 18637
+    },
+    {
+      "epoch": 0.7884761824181403,
+      "grad_norm": 0.20486436784267426,
+      "learning_rate": 0.001,
+      "loss": 3.0638,
+      "step": 18638
+    },
+    {
+      "epoch": 0.7885184871816566,
+      "grad_norm": 0.18025173246860504,
+      "learning_rate": 0.001,
+      "loss": 1.9768,
+      "step": 18639
+    },
+    {
+      "epoch": 0.788560791945173,
+      "grad_norm": 0.1740838885307312,
+      "learning_rate": 0.001,
+      "loss": 3.0294,
+      "step": 18640
+    },
+    {
+      "epoch": 0.7886030967086894,
+      "grad_norm": 0.18480761349201202,
+      "learning_rate": 0.001,
+      "loss": 2.3447,
+      "step": 18641
+    },
+    {
+      "epoch": 0.7886454014722057,
+      "grad_norm": 14.460678100585938,
+      "learning_rate": 0.001,
+      "loss": 2.7443,
+      "step": 18642
+    },
+    {
+      "epoch": 0.7886877062357222,
+      "grad_norm": 0.19740672409534454,
+      "learning_rate": 0.001,
+      "loss": 1.882,
+      "step": 18643
+    },
+    {
+      "epoch": 0.7887300109992386,
+      "grad_norm": 0.13065767288208008,
+      "learning_rate": 0.001,
+      "loss": 2.2647,
+      "step": 18644
+    },
+    {
+      "epoch": 0.7887723157627549,
+      "grad_norm": 0.49033114314079285,
+      "learning_rate": 0.001,
+      "loss": 2.3273,
+      "step": 18645
+    },
+    {
+      "epoch": 0.7888146205262713,
+      "grad_norm": 0.16719037294387817,
+      "learning_rate": 0.001,
+      "loss": 2.3332,
+      "step": 18646
+    },
+    {
+      "epoch": 0.7888569252897877,
+      "grad_norm": 0.16855984926223755,
+      "learning_rate": 0.001,
+      "loss": 3.0683,
+      "step": 18647
+    },
+    {
+      "epoch": 0.788899230053304,
+      "grad_norm": 0.1336914449930191,
+      "learning_rate": 0.001,
+      "loss": 1.381,
+      "step": 18648
+    },
+    {
+      "epoch": 0.7889415348168204,
+      "grad_norm": 0.16387201845645905,
+      "learning_rate": 0.001,
+      "loss": 2.3596,
+      "step": 18649
+    },
+    {
+      "epoch": 0.7889838395803368,
+      "grad_norm": 2.3988749980926514,
+      "learning_rate": 0.001,
+      "loss": 2.2531,
+      "step": 18650
+    },
+    {
+      "epoch": 0.7890261443438531,
+      "grad_norm": 0.18819689750671387,
+      "learning_rate": 0.001,
+      "loss": 2.171,
+      "step": 18651
+    },
+    {
+      "epoch": 0.7890684491073695,
+      "grad_norm": 0.16322028636932373,
+      "learning_rate": 0.001,
+      "loss": 2.2966,
+      "step": 18652
+    },
+    {
+      "epoch": 0.7891107538708859,
+      "grad_norm": 0.15296301245689392,
+      "learning_rate": 0.001,
+      "loss": 1.9929,
+      "step": 18653
+    },
+    {
+      "epoch": 0.7891530586344022,
+      "grad_norm": 0.14490869641304016,
+      "learning_rate": 0.001,
+      "loss": 1.8859,
+      "step": 18654
+    },
+    {
+      "epoch": 0.7891953633979186,
+      "grad_norm": 0.16145136952400208,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 18655
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.15146324038505554,
+      "learning_rate": 0.001,
+      "loss": 1.7612,
+      "step": 18656
+    },
+    {
+      "epoch": 0.7892799729249513,
+      "grad_norm": 0.15252503752708435,
+      "learning_rate": 0.001,
+      "loss": 3.3884,
+      "step": 18657
+    },
+    {
+      "epoch": 0.7893222776884677,
+      "grad_norm": 1.6242767572402954,
+      "learning_rate": 0.001,
+      "loss": 1.7785,
+      "step": 18658
+    },
+    {
+      "epoch": 0.789364582451984,
+      "grad_norm": 0.19868817925453186,
+      "learning_rate": 0.001,
+      "loss": 3.0005,
+      "step": 18659
+    },
+    {
+      "epoch": 0.7894068872155005,
+      "grad_norm": 0.24456727504730225,
+      "learning_rate": 0.001,
+      "loss": 1.73,
+      "step": 18660
+    },
+    {
+      "epoch": 0.7894491919790169,
+      "grad_norm": 0.23000819981098175,
+      "learning_rate": 0.001,
+      "loss": 2.2173,
+      "step": 18661
+    },
+    {
+      "epoch": 0.7894914967425332,
+      "grad_norm": 0.220015287399292,
+      "learning_rate": 0.001,
+      "loss": 2.1518,
+      "step": 18662
+    },
+    {
+      "epoch": 0.7895338015060496,
+      "grad_norm": 0.20717963576316833,
+      "learning_rate": 0.001,
+      "loss": 3.1284,
+      "step": 18663
+    },
+    {
+      "epoch": 0.789576106269566,
+      "grad_norm": 0.2126576453447342,
+      "learning_rate": 0.001,
+      "loss": 2.8512,
+      "step": 18664
+    },
+    {
+      "epoch": 0.7896184110330823,
+      "grad_norm": 0.2596065402030945,
+      "learning_rate": 0.001,
+      "loss": 2.0507,
+      "step": 18665
+    },
+    {
+      "epoch": 0.7896607157965987,
+      "grad_norm": 0.22077849507331848,
+      "learning_rate": 0.001,
+      "loss": 1.6807,
+      "step": 18666
+    },
+    {
+      "epoch": 0.7897030205601151,
+      "grad_norm": 0.1828576773405075,
+      "learning_rate": 0.001,
+      "loss": 2.0024,
+      "step": 18667
+    },
+    {
+      "epoch": 0.7897453253236314,
+      "grad_norm": 0.5003877282142639,
+      "learning_rate": 0.001,
+      "loss": 3.519,
+      "step": 18668
+    },
+    {
+      "epoch": 0.7897876300871478,
+      "grad_norm": 0.17294269800186157,
+      "learning_rate": 0.001,
+      "loss": 2.7761,
+      "step": 18669
+    },
+    {
+      "epoch": 0.7898299348506642,
+      "grad_norm": 0.19300001859664917,
+      "learning_rate": 0.001,
+      "loss": 1.9447,
+      "step": 18670
+    },
+    {
+      "epoch": 0.7898722396141805,
+      "grad_norm": 1.2025370597839355,
+      "learning_rate": 0.001,
+      "loss": 3.5368,
+      "step": 18671
+    },
+    {
+      "epoch": 0.7899145443776969,
+      "grad_norm": 0.14662951231002808,
+      "learning_rate": 0.001,
+      "loss": 2.5346,
+      "step": 18672
+    },
+    {
+      "epoch": 0.7899568491412133,
+      "grad_norm": 2.872141122817993,
+      "learning_rate": 0.001,
+      "loss": 1.9286,
+      "step": 18673
+    },
+    {
+      "epoch": 0.7899991539047296,
+      "grad_norm": 0.16066762804985046,
+      "learning_rate": 0.001,
+      "loss": 2.2664,
+      "step": 18674
+    },
+    {
+      "epoch": 0.790041458668246,
+      "grad_norm": 0.16592992842197418,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 18675
+    },
+    {
+      "epoch": 0.7900837634317625,
+      "grad_norm": 0.16850398480892181,
+      "learning_rate": 0.001,
+      "loss": 1.8177,
+      "step": 18676
+    },
+    {
+      "epoch": 0.7901260681952788,
+      "grad_norm": 0.16996052861213684,
+      "learning_rate": 0.001,
+      "loss": 1.9456,
+      "step": 18677
+    },
+    {
+      "epoch": 0.7901683729587952,
+      "grad_norm": 0.39866968989372253,
+      "learning_rate": 0.001,
+      "loss": 2.8809,
+      "step": 18678
+    },
+    {
+      "epoch": 0.7902106777223116,
+      "grad_norm": 0.16642092168331146,
+      "learning_rate": 0.001,
+      "loss": 2.6154,
+      "step": 18679
+    },
+    {
+      "epoch": 0.7902529824858279,
+      "grad_norm": 0.17593850195407867,
+      "learning_rate": 0.001,
+      "loss": 1.7041,
+      "step": 18680
+    },
+    {
+      "epoch": 0.7902952872493443,
+      "grad_norm": 0.17078939080238342,
+      "learning_rate": 0.001,
+      "loss": 2.9029,
+      "step": 18681
+    },
+    {
+      "epoch": 0.7903375920128607,
+      "grad_norm": 2.226186990737915,
+      "learning_rate": 0.001,
+      "loss": 1.6863,
+      "step": 18682
+    },
+    {
+      "epoch": 0.790379896776377,
+      "grad_norm": 1.4724563360214233,
+      "learning_rate": 0.001,
+      "loss": 2.1862,
+      "step": 18683
+    },
+    {
+      "epoch": 0.7904222015398934,
+      "grad_norm": 0.1770937144756317,
+      "learning_rate": 0.001,
+      "loss": 2.435,
+      "step": 18684
+    },
+    {
+      "epoch": 0.7904645063034098,
+      "grad_norm": 0.16543549299240112,
+      "learning_rate": 0.001,
+      "loss": 1.5855,
+      "step": 18685
+    },
+    {
+      "epoch": 0.7905068110669261,
+      "grad_norm": 0.19836363196372986,
+      "learning_rate": 0.001,
+      "loss": 1.9606,
+      "step": 18686
+    },
+    {
+      "epoch": 0.7905491158304425,
+      "grad_norm": 0.1877475082874298,
+      "learning_rate": 0.001,
+      "loss": 1.8057,
+      "step": 18687
+    },
+    {
+      "epoch": 0.7905914205939589,
+      "grad_norm": 0.14974345266819,
+      "learning_rate": 0.001,
+      "loss": 2.067,
+      "step": 18688
+    },
+    {
+      "epoch": 0.7906337253574752,
+      "grad_norm": 0.21924450993537903,
+      "learning_rate": 0.001,
+      "loss": 3.1355,
+      "step": 18689
+    },
+    {
+      "epoch": 0.7906760301209916,
+      "grad_norm": 0.17212629318237305,
+      "learning_rate": 0.001,
+      "loss": 1.5735,
+      "step": 18690
+    },
+    {
+      "epoch": 0.790718334884508,
+      "grad_norm": 0.4152930676937103,
+      "learning_rate": 0.001,
+      "loss": 2.7388,
+      "step": 18691
+    },
+    {
+      "epoch": 0.7907606396480243,
+      "grad_norm": 0.1306760460138321,
+      "learning_rate": 0.001,
+      "loss": 1.4797,
+      "step": 18692
+    },
+    {
+      "epoch": 0.7908029444115408,
+      "grad_norm": 0.166093111038208,
+      "learning_rate": 0.001,
+      "loss": 2.0058,
+      "step": 18693
+    },
+    {
+      "epoch": 0.7908452491750572,
+      "grad_norm": 1.3968002796173096,
+      "learning_rate": 0.001,
+      "loss": 1.5695,
+      "step": 18694
+    },
+    {
+      "epoch": 0.7908875539385735,
+      "grad_norm": 0.18365883827209473,
+      "learning_rate": 0.001,
+      "loss": 2.3175,
+      "step": 18695
+    },
+    {
+      "epoch": 0.7909298587020899,
+      "grad_norm": 0.14197491109371185,
+      "learning_rate": 0.001,
+      "loss": 1.7524,
+      "step": 18696
+    },
+    {
+      "epoch": 0.7909721634656063,
+      "grad_norm": 0.19356654584407806,
+      "learning_rate": 0.001,
+      "loss": 3.7002,
+      "step": 18697
+    },
+    {
+      "epoch": 0.7910144682291226,
+      "grad_norm": 0.1772981733083725,
+      "learning_rate": 0.001,
+      "loss": 2.2998,
+      "step": 18698
+    },
+    {
+      "epoch": 0.791056772992639,
+      "grad_norm": 0.2028152197599411,
+      "learning_rate": 0.001,
+      "loss": 1.9209,
+      "step": 18699
+    },
+    {
+      "epoch": 0.7910990777561553,
+      "grad_norm": 0.4528128504753113,
+      "learning_rate": 0.001,
+      "loss": 1.9742,
+      "step": 18700
+    },
+    {
+      "epoch": 0.7911413825196717,
+      "grad_norm": 0.19012227654457092,
+      "learning_rate": 0.001,
+      "loss": 2.475,
+      "step": 18701
+    },
+    {
+      "epoch": 0.7911836872831881,
+      "grad_norm": 0.17222073674201965,
+      "learning_rate": 0.001,
+      "loss": 1.9615,
+      "step": 18702
+    },
+    {
+      "epoch": 0.7912259920467044,
+      "grad_norm": 0.19183796644210815,
+      "learning_rate": 0.001,
+      "loss": 2.1465,
+      "step": 18703
+    },
+    {
+      "epoch": 0.7912682968102208,
+      "grad_norm": 0.1738269031047821,
+      "learning_rate": 0.001,
+      "loss": 2.6244,
+      "step": 18704
+    },
+    {
+      "epoch": 0.7913106015737372,
+      "grad_norm": 0.18650266528129578,
+      "learning_rate": 0.001,
+      "loss": 2.0074,
+      "step": 18705
+    },
+    {
+      "epoch": 0.7913529063372535,
+      "grad_norm": 0.1911695897579193,
+      "learning_rate": 0.001,
+      "loss": 2.1624,
+      "step": 18706
+    },
+    {
+      "epoch": 0.7913952111007699,
+      "grad_norm": 0.2884248197078705,
+      "learning_rate": 0.001,
+      "loss": 2.7871,
+      "step": 18707
+    },
+    {
+      "epoch": 0.7914375158642863,
+      "grad_norm": 0.15906710922718048,
+      "learning_rate": 0.001,
+      "loss": 2.9539,
+      "step": 18708
+    },
+    {
+      "epoch": 0.7914798206278026,
+      "grad_norm": 0.19278644025325775,
+      "learning_rate": 0.001,
+      "loss": 1.6402,
+      "step": 18709
+    },
+    {
+      "epoch": 0.791522125391319,
+      "grad_norm": 0.15218260884284973,
+      "learning_rate": 0.001,
+      "loss": 1.8649,
+      "step": 18710
+    },
+    {
+      "epoch": 0.7915644301548355,
+      "grad_norm": 0.1600116640329361,
+      "learning_rate": 0.001,
+      "loss": 1.6598,
+      "step": 18711
+    },
+    {
+      "epoch": 0.7916067349183518,
+      "grad_norm": 0.15089303255081177,
+      "learning_rate": 0.001,
+      "loss": 1.9441,
+      "step": 18712
+    },
+    {
+      "epoch": 0.7916490396818682,
+      "grad_norm": 0.15738871693611145,
+      "learning_rate": 0.001,
+      "loss": 1.888,
+      "step": 18713
+    },
+    {
+      "epoch": 0.7916913444453846,
+      "grad_norm": 0.15308193862438202,
+      "learning_rate": 0.001,
+      "loss": 1.8442,
+      "step": 18714
+    },
+    {
+      "epoch": 0.7917336492089009,
+      "grad_norm": 0.1537911295890808,
+      "learning_rate": 0.001,
+      "loss": 1.8999,
+      "step": 18715
+    },
+    {
+      "epoch": 0.7917759539724173,
+      "grad_norm": 0.16866432130336761,
+      "learning_rate": 0.001,
+      "loss": 2.5294,
+      "step": 18716
+    },
+    {
+      "epoch": 0.7918182587359337,
+      "grad_norm": 1.7140220403671265,
+      "learning_rate": 0.001,
+      "loss": 1.9505,
+      "step": 18717
+    },
+    {
+      "epoch": 0.79186056349945,
+      "grad_norm": 0.15509742498397827,
+      "learning_rate": 0.001,
+      "loss": 1.7273,
+      "step": 18718
+    },
+    {
+      "epoch": 0.7919028682629664,
+      "grad_norm": 0.16215214133262634,
+      "learning_rate": 0.001,
+      "loss": 3.0471,
+      "step": 18719
+    },
+    {
+      "epoch": 0.7919451730264828,
+      "grad_norm": 0.1872335970401764,
+      "learning_rate": 0.001,
+      "loss": 3.027,
+      "step": 18720
+    },
+    {
+      "epoch": 0.7919874777899991,
+      "grad_norm": 0.14136208593845367,
+      "learning_rate": 0.001,
+      "loss": 2.0332,
+      "step": 18721
+    },
+    {
+      "epoch": 0.7920297825535155,
+      "grad_norm": 0.2219260334968567,
+      "learning_rate": 0.001,
+      "loss": 2.2754,
+      "step": 18722
+    },
+    {
+      "epoch": 0.7920720873170319,
+      "grad_norm": 0.21660354733467102,
+      "learning_rate": 0.001,
+      "loss": 1.7934,
+      "step": 18723
+    },
+    {
+      "epoch": 0.7921143920805482,
+      "grad_norm": 0.16659744083881378,
+      "learning_rate": 0.001,
+      "loss": 2.2417,
+      "step": 18724
+    },
+    {
+      "epoch": 0.7921566968440646,
+      "grad_norm": 0.1875210851430893,
+      "learning_rate": 0.001,
+      "loss": 2.3443,
+      "step": 18725
+    },
+    {
+      "epoch": 0.792199001607581,
+      "grad_norm": 0.1662781685590744,
+      "learning_rate": 0.001,
+      "loss": 1.914,
+      "step": 18726
+    },
+    {
+      "epoch": 0.7922413063710974,
+      "grad_norm": 0.24961335957050323,
+      "learning_rate": 0.001,
+      "loss": 1.9302,
+      "step": 18727
+    },
+    {
+      "epoch": 0.7922836111346138,
+      "grad_norm": 0.17793142795562744,
+      "learning_rate": 0.001,
+      "loss": 2.1055,
+      "step": 18728
+    },
+    {
+      "epoch": 0.7923259158981302,
+      "grad_norm": 0.7194470763206482,
+      "learning_rate": 0.001,
+      "loss": 2.9523,
+      "step": 18729
+    },
+    {
+      "epoch": 0.7923682206616465,
+      "grad_norm": 0.19897310435771942,
+      "learning_rate": 0.001,
+      "loss": 1.8951,
+      "step": 18730
+    },
+    {
+      "epoch": 0.7924105254251629,
+      "grad_norm": 0.16758960485458374,
+      "learning_rate": 0.001,
+      "loss": 3.3784,
+      "step": 18731
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.18163642287254333,
+      "learning_rate": 0.001,
+      "loss": 2.5377,
+      "step": 18732
+    },
+    {
+      "epoch": 0.7924951349521956,
+      "grad_norm": 0.1600625365972519,
+      "learning_rate": 0.001,
+      "loss": 2.8553,
+      "step": 18733
+    },
+    {
+      "epoch": 0.792537439715712,
+      "grad_norm": 0.17819511890411377,
+      "learning_rate": 0.001,
+      "loss": 3.1244,
+      "step": 18734
+    },
+    {
+      "epoch": 0.7925797444792284,
+      "grad_norm": 0.16769671440124512,
+      "learning_rate": 0.001,
+      "loss": 1.4566,
+      "step": 18735
+    },
+    {
+      "epoch": 0.7926220492427447,
+      "grad_norm": 0.18893221020698547,
+      "learning_rate": 0.001,
+      "loss": 2.7459,
+      "step": 18736
+    },
+    {
+      "epoch": 0.7926643540062611,
+      "grad_norm": 0.16248752176761627,
+      "learning_rate": 0.001,
+      "loss": 1.829,
+      "step": 18737
+    },
+    {
+      "epoch": 0.7927066587697775,
+      "grad_norm": 0.1498391479253769,
+      "learning_rate": 0.001,
+      "loss": 2.218,
+      "step": 18738
+    },
+    {
+      "epoch": 0.7927489635332938,
+      "grad_norm": 0.2448561042547226,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 18739
+    },
+    {
+      "epoch": 0.7927912682968102,
+      "grad_norm": 0.17693863809108734,
+      "learning_rate": 0.001,
+      "loss": 1.7627,
+      "step": 18740
+    },
+    {
+      "epoch": 0.7928335730603266,
+      "grad_norm": 0.13363198935985565,
+      "learning_rate": 0.001,
+      "loss": 1.4754,
+      "step": 18741
+    },
+    {
+      "epoch": 0.792875877823843,
+      "grad_norm": 0.1259775161743164,
+      "learning_rate": 0.001,
+      "loss": 1.5831,
+      "step": 18742
+    },
+    {
+      "epoch": 0.7929181825873594,
+      "grad_norm": 0.18019893765449524,
+      "learning_rate": 0.001,
+      "loss": 2.3857,
+      "step": 18743
+    },
+    {
+      "epoch": 0.7929604873508757,
+      "grad_norm": 0.2442205846309662,
+      "learning_rate": 0.001,
+      "loss": 2.0327,
+      "step": 18744
+    },
+    {
+      "epoch": 0.7930027921143921,
+      "grad_norm": 1.5287914276123047,
+      "learning_rate": 0.001,
+      "loss": 2.4658,
+      "step": 18745
+    },
+    {
+      "epoch": 0.7930450968779085,
+      "grad_norm": 0.15980815887451172,
+      "learning_rate": 0.001,
+      "loss": 1.9778,
+      "step": 18746
+    },
+    {
+      "epoch": 0.7930874016414248,
+      "grad_norm": 0.1930980682373047,
+      "learning_rate": 0.001,
+      "loss": 2.2091,
+      "step": 18747
+    },
+    {
+      "epoch": 0.7931297064049412,
+      "grad_norm": 0.17652654647827148,
+      "learning_rate": 0.001,
+      "loss": 2.2903,
+      "step": 18748
+    },
+    {
+      "epoch": 0.7931720111684576,
+      "grad_norm": 0.22159257531166077,
+      "learning_rate": 0.001,
+      "loss": 2.6479,
+      "step": 18749
+    },
+    {
+      "epoch": 0.7932143159319739,
+      "grad_norm": 4.054377555847168,
+      "learning_rate": 0.001,
+      "loss": 1.5173,
+      "step": 18750
+    },
+    {
+      "epoch": 0.7932566206954903,
+      "grad_norm": 0.17010121047496796,
+      "learning_rate": 0.001,
+      "loss": 2.6045,
+      "step": 18751
+    },
+    {
+      "epoch": 0.7932989254590067,
+      "grad_norm": 0.15084238350391388,
+      "learning_rate": 0.001,
+      "loss": 2.4313,
+      "step": 18752
+    },
+    {
+      "epoch": 0.793341230222523,
+      "grad_norm": 0.18187488615512848,
+      "learning_rate": 0.001,
+      "loss": 2.0466,
+      "step": 18753
+    },
+    {
+      "epoch": 0.7933835349860394,
+      "grad_norm": 2.5695083141326904,
+      "learning_rate": 0.001,
+      "loss": 1.9052,
+      "step": 18754
+    },
+    {
+      "epoch": 0.7934258397495558,
+      "grad_norm": 0.1252131462097168,
+      "learning_rate": 0.001,
+      "loss": 2.476,
+      "step": 18755
+    },
+    {
+      "epoch": 0.7934681445130721,
+      "grad_norm": 0.194590225815773,
+      "learning_rate": 0.001,
+      "loss": 1.5803,
+      "step": 18756
+    },
+    {
+      "epoch": 0.7935104492765885,
+      "grad_norm": 0.2729746103286743,
+      "learning_rate": 0.001,
+      "loss": 2.2466,
+      "step": 18757
+    },
+    {
+      "epoch": 0.793552754040105,
+      "grad_norm": 0.2805977165699005,
+      "learning_rate": 0.001,
+      "loss": 2.5365,
+      "step": 18758
+    },
+    {
+      "epoch": 0.7935950588036212,
+      "grad_norm": 0.18586377799510956,
+      "learning_rate": 0.001,
+      "loss": 1.7274,
+      "step": 18759
+    },
+    {
+      "epoch": 0.7936373635671377,
+      "grad_norm": 0.16096092760562897,
+      "learning_rate": 0.001,
+      "loss": 3.2489,
+      "step": 18760
+    },
+    {
+      "epoch": 0.7936796683306541,
+      "grad_norm": 0.14794796705245972,
+      "learning_rate": 0.001,
+      "loss": 1.7881,
+      "step": 18761
+    },
+    {
+      "epoch": 0.7937219730941704,
+      "grad_norm": 0.17098478972911835,
+      "learning_rate": 0.001,
+      "loss": 3.5723,
+      "step": 18762
+    },
+    {
+      "epoch": 0.7937642778576868,
+      "grad_norm": 0.1851399689912796,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 18763
+    },
+    {
+      "epoch": 0.7938065826212032,
+      "grad_norm": 0.16206832230091095,
+      "learning_rate": 0.001,
+      "loss": 2.5929,
+      "step": 18764
+    },
+    {
+      "epoch": 0.7938488873847195,
+      "grad_norm": 0.16010499000549316,
+      "learning_rate": 0.001,
+      "loss": 1.6522,
+      "step": 18765
+    },
+    {
+      "epoch": 0.7938911921482359,
+      "grad_norm": 0.1725982278585434,
+      "learning_rate": 0.001,
+      "loss": 2.5854,
+      "step": 18766
+    },
+    {
+      "epoch": 0.7939334969117523,
+      "grad_norm": 0.13452595472335815,
+      "learning_rate": 0.001,
+      "loss": 1.9167,
+      "step": 18767
+    },
+    {
+      "epoch": 0.7939758016752686,
+      "grad_norm": 0.12271041423082352,
+      "learning_rate": 0.001,
+      "loss": 2.1516,
+      "step": 18768
+    },
+    {
+      "epoch": 0.794018106438785,
+      "grad_norm": 0.15723437070846558,
+      "learning_rate": 0.001,
+      "loss": 2.172,
+      "step": 18769
+    },
+    {
+      "epoch": 0.7940604112023014,
+      "grad_norm": 0.1402629315853119,
+      "learning_rate": 0.001,
+      "loss": 1.7417,
+      "step": 18770
+    },
+    {
+      "epoch": 0.7941027159658177,
+      "grad_norm": 0.11944486945867538,
+      "learning_rate": 0.001,
+      "loss": 1.5726,
+      "step": 18771
+    },
+    {
+      "epoch": 0.7941450207293341,
+      "grad_norm": 0.12228669226169586,
+      "learning_rate": 0.001,
+      "loss": 2.0272,
+      "step": 18772
+    },
+    {
+      "epoch": 0.7941873254928505,
+      "grad_norm": 0.18956278264522552,
+      "learning_rate": 0.001,
+      "loss": 3.0162,
+      "step": 18773
+    },
+    {
+      "epoch": 0.7942296302563668,
+      "grad_norm": 0.6293951272964478,
+      "learning_rate": 0.001,
+      "loss": 2.1262,
+      "step": 18774
+    },
+    {
+      "epoch": 0.7942719350198832,
+      "grad_norm": 0.14916300773620605,
+      "learning_rate": 0.001,
+      "loss": 1.9637,
+      "step": 18775
+    },
+    {
+      "epoch": 0.7943142397833997,
+      "grad_norm": 0.14974258840084076,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 18776
+    },
+    {
+      "epoch": 0.794356544546916,
+      "grad_norm": 0.151955246925354,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 18777
+    },
+    {
+      "epoch": 0.7943988493104324,
+      "grad_norm": 0.13923077285289764,
+      "learning_rate": 0.001,
+      "loss": 2.3116,
+      "step": 18778
+    },
+    {
+      "epoch": 0.7944411540739488,
+      "grad_norm": 0.14543022215366364,
+      "learning_rate": 0.001,
+      "loss": 2.7402,
+      "step": 18779
+    },
+    {
+      "epoch": 0.7944834588374651,
+      "grad_norm": 0.15743476152420044,
+      "learning_rate": 0.001,
+      "loss": 2.3271,
+      "step": 18780
+    },
+    {
+      "epoch": 0.7945257636009815,
+      "grad_norm": 0.1735035479068756,
+      "learning_rate": 0.001,
+      "loss": 1.9675,
+      "step": 18781
+    },
+    {
+      "epoch": 0.7945680683644979,
+      "grad_norm": 0.1507929265499115,
+      "learning_rate": 0.001,
+      "loss": 1.7867,
+      "step": 18782
+    },
+    {
+      "epoch": 0.7946103731280142,
+      "grad_norm": 0.47875481843948364,
+      "learning_rate": 0.001,
+      "loss": 1.931,
+      "step": 18783
+    },
+    {
+      "epoch": 0.7946526778915306,
+      "grad_norm": 1.3483095169067383,
+      "learning_rate": 0.001,
+      "loss": 1.6993,
+      "step": 18784
+    },
+    {
+      "epoch": 0.794694982655047,
+      "grad_norm": 0.13868524134159088,
+      "learning_rate": 0.001,
+      "loss": 1.9223,
+      "step": 18785
+    },
+    {
+      "epoch": 0.7947372874185633,
+      "grad_norm": 0.2855166494846344,
+      "learning_rate": 0.001,
+      "loss": 1.5803,
+      "step": 18786
+    },
+    {
+      "epoch": 0.7947795921820797,
+      "grad_norm": 0.16588838398456573,
+      "learning_rate": 0.001,
+      "loss": 2.3712,
+      "step": 18787
+    },
+    {
+      "epoch": 0.7948218969455961,
+      "grad_norm": 0.1705247461795807,
+      "learning_rate": 0.001,
+      "loss": 2.6876,
+      "step": 18788
+    },
+    {
+      "epoch": 0.7948642017091124,
+      "grad_norm": 0.19775214791297913,
+      "learning_rate": 0.001,
+      "loss": 2.2817,
+      "step": 18789
+    },
+    {
+      "epoch": 0.7949065064726288,
+      "grad_norm": 0.691185474395752,
+      "learning_rate": 0.001,
+      "loss": 1.8467,
+      "step": 18790
+    },
+    {
+      "epoch": 0.7949488112361451,
+      "grad_norm": 0.1467396765947342,
+      "learning_rate": 0.001,
+      "loss": 3.6287,
+      "step": 18791
+    },
+    {
+      "epoch": 0.7949911159996615,
+      "grad_norm": 0.1282058209180832,
+      "learning_rate": 0.001,
+      "loss": 2.5868,
+      "step": 18792
+    },
+    {
+      "epoch": 0.795033420763178,
+      "grad_norm": 0.16036106646060944,
+      "learning_rate": 0.001,
+      "loss": 1.739,
+      "step": 18793
+    },
+    {
+      "epoch": 0.7950757255266943,
+      "grad_norm": 0.17441564798355103,
+      "learning_rate": 0.001,
+      "loss": 1.7735,
+      "step": 18794
+    },
+    {
+      "epoch": 0.7951180302902107,
+      "grad_norm": 0.15197812020778656,
+      "learning_rate": 0.001,
+      "loss": 1.4933,
+      "step": 18795
+    },
+    {
+      "epoch": 0.7951603350537271,
+      "grad_norm": 0.16789105534553528,
+      "learning_rate": 0.001,
+      "loss": 1.5755,
+      "step": 18796
+    },
+    {
+      "epoch": 0.7952026398172434,
+      "grad_norm": 0.34609055519104004,
+      "learning_rate": 0.001,
+      "loss": 2.2578,
+      "step": 18797
+    },
+    {
+      "epoch": 0.7952449445807598,
+      "grad_norm": 0.1598827838897705,
+      "learning_rate": 0.001,
+      "loss": 2.316,
+      "step": 18798
+    },
+    {
+      "epoch": 0.7952872493442762,
+      "grad_norm": 0.15156984329223633,
+      "learning_rate": 0.001,
+      "loss": 2.456,
+      "step": 18799
+    },
+    {
+      "epoch": 0.7953295541077925,
+      "grad_norm": 0.151535302400589,
+      "learning_rate": 0.001,
+      "loss": 1.4494,
+      "step": 18800
+    },
+    {
+      "epoch": 0.7953718588713089,
+      "grad_norm": 0.19072596728801727,
+      "learning_rate": 0.001,
+      "loss": 2.3097,
+      "step": 18801
+    },
+    {
+      "epoch": 0.7954141636348253,
+      "grad_norm": 0.14297893643379211,
+      "learning_rate": 0.001,
+      "loss": 1.9724,
+      "step": 18802
+    },
+    {
+      "epoch": 0.7954564683983416,
+      "grad_norm": 25.156417846679688,
+      "learning_rate": 0.001,
+      "loss": 2.0334,
+      "step": 18803
+    },
+    {
+      "epoch": 0.795498773161858,
+      "grad_norm": 2.4311437606811523,
+      "learning_rate": 0.001,
+      "loss": 2.0976,
+      "step": 18804
+    },
+    {
+      "epoch": 0.7955410779253744,
+      "grad_norm": 0.24896959960460663,
+      "learning_rate": 0.001,
+      "loss": 2.2145,
+      "step": 18805
+    },
+    {
+      "epoch": 0.7955833826888907,
+      "grad_norm": 0.17482326924800873,
+      "learning_rate": 0.001,
+      "loss": 2.1033,
+      "step": 18806
+    },
+    {
+      "epoch": 0.7956256874524071,
+      "grad_norm": 0.13971750438213348,
+      "learning_rate": 0.001,
+      "loss": 1.864,
+      "step": 18807
+    },
+    {
+      "epoch": 0.7956679922159235,
+      "grad_norm": 0.227436363697052,
+      "learning_rate": 0.001,
+      "loss": 2.6199,
+      "step": 18808
+    },
+    {
+      "epoch": 0.7957102969794398,
+      "grad_norm": 0.14270329475402832,
+      "learning_rate": 0.001,
+      "loss": 2.267,
+      "step": 18809
+    },
+    {
+      "epoch": 0.7957526017429563,
+      "grad_norm": 0.17308275401592255,
+      "learning_rate": 0.001,
+      "loss": 2.0791,
+      "step": 18810
+    },
+    {
+      "epoch": 0.7957949065064727,
+      "grad_norm": 0.16473335027694702,
+      "learning_rate": 0.001,
+      "loss": 1.7909,
+      "step": 18811
+    },
+    {
+      "epoch": 0.795837211269989,
+      "grad_norm": 0.17767003178596497,
+      "learning_rate": 0.001,
+      "loss": 3.3289,
+      "step": 18812
+    },
+    {
+      "epoch": 0.7958795160335054,
+      "grad_norm": 0.182156041264534,
+      "learning_rate": 0.001,
+      "loss": 2.5886,
+      "step": 18813
+    },
+    {
+      "epoch": 0.7959218207970218,
+      "grad_norm": 0.6405094861984253,
+      "learning_rate": 0.001,
+      "loss": 2.5443,
+      "step": 18814
+    },
+    {
+      "epoch": 0.7959641255605381,
+      "grad_norm": 0.20531894266605377,
+      "learning_rate": 0.001,
+      "loss": 2.3202,
+      "step": 18815
+    },
+    {
+      "epoch": 0.7960064303240545,
+      "grad_norm": 0.1372930109500885,
+      "learning_rate": 0.001,
+      "loss": 2.0397,
+      "step": 18816
+    },
+    {
+      "epoch": 0.7960487350875709,
+      "grad_norm": 0.15792180597782135,
+      "learning_rate": 0.001,
+      "loss": 2.4564,
+      "step": 18817
+    },
+    {
+      "epoch": 0.7960910398510872,
+      "grad_norm": 1.705318808555603,
+      "learning_rate": 0.001,
+      "loss": 1.7707,
+      "step": 18818
+    },
+    {
+      "epoch": 0.7961333446146036,
+      "grad_norm": 0.1559934914112091,
+      "learning_rate": 0.001,
+      "loss": 1.7459,
+      "step": 18819
+    },
+    {
+      "epoch": 0.79617564937812,
+      "grad_norm": 0.1450294852256775,
+      "learning_rate": 0.001,
+      "loss": 2.4879,
+      "step": 18820
+    },
+    {
+      "epoch": 0.7962179541416363,
+      "grad_norm": 0.1646071970462799,
+      "learning_rate": 0.001,
+      "loss": 2.2083,
+      "step": 18821
+    },
+    {
+      "epoch": 0.7962602589051527,
+      "grad_norm": 6.388550758361816,
+      "learning_rate": 0.001,
+      "loss": 3.5271,
+      "step": 18822
+    },
+    {
+      "epoch": 0.7963025636686691,
+      "grad_norm": 0.13386476039886475,
+      "learning_rate": 0.001,
+      "loss": 2.1901,
+      "step": 18823
+    },
+    {
+      "epoch": 0.7963448684321854,
+      "grad_norm": 0.15483008325099945,
+      "learning_rate": 0.001,
+      "loss": 2.7349,
+      "step": 18824
+    },
+    {
+      "epoch": 0.7963871731957018,
+      "grad_norm": 0.16752897202968597,
+      "learning_rate": 0.001,
+      "loss": 2.2906,
+      "step": 18825
+    },
+    {
+      "epoch": 0.7964294779592183,
+      "grad_norm": 0.17875881493091583,
+      "learning_rate": 0.001,
+      "loss": 2.8077,
+      "step": 18826
+    },
+    {
+      "epoch": 0.7964717827227346,
+      "grad_norm": 0.16410550475120544,
+      "learning_rate": 0.001,
+      "loss": 2.6807,
+      "step": 18827
+    },
+    {
+      "epoch": 0.796514087486251,
+      "grad_norm": 0.15869401395320892,
+      "learning_rate": 0.001,
+      "loss": 2.016,
+      "step": 18828
+    },
+    {
+      "epoch": 0.7965563922497674,
+      "grad_norm": 0.24351370334625244,
+      "learning_rate": 0.001,
+      "loss": 2.5587,
+      "step": 18829
+    },
+    {
+      "epoch": 0.7965986970132837,
+      "grad_norm": 0.16663382947444916,
+      "learning_rate": 0.001,
+      "loss": 1.3997,
+      "step": 18830
+    },
+    {
+      "epoch": 0.7966410017768001,
+      "grad_norm": 0.15760239958763123,
+      "learning_rate": 0.001,
+      "loss": 1.8689,
+      "step": 18831
+    },
+    {
+      "epoch": 0.7966833065403165,
+      "grad_norm": 0.19611531496047974,
+      "learning_rate": 0.001,
+      "loss": 1.8662,
+      "step": 18832
+    },
+    {
+      "epoch": 0.7967256113038328,
+      "grad_norm": 0.1665828973054886,
+      "learning_rate": 0.001,
+      "loss": 1.6031,
+      "step": 18833
+    },
+    {
+      "epoch": 0.7967679160673492,
+      "grad_norm": 1.8677862882614136,
+      "learning_rate": 0.001,
+      "loss": 1.6596,
+      "step": 18834
+    },
+    {
+      "epoch": 0.7968102208308655,
+      "grad_norm": 16.954547882080078,
+      "learning_rate": 0.001,
+      "loss": 1.8566,
+      "step": 18835
+    },
+    {
+      "epoch": 0.7968525255943819,
+      "grad_norm": 0.18517085909843445,
+      "learning_rate": 0.001,
+      "loss": 2.7776,
+      "step": 18836
+    },
+    {
+      "epoch": 0.7968948303578983,
+      "grad_norm": 0.15459203720092773,
+      "learning_rate": 0.001,
+      "loss": 1.4028,
+      "step": 18837
+    },
+    {
+      "epoch": 0.7969371351214146,
+      "grad_norm": 1.6629470586776733,
+      "learning_rate": 0.001,
+      "loss": 2.8192,
+      "step": 18838
+    },
+    {
+      "epoch": 0.796979439884931,
+      "grad_norm": 0.20710648596286774,
+      "learning_rate": 0.001,
+      "loss": 2.1015,
+      "step": 18839
+    },
+    {
+      "epoch": 0.7970217446484474,
+      "grad_norm": 0.2469254583120346,
+      "learning_rate": 0.001,
+      "loss": 2.866,
+      "step": 18840
+    },
+    {
+      "epoch": 0.7970640494119637,
+      "grad_norm": 0.8012136816978455,
+      "learning_rate": 0.001,
+      "loss": 3.8139,
+      "step": 18841
+    },
+    {
+      "epoch": 0.7971063541754801,
+      "grad_norm": 0.4675610661506653,
+      "learning_rate": 0.001,
+      "loss": 2.2759,
+      "step": 18842
+    },
+    {
+      "epoch": 0.7971486589389966,
+      "grad_norm": 0.2675013244152069,
+      "learning_rate": 0.001,
+      "loss": 2.0761,
+      "step": 18843
+    },
+    {
+      "epoch": 0.7971909637025129,
+      "grad_norm": 0.18237780034542084,
+      "learning_rate": 0.001,
+      "loss": 2.3745,
+      "step": 18844
+    },
+    {
+      "epoch": 0.7972332684660293,
+      "grad_norm": 3.2315402030944824,
+      "learning_rate": 0.001,
+      "loss": 2.6036,
+      "step": 18845
+    },
+    {
+      "epoch": 0.7972755732295457,
+      "grad_norm": 0.17838868498802185,
+      "learning_rate": 0.001,
+      "loss": 2.091,
+      "step": 18846
+    },
+    {
+      "epoch": 0.797317877993062,
+      "grad_norm": 1.5021365880966187,
+      "learning_rate": 0.001,
+      "loss": 2.3548,
+      "step": 18847
+    },
+    {
+      "epoch": 0.7973601827565784,
+      "grad_norm": 0.1974629908800125,
+      "learning_rate": 0.001,
+      "loss": 1.9397,
+      "step": 18848
+    },
+    {
+      "epoch": 0.7974024875200948,
+      "grad_norm": 0.8569322228431702,
+      "learning_rate": 0.001,
+      "loss": 2.618,
+      "step": 18849
+    },
+    {
+      "epoch": 0.7974447922836111,
+      "grad_norm": 0.20803521573543549,
+      "learning_rate": 0.001,
+      "loss": 2.1371,
+      "step": 18850
+    },
+    {
+      "epoch": 0.7974870970471275,
+      "grad_norm": 0.24137185513973236,
+      "learning_rate": 0.001,
+      "loss": 2.1892,
+      "step": 18851
+    },
+    {
+      "epoch": 0.7975294018106439,
+      "grad_norm": 0.7290237545967102,
+      "learning_rate": 0.001,
+      "loss": 1.9481,
+      "step": 18852
+    },
+    {
+      "epoch": 0.7975717065741602,
+      "grad_norm": 0.1846807599067688,
+      "learning_rate": 0.001,
+      "loss": 1.9675,
+      "step": 18853
+    },
+    {
+      "epoch": 0.7976140113376766,
+      "grad_norm": 0.19551526010036469,
+      "learning_rate": 0.001,
+      "loss": 2.5217,
+      "step": 18854
+    },
+    {
+      "epoch": 0.797656316101193,
+      "grad_norm": 0.2642076909542084,
+      "learning_rate": 0.001,
+      "loss": 2.4063,
+      "step": 18855
+    },
+    {
+      "epoch": 0.7976986208647093,
+      "grad_norm": 0.15774700045585632,
+      "learning_rate": 0.001,
+      "loss": 2.3304,
+      "step": 18856
+    },
+    {
+      "epoch": 0.7977409256282257,
+      "grad_norm": 0.23456871509552002,
+      "learning_rate": 0.001,
+      "loss": 2.4911,
+      "step": 18857
+    },
+    {
+      "epoch": 0.7977832303917421,
+      "grad_norm": 0.15300516784191132,
+      "learning_rate": 0.001,
+      "loss": 1.6669,
+      "step": 18858
+    },
+    {
+      "epoch": 0.7978255351552584,
+      "grad_norm": 0.171369269490242,
+      "learning_rate": 0.001,
+      "loss": 2.9436,
+      "step": 18859
+    },
+    {
+      "epoch": 0.7978678399187749,
+      "grad_norm": 0.17169873416423798,
+      "learning_rate": 0.001,
+      "loss": 2.2797,
+      "step": 18860
+    },
+    {
+      "epoch": 0.7979101446822913,
+      "grad_norm": 0.17338159680366516,
+      "learning_rate": 0.001,
+      "loss": 1.5475,
+      "step": 18861
+    },
+    {
+      "epoch": 0.7979524494458076,
+      "grad_norm": 0.19149628281593323,
+      "learning_rate": 0.001,
+      "loss": 2.2884,
+      "step": 18862
+    },
+    {
+      "epoch": 0.797994754209324,
+      "grad_norm": 0.3046054244041443,
+      "learning_rate": 0.001,
+      "loss": 2.6577,
+      "step": 18863
+    },
+    {
+      "epoch": 0.7980370589728404,
+      "grad_norm": 0.19914156198501587,
+      "learning_rate": 0.001,
+      "loss": 3.0521,
+      "step": 18864
+    },
+    {
+      "epoch": 0.7980793637363567,
+      "grad_norm": 0.14342699944972992,
+      "learning_rate": 0.001,
+      "loss": 2.2552,
+      "step": 18865
+    },
+    {
+      "epoch": 0.7981216684998731,
+      "grad_norm": 0.16391606628894806,
+      "learning_rate": 0.001,
+      "loss": 1.8426,
+      "step": 18866
+    },
+    {
+      "epoch": 0.7981639732633895,
+      "grad_norm": 0.17975471913814545,
+      "learning_rate": 0.001,
+      "loss": 2.6264,
+      "step": 18867
+    },
+    {
+      "epoch": 0.7982062780269058,
+      "grad_norm": 0.2970002293586731,
+      "learning_rate": 0.001,
+      "loss": 2.2272,
+      "step": 18868
+    },
+    {
+      "epoch": 0.7982485827904222,
+      "grad_norm": 4.976659774780273,
+      "learning_rate": 0.001,
+      "loss": 2.746,
+      "step": 18869
+    },
+    {
+      "epoch": 0.7982908875539386,
+      "grad_norm": 0.34519144892692566,
+      "learning_rate": 0.001,
+      "loss": 2.1262,
+      "step": 18870
+    },
+    {
+      "epoch": 0.7983331923174549,
+      "grad_norm": 0.14414265751838684,
+      "learning_rate": 0.001,
+      "loss": 2.1009,
+      "step": 18871
+    },
+    {
+      "epoch": 0.7983754970809713,
+      "grad_norm": 0.15004147589206696,
+      "learning_rate": 0.001,
+      "loss": 1.9651,
+      "step": 18872
+    },
+    {
+      "epoch": 0.7984178018444877,
+      "grad_norm": 0.1422182321548462,
+      "learning_rate": 0.001,
+      "loss": 1.8298,
+      "step": 18873
+    },
+    {
+      "epoch": 0.798460106608004,
+      "grad_norm": 0.1388990730047226,
+      "learning_rate": 0.001,
+      "loss": 1.8034,
+      "step": 18874
+    },
+    {
+      "epoch": 0.7985024113715204,
+      "grad_norm": 0.14638282358646393,
+      "learning_rate": 0.001,
+      "loss": 2.0963,
+      "step": 18875
+    },
+    {
+      "epoch": 0.7985447161350369,
+      "grad_norm": 0.17456696927547455,
+      "learning_rate": 0.001,
+      "loss": 1.9739,
+      "step": 18876
+    },
+    {
+      "epoch": 0.7985870208985532,
+      "grad_norm": 0.17205384373664856,
+      "learning_rate": 0.001,
+      "loss": 3.2143,
+      "step": 18877
+    },
+    {
+      "epoch": 0.7986293256620696,
+      "grad_norm": 0.16180726885795593,
+      "learning_rate": 0.001,
+      "loss": 2.422,
+      "step": 18878
+    },
+    {
+      "epoch": 0.7986716304255859,
+      "grad_norm": 8.663548469543457,
+      "learning_rate": 0.001,
+      "loss": 2.9214,
+      "step": 18879
+    },
+    {
+      "epoch": 0.7987139351891023,
+      "grad_norm": 0.1585705280303955,
+      "learning_rate": 0.001,
+      "loss": 2.3293,
+      "step": 18880
+    },
+    {
+      "epoch": 0.7987562399526187,
+      "grad_norm": 0.19919650256633759,
+      "learning_rate": 0.001,
+      "loss": 1.7659,
+      "step": 18881
+    },
+    {
+      "epoch": 0.798798544716135,
+      "grad_norm": 0.18491598963737488,
+      "learning_rate": 0.001,
+      "loss": 1.8007,
+      "step": 18882
+    },
+    {
+      "epoch": 0.7988408494796514,
+      "grad_norm": 4.440988540649414,
+      "learning_rate": 0.001,
+      "loss": 2.3786,
+      "step": 18883
+    },
+    {
+      "epoch": 0.7988831542431678,
+      "grad_norm": 0.346278578042984,
+      "learning_rate": 0.001,
+      "loss": 2.312,
+      "step": 18884
+    },
+    {
+      "epoch": 0.7989254590066841,
+      "grad_norm": 0.16744297742843628,
+      "learning_rate": 0.001,
+      "loss": 3.3667,
+      "step": 18885
+    },
+    {
+      "epoch": 0.7989677637702005,
+      "grad_norm": 0.2730819880962372,
+      "learning_rate": 0.001,
+      "loss": 1.373,
+      "step": 18886
+    },
+    {
+      "epoch": 0.7990100685337169,
+      "grad_norm": 0.1686711460351944,
+      "learning_rate": 0.001,
+      "loss": 2.4004,
+      "step": 18887
+    },
+    {
+      "epoch": 0.7990523732972332,
+      "grad_norm": 0.32986345887184143,
+      "learning_rate": 0.001,
+      "loss": 2.1665,
+      "step": 18888
+    },
+    {
+      "epoch": 0.7990946780607496,
+      "grad_norm": 0.23133182525634766,
+      "learning_rate": 0.001,
+      "loss": 2.0615,
+      "step": 18889
+    },
+    {
+      "epoch": 0.799136982824266,
+      "grad_norm": 0.18338367342948914,
+      "learning_rate": 0.001,
+      "loss": 2.3496,
+      "step": 18890
+    },
+    {
+      "epoch": 0.7991792875877823,
+      "grad_norm": 0.15058721601963043,
+      "learning_rate": 0.001,
+      "loss": 1.7269,
+      "step": 18891
+    },
+    {
+      "epoch": 0.7992215923512987,
+      "grad_norm": 3.878596544265747,
+      "learning_rate": 0.001,
+      "loss": 2.3469,
+      "step": 18892
+    },
+    {
+      "epoch": 0.7992638971148152,
+      "grad_norm": 0.15892788767814636,
+      "learning_rate": 0.001,
+      "loss": 2.3426,
+      "step": 18893
+    },
+    {
+      "epoch": 0.7993062018783315,
+      "grad_norm": 0.18293994665145874,
+      "learning_rate": 0.001,
+      "loss": 1.5293,
+      "step": 18894
+    },
+    {
+      "epoch": 0.7993485066418479,
+      "grad_norm": 0.1805136799812317,
+      "learning_rate": 0.001,
+      "loss": 2.0846,
+      "step": 18895
+    },
+    {
+      "epoch": 0.7993908114053643,
+      "grad_norm": 0.16091568768024445,
+      "learning_rate": 0.001,
+      "loss": 1.4057,
+      "step": 18896
+    },
+    {
+      "epoch": 0.7994331161688806,
+      "grad_norm": 0.20409560203552246,
+      "learning_rate": 0.001,
+      "loss": 3.3809,
+      "step": 18897
+    },
+    {
+      "epoch": 0.799475420932397,
+      "grad_norm": 0.16749215126037598,
+      "learning_rate": 0.001,
+      "loss": 2.3456,
+      "step": 18898
+    },
+    {
+      "epoch": 0.7995177256959134,
+      "grad_norm": 0.165800079703331,
+      "learning_rate": 0.001,
+      "loss": 2.5899,
+      "step": 18899
+    },
+    {
+      "epoch": 0.7995600304594297,
+      "grad_norm": 0.17839626967906952,
+      "learning_rate": 0.001,
+      "loss": 1.6863,
+      "step": 18900
+    },
+    {
+      "epoch": 0.7996023352229461,
+      "grad_norm": 0.15647844970226288,
+      "learning_rate": 0.001,
+      "loss": 2.4963,
+      "step": 18901
+    },
+    {
+      "epoch": 0.7996446399864625,
+      "grad_norm": 0.22941941022872925,
+      "learning_rate": 0.001,
+      "loss": 2.7686,
+      "step": 18902
+    },
+    {
+      "epoch": 0.7996869447499788,
+      "grad_norm": 0.17130149900913239,
+      "learning_rate": 0.001,
+      "loss": 2.7059,
+      "step": 18903
+    },
+    {
+      "epoch": 0.7997292495134952,
+      "grad_norm": 0.1701461225748062,
+      "learning_rate": 0.001,
+      "loss": 1.87,
+      "step": 18904
+    },
+    {
+      "epoch": 0.7997715542770116,
+      "grad_norm": 0.514470100402832,
+      "learning_rate": 0.001,
+      "loss": 2.4732,
+      "step": 18905
+    },
+    {
+      "epoch": 0.7998138590405279,
+      "grad_norm": 0.13958191871643066,
+      "learning_rate": 0.001,
+      "loss": 2.6836,
+      "step": 18906
+    },
+    {
+      "epoch": 0.7998561638040443,
+      "grad_norm": 0.15818491578102112,
+      "learning_rate": 0.001,
+      "loss": 2.7445,
+      "step": 18907
+    },
+    {
+      "epoch": 0.7998984685675607,
+      "grad_norm": 0.18377156555652618,
+      "learning_rate": 0.001,
+      "loss": 1.9231,
+      "step": 18908
+    },
+    {
+      "epoch": 0.799940773331077,
+      "grad_norm": 0.35791802406311035,
+      "learning_rate": 0.001,
+      "loss": 2.0333,
+      "step": 18909
+    },
+    {
+      "epoch": 0.7999830780945935,
+      "grad_norm": 0.3409176170825958,
+      "learning_rate": 0.001,
+      "loss": 2.113,
+      "step": 18910
+    },
+    {
+      "epoch": 0.8000253828581099,
+      "grad_norm": 0.1757473349571228,
+      "learning_rate": 0.001,
+      "loss": 1.5102,
+      "step": 18911
+    },
+    {
+      "epoch": 0.8000676876216262,
+      "grad_norm": 1.6405963897705078,
+      "learning_rate": 0.001,
+      "loss": 2.0664,
+      "step": 18912
+    },
+    {
+      "epoch": 0.8001099923851426,
+      "grad_norm": 0.1534256935119629,
+      "learning_rate": 0.001,
+      "loss": 2.2177,
+      "step": 18913
+    },
+    {
+      "epoch": 0.800152297148659,
+      "grad_norm": 0.21698273718357086,
+      "learning_rate": 0.001,
+      "loss": 2.74,
+      "step": 18914
+    },
+    {
+      "epoch": 0.8001946019121753,
+      "grad_norm": 0.1605890989303589,
+      "learning_rate": 0.001,
+      "loss": 1.9252,
+      "step": 18915
+    },
+    {
+      "epoch": 0.8002369066756917,
+      "grad_norm": 0.16456551849842072,
+      "learning_rate": 0.001,
+      "loss": 1.993,
+      "step": 18916
+    },
+    {
+      "epoch": 0.8002792114392081,
+      "grad_norm": 0.1534430980682373,
+      "learning_rate": 0.001,
+      "loss": 1.7507,
+      "step": 18917
+    },
+    {
+      "epoch": 0.8003215162027244,
+      "grad_norm": 0.17245927453041077,
+      "learning_rate": 0.001,
+      "loss": 3.3313,
+      "step": 18918
+    },
+    {
+      "epoch": 0.8003638209662408,
+      "grad_norm": 0.2450478971004486,
+      "learning_rate": 0.001,
+      "loss": 2.6292,
+      "step": 18919
+    },
+    {
+      "epoch": 0.8004061257297572,
+      "grad_norm": 0.2959800958633423,
+      "learning_rate": 0.001,
+      "loss": 2.498,
+      "step": 18920
+    },
+    {
+      "epoch": 0.8004484304932735,
+      "grad_norm": 0.15317165851593018,
+      "learning_rate": 0.001,
+      "loss": 2.6568,
+      "step": 18921
+    },
+    {
+      "epoch": 0.8004907352567899,
+      "grad_norm": 0.1782752275466919,
+      "learning_rate": 0.001,
+      "loss": 1.7211,
+      "step": 18922
+    },
+    {
+      "epoch": 0.8005330400203063,
+      "grad_norm": 0.1783091425895691,
+      "learning_rate": 0.001,
+      "loss": 2.6381,
+      "step": 18923
+    },
+    {
+      "epoch": 0.8005753447838226,
+      "grad_norm": 0.17199842631816864,
+      "learning_rate": 0.001,
+      "loss": 2.553,
+      "step": 18924
+    },
+    {
+      "epoch": 0.800617649547339,
+      "grad_norm": 1.3231291770935059,
+      "learning_rate": 0.001,
+      "loss": 2.695,
+      "step": 18925
+    },
+    {
+      "epoch": 0.8006599543108553,
+      "grad_norm": 2.2281360626220703,
+      "learning_rate": 0.001,
+      "loss": 2.2209,
+      "step": 18926
+    },
+    {
+      "epoch": 0.8007022590743718,
+      "grad_norm": 0.20170575380325317,
+      "learning_rate": 0.001,
+      "loss": 2.7078,
+      "step": 18927
+    },
+    {
+      "epoch": 0.8007445638378882,
+      "grad_norm": 0.17073240876197815,
+      "learning_rate": 0.001,
+      "loss": 2.4441,
+      "step": 18928
+    },
+    {
+      "epoch": 0.8007868686014045,
+      "grad_norm": 0.15086358785629272,
+      "learning_rate": 0.001,
+      "loss": 2.0499,
+      "step": 18929
+    },
+    {
+      "epoch": 0.8008291733649209,
+      "grad_norm": 0.8255175352096558,
+      "learning_rate": 0.001,
+      "loss": 2.4192,
+      "step": 18930
+    },
+    {
+      "epoch": 0.8008714781284373,
+      "grad_norm": 0.1681467890739441,
+      "learning_rate": 0.001,
+      "loss": 2.2557,
+      "step": 18931
+    },
+    {
+      "epoch": 0.8009137828919536,
+      "grad_norm": 0.14386318624019623,
+      "learning_rate": 0.001,
+      "loss": 1.9512,
+      "step": 18932
+    },
+    {
+      "epoch": 0.80095608765547,
+      "grad_norm": 0.36584439873695374,
+      "learning_rate": 0.001,
+      "loss": 1.7842,
+      "step": 18933
+    },
+    {
+      "epoch": 0.8009983924189864,
+      "grad_norm": 0.14948461949825287,
+      "learning_rate": 0.001,
+      "loss": 1.7132,
+      "step": 18934
+    },
+    {
+      "epoch": 0.8010406971825027,
+      "grad_norm": 0.1546405851840973,
+      "learning_rate": 0.001,
+      "loss": 1.8271,
+      "step": 18935
+    },
+    {
+      "epoch": 0.8010830019460191,
+      "grad_norm": 0.15600258111953735,
+      "learning_rate": 0.001,
+      "loss": 2.0882,
+      "step": 18936
+    },
+    {
+      "epoch": 0.8011253067095355,
+      "grad_norm": 0.20449069142341614,
+      "learning_rate": 0.001,
+      "loss": 1.752,
+      "step": 18937
+    },
+    {
+      "epoch": 0.8011676114730518,
+      "grad_norm": 0.143666610121727,
+      "learning_rate": 0.001,
+      "loss": 2.4691,
+      "step": 18938
+    },
+    {
+      "epoch": 0.8012099162365682,
+      "grad_norm": 0.17219458520412445,
+      "learning_rate": 0.001,
+      "loss": 1.9348,
+      "step": 18939
+    },
+    {
+      "epoch": 0.8012522210000846,
+      "grad_norm": 0.14504066109657288,
+      "learning_rate": 0.001,
+      "loss": 2.5255,
+      "step": 18940
+    },
+    {
+      "epoch": 0.8012945257636009,
+      "grad_norm": 0.1552887260913849,
+      "learning_rate": 0.001,
+      "loss": 1.8034,
+      "step": 18941
+    },
+    {
+      "epoch": 0.8013368305271173,
+      "grad_norm": 0.3989625871181488,
+      "learning_rate": 0.001,
+      "loss": 1.6939,
+      "step": 18942
+    },
+    {
+      "epoch": 0.8013791352906338,
+      "grad_norm": 0.15734554827213287,
+      "learning_rate": 0.001,
+      "loss": 1.8861,
+      "step": 18943
+    },
+    {
+      "epoch": 0.8014214400541501,
+      "grad_norm": 0.159386545419693,
+      "learning_rate": 0.001,
+      "loss": 2.3796,
+      "step": 18944
+    },
+    {
+      "epoch": 0.8014637448176665,
+      "grad_norm": 0.1559273898601532,
+      "learning_rate": 0.001,
+      "loss": 3.0239,
+      "step": 18945
+    },
+    {
+      "epoch": 0.8015060495811829,
+      "grad_norm": 0.15622538328170776,
+      "learning_rate": 0.001,
+      "loss": 2.6106,
+      "step": 18946
+    },
+    {
+      "epoch": 0.8015483543446992,
+      "grad_norm": 0.17876319587230682,
+      "learning_rate": 0.001,
+      "loss": 1.8505,
+      "step": 18947
+    },
+    {
+      "epoch": 0.8015906591082156,
+      "grad_norm": 0.1366472691297531,
+      "learning_rate": 0.001,
+      "loss": 1.9985,
+      "step": 18948
+    },
+    {
+      "epoch": 0.801632963871732,
+      "grad_norm": 0.17091487348079681,
+      "learning_rate": 0.001,
+      "loss": 2.836,
+      "step": 18949
+    },
+    {
+      "epoch": 0.8016752686352483,
+      "grad_norm": 0.1562683880329132,
+      "learning_rate": 0.001,
+      "loss": 1.4412,
+      "step": 18950
+    },
+    {
+      "epoch": 0.8017175733987647,
+      "grad_norm": 0.1484891027212143,
+      "learning_rate": 0.001,
+      "loss": 1.8152,
+      "step": 18951
+    },
+    {
+      "epoch": 0.8017598781622811,
+      "grad_norm": 1.3076022863388062,
+      "learning_rate": 0.001,
+      "loss": 2.7306,
+      "step": 18952
+    },
+    {
+      "epoch": 0.8018021829257974,
+      "grad_norm": 0.16072845458984375,
+      "learning_rate": 0.001,
+      "loss": 2.0295,
+      "step": 18953
+    },
+    {
+      "epoch": 0.8018444876893138,
+      "grad_norm": 0.14599545300006866,
+      "learning_rate": 0.001,
+      "loss": 2.7667,
+      "step": 18954
+    },
+    {
+      "epoch": 0.8018867924528302,
+      "grad_norm": 0.15150395035743713,
+      "learning_rate": 0.001,
+      "loss": 3.1838,
+      "step": 18955
+    },
+    {
+      "epoch": 0.8019290972163465,
+      "grad_norm": 0.18270723521709442,
+      "learning_rate": 0.001,
+      "loss": 1.6511,
+      "step": 18956
+    },
+    {
+      "epoch": 0.8019714019798629,
+      "grad_norm": 0.1671307235956192,
+      "learning_rate": 0.001,
+      "loss": 2.9056,
+      "step": 18957
+    },
+    {
+      "epoch": 0.8020137067433794,
+      "grad_norm": 0.14215819537639618,
+      "learning_rate": 0.001,
+      "loss": 2.8229,
+      "step": 18958
+    },
+    {
+      "epoch": 0.8020560115068956,
+      "grad_norm": 0.14553506672382355,
+      "learning_rate": 0.001,
+      "loss": 1.3122,
+      "step": 18959
+    },
+    {
+      "epoch": 0.8020983162704121,
+      "grad_norm": 0.1251428723335266,
+      "learning_rate": 0.001,
+      "loss": 1.9551,
+      "step": 18960
+    },
+    {
+      "epoch": 0.8021406210339285,
+      "grad_norm": 0.7137900590896606,
+      "learning_rate": 0.001,
+      "loss": 2.8314,
+      "step": 18961
+    },
+    {
+      "epoch": 0.8021829257974448,
+      "grad_norm": 0.1495407074689865,
+      "learning_rate": 0.001,
+      "loss": 1.7774,
+      "step": 18962
+    },
+    {
+      "epoch": 0.8022252305609612,
+      "grad_norm": 0.14792723953723907,
+      "learning_rate": 0.001,
+      "loss": 2.1685,
+      "step": 18963
+    },
+    {
+      "epoch": 0.8022675353244776,
+      "grad_norm": 0.7943689823150635,
+      "learning_rate": 0.001,
+      "loss": 2.1938,
+      "step": 18964
+    },
+    {
+      "epoch": 0.8023098400879939,
+      "grad_norm": 19.49368667602539,
+      "learning_rate": 0.001,
+      "loss": 3.6481,
+      "step": 18965
+    },
+    {
+      "epoch": 0.8023521448515103,
+      "grad_norm": 2.110476493835449,
+      "learning_rate": 0.001,
+      "loss": 1.4227,
+      "step": 18966
+    },
+    {
+      "epoch": 0.8023944496150267,
+      "grad_norm": 0.13607414066791534,
+      "learning_rate": 0.001,
+      "loss": 1.6882,
+      "step": 18967
+    },
+    {
+      "epoch": 0.802436754378543,
+      "grad_norm": 1.1978744268417358,
+      "learning_rate": 0.001,
+      "loss": 1.8216,
+      "step": 18968
+    },
+    {
+      "epoch": 0.8024790591420594,
+      "grad_norm": 0.19706468284130096,
+      "learning_rate": 0.001,
+      "loss": 2.8467,
+      "step": 18969
+    },
+    {
+      "epoch": 0.8025213639055757,
+      "grad_norm": 0.14383366703987122,
+      "learning_rate": 0.001,
+      "loss": 1.8751,
+      "step": 18970
+    },
+    {
+      "epoch": 0.8025636686690921,
+      "grad_norm": 0.20489506423473358,
+      "learning_rate": 0.001,
+      "loss": 2.8281,
+      "step": 18971
+    },
+    {
+      "epoch": 0.8026059734326085,
+      "grad_norm": 1.1792211532592773,
+      "learning_rate": 0.001,
+      "loss": 2.9794,
+      "step": 18972
+    },
+    {
+      "epoch": 0.8026482781961248,
+      "grad_norm": 0.5277860164642334,
+      "learning_rate": 0.001,
+      "loss": 1.6409,
+      "step": 18973
+    },
+    {
+      "epoch": 0.8026905829596412,
+      "grad_norm": 0.13644924759864807,
+      "learning_rate": 0.001,
+      "loss": 1.4607,
+      "step": 18974
+    },
+    {
+      "epoch": 0.8027328877231577,
+      "grad_norm": 0.17572204768657684,
+      "learning_rate": 0.001,
+      "loss": 2.3758,
+      "step": 18975
+    },
+    {
+      "epoch": 0.802775192486674,
+      "grad_norm": 0.356366366147995,
+      "learning_rate": 0.001,
+      "loss": 1.5677,
+      "step": 18976
+    },
+    {
+      "epoch": 0.8028174972501904,
+      "grad_norm": 0.1653011590242386,
+      "learning_rate": 0.001,
+      "loss": 2.6266,
+      "step": 18977
+    },
+    {
+      "epoch": 0.8028598020137068,
+      "grad_norm": 0.1638076901435852,
+      "learning_rate": 0.001,
+      "loss": 1.9443,
+      "step": 18978
+    },
+    {
+      "epoch": 0.8029021067772231,
+      "grad_norm": 0.26030609011650085,
+      "learning_rate": 0.001,
+      "loss": 1.6647,
+      "step": 18979
+    },
+    {
+      "epoch": 0.8029444115407395,
+      "grad_norm": 0.16686412692070007,
+      "learning_rate": 0.001,
+      "loss": 2.2475,
+      "step": 18980
+    },
+    {
+      "epoch": 0.8029867163042559,
+      "grad_norm": 0.2093593031167984,
+      "learning_rate": 0.001,
+      "loss": 2.0305,
+      "step": 18981
+    },
+    {
+      "epoch": 0.8030290210677722,
+      "grad_norm": 0.17999699711799622,
+      "learning_rate": 0.001,
+      "loss": 2.2384,
+      "step": 18982
+    },
+    {
+      "epoch": 0.8030713258312886,
+      "grad_norm": 0.9324764013290405,
+      "learning_rate": 0.001,
+      "loss": 2.6971,
+      "step": 18983
+    },
+    {
+      "epoch": 0.803113630594805,
+      "grad_norm": 0.17794398963451385,
+      "learning_rate": 0.001,
+      "loss": 2.0274,
+      "step": 18984
+    },
+    {
+      "epoch": 0.8031559353583213,
+      "grad_norm": 0.18477022647857666,
+      "learning_rate": 0.001,
+      "loss": 2.2527,
+      "step": 18985
+    },
+    {
+      "epoch": 0.8031982401218377,
+      "grad_norm": 0.1455012857913971,
+      "learning_rate": 0.001,
+      "loss": 2.1315,
+      "step": 18986
+    },
+    {
+      "epoch": 0.8032405448853541,
+      "grad_norm": 1.0563513040542603,
+      "learning_rate": 0.001,
+      "loss": 2.3651,
+      "step": 18987
+    },
+    {
+      "epoch": 0.8032828496488704,
+      "grad_norm": 0.28026530146598816,
+      "learning_rate": 0.001,
+      "loss": 3.0058,
+      "step": 18988
+    },
+    {
+      "epoch": 0.8033251544123868,
+      "grad_norm": 58.185855865478516,
+      "learning_rate": 0.001,
+      "loss": 1.8485,
+      "step": 18989
+    },
+    {
+      "epoch": 0.8033674591759032,
+      "grad_norm": 0.35726267099380493,
+      "learning_rate": 0.001,
+      "loss": 1.8895,
+      "step": 18990
+    },
+    {
+      "epoch": 0.8034097639394195,
+      "grad_norm": 0.2682846486568451,
+      "learning_rate": 0.001,
+      "loss": 1.7967,
+      "step": 18991
+    },
+    {
+      "epoch": 0.803452068702936,
+      "grad_norm": 0.16665484011173248,
+      "learning_rate": 0.001,
+      "loss": 2.3617,
+      "step": 18992
+    },
+    {
+      "epoch": 0.8034943734664524,
+      "grad_norm": 0.16260682046413422,
+      "learning_rate": 0.001,
+      "loss": 2.4323,
+      "step": 18993
+    },
+    {
+      "epoch": 0.8035366782299687,
+      "grad_norm": 0.16122184693813324,
+      "learning_rate": 0.001,
+      "loss": 2.6508,
+      "step": 18994
+    },
+    {
+      "epoch": 0.8035789829934851,
+      "grad_norm": 0.16327929496765137,
+      "learning_rate": 0.001,
+      "loss": 1.5013,
+      "step": 18995
+    },
+    {
+      "epoch": 0.8036212877570015,
+      "grad_norm": 0.19706536829471588,
+      "learning_rate": 0.001,
+      "loss": 2.5097,
+      "step": 18996
+    },
+    {
+      "epoch": 0.8036635925205178,
+      "grad_norm": 0.3086377680301666,
+      "learning_rate": 0.001,
+      "loss": 1.9712,
+      "step": 18997
+    },
+    {
+      "epoch": 0.8037058972840342,
+      "grad_norm": 0.1745164543390274,
+      "learning_rate": 0.001,
+      "loss": 1.9054,
+      "step": 18998
+    },
+    {
+      "epoch": 0.8037482020475506,
+      "grad_norm": 0.20080901682376862,
+      "learning_rate": 0.001,
+      "loss": 2.6199,
+      "step": 18999
+    },
+    {
+      "epoch": 0.8037905068110669,
+      "grad_norm": 0.8673862814903259,
+      "learning_rate": 0.001,
+      "loss": 1.8435,
+      "step": 19000
+    },
+    {
+      "epoch": 0.8038328115745833,
+      "grad_norm": 0.17495474219322205,
+      "learning_rate": 0.001,
+      "loss": 2.3335,
+      "step": 19001
+    },
+    {
+      "epoch": 0.8038751163380997,
+      "grad_norm": 0.2492198944091797,
+      "learning_rate": 0.001,
+      "loss": 3.421,
+      "step": 19002
+    },
+    {
+      "epoch": 0.803917421101616,
+      "grad_norm": 0.2866990864276886,
+      "learning_rate": 0.001,
+      "loss": 2.3914,
+      "step": 19003
+    },
+    {
+      "epoch": 0.8039597258651324,
+      "grad_norm": 0.1283341497182846,
+      "learning_rate": 0.001,
+      "loss": 1.6294,
+      "step": 19004
+    },
+    {
+      "epoch": 0.8040020306286488,
+      "grad_norm": 0.18748348951339722,
+      "learning_rate": 0.001,
+      "loss": 3.779,
+      "step": 19005
+    },
+    {
+      "epoch": 0.8040443353921651,
+      "grad_norm": 0.2228836864233017,
+      "learning_rate": 0.001,
+      "loss": 2.1684,
+      "step": 19006
+    },
+    {
+      "epoch": 0.8040866401556815,
+      "grad_norm": 0.21044659614562988,
+      "learning_rate": 0.001,
+      "loss": 2.5293,
+      "step": 19007
+    },
+    {
+      "epoch": 0.804128944919198,
+      "grad_norm": 0.13109223544597626,
+      "learning_rate": 0.001,
+      "loss": 1.7461,
+      "step": 19008
+    },
+    {
+      "epoch": 0.8041712496827143,
+      "grad_norm": 0.16634885966777802,
+      "learning_rate": 0.001,
+      "loss": 1.6819,
+      "step": 19009
+    },
+    {
+      "epoch": 0.8042135544462307,
+      "grad_norm": 26.008522033691406,
+      "learning_rate": 0.001,
+      "loss": 1.8201,
+      "step": 19010
+    },
+    {
+      "epoch": 0.8042558592097471,
+      "grad_norm": 0.14524191617965698,
+      "learning_rate": 0.001,
+      "loss": 1.7735,
+      "step": 19011
+    },
+    {
+      "epoch": 0.8042981639732634,
+      "grad_norm": 0.14269325137138367,
+      "learning_rate": 0.001,
+      "loss": 2.4637,
+      "step": 19012
+    },
+    {
+      "epoch": 0.8043404687367798,
+      "grad_norm": 0.31678712368011475,
+      "learning_rate": 0.001,
+      "loss": 2.9312,
+      "step": 19013
+    },
+    {
+      "epoch": 0.8043827735002962,
+      "grad_norm": 0.16404123604297638,
+      "learning_rate": 0.001,
+      "loss": 1.8627,
+      "step": 19014
+    },
+    {
+      "epoch": 0.8044250782638125,
+      "grad_norm": 0.16034923493862152,
+      "learning_rate": 0.001,
+      "loss": 2.3652,
+      "step": 19015
+    },
+    {
+      "epoch": 0.8044673830273289,
+      "grad_norm": 0.1982741504907608,
+      "learning_rate": 0.001,
+      "loss": 2.198,
+      "step": 19016
+    },
+    {
+      "epoch": 0.8045096877908452,
+      "grad_norm": 0.1432885080575943,
+      "learning_rate": 0.001,
+      "loss": 1.7994,
+      "step": 19017
+    },
+    {
+      "epoch": 0.8045519925543616,
+      "grad_norm": 0.2005062848329544,
+      "learning_rate": 0.001,
+      "loss": 2.2732,
+      "step": 19018
+    },
+    {
+      "epoch": 0.804594297317878,
+      "grad_norm": 0.18627607822418213,
+      "learning_rate": 0.001,
+      "loss": 2.2855,
+      "step": 19019
+    },
+    {
+      "epoch": 0.8046366020813943,
+      "grad_norm": 0.17629285156726837,
+      "learning_rate": 0.001,
+      "loss": 2.3181,
+      "step": 19020
+    },
+    {
+      "epoch": 0.8046789068449107,
+      "grad_norm": 0.15572401881217957,
+      "learning_rate": 0.001,
+      "loss": 2.2486,
+      "step": 19021
+    },
+    {
+      "epoch": 0.8047212116084271,
+      "grad_norm": 0.13063278794288635,
+      "learning_rate": 0.001,
+      "loss": 2.0039,
+      "step": 19022
+    },
+    {
+      "epoch": 0.8047635163719434,
+      "grad_norm": 0.17360426485538483,
+      "learning_rate": 0.001,
+      "loss": 1.9815,
+      "step": 19023
+    },
+    {
+      "epoch": 0.8048058211354598,
+      "grad_norm": 0.14736105501651764,
+      "learning_rate": 0.001,
+      "loss": 1.7787,
+      "step": 19024
+    },
+    {
+      "epoch": 0.8048481258989763,
+      "grad_norm": 0.40037959814071655,
+      "learning_rate": 0.001,
+      "loss": 1.4608,
+      "step": 19025
+    },
+    {
+      "epoch": 0.8048904306624926,
+      "grad_norm": 0.13307571411132812,
+      "learning_rate": 0.001,
+      "loss": 2.472,
+      "step": 19026
+    },
+    {
+      "epoch": 0.804932735426009,
+      "grad_norm": 0.2580379545688629,
+      "learning_rate": 0.001,
+      "loss": 2.6533,
+      "step": 19027
+    },
+    {
+      "epoch": 0.8049750401895254,
+      "grad_norm": 1.3895004987716675,
+      "learning_rate": 0.001,
+      "loss": 2.1698,
+      "step": 19028
+    },
+    {
+      "epoch": 0.8050173449530417,
+      "grad_norm": 0.1485358327627182,
+      "learning_rate": 0.001,
+      "loss": 1.8783,
+      "step": 19029
+    },
+    {
+      "epoch": 0.8050596497165581,
+      "grad_norm": 33.825557708740234,
+      "learning_rate": 0.001,
+      "loss": 2.6339,
+      "step": 19030
+    },
+    {
+      "epoch": 0.8051019544800745,
+      "grad_norm": 0.15291500091552734,
+      "learning_rate": 0.001,
+      "loss": 1.5062,
+      "step": 19031
+    },
+    {
+      "epoch": 0.8051442592435908,
+      "grad_norm": 0.19544963538646698,
+      "learning_rate": 0.001,
+      "loss": 2.7773,
+      "step": 19032
+    },
+    {
+      "epoch": 0.8051865640071072,
+      "grad_norm": 0.7990001440048218,
+      "learning_rate": 0.001,
+      "loss": 2.6352,
+      "step": 19033
+    },
+    {
+      "epoch": 0.8052288687706236,
+      "grad_norm": 0.20977818965911865,
+      "learning_rate": 0.001,
+      "loss": 2.6468,
+      "step": 19034
+    },
+    {
+      "epoch": 0.8052711735341399,
+      "grad_norm": 0.3944449722766876,
+      "learning_rate": 0.001,
+      "loss": 2.3777,
+      "step": 19035
+    },
+    {
+      "epoch": 0.8053134782976563,
+      "grad_norm": 0.1769852191209793,
+      "learning_rate": 0.001,
+      "loss": 1.7299,
+      "step": 19036
+    },
+    {
+      "epoch": 0.8053557830611727,
+      "grad_norm": 0.2292184680700302,
+      "learning_rate": 0.001,
+      "loss": 2.9147,
+      "step": 19037
+    },
+    {
+      "epoch": 0.805398087824689,
+      "grad_norm": 0.20135535299777985,
+      "learning_rate": 0.001,
+      "loss": 1.8794,
+      "step": 19038
+    },
+    {
+      "epoch": 0.8054403925882054,
+      "grad_norm": 9.884795188903809,
+      "learning_rate": 0.001,
+      "loss": 2.4281,
+      "step": 19039
+    },
+    {
+      "epoch": 0.8054826973517218,
+      "grad_norm": 0.22030168771743774,
+      "learning_rate": 0.001,
+      "loss": 2.7934,
+      "step": 19040
+    },
+    {
+      "epoch": 0.8055250021152381,
+      "grad_norm": 0.3213954269886017,
+      "learning_rate": 0.001,
+      "loss": 1.9529,
+      "step": 19041
+    },
+    {
+      "epoch": 0.8055673068787546,
+      "grad_norm": 0.18723493814468384,
+      "learning_rate": 0.001,
+      "loss": 2.0792,
+      "step": 19042
+    },
+    {
+      "epoch": 0.805609611642271,
+      "grad_norm": 0.19658009707927704,
+      "learning_rate": 0.001,
+      "loss": 2.5037,
+      "step": 19043
+    },
+    {
+      "epoch": 0.8056519164057873,
+      "grad_norm": 0.18413256108760834,
+      "learning_rate": 0.001,
+      "loss": 2.0197,
+      "step": 19044
+    },
+    {
+      "epoch": 0.8056942211693037,
+      "grad_norm": 0.1551973521709442,
+      "learning_rate": 0.001,
+      "loss": 2.6713,
+      "step": 19045
+    },
+    {
+      "epoch": 0.8057365259328201,
+      "grad_norm": 0.18275663256645203,
+      "learning_rate": 0.001,
+      "loss": 2.6608,
+      "step": 19046
+    },
+    {
+      "epoch": 0.8057788306963364,
+      "grad_norm": 0.1611073613166809,
+      "learning_rate": 0.001,
+      "loss": 2.5312,
+      "step": 19047
+    },
+    {
+      "epoch": 0.8058211354598528,
+      "grad_norm": 0.160042405128479,
+      "learning_rate": 0.001,
+      "loss": 1.8463,
+      "step": 19048
+    },
+    {
+      "epoch": 0.8058634402233692,
+      "grad_norm": 0.19275878369808197,
+      "learning_rate": 0.001,
+      "loss": 1.818,
+      "step": 19049
+    },
+    {
+      "epoch": 0.8059057449868855,
+      "grad_norm": 1.6935782432556152,
+      "learning_rate": 0.001,
+      "loss": 4.1077,
+      "step": 19050
+    },
+    {
+      "epoch": 0.8059480497504019,
+      "grad_norm": 0.16775570809841156,
+      "learning_rate": 0.001,
+      "loss": 2.496,
+      "step": 19051
+    },
+    {
+      "epoch": 0.8059903545139183,
+      "grad_norm": 0.1673063337802887,
+      "learning_rate": 0.001,
+      "loss": 2.0237,
+      "step": 19052
+    },
+    {
+      "epoch": 0.8060326592774346,
+      "grad_norm": 0.17094041407108307,
+      "learning_rate": 0.001,
+      "loss": 1.7422,
+      "step": 19053
+    },
+    {
+      "epoch": 0.806074964040951,
+      "grad_norm": 0.18810485303401947,
+      "learning_rate": 0.001,
+      "loss": 2.8559,
+      "step": 19054
+    },
+    {
+      "epoch": 0.8061172688044674,
+      "grad_norm": 0.17345726490020752,
+      "learning_rate": 0.001,
+      "loss": 2.183,
+      "step": 19055
+    },
+    {
+      "epoch": 0.8061595735679837,
+      "grad_norm": 0.27425530552864075,
+      "learning_rate": 0.001,
+      "loss": 2.9688,
+      "step": 19056
+    },
+    {
+      "epoch": 0.8062018783315001,
+      "grad_norm": 16.195755004882812,
+      "learning_rate": 0.001,
+      "loss": 1.9708,
+      "step": 19057
+    },
+    {
+      "epoch": 0.8062441830950166,
+      "grad_norm": 0.15767496824264526,
+      "learning_rate": 0.001,
+      "loss": 1.8641,
+      "step": 19058
+    },
+    {
+      "epoch": 0.8062864878585329,
+      "grad_norm": 0.16588301956653595,
+      "learning_rate": 0.001,
+      "loss": 2.869,
+      "step": 19059
+    },
+    {
+      "epoch": 0.8063287926220493,
+      "grad_norm": 0.15558847784996033,
+      "learning_rate": 0.001,
+      "loss": 1.9133,
+      "step": 19060
+    },
+    {
+      "epoch": 0.8063710973855656,
+      "grad_norm": 0.16034120321273804,
+      "learning_rate": 0.001,
+      "loss": 3.0425,
+      "step": 19061
+    },
+    {
+      "epoch": 0.806413402149082,
+      "grad_norm": 0.1412864327430725,
+      "learning_rate": 0.001,
+      "loss": 2.0012,
+      "step": 19062
+    },
+    {
+      "epoch": 0.8064557069125984,
+      "grad_norm": 0.2045813649892807,
+      "learning_rate": 0.001,
+      "loss": 1.9367,
+      "step": 19063
+    },
+    {
+      "epoch": 0.8064980116761147,
+      "grad_norm": 0.20166616141796112,
+      "learning_rate": 0.001,
+      "loss": 1.8073,
+      "step": 19064
+    },
+    {
+      "epoch": 0.8065403164396311,
+      "grad_norm": 0.19779857993125916,
+      "learning_rate": 0.001,
+      "loss": 2.8934,
+      "step": 19065
+    },
+    {
+      "epoch": 0.8065826212031475,
+      "grad_norm": 0.15480828285217285,
+      "learning_rate": 0.001,
+      "loss": 2.1785,
+      "step": 19066
+    },
+    {
+      "epoch": 0.8066249259666638,
+      "grad_norm": 0.5121575593948364,
+      "learning_rate": 0.001,
+      "loss": 2.1465,
+      "step": 19067
+    },
+    {
+      "epoch": 0.8066672307301802,
+      "grad_norm": 0.25678855180740356,
+      "learning_rate": 0.001,
+      "loss": 2.5373,
+      "step": 19068
+    },
+    {
+      "epoch": 0.8067095354936966,
+      "grad_norm": 0.1567380726337433,
+      "learning_rate": 0.001,
+      "loss": 1.5213,
+      "step": 19069
+    },
+    {
+      "epoch": 0.8067518402572129,
+      "grad_norm": 0.15887407958507538,
+      "learning_rate": 0.001,
+      "loss": 2.0361,
+      "step": 19070
+    },
+    {
+      "epoch": 0.8067941450207293,
+      "grad_norm": 0.2849603593349457,
+      "learning_rate": 0.001,
+      "loss": 2.3657,
+      "step": 19071
+    },
+    {
+      "epoch": 0.8068364497842457,
+      "grad_norm": 0.27255427837371826,
+      "learning_rate": 0.001,
+      "loss": 2.005,
+      "step": 19072
+    },
+    {
+      "epoch": 0.806878754547762,
+      "grad_norm": 1.081209659576416,
+      "learning_rate": 0.001,
+      "loss": 2.7503,
+      "step": 19073
+    },
+    {
+      "epoch": 0.8069210593112784,
+      "grad_norm": 0.20307116210460663,
+      "learning_rate": 0.001,
+      "loss": 2.4047,
+      "step": 19074
+    },
+    {
+      "epoch": 0.8069633640747949,
+      "grad_norm": 0.3475703299045563,
+      "learning_rate": 0.001,
+      "loss": 2.1349,
+      "step": 19075
+    },
+    {
+      "epoch": 0.8070056688383112,
+      "grad_norm": 0.18069538474082947,
+      "learning_rate": 0.001,
+      "loss": 2.8262,
+      "step": 19076
+    },
+    {
+      "epoch": 0.8070479736018276,
+      "grad_norm": 0.16302266716957092,
+      "learning_rate": 0.001,
+      "loss": 2.1926,
+      "step": 19077
+    },
+    {
+      "epoch": 0.807090278365344,
+      "grad_norm": 0.13929545879364014,
+      "learning_rate": 0.001,
+      "loss": 1.7021,
+      "step": 19078
+    },
+    {
+      "epoch": 0.8071325831288603,
+      "grad_norm": 2.2254645824432373,
+      "learning_rate": 0.001,
+      "loss": 2.0774,
+      "step": 19079
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 0.12101297825574875,
+      "learning_rate": 0.001,
+      "loss": 2.1299,
+      "step": 19080
+    },
+    {
+      "epoch": 0.8072171926558931,
+      "grad_norm": 0.15485353767871857,
+      "learning_rate": 0.001,
+      "loss": 1.941,
+      "step": 19081
+    },
+    {
+      "epoch": 0.8072594974194094,
+      "grad_norm": 0.16078881919384003,
+      "learning_rate": 0.001,
+      "loss": 2.2291,
+      "step": 19082
+    },
+    {
+      "epoch": 0.8073018021829258,
+      "grad_norm": 1.5918997526168823,
+      "learning_rate": 0.001,
+      "loss": 3.3536,
+      "step": 19083
+    },
+    {
+      "epoch": 0.8073441069464422,
+      "grad_norm": 0.16375018656253815,
+      "learning_rate": 0.001,
+      "loss": 2.5549,
+      "step": 19084
+    },
+    {
+      "epoch": 0.8073864117099585,
+      "grad_norm": 1.0432242155075073,
+      "learning_rate": 0.001,
+      "loss": 2.6411,
+      "step": 19085
+    },
+    {
+      "epoch": 0.8074287164734749,
+      "grad_norm": 0.1408742517232895,
+      "learning_rate": 0.001,
+      "loss": 1.4977,
+      "step": 19086
+    },
+    {
+      "epoch": 0.8074710212369913,
+      "grad_norm": 0.13570459187030792,
+      "learning_rate": 0.001,
+      "loss": 3.3056,
+      "step": 19087
+    },
+    {
+      "epoch": 0.8075133260005076,
+      "grad_norm": 0.22087857127189636,
+      "learning_rate": 0.001,
+      "loss": 2.2098,
+      "step": 19088
+    },
+    {
+      "epoch": 0.807555630764024,
+      "grad_norm": 1.198239803314209,
+      "learning_rate": 0.001,
+      "loss": 3.1157,
+      "step": 19089
+    },
+    {
+      "epoch": 0.8075979355275404,
+      "grad_norm": 0.1383177936077118,
+      "learning_rate": 0.001,
+      "loss": 2.3563,
+      "step": 19090
+    },
+    {
+      "epoch": 0.8076402402910567,
+      "grad_norm": 0.14231859147548676,
+      "learning_rate": 0.001,
+      "loss": 1.863,
+      "step": 19091
+    },
+    {
+      "epoch": 0.8076825450545732,
+      "grad_norm": 0.17325039207935333,
+      "learning_rate": 0.001,
+      "loss": 3.1426,
+      "step": 19092
+    },
+    {
+      "epoch": 0.8077248498180896,
+      "grad_norm": 0.14337441325187683,
+      "learning_rate": 0.001,
+      "loss": 1.6843,
+      "step": 19093
+    },
+    {
+      "epoch": 0.8077671545816059,
+      "grad_norm": 0.15518362820148468,
+      "learning_rate": 0.001,
+      "loss": 2.3449,
+      "step": 19094
+    },
+    {
+      "epoch": 0.8078094593451223,
+      "grad_norm": 0.13496747612953186,
+      "learning_rate": 0.001,
+      "loss": 2.6607,
+      "step": 19095
+    },
+    {
+      "epoch": 0.8078517641086387,
+      "grad_norm": 0.17812597751617432,
+      "learning_rate": 0.001,
+      "loss": 2.6483,
+      "step": 19096
+    },
+    {
+      "epoch": 0.807894068872155,
+      "grad_norm": 0.24298779666423798,
+      "learning_rate": 0.001,
+      "loss": 1.6009,
+      "step": 19097
+    },
+    {
+      "epoch": 0.8079363736356714,
+      "grad_norm": 0.13833343982696533,
+      "learning_rate": 0.001,
+      "loss": 1.7103,
+      "step": 19098
+    },
+    {
+      "epoch": 0.8079786783991878,
+      "grad_norm": 0.17393560707569122,
+      "learning_rate": 0.001,
+      "loss": 3.2679,
+      "step": 19099
+    },
+    {
+      "epoch": 0.8080209831627041,
+      "grad_norm": 0.1347343772649765,
+      "learning_rate": 0.001,
+      "loss": 1.3848,
+      "step": 19100
+    },
+    {
+      "epoch": 0.8080632879262205,
+      "grad_norm": 0.3251754641532898,
+      "learning_rate": 0.001,
+      "loss": 2.6169,
+      "step": 19101
+    },
+    {
+      "epoch": 0.8081055926897369,
+      "grad_norm": 0.13301436603069305,
+      "learning_rate": 0.001,
+      "loss": 1.8152,
+      "step": 19102
+    },
+    {
+      "epoch": 0.8081478974532532,
+      "grad_norm": 0.35224834084510803,
+      "learning_rate": 0.001,
+      "loss": 1.9909,
+      "step": 19103
+    },
+    {
+      "epoch": 0.8081902022167696,
+      "grad_norm": 0.2184954732656479,
+      "learning_rate": 0.001,
+      "loss": 2.3797,
+      "step": 19104
+    },
+    {
+      "epoch": 0.8082325069802859,
+      "grad_norm": 0.14901390671730042,
+      "learning_rate": 0.001,
+      "loss": 2.0435,
+      "step": 19105
+    },
+    {
+      "epoch": 0.8082748117438023,
+      "grad_norm": 0.1670936793088913,
+      "learning_rate": 0.001,
+      "loss": 1.9192,
+      "step": 19106
+    },
+    {
+      "epoch": 0.8083171165073187,
+      "grad_norm": 0.1560143530368805,
+      "learning_rate": 0.001,
+      "loss": 1.6413,
+      "step": 19107
+    },
+    {
+      "epoch": 0.808359421270835,
+      "grad_norm": 0.14152587950229645,
+      "learning_rate": 0.001,
+      "loss": 1.7786,
+      "step": 19108
+    },
+    {
+      "epoch": 0.8084017260343515,
+      "grad_norm": 0.2246204912662506,
+      "learning_rate": 0.001,
+      "loss": 2.0439,
+      "step": 19109
+    },
+    {
+      "epoch": 0.8084440307978679,
+      "grad_norm": 0.17294353246688843,
+      "learning_rate": 0.001,
+      "loss": 2.3991,
+      "step": 19110
+    },
+    {
+      "epoch": 0.8084863355613842,
+      "grad_norm": 0.14539849758148193,
+      "learning_rate": 0.001,
+      "loss": 1.9719,
+      "step": 19111
+    },
+    {
+      "epoch": 0.8085286403249006,
+      "grad_norm": 0.16276989877223969,
+      "learning_rate": 0.001,
+      "loss": 1.7146,
+      "step": 19112
+    },
+    {
+      "epoch": 0.808570945088417,
+      "grad_norm": 9.473787307739258,
+      "learning_rate": 0.001,
+      "loss": 2.1896,
+      "step": 19113
+    },
+    {
+      "epoch": 0.8086132498519333,
+      "grad_norm": 0.1611110121011734,
+      "learning_rate": 0.001,
+      "loss": 1.8925,
+      "step": 19114
+    },
+    {
+      "epoch": 0.8086555546154497,
+      "grad_norm": 0.14507554471492767,
+      "learning_rate": 0.001,
+      "loss": 1.5477,
+      "step": 19115
+    },
+    {
+      "epoch": 0.8086978593789661,
+      "grad_norm": 0.17105506360530853,
+      "learning_rate": 0.001,
+      "loss": 2.5952,
+      "step": 19116
+    },
+    {
+      "epoch": 0.8087401641424824,
+      "grad_norm": 0.563596248626709,
+      "learning_rate": 0.001,
+      "loss": 2.5732,
+      "step": 19117
+    },
+    {
+      "epoch": 0.8087824689059988,
+      "grad_norm": 0.2284759283065796,
+      "learning_rate": 0.001,
+      "loss": 3.6089,
+      "step": 19118
+    },
+    {
+      "epoch": 0.8088247736695152,
+      "grad_norm": 0.2556639611721039,
+      "learning_rate": 0.001,
+      "loss": 1.9265,
+      "step": 19119
+    },
+    {
+      "epoch": 0.8088670784330315,
+      "grad_norm": 0.1958867907524109,
+      "learning_rate": 0.001,
+      "loss": 2.395,
+      "step": 19120
+    },
+    {
+      "epoch": 0.8089093831965479,
+      "grad_norm": 0.22015972435474396,
+      "learning_rate": 0.001,
+      "loss": 2.0847,
+      "step": 19121
+    },
+    {
+      "epoch": 0.8089516879600643,
+      "grad_norm": 0.17518767714500427,
+      "learning_rate": 0.001,
+      "loss": 2.8718,
+      "step": 19122
+    },
+    {
+      "epoch": 0.8089939927235806,
+      "grad_norm": 6.914678573608398,
+      "learning_rate": 0.001,
+      "loss": 1.9315,
+      "step": 19123
+    },
+    {
+      "epoch": 0.809036297487097,
+      "grad_norm": 1.030540108680725,
+      "learning_rate": 0.001,
+      "loss": 2.9412,
+      "step": 19124
+    },
+    {
+      "epoch": 0.8090786022506135,
+      "grad_norm": 28.061410903930664,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 19125
+    },
+    {
+      "epoch": 0.8091209070141298,
+      "grad_norm": 0.18755550682544708,
+      "learning_rate": 0.001,
+      "loss": 2.0356,
+      "step": 19126
+    },
+    {
+      "epoch": 0.8091632117776462,
+      "grad_norm": 0.15464842319488525,
+      "learning_rate": 0.001,
+      "loss": 1.7849,
+      "step": 19127
+    },
+    {
+      "epoch": 0.8092055165411626,
+      "grad_norm": 0.20664550364017487,
+      "learning_rate": 0.001,
+      "loss": 3.0477,
+      "step": 19128
+    },
+    {
+      "epoch": 0.8092478213046789,
+      "grad_norm": 0.19141417741775513,
+      "learning_rate": 0.001,
+      "loss": 2.0378,
+      "step": 19129
+    },
+    {
+      "epoch": 0.8092901260681953,
+      "grad_norm": 0.7262970805168152,
+      "learning_rate": 0.001,
+      "loss": 2.168,
+      "step": 19130
+    },
+    {
+      "epoch": 0.8093324308317117,
+      "grad_norm": 0.18341998755931854,
+      "learning_rate": 0.001,
+      "loss": 2.529,
+      "step": 19131
+    },
+    {
+      "epoch": 0.809374735595228,
+      "grad_norm": 0.16619554162025452,
+      "learning_rate": 0.001,
+      "loss": 1.9075,
+      "step": 19132
+    },
+    {
+      "epoch": 0.8094170403587444,
+      "grad_norm": 0.16377653181552887,
+      "learning_rate": 0.001,
+      "loss": 2.7905,
+      "step": 19133
+    },
+    {
+      "epoch": 0.8094593451222608,
+      "grad_norm": 0.15902559459209442,
+      "learning_rate": 0.001,
+      "loss": 2.0185,
+      "step": 19134
+    },
+    {
+      "epoch": 0.8095016498857771,
+      "grad_norm": 0.1559763252735138,
+      "learning_rate": 0.001,
+      "loss": 2.0704,
+      "step": 19135
+    },
+    {
+      "epoch": 0.8095439546492935,
+      "grad_norm": 0.1679486483335495,
+      "learning_rate": 0.001,
+      "loss": 2.4996,
+      "step": 19136
+    },
+    {
+      "epoch": 0.8095862594128099,
+      "grad_norm": 0.15577644109725952,
+      "learning_rate": 0.001,
+      "loss": 2.4594,
+      "step": 19137
+    },
+    {
+      "epoch": 0.8096285641763262,
+      "grad_norm": 0.14891107380390167,
+      "learning_rate": 0.001,
+      "loss": 1.9317,
+      "step": 19138
+    },
+    {
+      "epoch": 0.8096708689398426,
+      "grad_norm": 0.20740364491939545,
+      "learning_rate": 0.001,
+      "loss": 1.7758,
+      "step": 19139
+    },
+    {
+      "epoch": 0.809713173703359,
+      "grad_norm": 0.17380544543266296,
+      "learning_rate": 0.001,
+      "loss": 1.6315,
+      "step": 19140
+    },
+    {
+      "epoch": 0.8097554784668753,
+      "grad_norm": 0.15323974192142487,
+      "learning_rate": 0.001,
+      "loss": 2.2311,
+      "step": 19141
+    },
+    {
+      "epoch": 0.8097977832303918,
+      "grad_norm": 1.1222220659255981,
+      "learning_rate": 0.001,
+      "loss": 1.8476,
+      "step": 19142
+    },
+    {
+      "epoch": 0.8098400879939082,
+      "grad_norm": 0.17634250223636627,
+      "learning_rate": 0.001,
+      "loss": 2.2164,
+      "step": 19143
+    },
+    {
+      "epoch": 0.8098823927574245,
+      "grad_norm": 0.16410218179225922,
+      "learning_rate": 0.001,
+      "loss": 1.9059,
+      "step": 19144
+    },
+    {
+      "epoch": 0.8099246975209409,
+      "grad_norm": 0.16236481070518494,
+      "learning_rate": 0.001,
+      "loss": 1.8111,
+      "step": 19145
+    },
+    {
+      "epoch": 0.8099670022844573,
+      "grad_norm": 5.717217445373535,
+      "learning_rate": 0.001,
+      "loss": 2.1306,
+      "step": 19146
+    },
+    {
+      "epoch": 0.8100093070479736,
+      "grad_norm": 0.1764141023159027,
+      "learning_rate": 0.001,
+      "loss": 1.4532,
+      "step": 19147
+    },
+    {
+      "epoch": 0.81005161181149,
+      "grad_norm": 0.18373993039131165,
+      "learning_rate": 0.001,
+      "loss": 1.8148,
+      "step": 19148
+    },
+    {
+      "epoch": 0.8100939165750064,
+      "grad_norm": 0.1714664250612259,
+      "learning_rate": 0.001,
+      "loss": 2.0202,
+      "step": 19149
+    },
+    {
+      "epoch": 0.8101362213385227,
+      "grad_norm": 0.22828002274036407,
+      "learning_rate": 0.001,
+      "loss": 1.8922,
+      "step": 19150
+    },
+    {
+      "epoch": 0.8101785261020391,
+      "grad_norm": 0.37784022092819214,
+      "learning_rate": 0.001,
+      "loss": 3.023,
+      "step": 19151
+    },
+    {
+      "epoch": 0.8102208308655554,
+      "grad_norm": 0.17360255122184753,
+      "learning_rate": 0.001,
+      "loss": 2.1345,
+      "step": 19152
+    },
+    {
+      "epoch": 0.8102631356290718,
+      "grad_norm": 0.16689980030059814,
+      "learning_rate": 0.001,
+      "loss": 2.1847,
+      "step": 19153
+    },
+    {
+      "epoch": 0.8103054403925882,
+      "grad_norm": 0.23073500394821167,
+      "learning_rate": 0.001,
+      "loss": 2.0032,
+      "step": 19154
+    },
+    {
+      "epoch": 0.8103477451561045,
+      "grad_norm": 0.1540898084640503,
+      "learning_rate": 0.001,
+      "loss": 1.4885,
+      "step": 19155
+    },
+    {
+      "epoch": 0.8103900499196209,
+      "grad_norm": 2.4491209983825684,
+      "learning_rate": 0.001,
+      "loss": 2.0554,
+      "step": 19156
+    },
+    {
+      "epoch": 0.8104323546831373,
+      "grad_norm": 0.191018208861351,
+      "learning_rate": 0.001,
+      "loss": 2.0957,
+      "step": 19157
+    },
+    {
+      "epoch": 0.8104746594466536,
+      "grad_norm": 4.3804240226745605,
+      "learning_rate": 0.001,
+      "loss": 2.0839,
+      "step": 19158
+    },
+    {
+      "epoch": 0.81051696421017,
+      "grad_norm": 0.41173967719078064,
+      "learning_rate": 0.001,
+      "loss": 2.0613,
+      "step": 19159
+    },
+    {
+      "epoch": 0.8105592689736865,
+      "grad_norm": 0.13415801525115967,
+      "learning_rate": 0.001,
+      "loss": 1.7669,
+      "step": 19160
+    },
+    {
+      "epoch": 0.8106015737372028,
+      "grad_norm": 0.15621332824230194,
+      "learning_rate": 0.001,
+      "loss": 2.5883,
+      "step": 19161
+    },
+    {
+      "epoch": 0.8106438785007192,
+      "grad_norm": 2.067545175552368,
+      "learning_rate": 0.001,
+      "loss": 2.3052,
+      "step": 19162
+    },
+    {
+      "epoch": 0.8106861832642356,
+      "grad_norm": 0.15664492547512054,
+      "learning_rate": 0.001,
+      "loss": 2.0756,
+      "step": 19163
+    },
+    {
+      "epoch": 0.8107284880277519,
+      "grad_norm": 0.32015523314476013,
+      "learning_rate": 0.001,
+      "loss": 1.449,
+      "step": 19164
+    },
+    {
+      "epoch": 0.8107707927912683,
+      "grad_norm": 0.18561597168445587,
+      "learning_rate": 0.001,
+      "loss": 3.2036,
+      "step": 19165
+    },
+    {
+      "epoch": 0.8108130975547847,
+      "grad_norm": 0.1604049801826477,
+      "learning_rate": 0.001,
+      "loss": 2.1215,
+      "step": 19166
+    },
+    {
+      "epoch": 0.810855402318301,
+      "grad_norm": 0.13937008380889893,
+      "learning_rate": 0.001,
+      "loss": 2.1111,
+      "step": 19167
+    },
+    {
+      "epoch": 0.8108977070818174,
+      "grad_norm": 0.1511482447385788,
+      "learning_rate": 0.001,
+      "loss": 2.4686,
+      "step": 19168
+    },
+    {
+      "epoch": 0.8109400118453338,
+      "grad_norm": 0.4182822108268738,
+      "learning_rate": 0.001,
+      "loss": 2.0505,
+      "step": 19169
+    },
+    {
+      "epoch": 0.8109823166088501,
+      "grad_norm": 0.17562200129032135,
+      "learning_rate": 0.001,
+      "loss": 3.0039,
+      "step": 19170
+    },
+    {
+      "epoch": 0.8110246213723665,
+      "grad_norm": 0.14364513754844666,
+      "learning_rate": 0.001,
+      "loss": 1.9338,
+      "step": 19171
+    },
+    {
+      "epoch": 0.8110669261358829,
+      "grad_norm": 1.0508071184158325,
+      "learning_rate": 0.001,
+      "loss": 2.679,
+      "step": 19172
+    },
+    {
+      "epoch": 0.8111092308993992,
+      "grad_norm": 0.14670781791210175,
+      "learning_rate": 0.001,
+      "loss": 1.7561,
+      "step": 19173
+    },
+    {
+      "epoch": 0.8111515356629156,
+      "grad_norm": 0.12752504646778107,
+      "learning_rate": 0.001,
+      "loss": 1.1323,
+      "step": 19174
+    },
+    {
+      "epoch": 0.8111938404264321,
+      "grad_norm": 0.13774631917476654,
+      "learning_rate": 0.001,
+      "loss": 2.117,
+      "step": 19175
+    },
+    {
+      "epoch": 0.8112361451899484,
+      "grad_norm": 0.31936532258987427,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 19176
+    },
+    {
+      "epoch": 0.8112784499534648,
+      "grad_norm": 0.3710169196128845,
+      "learning_rate": 0.001,
+      "loss": 3.5819,
+      "step": 19177
+    },
+    {
+      "epoch": 0.8113207547169812,
+      "grad_norm": 0.1603461056947708,
+      "learning_rate": 0.001,
+      "loss": 2.0246,
+      "step": 19178
+    },
+    {
+      "epoch": 0.8113630594804975,
+      "grad_norm": 0.18489456176757812,
+      "learning_rate": 0.001,
+      "loss": 2.6565,
+      "step": 19179
+    },
+    {
+      "epoch": 0.8114053642440139,
+      "grad_norm": 0.16004648804664612,
+      "learning_rate": 0.001,
+      "loss": 1.6997,
+      "step": 19180
+    },
+    {
+      "epoch": 0.8114476690075303,
+      "grad_norm": 0.1679944545030594,
+      "learning_rate": 0.001,
+      "loss": 3.4693,
+      "step": 19181
+    },
+    {
+      "epoch": 0.8114899737710466,
+      "grad_norm": 0.1983286440372467,
+      "learning_rate": 0.001,
+      "loss": 1.9687,
+      "step": 19182
+    },
+    {
+      "epoch": 0.811532278534563,
+      "grad_norm": 0.5275279879570007,
+      "learning_rate": 0.001,
+      "loss": 1.5419,
+      "step": 19183
+    },
+    {
+      "epoch": 0.8115745832980794,
+      "grad_norm": 0.15500518679618835,
+      "learning_rate": 0.001,
+      "loss": 2.7367,
+      "step": 19184
+    },
+    {
+      "epoch": 0.8116168880615957,
+      "grad_norm": 0.16536766290664673,
+      "learning_rate": 0.001,
+      "loss": 2.1103,
+      "step": 19185
+    },
+    {
+      "epoch": 0.8116591928251121,
+      "grad_norm": 0.17966187000274658,
+      "learning_rate": 0.001,
+      "loss": 2.0974,
+      "step": 19186
+    },
+    {
+      "epoch": 0.8117014975886285,
+      "grad_norm": 4.907825946807861,
+      "learning_rate": 0.001,
+      "loss": 3.4044,
+      "step": 19187
+    },
+    {
+      "epoch": 0.8117438023521448,
+      "grad_norm": 0.18678595125675201,
+      "learning_rate": 0.001,
+      "loss": 2.5199,
+      "step": 19188
+    },
+    {
+      "epoch": 0.8117861071156612,
+      "grad_norm": 0.751160740852356,
+      "learning_rate": 0.001,
+      "loss": 2.3407,
+      "step": 19189
+    },
+    {
+      "epoch": 0.8118284118791776,
+      "grad_norm": 0.1894712597131729,
+      "learning_rate": 0.001,
+      "loss": 2.066,
+      "step": 19190
+    },
+    {
+      "epoch": 0.811870716642694,
+      "grad_norm": 0.174768328666687,
+      "learning_rate": 0.001,
+      "loss": 2.0651,
+      "step": 19191
+    },
+    {
+      "epoch": 0.8119130214062104,
+      "grad_norm": 0.17409004271030426,
+      "learning_rate": 0.001,
+      "loss": 2.4093,
+      "step": 19192
+    },
+    {
+      "epoch": 0.8119553261697268,
+      "grad_norm": 0.154340922832489,
+      "learning_rate": 0.001,
+      "loss": 2.1752,
+      "step": 19193
+    },
+    {
+      "epoch": 0.8119976309332431,
+      "grad_norm": 0.18005342781543732,
+      "learning_rate": 0.001,
+      "loss": 2.8527,
+      "step": 19194
+    },
+    {
+      "epoch": 0.8120399356967595,
+      "grad_norm": 0.1886720359325409,
+      "learning_rate": 0.001,
+      "loss": 2.052,
+      "step": 19195
+    },
+    {
+      "epoch": 0.8120822404602758,
+      "grad_norm": 0.15437401831150055,
+      "learning_rate": 0.001,
+      "loss": 2.0393,
+      "step": 19196
+    },
+    {
+      "epoch": 0.8121245452237922,
+      "grad_norm": 0.8311869502067566,
+      "learning_rate": 0.001,
+      "loss": 2.9646,
+      "step": 19197
+    },
+    {
+      "epoch": 0.8121668499873086,
+      "grad_norm": 0.26318734884262085,
+      "learning_rate": 0.001,
+      "loss": 3.0618,
+      "step": 19198
+    },
+    {
+      "epoch": 0.8122091547508249,
+      "grad_norm": 0.15267373621463776,
+      "learning_rate": 0.001,
+      "loss": 1.8216,
+      "step": 19199
+    },
+    {
+      "epoch": 0.8122514595143413,
+      "grad_norm": 0.20256157219409943,
+      "learning_rate": 0.001,
+      "loss": 2.2829,
+      "step": 19200
+    },
+    {
+      "epoch": 0.8122937642778577,
+      "grad_norm": 0.9764918088912964,
+      "learning_rate": 0.001,
+      "loss": 2.902,
+      "step": 19201
+    },
+    {
+      "epoch": 0.812336069041374,
+      "grad_norm": 0.5934033393859863,
+      "learning_rate": 0.001,
+      "loss": 2.9391,
+      "step": 19202
+    },
+    {
+      "epoch": 0.8123783738048904,
+      "grad_norm": 0.41964349150657654,
+      "learning_rate": 0.001,
+      "loss": 1.6686,
+      "step": 19203
+    },
+    {
+      "epoch": 0.8124206785684068,
+      "grad_norm": 0.15301333367824554,
+      "learning_rate": 0.001,
+      "loss": 2.829,
+      "step": 19204
+    },
+    {
+      "epoch": 0.8124629833319231,
+      "grad_norm": 0.15774710476398468,
+      "learning_rate": 0.001,
+      "loss": 2.467,
+      "step": 19205
+    },
+    {
+      "epoch": 0.8125052880954395,
+      "grad_norm": 0.3927057683467865,
+      "learning_rate": 0.001,
+      "loss": 2.3201,
+      "step": 19206
+    },
+    {
+      "epoch": 0.812547592858956,
+      "grad_norm": 4.193830490112305,
+      "learning_rate": 0.001,
+      "loss": 1.9153,
+      "step": 19207
+    },
+    {
+      "epoch": 0.8125898976224722,
+      "grad_norm": 0.16809159517288208,
+      "learning_rate": 0.001,
+      "loss": 2.6222,
+      "step": 19208
+    },
+    {
+      "epoch": 0.8126322023859887,
+      "grad_norm": 0.36612462997436523,
+      "learning_rate": 0.001,
+      "loss": 1.8079,
+      "step": 19209
+    },
+    {
+      "epoch": 0.8126745071495051,
+      "grad_norm": 0.17396114766597748,
+      "learning_rate": 0.001,
+      "loss": 1.7655,
+      "step": 19210
+    },
+    {
+      "epoch": 0.8127168119130214,
+      "grad_norm": 0.14558352530002594,
+      "learning_rate": 0.001,
+      "loss": 2.35,
+      "step": 19211
+    },
+    {
+      "epoch": 0.8127591166765378,
+      "grad_norm": 0.1535443514585495,
+      "learning_rate": 0.001,
+      "loss": 1.8861,
+      "step": 19212
+    },
+    {
+      "epoch": 0.8128014214400542,
+      "grad_norm": 0.14222368597984314,
+      "learning_rate": 0.001,
+      "loss": 1.6644,
+      "step": 19213
+    },
+    {
+      "epoch": 0.8128437262035705,
+      "grad_norm": 0.16522639989852905,
+      "learning_rate": 0.001,
+      "loss": 2.5941,
+      "step": 19214
+    },
+    {
+      "epoch": 0.8128860309670869,
+      "grad_norm": 1.800579309463501,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 19215
+    },
+    {
+      "epoch": 0.8129283357306033,
+      "grad_norm": 4.028626918792725,
+      "learning_rate": 0.001,
+      "loss": 1.5658,
+      "step": 19216
+    },
+    {
+      "epoch": 0.8129706404941196,
+      "grad_norm": 0.15518246591091156,
+      "learning_rate": 0.001,
+      "loss": 1.8562,
+      "step": 19217
+    },
+    {
+      "epoch": 0.813012945257636,
+      "grad_norm": 1.1874980926513672,
+      "learning_rate": 0.001,
+      "loss": 2.0366,
+      "step": 19218
+    },
+    {
+      "epoch": 0.8130552500211524,
+      "grad_norm": 0.16775746643543243,
+      "learning_rate": 0.001,
+      "loss": 1.7891,
+      "step": 19219
+    },
+    {
+      "epoch": 0.8130975547846687,
+      "grad_norm": 0.30443984270095825,
+      "learning_rate": 0.001,
+      "loss": 3.444,
+      "step": 19220
+    },
+    {
+      "epoch": 0.8131398595481851,
+      "grad_norm": 0.1578860729932785,
+      "learning_rate": 0.001,
+      "loss": 1.8633,
+      "step": 19221
+    },
+    {
+      "epoch": 0.8131821643117015,
+      "grad_norm": 17.17110252380371,
+      "learning_rate": 0.001,
+      "loss": 2.2716,
+      "step": 19222
+    },
+    {
+      "epoch": 0.8132244690752178,
+      "grad_norm": 0.16925646364688873,
+      "learning_rate": 0.001,
+      "loss": 2.1797,
+      "step": 19223
+    },
+    {
+      "epoch": 0.8132667738387342,
+      "grad_norm": 0.46783486008644104,
+      "learning_rate": 0.001,
+      "loss": 1.9686,
+      "step": 19224
+    },
+    {
+      "epoch": 0.8133090786022507,
+      "grad_norm": 0.21370871365070343,
+      "learning_rate": 0.001,
+      "loss": 2.3163,
+      "step": 19225
+    },
+    {
+      "epoch": 0.813351383365767,
+      "grad_norm": 2.3018229007720947,
+      "learning_rate": 0.001,
+      "loss": 3.1378,
+      "step": 19226
+    },
+    {
+      "epoch": 0.8133936881292834,
+      "grad_norm": 0.26849448680877686,
+      "learning_rate": 0.001,
+      "loss": 2.8301,
+      "step": 19227
+    },
+    {
+      "epoch": 0.8134359928927998,
+      "grad_norm": 0.36854100227355957,
+      "learning_rate": 0.001,
+      "loss": 2.2498,
+      "step": 19228
+    },
+    {
+      "epoch": 0.8134782976563161,
+      "grad_norm": 0.44008371233940125,
+      "learning_rate": 0.001,
+      "loss": 2.2804,
+      "step": 19229
+    },
+    {
+      "epoch": 0.8135206024198325,
+      "grad_norm": 0.2091284990310669,
+      "learning_rate": 0.001,
+      "loss": 2.2224,
+      "step": 19230
+    },
+    {
+      "epoch": 0.8135629071833489,
+      "grad_norm": 0.2483440637588501,
+      "learning_rate": 0.001,
+      "loss": 2.2146,
+      "step": 19231
+    },
+    {
+      "epoch": 0.8136052119468652,
+      "grad_norm": 0.19964353740215302,
+      "learning_rate": 0.001,
+      "loss": 1.861,
+      "step": 19232
+    },
+    {
+      "epoch": 0.8136475167103816,
+      "grad_norm": 0.22766467928886414,
+      "learning_rate": 0.001,
+      "loss": 2.3613,
+      "step": 19233
+    },
+    {
+      "epoch": 0.813689821473898,
+      "grad_norm": 0.20995555818080902,
+      "learning_rate": 0.001,
+      "loss": 2.4426,
+      "step": 19234
+    },
+    {
+      "epoch": 0.8137321262374143,
+      "grad_norm": 0.16860271990299225,
+      "learning_rate": 0.001,
+      "loss": 2.3357,
+      "step": 19235
+    },
+    {
+      "epoch": 0.8137744310009307,
+      "grad_norm": 0.18488185107707977,
+      "learning_rate": 0.001,
+      "loss": 2.7789,
+      "step": 19236
+    },
+    {
+      "epoch": 0.8138167357644471,
+      "grad_norm": 2.0512919425964355,
+      "learning_rate": 0.001,
+      "loss": 2.3825,
+      "step": 19237
+    },
+    {
+      "epoch": 0.8138590405279634,
+      "grad_norm": 0.39424288272857666,
+      "learning_rate": 0.001,
+      "loss": 2.3707,
+      "step": 19238
+    },
+    {
+      "epoch": 0.8139013452914798,
+      "grad_norm": 0.18912164866924286,
+      "learning_rate": 0.001,
+      "loss": 2.5157,
+      "step": 19239
+    },
+    {
+      "epoch": 0.8139436500549961,
+      "grad_norm": 0.17487472295761108,
+      "learning_rate": 0.001,
+      "loss": 2.0539,
+      "step": 19240
+    },
+    {
+      "epoch": 0.8139859548185125,
+      "grad_norm": 0.1670231968164444,
+      "learning_rate": 0.001,
+      "loss": 2.7521,
+      "step": 19241
+    },
+    {
+      "epoch": 0.814028259582029,
+      "grad_norm": 0.16060633957386017,
+      "learning_rate": 0.001,
+      "loss": 1.8548,
+      "step": 19242
+    },
+    {
+      "epoch": 0.8140705643455453,
+      "grad_norm": 0.1655425727367401,
+      "learning_rate": 0.001,
+      "loss": 2.6851,
+      "step": 19243
+    },
+    {
+      "epoch": 0.8141128691090617,
+      "grad_norm": 0.7323225736618042,
+      "learning_rate": 0.001,
+      "loss": 3.2264,
+      "step": 19244
+    },
+    {
+      "epoch": 0.8141551738725781,
+      "grad_norm": 0.16756169497966766,
+      "learning_rate": 0.001,
+      "loss": 2.0877,
+      "step": 19245
+    },
+    {
+      "epoch": 0.8141974786360944,
+      "grad_norm": 4.373884201049805,
+      "learning_rate": 0.001,
+      "loss": 3.0818,
+      "step": 19246
+    },
+    {
+      "epoch": 0.8142397833996108,
+      "grad_norm": 0.15877367556095123,
+      "learning_rate": 0.001,
+      "loss": 2.477,
+      "step": 19247
+    },
+    {
+      "epoch": 0.8142820881631272,
+      "grad_norm": 14.3695707321167,
+      "learning_rate": 0.001,
+      "loss": 3.6114,
+      "step": 19248
+    },
+    {
+      "epoch": 0.8143243929266435,
+      "grad_norm": 0.35859042406082153,
+      "learning_rate": 0.001,
+      "loss": 2.4061,
+      "step": 19249
+    },
+    {
+      "epoch": 0.8143666976901599,
+      "grad_norm": 0.17189420759677887,
+      "learning_rate": 0.001,
+      "loss": 2.6325,
+      "step": 19250
+    },
+    {
+      "epoch": 0.8144090024536763,
+      "grad_norm": 0.1583157330751419,
+      "learning_rate": 0.001,
+      "loss": 2.0351,
+      "step": 19251
+    },
+    {
+      "epoch": 0.8144513072171926,
+      "grad_norm": 0.1763736605644226,
+      "learning_rate": 0.001,
+      "loss": 1.9754,
+      "step": 19252
+    },
+    {
+      "epoch": 0.814493611980709,
+      "grad_norm": 0.16979268193244934,
+      "learning_rate": 0.001,
+      "loss": 2.0023,
+      "step": 19253
+    },
+    {
+      "epoch": 0.8145359167442254,
+      "grad_norm": 0.1730976700782776,
+      "learning_rate": 0.001,
+      "loss": 2.8671,
+      "step": 19254
+    },
+    {
+      "epoch": 0.8145782215077417,
+      "grad_norm": 0.17058543860912323,
+      "learning_rate": 0.001,
+      "loss": 1.4882,
+      "step": 19255
+    },
+    {
+      "epoch": 0.8146205262712581,
+      "grad_norm": 0.29904067516326904,
+      "learning_rate": 0.001,
+      "loss": 2.019,
+      "step": 19256
+    },
+    {
+      "epoch": 0.8146628310347745,
+      "grad_norm": 0.6264715194702148,
+      "learning_rate": 0.001,
+      "loss": 2.8114,
+      "step": 19257
+    },
+    {
+      "epoch": 0.8147051357982908,
+      "grad_norm": 0.14027170836925507,
+      "learning_rate": 0.001,
+      "loss": 3.2401,
+      "step": 19258
+    },
+    {
+      "epoch": 0.8147474405618073,
+      "grad_norm": 0.16307754814624786,
+      "learning_rate": 0.001,
+      "loss": 2.6808,
+      "step": 19259
+    },
+    {
+      "epoch": 0.8147897453253237,
+      "grad_norm": 0.9267063736915588,
+      "learning_rate": 0.001,
+      "loss": 2.3898,
+      "step": 19260
+    },
+    {
+      "epoch": 0.81483205008884,
+      "grad_norm": 0.1487603634595871,
+      "learning_rate": 0.001,
+      "loss": 2.9254,
+      "step": 19261
+    },
+    {
+      "epoch": 0.8148743548523564,
+      "grad_norm": 0.1394791156053543,
+      "learning_rate": 0.001,
+      "loss": 1.7672,
+      "step": 19262
+    },
+    {
+      "epoch": 0.8149166596158728,
+      "grad_norm": 45.422454833984375,
+      "learning_rate": 0.001,
+      "loss": 2.5624,
+      "step": 19263
+    },
+    {
+      "epoch": 0.8149589643793891,
+      "grad_norm": 0.4591314196586609,
+      "learning_rate": 0.001,
+      "loss": 2.3038,
+      "step": 19264
+    },
+    {
+      "epoch": 0.8150012691429055,
+      "grad_norm": 0.13458524644374847,
+      "learning_rate": 0.001,
+      "loss": 2.2434,
+      "step": 19265
+    },
+    {
+      "epoch": 0.8150435739064219,
+      "grad_norm": 0.14545901119709015,
+      "learning_rate": 0.001,
+      "loss": 2.3079,
+      "step": 19266
+    },
+    {
+      "epoch": 0.8150858786699382,
+      "grad_norm": 0.16260690987110138,
+      "learning_rate": 0.001,
+      "loss": 2.7739,
+      "step": 19267
+    },
+    {
+      "epoch": 0.8151281834334546,
+      "grad_norm": 0.15987445414066315,
+      "learning_rate": 0.001,
+      "loss": 1.6845,
+      "step": 19268
+    },
+    {
+      "epoch": 0.815170488196971,
+      "grad_norm": 0.16004207730293274,
+      "learning_rate": 0.001,
+      "loss": 2.0792,
+      "step": 19269
+    },
+    {
+      "epoch": 0.8152127929604873,
+      "grad_norm": 0.14027880132198334,
+      "learning_rate": 0.001,
+      "loss": 2.14,
+      "step": 19270
+    },
+    {
+      "epoch": 0.8152550977240037,
+      "grad_norm": 0.1478835940361023,
+      "learning_rate": 0.001,
+      "loss": 1.4469,
+      "step": 19271
+    },
+    {
+      "epoch": 0.8152974024875201,
+      "grad_norm": 0.6844944357872009,
+      "learning_rate": 0.001,
+      "loss": 2.0617,
+      "step": 19272
+    },
+    {
+      "epoch": 0.8153397072510364,
+      "grad_norm": 0.24570874869823456,
+      "learning_rate": 0.001,
+      "loss": 2.0208,
+      "step": 19273
+    },
+    {
+      "epoch": 0.8153820120145528,
+      "grad_norm": 0.15631866455078125,
+      "learning_rate": 0.001,
+      "loss": 1.826,
+      "step": 19274
+    },
+    {
+      "epoch": 0.8154243167780693,
+      "grad_norm": 0.18372394144535065,
+      "learning_rate": 0.001,
+      "loss": 2.138,
+      "step": 19275
+    },
+    {
+      "epoch": 0.8154666215415856,
+      "grad_norm": 0.6491008400917053,
+      "learning_rate": 0.001,
+      "loss": 3.212,
+      "step": 19276
+    },
+    {
+      "epoch": 0.815508926305102,
+      "grad_norm": 0.13584677875041962,
+      "learning_rate": 0.001,
+      "loss": 2.0065,
+      "step": 19277
+    },
+    {
+      "epoch": 0.8155512310686184,
+      "grad_norm": 0.16441628336906433,
+      "learning_rate": 0.001,
+      "loss": 1.7349,
+      "step": 19278
+    },
+    {
+      "epoch": 0.8155935358321347,
+      "grad_norm": 0.16459910571575165,
+      "learning_rate": 0.001,
+      "loss": 1.9561,
+      "step": 19279
+    },
+    {
+      "epoch": 0.8156358405956511,
+      "grad_norm": 0.8286261558532715,
+      "learning_rate": 0.001,
+      "loss": 1.744,
+      "step": 19280
+    },
+    {
+      "epoch": 0.8156781453591675,
+      "grad_norm": 0.18971623480319977,
+      "learning_rate": 0.001,
+      "loss": 1.968,
+      "step": 19281
+    },
+    {
+      "epoch": 0.8157204501226838,
+      "grad_norm": 0.15174627304077148,
+      "learning_rate": 0.001,
+      "loss": 1.5037,
+      "step": 19282
+    },
+    {
+      "epoch": 0.8157627548862002,
+      "grad_norm": 0.1971951127052307,
+      "learning_rate": 0.001,
+      "loss": 1.9095,
+      "step": 19283
+    },
+    {
+      "epoch": 0.8158050596497166,
+      "grad_norm": 0.6228017807006836,
+      "learning_rate": 0.001,
+      "loss": 3.6166,
+      "step": 19284
+    },
+    {
+      "epoch": 0.8158473644132329,
+      "grad_norm": 0.20216991007328033,
+      "learning_rate": 0.001,
+      "loss": 2.2737,
+      "step": 19285
+    },
+    {
+      "epoch": 0.8158896691767493,
+      "grad_norm": 0.15689660608768463,
+      "learning_rate": 0.001,
+      "loss": 2.5109,
+      "step": 19286
+    },
+    {
+      "epoch": 0.8159319739402656,
+      "grad_norm": 0.16045741736888885,
+      "learning_rate": 0.001,
+      "loss": 2.3619,
+      "step": 19287
+    },
+    {
+      "epoch": 0.815974278703782,
+      "grad_norm": 0.14327400922775269,
+      "learning_rate": 0.001,
+      "loss": 1.7969,
+      "step": 19288
+    },
+    {
+      "epoch": 0.8160165834672984,
+      "grad_norm": 0.16195663809776306,
+      "learning_rate": 0.001,
+      "loss": 1.8645,
+      "step": 19289
+    },
+    {
+      "epoch": 0.8160588882308147,
+      "grad_norm": 0.17734597623348236,
+      "learning_rate": 0.001,
+      "loss": 1.8547,
+      "step": 19290
+    },
+    {
+      "epoch": 0.8161011929943311,
+      "grad_norm": 0.1504170149564743,
+      "learning_rate": 0.001,
+      "loss": 3.0344,
+      "step": 19291
+    },
+    {
+      "epoch": 0.8161434977578476,
+      "grad_norm": 0.3913165032863617,
+      "learning_rate": 0.001,
+      "loss": 2.165,
+      "step": 19292
+    },
+    {
+      "epoch": 0.8161858025213639,
+      "grad_norm": 0.14916856586933136,
+      "learning_rate": 0.001,
+      "loss": 2.5323,
+      "step": 19293
+    },
+    {
+      "epoch": 0.8162281072848803,
+      "grad_norm": 0.18995776772499084,
+      "learning_rate": 0.001,
+      "loss": 3.1768,
+      "step": 19294
+    },
+    {
+      "epoch": 0.8162704120483967,
+      "grad_norm": 0.16553127765655518,
+      "learning_rate": 0.001,
+      "loss": 2.8746,
+      "step": 19295
+    },
+    {
+      "epoch": 0.816312716811913,
+      "grad_norm": 0.16702581942081451,
+      "learning_rate": 0.001,
+      "loss": 1.8839,
+      "step": 19296
+    },
+    {
+      "epoch": 0.8163550215754294,
+      "grad_norm": 0.1927644908428192,
+      "learning_rate": 0.001,
+      "loss": 2.6619,
+      "step": 19297
+    },
+    {
+      "epoch": 0.8163973263389458,
+      "grad_norm": 1.5043902397155762,
+      "learning_rate": 0.001,
+      "loss": 1.8784,
+      "step": 19298
+    },
+    {
+      "epoch": 0.8164396311024621,
+      "grad_norm": 0.15827623009681702,
+      "learning_rate": 0.001,
+      "loss": 1.8496,
+      "step": 19299
+    },
+    {
+      "epoch": 0.8164819358659785,
+      "grad_norm": 0.15658104419708252,
+      "learning_rate": 0.001,
+      "loss": 2.642,
+      "step": 19300
+    },
+    {
+      "epoch": 0.8165242406294949,
+      "grad_norm": 0.2382393330335617,
+      "learning_rate": 0.001,
+      "loss": 2.8102,
+      "step": 19301
+    },
+    {
+      "epoch": 0.8165665453930112,
+      "grad_norm": 0.7768169045448303,
+      "learning_rate": 0.001,
+      "loss": 2.6393,
+      "step": 19302
+    },
+    {
+      "epoch": 0.8166088501565276,
+      "grad_norm": 0.9647676944732666,
+      "learning_rate": 0.001,
+      "loss": 2.4991,
+      "step": 19303
+    },
+    {
+      "epoch": 0.816651154920044,
+      "grad_norm": 0.17201387882232666,
+      "learning_rate": 0.001,
+      "loss": 2.1453,
+      "step": 19304
+    },
+    {
+      "epoch": 0.8166934596835603,
+      "grad_norm": 0.18046537041664124,
+      "learning_rate": 0.001,
+      "loss": 1.9127,
+      "step": 19305
+    },
+    {
+      "epoch": 0.8167357644470767,
+      "grad_norm": 0.1487366259098053,
+      "learning_rate": 0.001,
+      "loss": 2.5597,
+      "step": 19306
+    },
+    {
+      "epoch": 0.8167780692105931,
+      "grad_norm": 0.36740607023239136,
+      "learning_rate": 0.001,
+      "loss": 2.5074,
+      "step": 19307
+    },
+    {
+      "epoch": 0.8168203739741094,
+      "grad_norm": 0.2924293875694275,
+      "learning_rate": 0.001,
+      "loss": 3.529,
+      "step": 19308
+    },
+    {
+      "epoch": 0.8168626787376259,
+      "grad_norm": 0.14903929829597473,
+      "learning_rate": 0.001,
+      "loss": 3.0469,
+      "step": 19309
+    },
+    {
+      "epoch": 0.8169049835011423,
+      "grad_norm": 0.1474577784538269,
+      "learning_rate": 0.001,
+      "loss": 2.6798,
+      "step": 19310
+    },
+    {
+      "epoch": 0.8169472882646586,
+      "grad_norm": 0.14111989736557007,
+      "learning_rate": 0.001,
+      "loss": 2.0767,
+      "step": 19311
+    },
+    {
+      "epoch": 0.816989593028175,
+      "grad_norm": 0.16078026592731476,
+      "learning_rate": 0.001,
+      "loss": 2.0343,
+      "step": 19312
+    },
+    {
+      "epoch": 0.8170318977916914,
+      "grad_norm": 0.14191174507141113,
+      "learning_rate": 0.001,
+      "loss": 1.7657,
+      "step": 19313
+    },
+    {
+      "epoch": 0.8170742025552077,
+      "grad_norm": 0.23834367096424103,
+      "learning_rate": 0.001,
+      "loss": 1.9448,
+      "step": 19314
+    },
+    {
+      "epoch": 0.8171165073187241,
+      "grad_norm": 0.15897387266159058,
+      "learning_rate": 0.001,
+      "loss": 2.4028,
+      "step": 19315
+    },
+    {
+      "epoch": 0.8171588120822405,
+      "grad_norm": 0.14142805337905884,
+      "learning_rate": 0.001,
+      "loss": 2.5824,
+      "step": 19316
+    },
+    {
+      "epoch": 0.8172011168457568,
+      "grad_norm": 0.13879472017288208,
+      "learning_rate": 0.001,
+      "loss": 1.7327,
+      "step": 19317
+    },
+    {
+      "epoch": 0.8172434216092732,
+      "grad_norm": 0.13679929077625275,
+      "learning_rate": 0.001,
+      "loss": 1.8926,
+      "step": 19318
+    },
+    {
+      "epoch": 0.8172857263727896,
+      "grad_norm": 0.13544319570064545,
+      "learning_rate": 0.001,
+      "loss": 3.3204,
+      "step": 19319
+    },
+    {
+      "epoch": 0.8173280311363059,
+      "grad_norm": 0.23128291964530945,
+      "learning_rate": 0.001,
+      "loss": 3.2543,
+      "step": 19320
+    },
+    {
+      "epoch": 0.8173703358998223,
+      "grad_norm": 0.16951590776443481,
+      "learning_rate": 0.001,
+      "loss": 2.3761,
+      "step": 19321
+    },
+    {
+      "epoch": 0.8174126406633387,
+      "grad_norm": 0.13225175440311432,
+      "learning_rate": 0.001,
+      "loss": 3.1764,
+      "step": 19322
+    },
+    {
+      "epoch": 0.817454945426855,
+      "grad_norm": 0.22651726007461548,
+      "learning_rate": 0.001,
+      "loss": 2.3727,
+      "step": 19323
+    },
+    {
+      "epoch": 0.8174972501903714,
+      "grad_norm": 0.14250196516513824,
+      "learning_rate": 0.001,
+      "loss": 2.555,
+      "step": 19324
+    },
+    {
+      "epoch": 0.8175395549538879,
+      "grad_norm": 0.13862445950508118,
+      "learning_rate": 0.001,
+      "loss": 2.1272,
+      "step": 19325
+    },
+    {
+      "epoch": 0.8175818597174042,
+      "grad_norm": 0.6835310459136963,
+      "learning_rate": 0.001,
+      "loss": 2.1929,
+      "step": 19326
+    },
+    {
+      "epoch": 0.8176241644809206,
+      "grad_norm": 0.49331387877464294,
+      "learning_rate": 0.001,
+      "loss": 1.605,
+      "step": 19327
+    },
+    {
+      "epoch": 0.817666469244437,
+      "grad_norm": 0.8337584733963013,
+      "learning_rate": 0.001,
+      "loss": 1.8696,
+      "step": 19328
+    },
+    {
+      "epoch": 0.8177087740079533,
+      "grad_norm": 0.13888110220432281,
+      "learning_rate": 0.001,
+      "loss": 1.7783,
+      "step": 19329
+    },
+    {
+      "epoch": 0.8177510787714697,
+      "grad_norm": 0.15153059363365173,
+      "learning_rate": 0.001,
+      "loss": 3.0953,
+      "step": 19330
+    },
+    {
+      "epoch": 0.817793383534986,
+      "grad_norm": 0.16064806282520294,
+      "learning_rate": 0.001,
+      "loss": 2.2135,
+      "step": 19331
+    },
+    {
+      "epoch": 0.8178356882985024,
+      "grad_norm": 0.18726807832717896,
+      "learning_rate": 0.001,
+      "loss": 1.5846,
+      "step": 19332
+    },
+    {
+      "epoch": 0.8178779930620188,
+      "grad_norm": 1.7091985940933228,
+      "learning_rate": 0.001,
+      "loss": 1.8989,
+      "step": 19333
+    },
+    {
+      "epoch": 0.8179202978255351,
+      "grad_norm": 10.048054695129395,
+      "learning_rate": 0.001,
+      "loss": 1.9446,
+      "step": 19334
+    },
+    {
+      "epoch": 0.8179626025890515,
+      "grad_norm": 0.1563667356967926,
+      "learning_rate": 0.001,
+      "loss": 1.8827,
+      "step": 19335
+    },
+    {
+      "epoch": 0.8180049073525679,
+      "grad_norm": 0.14625267684459686,
+      "learning_rate": 0.001,
+      "loss": 2.9246,
+      "step": 19336
+    },
+    {
+      "epoch": 0.8180472121160842,
+      "grad_norm": 0.14250653982162476,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 19337
+    },
+    {
+      "epoch": 0.8180895168796006,
+      "grad_norm": 0.1519334465265274,
+      "learning_rate": 0.001,
+      "loss": 1.6979,
+      "step": 19338
+    },
+    {
+      "epoch": 0.818131821643117,
+      "grad_norm": 0.18076607584953308,
+      "learning_rate": 0.001,
+      "loss": 1.4037,
+      "step": 19339
+    },
+    {
+      "epoch": 0.8181741264066333,
+      "grad_norm": 0.20953752100467682,
+      "learning_rate": 0.001,
+      "loss": 2.7733,
+      "step": 19340
+    },
+    {
+      "epoch": 0.8182164311701497,
+      "grad_norm": 0.17903457581996918,
+      "learning_rate": 0.001,
+      "loss": 1.9328,
+      "step": 19341
+    },
+    {
+      "epoch": 0.8182587359336662,
+      "grad_norm": 0.1693604737520218,
+      "learning_rate": 0.001,
+      "loss": 1.6344,
+      "step": 19342
+    },
+    {
+      "epoch": 0.8183010406971825,
+      "grad_norm": 0.18389926850795746,
+      "learning_rate": 0.001,
+      "loss": 1.6565,
+      "step": 19343
+    },
+    {
+      "epoch": 0.8183433454606989,
+      "grad_norm": 0.15521776676177979,
+      "learning_rate": 0.001,
+      "loss": 2.1015,
+      "step": 19344
+    },
+    {
+      "epoch": 0.8183856502242153,
+      "grad_norm": 0.1567741185426712,
+      "learning_rate": 0.001,
+      "loss": 2.3034,
+      "step": 19345
+    },
+    {
+      "epoch": 0.8184279549877316,
+      "grad_norm": 0.1486789584159851,
+      "learning_rate": 0.001,
+      "loss": 1.6981,
+      "step": 19346
+    },
+    {
+      "epoch": 0.818470259751248,
+      "grad_norm": 0.14541268348693848,
+      "learning_rate": 0.001,
+      "loss": 2.1189,
+      "step": 19347
+    },
+    {
+      "epoch": 0.8185125645147644,
+      "grad_norm": 0.1486058086156845,
+      "learning_rate": 0.001,
+      "loss": 1.6406,
+      "step": 19348
+    },
+    {
+      "epoch": 0.8185548692782807,
+      "grad_norm": 0.1677182912826538,
+      "learning_rate": 0.001,
+      "loss": 1.6275,
+      "step": 19349
+    },
+    {
+      "epoch": 0.8185971740417971,
+      "grad_norm": 0.1867937445640564,
+      "learning_rate": 0.001,
+      "loss": 2.5879,
+      "step": 19350
+    },
+    {
+      "epoch": 0.8186394788053135,
+      "grad_norm": 0.18571646511554718,
+      "learning_rate": 0.001,
+      "loss": 2.914,
+      "step": 19351
+    },
+    {
+      "epoch": 0.8186817835688298,
+      "grad_norm": 0.17345742881298065,
+      "learning_rate": 0.001,
+      "loss": 2.0374,
+      "step": 19352
+    },
+    {
+      "epoch": 0.8187240883323462,
+      "grad_norm": 0.164658322930336,
+      "learning_rate": 0.001,
+      "loss": 1.6735,
+      "step": 19353
+    },
+    {
+      "epoch": 0.8187663930958626,
+      "grad_norm": 15.121390342712402,
+      "learning_rate": 0.001,
+      "loss": 2.3059,
+      "step": 19354
+    },
+    {
+      "epoch": 0.8188086978593789,
+      "grad_norm": 0.16249455511569977,
+      "learning_rate": 0.001,
+      "loss": 2.387,
+      "step": 19355
+    },
+    {
+      "epoch": 0.8188510026228953,
+      "grad_norm": 0.21848152577877045,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 19356
+    },
+    {
+      "epoch": 0.8188933073864118,
+      "grad_norm": 0.18375152349472046,
+      "learning_rate": 0.001,
+      "loss": 1.7933,
+      "step": 19357
+    },
+    {
+      "epoch": 0.818935612149928,
+      "grad_norm": 0.1748533397912979,
+      "learning_rate": 0.001,
+      "loss": 3.2785,
+      "step": 19358
+    },
+    {
+      "epoch": 0.8189779169134445,
+      "grad_norm": 0.18232855200767517,
+      "learning_rate": 0.001,
+      "loss": 2.2611,
+      "step": 19359
+    },
+    {
+      "epoch": 0.8190202216769609,
+      "grad_norm": 0.811168909072876,
+      "learning_rate": 0.001,
+      "loss": 1.9481,
+      "step": 19360
+    },
+    {
+      "epoch": 0.8190625264404772,
+      "grad_norm": 40.37857437133789,
+      "learning_rate": 0.001,
+      "loss": 1.5613,
+      "step": 19361
+    },
+    {
+      "epoch": 0.8191048312039936,
+      "grad_norm": 1.3089685440063477,
+      "learning_rate": 0.001,
+      "loss": 3.6364,
+      "step": 19362
+    },
+    {
+      "epoch": 0.81914713596751,
+      "grad_norm": 0.1557542085647583,
+      "learning_rate": 0.001,
+      "loss": 2.8861,
+      "step": 19363
+    },
+    {
+      "epoch": 0.8191894407310263,
+      "grad_norm": 0.1572728157043457,
+      "learning_rate": 0.001,
+      "loss": 1.987,
+      "step": 19364
+    },
+    {
+      "epoch": 0.8192317454945427,
+      "grad_norm": 0.13320891559123993,
+      "learning_rate": 0.001,
+      "loss": 2.0874,
+      "step": 19365
+    },
+    {
+      "epoch": 0.8192740502580591,
+      "grad_norm": 0.18693287670612335,
+      "learning_rate": 0.001,
+      "loss": 1.5792,
+      "step": 19366
+    },
+    {
+      "epoch": 0.8193163550215754,
+      "grad_norm": 0.17227143049240112,
+      "learning_rate": 0.001,
+      "loss": 2.8165,
+      "step": 19367
+    },
+    {
+      "epoch": 0.8193586597850918,
+      "grad_norm": 0.19532877206802368,
+      "learning_rate": 0.001,
+      "loss": 2.0586,
+      "step": 19368
+    },
+    {
+      "epoch": 0.8194009645486082,
+      "grad_norm": 0.18054965138435364,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 19369
+    },
+    {
+      "epoch": 0.8194432693121245,
+      "grad_norm": 0.3110874593257904,
+      "learning_rate": 0.001,
+      "loss": 3.9523,
+      "step": 19370
+    },
+    {
+      "epoch": 0.8194855740756409,
+      "grad_norm": 0.2044413536787033,
+      "learning_rate": 0.001,
+      "loss": 2.5399,
+      "step": 19371
+    },
+    {
+      "epoch": 0.8195278788391573,
+      "grad_norm": 0.21336640417575836,
+      "learning_rate": 0.001,
+      "loss": 2.1525,
+      "step": 19372
+    },
+    {
+      "epoch": 0.8195701836026736,
+      "grad_norm": 0.22624215483665466,
+      "learning_rate": 0.001,
+      "loss": 2.2209,
+      "step": 19373
+    },
+    {
+      "epoch": 0.81961248836619,
+      "grad_norm": 0.23245839774608612,
+      "learning_rate": 0.001,
+      "loss": 2.8199,
+      "step": 19374
+    },
+    {
+      "epoch": 0.8196547931297065,
+      "grad_norm": 0.19043493270874023,
+      "learning_rate": 0.001,
+      "loss": 1.8749,
+      "step": 19375
+    },
+    {
+      "epoch": 0.8196970978932228,
+      "grad_norm": 0.16715998947620392,
+      "learning_rate": 0.001,
+      "loss": 3.1058,
+      "step": 19376
+    },
+    {
+      "epoch": 0.8197394026567392,
+      "grad_norm": 0.15719690918922424,
+      "learning_rate": 0.001,
+      "loss": 2.2472,
+      "step": 19377
+    },
+    {
+      "epoch": 0.8197817074202555,
+      "grad_norm": 0.15933258831501007,
+      "learning_rate": 0.001,
+      "loss": 2.5364,
+      "step": 19378
+    },
+    {
+      "epoch": 0.8198240121837719,
+      "grad_norm": 0.29013222455978394,
+      "learning_rate": 0.001,
+      "loss": 3.0306,
+      "step": 19379
+    },
+    {
+      "epoch": 0.8198663169472883,
+      "grad_norm": 0.17771287262439728,
+      "learning_rate": 0.001,
+      "loss": 2.3702,
+      "step": 19380
+    },
+    {
+      "epoch": 0.8199086217108046,
+      "grad_norm": 0.14083783328533173,
+      "learning_rate": 0.001,
+      "loss": 1.8384,
+      "step": 19381
+    },
+    {
+      "epoch": 0.819950926474321,
+      "grad_norm": 0.17068462073802948,
+      "learning_rate": 0.001,
+      "loss": 2.4943,
+      "step": 19382
+    },
+    {
+      "epoch": 0.8199932312378374,
+      "grad_norm": 0.1593535989522934,
+      "learning_rate": 0.001,
+      "loss": 2.0397,
+      "step": 19383
+    },
+    {
+      "epoch": 0.8200355360013537,
+      "grad_norm": 0.3052305579185486,
+      "learning_rate": 0.001,
+      "loss": 2.3592,
+      "step": 19384
+    },
+    {
+      "epoch": 0.8200778407648701,
+      "grad_norm": 0.1488254964351654,
+      "learning_rate": 0.001,
+      "loss": 1.8685,
+      "step": 19385
+    },
+    {
+      "epoch": 0.8201201455283865,
+      "grad_norm": 0.18521372973918915,
+      "learning_rate": 0.001,
+      "loss": 2.578,
+      "step": 19386
+    },
+    {
+      "epoch": 0.8201624502919028,
+      "grad_norm": 0.1931733936071396,
+      "learning_rate": 0.001,
+      "loss": 1.9068,
+      "step": 19387
+    },
+    {
+      "epoch": 0.8202047550554192,
+      "grad_norm": 0.16335810720920563,
+      "learning_rate": 0.001,
+      "loss": 2.0929,
+      "step": 19388
+    },
+    {
+      "epoch": 0.8202470598189356,
+      "grad_norm": 0.14352665841579437,
+      "learning_rate": 0.001,
+      "loss": 1.5255,
+      "step": 19389
+    },
+    {
+      "epoch": 0.8202893645824519,
+      "grad_norm": 0.22038647532463074,
+      "learning_rate": 0.001,
+      "loss": 3.3152,
+      "step": 19390
+    },
+    {
+      "epoch": 0.8203316693459684,
+      "grad_norm": 0.17173390090465546,
+      "learning_rate": 0.001,
+      "loss": 1.8423,
+      "step": 19391
+    },
+    {
+      "epoch": 0.8203739741094848,
+      "grad_norm": 0.1474774330854416,
+      "learning_rate": 0.001,
+      "loss": 1.6557,
+      "step": 19392
+    },
+    {
+      "epoch": 0.8204162788730011,
+      "grad_norm": 0.1488204449415207,
+      "learning_rate": 0.001,
+      "loss": 1.867,
+      "step": 19393
+    },
+    {
+      "epoch": 0.8204585836365175,
+      "grad_norm": 0.26279768347740173,
+      "learning_rate": 0.001,
+      "loss": 2.287,
+      "step": 19394
+    },
+    {
+      "epoch": 0.8205008884000339,
+      "grad_norm": 0.3085935413837433,
+      "learning_rate": 0.001,
+      "loss": 2.1505,
+      "step": 19395
+    },
+    {
+      "epoch": 0.8205431931635502,
+      "grad_norm": 0.14529921114444733,
+      "learning_rate": 0.001,
+      "loss": 1.7498,
+      "step": 19396
+    },
+    {
+      "epoch": 0.8205854979270666,
+      "grad_norm": 0.1698216050863266,
+      "learning_rate": 0.001,
+      "loss": 2.3125,
+      "step": 19397
+    },
+    {
+      "epoch": 0.820627802690583,
+      "grad_norm": 0.1525857001543045,
+      "learning_rate": 0.001,
+      "loss": 2.6507,
+      "step": 19398
+    },
+    {
+      "epoch": 0.8206701074540993,
+      "grad_norm": 0.48174238204956055,
+      "learning_rate": 0.001,
+      "loss": 1.9215,
+      "step": 19399
+    },
+    {
+      "epoch": 0.8207124122176157,
+      "grad_norm": 0.1638476401567459,
+      "learning_rate": 0.001,
+      "loss": 2.4539,
+      "step": 19400
+    },
+    {
+      "epoch": 0.8207547169811321,
+      "grad_norm": 0.1637507677078247,
+      "learning_rate": 0.001,
+      "loss": 1.5719,
+      "step": 19401
+    },
+    {
+      "epoch": 0.8207970217446484,
+      "grad_norm": 0.1512572169303894,
+      "learning_rate": 0.001,
+      "loss": 2.4486,
+      "step": 19402
+    },
+    {
+      "epoch": 0.8208393265081648,
+      "grad_norm": 0.12860354781150818,
+      "learning_rate": 0.001,
+      "loss": 2.5915,
+      "step": 19403
+    },
+    {
+      "epoch": 0.8208816312716812,
+      "grad_norm": 0.15637515485286713,
+      "learning_rate": 0.001,
+      "loss": 2.1001,
+      "step": 19404
+    },
+    {
+      "epoch": 0.8209239360351975,
+      "grad_norm": 0.13939064741134644,
+      "learning_rate": 0.001,
+      "loss": 2.1013,
+      "step": 19405
+    },
+    {
+      "epoch": 0.8209662407987139,
+      "grad_norm": 0.1056860163807869,
+      "learning_rate": 0.001,
+      "loss": 1.4263,
+      "step": 19406
+    },
+    {
+      "epoch": 0.8210085455622304,
+      "grad_norm": 0.14277103543281555,
+      "learning_rate": 0.001,
+      "loss": 2.5677,
+      "step": 19407
+    },
+    {
+      "epoch": 0.8210508503257467,
+      "grad_norm": 0.12135271728038788,
+      "learning_rate": 0.001,
+      "loss": 2.5086,
+      "step": 19408
+    },
+    {
+      "epoch": 0.8210931550892631,
+      "grad_norm": 0.6470211148262024,
+      "learning_rate": 0.001,
+      "loss": 2.8058,
+      "step": 19409
+    },
+    {
+      "epoch": 0.8211354598527795,
+      "grad_norm": 0.14677073061466217,
+      "learning_rate": 0.001,
+      "loss": 2.0381,
+      "step": 19410
+    },
+    {
+      "epoch": 0.8211777646162958,
+      "grad_norm": 0.14049549400806427,
+      "learning_rate": 0.001,
+      "loss": 2.1267,
+      "step": 19411
+    },
+    {
+      "epoch": 0.8212200693798122,
+      "grad_norm": 0.1972610056400299,
+      "learning_rate": 0.001,
+      "loss": 1.4748,
+      "step": 19412
+    },
+    {
+      "epoch": 0.8212623741433286,
+      "grad_norm": 0.16732430458068848,
+      "learning_rate": 0.001,
+      "loss": 2.0388,
+      "step": 19413
+    },
+    {
+      "epoch": 0.8213046789068449,
+      "grad_norm": 0.17487087845802307,
+      "learning_rate": 0.001,
+      "loss": 1.8259,
+      "step": 19414
+    },
+    {
+      "epoch": 0.8213469836703613,
+      "grad_norm": 0.2771112322807312,
+      "learning_rate": 0.001,
+      "loss": 2.8102,
+      "step": 19415
+    },
+    {
+      "epoch": 0.8213892884338777,
+      "grad_norm": 0.15633077919483185,
+      "learning_rate": 0.001,
+      "loss": 2.9745,
+      "step": 19416
+    },
+    {
+      "epoch": 0.821431593197394,
+      "grad_norm": 0.8696543574333191,
+      "learning_rate": 0.001,
+      "loss": 2.1251,
+      "step": 19417
+    },
+    {
+      "epoch": 0.8214738979609104,
+      "grad_norm": 0.7702298164367676,
+      "learning_rate": 0.001,
+      "loss": 2.4299,
+      "step": 19418
+    },
+    {
+      "epoch": 0.8215162027244268,
+      "grad_norm": 1.8494880199432373,
+      "learning_rate": 0.001,
+      "loss": 2.1975,
+      "step": 19419
+    },
+    {
+      "epoch": 0.8215585074879431,
+      "grad_norm": 0.14353471994400024,
+      "learning_rate": 0.001,
+      "loss": 1.602,
+      "step": 19420
+    },
+    {
+      "epoch": 0.8216008122514595,
+      "grad_norm": 0.152914896607399,
+      "learning_rate": 0.001,
+      "loss": 2.5665,
+      "step": 19421
+    },
+    {
+      "epoch": 0.8216431170149758,
+      "grad_norm": 0.1569298803806305,
+      "learning_rate": 0.001,
+      "loss": 2.2651,
+      "step": 19422
+    },
+    {
+      "epoch": 0.8216854217784922,
+      "grad_norm": 0.14512638747692108,
+      "learning_rate": 0.001,
+      "loss": 1.6607,
+      "step": 19423
+    },
+    {
+      "epoch": 0.8217277265420087,
+      "grad_norm": 0.16484789550304413,
+      "learning_rate": 0.001,
+      "loss": 2.1772,
+      "step": 19424
+    },
+    {
+      "epoch": 0.821770031305525,
+      "grad_norm": 0.1395651400089264,
+      "learning_rate": 0.001,
+      "loss": 1.821,
+      "step": 19425
+    },
+    {
+      "epoch": 0.8218123360690414,
+      "grad_norm": 0.675366997718811,
+      "learning_rate": 0.001,
+      "loss": 2.4906,
+      "step": 19426
+    },
+    {
+      "epoch": 0.8218546408325578,
+      "grad_norm": 0.14823488891124725,
+      "learning_rate": 0.001,
+      "loss": 2.0327,
+      "step": 19427
+    },
+    {
+      "epoch": 0.8218969455960741,
+      "grad_norm": 0.14115604758262634,
+      "learning_rate": 0.001,
+      "loss": 1.8341,
+      "step": 19428
+    },
+    {
+      "epoch": 0.8219392503595905,
+      "grad_norm": 0.14076372981071472,
+      "learning_rate": 0.001,
+      "loss": 2.0194,
+      "step": 19429
+    },
+    {
+      "epoch": 0.8219815551231069,
+      "grad_norm": 0.1530357003211975,
+      "learning_rate": 0.001,
+      "loss": 1.8015,
+      "step": 19430
+    },
+    {
+      "epoch": 0.8220238598866232,
+      "grad_norm": 0.1735893338918686,
+      "learning_rate": 0.001,
+      "loss": 2.0837,
+      "step": 19431
+    },
+    {
+      "epoch": 0.8220661646501396,
+      "grad_norm": 264.96417236328125,
+      "learning_rate": 0.001,
+      "loss": 2.6431,
+      "step": 19432
+    },
+    {
+      "epoch": 0.822108469413656,
+      "grad_norm": 0.16807694733142853,
+      "learning_rate": 0.001,
+      "loss": 2.2922,
+      "step": 19433
+    },
+    {
+      "epoch": 0.8221507741771723,
+      "grad_norm": 0.1564405858516693,
+      "learning_rate": 0.001,
+      "loss": 1.6603,
+      "step": 19434
+    },
+    {
+      "epoch": 0.8221930789406887,
+      "grad_norm": 0.1768580824136734,
+      "learning_rate": 0.001,
+      "loss": 1.8339,
+      "step": 19435
+    },
+    {
+      "epoch": 0.8222353837042051,
+      "grad_norm": 0.16592250764369965,
+      "learning_rate": 0.001,
+      "loss": 2.0608,
+      "step": 19436
+    },
+    {
+      "epoch": 0.8222776884677214,
+      "grad_norm": 0.8113108277320862,
+      "learning_rate": 0.001,
+      "loss": 2.8101,
+      "step": 19437
+    },
+    {
+      "epoch": 0.8223199932312378,
+      "grad_norm": 3.1174583435058594,
+      "learning_rate": 0.001,
+      "loss": 2.0837,
+      "step": 19438
+    },
+    {
+      "epoch": 0.8223622979947542,
+      "grad_norm": 0.1804606169462204,
+      "learning_rate": 0.001,
+      "loss": 1.8602,
+      "step": 19439
+    },
+    {
+      "epoch": 0.8224046027582705,
+      "grad_norm": 0.2099166363477707,
+      "learning_rate": 0.001,
+      "loss": 2.2021,
+      "step": 19440
+    },
+    {
+      "epoch": 0.822446907521787,
+      "grad_norm": 0.2024306058883667,
+      "learning_rate": 0.001,
+      "loss": 2.5865,
+      "step": 19441
+    },
+    {
+      "epoch": 0.8224892122853034,
+      "grad_norm": 1.9643110036849976,
+      "learning_rate": 0.001,
+      "loss": 2.6401,
+      "step": 19442
+    },
+    {
+      "epoch": 0.8225315170488197,
+      "grad_norm": 0.2750485837459564,
+      "learning_rate": 0.001,
+      "loss": 1.6251,
+      "step": 19443
+    },
+    {
+      "epoch": 0.8225738218123361,
+      "grad_norm": 0.26389703154563904,
+      "learning_rate": 0.001,
+      "loss": 1.9309,
+      "step": 19444
+    },
+    {
+      "epoch": 0.8226161265758525,
+      "grad_norm": 0.1789424866437912,
+      "learning_rate": 0.001,
+      "loss": 1.6066,
+      "step": 19445
+    },
+    {
+      "epoch": 0.8226584313393688,
+      "grad_norm": 0.2300262451171875,
+      "learning_rate": 0.001,
+      "loss": 2.7709,
+      "step": 19446
+    },
+    {
+      "epoch": 0.8227007361028852,
+      "grad_norm": 0.25446465611457825,
+      "learning_rate": 0.001,
+      "loss": 2.4887,
+      "step": 19447
+    },
+    {
+      "epoch": 0.8227430408664016,
+      "grad_norm": 0.2057051956653595,
+      "learning_rate": 0.001,
+      "loss": 1.8832,
+      "step": 19448
+    },
+    {
+      "epoch": 0.8227853456299179,
+      "grad_norm": 0.17963270843029022,
+      "learning_rate": 0.001,
+      "loss": 3.0981,
+      "step": 19449
+    },
+    {
+      "epoch": 0.8228276503934343,
+      "grad_norm": 0.17611320316791534,
+      "learning_rate": 0.001,
+      "loss": 2.6242,
+      "step": 19450
+    },
+    {
+      "epoch": 0.8228699551569507,
+      "grad_norm": 0.19677342474460602,
+      "learning_rate": 0.001,
+      "loss": 1.7286,
+      "step": 19451
+    },
+    {
+      "epoch": 0.822912259920467,
+      "grad_norm": 0.19172005355358124,
+      "learning_rate": 0.001,
+      "loss": 1.6635,
+      "step": 19452
+    },
+    {
+      "epoch": 0.8229545646839834,
+      "grad_norm": 0.17036627233028412,
+      "learning_rate": 0.001,
+      "loss": 1.5419,
+      "step": 19453
+    },
+    {
+      "epoch": 0.8229968694474998,
+      "grad_norm": 0.1823369711637497,
+      "learning_rate": 0.001,
+      "loss": 2.7447,
+      "step": 19454
+    },
+    {
+      "epoch": 0.8230391742110161,
+      "grad_norm": 0.30361148715019226,
+      "learning_rate": 0.001,
+      "loss": 2.7019,
+      "step": 19455
+    },
+    {
+      "epoch": 0.8230814789745325,
+      "grad_norm": 0.14085571467876434,
+      "learning_rate": 0.001,
+      "loss": 2.5237,
+      "step": 19456
+    },
+    {
+      "epoch": 0.823123783738049,
+      "grad_norm": 0.15281033515930176,
+      "learning_rate": 0.001,
+      "loss": 2.1091,
+      "step": 19457
+    },
+    {
+      "epoch": 0.8231660885015653,
+      "grad_norm": 0.14232198894023895,
+      "learning_rate": 0.001,
+      "loss": 2.4139,
+      "step": 19458
+    },
+    {
+      "epoch": 0.8232083932650817,
+      "grad_norm": 0.142531618475914,
+      "learning_rate": 0.001,
+      "loss": 2.5104,
+      "step": 19459
+    },
+    {
+      "epoch": 0.8232506980285981,
+      "grad_norm": 0.14800485968589783,
+      "learning_rate": 0.001,
+      "loss": 1.6023,
+      "step": 19460
+    },
+    {
+      "epoch": 0.8232930027921144,
+      "grad_norm": 0.15559349954128265,
+      "learning_rate": 0.001,
+      "loss": 2.9792,
+      "step": 19461
+    },
+    {
+      "epoch": 0.8233353075556308,
+      "grad_norm": 0.1521608531475067,
+      "learning_rate": 0.001,
+      "loss": 2.7687,
+      "step": 19462
+    },
+    {
+      "epoch": 0.8233776123191472,
+      "grad_norm": 0.14281997084617615,
+      "learning_rate": 0.001,
+      "loss": 1.9515,
+      "step": 19463
+    },
+    {
+      "epoch": 0.8234199170826635,
+      "grad_norm": 0.9176727533340454,
+      "learning_rate": 0.001,
+      "loss": 1.5359,
+      "step": 19464
+    },
+    {
+      "epoch": 0.8234622218461799,
+      "grad_norm": 0.14004619419574738,
+      "learning_rate": 0.001,
+      "loss": 2.0011,
+      "step": 19465
+    },
+    {
+      "epoch": 0.8235045266096962,
+      "grad_norm": 0.14893396198749542,
+      "learning_rate": 0.001,
+      "loss": 2.1055,
+      "step": 19466
+    },
+    {
+      "epoch": 0.8235468313732126,
+      "grad_norm": 0.14513948559761047,
+      "learning_rate": 0.001,
+      "loss": 2.2117,
+      "step": 19467
+    },
+    {
+      "epoch": 0.823589136136729,
+      "grad_norm": 0.12800917029380798,
+      "learning_rate": 0.001,
+      "loss": 2.3386,
+      "step": 19468
+    },
+    {
+      "epoch": 0.8236314409002453,
+      "grad_norm": 0.15164655447006226,
+      "learning_rate": 0.001,
+      "loss": 1.9978,
+      "step": 19469
+    },
+    {
+      "epoch": 0.8236737456637617,
+      "grad_norm": 0.18457886576652527,
+      "learning_rate": 0.001,
+      "loss": 1.6669,
+      "step": 19470
+    },
+    {
+      "epoch": 0.8237160504272781,
+      "grad_norm": 0.3100104331970215,
+      "learning_rate": 0.001,
+      "loss": 1.9054,
+      "step": 19471
+    },
+    {
+      "epoch": 0.8237583551907944,
+      "grad_norm": 0.16601592302322388,
+      "learning_rate": 0.001,
+      "loss": 2.0019,
+      "step": 19472
+    },
+    {
+      "epoch": 0.8238006599543108,
+      "grad_norm": 0.25218597054481506,
+      "learning_rate": 0.001,
+      "loss": 2.5409,
+      "step": 19473
+    },
+    {
+      "epoch": 0.8238429647178273,
+      "grad_norm": 0.15414927899837494,
+      "learning_rate": 0.001,
+      "loss": 2.0852,
+      "step": 19474
+    },
+    {
+      "epoch": 0.8238852694813436,
+      "grad_norm": 0.1316404789686203,
+      "learning_rate": 0.001,
+      "loss": 1.8446,
+      "step": 19475
+    },
+    {
+      "epoch": 0.82392757424486,
+      "grad_norm": 0.19027061760425568,
+      "learning_rate": 0.001,
+      "loss": 2.2132,
+      "step": 19476
+    },
+    {
+      "epoch": 0.8239698790083764,
+      "grad_norm": 1.3380327224731445,
+      "learning_rate": 0.001,
+      "loss": 1.9915,
+      "step": 19477
+    },
+    {
+      "epoch": 0.8240121837718927,
+      "grad_norm": 0.14829513430595398,
+      "learning_rate": 0.001,
+      "loss": 2.1132,
+      "step": 19478
+    },
+    {
+      "epoch": 0.8240544885354091,
+      "grad_norm": 0.18493659794330597,
+      "learning_rate": 0.001,
+      "loss": 2.2199,
+      "step": 19479
+    },
+    {
+      "epoch": 0.8240967932989255,
+      "grad_norm": 0.13995139300823212,
+      "learning_rate": 0.001,
+      "loss": 1.894,
+      "step": 19480
+    },
+    {
+      "epoch": 0.8241390980624418,
+      "grad_norm": 0.6962743401527405,
+      "learning_rate": 0.001,
+      "loss": 2.9984,
+      "step": 19481
+    },
+    {
+      "epoch": 0.8241814028259582,
+      "grad_norm": 0.17796729505062103,
+      "learning_rate": 0.001,
+      "loss": 2.2849,
+      "step": 19482
+    },
+    {
+      "epoch": 0.8242237075894746,
+      "grad_norm": 0.5168524980545044,
+      "learning_rate": 0.001,
+      "loss": 3.2221,
+      "step": 19483
+    },
+    {
+      "epoch": 0.8242660123529909,
+      "grad_norm": 0.13838616013526917,
+      "learning_rate": 0.001,
+      "loss": 1.861,
+      "step": 19484
+    },
+    {
+      "epoch": 0.8243083171165073,
+      "grad_norm": 1.7203575372695923,
+      "learning_rate": 0.001,
+      "loss": 1.9319,
+      "step": 19485
+    },
+    {
+      "epoch": 0.8243506218800237,
+      "grad_norm": 0.13858763873577118,
+      "learning_rate": 0.001,
+      "loss": 2.8774,
+      "step": 19486
+    },
+    {
+      "epoch": 0.82439292664354,
+      "grad_norm": 0.15123365819454193,
+      "learning_rate": 0.001,
+      "loss": 2.2039,
+      "step": 19487
+    },
+    {
+      "epoch": 0.8244352314070564,
+      "grad_norm": 0.46859481930732727,
+      "learning_rate": 0.001,
+      "loss": 2.0706,
+      "step": 19488
+    },
+    {
+      "epoch": 0.8244775361705728,
+      "grad_norm": 0.16294020414352417,
+      "learning_rate": 0.001,
+      "loss": 1.4793,
+      "step": 19489
+    },
+    {
+      "epoch": 0.8245198409340891,
+      "grad_norm": 0.13173837959766388,
+      "learning_rate": 0.001,
+      "loss": 1.8825,
+      "step": 19490
+    },
+    {
+      "epoch": 0.8245621456976056,
+      "grad_norm": 0.1741243302822113,
+      "learning_rate": 0.001,
+      "loss": 2.136,
+      "step": 19491
+    },
+    {
+      "epoch": 0.824604450461122,
+      "grad_norm": 0.16755490005016327,
+      "learning_rate": 0.001,
+      "loss": 1.7519,
+      "step": 19492
+    },
+    {
+      "epoch": 0.8246467552246383,
+      "grad_norm": 0.17870347201824188,
+      "learning_rate": 0.001,
+      "loss": 2.4803,
+      "step": 19493
+    },
+    {
+      "epoch": 0.8246890599881547,
+      "grad_norm": 0.15022897720336914,
+      "learning_rate": 0.001,
+      "loss": 1.5368,
+      "step": 19494
+    },
+    {
+      "epoch": 0.8247313647516711,
+      "grad_norm": 2.018383026123047,
+      "learning_rate": 0.001,
+      "loss": 1.3298,
+      "step": 19495
+    },
+    {
+      "epoch": 0.8247736695151874,
+      "grad_norm": 0.15447530150413513,
+      "learning_rate": 0.001,
+      "loss": 1.6869,
+      "step": 19496
+    },
+    {
+      "epoch": 0.8248159742787038,
+      "grad_norm": 6.437837600708008,
+      "learning_rate": 0.001,
+      "loss": 2.0111,
+      "step": 19497
+    },
+    {
+      "epoch": 0.8248582790422202,
+      "grad_norm": 0.14872145652770996,
+      "learning_rate": 0.001,
+      "loss": 2.4922,
+      "step": 19498
+    },
+    {
+      "epoch": 0.8249005838057365,
+      "grad_norm": 0.15467225015163422,
+      "learning_rate": 0.001,
+      "loss": 2.5399,
+      "step": 19499
+    },
+    {
+      "epoch": 0.8249428885692529,
+      "grad_norm": 0.16929113864898682,
+      "learning_rate": 0.001,
+      "loss": 1.8642,
+      "step": 19500
+    },
+    {
+      "epoch": 0.8249851933327693,
+      "grad_norm": 0.14469696581363678,
+      "learning_rate": 0.001,
+      "loss": 2.8897,
+      "step": 19501
+    },
+    {
+      "epoch": 0.8250274980962856,
+      "grad_norm": 0.1643838882446289,
+      "learning_rate": 0.001,
+      "loss": 2.0871,
+      "step": 19502
+    },
+    {
+      "epoch": 0.825069802859802,
+      "grad_norm": 0.14725084602832794,
+      "learning_rate": 0.001,
+      "loss": 1.7524,
+      "step": 19503
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.14372789859771729,
+      "learning_rate": 0.001,
+      "loss": 3.4166,
+      "step": 19504
+    },
+    {
+      "epoch": 0.8251544123868347,
+      "grad_norm": 0.155779629945755,
+      "learning_rate": 0.001,
+      "loss": 1.9682,
+      "step": 19505
+    },
+    {
+      "epoch": 0.8251967171503511,
+      "grad_norm": 1.8167226314544678,
+      "learning_rate": 0.001,
+      "loss": 2.1802,
+      "step": 19506
+    },
+    {
+      "epoch": 0.8252390219138676,
+      "grad_norm": 0.4720889925956726,
+      "learning_rate": 0.001,
+      "loss": 1.7518,
+      "step": 19507
+    },
+    {
+      "epoch": 0.8252813266773839,
+      "grad_norm": 1.381333351135254,
+      "learning_rate": 0.001,
+      "loss": 2.3623,
+      "step": 19508
+    },
+    {
+      "epoch": 0.8253236314409003,
+      "grad_norm": 0.3292122781276703,
+      "learning_rate": 0.001,
+      "loss": 1.4812,
+      "step": 19509
+    },
+    {
+      "epoch": 0.8253659362044167,
+      "grad_norm": 0.16913935542106628,
+      "learning_rate": 0.001,
+      "loss": 1.9194,
+      "step": 19510
+    },
+    {
+      "epoch": 0.825408240967933,
+      "grad_norm": 0.20442567765712738,
+      "learning_rate": 0.001,
+      "loss": 2.7706,
+      "step": 19511
+    },
+    {
+      "epoch": 0.8254505457314494,
+      "grad_norm": 0.12844796478748322,
+      "learning_rate": 0.001,
+      "loss": 1.7295,
+      "step": 19512
+    },
+    {
+      "epoch": 0.8254928504949657,
+      "grad_norm": 0.6568970084190369,
+      "learning_rate": 0.001,
+      "loss": 2.2554,
+      "step": 19513
+    },
+    {
+      "epoch": 0.8255351552584821,
+      "grad_norm": 0.3706272542476654,
+      "learning_rate": 0.001,
+      "loss": 1.5584,
+      "step": 19514
+    },
+    {
+      "epoch": 0.8255774600219985,
+      "grad_norm": 18.31789207458496,
+      "learning_rate": 0.001,
+      "loss": 2.2803,
+      "step": 19515
+    },
+    {
+      "epoch": 0.8256197647855148,
+      "grad_norm": 0.15570005774497986,
+      "learning_rate": 0.001,
+      "loss": 2.3987,
+      "step": 19516
+    },
+    {
+      "epoch": 0.8256620695490312,
+      "grad_norm": 0.1890856921672821,
+      "learning_rate": 0.001,
+      "loss": 2.9565,
+      "step": 19517
+    },
+    {
+      "epoch": 0.8257043743125476,
+      "grad_norm": 0.17735014855861664,
+      "learning_rate": 0.001,
+      "loss": 1.7442,
+      "step": 19518
+    },
+    {
+      "epoch": 0.8257466790760639,
+      "grad_norm": 0.5632053017616272,
+      "learning_rate": 0.001,
+      "loss": 3.1095,
+      "step": 19519
+    },
+    {
+      "epoch": 0.8257889838395803,
+      "grad_norm": 0.1494271159172058,
+      "learning_rate": 0.001,
+      "loss": 1.583,
+      "step": 19520
+    },
+    {
+      "epoch": 0.8258312886030967,
+      "grad_norm": 3.3606419563293457,
+      "learning_rate": 0.001,
+      "loss": 2.7092,
+      "step": 19521
+    },
+    {
+      "epoch": 0.825873593366613,
+      "grad_norm": 0.162941113114357,
+      "learning_rate": 0.001,
+      "loss": 2.5138,
+      "step": 19522
+    },
+    {
+      "epoch": 0.8259158981301294,
+      "grad_norm": 0.18008768558502197,
+      "learning_rate": 0.001,
+      "loss": 1.5397,
+      "step": 19523
+    },
+    {
+      "epoch": 0.8259582028936459,
+      "grad_norm": 0.2005215436220169,
+      "learning_rate": 0.001,
+      "loss": 2.3149,
+      "step": 19524
+    },
+    {
+      "epoch": 0.8260005076571622,
+      "grad_norm": 21.058948516845703,
+      "learning_rate": 0.001,
+      "loss": 1.64,
+      "step": 19525
+    },
+    {
+      "epoch": 0.8260428124206786,
+      "grad_norm": 0.18747086822986603,
+      "learning_rate": 0.001,
+      "loss": 1.8543,
+      "step": 19526
+    },
+    {
+      "epoch": 0.826085117184195,
+      "grad_norm": 0.16638407111167908,
+      "learning_rate": 0.001,
+      "loss": 3.0271,
+      "step": 19527
+    },
+    {
+      "epoch": 0.8261274219477113,
+      "grad_norm": 0.26135191321372986,
+      "learning_rate": 0.001,
+      "loss": 1.9732,
+      "step": 19528
+    },
+    {
+      "epoch": 0.8261697267112277,
+      "grad_norm": 0.906550943851471,
+      "learning_rate": 0.001,
+      "loss": 1.9329,
+      "step": 19529
+    },
+    {
+      "epoch": 0.8262120314747441,
+      "grad_norm": 0.19989506900310516,
+      "learning_rate": 0.001,
+      "loss": 2.3301,
+      "step": 19530
+    },
+    {
+      "epoch": 0.8262543362382604,
+      "grad_norm": 0.1638246327638626,
+      "learning_rate": 0.001,
+      "loss": 2.0817,
+      "step": 19531
+    },
+    {
+      "epoch": 0.8262966410017768,
+      "grad_norm": 0.17549832165241241,
+      "learning_rate": 0.001,
+      "loss": 3.5144,
+      "step": 19532
+    },
+    {
+      "epoch": 0.8263389457652932,
+      "grad_norm": 0.18301671743392944,
+      "learning_rate": 0.001,
+      "loss": 2.1419,
+      "step": 19533
+    },
+    {
+      "epoch": 0.8263812505288095,
+      "grad_norm": 0.17851994931697845,
+      "learning_rate": 0.001,
+      "loss": 2.0738,
+      "step": 19534
+    },
+    {
+      "epoch": 0.8264235552923259,
+      "grad_norm": 0.18530461192131042,
+      "learning_rate": 0.001,
+      "loss": 1.9832,
+      "step": 19535
+    },
+    {
+      "epoch": 0.8264658600558423,
+      "grad_norm": 1.8142375946044922,
+      "learning_rate": 0.001,
+      "loss": 2.8591,
+      "step": 19536
+    },
+    {
+      "epoch": 0.8265081648193586,
+      "grad_norm": 0.15847790241241455,
+      "learning_rate": 0.001,
+      "loss": 1.9201,
+      "step": 19537
+    },
+    {
+      "epoch": 0.826550469582875,
+      "grad_norm": 0.8735809922218323,
+      "learning_rate": 0.001,
+      "loss": 3.1536,
+      "step": 19538
+    },
+    {
+      "epoch": 0.8265927743463914,
+      "grad_norm": 0.18533921241760254,
+      "learning_rate": 0.001,
+      "loss": 2.4057,
+      "step": 19539
+    },
+    {
+      "epoch": 0.8266350791099077,
+      "grad_norm": 2.2356624603271484,
+      "learning_rate": 0.001,
+      "loss": 2.8036,
+      "step": 19540
+    },
+    {
+      "epoch": 0.8266773838734242,
+      "grad_norm": 0.1672806292772293,
+      "learning_rate": 0.001,
+      "loss": 2.093,
+      "step": 19541
+    },
+    {
+      "epoch": 0.8267196886369406,
+      "grad_norm": 0.26861119270324707,
+      "learning_rate": 0.001,
+      "loss": 2.5095,
+      "step": 19542
+    },
+    {
+      "epoch": 0.8267619934004569,
+      "grad_norm": 0.1656004935503006,
+      "learning_rate": 0.001,
+      "loss": 2.269,
+      "step": 19543
+    },
+    {
+      "epoch": 0.8268042981639733,
+      "grad_norm": 0.1717766374349594,
+      "learning_rate": 0.001,
+      "loss": 1.9956,
+      "step": 19544
+    },
+    {
+      "epoch": 0.8268466029274897,
+      "grad_norm": 0.21453844010829926,
+      "learning_rate": 0.001,
+      "loss": 3.3933,
+      "step": 19545
+    },
+    {
+      "epoch": 0.826888907691006,
+      "grad_norm": 0.15061205625534058,
+      "learning_rate": 0.001,
+      "loss": 2.3823,
+      "step": 19546
+    },
+    {
+      "epoch": 0.8269312124545224,
+      "grad_norm": 0.20553554594516754,
+      "learning_rate": 0.001,
+      "loss": 2.4435,
+      "step": 19547
+    },
+    {
+      "epoch": 0.8269735172180388,
+      "grad_norm": 0.41001060605049133,
+      "learning_rate": 0.001,
+      "loss": 2.516,
+      "step": 19548
+    },
+    {
+      "epoch": 0.8270158219815551,
+      "grad_norm": 0.16718702018260956,
+      "learning_rate": 0.001,
+      "loss": 2.3812,
+      "step": 19549
+    },
+    {
+      "epoch": 0.8270581267450715,
+      "grad_norm": 0.13911297917366028,
+      "learning_rate": 0.001,
+      "loss": 1.7068,
+      "step": 19550
+    },
+    {
+      "epoch": 0.8271004315085879,
+      "grad_norm": 4.676766872406006,
+      "learning_rate": 0.001,
+      "loss": 3.0646,
+      "step": 19551
+    },
+    {
+      "epoch": 0.8271427362721042,
+      "grad_norm": 0.1728333830833435,
+      "learning_rate": 0.001,
+      "loss": 1.8783,
+      "step": 19552
+    },
+    {
+      "epoch": 0.8271850410356206,
+      "grad_norm": 0.1679517775774002,
+      "learning_rate": 0.001,
+      "loss": 1.7856,
+      "step": 19553
+    },
+    {
+      "epoch": 0.827227345799137,
+      "grad_norm": 0.1749308705329895,
+      "learning_rate": 0.001,
+      "loss": 1.9916,
+      "step": 19554
+    },
+    {
+      "epoch": 0.8272696505626533,
+      "grad_norm": 0.6144496202468872,
+      "learning_rate": 0.001,
+      "loss": 2.504,
+      "step": 19555
+    },
+    {
+      "epoch": 0.8273119553261697,
+      "grad_norm": 0.398067444562912,
+      "learning_rate": 0.001,
+      "loss": 2.041,
+      "step": 19556
+    },
+    {
+      "epoch": 0.827354260089686,
+      "grad_norm": 0.15144911408424377,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 19557
+    },
+    {
+      "epoch": 0.8273965648532025,
+      "grad_norm": 0.1478147804737091,
+      "learning_rate": 0.001,
+      "loss": 2.3869,
+      "step": 19558
+    },
+    {
+      "epoch": 0.8274388696167189,
+      "grad_norm": 0.17387191951274872,
+      "learning_rate": 0.001,
+      "loss": 3.1704,
+      "step": 19559
+    },
+    {
+      "epoch": 0.8274811743802352,
+      "grad_norm": 0.7300922870635986,
+      "learning_rate": 0.001,
+      "loss": 2.1563,
+      "step": 19560
+    },
+    {
+      "epoch": 0.8275234791437516,
+      "grad_norm": 0.17235367000102997,
+      "learning_rate": 0.001,
+      "loss": 1.8931,
+      "step": 19561
+    },
+    {
+      "epoch": 0.827565783907268,
+      "grad_norm": 0.21332013607025146,
+      "learning_rate": 0.001,
+      "loss": 2.7753,
+      "step": 19562
+    },
+    {
+      "epoch": 0.8276080886707843,
+      "grad_norm": 0.18801867961883545,
+      "learning_rate": 0.001,
+      "loss": 1.9607,
+      "step": 19563
+    },
+    {
+      "epoch": 0.8276503934343007,
+      "grad_norm": 0.15900994837284088,
+      "learning_rate": 0.001,
+      "loss": 2.6224,
+      "step": 19564
+    },
+    {
+      "epoch": 0.8276926981978171,
+      "grad_norm": 0.1822887808084488,
+      "learning_rate": 0.001,
+      "loss": 2.2022,
+      "step": 19565
+    },
+    {
+      "epoch": 0.8277350029613334,
+      "grad_norm": 0.1728331744670868,
+      "learning_rate": 0.001,
+      "loss": 1.7164,
+      "step": 19566
+    },
+    {
+      "epoch": 0.8277773077248498,
+      "grad_norm": 0.18861548602581024,
+      "learning_rate": 0.001,
+      "loss": 2.6136,
+      "step": 19567
+    },
+    {
+      "epoch": 0.8278196124883662,
+      "grad_norm": 4.526989459991455,
+      "learning_rate": 0.001,
+      "loss": 2.0578,
+      "step": 19568
+    },
+    {
+      "epoch": 0.8278619172518825,
+      "grad_norm": 1.0483940839767456,
+      "learning_rate": 0.001,
+      "loss": 2.2614,
+      "step": 19569
+    },
+    {
+      "epoch": 0.8279042220153989,
+      "grad_norm": 6.750152587890625,
+      "learning_rate": 0.001,
+      "loss": 1.6585,
+      "step": 19570
+    },
+    {
+      "epoch": 0.8279465267789153,
+      "grad_norm": 0.15433065593242645,
+      "learning_rate": 0.001,
+      "loss": 1.8853,
+      "step": 19571
+    },
+    {
+      "epoch": 0.8279888315424316,
+      "grad_norm": 7.699106693267822,
+      "learning_rate": 0.001,
+      "loss": 1.9598,
+      "step": 19572
+    },
+    {
+      "epoch": 0.828031136305948,
+      "grad_norm": 0.1841927170753479,
+      "learning_rate": 0.001,
+      "loss": 1.8607,
+      "step": 19573
+    },
+    {
+      "epoch": 0.8280734410694645,
+      "grad_norm": 0.16091950237751007,
+      "learning_rate": 0.001,
+      "loss": 2.6873,
+      "step": 19574
+    },
+    {
+      "epoch": 0.8281157458329808,
+      "grad_norm": 0.188730388879776,
+      "learning_rate": 0.001,
+      "loss": 2.4136,
+      "step": 19575
+    },
+    {
+      "epoch": 0.8281580505964972,
+      "grad_norm": 0.18793466687202454,
+      "learning_rate": 0.001,
+      "loss": 2.9059,
+      "step": 19576
+    },
+    {
+      "epoch": 0.8282003553600136,
+      "grad_norm": 10.438246726989746,
+      "learning_rate": 0.001,
+      "loss": 2.4116,
+      "step": 19577
+    },
+    {
+      "epoch": 0.8282426601235299,
+      "grad_norm": 1.2651994228363037,
+      "learning_rate": 0.001,
+      "loss": 2.6492,
+      "step": 19578
+    },
+    {
+      "epoch": 0.8282849648870463,
+      "grad_norm": 0.23723463714122772,
+      "learning_rate": 0.001,
+      "loss": 1.6084,
+      "step": 19579
+    },
+    {
+      "epoch": 0.8283272696505627,
+      "grad_norm": 0.18241967260837555,
+      "learning_rate": 0.001,
+      "loss": 2.2461,
+      "step": 19580
+    },
+    {
+      "epoch": 0.828369574414079,
+      "grad_norm": 0.1818530410528183,
+      "learning_rate": 0.001,
+      "loss": 2.7537,
+      "step": 19581
+    },
+    {
+      "epoch": 0.8284118791775954,
+      "grad_norm": 0.20046694576740265,
+      "learning_rate": 0.001,
+      "loss": 1.8172,
+      "step": 19582
+    },
+    {
+      "epoch": 0.8284541839411118,
+      "grad_norm": 0.20764821767807007,
+      "learning_rate": 0.001,
+      "loss": 1.5514,
+      "step": 19583
+    },
+    {
+      "epoch": 0.8284964887046281,
+      "grad_norm": 0.19038403034210205,
+      "learning_rate": 0.001,
+      "loss": 2.4937,
+      "step": 19584
+    },
+    {
+      "epoch": 0.8285387934681445,
+      "grad_norm": 13.164509773254395,
+      "learning_rate": 0.001,
+      "loss": 2.2863,
+      "step": 19585
+    },
+    {
+      "epoch": 0.8285810982316609,
+      "grad_norm": 0.18057294189929962,
+      "learning_rate": 0.001,
+      "loss": 2.6837,
+      "step": 19586
+    },
+    {
+      "epoch": 0.8286234029951772,
+      "grad_norm": 0.20905058085918427,
+      "learning_rate": 0.001,
+      "loss": 2.1163,
+      "step": 19587
+    },
+    {
+      "epoch": 0.8286657077586936,
+      "grad_norm": 0.23860158026218414,
+      "learning_rate": 0.001,
+      "loss": 2.5112,
+      "step": 19588
+    },
+    {
+      "epoch": 0.82870801252221,
+      "grad_norm": 0.1781463325023651,
+      "learning_rate": 0.001,
+      "loss": 3.0306,
+      "step": 19589
+    },
+    {
+      "epoch": 0.8287503172857263,
+      "grad_norm": 0.7263326644897461,
+      "learning_rate": 0.001,
+      "loss": 1.8875,
+      "step": 19590
+    },
+    {
+      "epoch": 0.8287926220492428,
+      "grad_norm": 0.8033844232559204,
+      "learning_rate": 0.001,
+      "loss": 2.5529,
+      "step": 19591
+    },
+    {
+      "epoch": 0.8288349268127592,
+      "grad_norm": 0.18857400119304657,
+      "learning_rate": 0.001,
+      "loss": 2.7359,
+      "step": 19592
+    },
+    {
+      "epoch": 0.8288772315762755,
+      "grad_norm": 0.16333748400211334,
+      "learning_rate": 0.001,
+      "loss": 2.0609,
+      "step": 19593
+    },
+    {
+      "epoch": 0.8289195363397919,
+      "grad_norm": 0.19050641357898712,
+      "learning_rate": 0.001,
+      "loss": 1.7926,
+      "step": 19594
+    },
+    {
+      "epoch": 0.8289618411033083,
+      "grad_norm": 0.1565246731042862,
+      "learning_rate": 0.001,
+      "loss": 2.0922,
+      "step": 19595
+    },
+    {
+      "epoch": 0.8290041458668246,
+      "grad_norm": 0.13431625068187714,
+      "learning_rate": 0.001,
+      "loss": 1.9739,
+      "step": 19596
+    },
+    {
+      "epoch": 0.829046450630341,
+      "grad_norm": 4.049526214599609,
+      "learning_rate": 0.001,
+      "loss": 3.013,
+      "step": 19597
+    },
+    {
+      "epoch": 0.8290887553938574,
+      "grad_norm": 0.15808945894241333,
+      "learning_rate": 0.001,
+      "loss": 2.2172,
+      "step": 19598
+    },
+    {
+      "epoch": 0.8291310601573737,
+      "grad_norm": 0.15318995714187622,
+      "learning_rate": 0.001,
+      "loss": 2.4165,
+      "step": 19599
+    },
+    {
+      "epoch": 0.8291733649208901,
+      "grad_norm": 0.16532470285892487,
+      "learning_rate": 0.001,
+      "loss": 1.7708,
+      "step": 19600
+    },
+    {
+      "epoch": 0.8292156696844064,
+      "grad_norm": 0.21278615295886993,
+      "learning_rate": 0.001,
+      "loss": 1.9634,
+      "step": 19601
+    },
+    {
+      "epoch": 0.8292579744479228,
+      "grad_norm": 0.16973280906677246,
+      "learning_rate": 0.001,
+      "loss": 2.0971,
+      "step": 19602
+    },
+    {
+      "epoch": 0.8293002792114392,
+      "grad_norm": 0.191205233335495,
+      "learning_rate": 0.001,
+      "loss": 2.2638,
+      "step": 19603
+    },
+    {
+      "epoch": 0.8293425839749555,
+      "grad_norm": 0.18113046884536743,
+      "learning_rate": 0.001,
+      "loss": 1.898,
+      "step": 19604
+    },
+    {
+      "epoch": 0.8293848887384719,
+      "grad_norm": 7.267598628997803,
+      "learning_rate": 0.001,
+      "loss": 1.8977,
+      "step": 19605
+    },
+    {
+      "epoch": 0.8294271935019883,
+      "grad_norm": 0.16045303642749786,
+      "learning_rate": 0.001,
+      "loss": 1.5686,
+      "step": 19606
+    },
+    {
+      "epoch": 0.8294694982655046,
+      "grad_norm": 0.15106256306171417,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 19607
+    },
+    {
+      "epoch": 0.829511803029021,
+      "grad_norm": 0.16756461560726166,
+      "learning_rate": 0.001,
+      "loss": 1.7174,
+      "step": 19608
+    },
+    {
+      "epoch": 0.8295541077925375,
+      "grad_norm": 0.6214344501495361,
+      "learning_rate": 0.001,
+      "loss": 2.2515,
+      "step": 19609
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 0.1721925288438797,
+      "learning_rate": 0.001,
+      "loss": 1.7405,
+      "step": 19610
+    },
+    {
+      "epoch": 0.8296387173195702,
+      "grad_norm": 0.1743050515651703,
+      "learning_rate": 0.001,
+      "loss": 1.6597,
+      "step": 19611
+    },
+    {
+      "epoch": 0.8296810220830866,
+      "grad_norm": 0.5375502109527588,
+      "learning_rate": 0.001,
+      "loss": 2.6013,
+      "step": 19612
+    },
+    {
+      "epoch": 0.8297233268466029,
+      "grad_norm": 2.799589157104492,
+      "learning_rate": 0.001,
+      "loss": 1.9554,
+      "step": 19613
+    },
+    {
+      "epoch": 0.8297656316101193,
+      "grad_norm": 0.17461428046226501,
+      "learning_rate": 0.001,
+      "loss": 1.6489,
+      "step": 19614
+    },
+    {
+      "epoch": 0.8298079363736357,
+      "grad_norm": 0.16564109921455383,
+      "learning_rate": 0.001,
+      "loss": 1.7323,
+      "step": 19615
+    },
+    {
+      "epoch": 0.829850241137152,
+      "grad_norm": 0.2872896194458008,
+      "learning_rate": 0.001,
+      "loss": 2.0232,
+      "step": 19616
+    },
+    {
+      "epoch": 0.8298925459006684,
+      "grad_norm": 0.13868144154548645,
+      "learning_rate": 0.001,
+      "loss": 1.6235,
+      "step": 19617
+    },
+    {
+      "epoch": 0.8299348506641848,
+      "grad_norm": 0.17989665269851685,
+      "learning_rate": 0.001,
+      "loss": 1.8112,
+      "step": 19618
+    },
+    {
+      "epoch": 0.8299771554277011,
+      "grad_norm": 0.42621657252311707,
+      "learning_rate": 0.001,
+      "loss": 2.8575,
+      "step": 19619
+    },
+    {
+      "epoch": 0.8300194601912175,
+      "grad_norm": 0.24013011157512665,
+      "learning_rate": 0.001,
+      "loss": 2.0607,
+      "step": 19620
+    },
+    {
+      "epoch": 0.8300617649547339,
+      "grad_norm": 0.22159551084041595,
+      "learning_rate": 0.001,
+      "loss": 2.8096,
+      "step": 19621
+    },
+    {
+      "epoch": 0.8301040697182502,
+      "grad_norm": 0.5154104232788086,
+      "learning_rate": 0.001,
+      "loss": 2.2271,
+      "step": 19622
+    },
+    {
+      "epoch": 0.8301463744817666,
+      "grad_norm": 0.9761685729026794,
+      "learning_rate": 0.001,
+      "loss": 2.3908,
+      "step": 19623
+    },
+    {
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.36297187209129333,
+      "learning_rate": 0.001,
+      "loss": 3.5581,
+      "step": 19624
+    },
+    {
+      "epoch": 0.8302309840087994,
+      "grad_norm": 0.15795831382274628,
+      "learning_rate": 0.001,
+      "loss": 1.8869,
+      "step": 19625
+    },
+    {
+      "epoch": 0.8302732887723158,
+      "grad_norm": 0.19538545608520508,
+      "learning_rate": 0.001,
+      "loss": 2.2421,
+      "step": 19626
+    },
+    {
+      "epoch": 0.8303155935358322,
+      "grad_norm": 0.1804639995098114,
+      "learning_rate": 0.001,
+      "loss": 2.4326,
+      "step": 19627
+    },
+    {
+      "epoch": 0.8303578982993485,
+      "grad_norm": 0.14908581972122192,
+      "learning_rate": 0.001,
+      "loss": 2.9382,
+      "step": 19628
+    },
+    {
+      "epoch": 0.8304002030628649,
+      "grad_norm": 0.21959789097309113,
+      "learning_rate": 0.001,
+      "loss": 2.003,
+      "step": 19629
+    },
+    {
+      "epoch": 0.8304425078263813,
+      "grad_norm": 0.5916885733604431,
+      "learning_rate": 0.001,
+      "loss": 3.7824,
+      "step": 19630
+    },
+    {
+      "epoch": 0.8304848125898976,
+      "grad_norm": 0.1725091189146042,
+      "learning_rate": 0.001,
+      "loss": 1.8455,
+      "step": 19631
+    },
+    {
+      "epoch": 0.830527117353414,
+      "grad_norm": 1.093875765800476,
+      "learning_rate": 0.001,
+      "loss": 1.9038,
+      "step": 19632
+    },
+    {
+      "epoch": 0.8305694221169304,
+      "grad_norm": 0.1586592197418213,
+      "learning_rate": 0.001,
+      "loss": 1.7105,
+      "step": 19633
+    },
+    {
+      "epoch": 0.8306117268804467,
+      "grad_norm": 0.1464533805847168,
+      "learning_rate": 0.001,
+      "loss": 1.5658,
+      "step": 19634
+    },
+    {
+      "epoch": 0.8306540316439631,
+      "grad_norm": 0.2901553213596344,
+      "learning_rate": 0.001,
+      "loss": 1.9834,
+      "step": 19635
+    },
+    {
+      "epoch": 0.8306963364074795,
+      "grad_norm": 0.2217240035533905,
+      "learning_rate": 0.001,
+      "loss": 2.5072,
+      "step": 19636
+    },
+    {
+      "epoch": 0.8307386411709958,
+      "grad_norm": 0.24927441775798798,
+      "learning_rate": 0.001,
+      "loss": 1.8134,
+      "step": 19637
+    },
+    {
+      "epoch": 0.8307809459345122,
+      "grad_norm": 0.1451728343963623,
+      "learning_rate": 0.001,
+      "loss": 2.3247,
+      "step": 19638
+    },
+    {
+      "epoch": 0.8308232506980286,
+      "grad_norm": 0.1595974713563919,
+      "learning_rate": 0.001,
+      "loss": 1.5477,
+      "step": 19639
+    },
+    {
+      "epoch": 0.830865555461545,
+      "grad_norm": 0.4610111117362976,
+      "learning_rate": 0.001,
+      "loss": 1.5437,
+      "step": 19640
+    },
+    {
+      "epoch": 0.8309078602250614,
+      "grad_norm": 0.1525607854127884,
+      "learning_rate": 0.001,
+      "loss": 2.1338,
+      "step": 19641
+    },
+    {
+      "epoch": 0.8309501649885778,
+      "grad_norm": 0.16522721946239471,
+      "learning_rate": 0.001,
+      "loss": 3.4636,
+      "step": 19642
+    },
+    {
+      "epoch": 0.8309924697520941,
+      "grad_norm": 0.4826166033744812,
+      "learning_rate": 0.001,
+      "loss": 1.8839,
+      "step": 19643
+    },
+    {
+      "epoch": 0.8310347745156105,
+      "grad_norm": 0.15211009979248047,
+      "learning_rate": 0.001,
+      "loss": 1.7515,
+      "step": 19644
+    },
+    {
+      "epoch": 0.8310770792791269,
+      "grad_norm": 0.13759112358093262,
+      "learning_rate": 0.001,
+      "loss": 1.6183,
+      "step": 19645
+    },
+    {
+      "epoch": 0.8311193840426432,
+      "grad_norm": 0.1878431886434555,
+      "learning_rate": 0.001,
+      "loss": 1.7729,
+      "step": 19646
+    },
+    {
+      "epoch": 0.8311616888061596,
+      "grad_norm": 0.5903341174125671,
+      "learning_rate": 0.001,
+      "loss": 2.8315,
+      "step": 19647
+    },
+    {
+      "epoch": 0.8312039935696759,
+      "grad_norm": 0.16062727570533752,
+      "learning_rate": 0.001,
+      "loss": 2.2591,
+      "step": 19648
+    },
+    {
+      "epoch": 0.8312462983331923,
+      "grad_norm": 0.1513572633266449,
+      "learning_rate": 0.001,
+      "loss": 1.9523,
+      "step": 19649
+    },
+    {
+      "epoch": 0.8312886030967087,
+      "grad_norm": 3.5382885932922363,
+      "learning_rate": 0.001,
+      "loss": 1.8145,
+      "step": 19650
+    },
+    {
+      "epoch": 0.831330907860225,
+      "grad_norm": 0.6378043293952942,
+      "learning_rate": 0.001,
+      "loss": 2.7972,
+      "step": 19651
+    },
+    {
+      "epoch": 0.8313732126237414,
+      "grad_norm": 0.5603189468383789,
+      "learning_rate": 0.001,
+      "loss": 1.9918,
+      "step": 19652
+    },
+    {
+      "epoch": 0.8314155173872578,
+      "grad_norm": 0.7151978611946106,
+      "learning_rate": 0.001,
+      "loss": 1.8135,
+      "step": 19653
+    },
+    {
+      "epoch": 0.8314578221507741,
+      "grad_norm": 0.14867694675922394,
+      "learning_rate": 0.001,
+      "loss": 1.9327,
+      "step": 19654
+    },
+    {
+      "epoch": 0.8315001269142905,
+      "grad_norm": 0.15285193920135498,
+      "learning_rate": 0.001,
+      "loss": 1.8189,
+      "step": 19655
+    },
+    {
+      "epoch": 0.831542431677807,
+      "grad_norm": 0.26532498002052307,
+      "learning_rate": 0.001,
+      "loss": 2.4956,
+      "step": 19656
+    },
+    {
+      "epoch": 0.8315847364413232,
+      "grad_norm": 0.18189120292663574,
+      "learning_rate": 0.001,
+      "loss": 2.8454,
+      "step": 19657
+    },
+    {
+      "epoch": 0.8316270412048397,
+      "grad_norm": 0.335979700088501,
+      "learning_rate": 0.001,
+      "loss": 1.7169,
+      "step": 19658
+    },
+    {
+      "epoch": 0.8316693459683561,
+      "grad_norm": 1.2381075620651245,
+      "learning_rate": 0.001,
+      "loss": 1.5355,
+      "step": 19659
+    },
+    {
+      "epoch": 0.8317116507318724,
+      "grad_norm": 0.18235164880752563,
+      "learning_rate": 0.001,
+      "loss": 2.2521,
+      "step": 19660
+    },
+    {
+      "epoch": 0.8317539554953888,
+      "grad_norm": 0.5001907348632812,
+      "learning_rate": 0.001,
+      "loss": 2.8203,
+      "step": 19661
+    },
+    {
+      "epoch": 0.8317962602589052,
+      "grad_norm": 0.13656170666217804,
+      "learning_rate": 0.001,
+      "loss": 2.2657,
+      "step": 19662
+    },
+    {
+      "epoch": 0.8318385650224215,
+      "grad_norm": 0.1643732637166977,
+      "learning_rate": 0.001,
+      "loss": 1.7424,
+      "step": 19663
+    },
+    {
+      "epoch": 0.8318808697859379,
+      "grad_norm": 0.15448181331157684,
+      "learning_rate": 0.001,
+      "loss": 1.4355,
+      "step": 19664
+    },
+    {
+      "epoch": 0.8319231745494543,
+      "grad_norm": 2.5105960369110107,
+      "learning_rate": 0.001,
+      "loss": 1.913,
+      "step": 19665
+    },
+    {
+      "epoch": 0.8319654793129706,
+      "grad_norm": 0.16451245546340942,
+      "learning_rate": 0.001,
+      "loss": 2.2253,
+      "step": 19666
+    },
+    {
+      "epoch": 0.832007784076487,
+      "grad_norm": 0.15725699067115784,
+      "learning_rate": 0.001,
+      "loss": 1.9037,
+      "step": 19667
+    },
+    {
+      "epoch": 0.8320500888400034,
+      "grad_norm": 0.32337695360183716,
+      "learning_rate": 0.001,
+      "loss": 2.0358,
+      "step": 19668
+    },
+    {
+      "epoch": 0.8320923936035197,
+      "grad_norm": 0.21804969012737274,
+      "learning_rate": 0.001,
+      "loss": 1.8434,
+      "step": 19669
+    },
+    {
+      "epoch": 0.8321346983670361,
+      "grad_norm": 5.778602123260498,
+      "learning_rate": 0.001,
+      "loss": 1.8731,
+      "step": 19670
+    },
+    {
+      "epoch": 0.8321770031305525,
+      "grad_norm": 0.21234005689620972,
+      "learning_rate": 0.001,
+      "loss": 1.9701,
+      "step": 19671
+    },
+    {
+      "epoch": 0.8322193078940688,
+      "grad_norm": 2.040515661239624,
+      "learning_rate": 0.001,
+      "loss": 3.1024,
+      "step": 19672
+    },
+    {
+      "epoch": 0.8322616126575852,
+      "grad_norm": 0.14439794421195984,
+      "learning_rate": 0.001,
+      "loss": 1.4982,
+      "step": 19673
+    },
+    {
+      "epoch": 0.8323039174211017,
+      "grad_norm": 0.14553527534008026,
+      "learning_rate": 0.001,
+      "loss": 2.5106,
+      "step": 19674
+    },
+    {
+      "epoch": 0.832346222184618,
+      "grad_norm": 0.1754305213689804,
+      "learning_rate": 0.001,
+      "loss": 2.5787,
+      "step": 19675
+    },
+    {
+      "epoch": 0.8323885269481344,
+      "grad_norm": 0.13530834019184113,
+      "learning_rate": 0.001,
+      "loss": 1.6652,
+      "step": 19676
+    },
+    {
+      "epoch": 0.8324308317116508,
+      "grad_norm": 16.148717880249023,
+      "learning_rate": 0.001,
+      "loss": 1.9761,
+      "step": 19677
+    },
+    {
+      "epoch": 0.8324731364751671,
+      "grad_norm": 0.16768808662891388,
+      "learning_rate": 0.001,
+      "loss": 2.5138,
+      "step": 19678
+    },
+    {
+      "epoch": 0.8325154412386835,
+      "grad_norm": 0.16565167903900146,
+      "learning_rate": 0.001,
+      "loss": 1.8295,
+      "step": 19679
+    },
+    {
+      "epoch": 0.8325577460021999,
+      "grad_norm": 0.24499014019966125,
+      "learning_rate": 0.001,
+      "loss": 2.4988,
+      "step": 19680
+    },
+    {
+      "epoch": 0.8326000507657162,
+      "grad_norm": 0.1661049723625183,
+      "learning_rate": 0.001,
+      "loss": 2.2289,
+      "step": 19681
+    },
+    {
+      "epoch": 0.8326423555292326,
+      "grad_norm": 0.16528615355491638,
+      "learning_rate": 0.001,
+      "loss": 1.5157,
+      "step": 19682
+    },
+    {
+      "epoch": 0.832684660292749,
+      "grad_norm": 0.1665087193250656,
+      "learning_rate": 0.001,
+      "loss": 2.2805,
+      "step": 19683
+    },
+    {
+      "epoch": 0.8327269650562653,
+      "grad_norm": 0.1810268610715866,
+      "learning_rate": 0.001,
+      "loss": 2.3202,
+      "step": 19684
+    },
+    {
+      "epoch": 0.8327692698197817,
+      "grad_norm": 0.15981633961200714,
+      "learning_rate": 0.001,
+      "loss": 1.511,
+      "step": 19685
+    },
+    {
+      "epoch": 0.8328115745832981,
+      "grad_norm": 0.43922242522239685,
+      "learning_rate": 0.001,
+      "loss": 3.167,
+      "step": 19686
+    },
+    {
+      "epoch": 0.8328538793468144,
+      "grad_norm": 0.16174915432929993,
+      "learning_rate": 0.001,
+      "loss": 1.8022,
+      "step": 19687
+    },
+    {
+      "epoch": 0.8328961841103308,
+      "grad_norm": 0.12966187298297882,
+      "learning_rate": 0.001,
+      "loss": 2.7875,
+      "step": 19688
+    },
+    {
+      "epoch": 0.8329384888738472,
+      "grad_norm": 0.17948752641677856,
+      "learning_rate": 0.001,
+      "loss": 1.7173,
+      "step": 19689
+    },
+    {
+      "epoch": 0.8329807936373635,
+      "grad_norm": 0.17646683752536774,
+      "learning_rate": 0.001,
+      "loss": 1.9492,
+      "step": 19690
+    },
+    {
+      "epoch": 0.83302309840088,
+      "grad_norm": 0.17098137736320496,
+      "learning_rate": 0.001,
+      "loss": 1.5657,
+      "step": 19691
+    },
+    {
+      "epoch": 0.8330654031643963,
+      "grad_norm": 7.42794942855835,
+      "learning_rate": 0.001,
+      "loss": 1.7042,
+      "step": 19692
+    },
+    {
+      "epoch": 0.8331077079279127,
+      "grad_norm": 0.15731237828731537,
+      "learning_rate": 0.001,
+      "loss": 2.4516,
+      "step": 19693
+    },
+    {
+      "epoch": 0.8331500126914291,
+      "grad_norm": 0.16267119348049164,
+      "learning_rate": 0.001,
+      "loss": 1.8393,
+      "step": 19694
+    },
+    {
+      "epoch": 0.8331923174549454,
+      "grad_norm": 0.1380946934223175,
+      "learning_rate": 0.001,
+      "loss": 2.3719,
+      "step": 19695
+    },
+    {
+      "epoch": 0.8332346222184618,
+      "grad_norm": 0.1593260020017624,
+      "learning_rate": 0.001,
+      "loss": 1.9913,
+      "step": 19696
+    },
+    {
+      "epoch": 0.8332769269819782,
+      "grad_norm": 0.19638261198997498,
+      "learning_rate": 0.001,
+      "loss": 2.1259,
+      "step": 19697
+    },
+    {
+      "epoch": 0.8333192317454945,
+      "grad_norm": 0.1739358901977539,
+      "learning_rate": 0.001,
+      "loss": 2.602,
+      "step": 19698
+    },
+    {
+      "epoch": 0.8333615365090109,
+      "grad_norm": 0.28092655539512634,
+      "learning_rate": 0.001,
+      "loss": 1.8167,
+      "step": 19699
+    },
+    {
+      "epoch": 0.8334038412725273,
+      "grad_norm": 0.15453416109085083,
+      "learning_rate": 0.001,
+      "loss": 1.8884,
+      "step": 19700
+    },
+    {
+      "epoch": 0.8334461460360436,
+      "grad_norm": 0.2648554742336273,
+      "learning_rate": 0.001,
+      "loss": 3.4063,
+      "step": 19701
+    },
+    {
+      "epoch": 0.83348845079956,
+      "grad_norm": 0.15070085227489471,
+      "learning_rate": 0.001,
+      "loss": 1.9261,
+      "step": 19702
+    },
+    {
+      "epoch": 0.8335307555630764,
+      "grad_norm": 0.15136472880840302,
+      "learning_rate": 0.001,
+      "loss": 1.8056,
+      "step": 19703
+    },
+    {
+      "epoch": 0.8335730603265927,
+      "grad_norm": 0.17444024980068207,
+      "learning_rate": 0.001,
+      "loss": 2.3862,
+      "step": 19704
+    },
+    {
+      "epoch": 0.8336153650901091,
+      "grad_norm": 0.20091457664966583,
+      "learning_rate": 0.001,
+      "loss": 2.6777,
+      "step": 19705
+    },
+    {
+      "epoch": 0.8336576698536255,
+      "grad_norm": 0.17483146488666534,
+      "learning_rate": 0.001,
+      "loss": 3.9762,
+      "step": 19706
+    },
+    {
+      "epoch": 0.8336999746171418,
+      "grad_norm": 0.14179664850234985,
+      "learning_rate": 0.001,
+      "loss": 1.7762,
+      "step": 19707
+    },
+    {
+      "epoch": 0.8337422793806583,
+      "grad_norm": 0.14632867276668549,
+      "learning_rate": 0.001,
+      "loss": 2.4196,
+      "step": 19708
+    },
+    {
+      "epoch": 0.8337845841441747,
+      "grad_norm": 0.1618465930223465,
+      "learning_rate": 0.001,
+      "loss": 2.0228,
+      "step": 19709
+    },
+    {
+      "epoch": 0.833826888907691,
+      "grad_norm": 0.17605137825012207,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 19710
+    },
+    {
+      "epoch": 0.8338691936712074,
+      "grad_norm": 0.1627848893404007,
+      "learning_rate": 0.001,
+      "loss": 2.5271,
+      "step": 19711
+    },
+    {
+      "epoch": 0.8339114984347238,
+      "grad_norm": 0.15712319314479828,
+      "learning_rate": 0.001,
+      "loss": 1.6956,
+      "step": 19712
+    },
+    {
+      "epoch": 0.8339538031982401,
+      "grad_norm": 0.2377215325832367,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 19713
+    },
+    {
+      "epoch": 0.8339961079617565,
+      "grad_norm": 0.14248542487621307,
+      "learning_rate": 0.001,
+      "loss": 2.9125,
+      "step": 19714
+    },
+    {
+      "epoch": 0.8340384127252729,
+      "grad_norm": 0.1505517214536667,
+      "learning_rate": 0.001,
+      "loss": 1.4674,
+      "step": 19715
+    },
+    {
+      "epoch": 0.8340807174887892,
+      "grad_norm": 0.11878804862499237,
+      "learning_rate": 0.001,
+      "loss": 1.7173,
+      "step": 19716
+    },
+    {
+      "epoch": 0.8341230222523056,
+      "grad_norm": 0.1470642238855362,
+      "learning_rate": 0.001,
+      "loss": 1.3979,
+      "step": 19717
+    },
+    {
+      "epoch": 0.834165327015822,
+      "grad_norm": 0.14857931435108185,
+      "learning_rate": 0.001,
+      "loss": 2.3536,
+      "step": 19718
+    },
+    {
+      "epoch": 0.8342076317793383,
+      "grad_norm": 0.14904291927814484,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 19719
+    },
+    {
+      "epoch": 0.8342499365428547,
+      "grad_norm": 4.184340476989746,
+      "learning_rate": 0.001,
+      "loss": 2.0502,
+      "step": 19720
+    },
+    {
+      "epoch": 0.8342922413063711,
+      "grad_norm": 0.16382192075252533,
+      "learning_rate": 0.001,
+      "loss": 3.3113,
+      "step": 19721
+    },
+    {
+      "epoch": 0.8343345460698874,
+      "grad_norm": 2.076079845428467,
+      "learning_rate": 0.001,
+      "loss": 2.152,
+      "step": 19722
+    },
+    {
+      "epoch": 0.8343768508334038,
+      "grad_norm": 0.13855180144309998,
+      "learning_rate": 0.001,
+      "loss": 1.7025,
+      "step": 19723
+    },
+    {
+      "epoch": 0.8344191555969203,
+      "grad_norm": 0.1491064429283142,
+      "learning_rate": 0.001,
+      "loss": 1.7431,
+      "step": 19724
+    },
+    {
+      "epoch": 0.8344614603604366,
+      "grad_norm": 0.3308009207248688,
+      "learning_rate": 0.001,
+      "loss": 1.3676,
+      "step": 19725
+    },
+    {
+      "epoch": 0.834503765123953,
+      "grad_norm": 1.1953833103179932,
+      "learning_rate": 0.001,
+      "loss": 2.9253,
+      "step": 19726
+    },
+    {
+      "epoch": 0.8345460698874694,
+      "grad_norm": 0.11848998814821243,
+      "learning_rate": 0.001,
+      "loss": 1.6943,
+      "step": 19727
+    },
+    {
+      "epoch": 0.8345883746509857,
+      "grad_norm": 1.0493426322937012,
+      "learning_rate": 0.001,
+      "loss": 1.9968,
+      "step": 19728
+    },
+    {
+      "epoch": 0.8346306794145021,
+      "grad_norm": 0.13267287611961365,
+      "learning_rate": 0.001,
+      "loss": 1.927,
+      "step": 19729
+    },
+    {
+      "epoch": 0.8346729841780185,
+      "grad_norm": 0.1327977031469345,
+      "learning_rate": 0.001,
+      "loss": 2.9165,
+      "step": 19730
+    },
+    {
+      "epoch": 0.8347152889415348,
+      "grad_norm": 0.16689974069595337,
+      "learning_rate": 0.001,
+      "loss": 2.0789,
+      "step": 19731
+    },
+    {
+      "epoch": 0.8347575937050512,
+      "grad_norm": 0.17344988882541656,
+      "learning_rate": 0.001,
+      "loss": 2.5676,
+      "step": 19732
+    },
+    {
+      "epoch": 0.8347998984685676,
+      "grad_norm": 0.14860732853412628,
+      "learning_rate": 0.001,
+      "loss": 1.6259,
+      "step": 19733
+    },
+    {
+      "epoch": 0.8348422032320839,
+      "grad_norm": 0.15001067519187927,
+      "learning_rate": 0.001,
+      "loss": 1.5225,
+      "step": 19734
+    },
+    {
+      "epoch": 0.8348845079956003,
+      "grad_norm": 0.16379481554031372,
+      "learning_rate": 0.001,
+      "loss": 3.1107,
+      "step": 19735
+    },
+    {
+      "epoch": 0.8349268127591166,
+      "grad_norm": 0.16210788488388062,
+      "learning_rate": 0.001,
+      "loss": 1.9486,
+      "step": 19736
+    },
+    {
+      "epoch": 0.834969117522633,
+      "grad_norm": 0.17542105913162231,
+      "learning_rate": 0.001,
+      "loss": 1.36,
+      "step": 19737
+    },
+    {
+      "epoch": 0.8350114222861494,
+      "grad_norm": 0.168275848031044,
+      "learning_rate": 0.001,
+      "loss": 2.1156,
+      "step": 19738
+    },
+    {
+      "epoch": 0.8350537270496657,
+      "grad_norm": 0.153310164809227,
+      "learning_rate": 0.001,
+      "loss": 2.1324,
+      "step": 19739
+    },
+    {
+      "epoch": 0.8350960318131821,
+      "grad_norm": 1.9082057476043701,
+      "learning_rate": 0.001,
+      "loss": 2.4097,
+      "step": 19740
+    },
+    {
+      "epoch": 0.8351383365766986,
+      "grad_norm": 0.1443844437599182,
+      "learning_rate": 0.001,
+      "loss": 2.4301,
+      "step": 19741
+    },
+    {
+      "epoch": 0.8351806413402149,
+      "grad_norm": 0.14315207302570343,
+      "learning_rate": 0.001,
+      "loss": 1.3867,
+      "step": 19742
+    },
+    {
+      "epoch": 0.8352229461037313,
+      "grad_norm": 0.15107427537441254,
+      "learning_rate": 0.001,
+      "loss": 3.8243,
+      "step": 19743
+    },
+    {
+      "epoch": 0.8352652508672477,
+      "grad_norm": 0.1486080139875412,
+      "learning_rate": 0.001,
+      "loss": 1.3748,
+      "step": 19744
+    },
+    {
+      "epoch": 0.835307555630764,
+      "grad_norm": 0.13850051164627075,
+      "learning_rate": 0.001,
+      "loss": 2.7205,
+      "step": 19745
+    },
+    {
+      "epoch": 0.8353498603942804,
+      "grad_norm": 0.1894737184047699,
+      "learning_rate": 0.001,
+      "loss": 2.0814,
+      "step": 19746
+    },
+    {
+      "epoch": 0.8353921651577968,
+      "grad_norm": 3.34781551361084,
+      "learning_rate": 0.001,
+      "loss": 1.7776,
+      "step": 19747
+    },
+    {
+      "epoch": 0.8354344699213131,
+      "grad_norm": 0.14483730494976044,
+      "learning_rate": 0.001,
+      "loss": 2.1921,
+      "step": 19748
+    },
+    {
+      "epoch": 0.8354767746848295,
+      "grad_norm": 0.15554209053516388,
+      "learning_rate": 0.001,
+      "loss": 2.1626,
+      "step": 19749
+    },
+    {
+      "epoch": 0.8355190794483459,
+      "grad_norm": 0.2667386531829834,
+      "learning_rate": 0.001,
+      "loss": 2.3544,
+      "step": 19750
+    },
+    {
+      "epoch": 0.8355613842118622,
+      "grad_norm": 0.17081429064273834,
+      "learning_rate": 0.001,
+      "loss": 3.0333,
+      "step": 19751
+    },
+    {
+      "epoch": 0.8356036889753786,
+      "grad_norm": 0.18492451310157776,
+      "learning_rate": 0.001,
+      "loss": 2.3743,
+      "step": 19752
+    },
+    {
+      "epoch": 0.835645993738895,
+      "grad_norm": 1.0934829711914062,
+      "learning_rate": 0.001,
+      "loss": 2.1734,
+      "step": 19753
+    },
+    {
+      "epoch": 0.8356882985024113,
+      "grad_norm": 0.2422739565372467,
+      "learning_rate": 0.001,
+      "loss": 1.8901,
+      "step": 19754
+    },
+    {
+      "epoch": 0.8357306032659277,
+      "grad_norm": 0.39902499318122864,
+      "learning_rate": 0.001,
+      "loss": 2.4519,
+      "step": 19755
+    },
+    {
+      "epoch": 0.8357729080294442,
+      "grad_norm": 18.87704849243164,
+      "learning_rate": 0.001,
+      "loss": 4.1573,
+      "step": 19756
+    },
+    {
+      "epoch": 0.8358152127929604,
+      "grad_norm": 0.1568903774023056,
+      "learning_rate": 0.001,
+      "loss": 1.8906,
+      "step": 19757
+    },
+    {
+      "epoch": 0.8358575175564769,
+      "grad_norm": 0.2181655913591385,
+      "learning_rate": 0.001,
+      "loss": 2.7009,
+      "step": 19758
+    },
+    {
+      "epoch": 0.8358998223199933,
+      "grad_norm": 0.14595481753349304,
+      "learning_rate": 0.001,
+      "loss": 1.4677,
+      "step": 19759
+    },
+    {
+      "epoch": 0.8359421270835096,
+      "grad_norm": 0.18573489785194397,
+      "learning_rate": 0.001,
+      "loss": 1.9539,
+      "step": 19760
+    },
+    {
+      "epoch": 0.835984431847026,
+      "grad_norm": 0.19582517445087433,
+      "learning_rate": 0.001,
+      "loss": 1.9018,
+      "step": 19761
+    },
+    {
+      "epoch": 0.8360267366105424,
+      "grad_norm": 0.18559232354164124,
+      "learning_rate": 0.001,
+      "loss": 2.0573,
+      "step": 19762
+    },
+    {
+      "epoch": 0.8360690413740587,
+      "grad_norm": 0.2434457689523697,
+      "learning_rate": 0.001,
+      "loss": 2.4941,
+      "step": 19763
+    },
+    {
+      "epoch": 0.8361113461375751,
+      "grad_norm": 0.24671140313148499,
+      "learning_rate": 0.001,
+      "loss": 2.2932,
+      "step": 19764
+    },
+    {
+      "epoch": 0.8361536509010915,
+      "grad_norm": 55.95083999633789,
+      "learning_rate": 0.001,
+      "loss": 1.4447,
+      "step": 19765
+    },
+    {
+      "epoch": 0.8361959556646078,
+      "grad_norm": 0.18943527340888977,
+      "learning_rate": 0.001,
+      "loss": 1.8799,
+      "step": 19766
+    },
+    {
+      "epoch": 0.8362382604281242,
+      "grad_norm": 0.248708114027977,
+      "learning_rate": 0.001,
+      "loss": 2.1987,
+      "step": 19767
+    },
+    {
+      "epoch": 0.8362805651916406,
+      "grad_norm": 0.2755674719810486,
+      "learning_rate": 0.001,
+      "loss": 2.5924,
+      "step": 19768
+    },
+    {
+      "epoch": 0.8363228699551569,
+      "grad_norm": 0.8433470726013184,
+      "learning_rate": 0.001,
+      "loss": 2.656,
+      "step": 19769
+    },
+    {
+      "epoch": 0.8363651747186733,
+      "grad_norm": 0.25172102451324463,
+      "learning_rate": 0.001,
+      "loss": 2.0833,
+      "step": 19770
+    },
+    {
+      "epoch": 0.8364074794821897,
+      "grad_norm": 0.20284956693649292,
+      "learning_rate": 0.001,
+      "loss": 1.7729,
+      "step": 19771
+    },
+    {
+      "epoch": 0.836449784245706,
+      "grad_norm": 0.16717220842838287,
+      "learning_rate": 0.001,
+      "loss": 1.6458,
+      "step": 19772
+    },
+    {
+      "epoch": 0.8364920890092225,
+      "grad_norm": 0.18188956379890442,
+      "learning_rate": 0.001,
+      "loss": 2.6662,
+      "step": 19773
+    },
+    {
+      "epoch": 0.8365343937727389,
+      "grad_norm": 0.19936378300189972,
+      "learning_rate": 0.001,
+      "loss": 2.1813,
+      "step": 19774
+    },
+    {
+      "epoch": 0.8365766985362552,
+      "grad_norm": 0.24121548235416412,
+      "learning_rate": 0.001,
+      "loss": 2.8356,
+      "step": 19775
+    },
+    {
+      "epoch": 0.8366190032997716,
+      "grad_norm": 2.9213600158691406,
+      "learning_rate": 0.001,
+      "loss": 2.4832,
+      "step": 19776
+    },
+    {
+      "epoch": 0.836661308063288,
+      "grad_norm": 0.16870221495628357,
+      "learning_rate": 0.001,
+      "loss": 2.8586,
+      "step": 19777
+    },
+    {
+      "epoch": 0.8367036128268043,
+      "grad_norm": 0.1627093255519867,
+      "learning_rate": 0.001,
+      "loss": 1.8701,
+      "step": 19778
+    },
+    {
+      "epoch": 0.8367459175903207,
+      "grad_norm": 0.18144871294498444,
+      "learning_rate": 0.001,
+      "loss": 2.6356,
+      "step": 19779
+    },
+    {
+      "epoch": 0.8367882223538371,
+      "grad_norm": 0.20875905454158783,
+      "learning_rate": 0.001,
+      "loss": 2.1233,
+      "step": 19780
+    },
+    {
+      "epoch": 0.8368305271173534,
+      "grad_norm": 0.17586883902549744,
+      "learning_rate": 0.001,
+      "loss": 2.2282,
+      "step": 19781
+    },
+    {
+      "epoch": 0.8368728318808698,
+      "grad_norm": 0.17279373109340668,
+      "learning_rate": 0.001,
+      "loss": 1.6294,
+      "step": 19782
+    },
+    {
+      "epoch": 0.8369151366443861,
+      "grad_norm": 0.19814880192279816,
+      "learning_rate": 0.001,
+      "loss": 1.8659,
+      "step": 19783
+    },
+    {
+      "epoch": 0.8369574414079025,
+      "grad_norm": 0.3138713538646698,
+      "learning_rate": 0.001,
+      "loss": 2.3541,
+      "step": 19784
+    },
+    {
+      "epoch": 0.8369997461714189,
+      "grad_norm": 0.15233896672725677,
+      "learning_rate": 0.001,
+      "loss": 2.0767,
+      "step": 19785
+    },
+    {
+      "epoch": 0.8370420509349352,
+      "grad_norm": 9.646958351135254,
+      "learning_rate": 0.001,
+      "loss": 2.3433,
+      "step": 19786
+    },
+    {
+      "epoch": 0.8370843556984516,
+      "grad_norm": 0.3921230733394623,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 19787
+    },
+    {
+      "epoch": 0.837126660461968,
+      "grad_norm": 0.17926689982414246,
+      "learning_rate": 0.001,
+      "loss": 2.008,
+      "step": 19788
+    },
+    {
+      "epoch": 0.8371689652254843,
+      "grad_norm": 0.17443296313285828,
+      "learning_rate": 0.001,
+      "loss": 2.0443,
+      "step": 19789
+    },
+    {
+      "epoch": 0.8372112699890008,
+      "grad_norm": 0.1694057285785675,
+      "learning_rate": 0.001,
+      "loss": 1.7924,
+      "step": 19790
+    },
+    {
+      "epoch": 0.8372535747525172,
+      "grad_norm": 0.16491025686264038,
+      "learning_rate": 0.001,
+      "loss": 2.1166,
+      "step": 19791
+    },
+    {
+      "epoch": 0.8372958795160335,
+      "grad_norm": 0.14807593822479248,
+      "learning_rate": 0.001,
+      "loss": 1.6911,
+      "step": 19792
+    },
+    {
+      "epoch": 0.8373381842795499,
+      "grad_norm": 0.19752711057662964,
+      "learning_rate": 0.001,
+      "loss": 2.2279,
+      "step": 19793
+    },
+    {
+      "epoch": 0.8373804890430663,
+      "grad_norm": 0.1440856158733368,
+      "learning_rate": 0.001,
+      "loss": 2.3304,
+      "step": 19794
+    },
+    {
+      "epoch": 0.8374227938065826,
+      "grad_norm": 0.2171926349401474,
+      "learning_rate": 0.001,
+      "loss": 2.6035,
+      "step": 19795
+    },
+    {
+      "epoch": 0.837465098570099,
+      "grad_norm": 0.1901613473892212,
+      "learning_rate": 0.001,
+      "loss": 2.6214,
+      "step": 19796
+    },
+    {
+      "epoch": 0.8375074033336154,
+      "grad_norm": 0.15385551750659943,
+      "learning_rate": 0.001,
+      "loss": 1.7298,
+      "step": 19797
+    },
+    {
+      "epoch": 0.8375497080971317,
+      "grad_norm": 1.251185655593872,
+      "learning_rate": 0.001,
+      "loss": 2.613,
+      "step": 19798
+    },
+    {
+      "epoch": 0.8375920128606481,
+      "grad_norm": 0.14876984059810638,
+      "learning_rate": 0.001,
+      "loss": 2.2236,
+      "step": 19799
+    },
+    {
+      "epoch": 0.8376343176241645,
+      "grad_norm": 0.19195450842380524,
+      "learning_rate": 0.001,
+      "loss": 2.2847,
+      "step": 19800
+    },
+    {
+      "epoch": 0.8376766223876808,
+      "grad_norm": 0.14749985933303833,
+      "learning_rate": 0.001,
+      "loss": 1.8272,
+      "step": 19801
+    },
+    {
+      "epoch": 0.8377189271511972,
+      "grad_norm": 0.18800973892211914,
+      "learning_rate": 0.001,
+      "loss": 1.738,
+      "step": 19802
+    },
+    {
+      "epoch": 0.8377612319147136,
+      "grad_norm": 27.659873962402344,
+      "learning_rate": 0.001,
+      "loss": 2.2451,
+      "step": 19803
+    },
+    {
+      "epoch": 0.8378035366782299,
+      "grad_norm": 0.16692310571670532,
+      "learning_rate": 0.001,
+      "loss": 2.4913,
+      "step": 19804
+    },
+    {
+      "epoch": 0.8378458414417463,
+      "grad_norm": 0.2250969558954239,
+      "learning_rate": 0.001,
+      "loss": 2.685,
+      "step": 19805
+    },
+    {
+      "epoch": 0.8378881462052628,
+      "grad_norm": 0.13324400782585144,
+      "learning_rate": 0.001,
+      "loss": 2.6045,
+      "step": 19806
+    },
+    {
+      "epoch": 0.837930450968779,
+      "grad_norm": 2.1039130687713623,
+      "learning_rate": 0.001,
+      "loss": 2.5165,
+      "step": 19807
+    },
+    {
+      "epoch": 0.8379727557322955,
+      "grad_norm": 0.14113382995128632,
+      "learning_rate": 0.001,
+      "loss": 1.6369,
+      "step": 19808
+    },
+    {
+      "epoch": 0.8380150604958119,
+      "grad_norm": 0.15121187269687653,
+      "learning_rate": 0.001,
+      "loss": 1.6652,
+      "step": 19809
+    },
+    {
+      "epoch": 0.8380573652593282,
+      "grad_norm": 0.15335388481616974,
+      "learning_rate": 0.001,
+      "loss": 2.2542,
+      "step": 19810
+    },
+    {
+      "epoch": 0.8380996700228446,
+      "grad_norm": 3.9381353855133057,
+      "learning_rate": 0.001,
+      "loss": 2.8601,
+      "step": 19811
+    },
+    {
+      "epoch": 0.838141974786361,
+      "grad_norm": 0.7249288558959961,
+      "learning_rate": 0.001,
+      "loss": 2.4747,
+      "step": 19812
+    },
+    {
+      "epoch": 0.8381842795498773,
+      "grad_norm": 6.57703971862793,
+      "learning_rate": 0.001,
+      "loss": 1.7587,
+      "step": 19813
+    },
+    {
+      "epoch": 0.8382265843133937,
+      "grad_norm": 0.7364828586578369,
+      "learning_rate": 0.001,
+      "loss": 3.9219,
+      "step": 19814
+    },
+    {
+      "epoch": 0.8382688890769101,
+      "grad_norm": 0.18288707733154297,
+      "learning_rate": 0.001,
+      "loss": 2.7903,
+      "step": 19815
+    },
+    {
+      "epoch": 0.8383111938404264,
+      "grad_norm": 0.1727524846792221,
+      "learning_rate": 0.001,
+      "loss": 1.9508,
+      "step": 19816
+    },
+    {
+      "epoch": 0.8383534986039428,
+      "grad_norm": 0.1617797166109085,
+      "learning_rate": 0.001,
+      "loss": 3.1475,
+      "step": 19817
+    },
+    {
+      "epoch": 0.8383958033674592,
+      "grad_norm": 0.16539444029331207,
+      "learning_rate": 0.001,
+      "loss": 2.1155,
+      "step": 19818
+    },
+    {
+      "epoch": 0.8384381081309755,
+      "grad_norm": 4.960668563842773,
+      "learning_rate": 0.001,
+      "loss": 2.3467,
+      "step": 19819
+    },
+    {
+      "epoch": 0.8384804128944919,
+      "grad_norm": 0.20643018186092377,
+      "learning_rate": 0.001,
+      "loss": 2.0452,
+      "step": 19820
+    },
+    {
+      "epoch": 0.8385227176580083,
+      "grad_norm": 2.9467222690582275,
+      "learning_rate": 0.001,
+      "loss": 2.9591,
+      "step": 19821
+    },
+    {
+      "epoch": 0.8385650224215246,
+      "grad_norm": 0.13099977374076843,
+      "learning_rate": 0.001,
+      "loss": 1.5804,
+      "step": 19822
+    },
+    {
+      "epoch": 0.838607327185041,
+      "grad_norm": 0.15258918702602386,
+      "learning_rate": 0.001,
+      "loss": 2.4391,
+      "step": 19823
+    },
+    {
+      "epoch": 0.8386496319485575,
+      "grad_norm": 0.18933935463428497,
+      "learning_rate": 0.001,
+      "loss": 2.2122,
+      "step": 19824
+    },
+    {
+      "epoch": 0.8386919367120738,
+      "grad_norm": 0.14285174012184143,
+      "learning_rate": 0.001,
+      "loss": 2.3817,
+      "step": 19825
+    },
+    {
+      "epoch": 0.8387342414755902,
+      "grad_norm": 1.277612328529358,
+      "learning_rate": 0.001,
+      "loss": 2.2287,
+      "step": 19826
+    },
+    {
+      "epoch": 0.8387765462391065,
+      "grad_norm": 0.12771695852279663,
+      "learning_rate": 0.001,
+      "loss": 1.9689,
+      "step": 19827
+    },
+    {
+      "epoch": 0.8388188510026229,
+      "grad_norm": 0.16007982194423676,
+      "learning_rate": 0.001,
+      "loss": 2.2604,
+      "step": 19828
+    },
+    {
+      "epoch": 0.8388611557661393,
+      "grad_norm": 0.625321090221405,
+      "learning_rate": 0.001,
+      "loss": 1.6921,
+      "step": 19829
+    },
+    {
+      "epoch": 0.8389034605296556,
+      "grad_norm": 0.44362181425094604,
+      "learning_rate": 0.001,
+      "loss": 2.2605,
+      "step": 19830
+    },
+    {
+      "epoch": 0.838945765293172,
+      "grad_norm": 0.16147875785827637,
+      "learning_rate": 0.001,
+      "loss": 1.9878,
+      "step": 19831
+    },
+    {
+      "epoch": 0.8389880700566884,
+      "grad_norm": 0.37815141677856445,
+      "learning_rate": 0.001,
+      "loss": 2.8435,
+      "step": 19832
+    },
+    {
+      "epoch": 0.8390303748202047,
+      "grad_norm": 0.15180058777332306,
+      "learning_rate": 0.001,
+      "loss": 1.7832,
+      "step": 19833
+    },
+    {
+      "epoch": 0.8390726795837211,
+      "grad_norm": 0.16070739924907684,
+      "learning_rate": 0.001,
+      "loss": 2.4271,
+      "step": 19834
+    },
+    {
+      "epoch": 0.8391149843472375,
+      "grad_norm": 2.378019332885742,
+      "learning_rate": 0.001,
+      "loss": 2.1971,
+      "step": 19835
+    },
+    {
+      "epoch": 0.8391572891107538,
+      "grad_norm": 0.13618861138820648,
+      "learning_rate": 0.001,
+      "loss": 2.482,
+      "step": 19836
+    },
+    {
+      "epoch": 0.8391995938742702,
+      "grad_norm": 0.1882878690958023,
+      "learning_rate": 0.001,
+      "loss": 2.0153,
+      "step": 19837
+    },
+    {
+      "epoch": 0.8392418986377866,
+      "grad_norm": 0.24582351744174957,
+      "learning_rate": 0.001,
+      "loss": 2.3494,
+      "step": 19838
+    },
+    {
+      "epoch": 0.8392842034013029,
+      "grad_norm": 0.22917570173740387,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 19839
+    },
+    {
+      "epoch": 0.8393265081648194,
+      "grad_norm": 0.15284200012683868,
+      "learning_rate": 0.001,
+      "loss": 2.142,
+      "step": 19840
+    },
+    {
+      "epoch": 0.8393688129283358,
+      "grad_norm": 0.15016065537929535,
+      "learning_rate": 0.001,
+      "loss": 1.9708,
+      "step": 19841
+    },
+    {
+      "epoch": 0.8394111176918521,
+      "grad_norm": 0.17199917137622833,
+      "learning_rate": 0.001,
+      "loss": 2.0482,
+      "step": 19842
+    },
+    {
+      "epoch": 0.8394534224553685,
+      "grad_norm": 2.0709526538848877,
+      "learning_rate": 0.001,
+      "loss": 2.1923,
+      "step": 19843
+    },
+    {
+      "epoch": 0.8394957272188849,
+      "grad_norm": 0.14755693078041077,
+      "learning_rate": 0.001,
+      "loss": 2.07,
+      "step": 19844
+    },
+    {
+      "epoch": 0.8395380319824012,
+      "grad_norm": 0.45920902490615845,
+      "learning_rate": 0.001,
+      "loss": 3.1264,
+      "step": 19845
+    },
+    {
+      "epoch": 0.8395803367459176,
+      "grad_norm": 0.1509585976600647,
+      "learning_rate": 0.001,
+      "loss": 1.4513,
+      "step": 19846
+    },
+    {
+      "epoch": 0.839622641509434,
+      "grad_norm": 0.15342389047145844,
+      "learning_rate": 0.001,
+      "loss": 2.4979,
+      "step": 19847
+    },
+    {
+      "epoch": 0.8396649462729503,
+      "grad_norm": 0.16700981557369232,
+      "learning_rate": 0.001,
+      "loss": 1.6099,
+      "step": 19848
+    },
+    {
+      "epoch": 0.8397072510364667,
+      "grad_norm": 0.200181245803833,
+      "learning_rate": 0.001,
+      "loss": 1.9247,
+      "step": 19849
+    },
+    {
+      "epoch": 0.8397495557999831,
+      "grad_norm": 0.13869711756706238,
+      "learning_rate": 0.001,
+      "loss": 1.6663,
+      "step": 19850
+    },
+    {
+      "epoch": 0.8397918605634994,
+      "grad_norm": 0.19390735030174255,
+      "learning_rate": 0.001,
+      "loss": 2.6041,
+      "step": 19851
+    },
+    {
+      "epoch": 0.8398341653270158,
+      "grad_norm": 0.15957386791706085,
+      "learning_rate": 0.001,
+      "loss": 1.6804,
+      "step": 19852
+    },
+    {
+      "epoch": 0.8398764700905322,
+      "grad_norm": 0.1435195654630661,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 19853
+    },
+    {
+      "epoch": 0.8399187748540485,
+      "grad_norm": 0.4569486379623413,
+      "learning_rate": 0.001,
+      "loss": 1.7374,
+      "step": 19854
+    },
+    {
+      "epoch": 0.8399610796175649,
+      "grad_norm": 0.6527416706085205,
+      "learning_rate": 0.001,
+      "loss": 1.8554,
+      "step": 19855
+    },
+    {
+      "epoch": 0.8400033843810814,
+      "grad_norm": 0.36150479316711426,
+      "learning_rate": 0.001,
+      "loss": 2.1568,
+      "step": 19856
+    },
+    {
+      "epoch": 0.8400456891445977,
+      "grad_norm": 0.16183467209339142,
+      "learning_rate": 0.001,
+      "loss": 2.5408,
+      "step": 19857
+    },
+    {
+      "epoch": 0.8400879939081141,
+      "grad_norm": 0.14693614840507507,
+      "learning_rate": 0.001,
+      "loss": 1.7957,
+      "step": 19858
+    },
+    {
+      "epoch": 0.8401302986716305,
+      "grad_norm": 0.16061703860759735,
+      "learning_rate": 0.001,
+      "loss": 1.9237,
+      "step": 19859
+    },
+    {
+      "epoch": 0.8401726034351468,
+      "grad_norm": 0.1884782463312149,
+      "learning_rate": 0.001,
+      "loss": 1.936,
+      "step": 19860
+    },
+    {
+      "epoch": 0.8402149081986632,
+      "grad_norm": 0.15317420661449432,
+      "learning_rate": 0.001,
+      "loss": 1.857,
+      "step": 19861
+    },
+    {
+      "epoch": 0.8402572129621796,
+      "grad_norm": 0.17697061598300934,
+      "learning_rate": 0.001,
+      "loss": 2.4018,
+      "step": 19862
+    },
+    {
+      "epoch": 0.8402995177256959,
+      "grad_norm": 0.14646954834461212,
+      "learning_rate": 0.001,
+      "loss": 1.6583,
+      "step": 19863
+    },
+    {
+      "epoch": 0.8403418224892123,
+      "grad_norm": 0.1281052678823471,
+      "learning_rate": 0.001,
+      "loss": 3.3,
+      "step": 19864
+    },
+    {
+      "epoch": 0.8403841272527287,
+      "grad_norm": 0.12583167850971222,
+      "learning_rate": 0.001,
+      "loss": 2.1985,
+      "step": 19865
+    },
+    {
+      "epoch": 0.840426432016245,
+      "grad_norm": 0.16362233459949493,
+      "learning_rate": 0.001,
+      "loss": 1.9579,
+      "step": 19866
+    },
+    {
+      "epoch": 0.8404687367797614,
+      "grad_norm": 14.114072799682617,
+      "learning_rate": 0.001,
+      "loss": 2.0031,
+      "step": 19867
+    },
+    {
+      "epoch": 0.8405110415432778,
+      "grad_norm": 1.4896159172058105,
+      "learning_rate": 0.001,
+      "loss": 2.8004,
+      "step": 19868
+    },
+    {
+      "epoch": 0.8405533463067941,
+      "grad_norm": 0.15967680513858795,
+      "learning_rate": 0.001,
+      "loss": 2.0756,
+      "step": 19869
+    },
+    {
+      "epoch": 0.8405956510703105,
+      "grad_norm": 0.1979799121618271,
+      "learning_rate": 0.001,
+      "loss": 2.3233,
+      "step": 19870
+    },
+    {
+      "epoch": 0.8406379558338269,
+      "grad_norm": 0.3304239809513092,
+      "learning_rate": 0.001,
+      "loss": 2.0972,
+      "step": 19871
+    },
+    {
+      "epoch": 0.8406802605973432,
+      "grad_norm": 0.13525773584842682,
+      "learning_rate": 0.001,
+      "loss": 1.4994,
+      "step": 19872
+    },
+    {
+      "epoch": 0.8407225653608597,
+      "grad_norm": 1.1187829971313477,
+      "learning_rate": 0.001,
+      "loss": 2.6048,
+      "step": 19873
+    },
+    {
+      "epoch": 0.840764870124376,
+      "grad_norm": 1.9175190925598145,
+      "learning_rate": 0.001,
+      "loss": 2.1303,
+      "step": 19874
+    },
+    {
+      "epoch": 0.8408071748878924,
+      "grad_norm": 0.17562279105186462,
+      "learning_rate": 0.001,
+      "loss": 2.3723,
+      "step": 19875
+    },
+    {
+      "epoch": 0.8408494796514088,
+      "grad_norm": 0.14081628620624542,
+      "learning_rate": 0.001,
+      "loss": 2.3064,
+      "step": 19876
+    },
+    {
+      "epoch": 0.8408917844149251,
+      "grad_norm": 0.4894697368144989,
+      "learning_rate": 0.001,
+      "loss": 2.8948,
+      "step": 19877
+    },
+    {
+      "epoch": 0.8409340891784415,
+      "grad_norm": 0.19866247475147247,
+      "learning_rate": 0.001,
+      "loss": 2.9252,
+      "step": 19878
+    },
+    {
+      "epoch": 0.8409763939419579,
+      "grad_norm": 0.4068894684314728,
+      "learning_rate": 0.001,
+      "loss": 2.2374,
+      "step": 19879
+    },
+    {
+      "epoch": 0.8410186987054742,
+      "grad_norm": 0.1639883816242218,
+      "learning_rate": 0.001,
+      "loss": 2.0532,
+      "step": 19880
+    },
+    {
+      "epoch": 0.8410610034689906,
+      "grad_norm": 0.3562491536140442,
+      "learning_rate": 0.001,
+      "loss": 4.0336,
+      "step": 19881
+    },
+    {
+      "epoch": 0.841103308232507,
+      "grad_norm": 0.13724088668823242,
+      "learning_rate": 0.001,
+      "loss": 1.8362,
+      "step": 19882
+    },
+    {
+      "epoch": 0.8411456129960233,
+      "grad_norm": 0.13581812381744385,
+      "learning_rate": 0.001,
+      "loss": 1.3958,
+      "step": 19883
+    },
+    {
+      "epoch": 0.8411879177595397,
+      "grad_norm": 0.17569448053836823,
+      "learning_rate": 0.001,
+      "loss": 2.8263,
+      "step": 19884
+    },
+    {
+      "epoch": 0.8412302225230561,
+      "grad_norm": 0.14061707258224487,
+      "learning_rate": 0.001,
+      "loss": 2.2112,
+      "step": 19885
+    },
+    {
+      "epoch": 0.8412725272865724,
+      "grad_norm": 0.9066682457923889,
+      "learning_rate": 0.001,
+      "loss": 2.0064,
+      "step": 19886
+    },
+    {
+      "epoch": 0.8413148320500888,
+      "grad_norm": 0.13263536989688873,
+      "learning_rate": 0.001,
+      "loss": 2.3227,
+      "step": 19887
+    },
+    {
+      "epoch": 0.8413571368136052,
+      "grad_norm": 0.1647445261478424,
+      "learning_rate": 0.001,
+      "loss": 3.9107,
+      "step": 19888
+    },
+    {
+      "epoch": 0.8413994415771215,
+      "grad_norm": 0.3302064538002014,
+      "learning_rate": 0.001,
+      "loss": 2.5003,
+      "step": 19889
+    },
+    {
+      "epoch": 0.841441746340638,
+      "grad_norm": 0.14488548040390015,
+      "learning_rate": 0.001,
+      "loss": 2.0212,
+      "step": 19890
+    },
+    {
+      "epoch": 0.8414840511041544,
+      "grad_norm": 0.1621946543455124,
+      "learning_rate": 0.001,
+      "loss": 2.5805,
+      "step": 19891
+    },
+    {
+      "epoch": 0.8415263558676707,
+      "grad_norm": 0.13542355597019196,
+      "learning_rate": 0.001,
+      "loss": 1.8591,
+      "step": 19892
+    },
+    {
+      "epoch": 0.8415686606311871,
+      "grad_norm": 0.13882328569889069,
+      "learning_rate": 0.001,
+      "loss": 1.6565,
+      "step": 19893
+    },
+    {
+      "epoch": 0.8416109653947035,
+      "grad_norm": 0.18400336802005768,
+      "learning_rate": 0.001,
+      "loss": 2.3863,
+      "step": 19894
+    },
+    {
+      "epoch": 0.8416532701582198,
+      "grad_norm": 0.14745768904685974,
+      "learning_rate": 0.001,
+      "loss": 1.6539,
+      "step": 19895
+    },
+    {
+      "epoch": 0.8416955749217362,
+      "grad_norm": 0.5481185913085938,
+      "learning_rate": 0.001,
+      "loss": 2.4827,
+      "step": 19896
+    },
+    {
+      "epoch": 0.8417378796852526,
+      "grad_norm": 0.18567869067192078,
+      "learning_rate": 0.001,
+      "loss": 1.7485,
+      "step": 19897
+    },
+    {
+      "epoch": 0.8417801844487689,
+      "grad_norm": 40.28692626953125,
+      "learning_rate": 0.001,
+      "loss": 2.2203,
+      "step": 19898
+    },
+    {
+      "epoch": 0.8418224892122853,
+      "grad_norm": 0.17718671262264252,
+      "learning_rate": 0.001,
+      "loss": 2.2116,
+      "step": 19899
+    },
+    {
+      "epoch": 0.8418647939758017,
+      "grad_norm": 0.20770712196826935,
+      "learning_rate": 0.001,
+      "loss": 1.6705,
+      "step": 19900
+    },
+    {
+      "epoch": 0.841907098739318,
+      "grad_norm": 0.33446425199508667,
+      "learning_rate": 0.001,
+      "loss": 2.2231,
+      "step": 19901
+    },
+    {
+      "epoch": 0.8419494035028344,
+      "grad_norm": 0.15271897614002228,
+      "learning_rate": 0.001,
+      "loss": 2.0796,
+      "step": 19902
+    },
+    {
+      "epoch": 0.8419917082663508,
+      "grad_norm": 0.3776164650917053,
+      "learning_rate": 0.001,
+      "loss": 2.7789,
+      "step": 19903
+    },
+    {
+      "epoch": 0.8420340130298671,
+      "grad_norm": 0.18110236525535583,
+      "learning_rate": 0.001,
+      "loss": 3.1009,
+      "step": 19904
+    },
+    {
+      "epoch": 0.8420763177933835,
+      "grad_norm": 0.14142876863479614,
+      "learning_rate": 0.001,
+      "loss": 1.5932,
+      "step": 19905
+    },
+    {
+      "epoch": 0.8421186225569,
+      "grad_norm": 0.1594587117433548,
+      "learning_rate": 0.001,
+      "loss": 2.9,
+      "step": 19906
+    },
+    {
+      "epoch": 0.8421609273204163,
+      "grad_norm": 0.17610591650009155,
+      "learning_rate": 0.001,
+      "loss": 2.0221,
+      "step": 19907
+    },
+    {
+      "epoch": 0.8422032320839327,
+      "grad_norm": 0.1743817925453186,
+      "learning_rate": 0.001,
+      "loss": 2.0761,
+      "step": 19908
+    },
+    {
+      "epoch": 0.8422455368474491,
+      "grad_norm": 0.3120170831680298,
+      "learning_rate": 0.001,
+      "loss": 3.1248,
+      "step": 19909
+    },
+    {
+      "epoch": 0.8422878416109654,
+      "grad_norm": 0.14950038492679596,
+      "learning_rate": 0.001,
+      "loss": 3.4022,
+      "step": 19910
+    },
+    {
+      "epoch": 0.8423301463744818,
+      "grad_norm": 0.17454718053340912,
+      "learning_rate": 0.001,
+      "loss": 1.8271,
+      "step": 19911
+    },
+    {
+      "epoch": 0.8423724511379982,
+      "grad_norm": 0.18811197578907013,
+      "learning_rate": 0.001,
+      "loss": 2.814,
+      "step": 19912
+    },
+    {
+      "epoch": 0.8424147559015145,
+      "grad_norm": 0.13598325848579407,
+      "learning_rate": 0.001,
+      "loss": 2.043,
+      "step": 19913
+    },
+    {
+      "epoch": 0.8424570606650309,
+      "grad_norm": 0.14503377676010132,
+      "learning_rate": 0.001,
+      "loss": 1.8979,
+      "step": 19914
+    },
+    {
+      "epoch": 0.8424993654285473,
+      "grad_norm": 0.16746678948402405,
+      "learning_rate": 0.001,
+      "loss": 1.9977,
+      "step": 19915
+    },
+    {
+      "epoch": 0.8425416701920636,
+      "grad_norm": 8.220918655395508,
+      "learning_rate": 0.001,
+      "loss": 2.3796,
+      "step": 19916
+    },
+    {
+      "epoch": 0.84258397495558,
+      "grad_norm": 0.20891021192073822,
+      "learning_rate": 0.001,
+      "loss": 2.4267,
+      "step": 19917
+    },
+    {
+      "epoch": 0.8426262797190963,
+      "grad_norm": 0.15309026837348938,
+      "learning_rate": 0.001,
+      "loss": 2.0111,
+      "step": 19918
+    },
+    {
+      "epoch": 0.8426685844826127,
+      "grad_norm": 1.157206654548645,
+      "learning_rate": 0.001,
+      "loss": 3.1127,
+      "step": 19919
+    },
+    {
+      "epoch": 0.8427108892461291,
+      "grad_norm": 0.1353883147239685,
+      "learning_rate": 0.001,
+      "loss": 1.5835,
+      "step": 19920
+    },
+    {
+      "epoch": 0.8427531940096454,
+      "grad_norm": 0.16961653530597687,
+      "learning_rate": 0.001,
+      "loss": 1.4123,
+      "step": 19921
+    },
+    {
+      "epoch": 0.8427954987731618,
+      "grad_norm": 0.15629060566425323,
+      "learning_rate": 0.001,
+      "loss": 1.8995,
+      "step": 19922
+    },
+    {
+      "epoch": 0.8428378035366783,
+      "grad_norm": 14.496636390686035,
+      "learning_rate": 0.001,
+      "loss": 1.9417,
+      "step": 19923
+    },
+    {
+      "epoch": 0.8428801083001946,
+      "grad_norm": 0.16082890331745148,
+      "learning_rate": 0.001,
+      "loss": 1.7762,
+      "step": 19924
+    },
+    {
+      "epoch": 0.842922413063711,
+      "grad_norm": 0.4194250702857971,
+      "learning_rate": 0.001,
+      "loss": 1.7252,
+      "step": 19925
+    },
+    {
+      "epoch": 0.8429647178272274,
+      "grad_norm": 0.25590166449546814,
+      "learning_rate": 0.001,
+      "loss": 2.1403,
+      "step": 19926
+    },
+    {
+      "epoch": 0.8430070225907437,
+      "grad_norm": 0.6089497804641724,
+      "learning_rate": 0.001,
+      "loss": 2.328,
+      "step": 19927
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.17226269841194153,
+      "learning_rate": 0.001,
+      "loss": 1.7442,
+      "step": 19928
+    },
+    {
+      "epoch": 0.8430916321177765,
+      "grad_norm": 12.525866508483887,
+      "learning_rate": 0.001,
+      "loss": 2.9423,
+      "step": 19929
+    },
+    {
+      "epoch": 0.8431339368812928,
+      "grad_norm": 0.15503224730491638,
+      "learning_rate": 0.001,
+      "loss": 2.119,
+      "step": 19930
+    },
+    {
+      "epoch": 0.8431762416448092,
+      "grad_norm": 0.5280601978302002,
+      "learning_rate": 0.001,
+      "loss": 2.2592,
+      "step": 19931
+    },
+    {
+      "epoch": 0.8432185464083256,
+      "grad_norm": 0.19108590483665466,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 19932
+    },
+    {
+      "epoch": 0.8432608511718419,
+      "grad_norm": 0.2428058534860611,
+      "learning_rate": 0.001,
+      "loss": 3.0192,
+      "step": 19933
+    },
+    {
+      "epoch": 0.8433031559353583,
+      "grad_norm": 0.16223500669002533,
+      "learning_rate": 0.001,
+      "loss": 2.3037,
+      "step": 19934
+    },
+    {
+      "epoch": 0.8433454606988747,
+      "grad_norm": 5.290875434875488,
+      "learning_rate": 0.001,
+      "loss": 2.3735,
+      "step": 19935
+    },
+    {
+      "epoch": 0.843387765462391,
+      "grad_norm": 0.139143168926239,
+      "learning_rate": 0.001,
+      "loss": 1.7609,
+      "step": 19936
+    },
+    {
+      "epoch": 0.8434300702259074,
+      "grad_norm": 0.2556076645851135,
+      "learning_rate": 0.001,
+      "loss": 2.5694,
+      "step": 19937
+    },
+    {
+      "epoch": 0.8434723749894238,
+      "grad_norm": 0.17417319118976593,
+      "learning_rate": 0.001,
+      "loss": 2.2479,
+      "step": 19938
+    },
+    {
+      "epoch": 0.8435146797529401,
+      "grad_norm": 0.15048249065876007,
+      "learning_rate": 0.001,
+      "loss": 2.2647,
+      "step": 19939
+    },
+    {
+      "epoch": 0.8435569845164566,
+      "grad_norm": 0.32785776257514954,
+      "learning_rate": 0.001,
+      "loss": 2.3088,
+      "step": 19940
+    },
+    {
+      "epoch": 0.843599289279973,
+      "grad_norm": 0.16994161903858185,
+      "learning_rate": 0.001,
+      "loss": 2.9571,
+      "step": 19941
+    },
+    {
+      "epoch": 0.8436415940434893,
+      "grad_norm": 0.16292929649353027,
+      "learning_rate": 0.001,
+      "loss": 2.268,
+      "step": 19942
+    },
+    {
+      "epoch": 0.8436838988070057,
+      "grad_norm": 0.1685224175453186,
+      "learning_rate": 0.001,
+      "loss": 3.1586,
+      "step": 19943
+    },
+    {
+      "epoch": 0.8437262035705221,
+      "grad_norm": 0.2571621239185333,
+      "learning_rate": 0.001,
+      "loss": 2.2422,
+      "step": 19944
+    },
+    {
+      "epoch": 0.8437685083340384,
+      "grad_norm": 0.1383311152458191,
+      "learning_rate": 0.001,
+      "loss": 1.6768,
+      "step": 19945
+    },
+    {
+      "epoch": 0.8438108130975548,
+      "grad_norm": 0.1316850483417511,
+      "learning_rate": 0.001,
+      "loss": 1.3501,
+      "step": 19946
+    },
+    {
+      "epoch": 0.8438531178610712,
+      "grad_norm": 0.1473502814769745,
+      "learning_rate": 0.001,
+      "loss": 2.4567,
+      "step": 19947
+    },
+    {
+      "epoch": 0.8438954226245875,
+      "grad_norm": 0.2394011914730072,
+      "learning_rate": 0.001,
+      "loss": 1.5157,
+      "step": 19948
+    },
+    {
+      "epoch": 0.8439377273881039,
+      "grad_norm": 0.21005521714687347,
+      "learning_rate": 0.001,
+      "loss": 3.4788,
+      "step": 19949
+    },
+    {
+      "epoch": 0.8439800321516203,
+      "grad_norm": 0.22676526010036469,
+      "learning_rate": 0.001,
+      "loss": 1.5542,
+      "step": 19950
+    },
+    {
+      "epoch": 0.8440223369151366,
+      "grad_norm": 0.19313664734363556,
+      "learning_rate": 0.001,
+      "loss": 2.9371,
+      "step": 19951
+    },
+    {
+      "epoch": 0.844064641678653,
+      "grad_norm": 0.18418525159358978,
+      "learning_rate": 0.001,
+      "loss": 3.3624,
+      "step": 19952
+    },
+    {
+      "epoch": 0.8441069464421694,
+      "grad_norm": 0.15955480933189392,
+      "learning_rate": 0.001,
+      "loss": 2.4359,
+      "step": 19953
+    },
+    {
+      "epoch": 0.8441492512056857,
+      "grad_norm": 0.15745678544044495,
+      "learning_rate": 0.001,
+      "loss": 2.4083,
+      "step": 19954
+    },
+    {
+      "epoch": 0.8441915559692021,
+      "grad_norm": 0.1493915468454361,
+      "learning_rate": 0.001,
+      "loss": 2.0375,
+      "step": 19955
+    },
+    {
+      "epoch": 0.8442338607327186,
+      "grad_norm": 0.13950100541114807,
+      "learning_rate": 0.001,
+      "loss": 2.7912,
+      "step": 19956
+    },
+    {
+      "epoch": 0.8442761654962349,
+      "grad_norm": 0.47593954205513,
+      "learning_rate": 0.001,
+      "loss": 1.4144,
+      "step": 19957
+    },
+    {
+      "epoch": 0.8443184702597513,
+      "grad_norm": 0.17434100806713104,
+      "learning_rate": 0.001,
+      "loss": 2.265,
+      "step": 19958
+    },
+    {
+      "epoch": 0.8443607750232677,
+      "grad_norm": 0.12682586908340454,
+      "learning_rate": 0.001,
+      "loss": 2.5613,
+      "step": 19959
+    },
+    {
+      "epoch": 0.844403079786784,
+      "grad_norm": 1.8817963600158691,
+      "learning_rate": 0.001,
+      "loss": 1.676,
+      "step": 19960
+    },
+    {
+      "epoch": 0.8444453845503004,
+      "grad_norm": 0.19121527671813965,
+      "learning_rate": 0.001,
+      "loss": 1.512,
+      "step": 19961
+    },
+    {
+      "epoch": 0.8444876893138167,
+      "grad_norm": 0.1444321870803833,
+      "learning_rate": 0.001,
+      "loss": 1.92,
+      "step": 19962
+    },
+    {
+      "epoch": 0.8445299940773331,
+      "grad_norm": 0.16925615072250366,
+      "learning_rate": 0.001,
+      "loss": 1.6687,
+      "step": 19963
+    },
+    {
+      "epoch": 0.8445722988408495,
+      "grad_norm": 4.560879707336426,
+      "learning_rate": 0.001,
+      "loss": 2.2629,
+      "step": 19964
+    },
+    {
+      "epoch": 0.8446146036043658,
+      "grad_norm": 0.13227243721485138,
+      "learning_rate": 0.001,
+      "loss": 2.5459,
+      "step": 19965
+    },
+    {
+      "epoch": 0.8446569083678822,
+      "grad_norm": 0.15780355036258698,
+      "learning_rate": 0.001,
+      "loss": 2.7421,
+      "step": 19966
+    },
+    {
+      "epoch": 0.8446992131313986,
+      "grad_norm": 0.1389780193567276,
+      "learning_rate": 0.001,
+      "loss": 2.5234,
+      "step": 19967
+    },
+    {
+      "epoch": 0.8447415178949149,
+      "grad_norm": 0.15538786351680756,
+      "learning_rate": 0.001,
+      "loss": 2.0989,
+      "step": 19968
+    },
+    {
+      "epoch": 0.8447838226584313,
+      "grad_norm": 9.396347999572754,
+      "learning_rate": 0.001,
+      "loss": 1.7852,
+      "step": 19969
+    },
+    {
+      "epoch": 0.8448261274219477,
+      "grad_norm": 0.2267782837152481,
+      "learning_rate": 0.001,
+      "loss": 1.8979,
+      "step": 19970
+    },
+    {
+      "epoch": 0.844868432185464,
+      "grad_norm": 0.1871020346879959,
+      "learning_rate": 0.001,
+      "loss": 3.3206,
+      "step": 19971
+    },
+    {
+      "epoch": 0.8449107369489804,
+      "grad_norm": 0.17954017221927643,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 19972
+    },
+    {
+      "epoch": 0.8449530417124969,
+      "grad_norm": 0.146229088306427,
+      "learning_rate": 0.001,
+      "loss": 1.5802,
+      "step": 19973
+    },
+    {
+      "epoch": 0.8449953464760132,
+      "grad_norm": 0.1864139437675476,
+      "learning_rate": 0.001,
+      "loss": 2.0258,
+      "step": 19974
+    },
+    {
+      "epoch": 0.8450376512395296,
+      "grad_norm": 0.1569264978170395,
+      "learning_rate": 0.001,
+      "loss": 1.9869,
+      "step": 19975
+    },
+    {
+      "epoch": 0.845079956003046,
+      "grad_norm": 0.15957815945148468,
+      "learning_rate": 0.001,
+      "loss": 2.4842,
+      "step": 19976
+    },
+    {
+      "epoch": 0.8451222607665623,
+      "grad_norm": 0.16529959440231323,
+      "learning_rate": 0.001,
+      "loss": 1.9408,
+      "step": 19977
+    },
+    {
+      "epoch": 0.8451645655300787,
+      "grad_norm": 0.1697777956724167,
+      "learning_rate": 0.001,
+      "loss": 2.1509,
+      "step": 19978
+    },
+    {
+      "epoch": 0.8452068702935951,
+      "grad_norm": 1.7855709791183472,
+      "learning_rate": 0.001,
+      "loss": 3.2382,
+      "step": 19979
+    },
+    {
+      "epoch": 0.8452491750571114,
+      "grad_norm": 0.17001429200172424,
+      "learning_rate": 0.001,
+      "loss": 2.8689,
+      "step": 19980
+    },
+    {
+      "epoch": 0.8452914798206278,
+      "grad_norm": 0.17882667481899261,
+      "learning_rate": 0.001,
+      "loss": 1.466,
+      "step": 19981
+    },
+    {
+      "epoch": 0.8453337845841442,
+      "grad_norm": 3.9639627933502197,
+      "learning_rate": 0.001,
+      "loss": 1.8731,
+      "step": 19982
+    },
+    {
+      "epoch": 0.8453760893476605,
+      "grad_norm": 0.14277885854244232,
+      "learning_rate": 0.001,
+      "loss": 1.8745,
+      "step": 19983
+    },
+    {
+      "epoch": 0.8454183941111769,
+      "grad_norm": 0.17287899553775787,
+      "learning_rate": 0.001,
+      "loss": 2.4062,
+      "step": 19984
+    },
+    {
+      "epoch": 0.8454606988746933,
+      "grad_norm": 0.1661319136619568,
+      "learning_rate": 0.001,
+      "loss": 1.7102,
+      "step": 19985
+    },
+    {
+      "epoch": 0.8455030036382096,
+      "grad_norm": 0.1925491839647293,
+      "learning_rate": 0.001,
+      "loss": 2.2246,
+      "step": 19986
+    },
+    {
+      "epoch": 0.845545308401726,
+      "grad_norm": 0.15277276933193207,
+      "learning_rate": 0.001,
+      "loss": 2.5972,
+      "step": 19987
+    },
+    {
+      "epoch": 0.8455876131652424,
+      "grad_norm": 0.1781436949968338,
+      "learning_rate": 0.001,
+      "loss": 3.143,
+      "step": 19988
+    },
+    {
+      "epoch": 0.8456299179287587,
+      "grad_norm": 0.15217041969299316,
+      "learning_rate": 0.001,
+      "loss": 2.5483,
+      "step": 19989
+    },
+    {
+      "epoch": 0.8456722226922752,
+      "grad_norm": 0.15171130001544952,
+      "learning_rate": 0.001,
+      "loss": 1.8737,
+      "step": 19990
+    },
+    {
+      "epoch": 0.8457145274557916,
+      "grad_norm": 0.15747137367725372,
+      "learning_rate": 0.001,
+      "loss": 1.7848,
+      "step": 19991
+    },
+    {
+      "epoch": 0.8457568322193079,
+      "grad_norm": 0.22473227977752686,
+      "learning_rate": 0.001,
+      "loss": 2.192,
+      "step": 19992
+    },
+    {
+      "epoch": 0.8457991369828243,
+      "grad_norm": 0.19137416779994965,
+      "learning_rate": 0.001,
+      "loss": 4.3493,
+      "step": 19993
+    },
+    {
+      "epoch": 0.8458414417463407,
+      "grad_norm": 0.17920900881290436,
+      "learning_rate": 0.001,
+      "loss": 2.6782,
+      "step": 19994
+    },
+    {
+      "epoch": 0.845883746509857,
+      "grad_norm": 0.16247224807739258,
+      "learning_rate": 0.001,
+      "loss": 2.0712,
+      "step": 19995
+    },
+    {
+      "epoch": 0.8459260512733734,
+      "grad_norm": 2.3473522663116455,
+      "learning_rate": 0.001,
+      "loss": 2.1859,
+      "step": 19996
+    },
+    {
+      "epoch": 0.8459683560368898,
+      "grad_norm": 0.1629837155342102,
+      "learning_rate": 0.001,
+      "loss": 2.4058,
+      "step": 19997
+    },
+    {
+      "epoch": 0.8460106608004061,
+      "grad_norm": 0.1541110724210739,
+      "learning_rate": 0.001,
+      "loss": 2.6553,
+      "step": 19998
+    },
+    {
+      "epoch": 0.8460529655639225,
+      "grad_norm": 0.12468776851892471,
+      "learning_rate": 0.001,
+      "loss": 2.0798,
+      "step": 19999
+    },
+    {
+      "epoch": 0.8460952703274389,
+      "grad_norm": 0.13102620840072632,
+      "learning_rate": 0.001,
+      "loss": 3.1871,
+      "step": 20000
+    },
+    {
+      "epoch": 0.8461375750909552,
+      "grad_norm": 0.16048361361026764,
+      "learning_rate": 0.001,
+      "loss": 1.6277,
+      "step": 20001
+    },
+    {
+      "epoch": 0.8461798798544716,
+      "grad_norm": 0.12173058837652206,
+      "learning_rate": 0.001,
+      "loss": 1.9958,
+      "step": 20002
+    },
+    {
+      "epoch": 0.846222184617988,
+      "grad_norm": 0.1663282960653305,
+      "learning_rate": 0.001,
+      "loss": 1.9365,
+      "step": 20003
+    },
+    {
+      "epoch": 0.8462644893815043,
+      "grad_norm": 0.2031364142894745,
+      "learning_rate": 0.001,
+      "loss": 2.5175,
+      "step": 20004
+    },
+    {
+      "epoch": 0.8463067941450207,
+      "grad_norm": 0.17239537835121155,
+      "learning_rate": 0.001,
+      "loss": 1.9938,
+      "step": 20005
+    },
+    {
+      "epoch": 0.8463490989085372,
+      "grad_norm": 0.14862409234046936,
+      "learning_rate": 0.001,
+      "loss": 1.9449,
+      "step": 20006
+    },
+    {
+      "epoch": 0.8463914036720535,
+      "grad_norm": 0.18297027051448822,
+      "learning_rate": 0.001,
+      "loss": 2.2749,
+      "step": 20007
+    },
+    {
+      "epoch": 0.8464337084355699,
+      "grad_norm": 0.16779938340187073,
+      "learning_rate": 0.001,
+      "loss": 1.6555,
+      "step": 20008
+    },
+    {
+      "epoch": 0.8464760131990862,
+      "grad_norm": 0.21444658935070038,
+      "learning_rate": 0.001,
+      "loss": 2.8099,
+      "step": 20009
+    },
+    {
+      "epoch": 0.8465183179626026,
+      "grad_norm": 0.13690084218978882,
+      "learning_rate": 0.001,
+      "loss": 3.1176,
+      "step": 20010
+    },
+    {
+      "epoch": 0.846560622726119,
+      "grad_norm": 0.235006645321846,
+      "learning_rate": 0.001,
+      "loss": 2.1184,
+      "step": 20011
+    },
+    {
+      "epoch": 0.8466029274896353,
+      "grad_norm": 0.1450556516647339,
+      "learning_rate": 0.001,
+      "loss": 1.8434,
+      "step": 20012
+    },
+    {
+      "epoch": 0.8466452322531517,
+      "grad_norm": 0.14639632403850555,
+      "learning_rate": 0.001,
+      "loss": 1.8493,
+      "step": 20013
+    },
+    {
+      "epoch": 0.8466875370166681,
+      "grad_norm": 0.12522344291210175,
+      "learning_rate": 0.001,
+      "loss": 1.5215,
+      "step": 20014
+    },
+    {
+      "epoch": 0.8467298417801844,
+      "grad_norm": 0.19824448227882385,
+      "learning_rate": 0.001,
+      "loss": 3.5594,
+      "step": 20015
+    },
+    {
+      "epoch": 0.8467721465437008,
+      "grad_norm": 0.13851766288280487,
+      "learning_rate": 0.001,
+      "loss": 2.2443,
+      "step": 20016
+    },
+    {
+      "epoch": 0.8468144513072172,
+      "grad_norm": 0.13753733038902283,
+      "learning_rate": 0.001,
+      "loss": 2.3559,
+      "step": 20017
+    },
+    {
+      "epoch": 0.8468567560707335,
+      "grad_norm": 0.15897375345230103,
+      "learning_rate": 0.001,
+      "loss": 1.8587,
+      "step": 20018
+    },
+    {
+      "epoch": 0.8468990608342499,
+      "grad_norm": 0.13712716102600098,
+      "learning_rate": 0.001,
+      "loss": 2.0603,
+      "step": 20019
+    },
+    {
+      "epoch": 0.8469413655977663,
+      "grad_norm": 0.1391967236995697,
+      "learning_rate": 0.001,
+      "loss": 2.0169,
+      "step": 20020
+    },
+    {
+      "epoch": 0.8469836703612826,
+      "grad_norm": 0.1359555423259735,
+      "learning_rate": 0.001,
+      "loss": 2.2155,
+      "step": 20021
+    },
+    {
+      "epoch": 0.847025975124799,
+      "grad_norm": 0.7110766768455505,
+      "learning_rate": 0.001,
+      "loss": 3.0314,
+      "step": 20022
+    },
+    {
+      "epoch": 0.8470682798883155,
+      "grad_norm": 0.1405889242887497,
+      "learning_rate": 0.001,
+      "loss": 2.1806,
+      "step": 20023
+    },
+    {
+      "epoch": 0.8471105846518318,
+      "grad_norm": 0.14044038951396942,
+      "learning_rate": 0.001,
+      "loss": 1.9486,
+      "step": 20024
+    },
+    {
+      "epoch": 0.8471528894153482,
+      "grad_norm": 0.122348852455616,
+      "learning_rate": 0.001,
+      "loss": 1.5374,
+      "step": 20025
+    },
+    {
+      "epoch": 0.8471951941788646,
+      "grad_norm": 0.14094999432563782,
+      "learning_rate": 0.001,
+      "loss": 1.7621,
+      "step": 20026
+    },
+    {
+      "epoch": 0.8472374989423809,
+      "grad_norm": 0.6236584782600403,
+      "learning_rate": 0.001,
+      "loss": 1.964,
+      "step": 20027
+    },
+    {
+      "epoch": 0.8472798037058973,
+      "grad_norm": 0.16254521906375885,
+      "learning_rate": 0.001,
+      "loss": 2.0317,
+      "step": 20028
+    },
+    {
+      "epoch": 0.8473221084694137,
+      "grad_norm": 0.16386757791042328,
+      "learning_rate": 0.001,
+      "loss": 2.4891,
+      "step": 20029
+    },
+    {
+      "epoch": 0.84736441323293,
+      "grad_norm": 0.26580187678337097,
+      "learning_rate": 0.001,
+      "loss": 1.803,
+      "step": 20030
+    },
+    {
+      "epoch": 0.8474067179964464,
+      "grad_norm": 0.13852697610855103,
+      "learning_rate": 0.001,
+      "loss": 1.9764,
+      "step": 20031
+    },
+    {
+      "epoch": 0.8474490227599628,
+      "grad_norm": 0.13823279738426208,
+      "learning_rate": 0.001,
+      "loss": 1.6498,
+      "step": 20032
+    },
+    {
+      "epoch": 0.8474913275234791,
+      "grad_norm": 0.15287473797798157,
+      "learning_rate": 0.001,
+      "loss": 2.5265,
+      "step": 20033
+    },
+    {
+      "epoch": 0.8475336322869955,
+      "grad_norm": 0.661291241645813,
+      "learning_rate": 0.001,
+      "loss": 2.3638,
+      "step": 20034
+    },
+    {
+      "epoch": 0.8475759370505119,
+      "grad_norm": 0.16113339364528656,
+      "learning_rate": 0.001,
+      "loss": 2.0843,
+      "step": 20035
+    },
+    {
+      "epoch": 0.8476182418140282,
+      "grad_norm": 0.1465645283460617,
+      "learning_rate": 0.001,
+      "loss": 2.0648,
+      "step": 20036
+    },
+    {
+      "epoch": 0.8476605465775446,
+      "grad_norm": 1.3468478918075562,
+      "learning_rate": 0.001,
+      "loss": 2.1217,
+      "step": 20037
+    },
+    {
+      "epoch": 0.847702851341061,
+      "grad_norm": 1.280560851097107,
+      "learning_rate": 0.001,
+      "loss": 2.0666,
+      "step": 20038
+    },
+    {
+      "epoch": 0.8477451561045773,
+      "grad_norm": 0.17766667902469635,
+      "learning_rate": 0.001,
+      "loss": 2.116,
+      "step": 20039
+    },
+    {
+      "epoch": 0.8477874608680938,
+      "grad_norm": 0.20155932009220123,
+      "learning_rate": 0.001,
+      "loss": 1.9426,
+      "step": 20040
+    },
+    {
+      "epoch": 0.8478297656316102,
+      "grad_norm": 0.2062946856021881,
+      "learning_rate": 0.001,
+      "loss": 2.7091,
+      "step": 20041
+    },
+    {
+      "epoch": 0.8478720703951265,
+      "grad_norm": 0.1285688877105713,
+      "learning_rate": 0.001,
+      "loss": 2.7954,
+      "step": 20042
+    },
+    {
+      "epoch": 0.8479143751586429,
+      "grad_norm": 2.014345407485962,
+      "learning_rate": 0.001,
+      "loss": 2.2796,
+      "step": 20043
+    },
+    {
+      "epoch": 0.8479566799221593,
+      "grad_norm": 6.951234817504883,
+      "learning_rate": 0.001,
+      "loss": 3.4023,
+      "step": 20044
+    },
+    {
+      "epoch": 0.8479989846856756,
+      "grad_norm": 0.16423997282981873,
+      "learning_rate": 0.001,
+      "loss": 2.1239,
+      "step": 20045
+    },
+    {
+      "epoch": 0.848041289449192,
+      "grad_norm": 0.14751717448234558,
+      "learning_rate": 0.001,
+      "loss": 2.3457,
+      "step": 20046
+    },
+    {
+      "epoch": 0.8480835942127084,
+      "grad_norm": 0.1324821263551712,
+      "learning_rate": 0.001,
+      "loss": 1.9863,
+      "step": 20047
+    },
+    {
+      "epoch": 0.8481258989762247,
+      "grad_norm": 0.2402341365814209,
+      "learning_rate": 0.001,
+      "loss": 1.8448,
+      "step": 20048
+    },
+    {
+      "epoch": 0.8481682037397411,
+      "grad_norm": 7.083425045013428,
+      "learning_rate": 0.001,
+      "loss": 2.5456,
+      "step": 20049
+    },
+    {
+      "epoch": 0.8482105085032575,
+      "grad_norm": 0.1683296114206314,
+      "learning_rate": 0.001,
+      "loss": 2.432,
+      "step": 20050
+    },
+    {
+      "epoch": 0.8482528132667738,
+      "grad_norm": 0.18871085345745087,
+      "learning_rate": 0.001,
+      "loss": 1.9026,
+      "step": 20051
+    },
+    {
+      "epoch": 0.8482951180302902,
+      "grad_norm": 0.13885819911956787,
+      "learning_rate": 0.001,
+      "loss": 1.7946,
+      "step": 20052
+    },
+    {
+      "epoch": 0.8483374227938065,
+      "grad_norm": 0.13285697996616364,
+      "learning_rate": 0.001,
+      "loss": 1.395,
+      "step": 20053
+    },
+    {
+      "epoch": 0.8483797275573229,
+      "grad_norm": 0.1700686663389206,
+      "learning_rate": 0.001,
+      "loss": 2.1973,
+      "step": 20054
+    },
+    {
+      "epoch": 0.8484220323208393,
+      "grad_norm": 0.20374974608421326,
+      "learning_rate": 0.001,
+      "loss": 2.6799,
+      "step": 20055
+    },
+    {
+      "epoch": 0.8484643370843556,
+      "grad_norm": 0.2053978443145752,
+      "learning_rate": 0.001,
+      "loss": 1.7101,
+      "step": 20056
+    },
+    {
+      "epoch": 0.8485066418478721,
+      "grad_norm": 0.16068226099014282,
+      "learning_rate": 0.001,
+      "loss": 2.1583,
+      "step": 20057
+    },
+    {
+      "epoch": 0.8485489466113885,
+      "grad_norm": 0.1553301364183426,
+      "learning_rate": 0.001,
+      "loss": 2.122,
+      "step": 20058
+    },
+    {
+      "epoch": 0.8485912513749048,
+      "grad_norm": 0.16317340731620789,
+      "learning_rate": 0.001,
+      "loss": 1.9723,
+      "step": 20059
+    },
+    {
+      "epoch": 0.8486335561384212,
+      "grad_norm": 0.6233834028244019,
+      "learning_rate": 0.001,
+      "loss": 1.8776,
+      "step": 20060
+    },
+    {
+      "epoch": 0.8486758609019376,
+      "grad_norm": 0.1380307525396347,
+      "learning_rate": 0.001,
+      "loss": 2.1375,
+      "step": 20061
+    },
+    {
+      "epoch": 0.8487181656654539,
+      "grad_norm": 0.17783686518669128,
+      "learning_rate": 0.001,
+      "loss": 2.2367,
+      "step": 20062
+    },
+    {
+      "epoch": 0.8487604704289703,
+      "grad_norm": 0.14636710286140442,
+      "learning_rate": 0.001,
+      "loss": 1.6968,
+      "step": 20063
+    },
+    {
+      "epoch": 0.8488027751924867,
+      "grad_norm": 0.2252545803785324,
+      "learning_rate": 0.001,
+      "loss": 2.1908,
+      "step": 20064
+    },
+    {
+      "epoch": 0.848845079956003,
+      "grad_norm": 0.129207044839859,
+      "learning_rate": 0.001,
+      "loss": 2.0196,
+      "step": 20065
+    },
+    {
+      "epoch": 0.8488873847195194,
+      "grad_norm": 471.605712890625,
+      "learning_rate": 0.001,
+      "loss": 2.1384,
+      "step": 20066
+    },
+    {
+      "epoch": 0.8489296894830358,
+      "grad_norm": 0.18888062238693237,
+      "learning_rate": 0.001,
+      "loss": 2.9906,
+      "step": 20067
+    },
+    {
+      "epoch": 0.8489719942465521,
+      "grad_norm": 0.16716241836547852,
+      "learning_rate": 0.001,
+      "loss": 2.5418,
+      "step": 20068
+    },
+    {
+      "epoch": 0.8490142990100685,
+      "grad_norm": 0.23846466839313507,
+      "learning_rate": 0.001,
+      "loss": 1.9297,
+      "step": 20069
+    },
+    {
+      "epoch": 0.8490566037735849,
+      "grad_norm": 0.2950774133205414,
+      "learning_rate": 0.001,
+      "loss": 2.0279,
+      "step": 20070
+    },
+    {
+      "epoch": 0.8490989085371012,
+      "grad_norm": 0.18450148403644562,
+      "learning_rate": 0.001,
+      "loss": 1.901,
+      "step": 20071
+    },
+    {
+      "epoch": 0.8491412133006176,
+      "grad_norm": 0.15849687159061432,
+      "learning_rate": 0.001,
+      "loss": 1.8239,
+      "step": 20072
+    },
+    {
+      "epoch": 0.8491835180641341,
+      "grad_norm": 0.17029528319835663,
+      "learning_rate": 0.001,
+      "loss": 1.5521,
+      "step": 20073
+    },
+    {
+      "epoch": 0.8492258228276504,
+      "grad_norm": 0.1589442491531372,
+      "learning_rate": 0.001,
+      "loss": 1.79,
+      "step": 20074
+    },
+    {
+      "epoch": 0.8492681275911668,
+      "grad_norm": 0.19395272433757782,
+      "learning_rate": 0.001,
+      "loss": 2.1332,
+      "step": 20075
+    },
+    {
+      "epoch": 0.8493104323546832,
+      "grad_norm": 0.1506047248840332,
+      "learning_rate": 0.001,
+      "loss": 2.3834,
+      "step": 20076
+    },
+    {
+      "epoch": 0.8493527371181995,
+      "grad_norm": 0.1427820324897766,
+      "learning_rate": 0.001,
+      "loss": 1.5925,
+      "step": 20077
+    },
+    {
+      "epoch": 0.8493950418817159,
+      "grad_norm": 0.14985975623130798,
+      "learning_rate": 0.001,
+      "loss": 2.5144,
+      "step": 20078
+    },
+    {
+      "epoch": 0.8494373466452323,
+      "grad_norm": 10.775849342346191,
+      "learning_rate": 0.001,
+      "loss": 1.8041,
+      "step": 20079
+    },
+    {
+      "epoch": 0.8494796514087486,
+      "grad_norm": 0.16326069831848145,
+      "learning_rate": 0.001,
+      "loss": 1.98,
+      "step": 20080
+    },
+    {
+      "epoch": 0.849521956172265,
+      "grad_norm": 0.6944580674171448,
+      "learning_rate": 0.001,
+      "loss": 3.5848,
+      "step": 20081
+    },
+    {
+      "epoch": 0.8495642609357814,
+      "grad_norm": 0.17368917167186737,
+      "learning_rate": 0.001,
+      "loss": 2.6593,
+      "step": 20082
+    },
+    {
+      "epoch": 0.8496065656992977,
+      "grad_norm": 0.17858272790908813,
+      "learning_rate": 0.001,
+      "loss": 2.2117,
+      "step": 20083
+    },
+    {
+      "epoch": 0.8496488704628141,
+      "grad_norm": 0.2750988304615021,
+      "learning_rate": 0.001,
+      "loss": 2.6538,
+      "step": 20084
+    },
+    {
+      "epoch": 0.8496911752263305,
+      "grad_norm": 0.4688890874385834,
+      "learning_rate": 0.001,
+      "loss": 2.3822,
+      "step": 20085
+    },
+    {
+      "epoch": 0.8497334799898468,
+      "grad_norm": 0.43606501817703247,
+      "learning_rate": 0.001,
+      "loss": 3.1274,
+      "step": 20086
+    },
+    {
+      "epoch": 0.8497757847533632,
+      "grad_norm": 0.1552165150642395,
+      "learning_rate": 0.001,
+      "loss": 1.8759,
+      "step": 20087
+    },
+    {
+      "epoch": 0.8498180895168796,
+      "grad_norm": 0.14785443246364594,
+      "learning_rate": 0.001,
+      "loss": 2.4866,
+      "step": 20088
+    },
+    {
+      "epoch": 0.849860394280396,
+      "grad_norm": 0.14098820090293884,
+      "learning_rate": 0.001,
+      "loss": 1.97,
+      "step": 20089
+    },
+    {
+      "epoch": 0.8499026990439124,
+      "grad_norm": 0.14183367788791656,
+      "learning_rate": 0.001,
+      "loss": 1.7328,
+      "step": 20090
+    },
+    {
+      "epoch": 0.8499450038074288,
+      "grad_norm": 0.2905732989311218,
+      "learning_rate": 0.001,
+      "loss": 2.5444,
+      "step": 20091
+    },
+    {
+      "epoch": 0.8499873085709451,
+      "grad_norm": 6.777652263641357,
+      "learning_rate": 0.001,
+      "loss": 1.8801,
+      "step": 20092
+    },
+    {
+      "epoch": 0.8500296133344615,
+      "grad_norm": 0.1453995257616043,
+      "learning_rate": 0.001,
+      "loss": 1.6809,
+      "step": 20093
+    },
+    {
+      "epoch": 0.8500719180979779,
+      "grad_norm": 0.18101413547992706,
+      "learning_rate": 0.001,
+      "loss": 2.2516,
+      "step": 20094
+    },
+    {
+      "epoch": 0.8501142228614942,
+      "grad_norm": 7.015632152557373,
+      "learning_rate": 0.001,
+      "loss": 1.9355,
+      "step": 20095
+    },
+    {
+      "epoch": 0.8501565276250106,
+      "grad_norm": 0.21599209308624268,
+      "learning_rate": 0.001,
+      "loss": 1.6428,
+      "step": 20096
+    },
+    {
+      "epoch": 0.8501988323885269,
+      "grad_norm": 0.19514302909374237,
+      "learning_rate": 0.001,
+      "loss": 1.8462,
+      "step": 20097
+    },
+    {
+      "epoch": 0.8502411371520433,
+      "grad_norm": 0.3018536865711212,
+      "learning_rate": 0.001,
+      "loss": 2.9103,
+      "step": 20098
+    },
+    {
+      "epoch": 0.8502834419155597,
+      "grad_norm": 0.220729798078537,
+      "learning_rate": 0.001,
+      "loss": 2.7488,
+      "step": 20099
+    },
+    {
+      "epoch": 0.850325746679076,
+      "grad_norm": 0.9878958463668823,
+      "learning_rate": 0.001,
+      "loss": 1.8072,
+      "step": 20100
+    },
+    {
+      "epoch": 0.8503680514425924,
+      "grad_norm": 0.19620686769485474,
+      "learning_rate": 0.001,
+      "loss": 1.6413,
+      "step": 20101
+    },
+    {
+      "epoch": 0.8504103562061088,
+      "grad_norm": 0.4310097098350525,
+      "learning_rate": 0.001,
+      "loss": 1.7159,
+      "step": 20102
+    },
+    {
+      "epoch": 0.8504526609696251,
+      "grad_norm": 0.352130264043808,
+      "learning_rate": 0.001,
+      "loss": 2.5859,
+      "step": 20103
+    },
+    {
+      "epoch": 0.8504949657331415,
+      "grad_norm": 0.3044179081916809,
+      "learning_rate": 0.001,
+      "loss": 1.5859,
+      "step": 20104
+    },
+    {
+      "epoch": 0.850537270496658,
+      "grad_norm": 0.2029111385345459,
+      "learning_rate": 0.001,
+      "loss": 2.3172,
+      "step": 20105
+    },
+    {
+      "epoch": 0.8505795752601742,
+      "grad_norm": 1.908150315284729,
+      "learning_rate": 0.001,
+      "loss": 2.2062,
+      "step": 20106
+    },
+    {
+      "epoch": 0.8506218800236907,
+      "grad_norm": 0.14281626045703888,
+      "learning_rate": 0.001,
+      "loss": 1.5915,
+      "step": 20107
+    },
+    {
+      "epoch": 0.8506641847872071,
+      "grad_norm": 0.16866661608219147,
+      "learning_rate": 0.001,
+      "loss": 2.6013,
+      "step": 20108
+    },
+    {
+      "epoch": 0.8507064895507234,
+      "grad_norm": 0.20509420335292816,
+      "learning_rate": 0.001,
+      "loss": 2.5722,
+      "step": 20109
+    },
+    {
+      "epoch": 0.8507487943142398,
+      "grad_norm": 0.17332904040813446,
+      "learning_rate": 0.001,
+      "loss": 1.7543,
+      "step": 20110
+    },
+    {
+      "epoch": 0.8507910990777562,
+      "grad_norm": 0.1855224370956421,
+      "learning_rate": 0.001,
+      "loss": 2.1487,
+      "step": 20111
+    },
+    {
+      "epoch": 0.8508334038412725,
+      "grad_norm": 0.17093543708324432,
+      "learning_rate": 0.001,
+      "loss": 2.3077,
+      "step": 20112
+    },
+    {
+      "epoch": 0.8508757086047889,
+      "grad_norm": 10.589107513427734,
+      "learning_rate": 0.001,
+      "loss": 2.2467,
+      "step": 20113
+    },
+    {
+      "epoch": 0.8509180133683053,
+      "grad_norm": 0.18264487385749817,
+      "learning_rate": 0.001,
+      "loss": 2.2559,
+      "step": 20114
+    },
+    {
+      "epoch": 0.8509603181318216,
+      "grad_norm": 0.19633848965168,
+      "learning_rate": 0.001,
+      "loss": 2.1265,
+      "step": 20115
+    },
+    {
+      "epoch": 0.851002622895338,
+      "grad_norm": 0.1672111302614212,
+      "learning_rate": 0.001,
+      "loss": 2.9774,
+      "step": 20116
+    },
+    {
+      "epoch": 0.8510449276588544,
+      "grad_norm": 0.25148797035217285,
+      "learning_rate": 0.001,
+      "loss": 2.0833,
+      "step": 20117
+    },
+    {
+      "epoch": 0.8510872324223707,
+      "grad_norm": 0.22718261182308197,
+      "learning_rate": 0.001,
+      "loss": 3.6184,
+      "step": 20118
+    },
+    {
+      "epoch": 0.8511295371858871,
+      "grad_norm": 3.537106990814209,
+      "learning_rate": 0.001,
+      "loss": 2.5408,
+      "step": 20119
+    },
+    {
+      "epoch": 0.8511718419494035,
+      "grad_norm": 0.16952507197856903,
+      "learning_rate": 0.001,
+      "loss": 2.31,
+      "step": 20120
+    },
+    {
+      "epoch": 0.8512141467129198,
+      "grad_norm": 0.23125673830509186,
+      "learning_rate": 0.001,
+      "loss": 2.2545,
+      "step": 20121
+    },
+    {
+      "epoch": 0.8512564514764362,
+      "grad_norm": 0.19513604044914246,
+      "learning_rate": 0.001,
+      "loss": 3.295,
+      "step": 20122
+    },
+    {
+      "epoch": 0.8512987562399527,
+      "grad_norm": 0.20079001784324646,
+      "learning_rate": 0.001,
+      "loss": 1.5811,
+      "step": 20123
+    },
+    {
+      "epoch": 0.851341061003469,
+      "grad_norm": 0.22401869297027588,
+      "learning_rate": 0.001,
+      "loss": 1.9916,
+      "step": 20124
+    },
+    {
+      "epoch": 0.8513833657669854,
+      "grad_norm": 0.19173742830753326,
+      "learning_rate": 0.001,
+      "loss": 2.5162,
+      "step": 20125
+    },
+    {
+      "epoch": 0.8514256705305018,
+      "grad_norm": 5.623570919036865,
+      "learning_rate": 0.001,
+      "loss": 1.7293,
+      "step": 20126
+    },
+    {
+      "epoch": 0.8514679752940181,
+      "grad_norm": 0.2150885909795761,
+      "learning_rate": 0.001,
+      "loss": 2.5409,
+      "step": 20127
+    },
+    {
+      "epoch": 0.8515102800575345,
+      "grad_norm": 0.2799954116344452,
+      "learning_rate": 0.001,
+      "loss": 3.2083,
+      "step": 20128
+    },
+    {
+      "epoch": 0.8515525848210509,
+      "grad_norm": 0.19551977515220642,
+      "learning_rate": 0.001,
+      "loss": 2.7637,
+      "step": 20129
+    },
+    {
+      "epoch": 0.8515948895845672,
+      "grad_norm": 0.24339886009693146,
+      "learning_rate": 0.001,
+      "loss": 3.3285,
+      "step": 20130
+    },
+    {
+      "epoch": 0.8516371943480836,
+      "grad_norm": 0.25382792949676514,
+      "learning_rate": 0.001,
+      "loss": 1.9141,
+      "step": 20131
+    },
+    {
+      "epoch": 0.8516794991116,
+      "grad_norm": 0.17884451150894165,
+      "learning_rate": 0.001,
+      "loss": 2.4345,
+      "step": 20132
+    },
+    {
+      "epoch": 0.8517218038751163,
+      "grad_norm": 0.9406694769859314,
+      "learning_rate": 0.001,
+      "loss": 2.2004,
+      "step": 20133
+    },
+    {
+      "epoch": 0.8517641086386327,
+      "grad_norm": 0.874098002910614,
+      "learning_rate": 0.001,
+      "loss": 2.1619,
+      "step": 20134
+    },
+    {
+      "epoch": 0.8518064134021491,
+      "grad_norm": 0.17671692371368408,
+      "learning_rate": 0.001,
+      "loss": 2.7296,
+      "step": 20135
+    },
+    {
+      "epoch": 0.8518487181656654,
+      "grad_norm": 0.16355298459529877,
+      "learning_rate": 0.001,
+      "loss": 3.0569,
+      "step": 20136
+    },
+    {
+      "epoch": 0.8518910229291818,
+      "grad_norm": 0.24405351281166077,
+      "learning_rate": 0.001,
+      "loss": 1.9078,
+      "step": 20137
+    },
+    {
+      "epoch": 0.8519333276926983,
+      "grad_norm": 0.14516574144363403,
+      "learning_rate": 0.001,
+      "loss": 1.629,
+      "step": 20138
+    },
+    {
+      "epoch": 0.8519756324562145,
+      "grad_norm": 0.1458577811717987,
+      "learning_rate": 0.001,
+      "loss": 1.9627,
+      "step": 20139
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 5.451334476470947,
+      "learning_rate": 0.001,
+      "loss": 2.0621,
+      "step": 20140
+    },
+    {
+      "epoch": 0.8520602419832474,
+      "grad_norm": 0.15075600147247314,
+      "learning_rate": 0.001,
+      "loss": 1.9024,
+      "step": 20141
+    },
+    {
+      "epoch": 0.8521025467467637,
+      "grad_norm": 0.16598421335220337,
+      "learning_rate": 0.001,
+      "loss": 1.7182,
+      "step": 20142
+    },
+    {
+      "epoch": 0.8521448515102801,
+      "grad_norm": 0.14806927740573883,
+      "learning_rate": 0.001,
+      "loss": 1.511,
+      "step": 20143
+    },
+    {
+      "epoch": 0.8521871562737964,
+      "grad_norm": 0.8803728818893433,
+      "learning_rate": 0.001,
+      "loss": 2.5187,
+      "step": 20144
+    },
+    {
+      "epoch": 0.8522294610373128,
+      "grad_norm": 0.14872333407402039,
+      "learning_rate": 0.001,
+      "loss": 2.2161,
+      "step": 20145
+    },
+    {
+      "epoch": 0.8522717658008292,
+      "grad_norm": 0.14645099639892578,
+      "learning_rate": 0.001,
+      "loss": 1.8971,
+      "step": 20146
+    },
+    {
+      "epoch": 0.8523140705643455,
+      "grad_norm": 0.16836774349212646,
+      "learning_rate": 0.001,
+      "loss": 2.1939,
+      "step": 20147
+    },
+    {
+      "epoch": 0.8523563753278619,
+      "grad_norm": 0.24831850826740265,
+      "learning_rate": 0.001,
+      "loss": 2.191,
+      "step": 20148
+    },
+    {
+      "epoch": 0.8523986800913783,
+      "grad_norm": 0.1501324474811554,
+      "learning_rate": 0.001,
+      "loss": 2.3453,
+      "step": 20149
+    },
+    {
+      "epoch": 0.8524409848548946,
+      "grad_norm": 0.15773944556713104,
+      "learning_rate": 0.001,
+      "loss": 2.0649,
+      "step": 20150
+    },
+    {
+      "epoch": 0.852483289618411,
+      "grad_norm": 0.13998405635356903,
+      "learning_rate": 0.001,
+      "loss": 2.1773,
+      "step": 20151
+    },
+    {
+      "epoch": 0.8525255943819274,
+      "grad_norm": 0.13668106496334076,
+      "learning_rate": 0.001,
+      "loss": 3.5096,
+      "step": 20152
+    },
+    {
+      "epoch": 0.8525678991454437,
+      "grad_norm": 2.312943935394287,
+      "learning_rate": 0.001,
+      "loss": 2.0723,
+      "step": 20153
+    },
+    {
+      "epoch": 0.8526102039089601,
+      "grad_norm": 0.12327218800783157,
+      "learning_rate": 0.001,
+      "loss": 1.5741,
+      "step": 20154
+    },
+    {
+      "epoch": 0.8526525086724766,
+      "grad_norm": 0.13412897288799286,
+      "learning_rate": 0.001,
+      "loss": 1.9401,
+      "step": 20155
+    },
+    {
+      "epoch": 0.8526948134359928,
+      "grad_norm": 0.8391098976135254,
+      "learning_rate": 0.001,
+      "loss": 2.408,
+      "step": 20156
+    },
+    {
+      "epoch": 0.8527371181995093,
+      "grad_norm": 0.14044369757175446,
+      "learning_rate": 0.001,
+      "loss": 2.7366,
+      "step": 20157
+    },
+    {
+      "epoch": 0.8527794229630257,
+      "grad_norm": 0.15021243691444397,
+      "learning_rate": 0.001,
+      "loss": 1.9738,
+      "step": 20158
+    },
+    {
+      "epoch": 0.852821727726542,
+      "grad_norm": 0.14377515017986298,
+      "learning_rate": 0.001,
+      "loss": 2.2906,
+      "step": 20159
+    },
+    {
+      "epoch": 0.8528640324900584,
+      "grad_norm": 0.17275890707969666,
+      "learning_rate": 0.001,
+      "loss": 1.9053,
+      "step": 20160
+    },
+    {
+      "epoch": 0.8529063372535748,
+      "grad_norm": 0.6215050220489502,
+      "learning_rate": 0.001,
+      "loss": 2.247,
+      "step": 20161
+    },
+    {
+      "epoch": 0.8529486420170911,
+      "grad_norm": 0.16696125268936157,
+      "learning_rate": 0.001,
+      "loss": 2.4084,
+      "step": 20162
+    },
+    {
+      "epoch": 0.8529909467806075,
+      "grad_norm": 1.0884296894073486,
+      "learning_rate": 0.001,
+      "loss": 2.9437,
+      "step": 20163
+    },
+    {
+      "epoch": 0.8530332515441239,
+      "grad_norm": 0.1563209444284439,
+      "learning_rate": 0.001,
+      "loss": 1.662,
+      "step": 20164
+    },
+    {
+      "epoch": 0.8530755563076402,
+      "grad_norm": 0.1429872363805771,
+      "learning_rate": 0.001,
+      "loss": 2.4649,
+      "step": 20165
+    },
+    {
+      "epoch": 0.8531178610711566,
+      "grad_norm": 2.498642921447754,
+      "learning_rate": 0.001,
+      "loss": 3.1658,
+      "step": 20166
+    },
+    {
+      "epoch": 0.853160165834673,
+      "grad_norm": 0.20113472640514374,
+      "learning_rate": 0.001,
+      "loss": 2.6068,
+      "step": 20167
+    },
+    {
+      "epoch": 0.8532024705981893,
+      "grad_norm": 0.14465253055095673,
+      "learning_rate": 0.001,
+      "loss": 2.3038,
+      "step": 20168
+    },
+    {
+      "epoch": 0.8532447753617057,
+      "grad_norm": 0.16391976177692413,
+      "learning_rate": 0.001,
+      "loss": 2.1344,
+      "step": 20169
+    },
+    {
+      "epoch": 0.8532870801252221,
+      "grad_norm": 0.16122137010097504,
+      "learning_rate": 0.001,
+      "loss": 3.441,
+      "step": 20170
+    },
+    {
+      "epoch": 0.8533293848887384,
+      "grad_norm": 0.7779273390769958,
+      "learning_rate": 0.001,
+      "loss": 1.4662,
+      "step": 20171
+    },
+    {
+      "epoch": 0.8533716896522549,
+      "grad_norm": 24.723020553588867,
+      "learning_rate": 0.001,
+      "loss": 1.8251,
+      "step": 20172
+    },
+    {
+      "epoch": 0.8534139944157713,
+      "grad_norm": 0.2329087257385254,
+      "learning_rate": 0.001,
+      "loss": 2.7639,
+      "step": 20173
+    },
+    {
+      "epoch": 0.8534562991792876,
+      "grad_norm": 0.14637401700019836,
+      "learning_rate": 0.001,
+      "loss": 1.5176,
+      "step": 20174
+    },
+    {
+      "epoch": 0.853498603942804,
+      "grad_norm": 0.16772529482841492,
+      "learning_rate": 0.001,
+      "loss": 2.0491,
+      "step": 20175
+    },
+    {
+      "epoch": 0.8535409087063204,
+      "grad_norm": 0.22313785552978516,
+      "learning_rate": 0.001,
+      "loss": 2.9042,
+      "step": 20176
+    },
+    {
+      "epoch": 0.8535832134698367,
+      "grad_norm": 0.16870731115341187,
+      "learning_rate": 0.001,
+      "loss": 2.4619,
+      "step": 20177
+    },
+    {
+      "epoch": 0.8536255182333531,
+      "grad_norm": 0.1313748061656952,
+      "learning_rate": 0.001,
+      "loss": 1.8931,
+      "step": 20178
+    },
+    {
+      "epoch": 0.8536678229968695,
+      "grad_norm": 0.1411811113357544,
+      "learning_rate": 0.001,
+      "loss": 1.7761,
+      "step": 20179
+    },
+    {
+      "epoch": 0.8537101277603858,
+      "grad_norm": 0.17591390013694763,
+      "learning_rate": 0.001,
+      "loss": 2.2475,
+      "step": 20180
+    },
+    {
+      "epoch": 0.8537524325239022,
+      "grad_norm": 0.12935340404510498,
+      "learning_rate": 0.001,
+      "loss": 2.8795,
+      "step": 20181
+    },
+    {
+      "epoch": 0.8537947372874186,
+      "grad_norm": 0.16716565191745758,
+      "learning_rate": 0.001,
+      "loss": 1.745,
+      "step": 20182
+    },
+    {
+      "epoch": 0.8538370420509349,
+      "grad_norm": 0.4727073907852173,
+      "learning_rate": 0.001,
+      "loss": 1.8856,
+      "step": 20183
+    },
+    {
+      "epoch": 0.8538793468144513,
+      "grad_norm": 0.13519716262817383,
+      "learning_rate": 0.001,
+      "loss": 1.5485,
+      "step": 20184
+    },
+    {
+      "epoch": 0.8539216515779677,
+      "grad_norm": 0.14166143536567688,
+      "learning_rate": 0.001,
+      "loss": 1.9562,
+      "step": 20185
+    },
+    {
+      "epoch": 0.853963956341484,
+      "grad_norm": 0.6362976431846619,
+      "learning_rate": 0.001,
+      "loss": 1.7597,
+      "step": 20186
+    },
+    {
+      "epoch": 0.8540062611050004,
+      "grad_norm": 0.1704770028591156,
+      "learning_rate": 0.001,
+      "loss": 2.1233,
+      "step": 20187
+    },
+    {
+      "epoch": 0.8540485658685167,
+      "grad_norm": 0.1493573784828186,
+      "learning_rate": 0.001,
+      "loss": 1.9775,
+      "step": 20188
+    },
+    {
+      "epoch": 0.8540908706320332,
+      "grad_norm": 0.14581403136253357,
+      "learning_rate": 0.001,
+      "loss": 1.6185,
+      "step": 20189
+    },
+    {
+      "epoch": 0.8541331753955496,
+      "grad_norm": 4.783538341522217,
+      "learning_rate": 0.001,
+      "loss": 2.7544,
+      "step": 20190
+    },
+    {
+      "epoch": 0.8541754801590659,
+      "grad_norm": 1.13120436668396,
+      "learning_rate": 0.001,
+      "loss": 2.2622,
+      "step": 20191
+    },
+    {
+      "epoch": 0.8542177849225823,
+      "grad_norm": 0.18143315613269806,
+      "learning_rate": 0.001,
+      "loss": 2.1182,
+      "step": 20192
+    },
+    {
+      "epoch": 0.8542600896860987,
+      "grad_norm": 0.37129276990890503,
+      "learning_rate": 0.001,
+      "loss": 2.6587,
+      "step": 20193
+    },
+    {
+      "epoch": 0.854302394449615,
+      "grad_norm": 2.5433599948883057,
+      "learning_rate": 0.001,
+      "loss": 2.0521,
+      "step": 20194
+    },
+    {
+      "epoch": 0.8543446992131314,
+      "grad_norm": 0.16446053981781006,
+      "learning_rate": 0.001,
+      "loss": 1.7504,
+      "step": 20195
+    },
+    {
+      "epoch": 0.8543870039766478,
+      "grad_norm": 0.16676220297813416,
+      "learning_rate": 0.001,
+      "loss": 1.6936,
+      "step": 20196
+    },
+    {
+      "epoch": 0.8544293087401641,
+      "grad_norm": 0.15930965542793274,
+      "learning_rate": 0.001,
+      "loss": 1.8836,
+      "step": 20197
+    },
+    {
+      "epoch": 0.8544716135036805,
+      "grad_norm": 0.16597270965576172,
+      "learning_rate": 0.001,
+      "loss": 1.7821,
+      "step": 20198
+    },
+    {
+      "epoch": 0.8545139182671969,
+      "grad_norm": 0.18009643256664276,
+      "learning_rate": 0.001,
+      "loss": 2.971,
+      "step": 20199
+    },
+    {
+      "epoch": 0.8545562230307132,
+      "grad_norm": 0.6921260952949524,
+      "learning_rate": 0.001,
+      "loss": 2.1276,
+      "step": 20200
+    },
+    {
+      "epoch": 0.8545985277942296,
+      "grad_norm": 0.1574309915304184,
+      "learning_rate": 0.001,
+      "loss": 2.1452,
+      "step": 20201
+    },
+    {
+      "epoch": 0.854640832557746,
+      "grad_norm": 0.16404499113559723,
+      "learning_rate": 0.001,
+      "loss": 2.189,
+      "step": 20202
+    },
+    {
+      "epoch": 0.8546831373212623,
+      "grad_norm": 0.13803629577159882,
+      "learning_rate": 0.001,
+      "loss": 1.8653,
+      "step": 20203
+    },
+    {
+      "epoch": 0.8547254420847787,
+      "grad_norm": 0.21450106799602509,
+      "learning_rate": 0.001,
+      "loss": 2.7563,
+      "step": 20204
+    },
+    {
+      "epoch": 0.8547677468482952,
+      "grad_norm": 0.1293824017047882,
+      "learning_rate": 0.001,
+      "loss": 1.6771,
+      "step": 20205
+    },
+    {
+      "epoch": 0.8548100516118115,
+      "grad_norm": 0.14951960742473602,
+      "learning_rate": 0.001,
+      "loss": 2.6877,
+      "step": 20206
+    },
+    {
+      "epoch": 0.8548523563753279,
+      "grad_norm": 1.9558665752410889,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 20207
+    },
+    {
+      "epoch": 0.8548946611388443,
+      "grad_norm": 0.16537365317344666,
+      "learning_rate": 0.001,
+      "loss": 1.9014,
+      "step": 20208
+    },
+    {
+      "epoch": 0.8549369659023606,
+      "grad_norm": 0.15604209899902344,
+      "learning_rate": 0.001,
+      "loss": 2.6344,
+      "step": 20209
+    },
+    {
+      "epoch": 0.854979270665877,
+      "grad_norm": 0.18542635440826416,
+      "learning_rate": 0.001,
+      "loss": 2.0202,
+      "step": 20210
+    },
+    {
+      "epoch": 0.8550215754293934,
+      "grad_norm": 0.4348926246166229,
+      "learning_rate": 0.001,
+      "loss": 2.0383,
+      "step": 20211
+    },
+    {
+      "epoch": 0.8550638801929097,
+      "grad_norm": 0.24259275197982788,
+      "learning_rate": 0.001,
+      "loss": 2.2676,
+      "step": 20212
+    },
+    {
+      "epoch": 0.8551061849564261,
+      "grad_norm": 0.1625998318195343,
+      "learning_rate": 0.001,
+      "loss": 1.8147,
+      "step": 20213
+    },
+    {
+      "epoch": 0.8551484897199425,
+      "grad_norm": 0.1525462120771408,
+      "learning_rate": 0.001,
+      "loss": 2.0379,
+      "step": 20214
+    },
+    {
+      "epoch": 0.8551907944834588,
+      "grad_norm": 0.1710134744644165,
+      "learning_rate": 0.001,
+      "loss": 2.2464,
+      "step": 20215
+    },
+    {
+      "epoch": 0.8552330992469752,
+      "grad_norm": 0.16427335143089294,
+      "learning_rate": 0.001,
+      "loss": 1.8928,
+      "step": 20216
+    },
+    {
+      "epoch": 0.8552754040104916,
+      "grad_norm": 0.1514868587255478,
+      "learning_rate": 0.001,
+      "loss": 2.0247,
+      "step": 20217
+    },
+    {
+      "epoch": 0.8553177087740079,
+      "grad_norm": 0.16171973943710327,
+      "learning_rate": 0.001,
+      "loss": 1.8759,
+      "step": 20218
+    },
+    {
+      "epoch": 0.8553600135375243,
+      "grad_norm": 1.2077202796936035,
+      "learning_rate": 0.001,
+      "loss": 2.7401,
+      "step": 20219
+    },
+    {
+      "epoch": 0.8554023183010407,
+      "grad_norm": 0.2370886206626892,
+      "learning_rate": 0.001,
+      "loss": 3.1154,
+      "step": 20220
+    },
+    {
+      "epoch": 0.855444623064557,
+      "grad_norm": 0.43571701645851135,
+      "learning_rate": 0.001,
+      "loss": 1.787,
+      "step": 20221
+    },
+    {
+      "epoch": 0.8554869278280735,
+      "grad_norm": 0.358680784702301,
+      "learning_rate": 0.001,
+      "loss": 2.0282,
+      "step": 20222
+    },
+    {
+      "epoch": 0.8555292325915899,
+      "grad_norm": 0.18976160883903503,
+      "learning_rate": 0.001,
+      "loss": 2.3883,
+      "step": 20223
+    },
+    {
+      "epoch": 0.8555715373551062,
+      "grad_norm": 0.1478756070137024,
+      "learning_rate": 0.001,
+      "loss": 1.8335,
+      "step": 20224
+    },
+    {
+      "epoch": 0.8556138421186226,
+      "grad_norm": 0.31914398074150085,
+      "learning_rate": 0.001,
+      "loss": 2.1218,
+      "step": 20225
+    },
+    {
+      "epoch": 0.855656146882139,
+      "grad_norm": 0.16509748995304108,
+      "learning_rate": 0.001,
+      "loss": 2.9447,
+      "step": 20226
+    },
+    {
+      "epoch": 0.8556984516456553,
+      "grad_norm": 0.20428349077701569,
+      "learning_rate": 0.001,
+      "loss": 1.7563,
+      "step": 20227
+    },
+    {
+      "epoch": 0.8557407564091717,
+      "grad_norm": 0.15375010669231415,
+      "learning_rate": 0.001,
+      "loss": 2.957,
+      "step": 20228
+    },
+    {
+      "epoch": 0.8557830611726881,
+      "grad_norm": 0.1550106555223465,
+      "learning_rate": 0.001,
+      "loss": 2.0625,
+      "step": 20229
+    },
+    {
+      "epoch": 0.8558253659362044,
+      "grad_norm": 0.7794061303138733,
+      "learning_rate": 0.001,
+      "loss": 2.9718,
+      "step": 20230
+    },
+    {
+      "epoch": 0.8558676706997208,
+      "grad_norm": 0.1645902544260025,
+      "learning_rate": 0.001,
+      "loss": 1.8263,
+      "step": 20231
+    },
+    {
+      "epoch": 0.8559099754632372,
+      "grad_norm": 0.14967839419841766,
+      "learning_rate": 0.001,
+      "loss": 1.828,
+      "step": 20232
+    },
+    {
+      "epoch": 0.8559522802267535,
+      "grad_norm": 0.3559803366661072,
+      "learning_rate": 0.001,
+      "loss": 2.069,
+      "step": 20233
+    },
+    {
+      "epoch": 0.8559945849902699,
+      "grad_norm": 0.1499088853597641,
+      "learning_rate": 0.001,
+      "loss": 1.9467,
+      "step": 20234
+    },
+    {
+      "epoch": 0.8560368897537862,
+      "grad_norm": 0.15247589349746704,
+      "learning_rate": 0.001,
+      "loss": 2.4844,
+      "step": 20235
+    },
+    {
+      "epoch": 0.8560791945173026,
+      "grad_norm": 0.16640542447566986,
+      "learning_rate": 0.001,
+      "loss": 2.222,
+      "step": 20236
+    },
+    {
+      "epoch": 0.856121499280819,
+      "grad_norm": 0.18216025829315186,
+      "learning_rate": 0.001,
+      "loss": 2.7879,
+      "step": 20237
+    },
+    {
+      "epoch": 0.8561638040443353,
+      "grad_norm": 0.2965571880340576,
+      "learning_rate": 0.001,
+      "loss": 2.3667,
+      "step": 20238
+    },
+    {
+      "epoch": 0.8562061088078518,
+      "grad_norm": 193.41860961914062,
+      "learning_rate": 0.001,
+      "loss": 1.4623,
+      "step": 20239
+    },
+    {
+      "epoch": 0.8562484135713682,
+      "grad_norm": 0.18015336990356445,
+      "learning_rate": 0.001,
+      "loss": 1.3319,
+      "step": 20240
+    },
+    {
+      "epoch": 0.8562907183348845,
+      "grad_norm": 0.1855952888727188,
+      "learning_rate": 0.001,
+      "loss": 2.0279,
+      "step": 20241
+    },
+    {
+      "epoch": 0.8563330230984009,
+      "grad_norm": 0.14489391446113586,
+      "learning_rate": 0.001,
+      "loss": 1.9138,
+      "step": 20242
+    },
+    {
+      "epoch": 0.8563753278619173,
+      "grad_norm": 0.19585612416267395,
+      "learning_rate": 0.001,
+      "loss": 1.921,
+      "step": 20243
+    },
+    {
+      "epoch": 0.8564176326254336,
+      "grad_norm": 0.1694945991039276,
+      "learning_rate": 0.001,
+      "loss": 1.8727,
+      "step": 20244
+    },
+    {
+      "epoch": 0.85645993738895,
+      "grad_norm": 0.2455274611711502,
+      "learning_rate": 0.001,
+      "loss": 2.0229,
+      "step": 20245
+    },
+    {
+      "epoch": 0.8565022421524664,
+      "grad_norm": 0.5265862941741943,
+      "learning_rate": 0.001,
+      "loss": 2.5353,
+      "step": 20246
+    },
+    {
+      "epoch": 0.8565445469159827,
+      "grad_norm": 0.7586323022842407,
+      "learning_rate": 0.001,
+      "loss": 2.2412,
+      "step": 20247
+    },
+    {
+      "epoch": 0.8565868516794991,
+      "grad_norm": 0.1906781643629074,
+      "learning_rate": 0.001,
+      "loss": 2.3801,
+      "step": 20248
+    },
+    {
+      "epoch": 0.8566291564430155,
+      "grad_norm": 0.18281270563602448,
+      "learning_rate": 0.001,
+      "loss": 1.8586,
+      "step": 20249
+    },
+    {
+      "epoch": 0.8566714612065318,
+      "grad_norm": 0.2404775768518448,
+      "learning_rate": 0.001,
+      "loss": 3.2444,
+      "step": 20250
+    },
+    {
+      "epoch": 0.8567137659700482,
+      "grad_norm": 0.1609775424003601,
+      "learning_rate": 0.001,
+      "loss": 2.8216,
+      "step": 20251
+    },
+    {
+      "epoch": 0.8567560707335646,
+      "grad_norm": 2.4240500926971436,
+      "learning_rate": 0.001,
+      "loss": 1.9752,
+      "step": 20252
+    },
+    {
+      "epoch": 0.8567983754970809,
+      "grad_norm": 0.2600942552089691,
+      "learning_rate": 0.001,
+      "loss": 2.9503,
+      "step": 20253
+    },
+    {
+      "epoch": 0.8568406802605973,
+      "grad_norm": 0.19915421307086945,
+      "learning_rate": 0.001,
+      "loss": 2.0628,
+      "step": 20254
+    },
+    {
+      "epoch": 0.8568829850241138,
+      "grad_norm": 0.16194456815719604,
+      "learning_rate": 0.001,
+      "loss": 2.0255,
+      "step": 20255
+    },
+    {
+      "epoch": 0.85692528978763,
+      "grad_norm": 0.15854515135288239,
+      "learning_rate": 0.001,
+      "loss": 2.5253,
+      "step": 20256
+    },
+    {
+      "epoch": 0.8569675945511465,
+      "grad_norm": 0.13806360960006714,
+      "learning_rate": 0.001,
+      "loss": 1.7646,
+      "step": 20257
+    },
+    {
+      "epoch": 0.8570098993146629,
+      "grad_norm": 0.47766992449760437,
+      "learning_rate": 0.001,
+      "loss": 2.2224,
+      "step": 20258
+    },
+    {
+      "epoch": 0.8570522040781792,
+      "grad_norm": 0.168570414185524,
+      "learning_rate": 0.001,
+      "loss": 3.5105,
+      "step": 20259
+    },
+    {
+      "epoch": 0.8570945088416956,
+      "grad_norm": 0.1976046860218048,
+      "learning_rate": 0.001,
+      "loss": 2.3813,
+      "step": 20260
+    },
+    {
+      "epoch": 0.857136813605212,
+      "grad_norm": 0.21734581887722015,
+      "learning_rate": 0.001,
+      "loss": 2.301,
+      "step": 20261
+    },
+    {
+      "epoch": 0.8571791183687283,
+      "grad_norm": 0.1577748954296112,
+      "learning_rate": 0.001,
+      "loss": 1.7362,
+      "step": 20262
+    },
+    {
+      "epoch": 0.8572214231322447,
+      "grad_norm": 0.1552111953496933,
+      "learning_rate": 0.001,
+      "loss": 2.7548,
+      "step": 20263
+    },
+    {
+      "epoch": 0.8572637278957611,
+      "grad_norm": 0.15418484807014465,
+      "learning_rate": 0.001,
+      "loss": 2.4548,
+      "step": 20264
+    },
+    {
+      "epoch": 0.8573060326592774,
+      "grad_norm": 0.2172391414642334,
+      "learning_rate": 0.001,
+      "loss": 2.6717,
+      "step": 20265
+    },
+    {
+      "epoch": 0.8573483374227938,
+      "grad_norm": 0.15989260375499725,
+      "learning_rate": 0.001,
+      "loss": 1.8661,
+      "step": 20266
+    },
+    {
+      "epoch": 0.8573906421863102,
+      "grad_norm": 2.102097988128662,
+      "learning_rate": 0.001,
+      "loss": 1.6953,
+      "step": 20267
+    },
+    {
+      "epoch": 0.8574329469498265,
+      "grad_norm": 0.13255327939987183,
+      "learning_rate": 0.001,
+      "loss": 2.1407,
+      "step": 20268
+    },
+    {
+      "epoch": 0.8574752517133429,
+      "grad_norm": 0.1770564168691635,
+      "learning_rate": 0.001,
+      "loss": 2.2512,
+      "step": 20269
+    },
+    {
+      "epoch": 0.8575175564768593,
+      "grad_norm": 0.14766398072242737,
+      "learning_rate": 0.001,
+      "loss": 1.7971,
+      "step": 20270
+    },
+    {
+      "epoch": 0.8575598612403756,
+      "grad_norm": 0.22131910920143127,
+      "learning_rate": 0.001,
+      "loss": 2.2181,
+      "step": 20271
+    },
+    {
+      "epoch": 0.857602166003892,
+      "grad_norm": 0.3644495904445648,
+      "learning_rate": 0.001,
+      "loss": 2.25,
+      "step": 20272
+    },
+    {
+      "epoch": 0.8576444707674085,
+      "grad_norm": 0.14953762292861938,
+      "learning_rate": 0.001,
+      "loss": 2.2675,
+      "step": 20273
+    },
+    {
+      "epoch": 0.8576867755309248,
+      "grad_norm": 0.16283674538135529,
+      "learning_rate": 0.001,
+      "loss": 1.8331,
+      "step": 20274
+    },
+    {
+      "epoch": 0.8577290802944412,
+      "grad_norm": 0.1207035556435585,
+      "learning_rate": 0.001,
+      "loss": 2.1584,
+      "step": 20275
+    },
+    {
+      "epoch": 0.8577713850579576,
+      "grad_norm": 0.12849994003772736,
+      "learning_rate": 0.001,
+      "loss": 2.2343,
+      "step": 20276
+    },
+    {
+      "epoch": 0.8578136898214739,
+      "grad_norm": 2.081145763397217,
+      "learning_rate": 0.001,
+      "loss": 1.8637,
+      "step": 20277
+    },
+    {
+      "epoch": 0.8578559945849903,
+      "grad_norm": 0.1410299390554428,
+      "learning_rate": 0.001,
+      "loss": 2.8566,
+      "step": 20278
+    },
+    {
+      "epoch": 0.8578982993485066,
+      "grad_norm": 0.13738563656806946,
+      "learning_rate": 0.001,
+      "loss": 2.2659,
+      "step": 20279
+    },
+    {
+      "epoch": 0.857940604112023,
+      "grad_norm": 0.13959841430187225,
+      "learning_rate": 0.001,
+      "loss": 2.9711,
+      "step": 20280
+    },
+    {
+      "epoch": 0.8579829088755394,
+      "grad_norm": 0.1706565022468567,
+      "learning_rate": 0.001,
+      "loss": 2.1167,
+      "step": 20281
+    },
+    {
+      "epoch": 0.8580252136390557,
+      "grad_norm": 0.14428362250328064,
+      "learning_rate": 0.001,
+      "loss": 2.3054,
+      "step": 20282
+    },
+    {
+      "epoch": 0.8580675184025721,
+      "grad_norm": 0.1867976188659668,
+      "learning_rate": 0.001,
+      "loss": 1.8198,
+      "step": 20283
+    },
+    {
+      "epoch": 0.8581098231660885,
+      "grad_norm": 0.1552973836660385,
+      "learning_rate": 0.001,
+      "loss": 2.5998,
+      "step": 20284
+    },
+    {
+      "epoch": 0.8581521279296048,
+      "grad_norm": 0.1912197321653366,
+      "learning_rate": 0.001,
+      "loss": 1.9071,
+      "step": 20285
+    },
+    {
+      "epoch": 0.8581944326931212,
+      "grad_norm": 0.18686088919639587,
+      "learning_rate": 0.001,
+      "loss": 1.5572,
+      "step": 20286
+    },
+    {
+      "epoch": 0.8582367374566376,
+      "grad_norm": 0.21159033477306366,
+      "learning_rate": 0.001,
+      "loss": 2.1987,
+      "step": 20287
+    },
+    {
+      "epoch": 0.8582790422201539,
+      "grad_norm": 0.16880102455615997,
+      "learning_rate": 0.001,
+      "loss": 2.0401,
+      "step": 20288
+    },
+    {
+      "epoch": 0.8583213469836704,
+      "grad_norm": 0.16665174067020416,
+      "learning_rate": 0.001,
+      "loss": 1.763,
+      "step": 20289
+    },
+    {
+      "epoch": 0.8583636517471868,
+      "grad_norm": 0.5530775189399719,
+      "learning_rate": 0.001,
+      "loss": 2.763,
+      "step": 20290
+    },
+    {
+      "epoch": 0.8584059565107031,
+      "grad_norm": 0.1329568773508072,
+      "learning_rate": 0.001,
+      "loss": 1.8301,
+      "step": 20291
+    },
+    {
+      "epoch": 0.8584482612742195,
+      "grad_norm": 22.826807022094727,
+      "learning_rate": 0.001,
+      "loss": 2.5194,
+      "step": 20292
+    },
+    {
+      "epoch": 0.8584905660377359,
+      "grad_norm": 1.763818621635437,
+      "learning_rate": 0.001,
+      "loss": 3.4182,
+      "step": 20293
+    },
+    {
+      "epoch": 0.8585328708012522,
+      "grad_norm": 0.1884799599647522,
+      "learning_rate": 0.001,
+      "loss": 1.8589,
+      "step": 20294
+    },
+    {
+      "epoch": 0.8585751755647686,
+      "grad_norm": 0.1700194627046585,
+      "learning_rate": 0.001,
+      "loss": 2.0548,
+      "step": 20295
+    },
+    {
+      "epoch": 0.858617480328285,
+      "grad_norm": 0.17256033420562744,
+      "learning_rate": 0.001,
+      "loss": 2.0633,
+      "step": 20296
+    },
+    {
+      "epoch": 0.8586597850918013,
+      "grad_norm": 0.1553211808204651,
+      "learning_rate": 0.001,
+      "loss": 2.8999,
+      "step": 20297
+    },
+    {
+      "epoch": 0.8587020898553177,
+      "grad_norm": 0.17085745930671692,
+      "learning_rate": 0.001,
+      "loss": 2.191,
+      "step": 20298
+    },
+    {
+      "epoch": 0.8587443946188341,
+      "grad_norm": 0.18587633967399597,
+      "learning_rate": 0.001,
+      "loss": 2.1725,
+      "step": 20299
+    },
+    {
+      "epoch": 0.8587866993823504,
+      "grad_norm": 0.1755942404270172,
+      "learning_rate": 0.001,
+      "loss": 2.9481,
+      "step": 20300
+    },
+    {
+      "epoch": 0.8588290041458668,
+      "grad_norm": 0.6229716539382935,
+      "learning_rate": 0.001,
+      "loss": 4.1505,
+      "step": 20301
+    },
+    {
+      "epoch": 0.8588713089093832,
+      "grad_norm": 0.14304950833320618,
+      "learning_rate": 0.001,
+      "loss": 1.9846,
+      "step": 20302
+    },
+    {
+      "epoch": 0.8589136136728995,
+      "grad_norm": 0.1368221938610077,
+      "learning_rate": 0.001,
+      "loss": 2.0029,
+      "step": 20303
+    },
+    {
+      "epoch": 0.8589559184364159,
+      "grad_norm": 0.1675972193479538,
+      "learning_rate": 0.001,
+      "loss": 1.9319,
+      "step": 20304
+    },
+    {
+      "epoch": 0.8589982231999324,
+      "grad_norm": 0.1587626338005066,
+      "learning_rate": 0.001,
+      "loss": 1.855,
+      "step": 20305
+    },
+    {
+      "epoch": 0.8590405279634487,
+      "grad_norm": 0.3222064673900604,
+      "learning_rate": 0.001,
+      "loss": 3.0261,
+      "step": 20306
+    },
+    {
+      "epoch": 0.8590828327269651,
+      "grad_norm": 0.13862450420856476,
+      "learning_rate": 0.001,
+      "loss": 2.2649,
+      "step": 20307
+    },
+    {
+      "epoch": 0.8591251374904815,
+      "grad_norm": 0.20226208865642548,
+      "learning_rate": 0.001,
+      "loss": 2.0139,
+      "step": 20308
+    },
+    {
+      "epoch": 0.8591674422539978,
+      "grad_norm": 0.16037043929100037,
+      "learning_rate": 0.001,
+      "loss": 2.5581,
+      "step": 20309
+    },
+    {
+      "epoch": 0.8592097470175142,
+      "grad_norm": 0.17345164716243744,
+      "learning_rate": 0.001,
+      "loss": 1.9058,
+      "step": 20310
+    },
+    {
+      "epoch": 0.8592520517810306,
+      "grad_norm": 0.4555630385875702,
+      "learning_rate": 0.001,
+      "loss": 3.2089,
+      "step": 20311
+    },
+    {
+      "epoch": 0.8592943565445469,
+      "grad_norm": 0.6636331677436829,
+      "learning_rate": 0.001,
+      "loss": 1.7678,
+      "step": 20312
+    },
+    {
+      "epoch": 0.8593366613080633,
+      "grad_norm": 1.500398874282837,
+      "learning_rate": 0.001,
+      "loss": 1.8767,
+      "step": 20313
+    },
+    {
+      "epoch": 0.8593789660715797,
+      "grad_norm": 0.11064822226762772,
+      "learning_rate": 0.001,
+      "loss": 2.8582,
+      "step": 20314
+    },
+    {
+      "epoch": 0.859421270835096,
+      "grad_norm": 0.2151585817337036,
+      "learning_rate": 0.001,
+      "loss": 1.8232,
+      "step": 20315
+    },
+    {
+      "epoch": 0.8594635755986124,
+      "grad_norm": 0.14713852107524872,
+      "learning_rate": 0.001,
+      "loss": 2.7334,
+      "step": 20316
+    },
+    {
+      "epoch": 0.8595058803621288,
+      "grad_norm": 0.13766881823539734,
+      "learning_rate": 0.001,
+      "loss": 1.6997,
+      "step": 20317
+    },
+    {
+      "epoch": 0.8595481851256451,
+      "grad_norm": 0.18348439037799835,
+      "learning_rate": 0.001,
+      "loss": 2.0182,
+      "step": 20318
+    },
+    {
+      "epoch": 0.8595904898891615,
+      "grad_norm": 0.12168294191360474,
+      "learning_rate": 0.001,
+      "loss": 3.2898,
+      "step": 20319
+    },
+    {
+      "epoch": 0.859632794652678,
+      "grad_norm": 0.12945544719696045,
+      "learning_rate": 0.001,
+      "loss": 1.9306,
+      "step": 20320
+    },
+    {
+      "epoch": 0.8596750994161942,
+      "grad_norm": 0.13580989837646484,
+      "learning_rate": 0.001,
+      "loss": 1.7863,
+      "step": 20321
+    },
+    {
+      "epoch": 0.8597174041797107,
+      "grad_norm": 0.1241777166724205,
+      "learning_rate": 0.001,
+      "loss": 2.0355,
+      "step": 20322
+    },
+    {
+      "epoch": 0.859759708943227,
+      "grad_norm": 0.15077315270900726,
+      "learning_rate": 0.001,
+      "loss": 2.9516,
+      "step": 20323
+    },
+    {
+      "epoch": 0.8598020137067434,
+      "grad_norm": 0.1805119663476944,
+      "learning_rate": 0.001,
+      "loss": 2.0972,
+      "step": 20324
+    },
+    {
+      "epoch": 0.8598443184702598,
+      "grad_norm": 0.24641673266887665,
+      "learning_rate": 0.001,
+      "loss": 2.3205,
+      "step": 20325
+    },
+    {
+      "epoch": 0.8598866232337761,
+      "grad_norm": 0.15835225582122803,
+      "learning_rate": 0.001,
+      "loss": 2.644,
+      "step": 20326
+    },
+    {
+      "epoch": 0.8599289279972925,
+      "grad_norm": 0.1202114075422287,
+      "learning_rate": 0.001,
+      "loss": 1.4397,
+      "step": 20327
+    },
+    {
+      "epoch": 0.8599712327608089,
+      "grad_norm": 0.13486960530281067,
+      "learning_rate": 0.001,
+      "loss": 1.5115,
+      "step": 20328
+    },
+    {
+      "epoch": 0.8600135375243252,
+      "grad_norm": 0.14606483280658722,
+      "learning_rate": 0.001,
+      "loss": 1.9028,
+      "step": 20329
+    },
+    {
+      "epoch": 0.8600558422878416,
+      "grad_norm": 0.14410161972045898,
+      "learning_rate": 0.001,
+      "loss": 2.147,
+      "step": 20330
+    },
+    {
+      "epoch": 0.860098147051358,
+      "grad_norm": 0.14581367373466492,
+      "learning_rate": 0.001,
+      "loss": 2.5193,
+      "step": 20331
+    },
+    {
+      "epoch": 0.8601404518148743,
+      "grad_norm": 1.6199474334716797,
+      "learning_rate": 0.001,
+      "loss": 1.8664,
+      "step": 20332
+    },
+    {
+      "epoch": 0.8601827565783907,
+      "grad_norm": 0.14424443244934082,
+      "learning_rate": 0.001,
+      "loss": 2.7133,
+      "step": 20333
+    },
+    {
+      "epoch": 0.8602250613419071,
+      "grad_norm": 2.242705821990967,
+      "learning_rate": 0.001,
+      "loss": 1.5209,
+      "step": 20334
+    },
+    {
+      "epoch": 0.8602673661054234,
+      "grad_norm": 0.2191903442144394,
+      "learning_rate": 0.001,
+      "loss": 1.4768,
+      "step": 20335
+    },
+    {
+      "epoch": 0.8603096708689398,
+      "grad_norm": 0.1648568958044052,
+      "learning_rate": 0.001,
+      "loss": 1.5056,
+      "step": 20336
+    },
+    {
+      "epoch": 0.8603519756324562,
+      "grad_norm": 0.16343936324119568,
+      "learning_rate": 0.001,
+      "loss": 1.8221,
+      "step": 20337
+    },
+    {
+      "epoch": 0.8603942803959725,
+      "grad_norm": 0.20765286684036255,
+      "learning_rate": 0.001,
+      "loss": 1.8847,
+      "step": 20338
+    },
+    {
+      "epoch": 0.860436585159489,
+      "grad_norm": 0.16252531111240387,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 20339
+    },
+    {
+      "epoch": 0.8604788899230054,
+      "grad_norm": 0.4384823739528656,
+      "learning_rate": 0.001,
+      "loss": 2.4298,
+      "step": 20340
+    },
+    {
+      "epoch": 0.8605211946865217,
+      "grad_norm": 0.1340947151184082,
+      "learning_rate": 0.001,
+      "loss": 2.4211,
+      "step": 20341
+    },
+    {
+      "epoch": 0.8605634994500381,
+      "grad_norm": 4.556620121002197,
+      "learning_rate": 0.001,
+      "loss": 1.8013,
+      "step": 20342
+    },
+    {
+      "epoch": 0.8606058042135545,
+      "grad_norm": 0.2241327315568924,
+      "learning_rate": 0.001,
+      "loss": 2.6166,
+      "step": 20343
+    },
+    {
+      "epoch": 0.8606481089770708,
+      "grad_norm": 0.18115611374378204,
+      "learning_rate": 0.001,
+      "loss": 2.5444,
+      "step": 20344
+    },
+    {
+      "epoch": 0.8606904137405872,
+      "grad_norm": 0.15894751250743866,
+      "learning_rate": 0.001,
+      "loss": 2.2565,
+      "step": 20345
+    },
+    {
+      "epoch": 0.8607327185041036,
+      "grad_norm": 0.9839511513710022,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 20346
+    },
+    {
+      "epoch": 0.8607750232676199,
+      "grad_norm": 0.14979402720928192,
+      "learning_rate": 0.001,
+      "loss": 2.9431,
+      "step": 20347
+    },
+    {
+      "epoch": 0.8608173280311363,
+      "grad_norm": 0.22843679785728455,
+      "learning_rate": 0.001,
+      "loss": 2.2349,
+      "step": 20348
+    },
+    {
+      "epoch": 0.8608596327946527,
+      "grad_norm": 0.26081418991088867,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 20349
+    },
+    {
+      "epoch": 0.860901937558169,
+      "grad_norm": 0.20167173445224762,
+      "learning_rate": 0.001,
+      "loss": 3.1905,
+      "step": 20350
+    },
+    {
+      "epoch": 0.8609442423216854,
+      "grad_norm": 0.22882631421089172,
+      "learning_rate": 0.001,
+      "loss": 2.6694,
+      "step": 20351
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.23098324239253998,
+      "learning_rate": 0.001,
+      "loss": 1.9321,
+      "step": 20352
+    },
+    {
+      "epoch": 0.8610288518487181,
+      "grad_norm": 0.18798838555812836,
+      "learning_rate": 0.001,
+      "loss": 2.3873,
+      "step": 20353
+    },
+    {
+      "epoch": 0.8610711566122345,
+      "grad_norm": 0.20491915941238403,
+      "learning_rate": 0.001,
+      "loss": 1.764,
+      "step": 20354
+    },
+    {
+      "epoch": 0.861113461375751,
+      "grad_norm": 0.2087901085615158,
+      "learning_rate": 0.001,
+      "loss": 2.1135,
+      "step": 20355
+    },
+    {
+      "epoch": 0.8611557661392673,
+      "grad_norm": 0.2142881155014038,
+      "learning_rate": 0.001,
+      "loss": 2.0897,
+      "step": 20356
+    },
+    {
+      "epoch": 0.8611980709027837,
+      "grad_norm": 2.1609714031219482,
+      "learning_rate": 0.001,
+      "loss": 4.1002,
+      "step": 20357
+    },
+    {
+      "epoch": 0.8612403756663001,
+      "grad_norm": 0.21619194746017456,
+      "learning_rate": 0.001,
+      "loss": 2.0694,
+      "step": 20358
+    },
+    {
+      "epoch": 0.8612826804298164,
+      "grad_norm": 738.5827026367188,
+      "learning_rate": 0.001,
+      "loss": 3.4232,
+      "step": 20359
+    },
+    {
+      "epoch": 0.8613249851933328,
+      "grad_norm": 0.20825865864753723,
+      "learning_rate": 0.001,
+      "loss": 2.4266,
+      "step": 20360
+    },
+    {
+      "epoch": 0.8613672899568492,
+      "grad_norm": 0.41342079639434814,
+      "learning_rate": 0.001,
+      "loss": 3.1381,
+      "step": 20361
+    },
+    {
+      "epoch": 0.8614095947203655,
+      "grad_norm": 0.2036539614200592,
+      "learning_rate": 0.001,
+      "loss": 2.0657,
+      "step": 20362
+    },
+    {
+      "epoch": 0.8614518994838819,
+      "grad_norm": 0.9940122365951538,
+      "learning_rate": 0.001,
+      "loss": 2.2586,
+      "step": 20363
+    },
+    {
+      "epoch": 0.8614942042473983,
+      "grad_norm": 0.19654519855976105,
+      "learning_rate": 0.001,
+      "loss": 1.582,
+      "step": 20364
+    },
+    {
+      "epoch": 0.8615365090109146,
+      "grad_norm": 0.1715071052312851,
+      "learning_rate": 0.001,
+      "loss": 1.2585,
+      "step": 20365
+    },
+    {
+      "epoch": 0.861578813774431,
+      "grad_norm": 0.14859987795352936,
+      "learning_rate": 0.001,
+      "loss": 2.478,
+      "step": 20366
+    },
+    {
+      "epoch": 0.8616211185379474,
+      "grad_norm": 0.18359853327274323,
+      "learning_rate": 0.001,
+      "loss": 1.8721,
+      "step": 20367
+    },
+    {
+      "epoch": 0.8616634233014637,
+      "grad_norm": 0.24993368983268738,
+      "learning_rate": 0.001,
+      "loss": 2.4343,
+      "step": 20368
+    },
+    {
+      "epoch": 0.8617057280649801,
+      "grad_norm": 0.2071733921766281,
+      "learning_rate": 0.001,
+      "loss": 2.7097,
+      "step": 20369
+    },
+    {
+      "epoch": 0.8617480328284964,
+      "grad_norm": 2.571146249771118,
+      "learning_rate": 0.001,
+      "loss": 2.0202,
+      "step": 20370
+    },
+    {
+      "epoch": 0.8617903375920128,
+      "grad_norm": 0.5075177550315857,
+      "learning_rate": 0.001,
+      "loss": 2.1632,
+      "step": 20371
+    },
+    {
+      "epoch": 0.8618326423555293,
+      "grad_norm": 0.18654043972492218,
+      "learning_rate": 0.001,
+      "loss": 2.4338,
+      "step": 20372
+    },
+    {
+      "epoch": 0.8618749471190456,
+      "grad_norm": 0.2202577441930771,
+      "learning_rate": 0.001,
+      "loss": 3.0875,
+      "step": 20373
+    },
+    {
+      "epoch": 0.861917251882562,
+      "grad_norm": 3.0814690589904785,
+      "learning_rate": 0.001,
+      "loss": 2.5025,
+      "step": 20374
+    },
+    {
+      "epoch": 0.8619595566460784,
+      "grad_norm": 1.5214474201202393,
+      "learning_rate": 0.001,
+      "loss": 3.0009,
+      "step": 20375
+    },
+    {
+      "epoch": 0.8620018614095947,
+      "grad_norm": 0.14105013012886047,
+      "learning_rate": 0.001,
+      "loss": 1.2828,
+      "step": 20376
+    },
+    {
+      "epoch": 0.8620441661731111,
+      "grad_norm": 0.16624034941196442,
+      "learning_rate": 0.001,
+      "loss": 2.2575,
+      "step": 20377
+    },
+    {
+      "epoch": 0.8620864709366275,
+      "grad_norm": 0.14870575070381165,
+      "learning_rate": 0.001,
+      "loss": 2.1555,
+      "step": 20378
+    },
+    {
+      "epoch": 0.8621287757001438,
+      "grad_norm": 0.17573565244674683,
+      "learning_rate": 0.001,
+      "loss": 1.6783,
+      "step": 20379
+    },
+    {
+      "epoch": 0.8621710804636602,
+      "grad_norm": 0.3848056197166443,
+      "learning_rate": 0.001,
+      "loss": 2.9054,
+      "step": 20380
+    },
+    {
+      "epoch": 0.8622133852271766,
+      "grad_norm": 4.15639591217041,
+      "learning_rate": 0.001,
+      "loss": 3.0232,
+      "step": 20381
+    },
+    {
+      "epoch": 0.8622556899906929,
+      "grad_norm": 0.9076377749443054,
+      "learning_rate": 0.001,
+      "loss": 2.6062,
+      "step": 20382
+    },
+    {
+      "epoch": 0.8622979947542093,
+      "grad_norm": 0.19111178815364838,
+      "learning_rate": 0.001,
+      "loss": 2.8008,
+      "step": 20383
+    },
+    {
+      "epoch": 0.8623402995177257,
+      "grad_norm": 0.15568110346794128,
+      "learning_rate": 0.001,
+      "loss": 1.6041,
+      "step": 20384
+    },
+    {
+      "epoch": 0.862382604281242,
+      "grad_norm": 0.15372473001480103,
+      "learning_rate": 0.001,
+      "loss": 2.1381,
+      "step": 20385
+    },
+    {
+      "epoch": 0.8624249090447584,
+      "grad_norm": 6.081607341766357,
+      "learning_rate": 0.001,
+      "loss": 2.3537,
+      "step": 20386
+    },
+    {
+      "epoch": 0.8624672138082748,
+      "grad_norm": 0.15494304895401,
+      "learning_rate": 0.001,
+      "loss": 2.4234,
+      "step": 20387
+    },
+    {
+      "epoch": 0.8625095185717911,
+      "grad_norm": 0.2465604841709137,
+      "learning_rate": 0.001,
+      "loss": 2.8873,
+      "step": 20388
+    },
+    {
+      "epoch": 0.8625518233353076,
+      "grad_norm": 0.16855676472187042,
+      "learning_rate": 0.001,
+      "loss": 2.7802,
+      "step": 20389
+    },
+    {
+      "epoch": 0.862594128098824,
+      "grad_norm": 0.1656714677810669,
+      "learning_rate": 0.001,
+      "loss": 2.3807,
+      "step": 20390
+    },
+    {
+      "epoch": 0.8626364328623403,
+      "grad_norm": 0.38702452182769775,
+      "learning_rate": 0.001,
+      "loss": 2.1454,
+      "step": 20391
+    },
+    {
+      "epoch": 0.8626787376258567,
+      "grad_norm": 0.17900040745735168,
+      "learning_rate": 0.001,
+      "loss": 3.0575,
+      "step": 20392
+    },
+    {
+      "epoch": 0.8627210423893731,
+      "grad_norm": 0.17749975621700287,
+      "learning_rate": 0.001,
+      "loss": 1.8404,
+      "step": 20393
+    },
+    {
+      "epoch": 0.8627633471528894,
+      "grad_norm": 0.17908912897109985,
+      "learning_rate": 0.001,
+      "loss": 1.9364,
+      "step": 20394
+    },
+    {
+      "epoch": 0.8628056519164058,
+      "grad_norm": 0.7642483115196228,
+      "learning_rate": 0.001,
+      "loss": 3.3047,
+      "step": 20395
+    },
+    {
+      "epoch": 0.8628479566799222,
+      "grad_norm": 0.2033914178609848,
+      "learning_rate": 0.001,
+      "loss": 2.6474,
+      "step": 20396
+    },
+    {
+      "epoch": 0.8628902614434385,
+      "grad_norm": 0.7764859199523926,
+      "learning_rate": 0.001,
+      "loss": 2.9559,
+      "step": 20397
+    },
+    {
+      "epoch": 0.8629325662069549,
+      "grad_norm": 0.1633528769016266,
+      "learning_rate": 0.001,
+      "loss": 2.0213,
+      "step": 20398
+    },
+    {
+      "epoch": 0.8629748709704713,
+      "grad_norm": 0.16931723058223724,
+      "learning_rate": 0.001,
+      "loss": 1.7263,
+      "step": 20399
+    },
+    {
+      "epoch": 0.8630171757339876,
+      "grad_norm": 0.1585397720336914,
+      "learning_rate": 0.001,
+      "loss": 2.8618,
+      "step": 20400
+    },
+    {
+      "epoch": 0.863059480497504,
+      "grad_norm": 0.1430475264787674,
+      "learning_rate": 0.001,
+      "loss": 1.8755,
+      "step": 20401
+    },
+    {
+      "epoch": 0.8631017852610204,
+      "grad_norm": 0.12753239274024963,
+      "learning_rate": 0.001,
+      "loss": 2.3516,
+      "step": 20402
+    },
+    {
+      "epoch": 0.8631440900245367,
+      "grad_norm": 0.1464456021785736,
+      "learning_rate": 0.001,
+      "loss": 2.6126,
+      "step": 20403
+    },
+    {
+      "epoch": 0.8631863947880531,
+      "grad_norm": 0.20379814505577087,
+      "learning_rate": 0.001,
+      "loss": 1.7753,
+      "step": 20404
+    },
+    {
+      "epoch": 0.8632286995515696,
+      "grad_norm": 0.15957339107990265,
+      "learning_rate": 0.001,
+      "loss": 1.6083,
+      "step": 20405
+    },
+    {
+      "epoch": 0.8632710043150859,
+      "grad_norm": 0.2491006702184677,
+      "learning_rate": 0.001,
+      "loss": 2.6279,
+      "step": 20406
+    },
+    {
+      "epoch": 0.8633133090786023,
+      "grad_norm": 0.1599300354719162,
+      "learning_rate": 0.001,
+      "loss": 2.2997,
+      "step": 20407
+    },
+    {
+      "epoch": 0.8633556138421187,
+      "grad_norm": 0.14536763727664948,
+      "learning_rate": 0.001,
+      "loss": 2.2268,
+      "step": 20408
+    },
+    {
+      "epoch": 0.863397918605635,
+      "grad_norm": 0.12874636054039001,
+      "learning_rate": 0.001,
+      "loss": 2.0432,
+      "step": 20409
+    },
+    {
+      "epoch": 0.8634402233691514,
+      "grad_norm": 0.1630754917860031,
+      "learning_rate": 0.001,
+      "loss": 1.519,
+      "step": 20410
+    },
+    {
+      "epoch": 0.8634825281326678,
+      "grad_norm": 0.18456068634986877,
+      "learning_rate": 0.001,
+      "loss": 2.3714,
+      "step": 20411
+    },
+    {
+      "epoch": 0.8635248328961841,
+      "grad_norm": 0.3489452004432678,
+      "learning_rate": 0.001,
+      "loss": 2.1257,
+      "step": 20412
+    },
+    {
+      "epoch": 0.8635671376597005,
+      "grad_norm": 0.13253667950630188,
+      "learning_rate": 0.001,
+      "loss": 1.619,
+      "step": 20413
+    },
+    {
+      "epoch": 0.8636094424232168,
+      "grad_norm": 0.14252302050590515,
+      "learning_rate": 0.001,
+      "loss": 1.8472,
+      "step": 20414
+    },
+    {
+      "epoch": 0.8636517471867332,
+      "grad_norm": 0.3455297350883484,
+      "learning_rate": 0.001,
+      "loss": 1.9293,
+      "step": 20415
+    },
+    {
+      "epoch": 0.8636940519502496,
+      "grad_norm": 0.14407595992088318,
+      "learning_rate": 0.001,
+      "loss": 2.3692,
+      "step": 20416
+    },
+    {
+      "epoch": 0.8637363567137659,
+      "grad_norm": 0.1891782134771347,
+      "learning_rate": 0.001,
+      "loss": 2.3615,
+      "step": 20417
+    },
+    {
+      "epoch": 0.8637786614772823,
+      "grad_norm": 0.17591020464897156,
+      "learning_rate": 0.001,
+      "loss": 2.9479,
+      "step": 20418
+    },
+    {
+      "epoch": 0.8638209662407987,
+      "grad_norm": 2.7036447525024414,
+      "learning_rate": 0.001,
+      "loss": 1.6538,
+      "step": 20419
+    },
+    {
+      "epoch": 0.863863271004315,
+      "grad_norm": 0.14303186535835266,
+      "learning_rate": 0.001,
+      "loss": 2.2801,
+      "step": 20420
+    },
+    {
+      "epoch": 0.8639055757678314,
+      "grad_norm": 0.15167652070522308,
+      "learning_rate": 0.001,
+      "loss": 2.1755,
+      "step": 20421
+    },
+    {
+      "epoch": 0.8639478805313479,
+      "grad_norm": 0.1477757692337036,
+      "learning_rate": 0.001,
+      "loss": 1.8386,
+      "step": 20422
+    },
+    {
+      "epoch": 0.8639901852948642,
+      "grad_norm": 0.1253604143857956,
+      "learning_rate": 0.001,
+      "loss": 1.2549,
+      "step": 20423
+    },
+    {
+      "epoch": 0.8640324900583806,
+      "grad_norm": 0.13061819970607758,
+      "learning_rate": 0.001,
+      "loss": 1.4896,
+      "step": 20424
+    },
+    {
+      "epoch": 0.864074794821897,
+      "grad_norm": 0.21047991514205933,
+      "learning_rate": 0.001,
+      "loss": 2.2363,
+      "step": 20425
+    },
+    {
+      "epoch": 0.8641170995854133,
+      "grad_norm": 0.1477625072002411,
+      "learning_rate": 0.001,
+      "loss": 1.5357,
+      "step": 20426
+    },
+    {
+      "epoch": 0.8641594043489297,
+      "grad_norm": 0.19324536621570587,
+      "learning_rate": 0.001,
+      "loss": 1.9138,
+      "step": 20427
+    },
+    {
+      "epoch": 0.8642017091124461,
+      "grad_norm": 0.508693516254425,
+      "learning_rate": 0.001,
+      "loss": 2.217,
+      "step": 20428
+    },
+    {
+      "epoch": 0.8642440138759624,
+      "grad_norm": 0.15548713505268097,
+      "learning_rate": 0.001,
+      "loss": 1.7976,
+      "step": 20429
+    },
+    {
+      "epoch": 0.8642863186394788,
+      "grad_norm": 0.18489326536655426,
+      "learning_rate": 0.001,
+      "loss": 3.005,
+      "step": 20430
+    },
+    {
+      "epoch": 0.8643286234029952,
+      "grad_norm": 0.1529025137424469,
+      "learning_rate": 0.001,
+      "loss": 2.1427,
+      "step": 20431
+    },
+    {
+      "epoch": 0.8643709281665115,
+      "grad_norm": 0.1769225001335144,
+      "learning_rate": 0.001,
+      "loss": 1.596,
+      "step": 20432
+    },
+    {
+      "epoch": 0.8644132329300279,
+      "grad_norm": 0.12776906788349152,
+      "learning_rate": 0.001,
+      "loss": 1.8726,
+      "step": 20433
+    },
+    {
+      "epoch": 0.8644555376935443,
+      "grad_norm": 0.1688249409198761,
+      "learning_rate": 0.001,
+      "loss": 1.9496,
+      "step": 20434
+    },
+    {
+      "epoch": 0.8644978424570606,
+      "grad_norm": 0.6416119337081909,
+      "learning_rate": 0.001,
+      "loss": 2.3831,
+      "step": 20435
+    },
+    {
+      "epoch": 0.864540147220577,
+      "grad_norm": 0.14832055568695068,
+      "learning_rate": 0.001,
+      "loss": 1.8445,
+      "step": 20436
+    },
+    {
+      "epoch": 0.8645824519840934,
+      "grad_norm": 0.25075727701187134,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 20437
+    },
+    {
+      "epoch": 0.8646247567476097,
+      "grad_norm": 0.15989474952220917,
+      "learning_rate": 0.001,
+      "loss": 1.8512,
+      "step": 20438
+    },
+    {
+      "epoch": 0.8646670615111262,
+      "grad_norm": 0.1592397540807724,
+      "learning_rate": 0.001,
+      "loss": 2.8129,
+      "step": 20439
+    },
+    {
+      "epoch": 0.8647093662746426,
+      "grad_norm": 0.14796599745750427,
+      "learning_rate": 0.001,
+      "loss": 1.5769,
+      "step": 20440
+    },
+    {
+      "epoch": 0.8647516710381589,
+      "grad_norm": 0.1716962605714798,
+      "learning_rate": 0.001,
+      "loss": 2.3114,
+      "step": 20441
+    },
+    {
+      "epoch": 0.8647939758016753,
+      "grad_norm": 0.4446561336517334,
+      "learning_rate": 0.001,
+      "loss": 1.3337,
+      "step": 20442
+    },
+    {
+      "epoch": 0.8648362805651917,
+      "grad_norm": 0.15819323062896729,
+      "learning_rate": 0.001,
+      "loss": 2.7212,
+      "step": 20443
+    },
+    {
+      "epoch": 0.864878585328708,
+      "grad_norm": 1.2492092847824097,
+      "learning_rate": 0.001,
+      "loss": 2.112,
+      "step": 20444
+    },
+    {
+      "epoch": 0.8649208900922244,
+      "grad_norm": 0.1683206856250763,
+      "learning_rate": 0.001,
+      "loss": 2.6859,
+      "step": 20445
+    },
+    {
+      "epoch": 0.8649631948557408,
+      "grad_norm": 0.18615072965621948,
+      "learning_rate": 0.001,
+      "loss": 2.0135,
+      "step": 20446
+    },
+    {
+      "epoch": 0.8650054996192571,
+      "grad_norm": 0.16364848613739014,
+      "learning_rate": 0.001,
+      "loss": 2.1575,
+      "step": 20447
+    },
+    {
+      "epoch": 0.8650478043827735,
+      "grad_norm": 0.1338125467300415,
+      "learning_rate": 0.001,
+      "loss": 1.5287,
+      "step": 20448
+    },
+    {
+      "epoch": 0.8650901091462899,
+      "grad_norm": 0.14159221947193146,
+      "learning_rate": 0.001,
+      "loss": 2.4444,
+      "step": 20449
+    },
+    {
+      "epoch": 0.8651324139098062,
+      "grad_norm": 0.15110106766223907,
+      "learning_rate": 0.001,
+      "loss": 2.3197,
+      "step": 20450
+    },
+    {
+      "epoch": 0.8651747186733226,
+      "grad_norm": 0.13722842931747437,
+      "learning_rate": 0.001,
+      "loss": 1.7881,
+      "step": 20451
+    },
+    {
+      "epoch": 0.865217023436839,
+      "grad_norm": 0.13986808061599731,
+      "learning_rate": 0.001,
+      "loss": 1.699,
+      "step": 20452
+    },
+    {
+      "epoch": 0.8652593282003553,
+      "grad_norm": 0.19141452014446259,
+      "learning_rate": 0.001,
+      "loss": 1.6697,
+      "step": 20453
+    },
+    {
+      "epoch": 0.8653016329638717,
+      "grad_norm": 0.15517202019691467,
+      "learning_rate": 0.001,
+      "loss": 3.0925,
+      "step": 20454
+    },
+    {
+      "epoch": 0.8653439377273882,
+      "grad_norm": 0.13340319693088531,
+      "learning_rate": 0.001,
+      "loss": 1.3833,
+      "step": 20455
+    },
+    {
+      "epoch": 0.8653862424909045,
+      "grad_norm": 0.14518952369689941,
+      "learning_rate": 0.001,
+      "loss": 1.6842,
+      "step": 20456
+    },
+    {
+      "epoch": 0.8654285472544209,
+      "grad_norm": 0.15267087519168854,
+      "learning_rate": 0.001,
+      "loss": 1.7472,
+      "step": 20457
+    },
+    {
+      "epoch": 0.8654708520179372,
+      "grad_norm": 0.16295918822288513,
+      "learning_rate": 0.001,
+      "loss": 1.6337,
+      "step": 20458
+    },
+    {
+      "epoch": 0.8655131567814536,
+      "grad_norm": 0.3994349241256714,
+      "learning_rate": 0.001,
+      "loss": 2.1025,
+      "step": 20459
+    },
+    {
+      "epoch": 0.86555546154497,
+      "grad_norm": 0.14686226844787598,
+      "learning_rate": 0.001,
+      "loss": 2.2015,
+      "step": 20460
+    },
+    {
+      "epoch": 0.8655977663084863,
+      "grad_norm": 2.716871500015259,
+      "learning_rate": 0.001,
+      "loss": 3.038,
+      "step": 20461
+    },
+    {
+      "epoch": 0.8656400710720027,
+      "grad_norm": 0.152231365442276,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 20462
+    },
+    {
+      "epoch": 0.8656823758355191,
+      "grad_norm": 0.16370278596878052,
+      "learning_rate": 0.001,
+      "loss": 2.3014,
+      "step": 20463
+    },
+    {
+      "epoch": 0.8657246805990354,
+      "grad_norm": 0.18813279271125793,
+      "learning_rate": 0.001,
+      "loss": 3.1802,
+      "step": 20464
+    },
+    {
+      "epoch": 0.8657669853625518,
+      "grad_norm": 2.1478288173675537,
+      "learning_rate": 0.001,
+      "loss": 2.8626,
+      "step": 20465
+    },
+    {
+      "epoch": 0.8658092901260682,
+      "grad_norm": 0.6545858979225159,
+      "learning_rate": 0.001,
+      "loss": 1.976,
+      "step": 20466
+    },
+    {
+      "epoch": 0.8658515948895845,
+      "grad_norm": 0.163694828748703,
+      "learning_rate": 0.001,
+      "loss": 1.7941,
+      "step": 20467
+    },
+    {
+      "epoch": 0.8658938996531009,
+      "grad_norm": 0.15311507880687714,
+      "learning_rate": 0.001,
+      "loss": 1.8408,
+      "step": 20468
+    },
+    {
+      "epoch": 0.8659362044166173,
+      "grad_norm": 0.13591715693473816,
+      "learning_rate": 0.001,
+      "loss": 1.8756,
+      "step": 20469
+    },
+    {
+      "epoch": 0.8659785091801336,
+      "grad_norm": 0.12307523190975189,
+      "learning_rate": 0.001,
+      "loss": 2.4576,
+      "step": 20470
+    },
+    {
+      "epoch": 0.86602081394365,
+      "grad_norm": 0.23300588130950928,
+      "learning_rate": 0.001,
+      "loss": 1.5551,
+      "step": 20471
+    },
+    {
+      "epoch": 0.8660631187071665,
+      "grad_norm": 0.30247703194618225,
+      "learning_rate": 0.001,
+      "loss": 1.9095,
+      "step": 20472
+    },
+    {
+      "epoch": 0.8661054234706828,
+      "grad_norm": 0.12208034843206406,
+      "learning_rate": 0.001,
+      "loss": 2.1441,
+      "step": 20473
+    },
+    {
+      "epoch": 0.8661477282341992,
+      "grad_norm": 0.427654504776001,
+      "learning_rate": 0.001,
+      "loss": 3.1424,
+      "step": 20474
+    },
+    {
+      "epoch": 0.8661900329977156,
+      "grad_norm": 0.15135295689105988,
+      "learning_rate": 0.001,
+      "loss": 1.4906,
+      "step": 20475
+    },
+    {
+      "epoch": 0.8662323377612319,
+      "grad_norm": 0.15214575827121735,
+      "learning_rate": 0.001,
+      "loss": 2.8835,
+      "step": 20476
+    },
+    {
+      "epoch": 0.8662746425247483,
+      "grad_norm": 0.3086780309677124,
+      "learning_rate": 0.001,
+      "loss": 3.6189,
+      "step": 20477
+    },
+    {
+      "epoch": 0.8663169472882647,
+      "grad_norm": 0.2229657918214798,
+      "learning_rate": 0.001,
+      "loss": 2.7436,
+      "step": 20478
+    },
+    {
+      "epoch": 0.866359252051781,
+      "grad_norm": 0.21993772685527802,
+      "learning_rate": 0.001,
+      "loss": 2.4626,
+      "step": 20479
+    },
+    {
+      "epoch": 0.8664015568152974,
+      "grad_norm": 7.104289531707764,
+      "learning_rate": 0.001,
+      "loss": 2.2925,
+      "step": 20480
+    },
+    {
+      "epoch": 0.8664438615788138,
+      "grad_norm": 0.18431110680103302,
+      "learning_rate": 0.001,
+      "loss": 2.1564,
+      "step": 20481
+    },
+    {
+      "epoch": 0.8664861663423301,
+      "grad_norm": 0.17499950528144836,
+      "learning_rate": 0.001,
+      "loss": 2.6352,
+      "step": 20482
+    },
+    {
+      "epoch": 0.8665284711058465,
+      "grad_norm": 0.17781849205493927,
+      "learning_rate": 0.001,
+      "loss": 2.3366,
+      "step": 20483
+    },
+    {
+      "epoch": 0.8665707758693629,
+      "grad_norm": 0.1824718415737152,
+      "learning_rate": 0.001,
+      "loss": 1.9301,
+      "step": 20484
+    },
+    {
+      "epoch": 0.8666130806328792,
+      "grad_norm": 0.14647133648395538,
+      "learning_rate": 0.001,
+      "loss": 2.0635,
+      "step": 20485
+    },
+    {
+      "epoch": 0.8666553853963956,
+      "grad_norm": 0.17452256381511688,
+      "learning_rate": 0.001,
+      "loss": 2.4909,
+      "step": 20486
+    },
+    {
+      "epoch": 0.866697690159912,
+      "grad_norm": 0.15065862238407135,
+      "learning_rate": 0.001,
+      "loss": 2.3384,
+      "step": 20487
+    },
+    {
+      "epoch": 0.8667399949234283,
+      "grad_norm": 0.19424915313720703,
+      "learning_rate": 0.001,
+      "loss": 2.3333,
+      "step": 20488
+    },
+    {
+      "epoch": 0.8667822996869448,
+      "grad_norm": 0.15558700263500214,
+      "learning_rate": 0.001,
+      "loss": 2.7007,
+      "step": 20489
+    },
+    {
+      "epoch": 0.8668246044504612,
+      "grad_norm": 0.14827044308185577,
+      "learning_rate": 0.001,
+      "loss": 1.6913,
+      "step": 20490
+    },
+    {
+      "epoch": 0.8668669092139775,
+      "grad_norm": 0.48052114248275757,
+      "learning_rate": 0.001,
+      "loss": 2.0873,
+      "step": 20491
+    },
+    {
+      "epoch": 0.8669092139774939,
+      "grad_norm": 0.1724882572889328,
+      "learning_rate": 0.001,
+      "loss": 2.2251,
+      "step": 20492
+    },
+    {
+      "epoch": 0.8669515187410103,
+      "grad_norm": 0.2082466185092926,
+      "learning_rate": 0.001,
+      "loss": 2.881,
+      "step": 20493
+    },
+    {
+      "epoch": 0.8669938235045266,
+      "grad_norm": 0.14403583109378815,
+      "learning_rate": 0.001,
+      "loss": 1.8366,
+      "step": 20494
+    },
+    {
+      "epoch": 0.867036128268043,
+      "grad_norm": 0.1699734479188919,
+      "learning_rate": 0.001,
+      "loss": 3.2508,
+      "step": 20495
+    },
+    {
+      "epoch": 0.8670784330315594,
+      "grad_norm": 0.1627493053674698,
+      "learning_rate": 0.001,
+      "loss": 1.6992,
+      "step": 20496
+    },
+    {
+      "epoch": 0.8671207377950757,
+      "grad_norm": 0.1512979418039322,
+      "learning_rate": 0.001,
+      "loss": 1.8145,
+      "step": 20497
+    },
+    {
+      "epoch": 0.8671630425585921,
+      "grad_norm": 0.1480405628681183,
+      "learning_rate": 0.001,
+      "loss": 2.0845,
+      "step": 20498
+    },
+    {
+      "epoch": 0.8672053473221085,
+      "grad_norm": 0.12783098220825195,
+      "learning_rate": 0.001,
+      "loss": 3.377,
+      "step": 20499
+    },
+    {
+      "epoch": 0.8672476520856248,
+      "grad_norm": 0.13607032597064972,
+      "learning_rate": 0.001,
+      "loss": 1.555,
+      "step": 20500
+    },
+    {
+      "epoch": 0.8672899568491412,
+      "grad_norm": 0.1391061395406723,
+      "learning_rate": 0.001,
+      "loss": 1.4297,
+      "step": 20501
+    },
+    {
+      "epoch": 0.8673322616126576,
+      "grad_norm": 0.12696748971939087,
+      "learning_rate": 0.001,
+      "loss": 1.5243,
+      "step": 20502
+    },
+    {
+      "epoch": 0.8673745663761739,
+      "grad_norm": 0.15007275342941284,
+      "learning_rate": 0.001,
+      "loss": 2.754,
+      "step": 20503
+    },
+    {
+      "epoch": 0.8674168711396903,
+      "grad_norm": 0.12282463908195496,
+      "learning_rate": 0.001,
+      "loss": 2.6723,
+      "step": 20504
+    },
+    {
+      "epoch": 0.8674591759032066,
+      "grad_norm": 0.14588630199432373,
+      "learning_rate": 0.001,
+      "loss": 2.4033,
+      "step": 20505
+    },
+    {
+      "epoch": 0.8675014806667231,
+      "grad_norm": 0.518017053604126,
+      "learning_rate": 0.001,
+      "loss": 2.2406,
+      "step": 20506
+    },
+    {
+      "epoch": 0.8675437854302395,
+      "grad_norm": 0.14224489033222198,
+      "learning_rate": 0.001,
+      "loss": 1.4533,
+      "step": 20507
+    },
+    {
+      "epoch": 0.8675860901937558,
+      "grad_norm": 0.13003142178058624,
+      "learning_rate": 0.001,
+      "loss": 2.5315,
+      "step": 20508
+    },
+    {
+      "epoch": 0.8676283949572722,
+      "grad_norm": 0.14314842224121094,
+      "learning_rate": 0.001,
+      "loss": 3.1466,
+      "step": 20509
+    },
+    {
+      "epoch": 0.8676706997207886,
+      "grad_norm": 0.18065118789672852,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 20510
+    },
+    {
+      "epoch": 0.8677130044843049,
+      "grad_norm": 0.16202561557292938,
+      "learning_rate": 0.001,
+      "loss": 1.7793,
+      "step": 20511
+    },
+    {
+      "epoch": 0.8677553092478213,
+      "grad_norm": 0.13975529372692108,
+      "learning_rate": 0.001,
+      "loss": 2.85,
+      "step": 20512
+    },
+    {
+      "epoch": 0.8677976140113377,
+      "grad_norm": 0.15059174597263336,
+      "learning_rate": 0.001,
+      "loss": 1.3228,
+      "step": 20513
+    },
+    {
+      "epoch": 0.867839918774854,
+      "grad_norm": 0.15195417404174805,
+      "learning_rate": 0.001,
+      "loss": 1.7705,
+      "step": 20514
+    },
+    {
+      "epoch": 0.8678822235383704,
+      "grad_norm": 0.159585103392601,
+      "learning_rate": 0.001,
+      "loss": 2.5388,
+      "step": 20515
+    },
+    {
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.14191685616970062,
+      "learning_rate": 0.001,
+      "loss": 2.1223,
+      "step": 20516
+    },
+    {
+      "epoch": 0.8679668330654031,
+      "grad_norm": 0.18546472489833832,
+      "learning_rate": 0.001,
+      "loss": 2.4321,
+      "step": 20517
+    },
+    {
+      "epoch": 0.8680091378289195,
+      "grad_norm": 0.13588781654834747,
+      "learning_rate": 0.001,
+      "loss": 2.0784,
+      "step": 20518
+    },
+    {
+      "epoch": 0.8680514425924359,
+      "grad_norm": 15.1656494140625,
+      "learning_rate": 0.001,
+      "loss": 1.8606,
+      "step": 20519
+    },
+    {
+      "epoch": 0.8680937473559522,
+      "grad_norm": 0.1273522973060608,
+      "learning_rate": 0.001,
+      "loss": 1.9861,
+      "step": 20520
+    },
+    {
+      "epoch": 0.8681360521194686,
+      "grad_norm": 0.3258301317691803,
+      "learning_rate": 0.001,
+      "loss": 2.4265,
+      "step": 20521
+    },
+    {
+      "epoch": 0.8681783568829851,
+      "grad_norm": 0.1691277176141739,
+      "learning_rate": 0.001,
+      "loss": 1.8709,
+      "step": 20522
+    },
+    {
+      "epoch": 0.8682206616465014,
+      "grad_norm": 0.1776563823223114,
+      "learning_rate": 0.001,
+      "loss": 1.9714,
+      "step": 20523
+    },
+    {
+      "epoch": 0.8682629664100178,
+      "grad_norm": 0.19368451833724976,
+      "learning_rate": 0.001,
+      "loss": 3.038,
+      "step": 20524
+    },
+    {
+      "epoch": 0.8683052711735342,
+      "grad_norm": 0.15642546117305756,
+      "learning_rate": 0.001,
+      "loss": 2.7047,
+      "step": 20525
+    },
+    {
+      "epoch": 0.8683475759370505,
+      "grad_norm": 0.23724526166915894,
+      "learning_rate": 0.001,
+      "loss": 2.3636,
+      "step": 20526
+    },
+    {
+      "epoch": 0.8683898807005669,
+      "grad_norm": 0.14943532645702362,
+      "learning_rate": 0.001,
+      "loss": 1.8734,
+      "step": 20527
+    },
+    {
+      "epoch": 0.8684321854640833,
+      "grad_norm": 0.15623928606510162,
+      "learning_rate": 0.001,
+      "loss": 2.4747,
+      "step": 20528
+    },
+    {
+      "epoch": 0.8684744902275996,
+      "grad_norm": 0.24184876680374146,
+      "learning_rate": 0.001,
+      "loss": 1.9066,
+      "step": 20529
+    },
+    {
+      "epoch": 0.868516794991116,
+      "grad_norm": 0.523157000541687,
+      "learning_rate": 0.001,
+      "loss": 2.8953,
+      "step": 20530
+    },
+    {
+      "epoch": 0.8685590997546324,
+      "grad_norm": 0.1399133801460266,
+      "learning_rate": 0.001,
+      "loss": 2.914,
+      "step": 20531
+    },
+    {
+      "epoch": 0.8686014045181487,
+      "grad_norm": 0.3917766809463501,
+      "learning_rate": 0.001,
+      "loss": 1.447,
+      "step": 20532
+    },
+    {
+      "epoch": 0.8686437092816651,
+      "grad_norm": 0.15621374547481537,
+      "learning_rate": 0.001,
+      "loss": 2.3168,
+      "step": 20533
+    },
+    {
+      "epoch": 0.8686860140451815,
+      "grad_norm": 0.12716899812221527,
+      "learning_rate": 0.001,
+      "loss": 1.4502,
+      "step": 20534
+    },
+    {
+      "epoch": 0.8687283188086978,
+      "grad_norm": 0.14455370604991913,
+      "learning_rate": 0.001,
+      "loss": 2.2383,
+      "step": 20535
+    },
+    {
+      "epoch": 0.8687706235722142,
+      "grad_norm": 0.14087846875190735,
+      "learning_rate": 0.001,
+      "loss": 2.0278,
+      "step": 20536
+    },
+    {
+      "epoch": 0.8688129283357306,
+      "grad_norm": 0.17462292313575745,
+      "learning_rate": 0.001,
+      "loss": 1.7217,
+      "step": 20537
+    },
+    {
+      "epoch": 0.868855233099247,
+      "grad_norm": 0.16990859806537628,
+      "learning_rate": 0.001,
+      "loss": 3.8208,
+      "step": 20538
+    },
+    {
+      "epoch": 0.8688975378627634,
+      "grad_norm": 0.131483256816864,
+      "learning_rate": 0.001,
+      "loss": 1.4108,
+      "step": 20539
+    },
+    {
+      "epoch": 0.8689398426262798,
+      "grad_norm": 0.13738611340522766,
+      "learning_rate": 0.001,
+      "loss": 2.644,
+      "step": 20540
+    },
+    {
+      "epoch": 0.8689821473897961,
+      "grad_norm": 0.2787913978099823,
+      "learning_rate": 0.001,
+      "loss": 1.9129,
+      "step": 20541
+    },
+    {
+      "epoch": 0.8690244521533125,
+      "grad_norm": 0.8234505653381348,
+      "learning_rate": 0.001,
+      "loss": 1.6344,
+      "step": 20542
+    },
+    {
+      "epoch": 0.8690667569168289,
+      "grad_norm": 0.15260133147239685,
+      "learning_rate": 0.001,
+      "loss": 1.8331,
+      "step": 20543
+    },
+    {
+      "epoch": 0.8691090616803452,
+      "grad_norm": 0.12822067737579346,
+      "learning_rate": 0.001,
+      "loss": 2.4685,
+      "step": 20544
+    },
+    {
+      "epoch": 0.8691513664438616,
+      "grad_norm": 0.14594188332557678,
+      "learning_rate": 0.001,
+      "loss": 1.5119,
+      "step": 20545
+    },
+    {
+      "epoch": 0.869193671207378,
+      "grad_norm": 0.14818280935287476,
+      "learning_rate": 0.001,
+      "loss": 2.4656,
+      "step": 20546
+    },
+    {
+      "epoch": 0.8692359759708943,
+      "grad_norm": 0.26366791129112244,
+      "learning_rate": 0.001,
+      "loss": 2.9585,
+      "step": 20547
+    },
+    {
+      "epoch": 0.8692782807344107,
+      "grad_norm": 0.17102564871311188,
+      "learning_rate": 0.001,
+      "loss": 2.0877,
+      "step": 20548
+    },
+    {
+      "epoch": 0.869320585497927,
+      "grad_norm": 0.1857893019914627,
+      "learning_rate": 0.001,
+      "loss": 2.5178,
+      "step": 20549
+    },
+    {
+      "epoch": 0.8693628902614434,
+      "grad_norm": 0.26888343691825867,
+      "learning_rate": 0.001,
+      "loss": 1.9708,
+      "step": 20550
+    },
+    {
+      "epoch": 0.8694051950249598,
+      "grad_norm": 0.4817790687084198,
+      "learning_rate": 0.001,
+      "loss": 1.3568,
+      "step": 20551
+    },
+    {
+      "epoch": 0.8694474997884761,
+      "grad_norm": 0.6389137506484985,
+      "learning_rate": 0.001,
+      "loss": 1.9148,
+      "step": 20552
+    },
+    {
+      "epoch": 0.8694898045519925,
+      "grad_norm": 0.17424023151397705,
+      "learning_rate": 0.001,
+      "loss": 1.8895,
+      "step": 20553
+    },
+    {
+      "epoch": 0.869532109315509,
+      "grad_norm": 0.15143568813800812,
+      "learning_rate": 0.001,
+      "loss": 1.5483,
+      "step": 20554
+    },
+    {
+      "epoch": 0.8695744140790252,
+      "grad_norm": 0.13719114661216736,
+      "learning_rate": 0.001,
+      "loss": 1.8091,
+      "step": 20555
+    },
+    {
+      "epoch": 0.8696167188425417,
+      "grad_norm": 0.1803659200668335,
+      "learning_rate": 0.001,
+      "loss": 2.2232,
+      "step": 20556
+    },
+    {
+      "epoch": 0.8696590236060581,
+      "grad_norm": 1.9404886960983276,
+      "learning_rate": 0.001,
+      "loss": 1.9271,
+      "step": 20557
+    },
+    {
+      "epoch": 0.8697013283695744,
+      "grad_norm": 0.4086678624153137,
+      "learning_rate": 0.001,
+      "loss": 2.0375,
+      "step": 20558
+    },
+    {
+      "epoch": 0.8697436331330908,
+      "grad_norm": 0.1896570473909378,
+      "learning_rate": 0.001,
+      "loss": 2.4307,
+      "step": 20559
+    },
+    {
+      "epoch": 0.8697859378966072,
+      "grad_norm": 0.1653144508600235,
+      "learning_rate": 0.001,
+      "loss": 2.1597,
+      "step": 20560
+    },
+    {
+      "epoch": 0.8698282426601235,
+      "grad_norm": 0.38588979840278625,
+      "learning_rate": 0.001,
+      "loss": 1.9474,
+      "step": 20561
+    },
+    {
+      "epoch": 0.8698705474236399,
+      "grad_norm": 0.20403960347175598,
+      "learning_rate": 0.001,
+      "loss": 2.1097,
+      "step": 20562
+    },
+    {
+      "epoch": 0.8699128521871563,
+      "grad_norm": 69.25297546386719,
+      "learning_rate": 0.001,
+      "loss": 2.0653,
+      "step": 20563
+    },
+    {
+      "epoch": 0.8699551569506726,
+      "grad_norm": 7.444055557250977,
+      "learning_rate": 0.001,
+      "loss": 1.743,
+      "step": 20564
+    },
+    {
+      "epoch": 0.869997461714189,
+      "grad_norm": 0.1590234786272049,
+      "learning_rate": 0.001,
+      "loss": 1.8023,
+      "step": 20565
+    },
+    {
+      "epoch": 0.8700397664777054,
+      "grad_norm": 0.3205440640449524,
+      "learning_rate": 0.001,
+      "loss": 1.9231,
+      "step": 20566
+    },
+    {
+      "epoch": 0.8700820712412217,
+      "grad_norm": 0.19606617093086243,
+      "learning_rate": 0.001,
+      "loss": 2.6873,
+      "step": 20567
+    },
+    {
+      "epoch": 0.8701243760047381,
+      "grad_norm": 0.44805920124053955,
+      "learning_rate": 0.001,
+      "loss": 2.4899,
+      "step": 20568
+    },
+    {
+      "epoch": 0.8701666807682545,
+      "grad_norm": 0.22577597200870514,
+      "learning_rate": 0.001,
+      "loss": 2.2703,
+      "step": 20569
+    },
+    {
+      "epoch": 0.8702089855317708,
+      "grad_norm": 0.168071910738945,
+      "learning_rate": 0.001,
+      "loss": 1.7944,
+      "step": 20570
+    },
+    {
+      "epoch": 0.8702512902952872,
+      "grad_norm": 0.17601118981838226,
+      "learning_rate": 0.001,
+      "loss": 1.8357,
+      "step": 20571
+    },
+    {
+      "epoch": 0.8702935950588037,
+      "grad_norm": 0.1971052885055542,
+      "learning_rate": 0.001,
+      "loss": 1.9707,
+      "step": 20572
+    },
+    {
+      "epoch": 0.87033589982232,
+      "grad_norm": 0.16236917674541473,
+      "learning_rate": 0.001,
+      "loss": 1.9057,
+      "step": 20573
+    },
+    {
+      "epoch": 0.8703782045858364,
+      "grad_norm": 0.22024036943912506,
+      "learning_rate": 0.001,
+      "loss": 1.9232,
+      "step": 20574
+    },
+    {
+      "epoch": 0.8704205093493528,
+      "grad_norm": 0.18604035675525665,
+      "learning_rate": 0.001,
+      "loss": 2.3201,
+      "step": 20575
+    },
+    {
+      "epoch": 0.8704628141128691,
+      "grad_norm": 0.19944395124912262,
+      "learning_rate": 0.001,
+      "loss": 2.2175,
+      "step": 20576
+    },
+    {
+      "epoch": 0.8705051188763855,
+      "grad_norm": 6.948441982269287,
+      "learning_rate": 0.001,
+      "loss": 2.322,
+      "step": 20577
+    },
+    {
+      "epoch": 0.8705474236399019,
+      "grad_norm": 0.19903208315372467,
+      "learning_rate": 0.001,
+      "loss": 2.2643,
+      "step": 20578
+    },
+    {
+      "epoch": 0.8705897284034182,
+      "grad_norm": 0.15716981887817383,
+      "learning_rate": 0.001,
+      "loss": 1.4093,
+      "step": 20579
+    },
+    {
+      "epoch": 0.8706320331669346,
+      "grad_norm": 0.13574804365634918,
+      "learning_rate": 0.001,
+      "loss": 2.0119,
+      "step": 20580
+    },
+    {
+      "epoch": 0.870674337930451,
+      "grad_norm": 0.15148890018463135,
+      "learning_rate": 0.001,
+      "loss": 2.0046,
+      "step": 20581
+    },
+    {
+      "epoch": 0.8707166426939673,
+      "grad_norm": 0.15020988881587982,
+      "learning_rate": 0.001,
+      "loss": 2.022,
+      "step": 20582
+    },
+    {
+      "epoch": 0.8707589474574837,
+      "grad_norm": 0.15000440180301666,
+      "learning_rate": 0.001,
+      "loss": 2.3382,
+      "step": 20583
+    },
+    {
+      "epoch": 0.8708012522210001,
+      "grad_norm": 0.13530394434928894,
+      "learning_rate": 0.001,
+      "loss": 1.8668,
+      "step": 20584
+    },
+    {
+      "epoch": 0.8708435569845164,
+      "grad_norm": 0.17123256623744965,
+      "learning_rate": 0.001,
+      "loss": 2.1002,
+      "step": 20585
+    },
+    {
+      "epoch": 0.8708858617480328,
+      "grad_norm": 0.15746386349201202,
+      "learning_rate": 0.001,
+      "loss": 2.6803,
+      "step": 20586
+    },
+    {
+      "epoch": 0.8709281665115493,
+      "grad_norm": 3.3827056884765625,
+      "learning_rate": 0.001,
+      "loss": 2.1643,
+      "step": 20587
+    },
+    {
+      "epoch": 0.8709704712750655,
+      "grad_norm": 0.16982562839984894,
+      "learning_rate": 0.001,
+      "loss": 2.244,
+      "step": 20588
+    },
+    {
+      "epoch": 0.871012776038582,
+      "grad_norm": 2.7574925422668457,
+      "learning_rate": 0.001,
+      "loss": 2.6158,
+      "step": 20589
+    },
+    {
+      "epoch": 0.8710550808020984,
+      "grad_norm": 0.18137569725513458,
+      "learning_rate": 0.001,
+      "loss": 1.7065,
+      "step": 20590
+    },
+    {
+      "epoch": 0.8710973855656147,
+      "grad_norm": 0.18702997267246246,
+      "learning_rate": 0.001,
+      "loss": 1.7625,
+      "step": 20591
+    },
+    {
+      "epoch": 0.8711396903291311,
+      "grad_norm": 0.21095755696296692,
+      "learning_rate": 0.001,
+      "loss": 2.8752,
+      "step": 20592
+    },
+    {
+      "epoch": 0.8711819950926475,
+      "grad_norm": 0.410576730966568,
+      "learning_rate": 0.001,
+      "loss": 1.894,
+      "step": 20593
+    },
+    {
+      "epoch": 0.8712242998561638,
+      "grad_norm": 0.19124916195869446,
+      "learning_rate": 0.001,
+      "loss": 1.6841,
+      "step": 20594
+    },
+    {
+      "epoch": 0.8712666046196802,
+      "grad_norm": 0.1525120586156845,
+      "learning_rate": 0.001,
+      "loss": 2.486,
+      "step": 20595
+    },
+    {
+      "epoch": 0.8713089093831965,
+      "grad_norm": 0.2803274393081665,
+      "learning_rate": 0.001,
+      "loss": 2.2124,
+      "step": 20596
+    },
+    {
+      "epoch": 0.8713512141467129,
+      "grad_norm": 6.939259052276611,
+      "learning_rate": 0.001,
+      "loss": 1.5995,
+      "step": 20597
+    },
+    {
+      "epoch": 0.8713935189102293,
+      "grad_norm": 0.14391854405403137,
+      "learning_rate": 0.001,
+      "loss": 2.7667,
+      "step": 20598
+    },
+    {
+      "epoch": 0.8714358236737456,
+      "grad_norm": 0.15434549748897552,
+      "learning_rate": 0.001,
+      "loss": 1.9091,
+      "step": 20599
+    },
+    {
+      "epoch": 0.871478128437262,
+      "grad_norm": 0.17144039273262024,
+      "learning_rate": 0.001,
+      "loss": 1.7429,
+      "step": 20600
+    },
+    {
+      "epoch": 0.8715204332007784,
+      "grad_norm": 0.15034642815589905,
+      "learning_rate": 0.001,
+      "loss": 1.6829,
+      "step": 20601
+    },
+    {
+      "epoch": 0.8715627379642947,
+      "grad_norm": 0.17373819649219513,
+      "learning_rate": 0.001,
+      "loss": 1.9474,
+      "step": 20602
+    },
+    {
+      "epoch": 0.8716050427278111,
+      "grad_norm": 0.7324623465538025,
+      "learning_rate": 0.001,
+      "loss": 2.0428,
+      "step": 20603
+    },
+    {
+      "epoch": 0.8716473474913276,
+      "grad_norm": 1.5604692697525024,
+      "learning_rate": 0.001,
+      "loss": 3.489,
+      "step": 20604
+    },
+    {
+      "epoch": 0.8716896522548438,
+      "grad_norm": 0.2737126350402832,
+      "learning_rate": 0.001,
+      "loss": 2.5385,
+      "step": 20605
+    },
+    {
+      "epoch": 0.8717319570183603,
+      "grad_norm": 0.20344385504722595,
+      "learning_rate": 0.001,
+      "loss": 2.0625,
+      "step": 20606
+    },
+    {
+      "epoch": 0.8717742617818767,
+      "grad_norm": 0.16309478878974915,
+      "learning_rate": 0.001,
+      "loss": 2.4926,
+      "step": 20607
+    },
+    {
+      "epoch": 0.871816566545393,
+      "grad_norm": 0.17930370569229126,
+      "learning_rate": 0.001,
+      "loss": 1.9174,
+      "step": 20608
+    },
+    {
+      "epoch": 0.8718588713089094,
+      "grad_norm": 0.25973576307296753,
+      "learning_rate": 0.001,
+      "loss": 2.8351,
+      "step": 20609
+    },
+    {
+      "epoch": 0.8719011760724258,
+      "grad_norm": 2.8675005435943604,
+      "learning_rate": 0.001,
+      "loss": 3.2205,
+      "step": 20610
+    },
+    {
+      "epoch": 0.8719434808359421,
+      "grad_norm": 0.17134712636470795,
+      "learning_rate": 0.001,
+      "loss": 2.1032,
+      "step": 20611
+    },
+    {
+      "epoch": 0.8719857855994585,
+      "grad_norm": 0.14888092875480652,
+      "learning_rate": 0.001,
+      "loss": 2.7499,
+      "step": 20612
+    },
+    {
+      "epoch": 0.8720280903629749,
+      "grad_norm": 0.13400807976722717,
+      "learning_rate": 0.001,
+      "loss": 1.7189,
+      "step": 20613
+    },
+    {
+      "epoch": 0.8720703951264912,
+      "grad_norm": 0.12388193607330322,
+      "learning_rate": 0.001,
+      "loss": 2.486,
+      "step": 20614
+    },
+    {
+      "epoch": 0.8721126998900076,
+      "grad_norm": 0.1444062441587448,
+      "learning_rate": 0.001,
+      "loss": 1.9426,
+      "step": 20615
+    },
+    {
+      "epoch": 0.872155004653524,
+      "grad_norm": 0.13056041300296783,
+      "learning_rate": 0.001,
+      "loss": 1.5044,
+      "step": 20616
+    },
+    {
+      "epoch": 0.8721973094170403,
+      "grad_norm": 0.1537250131368637,
+      "learning_rate": 0.001,
+      "loss": 2.2731,
+      "step": 20617
+    },
+    {
+      "epoch": 0.8722396141805567,
+      "grad_norm": 0.21203458309173584,
+      "learning_rate": 0.001,
+      "loss": 2.0906,
+      "step": 20618
+    },
+    {
+      "epoch": 0.8722819189440731,
+      "grad_norm": 0.19286859035491943,
+      "learning_rate": 0.001,
+      "loss": 1.6875,
+      "step": 20619
+    },
+    {
+      "epoch": 0.8723242237075894,
+      "grad_norm": 2.2411351203918457,
+      "learning_rate": 0.001,
+      "loss": 2.294,
+      "step": 20620
+    },
+    {
+      "epoch": 0.8723665284711059,
+      "grad_norm": 0.12386021763086319,
+      "learning_rate": 0.001,
+      "loss": 1.9246,
+      "step": 20621
+    },
+    {
+      "epoch": 0.8724088332346223,
+      "grad_norm": 0.1696896106004715,
+      "learning_rate": 0.001,
+      "loss": 2.8406,
+      "step": 20622
+    },
+    {
+      "epoch": 0.8724511379981386,
+      "grad_norm": 0.14747262001037598,
+      "learning_rate": 0.001,
+      "loss": 2.1123,
+      "step": 20623
+    },
+    {
+      "epoch": 0.872493442761655,
+      "grad_norm": 0.6882927417755127,
+      "learning_rate": 0.001,
+      "loss": 2.9728,
+      "step": 20624
+    },
+    {
+      "epoch": 0.8725357475251714,
+      "grad_norm": 0.13660912215709686,
+      "learning_rate": 0.001,
+      "loss": 1.3383,
+      "step": 20625
+    },
+    {
+      "epoch": 0.8725780522886877,
+      "grad_norm": 0.15439388155937195,
+      "learning_rate": 0.001,
+      "loss": 2.3997,
+      "step": 20626
+    },
+    {
+      "epoch": 0.8726203570522041,
+      "grad_norm": 0.16791464388370514,
+      "learning_rate": 0.001,
+      "loss": 2.0219,
+      "step": 20627
+    },
+    {
+      "epoch": 0.8726626618157205,
+      "grad_norm": 0.17568817734718323,
+      "learning_rate": 0.001,
+      "loss": 2.2818,
+      "step": 20628
+    },
+    {
+      "epoch": 0.8727049665792368,
+      "grad_norm": 0.1622392237186432,
+      "learning_rate": 0.001,
+      "loss": 2.0132,
+      "step": 20629
+    },
+    {
+      "epoch": 0.8727472713427532,
+      "grad_norm": 0.15402914583683014,
+      "learning_rate": 0.001,
+      "loss": 1.813,
+      "step": 20630
+    },
+    {
+      "epoch": 0.8727895761062696,
+      "grad_norm": 0.13786296546459198,
+      "learning_rate": 0.001,
+      "loss": 1.9361,
+      "step": 20631
+    },
+    {
+      "epoch": 0.8728318808697859,
+      "grad_norm": 0.19948144257068634,
+      "learning_rate": 0.001,
+      "loss": 2.8847,
+      "step": 20632
+    },
+    {
+      "epoch": 0.8728741856333023,
+      "grad_norm": 0.1480330228805542,
+      "learning_rate": 0.001,
+      "loss": 1.8211,
+      "step": 20633
+    },
+    {
+      "epoch": 0.8729164903968187,
+      "grad_norm": 0.16000570356845856,
+      "learning_rate": 0.001,
+      "loss": 1.5714,
+      "step": 20634
+    },
+    {
+      "epoch": 0.872958795160335,
+      "grad_norm": 1.0622891187667847,
+      "learning_rate": 0.001,
+      "loss": 2.1747,
+      "step": 20635
+    },
+    {
+      "epoch": 0.8730010999238514,
+      "grad_norm": 0.192660391330719,
+      "learning_rate": 0.001,
+      "loss": 2.0402,
+      "step": 20636
+    },
+    {
+      "epoch": 0.8730434046873679,
+      "grad_norm": 0.17580977082252502,
+      "learning_rate": 0.001,
+      "loss": 1.9806,
+      "step": 20637
+    },
+    {
+      "epoch": 0.8730857094508842,
+      "grad_norm": 0.14421142637729645,
+      "learning_rate": 0.001,
+      "loss": 1.8746,
+      "step": 20638
+    },
+    {
+      "epoch": 0.8731280142144006,
+      "grad_norm": 0.14612741768360138,
+      "learning_rate": 0.001,
+      "loss": 1.8898,
+      "step": 20639
+    },
+    {
+      "epoch": 0.8731703189779169,
+      "grad_norm": 0.4607209861278534,
+      "learning_rate": 0.001,
+      "loss": 1.58,
+      "step": 20640
+    },
+    {
+      "epoch": 0.8732126237414333,
+      "grad_norm": 0.13212980329990387,
+      "learning_rate": 0.001,
+      "loss": 2.2389,
+      "step": 20641
+    },
+    {
+      "epoch": 0.8732549285049497,
+      "grad_norm": 0.14031372964382172,
+      "learning_rate": 0.001,
+      "loss": 2.0144,
+      "step": 20642
+    },
+    {
+      "epoch": 0.873297233268466,
+      "grad_norm": 0.1437501460313797,
+      "learning_rate": 0.001,
+      "loss": 1.7929,
+      "step": 20643
+    },
+    {
+      "epoch": 0.8733395380319824,
+      "grad_norm": 0.13045533001422882,
+      "learning_rate": 0.001,
+      "loss": 2.6569,
+      "step": 20644
+    },
+    {
+      "epoch": 0.8733818427954988,
+      "grad_norm": 0.16348956525325775,
+      "learning_rate": 0.001,
+      "loss": 2.2738,
+      "step": 20645
+    },
+    {
+      "epoch": 0.8734241475590151,
+      "grad_norm": 0.14668866991996765,
+      "learning_rate": 0.001,
+      "loss": 2.4409,
+      "step": 20646
+    },
+    {
+      "epoch": 0.8734664523225315,
+      "grad_norm": 0.12461234629154205,
+      "learning_rate": 0.001,
+      "loss": 2.0134,
+      "step": 20647
+    },
+    {
+      "epoch": 0.8735087570860479,
+      "grad_norm": 0.16764739155769348,
+      "learning_rate": 0.001,
+      "loss": 1.9933,
+      "step": 20648
+    },
+    {
+      "epoch": 0.8735510618495642,
+      "grad_norm": 0.15134289860725403,
+      "learning_rate": 0.001,
+      "loss": 2.7479,
+      "step": 20649
+    },
+    {
+      "epoch": 0.8735933666130806,
+      "grad_norm": 0.16396795213222504,
+      "learning_rate": 0.001,
+      "loss": 2.1933,
+      "step": 20650
+    },
+    {
+      "epoch": 0.873635671376597,
+      "grad_norm": 0.15628643333911896,
+      "learning_rate": 0.001,
+      "loss": 1.9693,
+      "step": 20651
+    },
+    {
+      "epoch": 0.8736779761401133,
+      "grad_norm": 0.15194100141525269,
+      "learning_rate": 0.001,
+      "loss": 2.0662,
+      "step": 20652
+    },
+    {
+      "epoch": 0.8737202809036297,
+      "grad_norm": 0.14548543095588684,
+      "learning_rate": 0.001,
+      "loss": 1.4378,
+      "step": 20653
+    },
+    {
+      "epoch": 0.8737625856671462,
+      "grad_norm": 0.4701218903064728,
+      "learning_rate": 0.001,
+      "loss": 2.1941,
+      "step": 20654
+    },
+    {
+      "epoch": 0.8738048904306625,
+      "grad_norm": 0.2454076111316681,
+      "learning_rate": 0.001,
+      "loss": 1.673,
+      "step": 20655
+    },
+    {
+      "epoch": 0.8738471951941789,
+      "grad_norm": 0.1675247997045517,
+      "learning_rate": 0.001,
+      "loss": 1.6599,
+      "step": 20656
+    },
+    {
+      "epoch": 0.8738894999576953,
+      "grad_norm": 0.1386534571647644,
+      "learning_rate": 0.001,
+      "loss": 2.4196,
+      "step": 20657
+    },
+    {
+      "epoch": 0.8739318047212116,
+      "grad_norm": 0.1279524713754654,
+      "learning_rate": 0.001,
+      "loss": 1.4271,
+      "step": 20658
+    },
+    {
+      "epoch": 0.873974109484728,
+      "grad_norm": 1.2931326627731323,
+      "learning_rate": 0.001,
+      "loss": 1.8761,
+      "step": 20659
+    },
+    {
+      "epoch": 0.8740164142482444,
+      "grad_norm": 0.18578656017780304,
+      "learning_rate": 0.001,
+      "loss": 1.5095,
+      "step": 20660
+    },
+    {
+      "epoch": 0.8740587190117607,
+      "grad_norm": 0.17007368803024292,
+      "learning_rate": 0.001,
+      "loss": 1.7172,
+      "step": 20661
+    },
+    {
+      "epoch": 0.8741010237752771,
+      "grad_norm": 0.1408626288175583,
+      "learning_rate": 0.001,
+      "loss": 2.1397,
+      "step": 20662
+    },
+    {
+      "epoch": 0.8741433285387935,
+      "grad_norm": 0.1541232317686081,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 20663
+    },
+    {
+      "epoch": 0.8741856333023098,
+      "grad_norm": 1.1258147954940796,
+      "learning_rate": 0.001,
+      "loss": 1.9507,
+      "step": 20664
+    },
+    {
+      "epoch": 0.8742279380658262,
+      "grad_norm": 0.13171681761741638,
+      "learning_rate": 0.001,
+      "loss": 2.7839,
+      "step": 20665
+    },
+    {
+      "epoch": 0.8742702428293426,
+      "grad_norm": 0.13253676891326904,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 20666
+    },
+    {
+      "epoch": 0.8743125475928589,
+      "grad_norm": 0.1618097722530365,
+      "learning_rate": 0.001,
+      "loss": 2.0003,
+      "step": 20667
+    },
+    {
+      "epoch": 0.8743548523563753,
+      "grad_norm": 0.1612173318862915,
+      "learning_rate": 0.001,
+      "loss": 1.7852,
+      "step": 20668
+    },
+    {
+      "epoch": 0.8743971571198917,
+      "grad_norm": 0.20455177128314972,
+      "learning_rate": 0.001,
+      "loss": 1.7746,
+      "step": 20669
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 0.2330559939146042,
+      "learning_rate": 0.001,
+      "loss": 2.5461,
+      "step": 20670
+    },
+    {
+      "epoch": 0.8744817666469245,
+      "grad_norm": 4.225989818572998,
+      "learning_rate": 0.001,
+      "loss": 2.0473,
+      "step": 20671
+    },
+    {
+      "epoch": 0.8745240714104409,
+      "grad_norm": 4.388603210449219,
+      "learning_rate": 0.001,
+      "loss": 2.6031,
+      "step": 20672
+    },
+    {
+      "epoch": 0.8745663761739572,
+      "grad_norm": 0.6356297135353088,
+      "learning_rate": 0.001,
+      "loss": 1.7837,
+      "step": 20673
+    },
+    {
+      "epoch": 0.8746086809374736,
+      "grad_norm": 0.5951091647148132,
+      "learning_rate": 0.001,
+      "loss": 2.1406,
+      "step": 20674
+    },
+    {
+      "epoch": 0.87465098570099,
+      "grad_norm": 0.1605328619480133,
+      "learning_rate": 0.001,
+      "loss": 1.8724,
+      "step": 20675
+    },
+    {
+      "epoch": 0.8746932904645063,
+      "grad_norm": 0.22600284218788147,
+      "learning_rate": 0.001,
+      "loss": 1.5535,
+      "step": 20676
+    },
+    {
+      "epoch": 0.8747355952280227,
+      "grad_norm": 0.15503069758415222,
+      "learning_rate": 0.001,
+      "loss": 1.7348,
+      "step": 20677
+    },
+    {
+      "epoch": 0.8747778999915391,
+      "grad_norm": 0.5110114216804504,
+      "learning_rate": 0.001,
+      "loss": 2.3343,
+      "step": 20678
+    },
+    {
+      "epoch": 0.8748202047550554,
+      "grad_norm": 0.16068077087402344,
+      "learning_rate": 0.001,
+      "loss": 2.4127,
+      "step": 20679
+    },
+    {
+      "epoch": 0.8748625095185718,
+      "grad_norm": 4.883456230163574,
+      "learning_rate": 0.001,
+      "loss": 2.3072,
+      "step": 20680
+    },
+    {
+      "epoch": 0.8749048142820882,
+      "grad_norm": 0.1795644909143448,
+      "learning_rate": 0.001,
+      "loss": 2.4872,
+      "step": 20681
+    },
+    {
+      "epoch": 0.8749471190456045,
+      "grad_norm": 0.14965955913066864,
+      "learning_rate": 0.001,
+      "loss": 2.6468,
+      "step": 20682
+    },
+    {
+      "epoch": 0.8749894238091209,
+      "grad_norm": 0.17840637266635895,
+      "learning_rate": 0.001,
+      "loss": 1.908,
+      "step": 20683
+    },
+    {
+      "epoch": 0.8750317285726372,
+      "grad_norm": 0.18810167908668518,
+      "learning_rate": 0.001,
+      "loss": 1.8151,
+      "step": 20684
+    },
+    {
+      "epoch": 0.8750740333361536,
+      "grad_norm": 0.20336748659610748,
+      "learning_rate": 0.001,
+      "loss": 2.9232,
+      "step": 20685
+    },
+    {
+      "epoch": 0.87511633809967,
+      "grad_norm": 0.2506280243396759,
+      "learning_rate": 0.001,
+      "loss": 2.7538,
+      "step": 20686
+    },
+    {
+      "epoch": 0.8751586428631863,
+      "grad_norm": 0.24045564234256744,
+      "learning_rate": 0.001,
+      "loss": 2.1233,
+      "step": 20687
+    },
+    {
+      "epoch": 0.8752009476267028,
+      "grad_norm": 0.1784004122018814,
+      "learning_rate": 0.001,
+      "loss": 2.0315,
+      "step": 20688
+    },
+    {
+      "epoch": 0.8752432523902192,
+      "grad_norm": 0.19089365005493164,
+      "learning_rate": 0.001,
+      "loss": 3.481,
+      "step": 20689
+    },
+    {
+      "epoch": 0.8752855571537355,
+      "grad_norm": 0.15477164089679718,
+      "learning_rate": 0.001,
+      "loss": 2.0263,
+      "step": 20690
+    },
+    {
+      "epoch": 0.8753278619172519,
+      "grad_norm": 0.22737683355808258,
+      "learning_rate": 0.001,
+      "loss": 3.2862,
+      "step": 20691
+    },
+    {
+      "epoch": 0.8753701666807683,
+      "grad_norm": 0.142948180437088,
+      "learning_rate": 0.001,
+      "loss": 2.0765,
+      "step": 20692
+    },
+    {
+      "epoch": 0.8754124714442846,
+      "grad_norm": 0.2728869616985321,
+      "learning_rate": 0.001,
+      "loss": 1.7859,
+      "step": 20693
+    },
+    {
+      "epoch": 0.875454776207801,
+      "grad_norm": 2.2189605236053467,
+      "learning_rate": 0.001,
+      "loss": 2.2061,
+      "step": 20694
+    },
+    {
+      "epoch": 0.8754970809713174,
+      "grad_norm": 0.28525206446647644,
+      "learning_rate": 0.001,
+      "loss": 2.3435,
+      "step": 20695
+    },
+    {
+      "epoch": 0.8755393857348337,
+      "grad_norm": 2.395468235015869,
+      "learning_rate": 0.001,
+      "loss": 3.1819,
+      "step": 20696
+    },
+    {
+      "epoch": 0.8755816904983501,
+      "grad_norm": 0.16036246716976166,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 20697
+    },
+    {
+      "epoch": 0.8756239952618665,
+      "grad_norm": 0.15271493792533875,
+      "learning_rate": 0.001,
+      "loss": 2.0569,
+      "step": 20698
+    },
+    {
+      "epoch": 0.8756663000253828,
+      "grad_norm": 0.2659335732460022,
+      "learning_rate": 0.001,
+      "loss": 2.4376,
+      "step": 20699
+    },
+    {
+      "epoch": 0.8757086047888992,
+      "grad_norm": 0.6031797528266907,
+      "learning_rate": 0.001,
+      "loss": 2.0957,
+      "step": 20700
+    },
+    {
+      "epoch": 0.8757509095524156,
+      "grad_norm": 0.16728293895721436,
+      "learning_rate": 0.001,
+      "loss": 2.6738,
+      "step": 20701
+    },
+    {
+      "epoch": 0.8757932143159319,
+      "grad_norm": 0.17319273948669434,
+      "learning_rate": 0.001,
+      "loss": 2.0715,
+      "step": 20702
+    },
+    {
+      "epoch": 0.8758355190794483,
+      "grad_norm": 0.1556321531534195,
+      "learning_rate": 0.001,
+      "loss": 2.0015,
+      "step": 20703
+    },
+    {
+      "epoch": 0.8758778238429648,
+      "grad_norm": 0.15744656324386597,
+      "learning_rate": 0.001,
+      "loss": 1.8593,
+      "step": 20704
+    },
+    {
+      "epoch": 0.875920128606481,
+      "grad_norm": 0.13687211275100708,
+      "learning_rate": 0.001,
+      "loss": 1.7606,
+      "step": 20705
+    },
+    {
+      "epoch": 0.8759624333699975,
+      "grad_norm": 0.20388025045394897,
+      "learning_rate": 0.001,
+      "loss": 2.3,
+      "step": 20706
+    },
+    {
+      "epoch": 0.8760047381335139,
+      "grad_norm": 0.1699550747871399,
+      "learning_rate": 0.001,
+      "loss": 2.1128,
+      "step": 20707
+    },
+    {
+      "epoch": 0.8760470428970302,
+      "grad_norm": 0.17213287949562073,
+      "learning_rate": 0.001,
+      "loss": 3.0813,
+      "step": 20708
+    },
+    {
+      "epoch": 0.8760893476605466,
+      "grad_norm": 0.14642708003520966,
+      "learning_rate": 0.001,
+      "loss": 1.7773,
+      "step": 20709
+    },
+    {
+      "epoch": 0.876131652424063,
+      "grad_norm": 0.24943861365318298,
+      "learning_rate": 0.001,
+      "loss": 1.8321,
+      "step": 20710
+    },
+    {
+      "epoch": 0.8761739571875793,
+      "grad_norm": 0.26976755261421204,
+      "learning_rate": 0.001,
+      "loss": 1.9682,
+      "step": 20711
+    },
+    {
+      "epoch": 0.8762162619510957,
+      "grad_norm": 0.2460852712392807,
+      "learning_rate": 0.001,
+      "loss": 1.6967,
+      "step": 20712
+    },
+    {
+      "epoch": 0.8762585667146121,
+      "grad_norm": 0.13001763820648193,
+      "learning_rate": 0.001,
+      "loss": 1.8121,
+      "step": 20713
+    },
+    {
+      "epoch": 0.8763008714781284,
+      "grad_norm": 0.17002592980861664,
+      "learning_rate": 0.001,
+      "loss": 2.4601,
+      "step": 20714
+    },
+    {
+      "epoch": 0.8763431762416448,
+      "grad_norm": 0.15308377146720886,
+      "learning_rate": 0.001,
+      "loss": 2.1304,
+      "step": 20715
+    },
+    {
+      "epoch": 0.8763854810051612,
+      "grad_norm": 0.19930648803710938,
+      "learning_rate": 0.001,
+      "loss": 2.1169,
+      "step": 20716
+    },
+    {
+      "epoch": 0.8764277857686775,
+      "grad_norm": 0.15591351687908173,
+      "learning_rate": 0.001,
+      "loss": 3.023,
+      "step": 20717
+    },
+    {
+      "epoch": 0.8764700905321939,
+      "grad_norm": 0.38943207263946533,
+      "learning_rate": 0.001,
+      "loss": 3.6015,
+      "step": 20718
+    },
+    {
+      "epoch": 0.8765123952957103,
+      "grad_norm": 0.16425389051437378,
+      "learning_rate": 0.001,
+      "loss": 1.75,
+      "step": 20719
+    },
+    {
+      "epoch": 0.8765547000592266,
+      "grad_norm": 0.1728183478116989,
+      "learning_rate": 0.001,
+      "loss": 2.144,
+      "step": 20720
+    },
+    {
+      "epoch": 0.876597004822743,
+      "grad_norm": 1.0912002325057983,
+      "learning_rate": 0.001,
+      "loss": 2.133,
+      "step": 20721
+    },
+    {
+      "epoch": 0.8766393095862595,
+      "grad_norm": 0.6945961117744446,
+      "learning_rate": 0.001,
+      "loss": 3.1325,
+      "step": 20722
+    },
+    {
+      "epoch": 0.8766816143497758,
+      "grad_norm": 0.1368747353553772,
+      "learning_rate": 0.001,
+      "loss": 1.4968,
+      "step": 20723
+    },
+    {
+      "epoch": 0.8767239191132922,
+      "grad_norm": 0.1539839655160904,
+      "learning_rate": 0.001,
+      "loss": 3.2301,
+      "step": 20724
+    },
+    {
+      "epoch": 0.8767662238768086,
+      "grad_norm": 0.1409660428762436,
+      "learning_rate": 0.001,
+      "loss": 1.2512,
+      "step": 20725
+    },
+    {
+      "epoch": 0.8768085286403249,
+      "grad_norm": 0.5920034050941467,
+      "learning_rate": 0.001,
+      "loss": 1.7211,
+      "step": 20726
+    },
+    {
+      "epoch": 0.8768508334038413,
+      "grad_norm": 0.7836800217628479,
+      "learning_rate": 0.001,
+      "loss": 2.5221,
+      "step": 20727
+    },
+    {
+      "epoch": 0.8768931381673577,
+      "grad_norm": 0.1688200980424881,
+      "learning_rate": 0.001,
+      "loss": 1.8585,
+      "step": 20728
+    },
+    {
+      "epoch": 0.876935442930874,
+      "grad_norm": 2.877446174621582,
+      "learning_rate": 0.001,
+      "loss": 1.2657,
+      "step": 20729
+    },
+    {
+      "epoch": 0.8769777476943904,
+      "grad_norm": 0.1525486409664154,
+      "learning_rate": 0.001,
+      "loss": 2.808,
+      "step": 20730
+    },
+    {
+      "epoch": 0.8770200524579067,
+      "grad_norm": 0.17019006609916687,
+      "learning_rate": 0.001,
+      "loss": 2.4638,
+      "step": 20731
+    },
+    {
+      "epoch": 0.8770623572214231,
+      "grad_norm": 5.263099670410156,
+      "learning_rate": 0.001,
+      "loss": 2.6641,
+      "step": 20732
+    },
+    {
+      "epoch": 0.8771046619849395,
+      "grad_norm": 0.18145911395549774,
+      "learning_rate": 0.001,
+      "loss": 2.248,
+      "step": 20733
+    },
+    {
+      "epoch": 0.8771469667484558,
+      "grad_norm": 0.1718670129776001,
+      "learning_rate": 0.001,
+      "loss": 2.284,
+      "step": 20734
+    },
+    {
+      "epoch": 0.8771892715119722,
+      "grad_norm": 0.16717854142189026,
+      "learning_rate": 0.001,
+      "loss": 2.2931,
+      "step": 20735
+    },
+    {
+      "epoch": 0.8772315762754886,
+      "grad_norm": 0.17688849568367004,
+      "learning_rate": 0.001,
+      "loss": 2.4102,
+      "step": 20736
+    },
+    {
+      "epoch": 0.8772738810390049,
+      "grad_norm": 0.19576434791088104,
+      "learning_rate": 0.001,
+      "loss": 2.9707,
+      "step": 20737
+    },
+    {
+      "epoch": 0.8773161858025214,
+      "grad_norm": 0.16361911594867706,
+      "learning_rate": 0.001,
+      "loss": 1.8927,
+      "step": 20738
+    },
+    {
+      "epoch": 0.8773584905660378,
+      "grad_norm": 0.20847147703170776,
+      "learning_rate": 0.001,
+      "loss": 2.5862,
+      "step": 20739
+    },
+    {
+      "epoch": 0.8774007953295541,
+      "grad_norm": 0.1973280906677246,
+      "learning_rate": 0.001,
+      "loss": 1.5039,
+      "step": 20740
+    },
+    {
+      "epoch": 0.8774431000930705,
+      "grad_norm": 0.18212057650089264,
+      "learning_rate": 0.001,
+      "loss": 2.205,
+      "step": 20741
+    },
+    {
+      "epoch": 0.8774854048565869,
+      "grad_norm": 0.7270743250846863,
+      "learning_rate": 0.001,
+      "loss": 1.9571,
+      "step": 20742
+    },
+    {
+      "epoch": 0.8775277096201032,
+      "grad_norm": 0.18873396515846252,
+      "learning_rate": 0.001,
+      "loss": 1.9786,
+      "step": 20743
+    },
+    {
+      "epoch": 0.8775700143836196,
+      "grad_norm": 0.18425945937633514,
+      "learning_rate": 0.001,
+      "loss": 2.1827,
+      "step": 20744
+    },
+    {
+      "epoch": 0.877612319147136,
+      "grad_norm": 0.23821745812892914,
+      "learning_rate": 0.001,
+      "loss": 2.3047,
+      "step": 20745
+    },
+    {
+      "epoch": 0.8776546239106523,
+      "grad_norm": 0.16767555475234985,
+      "learning_rate": 0.001,
+      "loss": 3.2195,
+      "step": 20746
+    },
+    {
+      "epoch": 0.8776969286741687,
+      "grad_norm": 0.2009771168231964,
+      "learning_rate": 0.001,
+      "loss": 2.3288,
+      "step": 20747
+    },
+    {
+      "epoch": 0.8777392334376851,
+      "grad_norm": 0.19222243130207062,
+      "learning_rate": 0.001,
+      "loss": 2.1529,
+      "step": 20748
+    },
+    {
+      "epoch": 0.8777815382012014,
+      "grad_norm": 0.28680652379989624,
+      "learning_rate": 0.001,
+      "loss": 1.7814,
+      "step": 20749
+    },
+    {
+      "epoch": 0.8778238429647178,
+      "grad_norm": 0.18338562548160553,
+      "learning_rate": 0.001,
+      "loss": 3.939,
+      "step": 20750
+    },
+    {
+      "epoch": 0.8778661477282342,
+      "grad_norm": 0.9251672029495239,
+      "learning_rate": 0.001,
+      "loss": 1.9422,
+      "step": 20751
+    },
+    {
+      "epoch": 0.8779084524917505,
+      "grad_norm": 0.5910675525665283,
+      "learning_rate": 0.001,
+      "loss": 1.9307,
+      "step": 20752
+    },
+    {
+      "epoch": 0.877950757255267,
+      "grad_norm": 0.2055887132883072,
+      "learning_rate": 0.001,
+      "loss": 2.2307,
+      "step": 20753
+    },
+    {
+      "epoch": 0.8779930620187834,
+      "grad_norm": 0.27294355630874634,
+      "learning_rate": 0.001,
+      "loss": 2.7468,
+      "step": 20754
+    },
+    {
+      "epoch": 0.8780353667822997,
+      "grad_norm": 0.14302602410316467,
+      "learning_rate": 0.001,
+      "loss": 1.4932,
+      "step": 20755
+    },
+    {
+      "epoch": 0.8780776715458161,
+      "grad_norm": 0.15693305432796478,
+      "learning_rate": 0.001,
+      "loss": 2.1727,
+      "step": 20756
+    },
+    {
+      "epoch": 0.8781199763093325,
+      "grad_norm": 2.7029690742492676,
+      "learning_rate": 0.001,
+      "loss": 2.2942,
+      "step": 20757
+    },
+    {
+      "epoch": 0.8781622810728488,
+      "grad_norm": 0.16679178178310394,
+      "learning_rate": 0.001,
+      "loss": 2.4154,
+      "step": 20758
+    },
+    {
+      "epoch": 0.8782045858363652,
+      "grad_norm": 0.29728999733924866,
+      "learning_rate": 0.001,
+      "loss": 1.5379,
+      "step": 20759
+    },
+    {
+      "epoch": 0.8782468905998816,
+      "grad_norm": 0.7710385918617249,
+      "learning_rate": 0.001,
+      "loss": 2.386,
+      "step": 20760
+    },
+    {
+      "epoch": 0.8782891953633979,
+      "grad_norm": 0.2641301453113556,
+      "learning_rate": 0.001,
+      "loss": 1.7395,
+      "step": 20761
+    },
+    {
+      "epoch": 0.8783315001269143,
+      "grad_norm": 0.15455834567546844,
+      "learning_rate": 0.001,
+      "loss": 2.7225,
+      "step": 20762
+    },
+    {
+      "epoch": 0.8783738048904307,
+      "grad_norm": 1.485619068145752,
+      "learning_rate": 0.001,
+      "loss": 2.3048,
+      "step": 20763
+    },
+    {
+      "epoch": 0.878416109653947,
+      "grad_norm": 0.675614595413208,
+      "learning_rate": 0.001,
+      "loss": 1.5047,
+      "step": 20764
+    },
+    {
+      "epoch": 0.8784584144174634,
+      "grad_norm": 0.18644104897975922,
+      "learning_rate": 0.001,
+      "loss": 1.9652,
+      "step": 20765
+    },
+    {
+      "epoch": 0.8785007191809798,
+      "grad_norm": 0.1697007417678833,
+      "learning_rate": 0.001,
+      "loss": 1.9008,
+      "step": 20766
+    },
+    {
+      "epoch": 0.8785430239444961,
+      "grad_norm": 0.1929423213005066,
+      "learning_rate": 0.001,
+      "loss": 2.574,
+      "step": 20767
+    },
+    {
+      "epoch": 0.8785853287080125,
+      "grad_norm": 0.3777572810649872,
+      "learning_rate": 0.001,
+      "loss": 2.3181,
+      "step": 20768
+    },
+    {
+      "epoch": 0.878627633471529,
+      "grad_norm": 0.17042919993400574,
+      "learning_rate": 0.001,
+      "loss": 2.2234,
+      "step": 20769
+    },
+    {
+      "epoch": 0.8786699382350452,
+      "grad_norm": 0.19270406663417816,
+      "learning_rate": 0.001,
+      "loss": 1.4735,
+      "step": 20770
+    },
+    {
+      "epoch": 0.8787122429985617,
+      "grad_norm": 0.2564689517021179,
+      "learning_rate": 0.001,
+      "loss": 1.9759,
+      "step": 20771
+    },
+    {
+      "epoch": 0.8787545477620781,
+      "grad_norm": 4.633014678955078,
+      "learning_rate": 0.001,
+      "loss": 1.7299,
+      "step": 20772
+    },
+    {
+      "epoch": 0.8787968525255944,
+      "grad_norm": 0.14340277016162872,
+      "learning_rate": 0.001,
+      "loss": 1.7605,
+      "step": 20773
+    },
+    {
+      "epoch": 0.8788391572891108,
+      "grad_norm": 0.1704220324754715,
+      "learning_rate": 0.001,
+      "loss": 2.0632,
+      "step": 20774
+    },
+    {
+      "epoch": 0.8788814620526271,
+      "grad_norm": 0.17304351925849915,
+      "learning_rate": 0.001,
+      "loss": 2.2619,
+      "step": 20775
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.1750032901763916,
+      "learning_rate": 0.001,
+      "loss": 1.918,
+      "step": 20776
+    },
+    {
+      "epoch": 0.8789660715796599,
+      "grad_norm": 4.89594030380249,
+      "learning_rate": 0.001,
+      "loss": 1.3372,
+      "step": 20777
+    },
+    {
+      "epoch": 0.8790083763431762,
+      "grad_norm": 0.172307088971138,
+      "learning_rate": 0.001,
+      "loss": 2.11,
+      "step": 20778
+    },
+    {
+      "epoch": 0.8790506811066926,
+      "grad_norm": 0.1498982012271881,
+      "learning_rate": 0.001,
+      "loss": 3.3868,
+      "step": 20779
+    },
+    {
+      "epoch": 0.879092985870209,
+      "grad_norm": 0.1848546415567398,
+      "learning_rate": 0.001,
+      "loss": 1.8795,
+      "step": 20780
+    },
+    {
+      "epoch": 0.8791352906337253,
+      "grad_norm": 2.189716100692749,
+      "learning_rate": 0.001,
+      "loss": 3.0549,
+      "step": 20781
+    },
+    {
+      "epoch": 0.8791775953972417,
+      "grad_norm": 0.7227365970611572,
+      "learning_rate": 0.001,
+      "loss": 2.0596,
+      "step": 20782
+    },
+    {
+      "epoch": 0.8792199001607581,
+      "grad_norm": 0.1363682597875595,
+      "learning_rate": 0.001,
+      "loss": 1.9833,
+      "step": 20783
+    },
+    {
+      "epoch": 0.8792622049242744,
+      "grad_norm": 0.4821024239063263,
+      "learning_rate": 0.001,
+      "loss": 2.1675,
+      "step": 20784
+    },
+    {
+      "epoch": 0.8793045096877908,
+      "grad_norm": 0.16702604293823242,
+      "learning_rate": 0.001,
+      "loss": 2.4858,
+      "step": 20785
+    },
+    {
+      "epoch": 0.8793468144513072,
+      "grad_norm": 9.462590217590332,
+      "learning_rate": 0.001,
+      "loss": 1.7947,
+      "step": 20786
+    },
+    {
+      "epoch": 0.8793891192148235,
+      "grad_norm": 0.14354799687862396,
+      "learning_rate": 0.001,
+      "loss": 2.4982,
+      "step": 20787
+    },
+    {
+      "epoch": 0.87943142397834,
+      "grad_norm": 0.17250943183898926,
+      "learning_rate": 0.001,
+      "loss": 2.817,
+      "step": 20788
+    },
+    {
+      "epoch": 0.8794737287418564,
+      "grad_norm": 0.13783836364746094,
+      "learning_rate": 0.001,
+      "loss": 2.5655,
+      "step": 20789
+    },
+    {
+      "epoch": 0.8795160335053727,
+      "grad_norm": 0.18853969871997833,
+      "learning_rate": 0.001,
+      "loss": 2.2248,
+      "step": 20790
+    },
+    {
+      "epoch": 0.8795583382688891,
+      "grad_norm": 0.1598181277513504,
+      "learning_rate": 0.001,
+      "loss": 2.6343,
+      "step": 20791
+    },
+    {
+      "epoch": 0.8796006430324055,
+      "grad_norm": 0.16585583984851837,
+      "learning_rate": 0.001,
+      "loss": 1.8722,
+      "step": 20792
+    },
+    {
+      "epoch": 0.8796429477959218,
+      "grad_norm": 0.20554226636886597,
+      "learning_rate": 0.001,
+      "loss": 2.6009,
+      "step": 20793
+    },
+    {
+      "epoch": 0.8796852525594382,
+      "grad_norm": 2.5491397380828857,
+      "learning_rate": 0.001,
+      "loss": 1.7133,
+      "step": 20794
+    },
+    {
+      "epoch": 0.8797275573229546,
+      "grad_norm": 0.1531909704208374,
+      "learning_rate": 0.001,
+      "loss": 2.1605,
+      "step": 20795
+    },
+    {
+      "epoch": 0.8797698620864709,
+      "grad_norm": 0.16960613429546356,
+      "learning_rate": 0.001,
+      "loss": 2.6522,
+      "step": 20796
+    },
+    {
+      "epoch": 0.8798121668499873,
+      "grad_norm": 0.20041488111019135,
+      "learning_rate": 0.001,
+      "loss": 1.9546,
+      "step": 20797
+    },
+    {
+      "epoch": 0.8798544716135037,
+      "grad_norm": 0.20710138976573944,
+      "learning_rate": 0.001,
+      "loss": 2.7599,
+      "step": 20798
+    },
+    {
+      "epoch": 0.87989677637702,
+      "grad_norm": 0.153824120759964,
+      "learning_rate": 0.001,
+      "loss": 2.2814,
+      "step": 20799
+    },
+    {
+      "epoch": 0.8799390811405364,
+      "grad_norm": 0.1634339690208435,
+      "learning_rate": 0.001,
+      "loss": 2.3355,
+      "step": 20800
+    },
+    {
+      "epoch": 0.8799813859040528,
+      "grad_norm": 0.19159682095050812,
+      "learning_rate": 0.001,
+      "loss": 1.9803,
+      "step": 20801
+    },
+    {
+      "epoch": 0.8800236906675691,
+      "grad_norm": 0.40518084168434143,
+      "learning_rate": 0.001,
+      "loss": 2.6101,
+      "step": 20802
+    },
+    {
+      "epoch": 0.8800659954310855,
+      "grad_norm": 0.19787724316120148,
+      "learning_rate": 0.001,
+      "loss": 2.0322,
+      "step": 20803
+    },
+    {
+      "epoch": 0.880108300194602,
+      "grad_norm": 0.18997430801391602,
+      "learning_rate": 0.001,
+      "loss": 1.7237,
+      "step": 20804
+    },
+    {
+      "epoch": 0.8801506049581183,
+      "grad_norm": 0.4623604118824005,
+      "learning_rate": 0.001,
+      "loss": 2.3957,
+      "step": 20805
+    },
+    {
+      "epoch": 0.8801929097216347,
+      "grad_norm": 0.14696352183818817,
+      "learning_rate": 0.001,
+      "loss": 2.1396,
+      "step": 20806
+    },
+    {
+      "epoch": 0.8802352144851511,
+      "grad_norm": 0.2328002154827118,
+      "learning_rate": 0.001,
+      "loss": 2.0078,
+      "step": 20807
+    },
+    {
+      "epoch": 0.8802775192486674,
+      "grad_norm": 0.16920128464698792,
+      "learning_rate": 0.001,
+      "loss": 1.444,
+      "step": 20808
+    },
+    {
+      "epoch": 0.8803198240121838,
+      "grad_norm": 0.30305415391921997,
+      "learning_rate": 0.001,
+      "loss": 3.162,
+      "step": 20809
+    },
+    {
+      "epoch": 0.8803621287757002,
+      "grad_norm": 0.2750433087348938,
+      "learning_rate": 0.001,
+      "loss": 1.8929,
+      "step": 20810
+    },
+    {
+      "epoch": 0.8804044335392165,
+      "grad_norm": 0.38231226801872253,
+      "learning_rate": 0.001,
+      "loss": 2.9644,
+      "step": 20811
+    },
+    {
+      "epoch": 0.8804467383027329,
+      "grad_norm": 0.14932866394519806,
+      "learning_rate": 0.001,
+      "loss": 2.08,
+      "step": 20812
+    },
+    {
+      "epoch": 0.8804890430662493,
+      "grad_norm": 0.17217305302619934,
+      "learning_rate": 0.001,
+      "loss": 2.4867,
+      "step": 20813
+    },
+    {
+      "epoch": 0.8805313478297656,
+      "grad_norm": 0.1518954336643219,
+      "learning_rate": 0.001,
+      "loss": 2.6287,
+      "step": 20814
+    },
+    {
+      "epoch": 0.880573652593282,
+      "grad_norm": 0.1420203298330307,
+      "learning_rate": 0.001,
+      "loss": 2.0718,
+      "step": 20815
+    },
+    {
+      "epoch": 0.8806159573567984,
+      "grad_norm": 0.15746936202049255,
+      "learning_rate": 0.001,
+      "loss": 1.3726,
+      "step": 20816
+    },
+    {
+      "epoch": 0.8806582621203147,
+      "grad_norm": 0.20516842603683472,
+      "learning_rate": 0.001,
+      "loss": 2.5636,
+      "step": 20817
+    },
+    {
+      "epoch": 0.8807005668838311,
+      "grad_norm": 0.18061251938343048,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 20818
+    },
+    {
+      "epoch": 0.8807428716473474,
+      "grad_norm": 0.19963960349559784,
+      "learning_rate": 0.001,
+      "loss": 1.8834,
+      "step": 20819
+    },
+    {
+      "epoch": 0.8807851764108638,
+      "grad_norm": 0.2850932776927948,
+      "learning_rate": 0.001,
+      "loss": 2.3297,
+      "step": 20820
+    },
+    {
+      "epoch": 0.8808274811743803,
+      "grad_norm": 0.16823464632034302,
+      "learning_rate": 0.001,
+      "loss": 2.6427,
+      "step": 20821
+    },
+    {
+      "epoch": 0.8808697859378966,
+      "grad_norm": 0.16549649834632874,
+      "learning_rate": 0.001,
+      "loss": 2.9588,
+      "step": 20822
+    },
+    {
+      "epoch": 0.880912090701413,
+      "grad_norm": 0.18909013271331787,
+      "learning_rate": 0.001,
+      "loss": 2.0608,
+      "step": 20823
+    },
+    {
+      "epoch": 0.8809543954649294,
+      "grad_norm": 0.16195768117904663,
+      "learning_rate": 0.001,
+      "loss": 2.2387,
+      "step": 20824
+    },
+    {
+      "epoch": 0.8809967002284457,
+      "grad_norm": 0.14382191002368927,
+      "learning_rate": 0.001,
+      "loss": 2.3996,
+      "step": 20825
+    },
+    {
+      "epoch": 0.8810390049919621,
+      "grad_norm": 0.1621260792016983,
+      "learning_rate": 0.001,
+      "loss": 2.7321,
+      "step": 20826
+    },
+    {
+      "epoch": 0.8810813097554785,
+      "grad_norm": 0.13346368074417114,
+      "learning_rate": 0.001,
+      "loss": 3.0278,
+      "step": 20827
+    },
+    {
+      "epoch": 0.8811236145189948,
+      "grad_norm": 0.14685030281543732,
+      "learning_rate": 0.001,
+      "loss": 1.7467,
+      "step": 20828
+    },
+    {
+      "epoch": 0.8811659192825112,
+      "grad_norm": 0.15578360855579376,
+      "learning_rate": 0.001,
+      "loss": 3.2155,
+      "step": 20829
+    },
+    {
+      "epoch": 0.8812082240460276,
+      "grad_norm": 1.1892778873443604,
+      "learning_rate": 0.001,
+      "loss": 2.3981,
+      "step": 20830
+    },
+    {
+      "epoch": 0.8812505288095439,
+      "grad_norm": 0.13254669308662415,
+      "learning_rate": 0.001,
+      "loss": 2.8435,
+      "step": 20831
+    },
+    {
+      "epoch": 0.8812928335730603,
+      "grad_norm": 0.1270717978477478,
+      "learning_rate": 0.001,
+      "loss": 2.3439,
+      "step": 20832
+    },
+    {
+      "epoch": 0.8813351383365767,
+      "grad_norm": 0.1563170850276947,
+      "learning_rate": 0.001,
+      "loss": 1.8762,
+      "step": 20833
+    },
+    {
+      "epoch": 0.881377443100093,
+      "grad_norm": 0.8175224661827087,
+      "learning_rate": 0.001,
+      "loss": 2.7527,
+      "step": 20834
+    },
+    {
+      "epoch": 0.8814197478636094,
+      "grad_norm": 0.13206033408641815,
+      "learning_rate": 0.001,
+      "loss": 2.0897,
+      "step": 20835
+    },
+    {
+      "epoch": 0.8814620526271258,
+      "grad_norm": 0.2126411646604538,
+      "learning_rate": 0.001,
+      "loss": 3.5215,
+      "step": 20836
+    },
+    {
+      "epoch": 0.8815043573906421,
+      "grad_norm": 0.12830336391925812,
+      "learning_rate": 0.001,
+      "loss": 1.3334,
+      "step": 20837
+    },
+    {
+      "epoch": 0.8815466621541586,
+      "grad_norm": 0.601836621761322,
+      "learning_rate": 0.001,
+      "loss": 1.9787,
+      "step": 20838
+    },
+    {
+      "epoch": 0.881588966917675,
+      "grad_norm": 0.13936357200145721,
+      "learning_rate": 0.001,
+      "loss": 2.0659,
+      "step": 20839
+    },
+    {
+      "epoch": 0.8816312716811913,
+      "grad_norm": 0.1285770684480667,
+      "learning_rate": 0.001,
+      "loss": 2.1898,
+      "step": 20840
+    },
+    {
+      "epoch": 0.8816735764447077,
+      "grad_norm": 0.12381594628095627,
+      "learning_rate": 0.001,
+      "loss": 1.7729,
+      "step": 20841
+    },
+    {
+      "epoch": 0.8817158812082241,
+      "grad_norm": 0.12983018159866333,
+      "learning_rate": 0.001,
+      "loss": 2.5275,
+      "step": 20842
+    },
+    {
+      "epoch": 0.8817581859717404,
+      "grad_norm": 0.1269802451133728,
+      "learning_rate": 0.001,
+      "loss": 2.2177,
+      "step": 20843
+    },
+    {
+      "epoch": 0.8818004907352568,
+      "grad_norm": 0.12933611869812012,
+      "learning_rate": 0.001,
+      "loss": 1.416,
+      "step": 20844
+    },
+    {
+      "epoch": 0.8818427954987732,
+      "grad_norm": 0.5921683311462402,
+      "learning_rate": 0.001,
+      "loss": 2.5021,
+      "step": 20845
+    },
+    {
+      "epoch": 0.8818851002622895,
+      "grad_norm": 0.15616628527641296,
+      "learning_rate": 0.001,
+      "loss": 1.7597,
+      "step": 20846
+    },
+    {
+      "epoch": 0.8819274050258059,
+      "grad_norm": 0.16037492454051971,
+      "learning_rate": 0.001,
+      "loss": 3.31,
+      "step": 20847
+    },
+    {
+      "epoch": 0.8819697097893223,
+      "grad_norm": 0.19823972880840302,
+      "learning_rate": 0.001,
+      "loss": 2.1503,
+      "step": 20848
+    },
+    {
+      "epoch": 0.8820120145528386,
+      "grad_norm": 0.17855338752269745,
+      "learning_rate": 0.001,
+      "loss": 2.6852,
+      "step": 20849
+    },
+    {
+      "epoch": 0.882054319316355,
+      "grad_norm": 0.15427084267139435,
+      "learning_rate": 0.001,
+      "loss": 2.0922,
+      "step": 20850
+    },
+    {
+      "epoch": 0.8820966240798714,
+      "grad_norm": 0.14406783878803253,
+      "learning_rate": 0.001,
+      "loss": 2.3578,
+      "step": 20851
+    },
+    {
+      "epoch": 0.8821389288433877,
+      "grad_norm": 0.16214242577552795,
+      "learning_rate": 0.001,
+      "loss": 1.3277,
+      "step": 20852
+    },
+    {
+      "epoch": 0.8821812336069041,
+      "grad_norm": 0.1684330701828003,
+      "learning_rate": 0.001,
+      "loss": 2.0406,
+      "step": 20853
+    },
+    {
+      "epoch": 0.8822235383704206,
+      "grad_norm": 8.746175765991211,
+      "learning_rate": 0.001,
+      "loss": 2.3627,
+      "step": 20854
+    },
+    {
+      "epoch": 0.8822658431339369,
+      "grad_norm": 0.17695072293281555,
+      "learning_rate": 0.001,
+      "loss": 1.8508,
+      "step": 20855
+    },
+    {
+      "epoch": 0.8823081478974533,
+      "grad_norm": 0.147260844707489,
+      "learning_rate": 0.001,
+      "loss": 1.9841,
+      "step": 20856
+    },
+    {
+      "epoch": 0.8823504526609697,
+      "grad_norm": 0.13849042356014252,
+      "learning_rate": 0.001,
+      "loss": 1.6593,
+      "step": 20857
+    },
+    {
+      "epoch": 0.882392757424486,
+      "grad_norm": 1.9527419805526733,
+      "learning_rate": 0.001,
+      "loss": 2.0942,
+      "step": 20858
+    },
+    {
+      "epoch": 0.8824350621880024,
+      "grad_norm": 0.18281447887420654,
+      "learning_rate": 0.001,
+      "loss": 3.9133,
+      "step": 20859
+    },
+    {
+      "epoch": 0.8824773669515188,
+      "grad_norm": 0.8285083174705505,
+      "learning_rate": 0.001,
+      "loss": 2.7313,
+      "step": 20860
+    },
+    {
+      "epoch": 0.8825196717150351,
+      "grad_norm": 0.163629949092865,
+      "learning_rate": 0.001,
+      "loss": 1.8258,
+      "step": 20861
+    },
+    {
+      "epoch": 0.8825619764785515,
+      "grad_norm": 0.6261329054832458,
+      "learning_rate": 0.001,
+      "loss": 1.9942,
+      "step": 20862
+    },
+    {
+      "epoch": 0.8826042812420679,
+      "grad_norm": 0.11821243911981583,
+      "learning_rate": 0.001,
+      "loss": 2.0009,
+      "step": 20863
+    },
+    {
+      "epoch": 0.8826465860055842,
+      "grad_norm": 0.17022605240345,
+      "learning_rate": 0.001,
+      "loss": 2.0427,
+      "step": 20864
+    },
+    {
+      "epoch": 0.8826888907691006,
+      "grad_norm": 0.28422749042510986,
+      "learning_rate": 0.001,
+      "loss": 2.141,
+      "step": 20865
+    },
+    {
+      "epoch": 0.8827311955326169,
+      "grad_norm": 0.1458466500043869,
+      "learning_rate": 0.001,
+      "loss": 1.9478,
+      "step": 20866
+    },
+    {
+      "epoch": 0.8827735002961333,
+      "grad_norm": 0.1322123408317566,
+      "learning_rate": 0.001,
+      "loss": 1.4996,
+      "step": 20867
+    },
+    {
+      "epoch": 0.8828158050596497,
+      "grad_norm": 0.1259404420852661,
+      "learning_rate": 0.001,
+      "loss": 1.9031,
+      "step": 20868
+    },
+    {
+      "epoch": 0.882858109823166,
+      "grad_norm": 0.23596730828285217,
+      "learning_rate": 0.001,
+      "loss": 1.8853,
+      "step": 20869
+    },
+    {
+      "epoch": 0.8829004145866824,
+      "grad_norm": 0.15984271466732025,
+      "learning_rate": 0.001,
+      "loss": 3.1109,
+      "step": 20870
+    },
+    {
+      "epoch": 0.8829427193501989,
+      "grad_norm": 0.1410568356513977,
+      "learning_rate": 0.001,
+      "loss": 2.5422,
+      "step": 20871
+    },
+    {
+      "epoch": 0.8829850241137152,
+      "grad_norm": 0.15058191120624542,
+      "learning_rate": 0.001,
+      "loss": 2.1368,
+      "step": 20872
+    },
+    {
+      "epoch": 0.8830273288772316,
+      "grad_norm": 0.1391940861940384,
+      "learning_rate": 0.001,
+      "loss": 2.1887,
+      "step": 20873
+    },
+    {
+      "epoch": 0.883069633640748,
+      "grad_norm": 0.1491466760635376,
+      "learning_rate": 0.001,
+      "loss": 1.7719,
+      "step": 20874
+    },
+    {
+      "epoch": 0.8831119384042643,
+      "grad_norm": 0.13309025764465332,
+      "learning_rate": 0.001,
+      "loss": 1.9236,
+      "step": 20875
+    },
+    {
+      "epoch": 0.8831542431677807,
+      "grad_norm": 0.1632324457168579,
+      "learning_rate": 0.001,
+      "loss": 2.0839,
+      "step": 20876
+    },
+    {
+      "epoch": 0.8831965479312971,
+      "grad_norm": 0.26213958859443665,
+      "learning_rate": 0.001,
+      "loss": 2.0513,
+      "step": 20877
+    },
+    {
+      "epoch": 0.8832388526948134,
+      "grad_norm": 0.14815469086170197,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 20878
+    },
+    {
+      "epoch": 0.8832811574583298,
+      "grad_norm": 0.243429496884346,
+      "learning_rate": 0.001,
+      "loss": 3.2333,
+      "step": 20879
+    },
+    {
+      "epoch": 0.8833234622218462,
+      "grad_norm": 0.5001935958862305,
+      "learning_rate": 0.001,
+      "loss": 2.6958,
+      "step": 20880
+    },
+    {
+      "epoch": 0.8833657669853625,
+      "grad_norm": 0.14644937217235565,
+      "learning_rate": 0.001,
+      "loss": 1.6308,
+      "step": 20881
+    },
+    {
+      "epoch": 0.8834080717488789,
+      "grad_norm": 0.1332141011953354,
+      "learning_rate": 0.001,
+      "loss": 1.7561,
+      "step": 20882
+    },
+    {
+      "epoch": 0.8834503765123953,
+      "grad_norm": 0.5288890600204468,
+      "learning_rate": 0.001,
+      "loss": 2.2926,
+      "step": 20883
+    },
+    {
+      "epoch": 0.8834926812759116,
+      "grad_norm": 0.14572833478450775,
+      "learning_rate": 0.001,
+      "loss": 2.7297,
+      "step": 20884
+    },
+    {
+      "epoch": 0.883534986039428,
+      "grad_norm": 0.16958162188529968,
+      "learning_rate": 0.001,
+      "loss": 2.3938,
+      "step": 20885
+    },
+    {
+      "epoch": 0.8835772908029444,
+      "grad_norm": 0.16472996771335602,
+      "learning_rate": 0.001,
+      "loss": 2.0189,
+      "step": 20886
+    },
+    {
+      "epoch": 0.8836195955664607,
+      "grad_norm": 0.14302243292331696,
+      "learning_rate": 0.001,
+      "loss": 1.9094,
+      "step": 20887
+    },
+    {
+      "epoch": 0.8836619003299772,
+      "grad_norm": 0.14594019949436188,
+      "learning_rate": 0.001,
+      "loss": 1.9324,
+      "step": 20888
+    },
+    {
+      "epoch": 0.8837042050934936,
+      "grad_norm": 0.1699601113796234,
+      "learning_rate": 0.001,
+      "loss": 2.5119,
+      "step": 20889
+    },
+    {
+      "epoch": 0.8837465098570099,
+      "grad_norm": 0.16255362331867218,
+      "learning_rate": 0.001,
+      "loss": 3.1678,
+      "step": 20890
+    },
+    {
+      "epoch": 0.8837888146205263,
+      "grad_norm": 0.2458498328924179,
+      "learning_rate": 0.001,
+      "loss": 1.8441,
+      "step": 20891
+    },
+    {
+      "epoch": 0.8838311193840427,
+      "grad_norm": 0.14497283101081848,
+      "learning_rate": 0.001,
+      "loss": 2.5701,
+      "step": 20892
+    },
+    {
+      "epoch": 0.883873424147559,
+      "grad_norm": 2.2221720218658447,
+      "learning_rate": 0.001,
+      "loss": 2.7294,
+      "step": 20893
+    },
+    {
+      "epoch": 0.8839157289110754,
+      "grad_norm": 0.17696496844291687,
+      "learning_rate": 0.001,
+      "loss": 2.1181,
+      "step": 20894
+    },
+    {
+      "epoch": 0.8839580336745918,
+      "grad_norm": 0.1702883541584015,
+      "learning_rate": 0.001,
+      "loss": 2.9156,
+      "step": 20895
+    },
+    {
+      "epoch": 0.8840003384381081,
+      "grad_norm": 0.1452932506799698,
+      "learning_rate": 0.001,
+      "loss": 1.5805,
+      "step": 20896
+    },
+    {
+      "epoch": 0.8840426432016245,
+      "grad_norm": 0.14172394573688507,
+      "learning_rate": 0.001,
+      "loss": 1.5672,
+      "step": 20897
+    },
+    {
+      "epoch": 0.8840849479651409,
+      "grad_norm": 0.1812184453010559,
+      "learning_rate": 0.001,
+      "loss": 2.7763,
+      "step": 20898
+    },
+    {
+      "epoch": 0.8841272527286572,
+      "grad_norm": 0.1486140936613083,
+      "learning_rate": 0.001,
+      "loss": 2.2094,
+      "step": 20899
+    },
+    {
+      "epoch": 0.8841695574921736,
+      "grad_norm": 0.17032857239246368,
+      "learning_rate": 0.001,
+      "loss": 1.8822,
+      "step": 20900
+    },
+    {
+      "epoch": 0.88421186225569,
+      "grad_norm": 0.21682803332805634,
+      "learning_rate": 0.001,
+      "loss": 2.0348,
+      "step": 20901
+    },
+    {
+      "epoch": 0.8842541670192063,
+      "grad_norm": 0.1323508471250534,
+      "learning_rate": 0.001,
+      "loss": 1.682,
+      "step": 20902
+    },
+    {
+      "epoch": 0.8842964717827227,
+      "grad_norm": 0.3486897647380829,
+      "learning_rate": 0.001,
+      "loss": 2.2666,
+      "step": 20903
+    },
+    {
+      "epoch": 0.8843387765462392,
+      "grad_norm": 3.172335386276245,
+      "learning_rate": 0.001,
+      "loss": 2.2502,
+      "step": 20904
+    },
+    {
+      "epoch": 0.8843810813097555,
+      "grad_norm": 0.14836838841438293,
+      "learning_rate": 0.001,
+      "loss": 1.4472,
+      "step": 20905
+    },
+    {
+      "epoch": 0.8844233860732719,
+      "grad_norm": 0.195405974984169,
+      "learning_rate": 0.001,
+      "loss": 1.8995,
+      "step": 20906
+    },
+    {
+      "epoch": 0.8844656908367883,
+      "grad_norm": 0.41760605573654175,
+      "learning_rate": 0.001,
+      "loss": 2.139,
+      "step": 20907
+    },
+    {
+      "epoch": 0.8845079956003046,
+      "grad_norm": 0.16416040062904358,
+      "learning_rate": 0.001,
+      "loss": 2.0343,
+      "step": 20908
+    },
+    {
+      "epoch": 0.884550300363821,
+      "grad_norm": 0.16729791462421417,
+      "learning_rate": 0.001,
+      "loss": 2.0194,
+      "step": 20909
+    },
+    {
+      "epoch": 0.8845926051273373,
+      "grad_norm": 0.15572674572467804,
+      "learning_rate": 0.001,
+      "loss": 1.7202,
+      "step": 20910
+    },
+    {
+      "epoch": 0.8846349098908537,
+      "grad_norm": 36.179466247558594,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 20911
+    },
+    {
+      "epoch": 0.8846772146543701,
+      "grad_norm": 0.15077073872089386,
+      "learning_rate": 0.001,
+      "loss": 1.6993,
+      "step": 20912
+    },
+    {
+      "epoch": 0.8847195194178864,
+      "grad_norm": 0.1646421253681183,
+      "learning_rate": 0.001,
+      "loss": 2.058,
+      "step": 20913
+    },
+    {
+      "epoch": 0.8847618241814028,
+      "grad_norm": 0.2249441295862198,
+      "learning_rate": 0.001,
+      "loss": 2.8581,
+      "step": 20914
+    },
+    {
+      "epoch": 0.8848041289449192,
+      "grad_norm": 0.19129936397075653,
+      "learning_rate": 0.001,
+      "loss": 2.1256,
+      "step": 20915
+    },
+    {
+      "epoch": 0.8848464337084355,
+      "grad_norm": 0.16032202541828156,
+      "learning_rate": 0.001,
+      "loss": 1.7309,
+      "step": 20916
+    },
+    {
+      "epoch": 0.8848887384719519,
+      "grad_norm": 0.21041136980056763,
+      "learning_rate": 0.001,
+      "loss": 3.1444,
+      "step": 20917
+    },
+    {
+      "epoch": 0.8849310432354683,
+      "grad_norm": 0.1504105031490326,
+      "learning_rate": 0.001,
+      "loss": 1.4867,
+      "step": 20918
+    },
+    {
+      "epoch": 0.8849733479989846,
+      "grad_norm": 0.16670987010002136,
+      "learning_rate": 0.001,
+      "loss": 1.6321,
+      "step": 20919
+    },
+    {
+      "epoch": 0.885015652762501,
+      "grad_norm": 0.22837378084659576,
+      "learning_rate": 0.001,
+      "loss": 2.2452,
+      "step": 20920
+    },
+    {
+      "epoch": 0.8850579575260175,
+      "grad_norm": 0.15080545842647552,
+      "learning_rate": 0.001,
+      "loss": 2.2631,
+      "step": 20921
+    },
+    {
+      "epoch": 0.8851002622895338,
+      "grad_norm": 0.16845811903476715,
+      "learning_rate": 0.001,
+      "loss": 1.9818,
+      "step": 20922
+    },
+    {
+      "epoch": 0.8851425670530502,
+      "grad_norm": 0.6838306784629822,
+      "learning_rate": 0.001,
+      "loss": 1.6603,
+      "step": 20923
+    },
+    {
+      "epoch": 0.8851848718165666,
+      "grad_norm": 0.12884469330310822,
+      "learning_rate": 0.001,
+      "loss": 1.7758,
+      "step": 20924
+    },
+    {
+      "epoch": 0.8852271765800829,
+      "grad_norm": 0.7844867706298828,
+      "learning_rate": 0.001,
+      "loss": 2.031,
+      "step": 20925
+    },
+    {
+      "epoch": 0.8852694813435993,
+      "grad_norm": 0.1945054531097412,
+      "learning_rate": 0.001,
+      "loss": 3.3753,
+      "step": 20926
+    },
+    {
+      "epoch": 0.8853117861071157,
+      "grad_norm": 30.440494537353516,
+      "learning_rate": 0.001,
+      "loss": 2.9578,
+      "step": 20927
+    },
+    {
+      "epoch": 0.885354090870632,
+      "grad_norm": 0.13637934625148773,
+      "learning_rate": 0.001,
+      "loss": 1.677,
+      "step": 20928
+    },
+    {
+      "epoch": 0.8853963956341484,
+      "grad_norm": 0.7692943811416626,
+      "learning_rate": 0.001,
+      "loss": 2.1675,
+      "step": 20929
+    },
+    {
+      "epoch": 0.8854387003976648,
+      "grad_norm": 0.9168503880500793,
+      "learning_rate": 0.001,
+      "loss": 2.1578,
+      "step": 20930
+    },
+    {
+      "epoch": 0.8854810051611811,
+      "grad_norm": 0.1996372491121292,
+      "learning_rate": 0.001,
+      "loss": 2.6837,
+      "step": 20931
+    },
+    {
+      "epoch": 0.8855233099246975,
+      "grad_norm": 0.15608003735542297,
+      "learning_rate": 0.001,
+      "loss": 3.4255,
+      "step": 20932
+    },
+    {
+      "epoch": 0.8855656146882139,
+      "grad_norm": 0.7648287415504456,
+      "learning_rate": 0.001,
+      "loss": 1.6965,
+      "step": 20933
+    },
+    {
+      "epoch": 0.8856079194517302,
+      "grad_norm": 8.121953964233398,
+      "learning_rate": 0.001,
+      "loss": 2.2537,
+      "step": 20934
+    },
+    {
+      "epoch": 0.8856502242152466,
+      "grad_norm": 2.677286148071289,
+      "learning_rate": 0.001,
+      "loss": 3.2583,
+      "step": 20935
+    },
+    {
+      "epoch": 0.885692528978763,
+      "grad_norm": 0.17523032426834106,
+      "learning_rate": 0.001,
+      "loss": 2.1655,
+      "step": 20936
+    },
+    {
+      "epoch": 0.8857348337422793,
+      "grad_norm": 0.16115929186344147,
+      "learning_rate": 0.001,
+      "loss": 1.5949,
+      "step": 20937
+    },
+    {
+      "epoch": 0.8857771385057958,
+      "grad_norm": 0.18050625920295715,
+      "learning_rate": 0.001,
+      "loss": 2.2334,
+      "step": 20938
+    },
+    {
+      "epoch": 0.8858194432693122,
+      "grad_norm": 0.19290222227573395,
+      "learning_rate": 0.001,
+      "loss": 2.4616,
+      "step": 20939
+    },
+    {
+      "epoch": 0.8858617480328285,
+      "grad_norm": 0.15666300058364868,
+      "learning_rate": 0.001,
+      "loss": 2.5885,
+      "step": 20940
+    },
+    {
+      "epoch": 0.8859040527963449,
+      "grad_norm": 0.21814775466918945,
+      "learning_rate": 0.001,
+      "loss": 1.8548,
+      "step": 20941
+    },
+    {
+      "epoch": 0.8859463575598613,
+      "grad_norm": 0.18262353539466858,
+      "learning_rate": 0.001,
+      "loss": 1.8053,
+      "step": 20942
+    },
+    {
+      "epoch": 0.8859886623233776,
+      "grad_norm": 0.4183075726032257,
+      "learning_rate": 0.001,
+      "loss": 2.103,
+      "step": 20943
+    },
+    {
+      "epoch": 0.886030967086894,
+      "grad_norm": 0.18610745668411255,
+      "learning_rate": 0.001,
+      "loss": 2.7605,
+      "step": 20944
+    },
+    {
+      "epoch": 0.8860732718504104,
+      "grad_norm": 10.397679328918457,
+      "learning_rate": 0.001,
+      "loss": 2.124,
+      "step": 20945
+    },
+    {
+      "epoch": 0.8861155766139267,
+      "grad_norm": 0.14724650979042053,
+      "learning_rate": 0.001,
+      "loss": 2.377,
+      "step": 20946
+    },
+    {
+      "epoch": 0.8861578813774431,
+      "grad_norm": 0.4719243049621582,
+      "learning_rate": 0.001,
+      "loss": 1.979,
+      "step": 20947
+    },
+    {
+      "epoch": 0.8862001861409595,
+      "grad_norm": 0.21508583426475525,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 20948
+    },
+    {
+      "epoch": 0.8862424909044758,
+      "grad_norm": 0.16444562375545502,
+      "learning_rate": 0.001,
+      "loss": 1.8839,
+      "step": 20949
+    },
+    {
+      "epoch": 0.8862847956679922,
+      "grad_norm": 0.22835667431354523,
+      "learning_rate": 0.001,
+      "loss": 2.5681,
+      "step": 20950
+    },
+    {
+      "epoch": 0.8863271004315086,
+      "grad_norm": 0.17170853912830353,
+      "learning_rate": 0.001,
+      "loss": 2.0785,
+      "step": 20951
+    },
+    {
+      "epoch": 0.8863694051950249,
+      "grad_norm": 0.135117307305336,
+      "learning_rate": 0.001,
+      "loss": 2.4973,
+      "step": 20952
+    },
+    {
+      "epoch": 0.8864117099585413,
+      "grad_norm": 1.2881138324737549,
+      "learning_rate": 0.001,
+      "loss": 1.8363,
+      "step": 20953
+    },
+    {
+      "epoch": 0.8864540147220576,
+      "grad_norm": 0.1489884853363037,
+      "learning_rate": 0.001,
+      "loss": 1.9252,
+      "step": 20954
+    },
+    {
+      "epoch": 0.8864963194855741,
+      "grad_norm": 0.14874590933322906,
+      "learning_rate": 0.001,
+      "loss": 1.4085,
+      "step": 20955
+    },
+    {
+      "epoch": 0.8865386242490905,
+      "grad_norm": 0.13370899856090546,
+      "learning_rate": 0.001,
+      "loss": 2.2094,
+      "step": 20956
+    },
+    {
+      "epoch": 0.8865809290126068,
+      "grad_norm": 4.428647041320801,
+      "learning_rate": 0.001,
+      "loss": 2.2555,
+      "step": 20957
+    },
+    {
+      "epoch": 0.8866232337761232,
+      "grad_norm": 0.12054305523633957,
+      "learning_rate": 0.001,
+      "loss": 2.2724,
+      "step": 20958
+    },
+    {
+      "epoch": 0.8866655385396396,
+      "grad_norm": 0.19234895706176758,
+      "learning_rate": 0.001,
+      "loss": 2.7066,
+      "step": 20959
+    },
+    {
+      "epoch": 0.8867078433031559,
+      "grad_norm": 0.18497143685817719,
+      "learning_rate": 0.001,
+      "loss": 2.3736,
+      "step": 20960
+    },
+    {
+      "epoch": 0.8867501480666723,
+      "grad_norm": 0.2834151089191437,
+      "learning_rate": 0.001,
+      "loss": 2.2192,
+      "step": 20961
+    },
+    {
+      "epoch": 0.8867924528301887,
+      "grad_norm": 0.18010100722312927,
+      "learning_rate": 0.001,
+      "loss": 2.4076,
+      "step": 20962
+    },
+    {
+      "epoch": 0.886834757593705,
+      "grad_norm": 0.3335997462272644,
+      "learning_rate": 0.001,
+      "loss": 3.2665,
+      "step": 20963
+    },
+    {
+      "epoch": 0.8868770623572214,
+      "grad_norm": 0.15331315994262695,
+      "learning_rate": 0.001,
+      "loss": 2.5892,
+      "step": 20964
+    },
+    {
+      "epoch": 0.8869193671207378,
+      "grad_norm": 0.16281548142433167,
+      "learning_rate": 0.001,
+      "loss": 2.1734,
+      "step": 20965
+    },
+    {
+      "epoch": 0.8869616718842541,
+      "grad_norm": 0.17195676267147064,
+      "learning_rate": 0.001,
+      "loss": 2.339,
+      "step": 20966
+    },
+    {
+      "epoch": 0.8870039766477705,
+      "grad_norm": 0.13720469176769257,
+      "learning_rate": 0.001,
+      "loss": 1.9478,
+      "step": 20967
+    },
+    {
+      "epoch": 0.8870462814112869,
+      "grad_norm": 0.17597544193267822,
+      "learning_rate": 0.001,
+      "loss": 1.9665,
+      "step": 20968
+    },
+    {
+      "epoch": 0.8870885861748032,
+      "grad_norm": 0.1732187271118164,
+      "learning_rate": 0.001,
+      "loss": 1.9538,
+      "step": 20969
+    },
+    {
+      "epoch": 0.8871308909383196,
+      "grad_norm": 0.14298303425312042,
+      "learning_rate": 0.001,
+      "loss": 2.8327,
+      "step": 20970
+    },
+    {
+      "epoch": 0.8871731957018361,
+      "grad_norm": 0.15061107277870178,
+      "learning_rate": 0.001,
+      "loss": 2.009,
+      "step": 20971
+    },
+    {
+      "epoch": 0.8872155004653524,
+      "grad_norm": 0.14526939392089844,
+      "learning_rate": 0.001,
+      "loss": 1.8338,
+      "step": 20972
+    },
+    {
+      "epoch": 0.8872578052288688,
+      "grad_norm": 0.1140129417181015,
+      "learning_rate": 0.001,
+      "loss": 1.2921,
+      "step": 20973
+    },
+    {
+      "epoch": 0.8873001099923852,
+      "grad_norm": 9.253741264343262,
+      "learning_rate": 0.001,
+      "loss": 2.1857,
+      "step": 20974
+    },
+    {
+      "epoch": 0.8873424147559015,
+      "grad_norm": 0.12813687324523926,
+      "learning_rate": 0.001,
+      "loss": 1.8503,
+      "step": 20975
+    },
+    {
+      "epoch": 0.8873847195194179,
+      "grad_norm": 0.1430562436580658,
+      "learning_rate": 0.001,
+      "loss": 1.9686,
+      "step": 20976
+    },
+    {
+      "epoch": 0.8874270242829343,
+      "grad_norm": 0.1441165953874588,
+      "learning_rate": 0.001,
+      "loss": 2.0366,
+      "step": 20977
+    },
+    {
+      "epoch": 0.8874693290464506,
+      "grad_norm": 0.6266450881958008,
+      "learning_rate": 0.001,
+      "loss": 1.4746,
+      "step": 20978
+    },
+    {
+      "epoch": 0.887511633809967,
+      "grad_norm": 0.13469688594341278,
+      "learning_rate": 0.001,
+      "loss": 2.3698,
+      "step": 20979
+    },
+    {
+      "epoch": 0.8875539385734834,
+      "grad_norm": 0.13266874849796295,
+      "learning_rate": 0.001,
+      "loss": 1.6635,
+      "step": 20980
+    },
+    {
+      "epoch": 0.8875962433369997,
+      "grad_norm": 25.529739379882812,
+      "learning_rate": 0.001,
+      "loss": 2.0676,
+      "step": 20981
+    },
+    {
+      "epoch": 0.8876385481005161,
+      "grad_norm": 0.2457568198442459,
+      "learning_rate": 0.001,
+      "loss": 3.1092,
+      "step": 20982
+    },
+    {
+      "epoch": 0.8876808528640325,
+      "grad_norm": 0.15447290241718292,
+      "learning_rate": 0.001,
+      "loss": 1.8904,
+      "step": 20983
+    },
+    {
+      "epoch": 0.8877231576275488,
+      "grad_norm": 0.280048131942749,
+      "learning_rate": 0.001,
+      "loss": 2.263,
+      "step": 20984
+    },
+    {
+      "epoch": 0.8877654623910652,
+      "grad_norm": 0.11680661141872406,
+      "learning_rate": 0.001,
+      "loss": 1.9202,
+      "step": 20985
+    },
+    {
+      "epoch": 0.8878077671545817,
+      "grad_norm": 16.68817138671875,
+      "learning_rate": 0.001,
+      "loss": 1.7731,
+      "step": 20986
+    },
+    {
+      "epoch": 0.887850071918098,
+      "grad_norm": 0.17112363874912262,
+      "learning_rate": 0.001,
+      "loss": 3.0417,
+      "step": 20987
+    },
+    {
+      "epoch": 0.8878923766816144,
+      "grad_norm": 0.19305647909641266,
+      "learning_rate": 0.001,
+      "loss": 1.4348,
+      "step": 20988
+    },
+    {
+      "epoch": 0.8879346814451308,
+      "grad_norm": 0.162348210811615,
+      "learning_rate": 0.001,
+      "loss": 1.9347,
+      "step": 20989
+    },
+    {
+      "epoch": 0.8879769862086471,
+      "grad_norm": 0.14823856949806213,
+      "learning_rate": 0.001,
+      "loss": 1.9564,
+      "step": 20990
+    },
+    {
+      "epoch": 0.8880192909721635,
+      "grad_norm": 29.342689514160156,
+      "learning_rate": 0.001,
+      "loss": 2.334,
+      "step": 20991
+    },
+    {
+      "epoch": 0.8880615957356799,
+      "grad_norm": 0.7693207859992981,
+      "learning_rate": 0.001,
+      "loss": 1.766,
+      "step": 20992
+    },
+    {
+      "epoch": 0.8881039004991962,
+      "grad_norm": 1.0036860704421997,
+      "learning_rate": 0.001,
+      "loss": 2.2152,
+      "step": 20993
+    },
+    {
+      "epoch": 0.8881462052627126,
+      "grad_norm": 3.9714372158050537,
+      "learning_rate": 0.001,
+      "loss": 2.1756,
+      "step": 20994
+    },
+    {
+      "epoch": 0.888188510026229,
+      "grad_norm": 0.21561984717845917,
+      "learning_rate": 0.001,
+      "loss": 1.7397,
+      "step": 20995
+    },
+    {
+      "epoch": 0.8882308147897453,
+      "grad_norm": 0.1669800579547882,
+      "learning_rate": 0.001,
+      "loss": 2.7163,
+      "step": 20996
+    },
+    {
+      "epoch": 0.8882731195532617,
+      "grad_norm": 0.19429895281791687,
+      "learning_rate": 0.001,
+      "loss": 2.321,
+      "step": 20997
+    },
+    {
+      "epoch": 0.8883154243167781,
+      "grad_norm": 0.21047398447990417,
+      "learning_rate": 0.001,
+      "loss": 1.6531,
+      "step": 20998
+    },
+    {
+      "epoch": 0.8883577290802944,
+      "grad_norm": 0.20937636494636536,
+      "learning_rate": 0.001,
+      "loss": 2.1078,
+      "step": 20999
+    },
+    {
+      "epoch": 0.8884000338438108,
+      "grad_norm": 0.4300979673862457,
+      "learning_rate": 0.001,
+      "loss": 2.4772,
+      "step": 21000
+    },
+    {
+      "epoch": 0.8884423386073271,
+      "grad_norm": 0.18528619408607483,
+      "learning_rate": 0.001,
+      "loss": 1.5542,
+      "step": 21001
+    },
+    {
+      "epoch": 0.8884846433708435,
+      "grad_norm": 0.989427387714386,
+      "learning_rate": 0.001,
+      "loss": 1.9901,
+      "step": 21002
+    },
+    {
+      "epoch": 0.88852694813436,
+      "grad_norm": 0.19814398884773254,
+      "learning_rate": 0.001,
+      "loss": 2.3798,
+      "step": 21003
+    },
+    {
+      "epoch": 0.8885692528978762,
+      "grad_norm": 0.16753339767456055,
+      "learning_rate": 0.001,
+      "loss": 2.8676,
+      "step": 21004
+    },
+    {
+      "epoch": 0.8886115576613927,
+      "grad_norm": 0.1605309545993805,
+      "learning_rate": 0.001,
+      "loss": 1.5876,
+      "step": 21005
+    },
+    {
+      "epoch": 0.8886538624249091,
+      "grad_norm": 1.0436333417892456,
+      "learning_rate": 0.001,
+      "loss": 3.0352,
+      "step": 21006
+    },
+    {
+      "epoch": 0.8886961671884254,
+      "grad_norm": 0.5114812850952148,
+      "learning_rate": 0.001,
+      "loss": 1.9538,
+      "step": 21007
+    },
+    {
+      "epoch": 0.8887384719519418,
+      "grad_norm": 0.16753728687763214,
+      "learning_rate": 0.001,
+      "loss": 1.6974,
+      "step": 21008
+    },
+    {
+      "epoch": 0.8887807767154582,
+      "grad_norm": 0.17720456421375275,
+      "learning_rate": 0.001,
+      "loss": 2.0584,
+      "step": 21009
+    },
+    {
+      "epoch": 0.8888230814789745,
+      "grad_norm": 0.23170483112335205,
+      "learning_rate": 0.001,
+      "loss": 3.2882,
+      "step": 21010
+    },
+    {
+      "epoch": 0.8888653862424909,
+      "grad_norm": 0.41039520502090454,
+      "learning_rate": 0.001,
+      "loss": 2.4214,
+      "step": 21011
+    },
+    {
+      "epoch": 0.8889076910060073,
+      "grad_norm": 0.1671202927827835,
+      "learning_rate": 0.001,
+      "loss": 1.606,
+      "step": 21012
+    },
+    {
+      "epoch": 0.8889499957695236,
+      "grad_norm": 0.2298729419708252,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 21013
+    },
+    {
+      "epoch": 0.88899230053304,
+      "grad_norm": 0.1773902326822281,
+      "learning_rate": 0.001,
+      "loss": 2.9743,
+      "step": 21014
+    },
+    {
+      "epoch": 0.8890346052965564,
+      "grad_norm": 0.15954598784446716,
+      "learning_rate": 0.001,
+      "loss": 1.7877,
+      "step": 21015
+    },
+    {
+      "epoch": 0.8890769100600727,
+      "grad_norm": 0.1797361969947815,
+      "learning_rate": 0.001,
+      "loss": 2.4128,
+      "step": 21016
+    },
+    {
+      "epoch": 0.8891192148235891,
+      "grad_norm": 0.1873590052127838,
+      "learning_rate": 0.001,
+      "loss": 2.0114,
+      "step": 21017
+    },
+    {
+      "epoch": 0.8891615195871055,
+      "grad_norm": 0.20286858081817627,
+      "learning_rate": 0.001,
+      "loss": 1.7583,
+      "step": 21018
+    },
+    {
+      "epoch": 0.8892038243506218,
+      "grad_norm": 0.17168866097927094,
+      "learning_rate": 0.001,
+      "loss": 1.9782,
+      "step": 21019
+    },
+    {
+      "epoch": 0.8892461291141383,
+      "grad_norm": 0.1769993156194687,
+      "learning_rate": 0.001,
+      "loss": 1.922,
+      "step": 21020
+    },
+    {
+      "epoch": 0.8892884338776547,
+      "grad_norm": 0.1596214920282364,
+      "learning_rate": 0.001,
+      "loss": 2.1562,
+      "step": 21021
+    },
+    {
+      "epoch": 0.889330738641171,
+      "grad_norm": 1.326114535331726,
+      "learning_rate": 0.001,
+      "loss": 1.683,
+      "step": 21022
+    },
+    {
+      "epoch": 0.8893730434046874,
+      "grad_norm": 0.19646044075489044,
+      "learning_rate": 0.001,
+      "loss": 2.3388,
+      "step": 21023
+    },
+    {
+      "epoch": 0.8894153481682038,
+      "grad_norm": 0.173504039645195,
+      "learning_rate": 0.001,
+      "loss": 1.6195,
+      "step": 21024
+    },
+    {
+      "epoch": 0.8894576529317201,
+      "grad_norm": 0.145315483212471,
+      "learning_rate": 0.001,
+      "loss": 2.3937,
+      "step": 21025
+    },
+    {
+      "epoch": 0.8894999576952365,
+      "grad_norm": 0.3077605366706848,
+      "learning_rate": 0.001,
+      "loss": 3.0489,
+      "step": 21026
+    },
+    {
+      "epoch": 0.8895422624587529,
+      "grad_norm": 0.15277047455310822,
+      "learning_rate": 0.001,
+      "loss": 1.9625,
+      "step": 21027
+    },
+    {
+      "epoch": 0.8895845672222692,
+      "grad_norm": 7.509362697601318,
+      "learning_rate": 0.001,
+      "loss": 3.2375,
+      "step": 21028
+    },
+    {
+      "epoch": 0.8896268719857856,
+      "grad_norm": 0.23799103498458862,
+      "learning_rate": 0.001,
+      "loss": 1.8929,
+      "step": 21029
+    },
+    {
+      "epoch": 0.889669176749302,
+      "grad_norm": 0.16391988098621368,
+      "learning_rate": 0.001,
+      "loss": 2.3133,
+      "step": 21030
+    },
+    {
+      "epoch": 0.8897114815128183,
+      "grad_norm": 0.13483858108520508,
+      "learning_rate": 0.001,
+      "loss": 1.8435,
+      "step": 21031
+    },
+    {
+      "epoch": 0.8897537862763347,
+      "grad_norm": 0.1675466150045395,
+      "learning_rate": 0.001,
+      "loss": 2.2034,
+      "step": 21032
+    },
+    {
+      "epoch": 0.8897960910398511,
+      "grad_norm": 0.5026352405548096,
+      "learning_rate": 0.001,
+      "loss": 3.6788,
+      "step": 21033
+    },
+    {
+      "epoch": 0.8898383958033674,
+      "grad_norm": 0.1782296597957611,
+      "learning_rate": 0.001,
+      "loss": 1.971,
+      "step": 21034
+    },
+    {
+      "epoch": 0.8898807005668838,
+      "grad_norm": 0.1635374128818512,
+      "learning_rate": 0.001,
+      "loss": 1.5724,
+      "step": 21035
+    },
+    {
+      "epoch": 0.8899230053304003,
+      "grad_norm": 0.17833854258060455,
+      "learning_rate": 0.001,
+      "loss": 2.2641,
+      "step": 21036
+    },
+    {
+      "epoch": 0.8899653100939166,
+      "grad_norm": 0.15048396587371826,
+      "learning_rate": 0.001,
+      "loss": 1.8958,
+      "step": 21037
+    },
+    {
+      "epoch": 0.890007614857433,
+      "grad_norm": 0.4295186996459961,
+      "learning_rate": 0.001,
+      "loss": 3.0759,
+      "step": 21038
+    },
+    {
+      "epoch": 0.8900499196209494,
+      "grad_norm": 0.26509931683540344,
+      "learning_rate": 0.001,
+      "loss": 3.175,
+      "step": 21039
+    },
+    {
+      "epoch": 0.8900922243844657,
+      "grad_norm": 0.21517863869667053,
+      "learning_rate": 0.001,
+      "loss": 2.5894,
+      "step": 21040
+    },
+    {
+      "epoch": 0.8901345291479821,
+      "grad_norm": 0.15803544223308563,
+      "learning_rate": 0.001,
+      "loss": 1.7448,
+      "step": 21041
+    },
+    {
+      "epoch": 0.8901768339114985,
+      "grad_norm": 0.6315363645553589,
+      "learning_rate": 0.001,
+      "loss": 2.1491,
+      "step": 21042
+    },
+    {
+      "epoch": 0.8902191386750148,
+      "grad_norm": 0.1378752887248993,
+      "learning_rate": 0.001,
+      "loss": 1.8168,
+      "step": 21043
+    },
+    {
+      "epoch": 0.8902614434385312,
+      "grad_norm": 0.16324613988399506,
+      "learning_rate": 0.001,
+      "loss": 3.0616,
+      "step": 21044
+    },
+    {
+      "epoch": 0.8903037482020475,
+      "grad_norm": 0.16827210783958435,
+      "learning_rate": 0.001,
+      "loss": 1.9812,
+      "step": 21045
+    },
+    {
+      "epoch": 0.8903460529655639,
+      "grad_norm": 0.16511641442775726,
+      "learning_rate": 0.001,
+      "loss": 2.405,
+      "step": 21046
+    },
+    {
+      "epoch": 0.8903883577290803,
+      "grad_norm": 0.14627037942409515,
+      "learning_rate": 0.001,
+      "loss": 1.4163,
+      "step": 21047
+    },
+    {
+      "epoch": 0.8904306624925966,
+      "grad_norm": 8.328338623046875,
+      "learning_rate": 0.001,
+      "loss": 2.2162,
+      "step": 21048
+    },
+    {
+      "epoch": 0.890472967256113,
+      "grad_norm": 0.2969225347042084,
+      "learning_rate": 0.001,
+      "loss": 1.7355,
+      "step": 21049
+    },
+    {
+      "epoch": 0.8905152720196294,
+      "grad_norm": 0.17002110183238983,
+      "learning_rate": 0.001,
+      "loss": 1.5175,
+      "step": 21050
+    },
+    {
+      "epoch": 0.8905575767831457,
+      "grad_norm": 4.17087459564209,
+      "learning_rate": 0.001,
+      "loss": 1.9892,
+      "step": 21051
+    },
+    {
+      "epoch": 0.8905998815466621,
+      "grad_norm": 0.1925853043794632,
+      "learning_rate": 0.001,
+      "loss": 1.745,
+      "step": 21052
+    },
+    {
+      "epoch": 0.8906421863101786,
+      "grad_norm": 0.19640986621379852,
+      "learning_rate": 0.001,
+      "loss": 2.2911,
+      "step": 21053
+    },
+    {
+      "epoch": 0.8906844910736949,
+      "grad_norm": 0.3781888484954834,
+      "learning_rate": 0.001,
+      "loss": 1.5467,
+      "step": 21054
+    },
+    {
+      "epoch": 0.8907267958372113,
+      "grad_norm": 0.20206791162490845,
+      "learning_rate": 0.001,
+      "loss": 1.8796,
+      "step": 21055
+    },
+    {
+      "epoch": 0.8907691006007277,
+      "grad_norm": 0.13937760889530182,
+      "learning_rate": 0.001,
+      "loss": 2.0243,
+      "step": 21056
+    },
+    {
+      "epoch": 0.890811405364244,
+      "grad_norm": 0.2623637318611145,
+      "learning_rate": 0.001,
+      "loss": 1.7412,
+      "step": 21057
+    },
+    {
+      "epoch": 0.8908537101277604,
+      "grad_norm": 0.21680723130702972,
+      "learning_rate": 0.001,
+      "loss": 2.2334,
+      "step": 21058
+    },
+    {
+      "epoch": 0.8908960148912768,
+      "grad_norm": 0.18895690143108368,
+      "learning_rate": 0.001,
+      "loss": 1.9865,
+      "step": 21059
+    },
+    {
+      "epoch": 0.8909383196547931,
+      "grad_norm": 0.1650429368019104,
+      "learning_rate": 0.001,
+      "loss": 2.9103,
+      "step": 21060
+    },
+    {
+      "epoch": 0.8909806244183095,
+      "grad_norm": 0.32935771346092224,
+      "learning_rate": 0.001,
+      "loss": 2.0757,
+      "step": 21061
+    },
+    {
+      "epoch": 0.8910229291818259,
+      "grad_norm": 0.14048244059085846,
+      "learning_rate": 0.001,
+      "loss": 1.6505,
+      "step": 21062
+    },
+    {
+      "epoch": 0.8910652339453422,
+      "grad_norm": 0.15878313779830933,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 21063
+    },
+    {
+      "epoch": 0.8911075387088586,
+      "grad_norm": 0.17732056975364685,
+      "learning_rate": 0.001,
+      "loss": 1.7493,
+      "step": 21064
+    },
+    {
+      "epoch": 0.891149843472375,
+      "grad_norm": 0.15534935891628265,
+      "learning_rate": 0.001,
+      "loss": 2.3862,
+      "step": 21065
+    },
+    {
+      "epoch": 0.8911921482358913,
+      "grad_norm": 0.5262129902839661,
+      "learning_rate": 0.001,
+      "loss": 1.3208,
+      "step": 21066
+    },
+    {
+      "epoch": 0.8912344529994077,
+      "grad_norm": 0.15670378506183624,
+      "learning_rate": 0.001,
+      "loss": 2.1793,
+      "step": 21067
+    },
+    {
+      "epoch": 0.8912767577629241,
+      "grad_norm": 1.0509247779846191,
+      "learning_rate": 0.001,
+      "loss": 1.7555,
+      "step": 21068
+    },
+    {
+      "epoch": 0.8913190625264404,
+      "grad_norm": 0.16009117662906647,
+      "learning_rate": 0.001,
+      "loss": 1.5403,
+      "step": 21069
+    },
+    {
+      "epoch": 0.8913613672899569,
+      "grad_norm": 0.19776473939418793,
+      "learning_rate": 0.001,
+      "loss": 1.93,
+      "step": 21070
+    },
+    {
+      "epoch": 0.8914036720534733,
+      "grad_norm": 0.12899529933929443,
+      "learning_rate": 0.001,
+      "loss": 1.7009,
+      "step": 21071
+    },
+    {
+      "epoch": 0.8914459768169896,
+      "grad_norm": 0.49440255761146545,
+      "learning_rate": 0.001,
+      "loss": 2.0824,
+      "step": 21072
+    },
+    {
+      "epoch": 0.891488281580506,
+      "grad_norm": 0.1522773802280426,
+      "learning_rate": 0.001,
+      "loss": 2.4829,
+      "step": 21073
+    },
+    {
+      "epoch": 0.8915305863440224,
+      "grad_norm": 0.16316981613636017,
+      "learning_rate": 0.001,
+      "loss": 1.7877,
+      "step": 21074
+    },
+    {
+      "epoch": 0.8915728911075387,
+      "grad_norm": 0.18345019221305847,
+      "learning_rate": 0.001,
+      "loss": 1.7857,
+      "step": 21075
+    },
+    {
+      "epoch": 0.8916151958710551,
+      "grad_norm": 0.19031278789043427,
+      "learning_rate": 0.001,
+      "loss": 2.7861,
+      "step": 21076
+    },
+    {
+      "epoch": 0.8916575006345715,
+      "grad_norm": 0.14421750605106354,
+      "learning_rate": 0.001,
+      "loss": 2.4813,
+      "step": 21077
+    },
+    {
+      "epoch": 0.8916998053980878,
+      "grad_norm": 0.13803726434707642,
+      "learning_rate": 0.001,
+      "loss": 1.7713,
+      "step": 21078
+    },
+    {
+      "epoch": 0.8917421101616042,
+      "grad_norm": 0.16265204548835754,
+      "learning_rate": 0.001,
+      "loss": 1.8057,
+      "step": 21079
+    },
+    {
+      "epoch": 0.8917844149251206,
+      "grad_norm": 0.17232954502105713,
+      "learning_rate": 0.001,
+      "loss": 1.8699,
+      "step": 21080
+    },
+    {
+      "epoch": 0.8918267196886369,
+      "grad_norm": 0.14032989740371704,
+      "learning_rate": 0.001,
+      "loss": 1.632,
+      "step": 21081
+    },
+    {
+      "epoch": 0.8918690244521533,
+      "grad_norm": 0.16456128656864166,
+      "learning_rate": 0.001,
+      "loss": 2.1099,
+      "step": 21082
+    },
+    {
+      "epoch": 0.8919113292156697,
+      "grad_norm": 0.6810503005981445,
+      "learning_rate": 0.001,
+      "loss": 2.6714,
+      "step": 21083
+    },
+    {
+      "epoch": 0.891953633979186,
+      "grad_norm": 0.12685897946357727,
+      "learning_rate": 0.001,
+      "loss": 1.3731,
+      "step": 21084
+    },
+    {
+      "epoch": 0.8919959387427024,
+      "grad_norm": 0.17308743298053741,
+      "learning_rate": 0.001,
+      "loss": 2.6058,
+      "step": 21085
+    },
+    {
+      "epoch": 0.8920382435062189,
+      "grad_norm": 0.17059041559696198,
+      "learning_rate": 0.001,
+      "loss": 2.7914,
+      "step": 21086
+    },
+    {
+      "epoch": 0.8920805482697352,
+      "grad_norm": 0.14783084392547607,
+      "learning_rate": 0.001,
+      "loss": 1.8769,
+      "step": 21087
+    },
+    {
+      "epoch": 0.8921228530332516,
+      "grad_norm": 0.1687219738960266,
+      "learning_rate": 0.001,
+      "loss": 2.3727,
+      "step": 21088
+    },
+    {
+      "epoch": 0.892165157796768,
+      "grad_norm": 0.1436719000339508,
+      "learning_rate": 0.001,
+      "loss": 1.8052,
+      "step": 21089
+    },
+    {
+      "epoch": 0.8922074625602843,
+      "grad_norm": 0.18090786039829254,
+      "learning_rate": 0.001,
+      "loss": 1.6288,
+      "step": 21090
+    },
+    {
+      "epoch": 0.8922497673238007,
+      "grad_norm": 0.2179936170578003,
+      "learning_rate": 0.001,
+      "loss": 1.6717,
+      "step": 21091
+    },
+    {
+      "epoch": 0.892292072087317,
+      "grad_norm": 0.36641108989715576,
+      "learning_rate": 0.001,
+      "loss": 2.6457,
+      "step": 21092
+    },
+    {
+      "epoch": 0.8923343768508334,
+      "grad_norm": 0.13992157578468323,
+      "learning_rate": 0.001,
+      "loss": 3.2825,
+      "step": 21093
+    },
+    {
+      "epoch": 0.8923766816143498,
+      "grad_norm": 2.9790940284729004,
+      "learning_rate": 0.001,
+      "loss": 2.9316,
+      "step": 21094
+    },
+    {
+      "epoch": 0.8924189863778661,
+      "grad_norm": 0.1301809698343277,
+      "learning_rate": 0.001,
+      "loss": 1.6709,
+      "step": 21095
+    },
+    {
+      "epoch": 0.8924612911413825,
+      "grad_norm": 0.13229301571846008,
+      "learning_rate": 0.001,
+      "loss": 1.9206,
+      "step": 21096
+    },
+    {
+      "epoch": 0.8925035959048989,
+      "grad_norm": 3.1202783584594727,
+      "learning_rate": 0.001,
+      "loss": 1.9438,
+      "step": 21097
+    },
+    {
+      "epoch": 0.8925459006684152,
+      "grad_norm": 0.17747105658054352,
+      "learning_rate": 0.001,
+      "loss": 2.134,
+      "step": 21098
+    },
+    {
+      "epoch": 0.8925882054319316,
+      "grad_norm": 0.16291233897209167,
+      "learning_rate": 0.001,
+      "loss": 2.8814,
+      "step": 21099
+    },
+    {
+      "epoch": 0.892630510195448,
+      "grad_norm": 0.13646738231182098,
+      "learning_rate": 0.001,
+      "loss": 2.265,
+      "step": 21100
+    },
+    {
+      "epoch": 0.8926728149589643,
+      "grad_norm": 0.19404670596122742,
+      "learning_rate": 0.001,
+      "loss": 2.3592,
+      "step": 21101
+    },
+    {
+      "epoch": 0.8927151197224807,
+      "grad_norm": 0.5409544110298157,
+      "learning_rate": 0.001,
+      "loss": 2.8194,
+      "step": 21102
+    },
+    {
+      "epoch": 0.8927574244859972,
+      "grad_norm": 36.899574279785156,
+      "learning_rate": 0.001,
+      "loss": 1.8905,
+      "step": 21103
+    },
+    {
+      "epoch": 0.8927997292495135,
+      "grad_norm": 0.705781877040863,
+      "learning_rate": 0.001,
+      "loss": 2.5204,
+      "step": 21104
+    },
+    {
+      "epoch": 0.8928420340130299,
+      "grad_norm": 0.14836551249027252,
+      "learning_rate": 0.001,
+      "loss": 1.9536,
+      "step": 21105
+    },
+    {
+      "epoch": 0.8928843387765463,
+      "grad_norm": 0.14388398826122284,
+      "learning_rate": 0.001,
+      "loss": 2.0019,
+      "step": 21106
+    },
+    {
+      "epoch": 0.8929266435400626,
+      "grad_norm": 0.15267707407474518,
+      "learning_rate": 0.001,
+      "loss": 1.9647,
+      "step": 21107
+    },
+    {
+      "epoch": 0.892968948303579,
+      "grad_norm": 0.8427422046661377,
+      "learning_rate": 0.001,
+      "loss": 2.3751,
+      "step": 21108
+    },
+    {
+      "epoch": 0.8930112530670954,
+      "grad_norm": 0.13757242262363434,
+      "learning_rate": 0.001,
+      "loss": 1.9751,
+      "step": 21109
+    },
+    {
+      "epoch": 0.8930535578306117,
+      "grad_norm": 4.548515319824219,
+      "learning_rate": 0.001,
+      "loss": 1.8397,
+      "step": 21110
+    },
+    {
+      "epoch": 0.8930958625941281,
+      "grad_norm": 0.14274819195270538,
+      "learning_rate": 0.001,
+      "loss": 1.7187,
+      "step": 21111
+    },
+    {
+      "epoch": 0.8931381673576445,
+      "grad_norm": 0.15041141211986542,
+      "learning_rate": 0.001,
+      "loss": 2.6026,
+      "step": 21112
+    },
+    {
+      "epoch": 0.8931804721211608,
+      "grad_norm": 0.1963375061750412,
+      "learning_rate": 0.001,
+      "loss": 2.2654,
+      "step": 21113
+    },
+    {
+      "epoch": 0.8932227768846772,
+      "grad_norm": 0.14106130599975586,
+      "learning_rate": 0.001,
+      "loss": 2.6673,
+      "step": 21114
+    },
+    {
+      "epoch": 0.8932650816481936,
+      "grad_norm": 0.15438930690288544,
+      "learning_rate": 0.001,
+      "loss": 1.556,
+      "step": 21115
+    },
+    {
+      "epoch": 0.8933073864117099,
+      "grad_norm": 0.18182194232940674,
+      "learning_rate": 0.001,
+      "loss": 2.803,
+      "step": 21116
+    },
+    {
+      "epoch": 0.8933496911752263,
+      "grad_norm": 2.966031074523926,
+      "learning_rate": 0.001,
+      "loss": 3.5384,
+      "step": 21117
+    },
+    {
+      "epoch": 0.8933919959387427,
+      "grad_norm": 0.16081485152244568,
+      "learning_rate": 0.001,
+      "loss": 2.8794,
+      "step": 21118
+    },
+    {
+      "epoch": 0.893434300702259,
+      "grad_norm": 0.15742535889148712,
+      "learning_rate": 0.001,
+      "loss": 1.7857,
+      "step": 21119
+    },
+    {
+      "epoch": 0.8934766054657755,
+      "grad_norm": 0.16597464680671692,
+      "learning_rate": 0.001,
+      "loss": 2.4486,
+      "step": 21120
+    },
+    {
+      "epoch": 0.8935189102292919,
+      "grad_norm": 0.1744956225156784,
+      "learning_rate": 0.001,
+      "loss": 2.2636,
+      "step": 21121
+    },
+    {
+      "epoch": 0.8935612149928082,
+      "grad_norm": 0.1758044958114624,
+      "learning_rate": 0.001,
+      "loss": 2.0722,
+      "step": 21122
+    },
+    {
+      "epoch": 0.8936035197563246,
+      "grad_norm": 0.1646287441253662,
+      "learning_rate": 0.001,
+      "loss": 2.4013,
+      "step": 21123
+    },
+    {
+      "epoch": 0.893645824519841,
+      "grad_norm": 1.3558330535888672,
+      "learning_rate": 0.001,
+      "loss": 1.6679,
+      "step": 21124
+    },
+    {
+      "epoch": 0.8936881292833573,
+      "grad_norm": 0.1626473218202591,
+      "learning_rate": 0.001,
+      "loss": 1.708,
+      "step": 21125
+    },
+    {
+      "epoch": 0.8937304340468737,
+      "grad_norm": 0.505549967288971,
+      "learning_rate": 0.001,
+      "loss": 1.9515,
+      "step": 21126
+    },
+    {
+      "epoch": 0.8937727388103901,
+      "grad_norm": 0.20900815725326538,
+      "learning_rate": 0.001,
+      "loss": 2.2761,
+      "step": 21127
+    },
+    {
+      "epoch": 0.8938150435739064,
+      "grad_norm": 0.1580853909254074,
+      "learning_rate": 0.001,
+      "loss": 2.0052,
+      "step": 21128
+    },
+    {
+      "epoch": 0.8938573483374228,
+      "grad_norm": 0.14948229491710663,
+      "learning_rate": 0.001,
+      "loss": 1.5362,
+      "step": 21129
+    },
+    {
+      "epoch": 0.8938996531009392,
+      "grad_norm": 0.17575781047344208,
+      "learning_rate": 0.001,
+      "loss": 1.8339,
+      "step": 21130
+    },
+    {
+      "epoch": 0.8939419578644555,
+      "grad_norm": 0.16846154630184174,
+      "learning_rate": 0.001,
+      "loss": 3.2954,
+      "step": 21131
+    },
+    {
+      "epoch": 0.8939842626279719,
+      "grad_norm": 0.18402104079723358,
+      "learning_rate": 0.001,
+      "loss": 2.1852,
+      "step": 21132
+    },
+    {
+      "epoch": 0.8940265673914883,
+      "grad_norm": 0.1554890275001526,
+      "learning_rate": 0.001,
+      "loss": 2.3382,
+      "step": 21133
+    },
+    {
+      "epoch": 0.8940688721550046,
+      "grad_norm": 0.13957303762435913,
+      "learning_rate": 0.001,
+      "loss": 2.1363,
+      "step": 21134
+    },
+    {
+      "epoch": 0.894111176918521,
+      "grad_norm": 0.16399943828582764,
+      "learning_rate": 0.001,
+      "loss": 1.6929,
+      "step": 21135
+    },
+    {
+      "epoch": 0.8941534816820373,
+      "grad_norm": 0.8293852210044861,
+      "learning_rate": 0.001,
+      "loss": 1.8677,
+      "step": 21136
+    },
+    {
+      "epoch": 0.8941957864455538,
+      "grad_norm": 0.15984512865543365,
+      "learning_rate": 0.001,
+      "loss": 1.7682,
+      "step": 21137
+    },
+    {
+      "epoch": 0.8942380912090702,
+      "grad_norm": 0.3316708505153656,
+      "learning_rate": 0.001,
+      "loss": 2.9118,
+      "step": 21138
+    },
+    {
+      "epoch": 0.8942803959725865,
+      "grad_norm": 0.1807178258895874,
+      "learning_rate": 0.001,
+      "loss": 2.949,
+      "step": 21139
+    },
+    {
+      "epoch": 0.8943227007361029,
+      "grad_norm": 0.13455355167388916,
+      "learning_rate": 0.001,
+      "loss": 1.867,
+      "step": 21140
+    },
+    {
+      "epoch": 0.8943650054996193,
+      "grad_norm": 0.1604076772928238,
+      "learning_rate": 0.001,
+      "loss": 3.1111,
+      "step": 21141
+    },
+    {
+      "epoch": 0.8944073102631356,
+      "grad_norm": 0.2634793221950531,
+      "learning_rate": 0.001,
+      "loss": 1.3338,
+      "step": 21142
+    },
+    {
+      "epoch": 0.894449615026652,
+      "grad_norm": 0.15633098781108856,
+      "learning_rate": 0.001,
+      "loss": 2.0984,
+      "step": 21143
+    },
+    {
+      "epoch": 0.8944919197901684,
+      "grad_norm": 0.16510465741157532,
+      "learning_rate": 0.001,
+      "loss": 1.7809,
+      "step": 21144
+    },
+    {
+      "epoch": 0.8945342245536847,
+      "grad_norm": 0.13871534168720245,
+      "learning_rate": 0.001,
+      "loss": 1.702,
+      "step": 21145
+    },
+    {
+      "epoch": 0.8945765293172011,
+      "grad_norm": 3.331571340560913,
+      "learning_rate": 0.001,
+      "loss": 3.3754,
+      "step": 21146
+    },
+    {
+      "epoch": 0.8946188340807175,
+      "grad_norm": 0.14018763601779938,
+      "learning_rate": 0.001,
+      "loss": 2.0596,
+      "step": 21147
+    },
+    {
+      "epoch": 0.8946611388442338,
+      "grad_norm": 0.14845266938209534,
+      "learning_rate": 0.001,
+      "loss": 1.8956,
+      "step": 21148
+    },
+    {
+      "epoch": 0.8947034436077502,
+      "grad_norm": 0.18321087956428528,
+      "learning_rate": 0.001,
+      "loss": 1.821,
+      "step": 21149
+    },
+    {
+      "epoch": 0.8947457483712666,
+      "grad_norm": 39.97751998901367,
+      "learning_rate": 0.001,
+      "loss": 2.6723,
+      "step": 21150
+    },
+    {
+      "epoch": 0.8947880531347829,
+      "grad_norm": 0.7367990612983704,
+      "learning_rate": 0.001,
+      "loss": 2.3226,
+      "step": 21151
+    },
+    {
+      "epoch": 0.8948303578982993,
+      "grad_norm": 0.15124103426933289,
+      "learning_rate": 0.001,
+      "loss": 1.99,
+      "step": 21152
+    },
+    {
+      "epoch": 0.8948726626618158,
+      "grad_norm": 0.14537779986858368,
+      "learning_rate": 0.001,
+      "loss": 2.4792,
+      "step": 21153
+    },
+    {
+      "epoch": 0.894914967425332,
+      "grad_norm": 0.16853304207324982,
+      "learning_rate": 0.001,
+      "loss": 2.233,
+      "step": 21154
+    },
+    {
+      "epoch": 0.8949572721888485,
+      "grad_norm": 0.17724493145942688,
+      "learning_rate": 0.001,
+      "loss": 2.6854,
+      "step": 21155
+    },
+    {
+      "epoch": 0.8949995769523649,
+      "grad_norm": 0.1677272468805313,
+      "learning_rate": 0.001,
+      "loss": 1.6028,
+      "step": 21156
+    },
+    {
+      "epoch": 0.8950418817158812,
+      "grad_norm": 0.17956793308258057,
+      "learning_rate": 0.001,
+      "loss": 1.9077,
+      "step": 21157
+    },
+    {
+      "epoch": 0.8950841864793976,
+      "grad_norm": 0.18584755063056946,
+      "learning_rate": 0.001,
+      "loss": 1.6491,
+      "step": 21158
+    },
+    {
+      "epoch": 0.895126491242914,
+      "grad_norm": 0.15615415573120117,
+      "learning_rate": 0.001,
+      "loss": 2.2369,
+      "step": 21159
+    },
+    {
+      "epoch": 0.8951687960064303,
+      "grad_norm": 0.16031022369861603,
+      "learning_rate": 0.001,
+      "loss": 1.9678,
+      "step": 21160
+    },
+    {
+      "epoch": 0.8952111007699467,
+      "grad_norm": 10.240532875061035,
+      "learning_rate": 0.001,
+      "loss": 3.2672,
+      "step": 21161
+    },
+    {
+      "epoch": 0.8952534055334631,
+      "grad_norm": 0.1851503998041153,
+      "learning_rate": 0.001,
+      "loss": 2.2191,
+      "step": 21162
+    },
+    {
+      "epoch": 0.8952957102969794,
+      "grad_norm": 0.17084771394729614,
+      "learning_rate": 0.001,
+      "loss": 1.9097,
+      "step": 21163
+    },
+    {
+      "epoch": 0.8953380150604958,
+      "grad_norm": 0.6595641374588013,
+      "learning_rate": 0.001,
+      "loss": 3.2942,
+      "step": 21164
+    },
+    {
+      "epoch": 0.8953803198240122,
+      "grad_norm": 0.12742707133293152,
+      "learning_rate": 0.001,
+      "loss": 2.523,
+      "step": 21165
+    },
+    {
+      "epoch": 0.8954226245875285,
+      "grad_norm": 3.2080767154693604,
+      "learning_rate": 0.001,
+      "loss": 2.5101,
+      "step": 21166
+    },
+    {
+      "epoch": 0.8954649293510449,
+      "grad_norm": 0.1244649738073349,
+      "learning_rate": 0.001,
+      "loss": 3.2143,
+      "step": 21167
+    },
+    {
+      "epoch": 0.8955072341145613,
+      "grad_norm": 0.14987188577651978,
+      "learning_rate": 0.001,
+      "loss": 2.8929,
+      "step": 21168
+    },
+    {
+      "epoch": 0.8955495388780776,
+      "grad_norm": 0.20873646438121796,
+      "learning_rate": 0.001,
+      "loss": 1.5905,
+      "step": 21169
+    },
+    {
+      "epoch": 0.895591843641594,
+      "grad_norm": 1.1376652717590332,
+      "learning_rate": 0.001,
+      "loss": 1.4977,
+      "step": 21170
+    },
+    {
+      "epoch": 0.8956341484051105,
+      "grad_norm": 0.16955767571926117,
+      "learning_rate": 0.001,
+      "loss": 2.2745,
+      "step": 21171
+    },
+    {
+      "epoch": 0.8956764531686268,
+      "grad_norm": 0.1822686642408371,
+      "learning_rate": 0.001,
+      "loss": 2.4192,
+      "step": 21172
+    },
+    {
+      "epoch": 0.8957187579321432,
+      "grad_norm": 0.1916595995426178,
+      "learning_rate": 0.001,
+      "loss": 1.7201,
+      "step": 21173
+    },
+    {
+      "epoch": 0.8957610626956596,
+      "grad_norm": 0.14500486850738525,
+      "learning_rate": 0.001,
+      "loss": 1.6737,
+      "step": 21174
+    },
+    {
+      "epoch": 0.8958033674591759,
+      "grad_norm": 0.15613558888435364,
+      "learning_rate": 0.001,
+      "loss": 1.9017,
+      "step": 21175
+    },
+    {
+      "epoch": 0.8958456722226923,
+      "grad_norm": 0.149849534034729,
+      "learning_rate": 0.001,
+      "loss": 2.7908,
+      "step": 21176
+    },
+    {
+      "epoch": 0.8958879769862087,
+      "grad_norm": 0.1522967666387558,
+      "learning_rate": 0.001,
+      "loss": 3.3811,
+      "step": 21177
+    },
+    {
+      "epoch": 0.895930281749725,
+      "grad_norm": 0.20839162170886993,
+      "learning_rate": 0.001,
+      "loss": 1.6453,
+      "step": 21178
+    },
+    {
+      "epoch": 0.8959725865132414,
+      "grad_norm": 0.18423724174499512,
+      "learning_rate": 0.001,
+      "loss": 2.3114,
+      "step": 21179
+    },
+    {
+      "epoch": 0.8960148912767577,
+      "grad_norm": 0.20367826521396637,
+      "learning_rate": 0.001,
+      "loss": 2.0914,
+      "step": 21180
+    },
+    {
+      "epoch": 0.8960571960402741,
+      "grad_norm": 2.9200098514556885,
+      "learning_rate": 0.001,
+      "loss": 1.8609,
+      "step": 21181
+    },
+    {
+      "epoch": 0.8960995008037905,
+      "grad_norm": 0.2010602205991745,
+      "learning_rate": 0.001,
+      "loss": 2.1336,
+      "step": 21182
+    },
+    {
+      "epoch": 0.8961418055673068,
+      "grad_norm": 0.15395566821098328,
+      "learning_rate": 0.001,
+      "loss": 3.506,
+      "step": 21183
+    },
+    {
+      "epoch": 0.8961841103308232,
+      "grad_norm": 0.17142242193222046,
+      "learning_rate": 0.001,
+      "loss": 3.1603,
+      "step": 21184
+    },
+    {
+      "epoch": 0.8962264150943396,
+      "grad_norm": 0.36323732137680054,
+      "learning_rate": 0.001,
+      "loss": 2.0423,
+      "step": 21185
+    },
+    {
+      "epoch": 0.8962687198578559,
+      "grad_norm": 1.7019706964492798,
+      "learning_rate": 0.001,
+      "loss": 3.0269,
+      "step": 21186
+    },
+    {
+      "epoch": 0.8963110246213724,
+      "grad_norm": 2.892049551010132,
+      "learning_rate": 0.001,
+      "loss": 2.2748,
+      "step": 21187
+    },
+    {
+      "epoch": 0.8963533293848888,
+      "grad_norm": 0.4944346845149994,
+      "learning_rate": 0.001,
+      "loss": 2.5055,
+      "step": 21188
+    },
+    {
+      "epoch": 0.8963956341484051,
+      "grad_norm": 0.19812704622745514,
+      "learning_rate": 0.001,
+      "loss": 1.639,
+      "step": 21189
+    },
+    {
+      "epoch": 0.8964379389119215,
+      "grad_norm": 0.32802027463912964,
+      "learning_rate": 0.001,
+      "loss": 2.4047,
+      "step": 21190
+    },
+    {
+      "epoch": 0.8964802436754379,
+      "grad_norm": 0.15056705474853516,
+      "learning_rate": 0.001,
+      "loss": 1.724,
+      "step": 21191
+    },
+    {
+      "epoch": 0.8965225484389542,
+      "grad_norm": 0.15505051612854004,
+      "learning_rate": 0.001,
+      "loss": 2.1101,
+      "step": 21192
+    },
+    {
+      "epoch": 0.8965648532024706,
+      "grad_norm": 0.16855968534946442,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 21193
+    },
+    {
+      "epoch": 0.896607157965987,
+      "grad_norm": 0.15163734555244446,
+      "learning_rate": 0.001,
+      "loss": 1.4858,
+      "step": 21194
+    },
+    {
+      "epoch": 0.8966494627295033,
+      "grad_norm": 0.21959041059017181,
+      "learning_rate": 0.001,
+      "loss": 2.2257,
+      "step": 21195
+    },
+    {
+      "epoch": 0.8966917674930197,
+      "grad_norm": 0.18358977138996124,
+      "learning_rate": 0.001,
+      "loss": 1.8863,
+      "step": 21196
+    },
+    {
+      "epoch": 0.8967340722565361,
+      "grad_norm": 0.19511397182941437,
+      "learning_rate": 0.001,
+      "loss": 1.9636,
+      "step": 21197
+    },
+    {
+      "epoch": 0.8967763770200524,
+      "grad_norm": 0.19627627730369568,
+      "learning_rate": 0.001,
+      "loss": 2.0717,
+      "step": 21198
+    },
+    {
+      "epoch": 0.8968186817835688,
+      "grad_norm": 0.22323928773403168,
+      "learning_rate": 0.001,
+      "loss": 1.7666,
+      "step": 21199
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.14983603358268738,
+      "learning_rate": 0.001,
+      "loss": 2.8837,
+      "step": 21200
+    },
+    {
+      "epoch": 0.8969032913106015,
+      "grad_norm": 0.13686136901378632,
+      "learning_rate": 0.001,
+      "loss": 2.385,
+      "step": 21201
+    },
+    {
+      "epoch": 0.896945596074118,
+      "grad_norm": 0.15880164504051208,
+      "learning_rate": 0.001,
+      "loss": 2.1539,
+      "step": 21202
+    },
+    {
+      "epoch": 0.8969879008376344,
+      "grad_norm": 0.13948020339012146,
+      "learning_rate": 0.001,
+      "loss": 1.6086,
+      "step": 21203
+    },
+    {
+      "epoch": 0.8970302056011507,
+      "grad_norm": 0.12735982239246368,
+      "learning_rate": 0.001,
+      "loss": 1.8913,
+      "step": 21204
+    },
+    {
+      "epoch": 0.8970725103646671,
+      "grad_norm": 0.1532490998506546,
+      "learning_rate": 0.001,
+      "loss": 1.7689,
+      "step": 21205
+    },
+    {
+      "epoch": 0.8971148151281835,
+      "grad_norm": 0.20498578250408173,
+      "learning_rate": 0.001,
+      "loss": 1.7279,
+      "step": 21206
+    },
+    {
+      "epoch": 0.8971571198916998,
+      "grad_norm": 0.13713718950748444,
+      "learning_rate": 0.001,
+      "loss": 1.7066,
+      "step": 21207
+    },
+    {
+      "epoch": 0.8971994246552162,
+      "grad_norm": 0.15375199913978577,
+      "learning_rate": 0.001,
+      "loss": 1.5455,
+      "step": 21208
+    },
+    {
+      "epoch": 0.8972417294187326,
+      "grad_norm": 0.12200403213500977,
+      "learning_rate": 0.001,
+      "loss": 1.8769,
+      "step": 21209
+    },
+    {
+      "epoch": 0.8972840341822489,
+      "grad_norm": 0.15674884617328644,
+      "learning_rate": 0.001,
+      "loss": 2.4422,
+      "step": 21210
+    },
+    {
+      "epoch": 0.8973263389457653,
+      "grad_norm": 0.7319082617759705,
+      "learning_rate": 0.001,
+      "loss": 2.3794,
+      "step": 21211
+    },
+    {
+      "epoch": 0.8973686437092817,
+      "grad_norm": 0.14720921218395233,
+      "learning_rate": 0.001,
+      "loss": 1.9387,
+      "step": 21212
+    },
+    {
+      "epoch": 0.897410948472798,
+      "grad_norm": 0.2884654402732849,
+      "learning_rate": 0.001,
+      "loss": 3.2008,
+      "step": 21213
+    },
+    {
+      "epoch": 0.8974532532363144,
+      "grad_norm": 1.7841389179229736,
+      "learning_rate": 0.001,
+      "loss": 3.6275,
+      "step": 21214
+    },
+    {
+      "epoch": 0.8974955579998308,
+      "grad_norm": 0.7872580885887146,
+      "learning_rate": 0.001,
+      "loss": 1.6937,
+      "step": 21215
+    },
+    {
+      "epoch": 0.8975378627633471,
+      "grad_norm": 0.16804750263690948,
+      "learning_rate": 0.001,
+      "loss": 2.4865,
+      "step": 21216
+    },
+    {
+      "epoch": 0.8975801675268635,
+      "grad_norm": 0.18174222111701965,
+      "learning_rate": 0.001,
+      "loss": 2.2794,
+      "step": 21217
+    },
+    {
+      "epoch": 0.89762247229038,
+      "grad_norm": 0.21821044385433197,
+      "learning_rate": 0.001,
+      "loss": 3.6223,
+      "step": 21218
+    },
+    {
+      "epoch": 0.8976647770538962,
+      "grad_norm": 0.2736088037490845,
+      "learning_rate": 0.001,
+      "loss": 2.0796,
+      "step": 21219
+    },
+    {
+      "epoch": 0.8977070818174127,
+      "grad_norm": 0.15616385638713837,
+      "learning_rate": 0.001,
+      "loss": 2.9314,
+      "step": 21220
+    },
+    {
+      "epoch": 0.8977493865809291,
+      "grad_norm": 0.13997872173786163,
+      "learning_rate": 0.001,
+      "loss": 1.6217,
+      "step": 21221
+    },
+    {
+      "epoch": 0.8977916913444454,
+      "grad_norm": 0.17286120355129242,
+      "learning_rate": 0.001,
+      "loss": 2.5522,
+      "step": 21222
+    },
+    {
+      "epoch": 0.8978339961079618,
+      "grad_norm": 0.16818682849407196,
+      "learning_rate": 0.001,
+      "loss": 2.1953,
+      "step": 21223
+    },
+    {
+      "epoch": 0.8978763008714782,
+      "grad_norm": 0.8613235354423523,
+      "learning_rate": 0.001,
+      "loss": 2.311,
+      "step": 21224
+    },
+    {
+      "epoch": 0.8979186056349945,
+      "grad_norm": 1.0731678009033203,
+      "learning_rate": 0.001,
+      "loss": 3.8418,
+      "step": 21225
+    },
+    {
+      "epoch": 0.8979609103985109,
+      "grad_norm": 0.1638665348291397,
+      "learning_rate": 0.001,
+      "loss": 2.2482,
+      "step": 21226
+    },
+    {
+      "epoch": 0.8980032151620272,
+      "grad_norm": 0.3347175419330597,
+      "learning_rate": 0.001,
+      "loss": 1.9502,
+      "step": 21227
+    },
+    {
+      "epoch": 0.8980455199255436,
+      "grad_norm": 1.1207183599472046,
+      "learning_rate": 0.001,
+      "loss": 1.5674,
+      "step": 21228
+    },
+    {
+      "epoch": 0.89808782468906,
+      "grad_norm": 0.13141308724880219,
+      "learning_rate": 0.001,
+      "loss": 2.1263,
+      "step": 21229
+    },
+    {
+      "epoch": 0.8981301294525763,
+      "grad_norm": 0.14101554453372955,
+      "learning_rate": 0.001,
+      "loss": 2.591,
+      "step": 21230
+    },
+    {
+      "epoch": 0.8981724342160927,
+      "grad_norm": 0.17062920331954956,
+      "learning_rate": 0.001,
+      "loss": 1.6321,
+      "step": 21231
+    },
+    {
+      "epoch": 0.8982147389796091,
+      "grad_norm": 0.18323302268981934,
+      "learning_rate": 0.001,
+      "loss": 2.9428,
+      "step": 21232
+    },
+    {
+      "epoch": 0.8982570437431254,
+      "grad_norm": 0.15976440906524658,
+      "learning_rate": 0.001,
+      "loss": 1.8825,
+      "step": 21233
+    },
+    {
+      "epoch": 0.8982993485066418,
+      "grad_norm": 0.15031205117702484,
+      "learning_rate": 0.001,
+      "loss": 2.5819,
+      "step": 21234
+    },
+    {
+      "epoch": 0.8983416532701582,
+      "grad_norm": 4.328239917755127,
+      "learning_rate": 0.001,
+      "loss": 1.6748,
+      "step": 21235
+    },
+    {
+      "epoch": 0.8983839580336745,
+      "grad_norm": 0.17019319534301758,
+      "learning_rate": 0.001,
+      "loss": 3.3208,
+      "step": 21236
+    },
+    {
+      "epoch": 0.898426262797191,
+      "grad_norm": 0.21404282748699188,
+      "learning_rate": 0.001,
+      "loss": 1.6199,
+      "step": 21237
+    },
+    {
+      "epoch": 0.8984685675607074,
+      "grad_norm": 0.15579356253147125,
+      "learning_rate": 0.001,
+      "loss": 1.7415,
+      "step": 21238
+    },
+    {
+      "epoch": 0.8985108723242237,
+      "grad_norm": 0.3757559359073639,
+      "learning_rate": 0.001,
+      "loss": 1.8607,
+      "step": 21239
+    },
+    {
+      "epoch": 0.8985531770877401,
+      "grad_norm": 0.16331882774829865,
+      "learning_rate": 0.001,
+      "loss": 2.3862,
+      "step": 21240
+    },
+    {
+      "epoch": 0.8985954818512565,
+      "grad_norm": 0.1645757555961609,
+      "learning_rate": 0.001,
+      "loss": 1.8158,
+      "step": 21241
+    },
+    {
+      "epoch": 0.8986377866147728,
+      "grad_norm": 0.159241184592247,
+      "learning_rate": 0.001,
+      "loss": 1.7989,
+      "step": 21242
+    },
+    {
+      "epoch": 0.8986800913782892,
+      "grad_norm": 0.1543315201997757,
+      "learning_rate": 0.001,
+      "loss": 3.289,
+      "step": 21243
+    },
+    {
+      "epoch": 0.8987223961418056,
+      "grad_norm": 0.14359703660011292,
+      "learning_rate": 0.001,
+      "loss": 2.5562,
+      "step": 21244
+    },
+    {
+      "epoch": 0.8987647009053219,
+      "grad_norm": 2.587345600128174,
+      "learning_rate": 0.001,
+      "loss": 2.6168,
+      "step": 21245
+    },
+    {
+      "epoch": 0.8988070056688383,
+      "grad_norm": 0.17357079684734344,
+      "learning_rate": 0.001,
+      "loss": 1.9072,
+      "step": 21246
+    },
+    {
+      "epoch": 0.8988493104323547,
+      "grad_norm": 0.1991180181503296,
+      "learning_rate": 0.001,
+      "loss": 1.8736,
+      "step": 21247
+    },
+    {
+      "epoch": 0.898891615195871,
+      "grad_norm": 0.14238642156124115,
+      "learning_rate": 0.001,
+      "loss": 1.1997,
+      "step": 21248
+    },
+    {
+      "epoch": 0.8989339199593874,
+      "grad_norm": 0.20077811181545258,
+      "learning_rate": 0.001,
+      "loss": 2.0872,
+      "step": 21249
+    },
+    {
+      "epoch": 0.8989762247229038,
+      "grad_norm": 0.7532477378845215,
+      "learning_rate": 0.001,
+      "loss": 2.3801,
+      "step": 21250
+    },
+    {
+      "epoch": 0.8990185294864201,
+      "grad_norm": 0.1633615493774414,
+      "learning_rate": 0.001,
+      "loss": 2.2373,
+      "step": 21251
+    },
+    {
+      "epoch": 0.8990608342499365,
+      "grad_norm": 0.13562917709350586,
+      "learning_rate": 0.001,
+      "loss": 1.5633,
+      "step": 21252
+    },
+    {
+      "epoch": 0.899103139013453,
+      "grad_norm": 0.14501778781414032,
+      "learning_rate": 0.001,
+      "loss": 1.7519,
+      "step": 21253
+    },
+    {
+      "epoch": 0.8991454437769693,
+      "grad_norm": 0.14665882289409637,
+      "learning_rate": 0.001,
+      "loss": 1.2838,
+      "step": 21254
+    },
+    {
+      "epoch": 0.8991877485404857,
+      "grad_norm": 0.14085936546325684,
+      "learning_rate": 0.001,
+      "loss": 1.6832,
+      "step": 21255
+    },
+    {
+      "epoch": 0.8992300533040021,
+      "grad_norm": 0.5537809133529663,
+      "learning_rate": 0.001,
+      "loss": 1.9436,
+      "step": 21256
+    },
+    {
+      "epoch": 0.8992723580675184,
+      "grad_norm": 0.14022964239120483,
+      "learning_rate": 0.001,
+      "loss": 2.0391,
+      "step": 21257
+    },
+    {
+      "epoch": 0.8993146628310348,
+      "grad_norm": 0.13678425550460815,
+      "learning_rate": 0.001,
+      "loss": 1.2781,
+      "step": 21258
+    },
+    {
+      "epoch": 0.8993569675945512,
+      "grad_norm": 0.1590747982263565,
+      "learning_rate": 0.001,
+      "loss": 2.4737,
+      "step": 21259
+    },
+    {
+      "epoch": 0.8993992723580675,
+      "grad_norm": 0.31500861048698425,
+      "learning_rate": 0.001,
+      "loss": 2.0949,
+      "step": 21260
+    },
+    {
+      "epoch": 0.8994415771215839,
+      "grad_norm": 0.7429576516151428,
+      "learning_rate": 0.001,
+      "loss": 2.6399,
+      "step": 21261
+    },
+    {
+      "epoch": 0.8994838818851003,
+      "grad_norm": 0.1560697704553604,
+      "learning_rate": 0.001,
+      "loss": 2.4382,
+      "step": 21262
+    },
+    {
+      "epoch": 0.8995261866486166,
+      "grad_norm": 0.1646103411912918,
+      "learning_rate": 0.001,
+      "loss": 1.8626,
+      "step": 21263
+    },
+    {
+      "epoch": 0.899568491412133,
+      "grad_norm": 0.4191820025444031,
+      "learning_rate": 0.001,
+      "loss": 2.2337,
+      "step": 21264
+    },
+    {
+      "epoch": 0.8996107961756494,
+      "grad_norm": 0.16874629259109497,
+      "learning_rate": 0.001,
+      "loss": 3.4052,
+      "step": 21265
+    },
+    {
+      "epoch": 0.8996531009391657,
+      "grad_norm": 0.12579210102558136,
+      "learning_rate": 0.001,
+      "loss": 2.4531,
+      "step": 21266
+    },
+    {
+      "epoch": 0.8996954057026821,
+      "grad_norm": 0.14168886840343475,
+      "learning_rate": 0.001,
+      "loss": 2.4614,
+      "step": 21267
+    },
+    {
+      "epoch": 0.8997377104661985,
+      "grad_norm": 0.16749995946884155,
+      "learning_rate": 0.001,
+      "loss": 1.3653,
+      "step": 21268
+    },
+    {
+      "epoch": 0.8997800152297148,
+      "grad_norm": 0.6988632678985596,
+      "learning_rate": 0.001,
+      "loss": 1.7769,
+      "step": 21269
+    },
+    {
+      "epoch": 0.8998223199932313,
+      "grad_norm": 0.22250531613826752,
+      "learning_rate": 0.001,
+      "loss": 1.8593,
+      "step": 21270
+    },
+    {
+      "epoch": 0.8998646247567476,
+      "grad_norm": 0.11966440081596375,
+      "learning_rate": 0.001,
+      "loss": 1.4307,
+      "step": 21271
+    },
+    {
+      "epoch": 0.899906929520264,
+      "grad_norm": 0.3543547987937927,
+      "learning_rate": 0.001,
+      "loss": 1.6872,
+      "step": 21272
+    },
+    {
+      "epoch": 0.8999492342837804,
+      "grad_norm": 0.15212112665176392,
+      "learning_rate": 0.001,
+      "loss": 2.4326,
+      "step": 21273
+    },
+    {
+      "epoch": 0.8999915390472967,
+      "grad_norm": 0.14390593767166138,
+      "learning_rate": 0.001,
+      "loss": 1.6194,
+      "step": 21274
+    },
+    {
+      "epoch": 0.9000338438108131,
+      "grad_norm": 0.1561872363090515,
+      "learning_rate": 0.001,
+      "loss": 1.9529,
+      "step": 21275
+    },
+    {
+      "epoch": 0.9000761485743295,
+      "grad_norm": 0.14127811789512634,
+      "learning_rate": 0.001,
+      "loss": 1.3719,
+      "step": 21276
+    },
+    {
+      "epoch": 0.9001184533378458,
+      "grad_norm": 0.13319812715053558,
+      "learning_rate": 0.001,
+      "loss": 1.8501,
+      "step": 21277
+    },
+    {
+      "epoch": 0.9001607581013622,
+      "grad_norm": 0.13405123353004456,
+      "learning_rate": 0.001,
+      "loss": 2.0724,
+      "step": 21278
+    },
+    {
+      "epoch": 0.9002030628648786,
+      "grad_norm": 0.18458141386508942,
+      "learning_rate": 0.001,
+      "loss": 2.1271,
+      "step": 21279
+    },
+    {
+      "epoch": 0.9002453676283949,
+      "grad_norm": 0.1432388573884964,
+      "learning_rate": 0.001,
+      "loss": 1.6285,
+      "step": 21280
+    },
+    {
+      "epoch": 0.9002876723919113,
+      "grad_norm": 0.12944474816322327,
+      "learning_rate": 0.001,
+      "loss": 2.7784,
+      "step": 21281
+    },
+    {
+      "epoch": 0.9003299771554277,
+      "grad_norm": 0.305156946182251,
+      "learning_rate": 0.001,
+      "loss": 2.5529,
+      "step": 21282
+    },
+    {
+      "epoch": 0.900372281918944,
+      "grad_norm": 104.79377746582031,
+      "learning_rate": 0.001,
+      "loss": 2.4094,
+      "step": 21283
+    },
+    {
+      "epoch": 0.9004145866824604,
+      "grad_norm": 0.16589900851249695,
+      "learning_rate": 0.001,
+      "loss": 2.0991,
+      "step": 21284
+    },
+    {
+      "epoch": 0.9004568914459768,
+      "grad_norm": 0.14582109451293945,
+      "learning_rate": 0.001,
+      "loss": 2.7329,
+      "step": 21285
+    },
+    {
+      "epoch": 0.9004991962094931,
+      "grad_norm": 0.17398712038993835,
+      "learning_rate": 0.001,
+      "loss": 1.9659,
+      "step": 21286
+    },
+    {
+      "epoch": 0.9005415009730096,
+      "grad_norm": 0.16176342964172363,
+      "learning_rate": 0.001,
+      "loss": 1.6683,
+      "step": 21287
+    },
+    {
+      "epoch": 0.900583805736526,
+      "grad_norm": 0.2542744278907776,
+      "learning_rate": 0.001,
+      "loss": 2.9904,
+      "step": 21288
+    },
+    {
+      "epoch": 0.9006261105000423,
+      "grad_norm": 0.1428564190864563,
+      "learning_rate": 0.001,
+      "loss": 2.8341,
+      "step": 21289
+    },
+    {
+      "epoch": 0.9006684152635587,
+      "grad_norm": 0.16790971159934998,
+      "learning_rate": 0.001,
+      "loss": 1.8499,
+      "step": 21290
+    },
+    {
+      "epoch": 0.9007107200270751,
+      "grad_norm": 0.32263800501823425,
+      "learning_rate": 0.001,
+      "loss": 2.4642,
+      "step": 21291
+    },
+    {
+      "epoch": 0.9007530247905914,
+      "grad_norm": 0.19915816187858582,
+      "learning_rate": 0.001,
+      "loss": 1.686,
+      "step": 21292
+    },
+    {
+      "epoch": 0.9007953295541078,
+      "grad_norm": 0.17057372629642487,
+      "learning_rate": 0.001,
+      "loss": 1.6805,
+      "step": 21293
+    },
+    {
+      "epoch": 0.9008376343176242,
+      "grad_norm": 0.22040168941020966,
+      "learning_rate": 0.001,
+      "loss": 2.4221,
+      "step": 21294
+    },
+    {
+      "epoch": 0.9008799390811405,
+      "grad_norm": 0.41536402702331543,
+      "learning_rate": 0.001,
+      "loss": 1.8372,
+      "step": 21295
+    },
+    {
+      "epoch": 0.9009222438446569,
+      "grad_norm": 2.5897037982940674,
+      "learning_rate": 0.001,
+      "loss": 2.0451,
+      "step": 21296
+    },
+    {
+      "epoch": 0.9009645486081733,
+      "grad_norm": 0.16217264533042908,
+      "learning_rate": 0.001,
+      "loss": 2.2002,
+      "step": 21297
+    },
+    {
+      "epoch": 0.9010068533716896,
+      "grad_norm": 0.16462668776512146,
+      "learning_rate": 0.001,
+      "loss": 2.3787,
+      "step": 21298
+    },
+    {
+      "epoch": 0.901049158135206,
+      "grad_norm": 0.13496233522891998,
+      "learning_rate": 0.001,
+      "loss": 1.757,
+      "step": 21299
+    },
+    {
+      "epoch": 0.9010914628987224,
+      "grad_norm": 0.15986299514770508,
+      "learning_rate": 0.001,
+      "loss": 2.5716,
+      "step": 21300
+    },
+    {
+      "epoch": 0.9011337676622387,
+      "grad_norm": 2.0072743892669678,
+      "learning_rate": 0.001,
+      "loss": 1.5631,
+      "step": 21301
+    },
+    {
+      "epoch": 0.9011760724257551,
+      "grad_norm": 0.15145206451416016,
+      "learning_rate": 0.001,
+      "loss": 1.5825,
+      "step": 21302
+    },
+    {
+      "epoch": 0.9012183771892716,
+      "grad_norm": 0.1307021826505661,
+      "learning_rate": 0.001,
+      "loss": 2.7586,
+      "step": 21303
+    },
+    {
+      "epoch": 0.9012606819527879,
+      "grad_norm": 1.082275152206421,
+      "learning_rate": 0.001,
+      "loss": 1.5857,
+      "step": 21304
+    },
+    {
+      "epoch": 0.9013029867163043,
+      "grad_norm": 0.13324807584285736,
+      "learning_rate": 0.001,
+      "loss": 1.9865,
+      "step": 21305
+    },
+    {
+      "epoch": 0.9013452914798207,
+      "grad_norm": 0.18132184445858002,
+      "learning_rate": 0.001,
+      "loss": 2.2404,
+      "step": 21306
+    },
+    {
+      "epoch": 0.901387596243337,
+      "grad_norm": 0.15229110419750214,
+      "learning_rate": 0.001,
+      "loss": 1.9643,
+      "step": 21307
+    },
+    {
+      "epoch": 0.9014299010068534,
+      "grad_norm": 0.15678609907627106,
+      "learning_rate": 0.001,
+      "loss": 1.4846,
+      "step": 21308
+    },
+    {
+      "epoch": 0.9014722057703698,
+      "grad_norm": 0.2569018006324768,
+      "learning_rate": 0.001,
+      "loss": 1.4851,
+      "step": 21309
+    },
+    {
+      "epoch": 0.9015145105338861,
+      "grad_norm": 0.2765081822872162,
+      "learning_rate": 0.001,
+      "loss": 2.6487,
+      "step": 21310
+    },
+    {
+      "epoch": 0.9015568152974025,
+      "grad_norm": 0.4249289631843567,
+      "learning_rate": 0.001,
+      "loss": 2.114,
+      "step": 21311
+    },
+    {
+      "epoch": 0.9015991200609189,
+      "grad_norm": 0.14523530006408691,
+      "learning_rate": 0.001,
+      "loss": 1.4716,
+      "step": 21312
+    },
+    {
+      "epoch": 0.9016414248244352,
+      "grad_norm": 0.14755897223949432,
+      "learning_rate": 0.001,
+      "loss": 1.7763,
+      "step": 21313
+    },
+    {
+      "epoch": 0.9016837295879516,
+      "grad_norm": 6.201207637786865,
+      "learning_rate": 0.001,
+      "loss": 1.6094,
+      "step": 21314
+    },
+    {
+      "epoch": 0.9017260343514679,
+      "grad_norm": 0.15470610558986664,
+      "learning_rate": 0.001,
+      "loss": 2.9894,
+      "step": 21315
+    },
+    {
+      "epoch": 0.9017683391149843,
+      "grad_norm": 0.2583865821361542,
+      "learning_rate": 0.001,
+      "loss": 2.2243,
+      "step": 21316
+    },
+    {
+      "epoch": 0.9018106438785007,
+      "grad_norm": 0.1813354194164276,
+      "learning_rate": 0.001,
+      "loss": 1.8183,
+      "step": 21317
+    },
+    {
+      "epoch": 0.901852948642017,
+      "grad_norm": 0.1495773047208786,
+      "learning_rate": 0.001,
+      "loss": 2.1804,
+      "step": 21318
+    },
+    {
+      "epoch": 0.9018952534055334,
+      "grad_norm": 0.15756896138191223,
+      "learning_rate": 0.001,
+      "loss": 1.8869,
+      "step": 21319
+    },
+    {
+      "epoch": 0.9019375581690499,
+      "grad_norm": 0.44224387407302856,
+      "learning_rate": 0.001,
+      "loss": 2.3278,
+      "step": 21320
+    },
+    {
+      "epoch": 0.9019798629325662,
+      "grad_norm": 0.23249466717243195,
+      "learning_rate": 0.001,
+      "loss": 1.8345,
+      "step": 21321
+    },
+    {
+      "epoch": 0.9020221676960826,
+      "grad_norm": 0.17175745964050293,
+      "learning_rate": 0.001,
+      "loss": 1.687,
+      "step": 21322
+    },
+    {
+      "epoch": 0.902064472459599,
+      "grad_norm": 0.15828604996204376,
+      "learning_rate": 0.001,
+      "loss": 2.0996,
+      "step": 21323
+    },
+    {
+      "epoch": 0.9021067772231153,
+      "grad_norm": 0.1508798450231552,
+      "learning_rate": 0.001,
+      "loss": 2.1381,
+      "step": 21324
+    },
+    {
+      "epoch": 0.9021490819866317,
+      "grad_norm": 0.1373659372329712,
+      "learning_rate": 0.001,
+      "loss": 2.6336,
+      "step": 21325
+    },
+    {
+      "epoch": 0.9021913867501481,
+      "grad_norm": 0.3546662926673889,
+      "learning_rate": 0.001,
+      "loss": 2.1206,
+      "step": 21326
+    },
+    {
+      "epoch": 0.9022336915136644,
+      "grad_norm": 0.17910721898078918,
+      "learning_rate": 0.001,
+      "loss": 3.109,
+      "step": 21327
+    },
+    {
+      "epoch": 0.9022759962771808,
+      "grad_norm": 2.0273354053497314,
+      "learning_rate": 0.001,
+      "loss": 1.6428,
+      "step": 21328
+    },
+    {
+      "epoch": 0.9023183010406972,
+      "grad_norm": 0.17860253155231476,
+      "learning_rate": 0.001,
+      "loss": 2.9683,
+      "step": 21329
+    },
+    {
+      "epoch": 0.9023606058042135,
+      "grad_norm": 0.16391193866729736,
+      "learning_rate": 0.001,
+      "loss": 1.66,
+      "step": 21330
+    },
+    {
+      "epoch": 0.9024029105677299,
+      "grad_norm": 0.41829827427864075,
+      "learning_rate": 0.001,
+      "loss": 2.3392,
+      "step": 21331
+    },
+    {
+      "epoch": 0.9024452153312463,
+      "grad_norm": 0.15510700643062592,
+      "learning_rate": 0.001,
+      "loss": 2.1811,
+      "step": 21332
+    },
+    {
+      "epoch": 0.9024875200947626,
+      "grad_norm": 0.16545583307743073,
+      "learning_rate": 0.001,
+      "loss": 2.1031,
+      "step": 21333
+    },
+    {
+      "epoch": 0.902529824858279,
+      "grad_norm": 1.293203592300415,
+      "learning_rate": 0.001,
+      "loss": 1.5144,
+      "step": 21334
+    },
+    {
+      "epoch": 0.9025721296217954,
+      "grad_norm": 0.14911191165447235,
+      "learning_rate": 0.001,
+      "loss": 2.7838,
+      "step": 21335
+    },
+    {
+      "epoch": 0.9026144343853117,
+      "grad_norm": 0.2668335437774658,
+      "learning_rate": 0.001,
+      "loss": 3.2185,
+      "step": 21336
+    },
+    {
+      "epoch": 0.9026567391488282,
+      "grad_norm": 0.2777593433856964,
+      "learning_rate": 0.001,
+      "loss": 2.5853,
+      "step": 21337
+    },
+    {
+      "epoch": 0.9026990439123446,
+      "grad_norm": 0.6844047904014587,
+      "learning_rate": 0.001,
+      "loss": 1.6514,
+      "step": 21338
+    },
+    {
+      "epoch": 0.9027413486758609,
+      "grad_norm": 0.1647360622882843,
+      "learning_rate": 0.001,
+      "loss": 2.2108,
+      "step": 21339
+    },
+    {
+      "epoch": 0.9027836534393773,
+      "grad_norm": 0.15791663527488708,
+      "learning_rate": 0.001,
+      "loss": 1.6112,
+      "step": 21340
+    },
+    {
+      "epoch": 0.9028259582028937,
+      "grad_norm": 0.1473037302494049,
+      "learning_rate": 0.001,
+      "loss": 2.266,
+      "step": 21341
+    },
+    {
+      "epoch": 0.90286826296641,
+      "grad_norm": 0.15689495205879211,
+      "learning_rate": 0.001,
+      "loss": 1.765,
+      "step": 21342
+    },
+    {
+      "epoch": 0.9029105677299264,
+      "grad_norm": 0.18425515294075012,
+      "learning_rate": 0.001,
+      "loss": 1.8901,
+      "step": 21343
+    },
+    {
+      "epoch": 0.9029528724934428,
+      "grad_norm": 0.2768228054046631,
+      "learning_rate": 0.001,
+      "loss": 1.9768,
+      "step": 21344
+    },
+    {
+      "epoch": 0.9029951772569591,
+      "grad_norm": 0.15262466669082642,
+      "learning_rate": 0.001,
+      "loss": 2.0877,
+      "step": 21345
+    },
+    {
+      "epoch": 0.9030374820204755,
+      "grad_norm": 2.121443748474121,
+      "learning_rate": 0.001,
+      "loss": 2.5901,
+      "step": 21346
+    },
+    {
+      "epoch": 0.9030797867839919,
+      "grad_norm": 0.14500996470451355,
+      "learning_rate": 0.001,
+      "loss": 2.0742,
+      "step": 21347
+    },
+    {
+      "epoch": 0.9031220915475082,
+      "grad_norm": 0.19610632956027985,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 21348
+    },
+    {
+      "epoch": 0.9031643963110246,
+      "grad_norm": 0.17215634882450104,
+      "learning_rate": 0.001,
+      "loss": 3.0487,
+      "step": 21349
+    },
+    {
+      "epoch": 0.903206701074541,
+      "grad_norm": 0.20982448756694794,
+      "learning_rate": 0.001,
+      "loss": 2.9082,
+      "step": 21350
+    },
+    {
+      "epoch": 0.9032490058380573,
+      "grad_norm": 4.3952484130859375,
+      "learning_rate": 0.001,
+      "loss": 1.7265,
+      "step": 21351
+    },
+    {
+      "epoch": 0.9032913106015737,
+      "grad_norm": 0.21057921648025513,
+      "learning_rate": 0.001,
+      "loss": 1.9388,
+      "step": 21352
+    },
+    {
+      "epoch": 0.9033336153650902,
+      "grad_norm": 0.17475013434886932,
+      "learning_rate": 0.001,
+      "loss": 2.6006,
+      "step": 21353
+    },
+    {
+      "epoch": 0.9033759201286065,
+      "grad_norm": 4.874578475952148,
+      "learning_rate": 0.001,
+      "loss": 2.2432,
+      "step": 21354
+    },
+    {
+      "epoch": 0.9034182248921229,
+      "grad_norm": 0.16315825283527374,
+      "learning_rate": 0.001,
+      "loss": 1.6954,
+      "step": 21355
+    },
+    {
+      "epoch": 0.9034605296556393,
+      "grad_norm": 0.1910039633512497,
+      "learning_rate": 0.001,
+      "loss": 2.0015,
+      "step": 21356
+    },
+    {
+      "epoch": 0.9035028344191556,
+      "grad_norm": 0.14265620708465576,
+      "learning_rate": 0.001,
+      "loss": 2.2112,
+      "step": 21357
+    },
+    {
+      "epoch": 0.903545139182672,
+      "grad_norm": 0.4426799714565277,
+      "learning_rate": 0.001,
+      "loss": 2.4595,
+      "step": 21358
+    },
+    {
+      "epoch": 0.9035874439461884,
+      "grad_norm": 0.13497935235500336,
+      "learning_rate": 0.001,
+      "loss": 2.0469,
+      "step": 21359
+    },
+    {
+      "epoch": 0.9036297487097047,
+      "grad_norm": 0.2634698152542114,
+      "learning_rate": 0.001,
+      "loss": 1.2785,
+      "step": 21360
+    },
+    {
+      "epoch": 0.9036720534732211,
+      "grad_norm": 7.950343608856201,
+      "learning_rate": 0.001,
+      "loss": 3.839,
+      "step": 21361
+    },
+    {
+      "epoch": 0.9037143582367374,
+      "grad_norm": 0.30899831652641296,
+      "learning_rate": 0.001,
+      "loss": 2.0495,
+      "step": 21362
+    },
+    {
+      "epoch": 0.9037566630002538,
+      "grad_norm": 0.5557863116264343,
+      "learning_rate": 0.001,
+      "loss": 2.8,
+      "step": 21363
+    },
+    {
+      "epoch": 0.9037989677637702,
+      "grad_norm": 0.14372897148132324,
+      "learning_rate": 0.001,
+      "loss": 1.3838,
+      "step": 21364
+    },
+    {
+      "epoch": 0.9038412725272865,
+      "grad_norm": 0.17638888955116272,
+      "learning_rate": 0.001,
+      "loss": 1.7204,
+      "step": 21365
+    },
+    {
+      "epoch": 0.9038835772908029,
+      "grad_norm": 0.5824902653694153,
+      "learning_rate": 0.001,
+      "loss": 2.1462,
+      "step": 21366
+    },
+    {
+      "epoch": 0.9039258820543193,
+      "grad_norm": 0.18031518161296844,
+      "learning_rate": 0.001,
+      "loss": 1.7274,
+      "step": 21367
+    },
+    {
+      "epoch": 0.9039681868178356,
+      "grad_norm": 0.1800137311220169,
+      "learning_rate": 0.001,
+      "loss": 2.7204,
+      "step": 21368
+    },
+    {
+      "epoch": 0.904010491581352,
+      "grad_norm": 1.138705849647522,
+      "learning_rate": 0.001,
+      "loss": 1.8414,
+      "step": 21369
+    },
+    {
+      "epoch": 0.9040527963448685,
+      "grad_norm": 0.164835125207901,
+      "learning_rate": 0.001,
+      "loss": 2.7179,
+      "step": 21370
+    },
+    {
+      "epoch": 0.9040951011083848,
+      "grad_norm": 0.15999598801136017,
+      "learning_rate": 0.001,
+      "loss": 2.2956,
+      "step": 21371
+    },
+    {
+      "epoch": 0.9041374058719012,
+      "grad_norm": 0.17977473139762878,
+      "learning_rate": 0.001,
+      "loss": 3.2465,
+      "step": 21372
+    },
+    {
+      "epoch": 0.9041797106354176,
+      "grad_norm": 0.15177109837532043,
+      "learning_rate": 0.001,
+      "loss": 1.9558,
+      "step": 21373
+    },
+    {
+      "epoch": 0.9042220153989339,
+      "grad_norm": 0.9045629501342773,
+      "learning_rate": 0.001,
+      "loss": 2.9101,
+      "step": 21374
+    },
+    {
+      "epoch": 0.9042643201624503,
+      "grad_norm": 0.15818485617637634,
+      "learning_rate": 0.001,
+      "loss": 2.2249,
+      "step": 21375
+    },
+    {
+      "epoch": 0.9043066249259667,
+      "grad_norm": 0.16541606187820435,
+      "learning_rate": 0.001,
+      "loss": 2.3738,
+      "step": 21376
+    },
+    {
+      "epoch": 0.904348929689483,
+      "grad_norm": 0.9918728470802307,
+      "learning_rate": 0.001,
+      "loss": 2.5462,
+      "step": 21377
+    },
+    {
+      "epoch": 0.9043912344529994,
+      "grad_norm": 0.12892168760299683,
+      "learning_rate": 0.001,
+      "loss": 3.0338,
+      "step": 21378
+    },
+    {
+      "epoch": 0.9044335392165158,
+      "grad_norm": 0.1821994036436081,
+      "learning_rate": 0.001,
+      "loss": 2.3973,
+      "step": 21379
+    },
+    {
+      "epoch": 0.9044758439800321,
+      "grad_norm": 0.3445868492126465,
+      "learning_rate": 0.001,
+      "loss": 2.9188,
+      "step": 21380
+    },
+    {
+      "epoch": 0.9045181487435485,
+      "grad_norm": 0.14604288339614868,
+      "learning_rate": 0.001,
+      "loss": 2.1574,
+      "step": 21381
+    },
+    {
+      "epoch": 0.9045604535070649,
+      "grad_norm": 0.14212435483932495,
+      "learning_rate": 0.001,
+      "loss": 2.1932,
+      "step": 21382
+    },
+    {
+      "epoch": 0.9046027582705812,
+      "grad_norm": 0.12347246706485748,
+      "learning_rate": 0.001,
+      "loss": 1.7564,
+      "step": 21383
+    },
+    {
+      "epoch": 0.9046450630340976,
+      "grad_norm": 0.3492525517940521,
+      "learning_rate": 0.001,
+      "loss": 1.7874,
+      "step": 21384
+    },
+    {
+      "epoch": 0.904687367797614,
+      "grad_norm": 0.1499442309141159,
+      "learning_rate": 0.001,
+      "loss": 2.3026,
+      "step": 21385
+    },
+    {
+      "epoch": 0.9047296725611303,
+      "grad_norm": 0.15123094618320465,
+      "learning_rate": 0.001,
+      "loss": 2.6157,
+      "step": 21386
+    },
+    {
+      "epoch": 0.9047719773246468,
+      "grad_norm": 0.18827593326568604,
+      "learning_rate": 0.001,
+      "loss": 2.5425,
+      "step": 21387
+    },
+    {
+      "epoch": 0.9048142820881632,
+      "grad_norm": 0.16215500235557556,
+      "learning_rate": 0.001,
+      "loss": 2.1401,
+      "step": 21388
+    },
+    {
+      "epoch": 0.9048565868516795,
+      "grad_norm": 0.15039432048797607,
+      "learning_rate": 0.001,
+      "loss": 2.3611,
+      "step": 21389
+    },
+    {
+      "epoch": 0.9048988916151959,
+      "grad_norm": 0.183966264128685,
+      "learning_rate": 0.001,
+      "loss": 2.4818,
+      "step": 21390
+    },
+    {
+      "epoch": 0.9049411963787123,
+      "grad_norm": 0.1891811490058899,
+      "learning_rate": 0.001,
+      "loss": 3.7626,
+      "step": 21391
+    },
+    {
+      "epoch": 0.9049835011422286,
+      "grad_norm": 0.18254221975803375,
+      "learning_rate": 0.001,
+      "loss": 1.7773,
+      "step": 21392
+    },
+    {
+      "epoch": 0.905025805905745,
+      "grad_norm": 0.14208267629146576,
+      "learning_rate": 0.001,
+      "loss": 1.8959,
+      "step": 21393
+    },
+    {
+      "epoch": 0.9050681106692614,
+      "grad_norm": 0.3248007297515869,
+      "learning_rate": 0.001,
+      "loss": 2.2422,
+      "step": 21394
+    },
+    {
+      "epoch": 0.9051104154327777,
+      "grad_norm": 0.15163081884384155,
+      "learning_rate": 0.001,
+      "loss": 2.1537,
+      "step": 21395
+    },
+    {
+      "epoch": 0.9051527201962941,
+      "grad_norm": 0.16903750598430634,
+      "learning_rate": 0.001,
+      "loss": 1.838,
+      "step": 21396
+    },
+    {
+      "epoch": 0.9051950249598105,
+      "grad_norm": 0.13609051704406738,
+      "learning_rate": 0.001,
+      "loss": 1.3032,
+      "step": 21397
+    },
+    {
+      "epoch": 0.9052373297233268,
+      "grad_norm": 0.21412265300750732,
+      "learning_rate": 0.001,
+      "loss": 2.4583,
+      "step": 21398
+    },
+    {
+      "epoch": 0.9052796344868432,
+      "grad_norm": 0.6192092895507812,
+      "learning_rate": 0.001,
+      "loss": 3.1105,
+      "step": 21399
+    },
+    {
+      "epoch": 0.9053219392503596,
+      "grad_norm": 1.6677172183990479,
+      "learning_rate": 0.001,
+      "loss": 1.8853,
+      "step": 21400
+    },
+    {
+      "epoch": 0.9053642440138759,
+      "grad_norm": 0.5495797395706177,
+      "learning_rate": 0.001,
+      "loss": 2.3331,
+      "step": 21401
+    },
+    {
+      "epoch": 0.9054065487773924,
+      "grad_norm": 0.5030525922775269,
+      "learning_rate": 0.001,
+      "loss": 1.9881,
+      "step": 21402
+    },
+    {
+      "epoch": 0.9054488535409088,
+      "grad_norm": 0.14512759447097778,
+      "learning_rate": 0.001,
+      "loss": 2.2818,
+      "step": 21403
+    },
+    {
+      "epoch": 0.9054911583044251,
+      "grad_norm": 0.12264164537191391,
+      "learning_rate": 0.001,
+      "loss": 1.9594,
+      "step": 21404
+    },
+    {
+      "epoch": 0.9055334630679415,
+      "grad_norm": 0.14045067131519318,
+      "learning_rate": 0.001,
+      "loss": 1.8512,
+      "step": 21405
+    },
+    {
+      "epoch": 0.9055757678314578,
+      "grad_norm": 0.4061392843723297,
+      "learning_rate": 0.001,
+      "loss": 2.3092,
+      "step": 21406
+    },
+    {
+      "epoch": 0.9056180725949742,
+      "grad_norm": 1.0429646968841553,
+      "learning_rate": 0.001,
+      "loss": 1.7809,
+      "step": 21407
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.14219164848327637,
+      "learning_rate": 0.001,
+      "loss": 1.641,
+      "step": 21408
+    },
+    {
+      "epoch": 0.9057026821220069,
+      "grad_norm": 0.14022192358970642,
+      "learning_rate": 0.001,
+      "loss": 2.4478,
+      "step": 21409
+    },
+    {
+      "epoch": 0.9057449868855233,
+      "grad_norm": 0.20621517300605774,
+      "learning_rate": 0.001,
+      "loss": 2.3145,
+      "step": 21410
+    },
+    {
+      "epoch": 0.9057872916490397,
+      "grad_norm": 0.1814170777797699,
+      "learning_rate": 0.001,
+      "loss": 1.846,
+      "step": 21411
+    },
+    {
+      "epoch": 0.905829596412556,
+      "grad_norm": 0.17007854580879211,
+      "learning_rate": 0.001,
+      "loss": 2.2049,
+      "step": 21412
+    },
+    {
+      "epoch": 0.9058719011760724,
+      "grad_norm": 0.16254855692386627,
+      "learning_rate": 0.001,
+      "loss": 2.3874,
+      "step": 21413
+    },
+    {
+      "epoch": 0.9059142059395888,
+      "grad_norm": 0.1488046646118164,
+      "learning_rate": 0.001,
+      "loss": 2.5976,
+      "step": 21414
+    },
+    {
+      "epoch": 0.9059565107031051,
+      "grad_norm": 0.18739381432533264,
+      "learning_rate": 0.001,
+      "loss": 2.8319,
+      "step": 21415
+    },
+    {
+      "epoch": 0.9059988154666215,
+      "grad_norm": 0.3353780210018158,
+      "learning_rate": 0.001,
+      "loss": 2.6556,
+      "step": 21416
+    },
+    {
+      "epoch": 0.9060411202301379,
+      "grad_norm": 0.9559656381607056,
+      "learning_rate": 0.001,
+      "loss": 2.9665,
+      "step": 21417
+    },
+    {
+      "epoch": 0.9060834249936542,
+      "grad_norm": 0.15689292550086975,
+      "learning_rate": 0.001,
+      "loss": 2.0337,
+      "step": 21418
+    },
+    {
+      "epoch": 0.9061257297571707,
+      "grad_norm": 0.13871730864048004,
+      "learning_rate": 0.001,
+      "loss": 1.9963,
+      "step": 21419
+    },
+    {
+      "epoch": 0.9061680345206871,
+      "grad_norm": 0.16745619475841522,
+      "learning_rate": 0.001,
+      "loss": 1.7658,
+      "step": 21420
+    },
+    {
+      "epoch": 0.9062103392842034,
+      "grad_norm": 0.14893504977226257,
+      "learning_rate": 0.001,
+      "loss": 1.7183,
+      "step": 21421
+    },
+    {
+      "epoch": 0.9062526440477198,
+      "grad_norm": 0.18129420280456543,
+      "learning_rate": 0.001,
+      "loss": 1.9251,
+      "step": 21422
+    },
+    {
+      "epoch": 0.9062949488112362,
+      "grad_norm": 0.22704291343688965,
+      "learning_rate": 0.001,
+      "loss": 1.8624,
+      "step": 21423
+    },
+    {
+      "epoch": 0.9063372535747525,
+      "grad_norm": 0.16960926353931427,
+      "learning_rate": 0.001,
+      "loss": 1.7591,
+      "step": 21424
+    },
+    {
+      "epoch": 0.9063795583382689,
+      "grad_norm": 0.1893257051706314,
+      "learning_rate": 0.001,
+      "loss": 1.8019,
+      "step": 21425
+    },
+    {
+      "epoch": 0.9064218631017853,
+      "grad_norm": 8.908495903015137,
+      "learning_rate": 0.001,
+      "loss": 1.3735,
+      "step": 21426
+    },
+    {
+      "epoch": 0.9064641678653016,
+      "grad_norm": 0.16186918318271637,
+      "learning_rate": 0.001,
+      "loss": 1.9056,
+      "step": 21427
+    },
+    {
+      "epoch": 0.906506472628818,
+      "grad_norm": 0.16607151925563812,
+      "learning_rate": 0.001,
+      "loss": 2.8435,
+      "step": 21428
+    },
+    {
+      "epoch": 0.9065487773923344,
+      "grad_norm": 0.34751367568969727,
+      "learning_rate": 0.001,
+      "loss": 1.9356,
+      "step": 21429
+    },
+    {
+      "epoch": 0.9065910821558507,
+      "grad_norm": 0.3750505745410919,
+      "learning_rate": 0.001,
+      "loss": 2.2931,
+      "step": 21430
+    },
+    {
+      "epoch": 0.9066333869193671,
+      "grad_norm": 0.1678369790315628,
+      "learning_rate": 0.001,
+      "loss": 1.6514,
+      "step": 21431
+    },
+    {
+      "epoch": 0.9066756916828835,
+      "grad_norm": 0.5519915223121643,
+      "learning_rate": 0.001,
+      "loss": 2.7002,
+      "step": 21432
+    },
+    {
+      "epoch": 0.9067179964463998,
+      "grad_norm": 0.1600515991449356,
+      "learning_rate": 0.001,
+      "loss": 1.8045,
+      "step": 21433
+    },
+    {
+      "epoch": 0.9067603012099162,
+      "grad_norm": 0.15844613313674927,
+      "learning_rate": 0.001,
+      "loss": 1.8964,
+      "step": 21434
+    },
+    {
+      "epoch": 0.9068026059734327,
+      "grad_norm": 0.1344721019268036,
+      "learning_rate": 0.001,
+      "loss": 1.9177,
+      "step": 21435
+    },
+    {
+      "epoch": 0.906844910736949,
+      "grad_norm": 0.16520515084266663,
+      "learning_rate": 0.001,
+      "loss": 2.6828,
+      "step": 21436
+    },
+    {
+      "epoch": 0.9068872155004654,
+      "grad_norm": 0.15085281431674957,
+      "learning_rate": 0.001,
+      "loss": 1.7172,
+      "step": 21437
+    },
+    {
+      "epoch": 0.9069295202639818,
+      "grad_norm": 0.23470507562160492,
+      "learning_rate": 0.001,
+      "loss": 2.8068,
+      "step": 21438
+    },
+    {
+      "epoch": 0.9069718250274981,
+      "grad_norm": 0.15734681487083435,
+      "learning_rate": 0.001,
+      "loss": 2.1509,
+      "step": 21439
+    },
+    {
+      "epoch": 0.9070141297910145,
+      "grad_norm": 0.19993236660957336,
+      "learning_rate": 0.001,
+      "loss": 1.9934,
+      "step": 21440
+    },
+    {
+      "epoch": 0.9070564345545309,
+      "grad_norm": 0.1488742232322693,
+      "learning_rate": 0.001,
+      "loss": 2.1611,
+      "step": 21441
+    },
+    {
+      "epoch": 0.9070987393180472,
+      "grad_norm": 0.15275177359580994,
+      "learning_rate": 0.001,
+      "loss": 1.9241,
+      "step": 21442
+    },
+    {
+      "epoch": 0.9071410440815636,
+      "grad_norm": 0.1733865737915039,
+      "learning_rate": 0.001,
+      "loss": 1.6173,
+      "step": 21443
+    },
+    {
+      "epoch": 0.90718334884508,
+      "grad_norm": 43.84885025024414,
+      "learning_rate": 0.001,
+      "loss": 3.4629,
+      "step": 21444
+    },
+    {
+      "epoch": 0.9072256536085963,
+      "grad_norm": 0.14967910945415497,
+      "learning_rate": 0.001,
+      "loss": 1.882,
+      "step": 21445
+    },
+    {
+      "epoch": 0.9072679583721127,
+      "grad_norm": 0.1916898787021637,
+      "learning_rate": 0.001,
+      "loss": 1.8736,
+      "step": 21446
+    },
+    {
+      "epoch": 0.9073102631356291,
+      "grad_norm": 3.939188003540039,
+      "learning_rate": 0.001,
+      "loss": 3.2875,
+      "step": 21447
+    },
+    {
+      "epoch": 0.9073525678991454,
+      "grad_norm": 0.16006742417812347,
+      "learning_rate": 0.001,
+      "loss": 1.9604,
+      "step": 21448
+    },
+    {
+      "epoch": 0.9073948726626618,
+      "grad_norm": 0.20729239284992218,
+      "learning_rate": 0.001,
+      "loss": 2.4355,
+      "step": 21449
+    },
+    {
+      "epoch": 0.9074371774261782,
+      "grad_norm": 0.15743091702461243,
+      "learning_rate": 0.001,
+      "loss": 2.3285,
+      "step": 21450
+    },
+    {
+      "epoch": 0.9074794821896945,
+      "grad_norm": 0.15352989733219147,
+      "learning_rate": 0.001,
+      "loss": 2.2617,
+      "step": 21451
+    },
+    {
+      "epoch": 0.907521786953211,
+      "grad_norm": 1.4606817960739136,
+      "learning_rate": 0.001,
+      "loss": 2.2048,
+      "step": 21452
+    },
+    {
+      "epoch": 0.9075640917167273,
+      "grad_norm": 0.14188838005065918,
+      "learning_rate": 0.001,
+      "loss": 1.9296,
+      "step": 21453
+    },
+    {
+      "epoch": 0.9076063964802437,
+      "grad_norm": 0.1336817890405655,
+      "learning_rate": 0.001,
+      "loss": 1.9424,
+      "step": 21454
+    },
+    {
+      "epoch": 0.9076487012437601,
+      "grad_norm": 1.9917988777160645,
+      "learning_rate": 0.001,
+      "loss": 1.8911,
+      "step": 21455
+    },
+    {
+      "epoch": 0.9076910060072764,
+      "grad_norm": 0.7746376395225525,
+      "learning_rate": 0.001,
+      "loss": 3.08,
+      "step": 21456
+    },
+    {
+      "epoch": 0.9077333107707928,
+      "grad_norm": 0.18907096982002258,
+      "learning_rate": 0.001,
+      "loss": 1.9718,
+      "step": 21457
+    },
+    {
+      "epoch": 0.9077756155343092,
+      "grad_norm": 0.1984204649925232,
+      "learning_rate": 0.001,
+      "loss": 1.9067,
+      "step": 21458
+    },
+    {
+      "epoch": 0.9078179202978255,
+      "grad_norm": 0.13996772468090057,
+      "learning_rate": 0.001,
+      "loss": 4.0638,
+      "step": 21459
+    },
+    {
+      "epoch": 0.9078602250613419,
+      "grad_norm": 0.16142001748085022,
+      "learning_rate": 0.001,
+      "loss": 2.3478,
+      "step": 21460
+    },
+    {
+      "epoch": 0.9079025298248583,
+      "grad_norm": 0.17676426470279694,
+      "learning_rate": 0.001,
+      "loss": 4.1074,
+      "step": 21461
+    },
+    {
+      "epoch": 0.9079448345883746,
+      "grad_norm": 0.1353548765182495,
+      "learning_rate": 0.001,
+      "loss": 2.4199,
+      "step": 21462
+    },
+    {
+      "epoch": 0.907987139351891,
+      "grad_norm": 0.18205760419368744,
+      "learning_rate": 0.001,
+      "loss": 2.2659,
+      "step": 21463
+    },
+    {
+      "epoch": 0.9080294441154074,
+      "grad_norm": 0.16253677010536194,
+      "learning_rate": 0.001,
+      "loss": 2.387,
+      "step": 21464
+    },
+    {
+      "epoch": 0.9080717488789237,
+      "grad_norm": 0.18944591283798218,
+      "learning_rate": 0.001,
+      "loss": 2.3769,
+      "step": 21465
+    },
+    {
+      "epoch": 0.9081140536424401,
+      "grad_norm": 0.1672883927822113,
+      "learning_rate": 0.001,
+      "loss": 2.4218,
+      "step": 21466
+    },
+    {
+      "epoch": 0.9081563584059565,
+      "grad_norm": 0.1499961018562317,
+      "learning_rate": 0.001,
+      "loss": 2.0867,
+      "step": 21467
+    },
+    {
+      "epoch": 0.9081986631694728,
+      "grad_norm": 0.2781917452812195,
+      "learning_rate": 0.001,
+      "loss": 2.1938,
+      "step": 21468
+    },
+    {
+      "epoch": 0.9082409679329893,
+      "grad_norm": 0.2037796825170517,
+      "learning_rate": 0.001,
+      "loss": 3.4381,
+      "step": 21469
+    },
+    {
+      "epoch": 0.9082832726965057,
+      "grad_norm": 0.2747834622859955,
+      "learning_rate": 0.001,
+      "loss": 2.2192,
+      "step": 21470
+    },
+    {
+      "epoch": 0.908325577460022,
+      "grad_norm": 0.128767728805542,
+      "learning_rate": 0.001,
+      "loss": 1.6764,
+      "step": 21471
+    },
+    {
+      "epoch": 0.9083678822235384,
+      "grad_norm": 0.13520826399326324,
+      "learning_rate": 0.001,
+      "loss": 1.7189,
+      "step": 21472
+    },
+    {
+      "epoch": 0.9084101869870548,
+      "grad_norm": 0.16434721648693085,
+      "learning_rate": 0.001,
+      "loss": 1.1854,
+      "step": 21473
+    },
+    {
+      "epoch": 0.9084524917505711,
+      "grad_norm": 0.7008240818977356,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 21474
+    },
+    {
+      "epoch": 0.9084947965140875,
+      "grad_norm": 0.1205538809299469,
+      "learning_rate": 0.001,
+      "loss": 2.7022,
+      "step": 21475
+    },
+    {
+      "epoch": 0.9085371012776039,
+      "grad_norm": 0.17018336057662964,
+      "learning_rate": 0.001,
+      "loss": 2.3906,
+      "step": 21476
+    },
+    {
+      "epoch": 0.9085794060411202,
+      "grad_norm": 0.2808079421520233,
+      "learning_rate": 0.001,
+      "loss": 3.2353,
+      "step": 21477
+    },
+    {
+      "epoch": 0.9086217108046366,
+      "grad_norm": 0.18852543830871582,
+      "learning_rate": 0.001,
+      "loss": 2.035,
+      "step": 21478
+    },
+    {
+      "epoch": 0.908664015568153,
+      "grad_norm": 0.12925562262535095,
+      "learning_rate": 0.001,
+      "loss": 1.7236,
+      "step": 21479
+    },
+    {
+      "epoch": 0.9087063203316693,
+      "grad_norm": 4.139763832092285,
+      "learning_rate": 0.001,
+      "loss": 1.6331,
+      "step": 21480
+    },
+    {
+      "epoch": 0.9087486250951857,
+      "grad_norm": 0.15899895131587982,
+      "learning_rate": 0.001,
+      "loss": 1.7789,
+      "step": 21481
+    },
+    {
+      "epoch": 0.9087909298587021,
+      "grad_norm": 0.1730024665594101,
+      "learning_rate": 0.001,
+      "loss": 3.3566,
+      "step": 21482
+    },
+    {
+      "epoch": 0.9088332346222184,
+      "grad_norm": 0.1567884385585785,
+      "learning_rate": 0.001,
+      "loss": 1.9054,
+      "step": 21483
+    },
+    {
+      "epoch": 0.9088755393857348,
+      "grad_norm": 2.438006639480591,
+      "learning_rate": 0.001,
+      "loss": 2.6868,
+      "step": 21484
+    },
+    {
+      "epoch": 0.9089178441492513,
+      "grad_norm": 0.5900127291679382,
+      "learning_rate": 0.001,
+      "loss": 2.5526,
+      "step": 21485
+    },
+    {
+      "epoch": 0.9089601489127676,
+      "grad_norm": 0.7061132788658142,
+      "learning_rate": 0.001,
+      "loss": 2.5784,
+      "step": 21486
+    },
+    {
+      "epoch": 0.909002453676284,
+      "grad_norm": 0.7392309308052063,
+      "learning_rate": 0.001,
+      "loss": 2.1671,
+      "step": 21487
+    },
+    {
+      "epoch": 0.9090447584398004,
+      "grad_norm": 0.18610547482967377,
+      "learning_rate": 0.001,
+      "loss": 2.4229,
+      "step": 21488
+    },
+    {
+      "epoch": 0.9090870632033167,
+      "grad_norm": 0.15791387856006622,
+      "learning_rate": 0.001,
+      "loss": 2.6113,
+      "step": 21489
+    },
+    {
+      "epoch": 0.9091293679668331,
+      "grad_norm": 13.427140235900879,
+      "learning_rate": 0.001,
+      "loss": 1.7686,
+      "step": 21490
+    },
+    {
+      "epoch": 0.9091716727303495,
+      "grad_norm": 0.1820589154958725,
+      "learning_rate": 0.001,
+      "loss": 2.2658,
+      "step": 21491
+    },
+    {
+      "epoch": 0.9092139774938658,
+      "grad_norm": 0.5405210256576538,
+      "learning_rate": 0.001,
+      "loss": 2.6741,
+      "step": 21492
+    },
+    {
+      "epoch": 0.9092562822573822,
+      "grad_norm": 0.9230998754501343,
+      "learning_rate": 0.001,
+      "loss": 2.0778,
+      "step": 21493
+    },
+    {
+      "epoch": 0.9092985870208986,
+      "grad_norm": 0.1825629621744156,
+      "learning_rate": 0.001,
+      "loss": 2.0666,
+      "step": 21494
+    },
+    {
+      "epoch": 0.9093408917844149,
+      "grad_norm": 0.21903732419013977,
+      "learning_rate": 0.001,
+      "loss": 1.5503,
+      "step": 21495
+    },
+    {
+      "epoch": 0.9093831965479313,
+      "grad_norm": 0.30625858902931213,
+      "learning_rate": 0.001,
+      "loss": 1.7506,
+      "step": 21496
+    },
+    {
+      "epoch": 0.9094255013114476,
+      "grad_norm": 0.23304063081741333,
+      "learning_rate": 0.001,
+      "loss": 2.6598,
+      "step": 21497
+    },
+    {
+      "epoch": 0.909467806074964,
+      "grad_norm": 0.18999595940113068,
+      "learning_rate": 0.001,
+      "loss": 1.8506,
+      "step": 21498
+    },
+    {
+      "epoch": 0.9095101108384804,
+      "grad_norm": 3.7173240184783936,
+      "learning_rate": 0.001,
+      "loss": 2.9739,
+      "step": 21499
+    },
+    {
+      "epoch": 0.9095524156019967,
+      "grad_norm": 0.33216527104377747,
+      "learning_rate": 0.001,
+      "loss": 1.2953,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9095947203655131,
+      "grad_norm": 0.1976582258939743,
+      "learning_rate": 0.001,
+      "loss": 2.2576,
+      "step": 21501
+    },
+    {
+      "epoch": 0.9096370251290296,
+      "grad_norm": 5.519730091094971,
+      "learning_rate": 0.001,
+      "loss": 1.8796,
+      "step": 21502
+    },
+    {
+      "epoch": 0.9096793298925459,
+      "grad_norm": 0.31326529383659363,
+      "learning_rate": 0.001,
+      "loss": 2.1128,
+      "step": 21503
+    },
+    {
+      "epoch": 0.9097216346560623,
+      "grad_norm": 0.15805281698703766,
+      "learning_rate": 0.001,
+      "loss": 1.3282,
+      "step": 21504
+    },
+    {
+      "epoch": 0.9097639394195787,
+      "grad_norm": 1.21946120262146,
+      "learning_rate": 0.001,
+      "loss": 2.203,
+      "step": 21505
+    },
+    {
+      "epoch": 0.909806244183095,
+      "grad_norm": 0.13598865270614624,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 21506
+    },
+    {
+      "epoch": 0.9098485489466114,
+      "grad_norm": 0.30996546149253845,
+      "learning_rate": 0.001,
+      "loss": 1.5201,
+      "step": 21507
+    },
+    {
+      "epoch": 0.9098908537101278,
+      "grad_norm": 0.21078462898731232,
+      "learning_rate": 0.001,
+      "loss": 2.2036,
+      "step": 21508
+    },
+    {
+      "epoch": 0.9099331584736441,
+      "grad_norm": 0.1793532818555832,
+      "learning_rate": 0.001,
+      "loss": 2.754,
+      "step": 21509
+    },
+    {
+      "epoch": 0.9099754632371605,
+      "grad_norm": 0.16735830903053284,
+      "learning_rate": 0.001,
+      "loss": 3.22,
+      "step": 21510
+    },
+    {
+      "epoch": 0.9100177680006769,
+      "grad_norm": 0.15920020639896393,
+      "learning_rate": 0.001,
+      "loss": 2.8717,
+      "step": 21511
+    },
+    {
+      "epoch": 0.9100600727641932,
+      "grad_norm": 0.17090164124965668,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 21512
+    },
+    {
+      "epoch": 0.9101023775277096,
+      "grad_norm": 0.1805136799812317,
+      "learning_rate": 0.001,
+      "loss": 1.7801,
+      "step": 21513
+    },
+    {
+      "epoch": 0.910144682291226,
+      "grad_norm": 0.20294389128684998,
+      "learning_rate": 0.001,
+      "loss": 2.2071,
+      "step": 21514
+    },
+    {
+      "epoch": 0.9101869870547423,
+      "grad_norm": 0.18520846962928772,
+      "learning_rate": 0.001,
+      "loss": 1.3472,
+      "step": 21515
+    },
+    {
+      "epoch": 0.9102292918182587,
+      "grad_norm": 0.8986330032348633,
+      "learning_rate": 0.001,
+      "loss": 2.8275,
+      "step": 21516
+    },
+    {
+      "epoch": 0.9102715965817751,
+      "grad_norm": 0.16281363368034363,
+      "learning_rate": 0.001,
+      "loss": 3.0088,
+      "step": 21517
+    },
+    {
+      "epoch": 0.9103139013452914,
+      "grad_norm": 0.13993607461452484,
+      "learning_rate": 0.001,
+      "loss": 2.1769,
+      "step": 21518
+    },
+    {
+      "epoch": 0.9103562061088079,
+      "grad_norm": 0.13132378458976746,
+      "learning_rate": 0.001,
+      "loss": 1.4656,
+      "step": 21519
+    },
+    {
+      "epoch": 0.9103985108723243,
+      "grad_norm": 0.1520097702741623,
+      "learning_rate": 0.001,
+      "loss": 1.8638,
+      "step": 21520
+    },
+    {
+      "epoch": 0.9104408156358406,
+      "grad_norm": 1.1580663919448853,
+      "learning_rate": 0.001,
+      "loss": 2.0435,
+      "step": 21521
+    },
+    {
+      "epoch": 0.910483120399357,
+      "grad_norm": 0.1821383833885193,
+      "learning_rate": 0.001,
+      "loss": 2.5283,
+      "step": 21522
+    },
+    {
+      "epoch": 0.9105254251628734,
+      "grad_norm": 0.1639668494462967,
+      "learning_rate": 0.001,
+      "loss": 2.0908,
+      "step": 21523
+    },
+    {
+      "epoch": 0.9105677299263897,
+      "grad_norm": 0.18007436394691467,
+      "learning_rate": 0.001,
+      "loss": 1.8416,
+      "step": 21524
+    },
+    {
+      "epoch": 0.9106100346899061,
+      "grad_norm": 0.13596655428409576,
+      "learning_rate": 0.001,
+      "loss": 1.352,
+      "step": 21525
+    },
+    {
+      "epoch": 0.9106523394534225,
+      "grad_norm": 0.19696587324142456,
+      "learning_rate": 0.001,
+      "loss": 1.9196,
+      "step": 21526
+    },
+    {
+      "epoch": 0.9106946442169388,
+      "grad_norm": 0.15187956392765045,
+      "learning_rate": 0.001,
+      "loss": 2.5806,
+      "step": 21527
+    },
+    {
+      "epoch": 0.9107369489804552,
+      "grad_norm": 0.15384723246097565,
+      "learning_rate": 0.001,
+      "loss": 3.9416,
+      "step": 21528
+    },
+    {
+      "epoch": 0.9107792537439716,
+      "grad_norm": 0.21854868531227112,
+      "learning_rate": 0.001,
+      "loss": 2.478,
+      "step": 21529
+    },
+    {
+      "epoch": 0.9108215585074879,
+      "grad_norm": 0.1801101565361023,
+      "learning_rate": 0.001,
+      "loss": 2.2489,
+      "step": 21530
+    },
+    {
+      "epoch": 0.9108638632710043,
+      "grad_norm": 0.19910669326782227,
+      "learning_rate": 0.001,
+      "loss": 1.7515,
+      "step": 21531
+    },
+    {
+      "epoch": 0.9109061680345207,
+      "grad_norm": 0.18891854584217072,
+      "learning_rate": 0.001,
+      "loss": 1.8261,
+      "step": 21532
+    },
+    {
+      "epoch": 0.910948472798037,
+      "grad_norm": 0.1557842195034027,
+      "learning_rate": 0.001,
+      "loss": 2.175,
+      "step": 21533
+    },
+    {
+      "epoch": 0.9109907775615534,
+      "grad_norm": 0.15806511044502258,
+      "learning_rate": 0.001,
+      "loss": 1.7377,
+      "step": 21534
+    },
+    {
+      "epoch": 0.9110330823250699,
+      "grad_norm": 5.196726322174072,
+      "learning_rate": 0.001,
+      "loss": 2.0587,
+      "step": 21535
+    },
+    {
+      "epoch": 0.9110753870885862,
+      "grad_norm": 0.15925204753875732,
+      "learning_rate": 0.001,
+      "loss": 3.1414,
+      "step": 21536
+    },
+    {
+      "epoch": 0.9111176918521026,
+      "grad_norm": 0.16708320379257202,
+      "learning_rate": 0.001,
+      "loss": 2.3862,
+      "step": 21537
+    },
+    {
+      "epoch": 0.911159996615619,
+      "grad_norm": 0.14083898067474365,
+      "learning_rate": 0.001,
+      "loss": 1.5379,
+      "step": 21538
+    },
+    {
+      "epoch": 0.9112023013791353,
+      "grad_norm": 2.581650733947754,
+      "learning_rate": 0.001,
+      "loss": 2.6506,
+      "step": 21539
+    },
+    {
+      "epoch": 0.9112446061426517,
+      "grad_norm": 0.15130378305912018,
+      "learning_rate": 0.001,
+      "loss": 2.7111,
+      "step": 21540
+    },
+    {
+      "epoch": 0.911286910906168,
+      "grad_norm": 0.172215536236763,
+      "learning_rate": 0.001,
+      "loss": 1.4651,
+      "step": 21541
+    },
+    {
+      "epoch": 0.9113292156696844,
+      "grad_norm": 0.2063896358013153,
+      "learning_rate": 0.001,
+      "loss": 1.5084,
+      "step": 21542
+    },
+    {
+      "epoch": 0.9113715204332008,
+      "grad_norm": 0.48414063453674316,
+      "learning_rate": 0.001,
+      "loss": 3.9561,
+      "step": 21543
+    },
+    {
+      "epoch": 0.9114138251967171,
+      "grad_norm": 0.31004536151885986,
+      "learning_rate": 0.001,
+      "loss": 2.4255,
+      "step": 21544
+    },
+    {
+      "epoch": 0.9114561299602335,
+      "grad_norm": 0.17649905383586884,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 21545
+    },
+    {
+      "epoch": 0.9114984347237499,
+      "grad_norm": 0.4742259085178375,
+      "learning_rate": 0.001,
+      "loss": 3.0758,
+      "step": 21546
+    },
+    {
+      "epoch": 0.9115407394872662,
+      "grad_norm": 0.21102559566497803,
+      "learning_rate": 0.001,
+      "loss": 2.8004,
+      "step": 21547
+    },
+    {
+      "epoch": 0.9115830442507826,
+      "grad_norm": 0.47817152738571167,
+      "learning_rate": 0.001,
+      "loss": 2.0812,
+      "step": 21548
+    },
+    {
+      "epoch": 0.911625349014299,
+      "grad_norm": 0.156982883810997,
+      "learning_rate": 0.001,
+      "loss": 2.3945,
+      "step": 21549
+    },
+    {
+      "epoch": 0.9116676537778153,
+      "grad_norm": 0.14067421853542328,
+      "learning_rate": 0.001,
+      "loss": 2.0697,
+      "step": 21550
+    },
+    {
+      "epoch": 0.9117099585413317,
+      "grad_norm": 0.17669621109962463,
+      "learning_rate": 0.001,
+      "loss": 2.3237,
+      "step": 21551
+    },
+    {
+      "epoch": 0.9117522633048482,
+      "grad_norm": 0.16611798107624054,
+      "learning_rate": 0.001,
+      "loss": 1.5483,
+      "step": 21552
+    },
+    {
+      "epoch": 0.9117945680683645,
+      "grad_norm": 0.2230168581008911,
+      "learning_rate": 0.001,
+      "loss": 2.6282,
+      "step": 21553
+    },
+    {
+      "epoch": 0.9118368728318809,
+      "grad_norm": 4.087730884552002,
+      "learning_rate": 0.001,
+      "loss": 1.8573,
+      "step": 21554
+    },
+    {
+      "epoch": 0.9118791775953973,
+      "grad_norm": 0.1587943285703659,
+      "learning_rate": 0.001,
+      "loss": 3.0096,
+      "step": 21555
+    },
+    {
+      "epoch": 0.9119214823589136,
+      "grad_norm": 0.13796477019786835,
+      "learning_rate": 0.001,
+      "loss": 2.5867,
+      "step": 21556
+    },
+    {
+      "epoch": 0.91196378712243,
+      "grad_norm": 0.17402009665966034,
+      "learning_rate": 0.001,
+      "loss": 2.2494,
+      "step": 21557
+    },
+    {
+      "epoch": 0.9120060918859464,
+      "grad_norm": 0.20480291545391083,
+      "learning_rate": 0.001,
+      "loss": 2.7589,
+      "step": 21558
+    },
+    {
+      "epoch": 0.9120483966494627,
+      "grad_norm": 0.45675426721572876,
+      "learning_rate": 0.001,
+      "loss": 2.9796,
+      "step": 21559
+    },
+    {
+      "epoch": 0.9120907014129791,
+      "grad_norm": 0.26858261227607727,
+      "learning_rate": 0.001,
+      "loss": 2.7131,
+      "step": 21560
+    },
+    {
+      "epoch": 0.9121330061764955,
+      "grad_norm": 0.1663147658109665,
+      "learning_rate": 0.001,
+      "loss": 1.9666,
+      "step": 21561
+    },
+    {
+      "epoch": 0.9121753109400118,
+      "grad_norm": 0.16249950230121613,
+      "learning_rate": 0.001,
+      "loss": 2.6662,
+      "step": 21562
+    },
+    {
+      "epoch": 0.9122176157035282,
+      "grad_norm": 0.22289776802062988,
+      "learning_rate": 0.001,
+      "loss": 2.3533,
+      "step": 21563
+    },
+    {
+      "epoch": 0.9122599204670446,
+      "grad_norm": 0.20994803309440613,
+      "learning_rate": 0.001,
+      "loss": 1.978,
+      "step": 21564
+    },
+    {
+      "epoch": 0.9123022252305609,
+      "grad_norm": 0.4022495448589325,
+      "learning_rate": 0.001,
+      "loss": 2.1195,
+      "step": 21565
+    },
+    {
+      "epoch": 0.9123445299940773,
+      "grad_norm": 0.17992646992206573,
+      "learning_rate": 0.001,
+      "loss": 1.934,
+      "step": 21566
+    },
+    {
+      "epoch": 0.9123868347575937,
+      "grad_norm": 9.34231948852539,
+      "learning_rate": 0.001,
+      "loss": 2.3568,
+      "step": 21567
+    },
+    {
+      "epoch": 0.91242913952111,
+      "grad_norm": 0.19124965369701385,
+      "learning_rate": 0.001,
+      "loss": 1.8253,
+      "step": 21568
+    },
+    {
+      "epoch": 0.9124714442846265,
+      "grad_norm": 0.1695529967546463,
+      "learning_rate": 0.001,
+      "loss": 2.892,
+      "step": 21569
+    },
+    {
+      "epoch": 0.9125137490481429,
+      "grad_norm": 0.1486518681049347,
+      "learning_rate": 0.001,
+      "loss": 1.868,
+      "step": 21570
+    },
+    {
+      "epoch": 0.9125560538116592,
+      "grad_norm": 20.817100524902344,
+      "learning_rate": 0.001,
+      "loss": 1.8457,
+      "step": 21571
+    },
+    {
+      "epoch": 0.9125983585751756,
+      "grad_norm": 12.602542877197266,
+      "learning_rate": 0.001,
+      "loss": 2.1331,
+      "step": 21572
+    },
+    {
+      "epoch": 0.912640663338692,
+      "grad_norm": 0.16852043569087982,
+      "learning_rate": 0.001,
+      "loss": 2.6505,
+      "step": 21573
+    },
+    {
+      "epoch": 0.9126829681022083,
+      "grad_norm": 0.17694880068302155,
+      "learning_rate": 0.001,
+      "loss": 1.5316,
+      "step": 21574
+    },
+    {
+      "epoch": 0.9127252728657247,
+      "grad_norm": 0.19821469485759735,
+      "learning_rate": 0.001,
+      "loss": 2.0534,
+      "step": 21575
+    },
+    {
+      "epoch": 0.9127675776292411,
+      "grad_norm": 0.21617907285690308,
+      "learning_rate": 0.001,
+      "loss": 2.5649,
+      "step": 21576
+    },
+    {
+      "epoch": 0.9128098823927574,
+      "grad_norm": 0.17548352479934692,
+      "learning_rate": 0.001,
+      "loss": 1.9663,
+      "step": 21577
+    },
+    {
+      "epoch": 0.9128521871562738,
+      "grad_norm": 0.22286243736743927,
+      "learning_rate": 0.001,
+      "loss": 1.97,
+      "step": 21578
+    },
+    {
+      "epoch": 0.9128944919197902,
+      "grad_norm": 0.19249877333641052,
+      "learning_rate": 0.001,
+      "loss": 2.8943,
+      "step": 21579
+    },
+    {
+      "epoch": 0.9129367966833065,
+      "grad_norm": 0.14703737199306488,
+      "learning_rate": 0.001,
+      "loss": 2.2766,
+      "step": 21580
+    },
+    {
+      "epoch": 0.9129791014468229,
+      "grad_norm": 0.17037394642829895,
+      "learning_rate": 0.001,
+      "loss": 1.5034,
+      "step": 21581
+    },
+    {
+      "epoch": 0.9130214062103393,
+      "grad_norm": 0.18207064270973206,
+      "learning_rate": 0.001,
+      "loss": 2.1486,
+      "step": 21582
+    },
+    {
+      "epoch": 0.9130637109738556,
+      "grad_norm": 0.1920434832572937,
+      "learning_rate": 0.001,
+      "loss": 1.6474,
+      "step": 21583
+    },
+    {
+      "epoch": 0.913106015737372,
+      "grad_norm": 1.184415578842163,
+      "learning_rate": 0.001,
+      "loss": 3.1735,
+      "step": 21584
+    },
+    {
+      "epoch": 0.9131483205008885,
+      "grad_norm": 0.15448032319545746,
+      "learning_rate": 0.001,
+      "loss": 1.736,
+      "step": 21585
+    },
+    {
+      "epoch": 0.9131906252644048,
+      "grad_norm": 0.18459028005599976,
+      "learning_rate": 0.001,
+      "loss": 2.0065,
+      "step": 21586
+    },
+    {
+      "epoch": 0.9132329300279212,
+      "grad_norm": 0.14690928161144257,
+      "learning_rate": 0.001,
+      "loss": 2.3362,
+      "step": 21587
+    },
+    {
+      "epoch": 0.9132752347914375,
+      "grad_norm": 0.15525342524051666,
+      "learning_rate": 0.001,
+      "loss": 1.691,
+      "step": 21588
+    },
+    {
+      "epoch": 0.9133175395549539,
+      "grad_norm": 0.15081803500652313,
+      "learning_rate": 0.001,
+      "loss": 1.3852,
+      "step": 21589
+    },
+    {
+      "epoch": 0.9133598443184703,
+      "grad_norm": 0.14801929891109467,
+      "learning_rate": 0.001,
+      "loss": 2.3841,
+      "step": 21590
+    },
+    {
+      "epoch": 0.9134021490819866,
+      "grad_norm": 0.1725027710199356,
+      "learning_rate": 0.001,
+      "loss": 2.52,
+      "step": 21591
+    },
+    {
+      "epoch": 0.913444453845503,
+      "grad_norm": 0.11509247869253159,
+      "learning_rate": 0.001,
+      "loss": 1.6571,
+      "step": 21592
+    },
+    {
+      "epoch": 0.9134867586090194,
+      "grad_norm": 0.14290519058704376,
+      "learning_rate": 0.001,
+      "loss": 1.8487,
+      "step": 21593
+    },
+    {
+      "epoch": 0.9135290633725357,
+      "grad_norm": 0.21351279318332672,
+      "learning_rate": 0.001,
+      "loss": 1.9015,
+      "step": 21594
+    },
+    {
+      "epoch": 0.9135713681360521,
+      "grad_norm": 0.391020268201828,
+      "learning_rate": 0.001,
+      "loss": 2.0749,
+      "step": 21595
+    },
+    {
+      "epoch": 0.9136136728995685,
+      "grad_norm": 0.14413462579250336,
+      "learning_rate": 0.001,
+      "loss": 1.8242,
+      "step": 21596
+    },
+    {
+      "epoch": 0.9136559776630848,
+      "grad_norm": 0.18715713918209076,
+      "learning_rate": 0.001,
+      "loss": 3.9564,
+      "step": 21597
+    },
+    {
+      "epoch": 0.9136982824266012,
+      "grad_norm": 0.17989511787891388,
+      "learning_rate": 0.001,
+      "loss": 2.6794,
+      "step": 21598
+    },
+    {
+      "epoch": 0.9137405871901176,
+      "grad_norm": 0.1721455156803131,
+      "learning_rate": 0.001,
+      "loss": 1.6095,
+      "step": 21599
+    },
+    {
+      "epoch": 0.9137828919536339,
+      "grad_norm": 1.7878568172454834,
+      "learning_rate": 0.001,
+      "loss": 2.6891,
+      "step": 21600
+    },
+    {
+      "epoch": 0.9138251967171503,
+      "grad_norm": 0.14471499621868134,
+      "learning_rate": 0.001,
+      "loss": 1.3701,
+      "step": 21601
+    },
+    {
+      "epoch": 0.9138675014806668,
+      "grad_norm": 2.2740018367767334,
+      "learning_rate": 0.001,
+      "loss": 2.1687,
+      "step": 21602
+    },
+    {
+      "epoch": 0.913909806244183,
+      "grad_norm": 0.13715597987174988,
+      "learning_rate": 0.001,
+      "loss": 1.995,
+      "step": 21603
+    },
+    {
+      "epoch": 0.9139521110076995,
+      "grad_norm": 0.13513702154159546,
+      "learning_rate": 0.001,
+      "loss": 1.735,
+      "step": 21604
+    },
+    {
+      "epoch": 0.9139944157712159,
+      "grad_norm": 0.1583508551120758,
+      "learning_rate": 0.001,
+      "loss": 1.746,
+      "step": 21605
+    },
+    {
+      "epoch": 0.9140367205347322,
+      "grad_norm": 0.1774708777666092,
+      "learning_rate": 0.001,
+      "loss": 2.3999,
+      "step": 21606
+    },
+    {
+      "epoch": 0.9140790252982486,
+      "grad_norm": 0.15882831811904907,
+      "learning_rate": 0.001,
+      "loss": 1.6509,
+      "step": 21607
+    },
+    {
+      "epoch": 0.914121330061765,
+      "grad_norm": 0.1779770851135254,
+      "learning_rate": 0.001,
+      "loss": 2.1135,
+      "step": 21608
+    },
+    {
+      "epoch": 0.9141636348252813,
+      "grad_norm": 0.33798548579216003,
+      "learning_rate": 0.001,
+      "loss": 2.7735,
+      "step": 21609
+    },
+    {
+      "epoch": 0.9142059395887977,
+      "grad_norm": 0.31764113903045654,
+      "learning_rate": 0.001,
+      "loss": 2.0155,
+      "step": 21610
+    },
+    {
+      "epoch": 0.9142482443523141,
+      "grad_norm": 0.17167538404464722,
+      "learning_rate": 0.001,
+      "loss": 1.7689,
+      "step": 21611
+    },
+    {
+      "epoch": 0.9142905491158304,
+      "grad_norm": 0.16581086814403534,
+      "learning_rate": 0.001,
+      "loss": 1.7205,
+      "step": 21612
+    },
+    {
+      "epoch": 0.9143328538793468,
+      "grad_norm": 0.16308166086673737,
+      "learning_rate": 0.001,
+      "loss": 2.3171,
+      "step": 21613
+    },
+    {
+      "epoch": 0.9143751586428632,
+      "grad_norm": 0.1509043574333191,
+      "learning_rate": 0.001,
+      "loss": 1.4007,
+      "step": 21614
+    },
+    {
+      "epoch": 0.9144174634063795,
+      "grad_norm": 0.18043819069862366,
+      "learning_rate": 0.001,
+      "loss": 3.3869,
+      "step": 21615
+    },
+    {
+      "epoch": 0.9144597681698959,
+      "grad_norm": 0.19052763283252716,
+      "learning_rate": 0.001,
+      "loss": 2.2411,
+      "step": 21616
+    },
+    {
+      "epoch": 0.9145020729334123,
+      "grad_norm": 0.13170795142650604,
+      "learning_rate": 0.001,
+      "loss": 2.1545,
+      "step": 21617
+    },
+    {
+      "epoch": 0.9145443776969286,
+      "grad_norm": 0.16418077051639557,
+      "learning_rate": 0.001,
+      "loss": 2.2063,
+      "step": 21618
+    },
+    {
+      "epoch": 0.914586682460445,
+      "grad_norm": 0.5688857436180115,
+      "learning_rate": 0.001,
+      "loss": 2.1972,
+      "step": 21619
+    },
+    {
+      "epoch": 0.9146289872239615,
+      "grad_norm": 0.1831405758857727,
+      "learning_rate": 0.001,
+      "loss": 1.6587,
+      "step": 21620
+    },
+    {
+      "epoch": 0.9146712919874778,
+      "grad_norm": 0.4554640054702759,
+      "learning_rate": 0.001,
+      "loss": 1.7407,
+      "step": 21621
+    },
+    {
+      "epoch": 0.9147135967509942,
+      "grad_norm": 0.17047862708568573,
+      "learning_rate": 0.001,
+      "loss": 1.6915,
+      "step": 21622
+    },
+    {
+      "epoch": 0.9147559015145106,
+      "grad_norm": 0.7939593195915222,
+      "learning_rate": 0.001,
+      "loss": 1.8853,
+      "step": 21623
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 0.17119646072387695,
+      "learning_rate": 0.001,
+      "loss": 3.3695,
+      "step": 21624
+    },
+    {
+      "epoch": 0.9148405110415433,
+      "grad_norm": 0.17291773855686188,
+      "learning_rate": 0.001,
+      "loss": 1.8255,
+      "step": 21625
+    },
+    {
+      "epoch": 0.9148828158050597,
+      "grad_norm": 0.49092814326286316,
+      "learning_rate": 0.001,
+      "loss": 2.3737,
+      "step": 21626
+    },
+    {
+      "epoch": 0.914925120568576,
+      "grad_norm": 0.13538378477096558,
+      "learning_rate": 0.001,
+      "loss": 1.5565,
+      "step": 21627
+    },
+    {
+      "epoch": 0.9149674253320924,
+      "grad_norm": 0.14585568010807037,
+      "learning_rate": 0.001,
+      "loss": 2.0289,
+      "step": 21628
+    },
+    {
+      "epoch": 0.9150097300956088,
+      "grad_norm": 1.05545175075531,
+      "learning_rate": 0.001,
+      "loss": 2.3307,
+      "step": 21629
+    },
+    {
+      "epoch": 0.9150520348591251,
+      "grad_norm": 0.13337115943431854,
+      "learning_rate": 0.001,
+      "loss": 2.3943,
+      "step": 21630
+    },
+    {
+      "epoch": 0.9150943396226415,
+      "grad_norm": 0.13960030674934387,
+      "learning_rate": 0.001,
+      "loss": 2.0699,
+      "step": 21631
+    },
+    {
+      "epoch": 0.9151366443861578,
+      "grad_norm": 0.1596209853887558,
+      "learning_rate": 0.001,
+      "loss": 2.9616,
+      "step": 21632
+    },
+    {
+      "epoch": 0.9151789491496742,
+      "grad_norm": 1.2385811805725098,
+      "learning_rate": 0.001,
+      "loss": 2.1809,
+      "step": 21633
+    },
+    {
+      "epoch": 0.9152212539131906,
+      "grad_norm": 0.15184301137924194,
+      "learning_rate": 0.001,
+      "loss": 1.4528,
+      "step": 21634
+    },
+    {
+      "epoch": 0.915263558676707,
+      "grad_norm": 3.3898375034332275,
+      "learning_rate": 0.001,
+      "loss": 1.8417,
+      "step": 21635
+    },
+    {
+      "epoch": 0.9153058634402234,
+      "grad_norm": 0.1527806520462036,
+      "learning_rate": 0.001,
+      "loss": 2.1561,
+      "step": 21636
+    },
+    {
+      "epoch": 0.9153481682037398,
+      "grad_norm": 0.22297519445419312,
+      "learning_rate": 0.001,
+      "loss": 2.2794,
+      "step": 21637
+    },
+    {
+      "epoch": 0.9153904729672561,
+      "grad_norm": 0.21347364783287048,
+      "learning_rate": 0.001,
+      "loss": 3.186,
+      "step": 21638
+    },
+    {
+      "epoch": 0.9154327777307725,
+      "grad_norm": 0.15860065817832947,
+      "learning_rate": 0.001,
+      "loss": 1.7825,
+      "step": 21639
+    },
+    {
+      "epoch": 0.9154750824942889,
+      "grad_norm": 0.4748792052268982,
+      "learning_rate": 0.001,
+      "loss": 2.4484,
+      "step": 21640
+    },
+    {
+      "epoch": 0.9155173872578052,
+      "grad_norm": 0.15251575410366058,
+      "learning_rate": 0.001,
+      "loss": 1.5772,
+      "step": 21641
+    },
+    {
+      "epoch": 0.9155596920213216,
+      "grad_norm": 0.21372830867767334,
+      "learning_rate": 0.001,
+      "loss": 2.8094,
+      "step": 21642
+    },
+    {
+      "epoch": 0.915601996784838,
+      "grad_norm": 0.19164182245731354,
+      "learning_rate": 0.001,
+      "loss": 3.3052,
+      "step": 21643
+    },
+    {
+      "epoch": 0.9156443015483543,
+      "grad_norm": 0.15785560011863708,
+      "learning_rate": 0.001,
+      "loss": 2.3681,
+      "step": 21644
+    },
+    {
+      "epoch": 0.9156866063118707,
+      "grad_norm": 1.266538143157959,
+      "learning_rate": 0.001,
+      "loss": 2.0366,
+      "step": 21645
+    },
+    {
+      "epoch": 0.9157289110753871,
+      "grad_norm": 0.16436876356601715,
+      "learning_rate": 0.001,
+      "loss": 2.3029,
+      "step": 21646
+    },
+    {
+      "epoch": 0.9157712158389034,
+      "grad_norm": 0.1448490023612976,
+      "learning_rate": 0.001,
+      "loss": 2.3137,
+      "step": 21647
+    },
+    {
+      "epoch": 0.9158135206024198,
+      "grad_norm": 0.1606566458940506,
+      "learning_rate": 0.001,
+      "loss": 1.5938,
+      "step": 21648
+    },
+    {
+      "epoch": 0.9158558253659362,
+      "grad_norm": 0.6070122122764587,
+      "learning_rate": 0.001,
+      "loss": 3.4232,
+      "step": 21649
+    },
+    {
+      "epoch": 0.9158981301294525,
+      "grad_norm": 0.15284164249897003,
+      "learning_rate": 0.001,
+      "loss": 2.2872,
+      "step": 21650
+    },
+    {
+      "epoch": 0.915940434892969,
+      "grad_norm": 0.18675555288791656,
+      "learning_rate": 0.001,
+      "loss": 3.7336,
+      "step": 21651
+    },
+    {
+      "epoch": 0.9159827396564854,
+      "grad_norm": 0.20005974173545837,
+      "learning_rate": 0.001,
+      "loss": 2.4288,
+      "step": 21652
+    },
+    {
+      "epoch": 0.9160250444200017,
+      "grad_norm": 0.17080628871917725,
+      "learning_rate": 0.001,
+      "loss": 2.0765,
+      "step": 21653
+    },
+    {
+      "epoch": 0.9160673491835181,
+      "grad_norm": 0.17919304966926575,
+      "learning_rate": 0.001,
+      "loss": 1.6258,
+      "step": 21654
+    },
+    {
+      "epoch": 0.9161096539470345,
+      "grad_norm": 4.160139560699463,
+      "learning_rate": 0.001,
+      "loss": 3.0095,
+      "step": 21655
+    },
+    {
+      "epoch": 0.9161519587105508,
+      "grad_norm": 0.12385336309671402,
+      "learning_rate": 0.001,
+      "loss": 1.9264,
+      "step": 21656
+    },
+    {
+      "epoch": 0.9161942634740672,
+      "grad_norm": 0.19847314059734344,
+      "learning_rate": 0.001,
+      "loss": 2.0386,
+      "step": 21657
+    },
+    {
+      "epoch": 0.9162365682375836,
+      "grad_norm": 0.14693698287010193,
+      "learning_rate": 0.001,
+      "loss": 1.6431,
+      "step": 21658
+    },
+    {
+      "epoch": 0.9162788730010999,
+      "grad_norm": 0.1676744669675827,
+      "learning_rate": 0.001,
+      "loss": 1.7444,
+      "step": 21659
+    },
+    {
+      "epoch": 0.9163211777646163,
+      "grad_norm": 0.17183883488178253,
+      "learning_rate": 0.001,
+      "loss": 2.122,
+      "step": 21660
+    },
+    {
+      "epoch": 0.9163634825281327,
+      "grad_norm": 0.16480883955955505,
+      "learning_rate": 0.001,
+      "loss": 2.9668,
+      "step": 21661
+    },
+    {
+      "epoch": 0.916405787291649,
+      "grad_norm": 0.16469353437423706,
+      "learning_rate": 0.001,
+      "loss": 2.4785,
+      "step": 21662
+    },
+    {
+      "epoch": 0.9164480920551654,
+      "grad_norm": 0.1656375229358673,
+      "learning_rate": 0.001,
+      "loss": 1.7279,
+      "step": 21663
+    },
+    {
+      "epoch": 0.9164903968186818,
+      "grad_norm": 0.15687058866024017,
+      "learning_rate": 0.001,
+      "loss": 2.5898,
+      "step": 21664
+    },
+    {
+      "epoch": 0.9165327015821981,
+      "grad_norm": 0.585669994354248,
+      "learning_rate": 0.001,
+      "loss": 2.0657,
+      "step": 21665
+    },
+    {
+      "epoch": 0.9165750063457145,
+      "grad_norm": 0.5769844651222229,
+      "learning_rate": 0.001,
+      "loss": 2.0928,
+      "step": 21666
+    },
+    {
+      "epoch": 0.916617311109231,
+      "grad_norm": 0.18087390065193176,
+      "learning_rate": 0.001,
+      "loss": 1.8316,
+      "step": 21667
+    },
+    {
+      "epoch": 0.9166596158727472,
+      "grad_norm": 0.15963704884052277,
+      "learning_rate": 0.001,
+      "loss": 1.6256,
+      "step": 21668
+    },
+    {
+      "epoch": 0.9167019206362637,
+      "grad_norm": 0.15890270471572876,
+      "learning_rate": 0.001,
+      "loss": 1.5167,
+      "step": 21669
+    },
+    {
+      "epoch": 0.9167442253997801,
+      "grad_norm": 0.15813934803009033,
+      "learning_rate": 0.001,
+      "loss": 2.0896,
+      "step": 21670
+    },
+    {
+      "epoch": 0.9167865301632964,
+      "grad_norm": 0.17641639709472656,
+      "learning_rate": 0.001,
+      "loss": 1.7598,
+      "step": 21671
+    },
+    {
+      "epoch": 0.9168288349268128,
+      "grad_norm": 0.14289072155952454,
+      "learning_rate": 0.001,
+      "loss": 2.1995,
+      "step": 21672
+    },
+    {
+      "epoch": 0.9168711396903292,
+      "grad_norm": 0.14103195071220398,
+      "learning_rate": 0.001,
+      "loss": 1.6827,
+      "step": 21673
+    },
+    {
+      "epoch": 0.9169134444538455,
+      "grad_norm": 1.1286795139312744,
+      "learning_rate": 0.001,
+      "loss": 1.7365,
+      "step": 21674
+    },
+    {
+      "epoch": 0.9169557492173619,
+      "grad_norm": 0.1436692327260971,
+      "learning_rate": 0.001,
+      "loss": 1.5473,
+      "step": 21675
+    },
+    {
+      "epoch": 0.9169980539808782,
+      "grad_norm": 0.17740869522094727,
+      "learning_rate": 0.001,
+      "loss": 2.0802,
+      "step": 21676
+    },
+    {
+      "epoch": 0.9170403587443946,
+      "grad_norm": 0.21841758489608765,
+      "learning_rate": 0.001,
+      "loss": 2.4257,
+      "step": 21677
+    },
+    {
+      "epoch": 0.917082663507911,
+      "grad_norm": 0.1688189059495926,
+      "learning_rate": 0.001,
+      "loss": 1.6848,
+      "step": 21678
+    },
+    {
+      "epoch": 0.9171249682714273,
+      "grad_norm": 7.481250762939453,
+      "learning_rate": 0.001,
+      "loss": 1.8009,
+      "step": 21679
+    },
+    {
+      "epoch": 0.9171672730349437,
+      "grad_norm": 0.16650600731372833,
+      "learning_rate": 0.001,
+      "loss": 1.9332,
+      "step": 21680
+    },
+    {
+      "epoch": 0.9172095777984601,
+      "grad_norm": 0.16894063353538513,
+      "learning_rate": 0.001,
+      "loss": 1.8266,
+      "step": 21681
+    },
+    {
+      "epoch": 0.9172518825619764,
+      "grad_norm": 1.4320636987686157,
+      "learning_rate": 0.001,
+      "loss": 1.7303,
+      "step": 21682
+    },
+    {
+      "epoch": 0.9172941873254928,
+      "grad_norm": 0.15776890516281128,
+      "learning_rate": 0.001,
+      "loss": 1.483,
+      "step": 21683
+    },
+    {
+      "epoch": 0.9173364920890092,
+      "grad_norm": 0.5583562254905701,
+      "learning_rate": 0.001,
+      "loss": 2.6289,
+      "step": 21684
+    },
+    {
+      "epoch": 0.9173787968525255,
+      "grad_norm": 2.3220925331115723,
+      "learning_rate": 0.001,
+      "loss": 2.3595,
+      "step": 21685
+    },
+    {
+      "epoch": 0.917421101616042,
+      "grad_norm": 0.19249920547008514,
+      "learning_rate": 0.001,
+      "loss": 2.1336,
+      "step": 21686
+    },
+    {
+      "epoch": 0.9174634063795584,
+      "grad_norm": 0.24140247702598572,
+      "learning_rate": 0.001,
+      "loss": 2.4654,
+      "step": 21687
+    },
+    {
+      "epoch": 0.9175057111430747,
+      "grad_norm": 0.17343805730342865,
+      "learning_rate": 0.001,
+      "loss": 2.3035,
+      "step": 21688
+    },
+    {
+      "epoch": 0.9175480159065911,
+      "grad_norm": 0.19953952729701996,
+      "learning_rate": 0.001,
+      "loss": 1.9234,
+      "step": 21689
+    },
+    {
+      "epoch": 0.9175903206701075,
+      "grad_norm": 1.9272069931030273,
+      "learning_rate": 0.001,
+      "loss": 3.4397,
+      "step": 21690
+    },
+    {
+      "epoch": 0.9176326254336238,
+      "grad_norm": 0.1924961805343628,
+      "learning_rate": 0.001,
+      "loss": 2.4553,
+      "step": 21691
+    },
+    {
+      "epoch": 0.9176749301971402,
+      "grad_norm": 0.20997212827205658,
+      "learning_rate": 0.001,
+      "loss": 1.7757,
+      "step": 21692
+    },
+    {
+      "epoch": 0.9177172349606566,
+      "grad_norm": 0.18580669164657593,
+      "learning_rate": 0.001,
+      "loss": 2.1987,
+      "step": 21693
+    },
+    {
+      "epoch": 0.9177595397241729,
+      "grad_norm": 0.2456483542919159,
+      "learning_rate": 0.001,
+      "loss": 2.1412,
+      "step": 21694
+    },
+    {
+      "epoch": 0.9178018444876893,
+      "grad_norm": 0.2759229242801666,
+      "learning_rate": 0.001,
+      "loss": 2.3739,
+      "step": 21695
+    },
+    {
+      "epoch": 0.9178441492512057,
+      "grad_norm": 0.19626349210739136,
+      "learning_rate": 0.001,
+      "loss": 2.4,
+      "step": 21696
+    },
+    {
+      "epoch": 0.917886454014722,
+      "grad_norm": 0.19402183592319489,
+      "learning_rate": 0.001,
+      "loss": 3.3583,
+      "step": 21697
+    },
+    {
+      "epoch": 0.9179287587782384,
+      "grad_norm": 0.21589398384094238,
+      "learning_rate": 0.001,
+      "loss": 1.6545,
+      "step": 21698
+    },
+    {
+      "epoch": 0.9179710635417548,
+      "grad_norm": 0.19767926633358002,
+      "learning_rate": 0.001,
+      "loss": 2.963,
+      "step": 21699
+    },
+    {
+      "epoch": 0.9180133683052711,
+      "grad_norm": 0.6116325855255127,
+      "learning_rate": 0.001,
+      "loss": 2.4778,
+      "step": 21700
+    },
+    {
+      "epoch": 0.9180556730687875,
+      "grad_norm": 0.18952438235282898,
+      "learning_rate": 0.001,
+      "loss": 2.9357,
+      "step": 21701
+    },
+    {
+      "epoch": 0.918097977832304,
+      "grad_norm": 0.20334750413894653,
+      "learning_rate": 0.001,
+      "loss": 1.9754,
+      "step": 21702
+    },
+    {
+      "epoch": 0.9181402825958203,
+      "grad_norm": 0.17652149498462677,
+      "learning_rate": 0.001,
+      "loss": 3.0083,
+      "step": 21703
+    },
+    {
+      "epoch": 0.9181825873593367,
+      "grad_norm": 0.1732001155614853,
+      "learning_rate": 0.001,
+      "loss": 2.601,
+      "step": 21704
+    },
+    {
+      "epoch": 0.9182248921228531,
+      "grad_norm": 0.16499556601047516,
+      "learning_rate": 0.001,
+      "loss": 2.3997,
+      "step": 21705
+    },
+    {
+      "epoch": 0.9182671968863694,
+      "grad_norm": 0.1443309634923935,
+      "learning_rate": 0.001,
+      "loss": 3.3834,
+      "step": 21706
+    },
+    {
+      "epoch": 0.9183095016498858,
+      "grad_norm": 0.14484232664108276,
+      "learning_rate": 0.001,
+      "loss": 1.6655,
+      "step": 21707
+    },
+    {
+      "epoch": 0.9183518064134022,
+      "grad_norm": 0.13555298745632172,
+      "learning_rate": 0.001,
+      "loss": 1.6735,
+      "step": 21708
+    },
+    {
+      "epoch": 0.9183941111769185,
+      "grad_norm": 0.2232842892408371,
+      "learning_rate": 0.001,
+      "loss": 2.4252,
+      "step": 21709
+    },
+    {
+      "epoch": 0.9184364159404349,
+      "grad_norm": 0.175230473279953,
+      "learning_rate": 0.001,
+      "loss": 2.0342,
+      "step": 21710
+    },
+    {
+      "epoch": 0.9184787207039513,
+      "grad_norm": 0.1450617015361786,
+      "learning_rate": 0.001,
+      "loss": 1.7858,
+      "step": 21711
+    },
+    {
+      "epoch": 0.9185210254674676,
+      "grad_norm": 0.37246161699295044,
+      "learning_rate": 0.001,
+      "loss": 3.1194,
+      "step": 21712
+    },
+    {
+      "epoch": 0.918563330230984,
+      "grad_norm": 0.1520744115114212,
+      "learning_rate": 0.001,
+      "loss": 2.9826,
+      "step": 21713
+    },
+    {
+      "epoch": 0.9186056349945004,
+      "grad_norm": 0.17187443375587463,
+      "learning_rate": 0.001,
+      "loss": 2.6739,
+      "step": 21714
+    },
+    {
+      "epoch": 0.9186479397580167,
+      "grad_norm": 0.17397736012935638,
+      "learning_rate": 0.001,
+      "loss": 2.6349,
+      "step": 21715
+    },
+    {
+      "epoch": 0.9186902445215331,
+      "grad_norm": 2.4483251571655273,
+      "learning_rate": 0.001,
+      "loss": 2.8474,
+      "step": 21716
+    },
+    {
+      "epoch": 0.9187325492850495,
+      "grad_norm": 0.9830498695373535,
+      "learning_rate": 0.001,
+      "loss": 1.3181,
+      "step": 21717
+    },
+    {
+      "epoch": 0.9187748540485658,
+      "grad_norm": 0.1355532854795456,
+      "learning_rate": 0.001,
+      "loss": 2.4041,
+      "step": 21718
+    },
+    {
+      "epoch": 0.9188171588120823,
+      "grad_norm": 0.161952406167984,
+      "learning_rate": 0.001,
+      "loss": 2.9089,
+      "step": 21719
+    },
+    {
+      "epoch": 0.9188594635755987,
+      "grad_norm": 0.1725846230983734,
+      "learning_rate": 0.001,
+      "loss": 2.2824,
+      "step": 21720
+    },
+    {
+      "epoch": 0.918901768339115,
+      "grad_norm": 0.21761982142925262,
+      "learning_rate": 0.001,
+      "loss": 1.7065,
+      "step": 21721
+    },
+    {
+      "epoch": 0.9189440731026314,
+      "grad_norm": 0.16416600346565247,
+      "learning_rate": 0.001,
+      "loss": 1.8811,
+      "step": 21722
+    },
+    {
+      "epoch": 0.9189863778661477,
+      "grad_norm": 0.1734263002872467,
+      "learning_rate": 0.001,
+      "loss": 1.956,
+      "step": 21723
+    },
+    {
+      "epoch": 0.9190286826296641,
+      "grad_norm": 0.3407423198223114,
+      "learning_rate": 0.001,
+      "loss": 1.9373,
+      "step": 21724
+    },
+    {
+      "epoch": 0.9190709873931805,
+      "grad_norm": 0.1345268189907074,
+      "learning_rate": 0.001,
+      "loss": 1.7216,
+      "step": 21725
+    },
+    {
+      "epoch": 0.9191132921566968,
+      "grad_norm": 0.1486416757106781,
+      "learning_rate": 0.001,
+      "loss": 1.7895,
+      "step": 21726
+    },
+    {
+      "epoch": 0.9191555969202132,
+      "grad_norm": 0.14210502803325653,
+      "learning_rate": 0.001,
+      "loss": 2.0866,
+      "step": 21727
+    },
+    {
+      "epoch": 0.9191979016837296,
+      "grad_norm": 0.13867826759815216,
+      "learning_rate": 0.001,
+      "loss": 2.6647,
+      "step": 21728
+    },
+    {
+      "epoch": 0.9192402064472459,
+      "grad_norm": 0.4167260527610779,
+      "learning_rate": 0.001,
+      "loss": 2.3402,
+      "step": 21729
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 0.16781596839427948,
+      "learning_rate": 0.001,
+      "loss": 1.7369,
+      "step": 21730
+    },
+    {
+      "epoch": 0.9193248159742787,
+      "grad_norm": 0.16833274066448212,
+      "learning_rate": 0.001,
+      "loss": 2.136,
+      "step": 21731
+    },
+    {
+      "epoch": 0.919367120737795,
+      "grad_norm": 15.674321174621582,
+      "learning_rate": 0.001,
+      "loss": 2.771,
+      "step": 21732
+    },
+    {
+      "epoch": 0.9194094255013114,
+      "grad_norm": 1.287104606628418,
+      "learning_rate": 0.001,
+      "loss": 1.7155,
+      "step": 21733
+    },
+    {
+      "epoch": 0.9194517302648278,
+      "grad_norm": 0.14855779707431793,
+      "learning_rate": 0.001,
+      "loss": 2.1788,
+      "step": 21734
+    },
+    {
+      "epoch": 0.9194940350283441,
+      "grad_norm": 0.18779130280017853,
+      "learning_rate": 0.001,
+      "loss": 1.9621,
+      "step": 21735
+    },
+    {
+      "epoch": 0.9195363397918606,
+      "grad_norm": 0.18050996959209442,
+      "learning_rate": 0.001,
+      "loss": 1.7367,
+      "step": 21736
+    },
+    {
+      "epoch": 0.919578644555377,
+      "grad_norm": 0.2634378671646118,
+      "learning_rate": 0.001,
+      "loss": 2.6046,
+      "step": 21737
+    },
+    {
+      "epoch": 0.9196209493188933,
+      "grad_norm": 0.2787913978099823,
+      "learning_rate": 0.001,
+      "loss": 3.1114,
+      "step": 21738
+    },
+    {
+      "epoch": 0.9196632540824097,
+      "grad_norm": 0.15546195209026337,
+      "learning_rate": 0.001,
+      "loss": 2.7819,
+      "step": 21739
+    },
+    {
+      "epoch": 0.9197055588459261,
+      "grad_norm": 0.21307052671909332,
+      "learning_rate": 0.001,
+      "loss": 2.5438,
+      "step": 21740
+    },
+    {
+      "epoch": 0.9197478636094424,
+      "grad_norm": 0.18763384222984314,
+      "learning_rate": 0.001,
+      "loss": 1.7654,
+      "step": 21741
+    },
+    {
+      "epoch": 0.9197901683729588,
+      "grad_norm": 0.12560519576072693,
+      "learning_rate": 0.001,
+      "loss": 2.6024,
+      "step": 21742
+    },
+    {
+      "epoch": 0.9198324731364752,
+      "grad_norm": 0.18310007452964783,
+      "learning_rate": 0.001,
+      "loss": 2.0988,
+      "step": 21743
+    },
+    {
+      "epoch": 0.9198747778999915,
+      "grad_norm": 0.2050357460975647,
+      "learning_rate": 0.001,
+      "loss": 2.5167,
+      "step": 21744
+    },
+    {
+      "epoch": 0.9199170826635079,
+      "grad_norm": 0.21944154798984528,
+      "learning_rate": 0.001,
+      "loss": 2.9099,
+      "step": 21745
+    },
+    {
+      "epoch": 0.9199593874270243,
+      "grad_norm": 3.412935495376587,
+      "learning_rate": 0.001,
+      "loss": 2.1277,
+      "step": 21746
+    },
+    {
+      "epoch": 0.9200016921905406,
+      "grad_norm": 0.21068432927131653,
+      "learning_rate": 0.001,
+      "loss": 1.6681,
+      "step": 21747
+    },
+    {
+      "epoch": 0.920043996954057,
+      "grad_norm": 0.18886512517929077,
+      "learning_rate": 0.001,
+      "loss": 2.294,
+      "step": 21748
+    },
+    {
+      "epoch": 0.9200863017175734,
+      "grad_norm": 1.677681803703308,
+      "learning_rate": 0.001,
+      "loss": 2.0359,
+      "step": 21749
+    },
+    {
+      "epoch": 0.9201286064810897,
+      "grad_norm": 0.194906085729599,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 21750
+    },
+    {
+      "epoch": 0.9201709112446061,
+      "grad_norm": 0.15261149406433105,
+      "learning_rate": 0.001,
+      "loss": 2.8701,
+      "step": 21751
+    },
+    {
+      "epoch": 0.9202132160081226,
+      "grad_norm": 0.16249825060367584,
+      "learning_rate": 0.001,
+      "loss": 2.4157,
+      "step": 21752
+    },
+    {
+      "epoch": 0.9202555207716389,
+      "grad_norm": 0.20041526854038239,
+      "learning_rate": 0.001,
+      "loss": 2.5838,
+      "step": 21753
+    },
+    {
+      "epoch": 0.9202978255351553,
+      "grad_norm": 0.2039480060338974,
+      "learning_rate": 0.001,
+      "loss": 1.1561,
+      "step": 21754
+    },
+    {
+      "epoch": 0.9203401302986717,
+      "grad_norm": 0.1664251685142517,
+      "learning_rate": 0.001,
+      "loss": 2.5506,
+      "step": 21755
+    },
+    {
+      "epoch": 0.920382435062188,
+      "grad_norm": 0.19452784955501556,
+      "learning_rate": 0.001,
+      "loss": 2.1064,
+      "step": 21756
+    },
+    {
+      "epoch": 0.9204247398257044,
+      "grad_norm": 0.19989818334579468,
+      "learning_rate": 0.001,
+      "loss": 1.9846,
+      "step": 21757
+    },
+    {
+      "epoch": 0.9204670445892208,
+      "grad_norm": 0.18792906403541565,
+      "learning_rate": 0.001,
+      "loss": 1.8072,
+      "step": 21758
+    },
+    {
+      "epoch": 0.9205093493527371,
+      "grad_norm": 0.4300324618816376,
+      "learning_rate": 0.001,
+      "loss": 3.106,
+      "step": 21759
+    },
+    {
+      "epoch": 0.9205516541162535,
+      "grad_norm": 0.20472699403762817,
+      "learning_rate": 0.001,
+      "loss": 2.8519,
+      "step": 21760
+    },
+    {
+      "epoch": 0.9205939588797699,
+      "grad_norm": 0.16424056887626648,
+      "learning_rate": 0.001,
+      "loss": 2.8775,
+      "step": 21761
+    },
+    {
+      "epoch": 0.9206362636432862,
+      "grad_norm": 0.2106097787618637,
+      "learning_rate": 0.001,
+      "loss": 2.7263,
+      "step": 21762
+    },
+    {
+      "epoch": 0.9206785684068026,
+      "grad_norm": 0.18007276952266693,
+      "learning_rate": 0.001,
+      "loss": 2.1168,
+      "step": 21763
+    },
+    {
+      "epoch": 0.920720873170319,
+      "grad_norm": 0.8814851641654968,
+      "learning_rate": 0.001,
+      "loss": 2.6329,
+      "step": 21764
+    },
+    {
+      "epoch": 0.9207631779338353,
+      "grad_norm": 0.16528195142745972,
+      "learning_rate": 0.001,
+      "loss": 2.3614,
+      "step": 21765
+    },
+    {
+      "epoch": 0.9208054826973517,
+      "grad_norm": 0.14733877778053284,
+      "learning_rate": 0.001,
+      "loss": 1.9466,
+      "step": 21766
+    },
+    {
+      "epoch": 0.920847787460868,
+      "grad_norm": 0.15155373513698578,
+      "learning_rate": 0.001,
+      "loss": 2.0437,
+      "step": 21767
+    },
+    {
+      "epoch": 0.9208900922243844,
+      "grad_norm": 0.7689833045005798,
+      "learning_rate": 0.001,
+      "loss": 1.5882,
+      "step": 21768
+    },
+    {
+      "epoch": 0.9209323969879009,
+      "grad_norm": 0.35222962498664856,
+      "learning_rate": 0.001,
+      "loss": 2.6627,
+      "step": 21769
+    },
+    {
+      "epoch": 0.9209747017514172,
+      "grad_norm": 0.22019173204898834,
+      "learning_rate": 0.001,
+      "loss": 2.2336,
+      "step": 21770
+    },
+    {
+      "epoch": 0.9210170065149336,
+      "grad_norm": 1.362524151802063,
+      "learning_rate": 0.001,
+      "loss": 1.5802,
+      "step": 21771
+    },
+    {
+      "epoch": 0.92105931127845,
+      "grad_norm": 0.18947310745716095,
+      "learning_rate": 0.001,
+      "loss": 2.1529,
+      "step": 21772
+    },
+    {
+      "epoch": 0.9211016160419663,
+      "grad_norm": 0.5228827595710754,
+      "learning_rate": 0.001,
+      "loss": 2.2288,
+      "step": 21773
+    },
+    {
+      "epoch": 0.9211439208054827,
+      "grad_norm": 0.19477491080760956,
+      "learning_rate": 0.001,
+      "loss": 1.7683,
+      "step": 21774
+    },
+    {
+      "epoch": 0.9211862255689991,
+      "grad_norm": 0.1812824010848999,
+      "learning_rate": 0.001,
+      "loss": 1.9203,
+      "step": 21775
+    },
+    {
+      "epoch": 0.9212285303325154,
+      "grad_norm": 0.5380221009254456,
+      "learning_rate": 0.001,
+      "loss": 1.5879,
+      "step": 21776
+    },
+    {
+      "epoch": 0.9212708350960318,
+      "grad_norm": 2.430056095123291,
+      "learning_rate": 0.001,
+      "loss": 1.8626,
+      "step": 21777
+    },
+    {
+      "epoch": 0.9213131398595482,
+      "grad_norm": 0.14518165588378906,
+      "learning_rate": 0.001,
+      "loss": 1.956,
+      "step": 21778
+    },
+    {
+      "epoch": 0.9213554446230645,
+      "grad_norm": 0.16930201649665833,
+      "learning_rate": 0.001,
+      "loss": 2.0088,
+      "step": 21779
+    },
+    {
+      "epoch": 0.9213977493865809,
+      "grad_norm": 0.16156643629074097,
+      "learning_rate": 0.001,
+      "loss": 2.6447,
+      "step": 21780
+    },
+    {
+      "epoch": 0.9214400541500973,
+      "grad_norm": 0.2412625253200531,
+      "learning_rate": 0.001,
+      "loss": 1.5386,
+      "step": 21781
+    },
+    {
+      "epoch": 0.9214823589136136,
+      "grad_norm": 0.16392214596271515,
+      "learning_rate": 0.001,
+      "loss": 1.8491,
+      "step": 21782
+    },
+    {
+      "epoch": 0.92152466367713,
+      "grad_norm": 0.3166408836841583,
+      "learning_rate": 0.001,
+      "loss": 1.8965,
+      "step": 21783
+    },
+    {
+      "epoch": 0.9215669684406465,
+      "grad_norm": 0.6187849044799805,
+      "learning_rate": 0.001,
+      "loss": 3.2798,
+      "step": 21784
+    },
+    {
+      "epoch": 0.9216092732041627,
+      "grad_norm": 0.16876527667045593,
+      "learning_rate": 0.001,
+      "loss": 3.2222,
+      "step": 21785
+    },
+    {
+      "epoch": 0.9216515779676792,
+      "grad_norm": 0.21848821640014648,
+      "learning_rate": 0.001,
+      "loss": 2.4569,
+      "step": 21786
+    },
+    {
+      "epoch": 0.9216938827311956,
+      "grad_norm": 0.2898956537246704,
+      "learning_rate": 0.001,
+      "loss": 1.9074,
+      "step": 21787
+    },
+    {
+      "epoch": 0.9217361874947119,
+      "grad_norm": 0.9921806454658508,
+      "learning_rate": 0.001,
+      "loss": 2.6345,
+      "step": 21788
+    },
+    {
+      "epoch": 0.9217784922582283,
+      "grad_norm": 0.2962776720523834,
+      "learning_rate": 0.001,
+      "loss": 1.9957,
+      "step": 21789
+    },
+    {
+      "epoch": 0.9218207970217447,
+      "grad_norm": 4.324014663696289,
+      "learning_rate": 0.001,
+      "loss": 2.4474,
+      "step": 21790
+    },
+    {
+      "epoch": 0.921863101785261,
+      "grad_norm": 2.0869667530059814,
+      "learning_rate": 0.001,
+      "loss": 2.0607,
+      "step": 21791
+    },
+    {
+      "epoch": 0.9219054065487774,
+      "grad_norm": 0.1384233683347702,
+      "learning_rate": 0.001,
+      "loss": 2.6952,
+      "step": 21792
+    },
+    {
+      "epoch": 0.9219477113122938,
+      "grad_norm": 0.2545308470726013,
+      "learning_rate": 0.001,
+      "loss": 1.5397,
+      "step": 21793
+    },
+    {
+      "epoch": 0.9219900160758101,
+      "grad_norm": 0.1824929267168045,
+      "learning_rate": 0.001,
+      "loss": 2.8564,
+      "step": 21794
+    },
+    {
+      "epoch": 0.9220323208393265,
+      "grad_norm": 0.6214986443519592,
+      "learning_rate": 0.001,
+      "loss": 1.8838,
+      "step": 21795
+    },
+    {
+      "epoch": 0.9220746256028429,
+      "grad_norm": 0.1345427930355072,
+      "learning_rate": 0.001,
+      "loss": 1.5997,
+      "step": 21796
+    },
+    {
+      "epoch": 0.9221169303663592,
+      "grad_norm": 0.3405480980873108,
+      "learning_rate": 0.001,
+      "loss": 3.5338,
+      "step": 21797
+    },
+    {
+      "epoch": 0.9221592351298756,
+      "grad_norm": 1.3706905841827393,
+      "learning_rate": 0.001,
+      "loss": 1.5665,
+      "step": 21798
+    },
+    {
+      "epoch": 0.922201539893392,
+      "grad_norm": 3.3803353309631348,
+      "learning_rate": 0.001,
+      "loss": 2.0037,
+      "step": 21799
+    },
+    {
+      "epoch": 0.9222438446569083,
+      "grad_norm": 2.6381101608276367,
+      "learning_rate": 0.001,
+      "loss": 1.8247,
+      "step": 21800
+    },
+    {
+      "epoch": 0.9222861494204248,
+      "grad_norm": 0.15839798748493195,
+      "learning_rate": 0.001,
+      "loss": 2.7898,
+      "step": 21801
+    },
+    {
+      "epoch": 0.9223284541839412,
+      "grad_norm": 0.3356753885746002,
+      "learning_rate": 0.001,
+      "loss": 2.5539,
+      "step": 21802
+    },
+    {
+      "epoch": 0.9223707589474575,
+      "grad_norm": 0.1914248913526535,
+      "learning_rate": 0.001,
+      "loss": 1.6862,
+      "step": 21803
+    },
+    {
+      "epoch": 0.9224130637109739,
+      "grad_norm": 10.729948997497559,
+      "learning_rate": 0.001,
+      "loss": 1.7948,
+      "step": 21804
+    },
+    {
+      "epoch": 0.9224553684744903,
+      "grad_norm": 0.5791533589363098,
+      "learning_rate": 0.001,
+      "loss": 2.1725,
+      "step": 21805
+    },
+    {
+      "epoch": 0.9224976732380066,
+      "grad_norm": 0.20189444720745087,
+      "learning_rate": 0.001,
+      "loss": 1.7816,
+      "step": 21806
+    },
+    {
+      "epoch": 0.922539978001523,
+      "grad_norm": 5.010307788848877,
+      "learning_rate": 0.001,
+      "loss": 3.1041,
+      "step": 21807
+    },
+    {
+      "epoch": 0.9225822827650394,
+      "grad_norm": 0.21783508360385895,
+      "learning_rate": 0.001,
+      "loss": 2.4775,
+      "step": 21808
+    },
+    {
+      "epoch": 0.9226245875285557,
+      "grad_norm": 0.20718340575695038,
+      "learning_rate": 0.001,
+      "loss": 2.3186,
+      "step": 21809
+    },
+    {
+      "epoch": 0.9226668922920721,
+      "grad_norm": 0.18991310894489288,
+      "learning_rate": 0.001,
+      "loss": 1.8765,
+      "step": 21810
+    },
+    {
+      "epoch": 0.9227091970555885,
+      "grad_norm": 0.1602325290441513,
+      "learning_rate": 0.001,
+      "loss": 1.8977,
+      "step": 21811
+    },
+    {
+      "epoch": 0.9227515018191048,
+      "grad_norm": 0.40473005175590515,
+      "learning_rate": 0.001,
+      "loss": 2.3219,
+      "step": 21812
+    },
+    {
+      "epoch": 0.9227938065826212,
+      "grad_norm": 0.46014153957366943,
+      "learning_rate": 0.001,
+      "loss": 3.8687,
+      "step": 21813
+    },
+    {
+      "epoch": 0.9228361113461375,
+      "grad_norm": 53.86265182495117,
+      "learning_rate": 0.001,
+      "loss": 3.7695,
+      "step": 21814
+    },
+    {
+      "epoch": 0.9228784161096539,
+      "grad_norm": 0.18623895943164825,
+      "learning_rate": 0.001,
+      "loss": 2.0448,
+      "step": 21815
+    },
+    {
+      "epoch": 0.9229207208731703,
+      "grad_norm": 0.33342859148979187,
+      "learning_rate": 0.001,
+      "loss": 2.0159,
+      "step": 21816
+    },
+    {
+      "epoch": 0.9229630256366866,
+      "grad_norm": 17.7737979888916,
+      "learning_rate": 0.001,
+      "loss": 3.0125,
+      "step": 21817
+    },
+    {
+      "epoch": 0.923005330400203,
+      "grad_norm": 0.12280112504959106,
+      "learning_rate": 0.001,
+      "loss": 1.5704,
+      "step": 21818
+    },
+    {
+      "epoch": 0.9230476351637195,
+      "grad_norm": 0.21494708955287933,
+      "learning_rate": 0.001,
+      "loss": 1.6715,
+      "step": 21819
+    },
+    {
+      "epoch": 0.9230899399272358,
+      "grad_norm": 0.2017885446548462,
+      "learning_rate": 0.001,
+      "loss": 2.2747,
+      "step": 21820
+    },
+    {
+      "epoch": 0.9231322446907522,
+      "grad_norm": 0.2331605851650238,
+      "learning_rate": 0.001,
+      "loss": 1.9135,
+      "step": 21821
+    },
+    {
+      "epoch": 0.9231745494542686,
+      "grad_norm": 0.16537924110889435,
+      "learning_rate": 0.001,
+      "loss": 2.3864,
+      "step": 21822
+    },
+    {
+      "epoch": 0.9232168542177849,
+      "grad_norm": 0.1348440796136856,
+      "learning_rate": 0.001,
+      "loss": 1.8148,
+      "step": 21823
+    },
+    {
+      "epoch": 0.9232591589813013,
+      "grad_norm": 0.24906490743160248,
+      "learning_rate": 0.001,
+      "loss": 2.2961,
+      "step": 21824
+    },
+    {
+      "epoch": 0.9233014637448177,
+      "grad_norm": 0.16530346870422363,
+      "learning_rate": 0.001,
+      "loss": 2.0339,
+      "step": 21825
+    },
+    {
+      "epoch": 0.923343768508334,
+      "grad_norm": 0.13655491173267365,
+      "learning_rate": 0.001,
+      "loss": 2.1709,
+      "step": 21826
+    },
+    {
+      "epoch": 0.9233860732718504,
+      "grad_norm": 0.13853634893894196,
+      "learning_rate": 0.001,
+      "loss": 1.7936,
+      "step": 21827
+    },
+    {
+      "epoch": 0.9234283780353668,
+      "grad_norm": 0.14643678069114685,
+      "learning_rate": 0.001,
+      "loss": 2.1668,
+      "step": 21828
+    },
+    {
+      "epoch": 0.9234706827988831,
+      "grad_norm": 0.2975621521472931,
+      "learning_rate": 0.001,
+      "loss": 1.9856,
+      "step": 21829
+    },
+    {
+      "epoch": 0.9235129875623995,
+      "grad_norm": 0.1982540637254715,
+      "learning_rate": 0.001,
+      "loss": 2.4848,
+      "step": 21830
+    },
+    {
+      "epoch": 0.9235552923259159,
+      "grad_norm": 0.22689512372016907,
+      "learning_rate": 0.001,
+      "loss": 2.57,
+      "step": 21831
+    },
+    {
+      "epoch": 0.9235975970894322,
+      "grad_norm": 0.1401916742324829,
+      "learning_rate": 0.001,
+      "loss": 1.6943,
+      "step": 21832
+    },
+    {
+      "epoch": 0.9236399018529486,
+      "grad_norm": 2.501429796218872,
+      "learning_rate": 0.001,
+      "loss": 1.5686,
+      "step": 21833
+    },
+    {
+      "epoch": 0.923682206616465,
+      "grad_norm": 0.15667013823986053,
+      "learning_rate": 0.001,
+      "loss": 1.5747,
+      "step": 21834
+    },
+    {
+      "epoch": 0.9237245113799814,
+      "grad_norm": 0.19675031304359436,
+      "learning_rate": 0.001,
+      "loss": 2.1214,
+      "step": 21835
+    },
+    {
+      "epoch": 0.9237668161434978,
+      "grad_norm": 0.28486713767051697,
+      "learning_rate": 0.001,
+      "loss": 2.2525,
+      "step": 21836
+    },
+    {
+      "epoch": 0.9238091209070142,
+      "grad_norm": 0.28747811913490295,
+      "learning_rate": 0.001,
+      "loss": 2.6731,
+      "step": 21837
+    },
+    {
+      "epoch": 0.9238514256705305,
+      "grad_norm": 0.16760526597499847,
+      "learning_rate": 0.001,
+      "loss": 2.2244,
+      "step": 21838
+    },
+    {
+      "epoch": 0.9238937304340469,
+      "grad_norm": 0.23037217557430267,
+      "learning_rate": 0.001,
+      "loss": 2.328,
+      "step": 21839
+    },
+    {
+      "epoch": 0.9239360351975633,
+      "grad_norm": 0.13840559124946594,
+      "learning_rate": 0.001,
+      "loss": 1.8101,
+      "step": 21840
+    },
+    {
+      "epoch": 0.9239783399610796,
+      "grad_norm": 0.25135916471481323,
+      "learning_rate": 0.001,
+      "loss": 2.3498,
+      "step": 21841
+    },
+    {
+      "epoch": 0.924020644724596,
+      "grad_norm": 55.877220153808594,
+      "learning_rate": 0.001,
+      "loss": 2.1315,
+      "step": 21842
+    },
+    {
+      "epoch": 0.9240629494881124,
+      "grad_norm": 0.12954889237880707,
+      "learning_rate": 0.001,
+      "loss": 2.0259,
+      "step": 21843
+    },
+    {
+      "epoch": 0.9241052542516287,
+      "grad_norm": 3.64554762840271,
+      "learning_rate": 0.001,
+      "loss": 2.1561,
+      "step": 21844
+    },
+    {
+      "epoch": 0.9241475590151451,
+      "grad_norm": 0.13647957146167755,
+      "learning_rate": 0.001,
+      "loss": 2.0474,
+      "step": 21845
+    },
+    {
+      "epoch": 0.9241898637786615,
+      "grad_norm": 0.1622430980205536,
+      "learning_rate": 0.001,
+      "loss": 2.051,
+      "step": 21846
+    },
+    {
+      "epoch": 0.9242321685421778,
+      "grad_norm": 0.17634640634059906,
+      "learning_rate": 0.001,
+      "loss": 2.8697,
+      "step": 21847
+    },
+    {
+      "epoch": 0.9242744733056942,
+      "grad_norm": 0.14419732987880707,
+      "learning_rate": 0.001,
+      "loss": 1.8025,
+      "step": 21848
+    },
+    {
+      "epoch": 0.9243167780692106,
+      "grad_norm": 37.351749420166016,
+      "learning_rate": 0.001,
+      "loss": 1.999,
+      "step": 21849
+    },
+    {
+      "epoch": 0.9243590828327269,
+      "grad_norm": 0.20993314683437347,
+      "learning_rate": 0.001,
+      "loss": 2.2057,
+      "step": 21850
+    },
+    {
+      "epoch": 0.9244013875962434,
+      "grad_norm": 0.29707229137420654,
+      "learning_rate": 0.001,
+      "loss": 3.4083,
+      "step": 21851
+    },
+    {
+      "epoch": 0.9244436923597598,
+      "grad_norm": 0.2973187565803528,
+      "learning_rate": 0.001,
+      "loss": 2.5179,
+      "step": 21852
+    },
+    {
+      "epoch": 0.9244859971232761,
+      "grad_norm": 0.26736098527908325,
+      "learning_rate": 0.001,
+      "loss": 1.8084,
+      "step": 21853
+    },
+    {
+      "epoch": 0.9245283018867925,
+      "grad_norm": 0.22918181121349335,
+      "learning_rate": 0.001,
+      "loss": 2.0938,
+      "step": 21854
+    },
+    {
+      "epoch": 0.9245706066503089,
+      "grad_norm": 0.2083725780248642,
+      "learning_rate": 0.001,
+      "loss": 3.048,
+      "step": 21855
+    },
+    {
+      "epoch": 0.9246129114138252,
+      "grad_norm": 1.0710909366607666,
+      "learning_rate": 0.001,
+      "loss": 1.9642,
+      "step": 21856
+    },
+    {
+      "epoch": 0.9246552161773416,
+      "grad_norm": 0.6909781694412231,
+      "learning_rate": 0.001,
+      "loss": 3.8566,
+      "step": 21857
+    },
+    {
+      "epoch": 0.9246975209408579,
+      "grad_norm": 0.310916543006897,
+      "learning_rate": 0.001,
+      "loss": 2.163,
+      "step": 21858
+    },
+    {
+      "epoch": 0.9247398257043743,
+      "grad_norm": 0.15251141786575317,
+      "learning_rate": 0.001,
+      "loss": 2.3132,
+      "step": 21859
+    },
+    {
+      "epoch": 0.9247821304678907,
+      "grad_norm": 0.19609764218330383,
+      "learning_rate": 0.001,
+      "loss": 1.6755,
+      "step": 21860
+    },
+    {
+      "epoch": 0.924824435231407,
+      "grad_norm": 0.7422865033149719,
+      "learning_rate": 0.001,
+      "loss": 2.823,
+      "step": 21861
+    },
+    {
+      "epoch": 0.9248667399949234,
+      "grad_norm": 0.14667384326457977,
+      "learning_rate": 0.001,
+      "loss": 2.1124,
+      "step": 21862
+    },
+    {
+      "epoch": 0.9249090447584398,
+      "grad_norm": 0.16594403982162476,
+      "learning_rate": 0.001,
+      "loss": 2.0591,
+      "step": 21863
+    },
+    {
+      "epoch": 0.9249513495219561,
+      "grad_norm": 0.18323829770088196,
+      "learning_rate": 0.001,
+      "loss": 1.6222,
+      "step": 21864
+    },
+    {
+      "epoch": 0.9249936542854725,
+      "grad_norm": 0.22297947108745575,
+      "learning_rate": 0.001,
+      "loss": 2.2291,
+      "step": 21865
+    },
+    {
+      "epoch": 0.9250359590489889,
+      "grad_norm": 0.16403332352638245,
+      "learning_rate": 0.001,
+      "loss": 2.3493,
+      "step": 21866
+    },
+    {
+      "epoch": 0.9250782638125052,
+      "grad_norm": 0.23182706534862518,
+      "learning_rate": 0.001,
+      "loss": 2.0805,
+      "step": 21867
+    },
+    {
+      "epoch": 0.9251205685760217,
+      "grad_norm": 0.22759734094142914,
+      "learning_rate": 0.001,
+      "loss": 2.4908,
+      "step": 21868
+    },
+    {
+      "epoch": 0.9251628733395381,
+      "grad_norm": 0.6803989410400391,
+      "learning_rate": 0.001,
+      "loss": 2.7571,
+      "step": 21869
+    },
+    {
+      "epoch": 0.9252051781030544,
+      "grad_norm": 1.2668521404266357,
+      "learning_rate": 0.001,
+      "loss": 1.9701,
+      "step": 21870
+    },
+    {
+      "epoch": 0.9252474828665708,
+      "grad_norm": 0.1656140238046646,
+      "learning_rate": 0.001,
+      "loss": 1.9307,
+      "step": 21871
+    },
+    {
+      "epoch": 0.9252897876300872,
+      "grad_norm": 0.14242412149906158,
+      "learning_rate": 0.001,
+      "loss": 2.5352,
+      "step": 21872
+    },
+    {
+      "epoch": 0.9253320923936035,
+      "grad_norm": 0.1434088498353958,
+      "learning_rate": 0.001,
+      "loss": 2.284,
+      "step": 21873
+    },
+    {
+      "epoch": 0.9253743971571199,
+      "grad_norm": 0.21865014731884003,
+      "learning_rate": 0.001,
+      "loss": 2.3557,
+      "step": 21874
+    },
+    {
+      "epoch": 0.9254167019206363,
+      "grad_norm": 0.1559746265411377,
+      "learning_rate": 0.001,
+      "loss": 1.9937,
+      "step": 21875
+    },
+    {
+      "epoch": 0.9254590066841526,
+      "grad_norm": 2.3426716327667236,
+      "learning_rate": 0.001,
+      "loss": 2.8022,
+      "step": 21876
+    },
+    {
+      "epoch": 0.925501311447669,
+      "grad_norm": 0.18600401282310486,
+      "learning_rate": 0.001,
+      "loss": 2.062,
+      "step": 21877
+    },
+    {
+      "epoch": 0.9255436162111854,
+      "grad_norm": 0.37630805373191833,
+      "learning_rate": 0.001,
+      "loss": 2.308,
+      "step": 21878
+    },
+    {
+      "epoch": 0.9255859209747017,
+      "grad_norm": 0.26190125942230225,
+      "learning_rate": 0.001,
+      "loss": 1.5722,
+      "step": 21879
+    },
+    {
+      "epoch": 0.9256282257382181,
+      "grad_norm": 0.17115432024002075,
+      "learning_rate": 0.001,
+      "loss": 2.0632,
+      "step": 21880
+    },
+    {
+      "epoch": 0.9256705305017345,
+      "grad_norm": 0.3577999472618103,
+      "learning_rate": 0.001,
+      "loss": 2.8232,
+      "step": 21881
+    },
+    {
+      "epoch": 0.9257128352652508,
+      "grad_norm": 0.15650738775730133,
+      "learning_rate": 0.001,
+      "loss": 2.1344,
+      "step": 21882
+    },
+    {
+      "epoch": 0.9257551400287672,
+      "grad_norm": 1.1321587562561035,
+      "learning_rate": 0.001,
+      "loss": 3.678,
+      "step": 21883
+    },
+    {
+      "epoch": 0.9257974447922837,
+      "grad_norm": 0.2203882336616516,
+      "learning_rate": 0.001,
+      "loss": 2.5869,
+      "step": 21884
+    },
+    {
+      "epoch": 0.9258397495558,
+      "grad_norm": 0.20148371160030365,
+      "learning_rate": 0.001,
+      "loss": 1.5426,
+      "step": 21885
+    },
+    {
+      "epoch": 0.9258820543193164,
+      "grad_norm": 0.18005190789699554,
+      "learning_rate": 0.001,
+      "loss": 1.9097,
+      "step": 21886
+    },
+    {
+      "epoch": 0.9259243590828328,
+      "grad_norm": 0.9154829382896423,
+      "learning_rate": 0.001,
+      "loss": 2.5761,
+      "step": 21887
+    },
+    {
+      "epoch": 0.9259666638463491,
+      "grad_norm": 0.16887728869915009,
+      "learning_rate": 0.001,
+      "loss": 2.6111,
+      "step": 21888
+    },
+    {
+      "epoch": 0.9260089686098655,
+      "grad_norm": 0.16773365437984467,
+      "learning_rate": 0.001,
+      "loss": 1.4804,
+      "step": 21889
+    },
+    {
+      "epoch": 0.9260512733733819,
+      "grad_norm": 0.672725260257721,
+      "learning_rate": 0.001,
+      "loss": 2.6119,
+      "step": 21890
+    },
+    {
+      "epoch": 0.9260935781368982,
+      "grad_norm": 2.3344016075134277,
+      "learning_rate": 0.001,
+      "loss": 1.5694,
+      "step": 21891
+    },
+    {
+      "epoch": 0.9261358829004146,
+      "grad_norm": 0.21035782992839813,
+      "learning_rate": 0.001,
+      "loss": 3.0887,
+      "step": 21892
+    },
+    {
+      "epoch": 0.926178187663931,
+      "grad_norm": 0.1845501810312271,
+      "learning_rate": 0.001,
+      "loss": 2.0434,
+      "step": 21893
+    },
+    {
+      "epoch": 0.9262204924274473,
+      "grad_norm": 0.20723649859428406,
+      "learning_rate": 0.001,
+      "loss": 1.8821,
+      "step": 21894
+    },
+    {
+      "epoch": 0.9262627971909637,
+      "grad_norm": 0.4533455967903137,
+      "learning_rate": 0.001,
+      "loss": 2.3809,
+      "step": 21895
+    },
+    {
+      "epoch": 0.9263051019544801,
+      "grad_norm": 1.0757017135620117,
+      "learning_rate": 0.001,
+      "loss": 2.2222,
+      "step": 21896
+    },
+    {
+      "epoch": 0.9263474067179964,
+      "grad_norm": 0.2541113495826721,
+      "learning_rate": 0.001,
+      "loss": 2.3574,
+      "step": 21897
+    },
+    {
+      "epoch": 0.9263897114815128,
+      "grad_norm": 1.4928815364837646,
+      "learning_rate": 0.001,
+      "loss": 2.738,
+      "step": 21898
+    },
+    {
+      "epoch": 0.9264320162450292,
+      "grad_norm": 0.17061060667037964,
+      "learning_rate": 0.001,
+      "loss": 1.8219,
+      "step": 21899
+    },
+    {
+      "epoch": 0.9264743210085455,
+      "grad_norm": 2.1404104232788086,
+      "learning_rate": 0.001,
+      "loss": 3.1614,
+      "step": 21900
+    },
+    {
+      "epoch": 0.926516625772062,
+      "grad_norm": 0.4184578061103821,
+      "learning_rate": 0.001,
+      "loss": 2.5873,
+      "step": 21901
+    },
+    {
+      "epoch": 0.9265589305355783,
+      "grad_norm": 0.758017361164093,
+      "learning_rate": 0.001,
+      "loss": 2.5198,
+      "step": 21902
+    },
+    {
+      "epoch": 0.9266012352990947,
+      "grad_norm": 0.2236809879541397,
+      "learning_rate": 0.001,
+      "loss": 2.0093,
+      "step": 21903
+    },
+    {
+      "epoch": 0.9266435400626111,
+      "grad_norm": 0.1709446907043457,
+      "learning_rate": 0.001,
+      "loss": 2.2584,
+      "step": 21904
+    },
+    {
+      "epoch": 0.9266858448261274,
+      "grad_norm": 0.18377362191677094,
+      "learning_rate": 0.001,
+      "loss": 1.7896,
+      "step": 21905
+    },
+    {
+      "epoch": 0.9267281495896438,
+      "grad_norm": 0.17706288397312164,
+      "learning_rate": 0.001,
+      "loss": 2.1536,
+      "step": 21906
+    },
+    {
+      "epoch": 0.9267704543531602,
+      "grad_norm": 3.2292115688323975,
+      "learning_rate": 0.001,
+      "loss": 2.3149,
+      "step": 21907
+    },
+    {
+      "epoch": 0.9268127591166765,
+      "grad_norm": 0.1633622944355011,
+      "learning_rate": 0.001,
+      "loss": 2.2869,
+      "step": 21908
+    },
+    {
+      "epoch": 0.9268550638801929,
+      "grad_norm": 0.18171913921833038,
+      "learning_rate": 0.001,
+      "loss": 1.4094,
+      "step": 21909
+    },
+    {
+      "epoch": 0.9268973686437093,
+      "grad_norm": 0.20554277300834656,
+      "learning_rate": 0.001,
+      "loss": 1.8559,
+      "step": 21910
+    },
+    {
+      "epoch": 0.9269396734072256,
+      "grad_norm": 0.4027025103569031,
+      "learning_rate": 0.001,
+      "loss": 3.185,
+      "step": 21911
+    },
+    {
+      "epoch": 0.926981978170742,
+      "grad_norm": 0.23326435685157776,
+      "learning_rate": 0.001,
+      "loss": 2.6125,
+      "step": 21912
+    },
+    {
+      "epoch": 0.9270242829342584,
+      "grad_norm": 0.13189657032489777,
+      "learning_rate": 0.001,
+      "loss": 1.8056,
+      "step": 21913
+    },
+    {
+      "epoch": 0.9270665876977747,
+      "grad_norm": 0.15088336169719696,
+      "learning_rate": 0.001,
+      "loss": 2.4187,
+      "step": 21914
+    },
+    {
+      "epoch": 0.9271088924612911,
+      "grad_norm": 0.275808185338974,
+      "learning_rate": 0.001,
+      "loss": 2.0664,
+      "step": 21915
+    },
+    {
+      "epoch": 0.9271511972248075,
+      "grad_norm": 0.16052496433258057,
+      "learning_rate": 0.001,
+      "loss": 1.911,
+      "step": 21916
+    },
+    {
+      "epoch": 0.9271935019883238,
+      "grad_norm": 0.14898687601089478,
+      "learning_rate": 0.001,
+      "loss": 1.9662,
+      "step": 21917
+    },
+    {
+      "epoch": 0.9272358067518403,
+      "grad_norm": 0.13894212245941162,
+      "learning_rate": 0.001,
+      "loss": 2.3613,
+      "step": 21918
+    },
+    {
+      "epoch": 0.9272781115153567,
+      "grad_norm": 0.12985843420028687,
+      "learning_rate": 0.001,
+      "loss": 1.5966,
+      "step": 21919
+    },
+    {
+      "epoch": 0.927320416278873,
+      "grad_norm": 13.45024299621582,
+      "learning_rate": 0.001,
+      "loss": 2.3785,
+      "step": 21920
+    },
+    {
+      "epoch": 0.9273627210423894,
+      "grad_norm": 0.14908571541309357,
+      "learning_rate": 0.001,
+      "loss": 2.8162,
+      "step": 21921
+    },
+    {
+      "epoch": 0.9274050258059058,
+      "grad_norm": 0.1413915455341339,
+      "learning_rate": 0.001,
+      "loss": 2.1948,
+      "step": 21922
+    },
+    {
+      "epoch": 0.9274473305694221,
+      "grad_norm": 0.13808265328407288,
+      "learning_rate": 0.001,
+      "loss": 1.619,
+      "step": 21923
+    },
+    {
+      "epoch": 0.9274896353329385,
+      "grad_norm": 1.1322904825210571,
+      "learning_rate": 0.001,
+      "loss": 3.322,
+      "step": 21924
+    },
+    {
+      "epoch": 0.9275319400964549,
+      "grad_norm": 0.19078315794467926,
+      "learning_rate": 0.001,
+      "loss": 2.4069,
+      "step": 21925
+    },
+    {
+      "epoch": 0.9275742448599712,
+      "grad_norm": 0.341876357793808,
+      "learning_rate": 0.001,
+      "loss": 1.9122,
+      "step": 21926
+    },
+    {
+      "epoch": 0.9276165496234876,
+      "grad_norm": 0.1864626556634903,
+      "learning_rate": 0.001,
+      "loss": 1.8488,
+      "step": 21927
+    },
+    {
+      "epoch": 0.927658854387004,
+      "grad_norm": 0.19447799026966095,
+      "learning_rate": 0.001,
+      "loss": 2.0424,
+      "step": 21928
+    },
+    {
+      "epoch": 0.9277011591505203,
+      "grad_norm": 0.534332275390625,
+      "learning_rate": 0.001,
+      "loss": 2.9907,
+      "step": 21929
+    },
+    {
+      "epoch": 0.9277434639140367,
+      "grad_norm": 0.16288526356220245,
+      "learning_rate": 0.001,
+      "loss": 2.4716,
+      "step": 21930
+    },
+    {
+      "epoch": 0.9277857686775531,
+      "grad_norm": 0.1490895301103592,
+      "learning_rate": 0.001,
+      "loss": 1.6003,
+      "step": 21931
+    },
+    {
+      "epoch": 0.9278280734410694,
+      "grad_norm": 0.16813409328460693,
+      "learning_rate": 0.001,
+      "loss": 2.4994,
+      "step": 21932
+    },
+    {
+      "epoch": 0.9278703782045858,
+      "grad_norm": 0.14073902368545532,
+      "learning_rate": 0.001,
+      "loss": 2.4238,
+      "step": 21933
+    },
+    {
+      "epoch": 0.9279126829681023,
+      "grad_norm": 0.1578540802001953,
+      "learning_rate": 0.001,
+      "loss": 1.8206,
+      "step": 21934
+    },
+    {
+      "epoch": 0.9279549877316186,
+      "grad_norm": 0.3728206753730774,
+      "learning_rate": 0.001,
+      "loss": 2.1717,
+      "step": 21935
+    },
+    {
+      "epoch": 0.927997292495135,
+      "grad_norm": 0.16407276690006256,
+      "learning_rate": 0.001,
+      "loss": 1.3542,
+      "step": 21936
+    },
+    {
+      "epoch": 0.9280395972586514,
+      "grad_norm": 0.5730623006820679,
+      "learning_rate": 0.001,
+      "loss": 2.5158,
+      "step": 21937
+    },
+    {
+      "epoch": 0.9280819020221677,
+      "grad_norm": 0.4225897789001465,
+      "learning_rate": 0.001,
+      "loss": 2.63,
+      "step": 21938
+    },
+    {
+      "epoch": 0.9281242067856841,
+      "grad_norm": 0.19734539091587067,
+      "learning_rate": 0.001,
+      "loss": 1.9227,
+      "step": 21939
+    },
+    {
+      "epoch": 0.9281665115492005,
+      "grad_norm": 0.22450488805770874,
+      "learning_rate": 0.001,
+      "loss": 2.4276,
+      "step": 21940
+    },
+    {
+      "epoch": 0.9282088163127168,
+      "grad_norm": 0.174544557929039,
+      "learning_rate": 0.001,
+      "loss": 2.449,
+      "step": 21941
+    },
+    {
+      "epoch": 0.9282511210762332,
+      "grad_norm": 0.16052083671092987,
+      "learning_rate": 0.001,
+      "loss": 2.0025,
+      "step": 21942
+    },
+    {
+      "epoch": 0.9282934258397496,
+      "grad_norm": 0.14678356051445007,
+      "learning_rate": 0.001,
+      "loss": 2.2724,
+      "step": 21943
+    },
+    {
+      "epoch": 0.9283357306032659,
+      "grad_norm": 0.1925598680973053,
+      "learning_rate": 0.001,
+      "loss": 2.3425,
+      "step": 21944
+    },
+    {
+      "epoch": 0.9283780353667823,
+      "grad_norm": 0.1436510980129242,
+      "learning_rate": 0.001,
+      "loss": 1.4963,
+      "step": 21945
+    },
+    {
+      "epoch": 0.9284203401302987,
+      "grad_norm": 0.1730327010154724,
+      "learning_rate": 0.001,
+      "loss": 2.7398,
+      "step": 21946
+    },
+    {
+      "epoch": 0.928462644893815,
+      "grad_norm": 0.19589565694332123,
+      "learning_rate": 0.001,
+      "loss": 1.7817,
+      "step": 21947
+    },
+    {
+      "epoch": 0.9285049496573314,
+      "grad_norm": 0.4311773180961609,
+      "learning_rate": 0.001,
+      "loss": 2.5202,
+      "step": 21948
+    },
+    {
+      "epoch": 0.9285472544208477,
+      "grad_norm": 0.16326993703842163,
+      "learning_rate": 0.001,
+      "loss": 1.5574,
+      "step": 21949
+    },
+    {
+      "epoch": 0.9285895591843641,
+      "grad_norm": 0.17932333052158356,
+      "learning_rate": 0.001,
+      "loss": 1.8842,
+      "step": 21950
+    },
+    {
+      "epoch": 0.9286318639478806,
+      "grad_norm": 0.18280398845672607,
+      "learning_rate": 0.001,
+      "loss": 1.8407,
+      "step": 21951
+    },
+    {
+      "epoch": 0.9286741687113969,
+      "grad_norm": 0.2984718978404999,
+      "learning_rate": 0.001,
+      "loss": 2.01,
+      "step": 21952
+    },
+    {
+      "epoch": 0.9287164734749133,
+      "grad_norm": 0.1796807199716568,
+      "learning_rate": 0.001,
+      "loss": 2.3846,
+      "step": 21953
+    },
+    {
+      "epoch": 0.9287587782384297,
+      "grad_norm": 0.21217189729213715,
+      "learning_rate": 0.001,
+      "loss": 1.991,
+      "step": 21954
+    },
+    {
+      "epoch": 0.928801083001946,
+      "grad_norm": 0.19677573442459106,
+      "learning_rate": 0.001,
+      "loss": 1.8435,
+      "step": 21955
+    },
+    {
+      "epoch": 0.9288433877654624,
+      "grad_norm": 0.15436428785324097,
+      "learning_rate": 0.001,
+      "loss": 2.0693,
+      "step": 21956
+    },
+    {
+      "epoch": 0.9288856925289788,
+      "grad_norm": 0.15341611206531525,
+      "learning_rate": 0.001,
+      "loss": 1.5144,
+      "step": 21957
+    },
+    {
+      "epoch": 0.9289279972924951,
+      "grad_norm": 0.16485430300235748,
+      "learning_rate": 0.001,
+      "loss": 2.2537,
+      "step": 21958
+    },
+    {
+      "epoch": 0.9289703020560115,
+      "grad_norm": 0.20801179111003876,
+      "learning_rate": 0.001,
+      "loss": 1.6632,
+      "step": 21959
+    },
+    {
+      "epoch": 0.9290126068195279,
+      "grad_norm": 0.1642443984746933,
+      "learning_rate": 0.001,
+      "loss": 2.366,
+      "step": 21960
+    },
+    {
+      "epoch": 0.9290549115830442,
+      "grad_norm": 0.22680698335170746,
+      "learning_rate": 0.001,
+      "loss": 1.711,
+      "step": 21961
+    },
+    {
+      "epoch": 0.9290972163465606,
+      "grad_norm": 0.16985486447811127,
+      "learning_rate": 0.001,
+      "loss": 2.383,
+      "step": 21962
+    },
+    {
+      "epoch": 0.929139521110077,
+      "grad_norm": 0.1616482138633728,
+      "learning_rate": 0.001,
+      "loss": 1.9125,
+      "step": 21963
+    },
+    {
+      "epoch": 0.9291818258735933,
+      "grad_norm": 0.15879781544208527,
+      "learning_rate": 0.001,
+      "loss": 2.075,
+      "step": 21964
+    },
+    {
+      "epoch": 0.9292241306371097,
+      "grad_norm": 0.12933063507080078,
+      "learning_rate": 0.001,
+      "loss": 2.4285,
+      "step": 21965
+    },
+    {
+      "epoch": 0.9292664354006261,
+      "grad_norm": 0.1703050583600998,
+      "learning_rate": 0.001,
+      "loss": 1.8814,
+      "step": 21966
+    },
+    {
+      "epoch": 0.9293087401641424,
+      "grad_norm": 0.15759029984474182,
+      "learning_rate": 0.001,
+      "loss": 1.975,
+      "step": 21967
+    },
+    {
+      "epoch": 0.9293510449276589,
+      "grad_norm": 0.16814181208610535,
+      "learning_rate": 0.001,
+      "loss": 3.0872,
+      "step": 21968
+    },
+    {
+      "epoch": 0.9293933496911753,
+      "grad_norm": 1.8021339178085327,
+      "learning_rate": 0.001,
+      "loss": 2.3614,
+      "step": 21969
+    },
+    {
+      "epoch": 0.9294356544546916,
+      "grad_norm": 0.16891023516654968,
+      "learning_rate": 0.001,
+      "loss": 1.7206,
+      "step": 21970
+    },
+    {
+      "epoch": 0.929477959218208,
+      "grad_norm": 0.297505259513855,
+      "learning_rate": 0.001,
+      "loss": 2.7611,
+      "step": 21971
+    },
+    {
+      "epoch": 0.9295202639817244,
+      "grad_norm": 0.1714632362127304,
+      "learning_rate": 0.001,
+      "loss": 2.1129,
+      "step": 21972
+    },
+    {
+      "epoch": 0.9295625687452407,
+      "grad_norm": 0.19479063153266907,
+      "learning_rate": 0.001,
+      "loss": 2.0595,
+      "step": 21973
+    },
+    {
+      "epoch": 0.9296048735087571,
+      "grad_norm": 0.180049866437912,
+      "learning_rate": 0.001,
+      "loss": 2.3096,
+      "step": 21974
+    },
+    {
+      "epoch": 0.9296471782722735,
+      "grad_norm": 0.6898881196975708,
+      "learning_rate": 0.001,
+      "loss": 2.1587,
+      "step": 21975
+    },
+    {
+      "epoch": 0.9296894830357898,
+      "grad_norm": 1.291130781173706,
+      "learning_rate": 0.001,
+      "loss": 2.1467,
+      "step": 21976
+    },
+    {
+      "epoch": 0.9297317877993062,
+      "grad_norm": 0.16307774186134338,
+      "learning_rate": 0.001,
+      "loss": 2.5267,
+      "step": 21977
+    },
+    {
+      "epoch": 0.9297740925628226,
+      "grad_norm": 0.2098264843225479,
+      "learning_rate": 0.001,
+      "loss": 2.5711,
+      "step": 21978
+    },
+    {
+      "epoch": 0.9298163973263389,
+      "grad_norm": 2.183814287185669,
+      "learning_rate": 0.001,
+      "loss": 1.6738,
+      "step": 21979
+    },
+    {
+      "epoch": 0.9298587020898553,
+      "grad_norm": 0.1923678070306778,
+      "learning_rate": 0.001,
+      "loss": 3.2464,
+      "step": 21980
+    },
+    {
+      "epoch": 0.9299010068533717,
+      "grad_norm": 0.13686540722846985,
+      "learning_rate": 0.001,
+      "loss": 2.1387,
+      "step": 21981
+    },
+    {
+      "epoch": 0.929943311616888,
+      "grad_norm": 4.34953498840332,
+      "learning_rate": 0.001,
+      "loss": 1.6731,
+      "step": 21982
+    },
+    {
+      "epoch": 0.9299856163804044,
+      "grad_norm": 0.16118478775024414,
+      "learning_rate": 0.001,
+      "loss": 2.368,
+      "step": 21983
+    },
+    {
+      "epoch": 0.9300279211439209,
+      "grad_norm": 0.589535117149353,
+      "learning_rate": 0.001,
+      "loss": 2.4586,
+      "step": 21984
+    },
+    {
+      "epoch": 0.9300702259074372,
+      "grad_norm": 0.535638689994812,
+      "learning_rate": 0.001,
+      "loss": 2.6041,
+      "step": 21985
+    },
+    {
+      "epoch": 0.9301125306709536,
+      "grad_norm": 0.1409117430448532,
+      "learning_rate": 0.001,
+      "loss": 2.2921,
+      "step": 21986
+    },
+    {
+      "epoch": 0.93015483543447,
+      "grad_norm": 0.27417516708374023,
+      "learning_rate": 0.001,
+      "loss": 1.468,
+      "step": 21987
+    },
+    {
+      "epoch": 0.9301971401979863,
+      "grad_norm": 0.4808787405490875,
+      "learning_rate": 0.001,
+      "loss": 1.8091,
+      "step": 21988
+    },
+    {
+      "epoch": 0.9302394449615027,
+      "grad_norm": 3.5292954444885254,
+      "learning_rate": 0.001,
+      "loss": 2.1259,
+      "step": 21989
+    },
+    {
+      "epoch": 0.9302817497250191,
+      "grad_norm": 0.1566908210515976,
+      "learning_rate": 0.001,
+      "loss": 2.3119,
+      "step": 21990
+    },
+    {
+      "epoch": 0.9303240544885354,
+      "grad_norm": 0.14310206472873688,
+      "learning_rate": 0.001,
+      "loss": 3.0195,
+      "step": 21991
+    },
+    {
+      "epoch": 0.9303663592520518,
+      "grad_norm": 0.16697970032691956,
+      "learning_rate": 0.001,
+      "loss": 2.2591,
+      "step": 21992
+    },
+    {
+      "epoch": 0.9304086640155681,
+      "grad_norm": 0.1776588410139084,
+      "learning_rate": 0.001,
+      "loss": 2.4498,
+      "step": 21993
+    },
+    {
+      "epoch": 0.9304509687790845,
+      "grad_norm": 0.6791718602180481,
+      "learning_rate": 0.001,
+      "loss": 1.7671,
+      "step": 21994
+    },
+    {
+      "epoch": 0.9304932735426009,
+      "grad_norm": 11.523677825927734,
+      "learning_rate": 0.001,
+      "loss": 2.4538,
+      "step": 21995
+    },
+    {
+      "epoch": 0.9305355783061172,
+      "grad_norm": 0.16806228458881378,
+      "learning_rate": 0.001,
+      "loss": 2.1725,
+      "step": 21996
+    },
+    {
+      "epoch": 0.9305778830696336,
+      "grad_norm": 0.24460969865322113,
+      "learning_rate": 0.001,
+      "loss": 2.6757,
+      "step": 21997
+    },
+    {
+      "epoch": 0.93062018783315,
+      "grad_norm": 0.19408555328845978,
+      "learning_rate": 0.001,
+      "loss": 2.5526,
+      "step": 21998
+    },
+    {
+      "epoch": 0.9306624925966663,
+      "grad_norm": 0.18568742275238037,
+      "learning_rate": 0.001,
+      "loss": 2.6363,
+      "step": 21999
+    },
+    {
+      "epoch": 0.9307047973601827,
+      "grad_norm": 0.16618452966213226,
+      "learning_rate": 0.001,
+      "loss": 2.1546,
+      "step": 22000
+    },
+    {
+      "epoch": 0.9307471021236992,
+      "grad_norm": 0.16462920606136322,
+      "learning_rate": 0.001,
+      "loss": 1.5723,
+      "step": 22001
+    },
+    {
+      "epoch": 0.9307894068872155,
+      "grad_norm": 0.12047246843576431,
+      "learning_rate": 0.001,
+      "loss": 1.9203,
+      "step": 22002
+    },
+    {
+      "epoch": 0.9308317116507319,
+      "grad_norm": 0.22285322844982147,
+      "learning_rate": 0.001,
+      "loss": 3.2207,
+      "step": 22003
+    },
+    {
+      "epoch": 0.9308740164142483,
+      "grad_norm": 0.14793570339679718,
+      "learning_rate": 0.001,
+      "loss": 2.233,
+      "step": 22004
+    },
+    {
+      "epoch": 0.9309163211777646,
+      "grad_norm": 0.16332119703292847,
+      "learning_rate": 0.001,
+      "loss": 1.9054,
+      "step": 22005
+    },
+    {
+      "epoch": 0.930958625941281,
+      "grad_norm": 0.1610136777162552,
+      "learning_rate": 0.001,
+      "loss": 2.1695,
+      "step": 22006
+    },
+    {
+      "epoch": 0.9310009307047974,
+      "grad_norm": 0.8120068907737732,
+      "learning_rate": 0.001,
+      "loss": 2.0812,
+      "step": 22007
+    },
+    {
+      "epoch": 0.9310432354683137,
+      "grad_norm": 0.1367693543434143,
+      "learning_rate": 0.001,
+      "loss": 1.6623,
+      "step": 22008
+    },
+    {
+      "epoch": 0.9310855402318301,
+      "grad_norm": 0.2394467443227768,
+      "learning_rate": 0.001,
+      "loss": 2.7049,
+      "step": 22009
+    },
+    {
+      "epoch": 0.9311278449953465,
+      "grad_norm": 0.4347785413265228,
+      "learning_rate": 0.001,
+      "loss": 2.373,
+      "step": 22010
+    },
+    {
+      "epoch": 0.9311701497588628,
+      "grad_norm": 0.1576579064130783,
+      "learning_rate": 0.001,
+      "loss": 2.3929,
+      "step": 22011
+    },
+    {
+      "epoch": 0.9312124545223792,
+      "grad_norm": 0.41782185435295105,
+      "learning_rate": 0.001,
+      "loss": 1.6985,
+      "step": 22012
+    },
+    {
+      "epoch": 0.9312547592858956,
+      "grad_norm": 0.16780440509319305,
+      "learning_rate": 0.001,
+      "loss": 1.8258,
+      "step": 22013
+    },
+    {
+      "epoch": 0.9312970640494119,
+      "grad_norm": 0.15203818678855896,
+      "learning_rate": 0.001,
+      "loss": 2.5598,
+      "step": 22014
+    },
+    {
+      "epoch": 0.9313393688129283,
+      "grad_norm": 0.1656513512134552,
+      "learning_rate": 0.001,
+      "loss": 2.0156,
+      "step": 22015
+    },
+    {
+      "epoch": 0.9313816735764447,
+      "grad_norm": 0.2978683412075043,
+      "learning_rate": 0.001,
+      "loss": 1.8497,
+      "step": 22016
+    },
+    {
+      "epoch": 0.931423978339961,
+      "grad_norm": 2.85372257232666,
+      "learning_rate": 0.001,
+      "loss": 2.6914,
+      "step": 22017
+    },
+    {
+      "epoch": 0.9314662831034775,
+      "grad_norm": 0.16330118477344513,
+      "learning_rate": 0.001,
+      "loss": 2.3187,
+      "step": 22018
+    },
+    {
+      "epoch": 0.9315085878669939,
+      "grad_norm": 0.6117385029792786,
+      "learning_rate": 0.001,
+      "loss": 2.3308,
+      "step": 22019
+    },
+    {
+      "epoch": 0.9315508926305102,
+      "grad_norm": 0.15620072185993195,
+      "learning_rate": 0.001,
+      "loss": 2.5478,
+      "step": 22020
+    },
+    {
+      "epoch": 0.9315931973940266,
+      "grad_norm": 0.167611226439476,
+      "learning_rate": 0.001,
+      "loss": 2.4346,
+      "step": 22021
+    },
+    {
+      "epoch": 0.931635502157543,
+      "grad_norm": 0.36283206939697266,
+      "learning_rate": 0.001,
+      "loss": 2.1529,
+      "step": 22022
+    },
+    {
+      "epoch": 0.9316778069210593,
+      "grad_norm": 0.15195079147815704,
+      "learning_rate": 0.001,
+      "loss": 1.6573,
+      "step": 22023
+    },
+    {
+      "epoch": 0.9317201116845757,
+      "grad_norm": 0.16134585440158844,
+      "learning_rate": 0.001,
+      "loss": 2.0162,
+      "step": 22024
+    },
+    {
+      "epoch": 0.9317624164480921,
+      "grad_norm": 0.20685450732707977,
+      "learning_rate": 0.001,
+      "loss": 1.9848,
+      "step": 22025
+    },
+    {
+      "epoch": 0.9318047212116084,
+      "grad_norm": 0.24313196539878845,
+      "learning_rate": 0.001,
+      "loss": 2.4929,
+      "step": 22026
+    },
+    {
+      "epoch": 0.9318470259751248,
+      "grad_norm": 0.20717033743858337,
+      "learning_rate": 0.001,
+      "loss": 1.6051,
+      "step": 22027
+    },
+    {
+      "epoch": 0.9318893307386412,
+      "grad_norm": 0.1792394369840622,
+      "learning_rate": 0.001,
+      "loss": 2.2256,
+      "step": 22028
+    },
+    {
+      "epoch": 0.9319316355021575,
+      "grad_norm": 0.17706766724586487,
+      "learning_rate": 0.001,
+      "loss": 3.126,
+      "step": 22029
+    },
+    {
+      "epoch": 0.9319739402656739,
+      "grad_norm": 0.21002162992954254,
+      "learning_rate": 0.001,
+      "loss": 2.634,
+      "step": 22030
+    },
+    {
+      "epoch": 0.9320162450291903,
+      "grad_norm": 0.16654951870441437,
+      "learning_rate": 0.001,
+      "loss": 2.363,
+      "step": 22031
+    },
+    {
+      "epoch": 0.9320585497927066,
+      "grad_norm": 0.19493767619132996,
+      "learning_rate": 0.001,
+      "loss": 2.8382,
+      "step": 22032
+    },
+    {
+      "epoch": 0.932100854556223,
+      "grad_norm": 0.24346645176410675,
+      "learning_rate": 0.001,
+      "loss": 1.4973,
+      "step": 22033
+    },
+    {
+      "epoch": 0.9321431593197395,
+      "grad_norm": 0.19552533328533173,
+      "learning_rate": 0.001,
+      "loss": 2.1005,
+      "step": 22034
+    },
+    {
+      "epoch": 0.9321854640832558,
+      "grad_norm": 0.4952528774738312,
+      "learning_rate": 0.001,
+      "loss": 2.4358,
+      "step": 22035
+    },
+    {
+      "epoch": 0.9322277688467722,
+      "grad_norm": 0.16397695243358612,
+      "learning_rate": 0.001,
+      "loss": 1.7882,
+      "step": 22036
+    },
+    {
+      "epoch": 0.9322700736102885,
+      "grad_norm": 0.16197176277637482,
+      "learning_rate": 0.001,
+      "loss": 2.0223,
+      "step": 22037
+    },
+    {
+      "epoch": 0.9323123783738049,
+      "grad_norm": 0.5470212697982788,
+      "learning_rate": 0.001,
+      "loss": 3.6555,
+      "step": 22038
+    },
+    {
+      "epoch": 0.9323546831373213,
+      "grad_norm": 0.1571442037820816,
+      "learning_rate": 0.001,
+      "loss": 2.9497,
+      "step": 22039
+    },
+    {
+      "epoch": 0.9323969879008376,
+      "grad_norm": 0.16548150777816772,
+      "learning_rate": 0.001,
+      "loss": 2.1686,
+      "step": 22040
+    },
+    {
+      "epoch": 0.932439292664354,
+      "grad_norm": 22.103620529174805,
+      "learning_rate": 0.001,
+      "loss": 2.1238,
+      "step": 22041
+    },
+    {
+      "epoch": 0.9324815974278704,
+      "grad_norm": 0.16696321964263916,
+      "learning_rate": 0.001,
+      "loss": 2.387,
+      "step": 22042
+    },
+    {
+      "epoch": 0.9325239021913867,
+      "grad_norm": 31.06536102294922,
+      "learning_rate": 0.001,
+      "loss": 2.0811,
+      "step": 22043
+    },
+    {
+      "epoch": 0.9325662069549031,
+      "grad_norm": 0.3314652144908905,
+      "learning_rate": 0.001,
+      "loss": 2.2734,
+      "step": 22044
+    },
+    {
+      "epoch": 0.9326085117184195,
+      "grad_norm": 0.17265798151493073,
+      "learning_rate": 0.001,
+      "loss": 2.7165,
+      "step": 22045
+    },
+    {
+      "epoch": 0.9326508164819358,
+      "grad_norm": 0.1684129536151886,
+      "learning_rate": 0.001,
+      "loss": 1.8214,
+      "step": 22046
+    },
+    {
+      "epoch": 0.9326931212454522,
+      "grad_norm": 3.3001720905303955,
+      "learning_rate": 0.001,
+      "loss": 2.131,
+      "step": 22047
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.13358072936534882,
+      "learning_rate": 0.001,
+      "loss": 2.129,
+      "step": 22048
+    },
+    {
+      "epoch": 0.9327777307724849,
+      "grad_norm": 0.15245433151721954,
+      "learning_rate": 0.001,
+      "loss": 2.2186,
+      "step": 22049
+    },
+    {
+      "epoch": 0.9328200355360013,
+      "grad_norm": 0.16228505969047546,
+      "learning_rate": 0.001,
+      "loss": 1.5467,
+      "step": 22050
+    },
+    {
+      "epoch": 0.9328623402995178,
+      "grad_norm": 0.18052756786346436,
+      "learning_rate": 0.001,
+      "loss": 2.555,
+      "step": 22051
+    },
+    {
+      "epoch": 0.932904645063034,
+      "grad_norm": 0.20280319452285767,
+      "learning_rate": 0.001,
+      "loss": 1.93,
+      "step": 22052
+    },
+    {
+      "epoch": 0.9329469498265505,
+      "grad_norm": 0.15624532103538513,
+      "learning_rate": 0.001,
+      "loss": 2.3892,
+      "step": 22053
+    },
+    {
+      "epoch": 0.9329892545900669,
+      "grad_norm": 0.15655791759490967,
+      "learning_rate": 0.001,
+      "loss": 2.7288,
+      "step": 22054
+    },
+    {
+      "epoch": 0.9330315593535832,
+      "grad_norm": 0.40425026416778564,
+      "learning_rate": 0.001,
+      "loss": 3.1181,
+      "step": 22055
+    },
+    {
+      "epoch": 0.9330738641170996,
+      "grad_norm": 0.41149187088012695,
+      "learning_rate": 0.001,
+      "loss": 4.535,
+      "step": 22056
+    },
+    {
+      "epoch": 0.933116168880616,
+      "grad_norm": 6.105749130249023,
+      "learning_rate": 0.001,
+      "loss": 2.8493,
+      "step": 22057
+    },
+    {
+      "epoch": 0.9331584736441323,
+      "grad_norm": 0.14242605865001678,
+      "learning_rate": 0.001,
+      "loss": 2.2115,
+      "step": 22058
+    },
+    {
+      "epoch": 0.9332007784076487,
+      "grad_norm": 0.14453691244125366,
+      "learning_rate": 0.001,
+      "loss": 2.0789,
+      "step": 22059
+    },
+    {
+      "epoch": 0.9332430831711651,
+      "grad_norm": 0.1681307554244995,
+      "learning_rate": 0.001,
+      "loss": 2.631,
+      "step": 22060
+    },
+    {
+      "epoch": 0.9332853879346814,
+      "grad_norm": 0.14247244596481323,
+      "learning_rate": 0.001,
+      "loss": 2.0911,
+      "step": 22061
+    },
+    {
+      "epoch": 0.9333276926981978,
+      "grad_norm": 0.16670750081539154,
+      "learning_rate": 0.001,
+      "loss": 1.9668,
+      "step": 22062
+    },
+    {
+      "epoch": 0.9333699974617142,
+      "grad_norm": 0.1602124273777008,
+      "learning_rate": 0.001,
+      "loss": 2.805,
+      "step": 22063
+    },
+    {
+      "epoch": 0.9334123022252305,
+      "grad_norm": 0.4584788978099823,
+      "learning_rate": 0.001,
+      "loss": 2.2362,
+      "step": 22064
+    },
+    {
+      "epoch": 0.9334546069887469,
+      "grad_norm": 0.16626909375190735,
+      "learning_rate": 0.001,
+      "loss": 2.8254,
+      "step": 22065
+    },
+    {
+      "epoch": 0.9334969117522633,
+      "grad_norm": 0.1582920104265213,
+      "learning_rate": 0.001,
+      "loss": 2.3247,
+      "step": 22066
+    },
+    {
+      "epoch": 0.9335392165157796,
+      "grad_norm": 0.15107718110084534,
+      "learning_rate": 0.001,
+      "loss": 1.2551,
+      "step": 22067
+    },
+    {
+      "epoch": 0.9335815212792961,
+      "grad_norm": 0.1743275374174118,
+      "learning_rate": 0.001,
+      "loss": 1.9243,
+      "step": 22068
+    },
+    {
+      "epoch": 0.9336238260428125,
+      "grad_norm": 0.148182213306427,
+      "learning_rate": 0.001,
+      "loss": 1.3991,
+      "step": 22069
+    },
+    {
+      "epoch": 0.9336661308063288,
+      "grad_norm": 0.1482088416814804,
+      "learning_rate": 0.001,
+      "loss": 1.905,
+      "step": 22070
+    },
+    {
+      "epoch": 0.9337084355698452,
+      "grad_norm": 0.17089058458805084,
+      "learning_rate": 0.001,
+      "loss": 2.596,
+      "step": 22071
+    },
+    {
+      "epoch": 0.9337507403333616,
+      "grad_norm": 0.15498429536819458,
+      "learning_rate": 0.001,
+      "loss": 1.7609,
+      "step": 22072
+    },
+    {
+      "epoch": 0.9337930450968779,
+      "grad_norm": 0.1602306365966797,
+      "learning_rate": 0.001,
+      "loss": 1.7998,
+      "step": 22073
+    },
+    {
+      "epoch": 0.9338353498603943,
+      "grad_norm": 0.816593587398529,
+      "learning_rate": 0.001,
+      "loss": 1.4125,
+      "step": 22074
+    },
+    {
+      "epoch": 0.9338776546239107,
+      "grad_norm": 3.837941884994507,
+      "learning_rate": 0.001,
+      "loss": 2.5317,
+      "step": 22075
+    },
+    {
+      "epoch": 0.933919959387427,
+      "grad_norm": 0.1522829383611679,
+      "learning_rate": 0.001,
+      "loss": 1.5934,
+      "step": 22076
+    },
+    {
+      "epoch": 0.9339622641509434,
+      "grad_norm": 0.1627858579158783,
+      "learning_rate": 0.001,
+      "loss": 2.3209,
+      "step": 22077
+    },
+    {
+      "epoch": 0.9340045689144598,
+      "grad_norm": 0.22749550640583038,
+      "learning_rate": 0.001,
+      "loss": 1.7942,
+      "step": 22078
+    },
+    {
+      "epoch": 0.9340468736779761,
+      "grad_norm": 0.2853844463825226,
+      "learning_rate": 0.001,
+      "loss": 2.4773,
+      "step": 22079
+    },
+    {
+      "epoch": 0.9340891784414925,
+      "grad_norm": 0.16091381013393402,
+      "learning_rate": 0.001,
+      "loss": 2.1776,
+      "step": 22080
+    },
+    {
+      "epoch": 0.9341314832050089,
+      "grad_norm": 0.15295559167861938,
+      "learning_rate": 0.001,
+      "loss": 2.8482,
+      "step": 22081
+    },
+    {
+      "epoch": 0.9341737879685252,
+      "grad_norm": 0.1788291186094284,
+      "learning_rate": 0.001,
+      "loss": 1.6635,
+      "step": 22082
+    },
+    {
+      "epoch": 0.9342160927320416,
+      "grad_norm": 0.1991390734910965,
+      "learning_rate": 0.001,
+      "loss": 2.5794,
+      "step": 22083
+    },
+    {
+      "epoch": 0.934258397495558,
+      "grad_norm": 0.32976192235946655,
+      "learning_rate": 0.001,
+      "loss": 2.057,
+      "step": 22084
+    },
+    {
+      "epoch": 0.9343007022590744,
+      "grad_norm": 0.1380927860736847,
+      "learning_rate": 0.001,
+      "loss": 1.8436,
+      "step": 22085
+    },
+    {
+      "epoch": 0.9343430070225908,
+      "grad_norm": 0.13859638571739197,
+      "learning_rate": 0.001,
+      "loss": 1.8601,
+      "step": 22086
+    },
+    {
+      "epoch": 0.9343853117861071,
+      "grad_norm": 3.625734567642212,
+      "learning_rate": 0.001,
+      "loss": 2.2505,
+      "step": 22087
+    },
+    {
+      "epoch": 0.9344276165496235,
+      "grad_norm": 0.1253889799118042,
+      "learning_rate": 0.001,
+      "loss": 2.8063,
+      "step": 22088
+    },
+    {
+      "epoch": 0.9344699213131399,
+      "grad_norm": 0.1292024403810501,
+      "learning_rate": 0.001,
+      "loss": 1.7487,
+      "step": 22089
+    },
+    {
+      "epoch": 0.9345122260766562,
+      "grad_norm": 2.1888620853424072,
+      "learning_rate": 0.001,
+      "loss": 1.8234,
+      "step": 22090
+    },
+    {
+      "epoch": 0.9345545308401726,
+      "grad_norm": 0.8209211230278015,
+      "learning_rate": 0.001,
+      "loss": 2.2189,
+      "step": 22091
+    },
+    {
+      "epoch": 0.934596835603689,
+      "grad_norm": 0.14426971971988678,
+      "learning_rate": 0.001,
+      "loss": 1.8472,
+      "step": 22092
+    },
+    {
+      "epoch": 0.9346391403672053,
+      "grad_norm": 0.1676510125398636,
+      "learning_rate": 0.001,
+      "loss": 2.6767,
+      "step": 22093
+    },
+    {
+      "epoch": 0.9346814451307217,
+      "grad_norm": 0.1549409031867981,
+      "learning_rate": 0.001,
+      "loss": 2.5355,
+      "step": 22094
+    },
+    {
+      "epoch": 0.9347237498942381,
+      "grad_norm": 0.15506504476070404,
+      "learning_rate": 0.001,
+      "loss": 2.0499,
+      "step": 22095
+    },
+    {
+      "epoch": 0.9347660546577544,
+      "grad_norm": 0.16985729336738586,
+      "learning_rate": 0.001,
+      "loss": 2.1923,
+      "step": 22096
+    },
+    {
+      "epoch": 0.9348083594212708,
+      "grad_norm": 0.15692901611328125,
+      "learning_rate": 0.001,
+      "loss": 1.603,
+      "step": 22097
+    },
+    {
+      "epoch": 0.9348506641847872,
+      "grad_norm": 0.17224203050136566,
+      "learning_rate": 0.001,
+      "loss": 2.0071,
+      "step": 22098
+    },
+    {
+      "epoch": 0.9348929689483035,
+      "grad_norm": 1.8327698707580566,
+      "learning_rate": 0.001,
+      "loss": 1.5114,
+      "step": 22099
+    },
+    {
+      "epoch": 0.93493527371182,
+      "grad_norm": 0.16607293486595154,
+      "learning_rate": 0.001,
+      "loss": 1.9993,
+      "step": 22100
+    },
+    {
+      "epoch": 0.9349775784753364,
+      "grad_norm": 1.2528990507125854,
+      "learning_rate": 0.001,
+      "loss": 2.7158,
+      "step": 22101
+    },
+    {
+      "epoch": 0.9350198832388527,
+      "grad_norm": 1.1872773170471191,
+      "learning_rate": 0.001,
+      "loss": 2.4694,
+      "step": 22102
+    },
+    {
+      "epoch": 0.9350621880023691,
+      "grad_norm": 0.6562841534614563,
+      "learning_rate": 0.001,
+      "loss": 2.5087,
+      "step": 22103
+    },
+    {
+      "epoch": 0.9351044927658855,
+      "grad_norm": 0.16749240458011627,
+      "learning_rate": 0.001,
+      "loss": 2.0348,
+      "step": 22104
+    },
+    {
+      "epoch": 0.9351467975294018,
+      "grad_norm": 0.7875110507011414,
+      "learning_rate": 0.001,
+      "loss": 2.6865,
+      "step": 22105
+    },
+    {
+      "epoch": 0.9351891022929182,
+      "grad_norm": 0.15838965773582458,
+      "learning_rate": 0.001,
+      "loss": 2.4356,
+      "step": 22106
+    },
+    {
+      "epoch": 0.9352314070564346,
+      "grad_norm": 0.23913998901844025,
+      "learning_rate": 0.001,
+      "loss": 2.4341,
+      "step": 22107
+    },
+    {
+      "epoch": 0.9352737118199509,
+      "grad_norm": 0.19448761641979218,
+      "learning_rate": 0.001,
+      "loss": 2.0626,
+      "step": 22108
+    },
+    {
+      "epoch": 0.9353160165834673,
+      "grad_norm": 0.16222184896469116,
+      "learning_rate": 0.001,
+      "loss": 1.2972,
+      "step": 22109
+    },
+    {
+      "epoch": 0.9353583213469837,
+      "grad_norm": 0.17358072102069855,
+      "learning_rate": 0.001,
+      "loss": 2.4516,
+      "step": 22110
+    },
+    {
+      "epoch": 0.9354006261105,
+      "grad_norm": 0.22711333632469177,
+      "learning_rate": 0.001,
+      "loss": 2.397,
+      "step": 22111
+    },
+    {
+      "epoch": 0.9354429308740164,
+      "grad_norm": 0.151783287525177,
+      "learning_rate": 0.001,
+      "loss": 2.1857,
+      "step": 22112
+    },
+    {
+      "epoch": 0.9354852356375328,
+      "grad_norm": 0.7333983182907104,
+      "learning_rate": 0.001,
+      "loss": 1.6396,
+      "step": 22113
+    },
+    {
+      "epoch": 0.9355275404010491,
+      "grad_norm": 0.5164382457733154,
+      "learning_rate": 0.001,
+      "loss": 1.9616,
+      "step": 22114
+    },
+    {
+      "epoch": 0.9355698451645655,
+      "grad_norm": 0.21135050058364868,
+      "learning_rate": 0.001,
+      "loss": 1.9535,
+      "step": 22115
+    },
+    {
+      "epoch": 0.935612149928082,
+      "grad_norm": 0.19637274742126465,
+      "learning_rate": 0.001,
+      "loss": 2.3697,
+      "step": 22116
+    },
+    {
+      "epoch": 0.9356544546915982,
+      "grad_norm": 0.2197602093219757,
+      "learning_rate": 0.001,
+      "loss": 2.5342,
+      "step": 22117
+    },
+    {
+      "epoch": 0.9356967594551147,
+      "grad_norm": 11.109819412231445,
+      "learning_rate": 0.001,
+      "loss": 1.8034,
+      "step": 22118
+    },
+    {
+      "epoch": 0.9357390642186311,
+      "grad_norm": 0.5363801121711731,
+      "learning_rate": 0.001,
+      "loss": 1.6609,
+      "step": 22119
+    },
+    {
+      "epoch": 0.9357813689821474,
+      "grad_norm": 0.7339990735054016,
+      "learning_rate": 0.001,
+      "loss": 1.8741,
+      "step": 22120
+    },
+    {
+      "epoch": 0.9358236737456638,
+      "grad_norm": 0.1876104772090912,
+      "learning_rate": 0.001,
+      "loss": 1.6976,
+      "step": 22121
+    },
+    {
+      "epoch": 0.9358659785091802,
+      "grad_norm": 6.681327819824219,
+      "learning_rate": 0.001,
+      "loss": 2.7797,
+      "step": 22122
+    },
+    {
+      "epoch": 0.9359082832726965,
+      "grad_norm": 0.15576091408729553,
+      "learning_rate": 0.001,
+      "loss": 1.8506,
+      "step": 22123
+    },
+    {
+      "epoch": 0.9359505880362129,
+      "grad_norm": 0.15668316185474396,
+      "learning_rate": 0.001,
+      "loss": 1.701,
+      "step": 22124
+    },
+    {
+      "epoch": 0.9359928927997293,
+      "grad_norm": 0.19054444134235382,
+      "learning_rate": 0.001,
+      "loss": 1.8375,
+      "step": 22125
+    },
+    {
+      "epoch": 0.9360351975632456,
+      "grad_norm": 0.14556585252285004,
+      "learning_rate": 0.001,
+      "loss": 1.7359,
+      "step": 22126
+    },
+    {
+      "epoch": 0.936077502326762,
+      "grad_norm": 1.7815239429473877,
+      "learning_rate": 0.001,
+      "loss": 2.0106,
+      "step": 22127
+    },
+    {
+      "epoch": 0.9361198070902783,
+      "grad_norm": 0.13777472078800201,
+      "learning_rate": 0.001,
+      "loss": 2.0197,
+      "step": 22128
+    },
+    {
+      "epoch": 0.9361621118537947,
+      "grad_norm": 0.1779537945985794,
+      "learning_rate": 0.001,
+      "loss": 2.0832,
+      "step": 22129
+    },
+    {
+      "epoch": 0.9362044166173111,
+      "grad_norm": 0.1431722640991211,
+      "learning_rate": 0.001,
+      "loss": 2.2073,
+      "step": 22130
+    },
+    {
+      "epoch": 0.9362467213808274,
+      "grad_norm": 0.5513030290603638,
+      "learning_rate": 0.001,
+      "loss": 1.8146,
+      "step": 22131
+    },
+    {
+      "epoch": 0.9362890261443438,
+      "grad_norm": 0.14709900319576263,
+      "learning_rate": 0.001,
+      "loss": 2.2721,
+      "step": 22132
+    },
+    {
+      "epoch": 0.9363313309078602,
+      "grad_norm": 0.13920997083187103,
+      "learning_rate": 0.001,
+      "loss": 1.7674,
+      "step": 22133
+    },
+    {
+      "epoch": 0.9363736356713765,
+      "grad_norm": 0.13396479189395905,
+      "learning_rate": 0.001,
+      "loss": 1.5546,
+      "step": 22134
+    },
+    {
+      "epoch": 0.936415940434893,
+      "grad_norm": 0.14524602890014648,
+      "learning_rate": 0.001,
+      "loss": 1.9367,
+      "step": 22135
+    },
+    {
+      "epoch": 0.9364582451984094,
+      "grad_norm": 0.14268428087234497,
+      "learning_rate": 0.001,
+      "loss": 1.7628,
+      "step": 22136
+    },
+    {
+      "epoch": 0.9365005499619257,
+      "grad_norm": 0.16720207035541534,
+      "learning_rate": 0.001,
+      "loss": 2.0936,
+      "step": 22137
+    },
+    {
+      "epoch": 0.9365428547254421,
+      "grad_norm": 0.1389460563659668,
+      "learning_rate": 0.001,
+      "loss": 1.5024,
+      "step": 22138
+    },
+    {
+      "epoch": 0.9365851594889585,
+      "grad_norm": 0.1430455595254898,
+      "learning_rate": 0.001,
+      "loss": 3.3252,
+      "step": 22139
+    },
+    {
+      "epoch": 0.9366274642524748,
+      "grad_norm": 0.2216232270002365,
+      "learning_rate": 0.001,
+      "loss": 1.4568,
+      "step": 22140
+    },
+    {
+      "epoch": 0.9366697690159912,
+      "grad_norm": 0.1489740014076233,
+      "learning_rate": 0.001,
+      "loss": 2.1509,
+      "step": 22141
+    },
+    {
+      "epoch": 0.9367120737795076,
+      "grad_norm": 0.16746602952480316,
+      "learning_rate": 0.001,
+      "loss": 1.7385,
+      "step": 22142
+    },
+    {
+      "epoch": 0.9367543785430239,
+      "grad_norm": 1.196807861328125,
+      "learning_rate": 0.001,
+      "loss": 2.4886,
+      "step": 22143
+    },
+    {
+      "epoch": 0.9367966833065403,
+      "grad_norm": 0.15129493176937103,
+      "learning_rate": 0.001,
+      "loss": 2.6034,
+      "step": 22144
+    },
+    {
+      "epoch": 0.9368389880700567,
+      "grad_norm": 0.17672112584114075,
+      "learning_rate": 0.001,
+      "loss": 3.1263,
+      "step": 22145
+    },
+    {
+      "epoch": 0.936881292833573,
+      "grad_norm": 0.16527965664863586,
+      "learning_rate": 0.001,
+      "loss": 1.997,
+      "step": 22146
+    },
+    {
+      "epoch": 0.9369235975970894,
+      "grad_norm": 0.8040674328804016,
+      "learning_rate": 0.001,
+      "loss": 2.0254,
+      "step": 22147
+    },
+    {
+      "epoch": 0.9369659023606058,
+      "grad_norm": 0.1395406723022461,
+      "learning_rate": 0.001,
+      "loss": 2.0166,
+      "step": 22148
+    },
+    {
+      "epoch": 0.9370082071241221,
+      "grad_norm": 0.14435064792633057,
+      "learning_rate": 0.001,
+      "loss": 1.8352,
+      "step": 22149
+    },
+    {
+      "epoch": 0.9370505118876385,
+      "grad_norm": 0.14937026798725128,
+      "learning_rate": 0.001,
+      "loss": 4.2509,
+      "step": 22150
+    },
+    {
+      "epoch": 0.937092816651155,
+      "grad_norm": 0.22345872223377228,
+      "learning_rate": 0.001,
+      "loss": 2.0077,
+      "step": 22151
+    },
+    {
+      "epoch": 0.9371351214146713,
+      "grad_norm": 0.14060348272323608,
+      "learning_rate": 0.001,
+      "loss": 1.376,
+      "step": 22152
+    },
+    {
+      "epoch": 0.9371774261781877,
+      "grad_norm": 0.3378833532333374,
+      "learning_rate": 0.001,
+      "loss": 2.4393,
+      "step": 22153
+    },
+    {
+      "epoch": 0.9372197309417041,
+      "grad_norm": 0.16241557896137238,
+      "learning_rate": 0.001,
+      "loss": 2.1142,
+      "step": 22154
+    },
+    {
+      "epoch": 0.9372620357052204,
+      "grad_norm": 0.15988023579120636,
+      "learning_rate": 0.001,
+      "loss": 1.9314,
+      "step": 22155
+    },
+    {
+      "epoch": 0.9373043404687368,
+      "grad_norm": 0.15488198399543762,
+      "learning_rate": 0.001,
+      "loss": 1.4996,
+      "step": 22156
+    },
+    {
+      "epoch": 0.9373466452322532,
+      "grad_norm": 0.157693013548851,
+      "learning_rate": 0.001,
+      "loss": 2.2934,
+      "step": 22157
+    },
+    {
+      "epoch": 0.9373889499957695,
+      "grad_norm": 0.16368556022644043,
+      "learning_rate": 0.001,
+      "loss": 1.5275,
+      "step": 22158
+    },
+    {
+      "epoch": 0.9374312547592859,
+      "grad_norm": 0.20545239746570587,
+      "learning_rate": 0.001,
+      "loss": 1.5569,
+      "step": 22159
+    },
+    {
+      "epoch": 0.9374735595228023,
+      "grad_norm": 0.2001711130142212,
+      "learning_rate": 0.001,
+      "loss": 2.5836,
+      "step": 22160
+    },
+    {
+      "epoch": 0.9375158642863186,
+      "grad_norm": 0.3313537538051605,
+      "learning_rate": 0.001,
+      "loss": 3.5361,
+      "step": 22161
+    },
+    {
+      "epoch": 0.937558169049835,
+      "grad_norm": 0.1719702184200287,
+      "learning_rate": 0.001,
+      "loss": 1.6831,
+      "step": 22162
+    },
+    {
+      "epoch": 0.9376004738133514,
+      "grad_norm": 0.1592639684677124,
+      "learning_rate": 0.001,
+      "loss": 1.5466,
+      "step": 22163
+    },
+    {
+      "epoch": 0.9376427785768677,
+      "grad_norm": 0.13352732360363007,
+      "learning_rate": 0.001,
+      "loss": 1.5322,
+      "step": 22164
+    },
+    {
+      "epoch": 0.9376850833403841,
+      "grad_norm": 0.18848414719104767,
+      "learning_rate": 0.001,
+      "loss": 2.1188,
+      "step": 22165
+    },
+    {
+      "epoch": 0.9377273881039005,
+      "grad_norm": 0.14446255564689636,
+      "learning_rate": 0.001,
+      "loss": 2.4954,
+      "step": 22166
+    },
+    {
+      "epoch": 0.9377696928674168,
+      "grad_norm": 0.1566217541694641,
+      "learning_rate": 0.001,
+      "loss": 1.6525,
+      "step": 22167
+    },
+    {
+      "epoch": 0.9378119976309333,
+      "grad_norm": 0.15236781537532806,
+      "learning_rate": 0.001,
+      "loss": 1.9077,
+      "step": 22168
+    },
+    {
+      "epoch": 0.9378543023944497,
+      "grad_norm": 0.4224134385585785,
+      "learning_rate": 0.001,
+      "loss": 2.2906,
+      "step": 22169
+    },
+    {
+      "epoch": 0.937896607157966,
+      "grad_norm": 0.6217629313468933,
+      "learning_rate": 0.001,
+      "loss": 2.5966,
+      "step": 22170
+    },
+    {
+      "epoch": 0.9379389119214824,
+      "grad_norm": 0.17094938457012177,
+      "learning_rate": 0.001,
+      "loss": 2.0306,
+      "step": 22171
+    },
+    {
+      "epoch": 0.9379812166849987,
+      "grad_norm": 0.18009664118289948,
+      "learning_rate": 0.001,
+      "loss": 1.8595,
+      "step": 22172
+    },
+    {
+      "epoch": 0.9380235214485151,
+      "grad_norm": 0.1493372768163681,
+      "learning_rate": 0.001,
+      "loss": 1.5855,
+      "step": 22173
+    },
+    {
+      "epoch": 0.9380658262120315,
+      "grad_norm": 0.13836872577667236,
+      "learning_rate": 0.001,
+      "loss": 2.7204,
+      "step": 22174
+    },
+    {
+      "epoch": 0.9381081309755478,
+      "grad_norm": 0.15783464908599854,
+      "learning_rate": 0.001,
+      "loss": 2.2411,
+      "step": 22175
+    },
+    {
+      "epoch": 0.9381504357390642,
+      "grad_norm": 0.2531960606575012,
+      "learning_rate": 0.001,
+      "loss": 2.9065,
+      "step": 22176
+    },
+    {
+      "epoch": 0.9381927405025806,
+      "grad_norm": 0.18750788271427155,
+      "learning_rate": 0.001,
+      "loss": 2.6953,
+      "step": 22177
+    },
+    {
+      "epoch": 0.9382350452660969,
+      "grad_norm": 0.1514870524406433,
+      "learning_rate": 0.001,
+      "loss": 2.5924,
+      "step": 22178
+    },
+    {
+      "epoch": 0.9382773500296133,
+      "grad_norm": 0.1663045436143875,
+      "learning_rate": 0.001,
+      "loss": 3.2424,
+      "step": 22179
+    },
+    {
+      "epoch": 0.9383196547931297,
+      "grad_norm": 0.12670683860778809,
+      "learning_rate": 0.001,
+      "loss": 1.8389,
+      "step": 22180
+    },
+    {
+      "epoch": 0.938361959556646,
+      "grad_norm": 0.16413888335227966,
+      "learning_rate": 0.001,
+      "loss": 1.6121,
+      "step": 22181
+    },
+    {
+      "epoch": 0.9384042643201624,
+      "grad_norm": 0.1846090853214264,
+      "learning_rate": 0.001,
+      "loss": 2.6834,
+      "step": 22182
+    },
+    {
+      "epoch": 0.9384465690836788,
+      "grad_norm": 0.17222410440444946,
+      "learning_rate": 0.001,
+      "loss": 1.5327,
+      "step": 22183
+    },
+    {
+      "epoch": 0.9384888738471951,
+      "grad_norm": 0.18798105418682098,
+      "learning_rate": 0.001,
+      "loss": 2.3528,
+      "step": 22184
+    },
+    {
+      "epoch": 0.9385311786107116,
+      "grad_norm": 0.12880779802799225,
+      "learning_rate": 0.001,
+      "loss": 3.7122,
+      "step": 22185
+    },
+    {
+      "epoch": 0.938573483374228,
+      "grad_norm": 0.15221832692623138,
+      "learning_rate": 0.001,
+      "loss": 2.1376,
+      "step": 22186
+    },
+    {
+      "epoch": 0.9386157881377443,
+      "grad_norm": 0.2237653285264969,
+      "learning_rate": 0.001,
+      "loss": 3.0669,
+      "step": 22187
+    },
+    {
+      "epoch": 0.9386580929012607,
+      "grad_norm": 0.17113400995731354,
+      "learning_rate": 0.001,
+      "loss": 1.6939,
+      "step": 22188
+    },
+    {
+      "epoch": 0.9387003976647771,
+      "grad_norm": 0.1556374728679657,
+      "learning_rate": 0.001,
+      "loss": 2.1034,
+      "step": 22189
+    },
+    {
+      "epoch": 0.9387427024282934,
+      "grad_norm": 0.14844928681850433,
+      "learning_rate": 0.001,
+      "loss": 1.7177,
+      "step": 22190
+    },
+    {
+      "epoch": 0.9387850071918098,
+      "grad_norm": 0.7886921763420105,
+      "learning_rate": 0.001,
+      "loss": 2.4152,
+      "step": 22191
+    },
+    {
+      "epoch": 0.9388273119553262,
+      "grad_norm": 0.17194926738739014,
+      "learning_rate": 0.001,
+      "loss": 2.1804,
+      "step": 22192
+    },
+    {
+      "epoch": 0.9388696167188425,
+      "grad_norm": 0.17443875968456268,
+      "learning_rate": 0.001,
+      "loss": 1.7619,
+      "step": 22193
+    },
+    {
+      "epoch": 0.9389119214823589,
+      "grad_norm": 0.17937207221984863,
+      "learning_rate": 0.001,
+      "loss": 1.8103,
+      "step": 22194
+    },
+    {
+      "epoch": 0.9389542262458753,
+      "grad_norm": 0.16381466388702393,
+      "learning_rate": 0.001,
+      "loss": 2.595,
+      "step": 22195
+    },
+    {
+      "epoch": 0.9389965310093916,
+      "grad_norm": 0.15668261051177979,
+      "learning_rate": 0.001,
+      "loss": 2.379,
+      "step": 22196
+    },
+    {
+      "epoch": 0.939038835772908,
+      "grad_norm": 0.1664295494556427,
+      "learning_rate": 0.001,
+      "loss": 1.8999,
+      "step": 22197
+    },
+    {
+      "epoch": 0.9390811405364244,
+      "grad_norm": 0.20704372227191925,
+      "learning_rate": 0.001,
+      "loss": 2.2593,
+      "step": 22198
+    },
+    {
+      "epoch": 0.9391234452999407,
+      "grad_norm": 0.17788812518119812,
+      "learning_rate": 0.001,
+      "loss": 2.0697,
+      "step": 22199
+    },
+    {
+      "epoch": 0.9391657500634571,
+      "grad_norm": 0.20091095566749573,
+      "learning_rate": 0.001,
+      "loss": 2.408,
+      "step": 22200
+    },
+    {
+      "epoch": 0.9392080548269736,
+      "grad_norm": 0.16119319200515747,
+      "learning_rate": 0.001,
+      "loss": 2.2376,
+      "step": 22201
+    },
+    {
+      "epoch": 0.9392503595904899,
+      "grad_norm": 0.16883371770381927,
+      "learning_rate": 0.001,
+      "loss": 1.7975,
+      "step": 22202
+    },
+    {
+      "epoch": 0.9392926643540063,
+      "grad_norm": 0.15238231420516968,
+      "learning_rate": 0.001,
+      "loss": 2.2641,
+      "step": 22203
+    },
+    {
+      "epoch": 0.9393349691175227,
+      "grad_norm": 0.14293359220027924,
+      "learning_rate": 0.001,
+      "loss": 1.9112,
+      "step": 22204
+    },
+    {
+      "epoch": 0.939377273881039,
+      "grad_norm": 0.41308876872062683,
+      "learning_rate": 0.001,
+      "loss": 2.4264,
+      "step": 22205
+    },
+    {
+      "epoch": 0.9394195786445554,
+      "grad_norm": 0.1429714858531952,
+      "learning_rate": 0.001,
+      "loss": 2.8803,
+      "step": 22206
+    },
+    {
+      "epoch": 0.9394618834080718,
+      "grad_norm": 0.16283197700977325,
+      "learning_rate": 0.001,
+      "loss": 1.6608,
+      "step": 22207
+    },
+    {
+      "epoch": 0.9395041881715881,
+      "grad_norm": 0.18357500433921814,
+      "learning_rate": 0.001,
+      "loss": 2.3365,
+      "step": 22208
+    },
+    {
+      "epoch": 0.9395464929351045,
+      "grad_norm": 0.1459013670682907,
+      "learning_rate": 0.001,
+      "loss": 1.9035,
+      "step": 22209
+    },
+    {
+      "epoch": 0.9395887976986209,
+      "grad_norm": 0.13705354928970337,
+      "learning_rate": 0.001,
+      "loss": 2.3804,
+      "step": 22210
+    },
+    {
+      "epoch": 0.9396311024621372,
+      "grad_norm": 0.15534013509750366,
+      "learning_rate": 0.001,
+      "loss": 2.1509,
+      "step": 22211
+    },
+    {
+      "epoch": 0.9396734072256536,
+      "grad_norm": 0.19198080897331238,
+      "learning_rate": 0.001,
+      "loss": 1.5588,
+      "step": 22212
+    },
+    {
+      "epoch": 0.93971571198917,
+      "grad_norm": 0.14550134539604187,
+      "learning_rate": 0.001,
+      "loss": 1.6061,
+      "step": 22213
+    },
+    {
+      "epoch": 0.9397580167526863,
+      "grad_norm": 0.1513293832540512,
+      "learning_rate": 0.001,
+      "loss": 2.114,
+      "step": 22214
+    },
+    {
+      "epoch": 0.9398003215162027,
+      "grad_norm": 6.008566379547119,
+      "learning_rate": 0.001,
+      "loss": 1.6516,
+      "step": 22215
+    },
+    {
+      "epoch": 0.9398426262797192,
+      "grad_norm": 0.14722181856632233,
+      "learning_rate": 0.001,
+      "loss": 1.2652,
+      "step": 22216
+    },
+    {
+      "epoch": 0.9398849310432354,
+      "grad_norm": 0.16536396741867065,
+      "learning_rate": 0.001,
+      "loss": 1.9738,
+      "step": 22217
+    },
+    {
+      "epoch": 0.9399272358067519,
+      "grad_norm": 0.1685878336429596,
+      "learning_rate": 0.001,
+      "loss": 1.5726,
+      "step": 22218
+    },
+    {
+      "epoch": 0.9399695405702682,
+      "grad_norm": 0.19034594297409058,
+      "learning_rate": 0.001,
+      "loss": 2.2402,
+      "step": 22219
+    },
+    {
+      "epoch": 0.9400118453337846,
+      "grad_norm": 0.16235460340976715,
+      "learning_rate": 0.001,
+      "loss": 2.2383,
+      "step": 22220
+    },
+    {
+      "epoch": 0.940054150097301,
+      "grad_norm": 0.1350528448820114,
+      "learning_rate": 0.001,
+      "loss": 3.5003,
+      "step": 22221
+    },
+    {
+      "epoch": 0.9400964548608173,
+      "grad_norm": 0.5780348181724548,
+      "learning_rate": 0.001,
+      "loss": 3.6418,
+      "step": 22222
+    },
+    {
+      "epoch": 0.9401387596243337,
+      "grad_norm": 0.9433380365371704,
+      "learning_rate": 0.001,
+      "loss": 2.0663,
+      "step": 22223
+    },
+    {
+      "epoch": 0.9401810643878501,
+      "grad_norm": 1.4721174240112305,
+      "learning_rate": 0.001,
+      "loss": 2.5166,
+      "step": 22224
+    },
+    {
+      "epoch": 0.9402233691513664,
+      "grad_norm": 0.18319587409496307,
+      "learning_rate": 0.001,
+      "loss": 2.239,
+      "step": 22225
+    },
+    {
+      "epoch": 0.9402656739148828,
+      "grad_norm": 0.14535488188266754,
+      "learning_rate": 0.001,
+      "loss": 2.5118,
+      "step": 22226
+    },
+    {
+      "epoch": 0.9403079786783992,
+      "grad_norm": 0.1889243721961975,
+      "learning_rate": 0.001,
+      "loss": 2.082,
+      "step": 22227
+    },
+    {
+      "epoch": 0.9403502834419155,
+      "grad_norm": 0.15350744128227234,
+      "learning_rate": 0.001,
+      "loss": 2.3159,
+      "step": 22228
+    },
+    {
+      "epoch": 0.9403925882054319,
+      "grad_norm": 0.9224516749382019,
+      "learning_rate": 0.001,
+      "loss": 1.3976,
+      "step": 22229
+    },
+    {
+      "epoch": 0.9404348929689483,
+      "grad_norm": 0.132807195186615,
+      "learning_rate": 0.001,
+      "loss": 2.8195,
+      "step": 22230
+    },
+    {
+      "epoch": 0.9404771977324646,
+      "grad_norm": 0.17904257774353027,
+      "learning_rate": 0.001,
+      "loss": 2.3766,
+      "step": 22231
+    },
+    {
+      "epoch": 0.940519502495981,
+      "grad_norm": 0.15651574730873108,
+      "learning_rate": 0.001,
+      "loss": 1.5067,
+      "step": 22232
+    },
+    {
+      "epoch": 0.9405618072594975,
+      "grad_norm": 0.1492500901222229,
+      "learning_rate": 0.001,
+      "loss": 1.3076,
+      "step": 22233
+    },
+    {
+      "epoch": 0.9406041120230137,
+      "grad_norm": 0.177638441324234,
+      "learning_rate": 0.001,
+      "loss": 2.5056,
+      "step": 22234
+    },
+    {
+      "epoch": 0.9406464167865302,
+      "grad_norm": 0.1903133988380432,
+      "learning_rate": 0.001,
+      "loss": 2.0664,
+      "step": 22235
+    },
+    {
+      "epoch": 0.9406887215500466,
+      "grad_norm": 0.2579449415206909,
+      "learning_rate": 0.001,
+      "loss": 1.6951,
+      "step": 22236
+    },
+    {
+      "epoch": 0.9407310263135629,
+      "grad_norm": 0.19664129614830017,
+      "learning_rate": 0.001,
+      "loss": 2.2619,
+      "step": 22237
+    },
+    {
+      "epoch": 0.9407733310770793,
+      "grad_norm": 0.1550266593694687,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 22238
+    },
+    {
+      "epoch": 0.9408156358405957,
+      "grad_norm": 0.3928470015525818,
+      "learning_rate": 0.001,
+      "loss": 2.0354,
+      "step": 22239
+    },
+    {
+      "epoch": 0.940857940604112,
+      "grad_norm": 0.16893982887268066,
+      "learning_rate": 0.001,
+      "loss": 2.6994,
+      "step": 22240
+    },
+    {
+      "epoch": 0.9409002453676284,
+      "grad_norm": 0.16361390054225922,
+      "learning_rate": 0.001,
+      "loss": 2.2244,
+      "step": 22241
+    },
+    {
+      "epoch": 0.9409425501311448,
+      "grad_norm": 0.14021305739879608,
+      "learning_rate": 0.001,
+      "loss": 2.2345,
+      "step": 22242
+    },
+    {
+      "epoch": 0.9409848548946611,
+      "grad_norm": 0.1266973614692688,
+      "learning_rate": 0.001,
+      "loss": 1.6334,
+      "step": 22243
+    },
+    {
+      "epoch": 0.9410271596581775,
+      "grad_norm": 0.16998490691184998,
+      "learning_rate": 0.001,
+      "loss": 1.9222,
+      "step": 22244
+    },
+    {
+      "epoch": 0.9410694644216939,
+      "grad_norm": 0.17645969986915588,
+      "learning_rate": 0.001,
+      "loss": 2.8088,
+      "step": 22245
+    },
+    {
+      "epoch": 0.9411117691852102,
+      "grad_norm": 0.14849461615085602,
+      "learning_rate": 0.001,
+      "loss": 1.4422,
+      "step": 22246
+    },
+    {
+      "epoch": 0.9411540739487266,
+      "grad_norm": 0.12691397964954376,
+      "learning_rate": 0.001,
+      "loss": 1.4996,
+      "step": 22247
+    },
+    {
+      "epoch": 0.941196378712243,
+      "grad_norm": 0.13772864639759064,
+      "learning_rate": 0.001,
+      "loss": 2.2293,
+      "step": 22248
+    },
+    {
+      "epoch": 0.9412386834757593,
+      "grad_norm": 0.15115776658058167,
+      "learning_rate": 0.001,
+      "loss": 1.4696,
+      "step": 22249
+    },
+    {
+      "epoch": 0.9412809882392758,
+      "grad_norm": 0.14529401063919067,
+      "learning_rate": 0.001,
+      "loss": 2.4046,
+      "step": 22250
+    },
+    {
+      "epoch": 0.9413232930027922,
+      "grad_norm": 0.14119921624660492,
+      "learning_rate": 0.001,
+      "loss": 1.5726,
+      "step": 22251
+    },
+    {
+      "epoch": 0.9413655977663085,
+      "grad_norm": 0.31882578134536743,
+      "learning_rate": 0.001,
+      "loss": 2.2414,
+      "step": 22252
+    },
+    {
+      "epoch": 0.9414079025298249,
+      "grad_norm": 0.20864500105381012,
+      "learning_rate": 0.001,
+      "loss": 2.1722,
+      "step": 22253
+    },
+    {
+      "epoch": 0.9414502072933413,
+      "grad_norm": 0.15085260570049286,
+      "learning_rate": 0.001,
+      "loss": 1.5922,
+      "step": 22254
+    },
+    {
+      "epoch": 0.9414925120568576,
+      "grad_norm": 0.15553170442581177,
+      "learning_rate": 0.001,
+      "loss": 1.7225,
+      "step": 22255
+    },
+    {
+      "epoch": 0.941534816820374,
+      "grad_norm": 41.11113739013672,
+      "learning_rate": 0.001,
+      "loss": 1.9813,
+      "step": 22256
+    },
+    {
+      "epoch": 0.9415771215838904,
+      "grad_norm": 0.1472524106502533,
+      "learning_rate": 0.001,
+      "loss": 1.5666,
+      "step": 22257
+    },
+    {
+      "epoch": 0.9416194263474067,
+      "grad_norm": 0.15531371533870697,
+      "learning_rate": 0.001,
+      "loss": 2.4356,
+      "step": 22258
+    },
+    {
+      "epoch": 0.9416617311109231,
+      "grad_norm": 0.158295139670372,
+      "learning_rate": 0.001,
+      "loss": 1.8226,
+      "step": 22259
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 0.1623242199420929,
+      "learning_rate": 0.001,
+      "loss": 2.2772,
+      "step": 22260
+    },
+    {
+      "epoch": 0.9417463406379558,
+      "grad_norm": 0.14404623210430145,
+      "learning_rate": 0.001,
+      "loss": 2.1139,
+      "step": 22261
+    },
+    {
+      "epoch": 0.9417886454014722,
+      "grad_norm": 0.15782445669174194,
+      "learning_rate": 0.001,
+      "loss": 1.8102,
+      "step": 22262
+    },
+    {
+      "epoch": 0.9418309501649885,
+      "grad_norm": 0.15719851851463318,
+      "learning_rate": 0.001,
+      "loss": 1.3974,
+      "step": 22263
+    },
+    {
+      "epoch": 0.9418732549285049,
+      "grad_norm": 0.17760580778121948,
+      "learning_rate": 0.001,
+      "loss": 1.7701,
+      "step": 22264
+    },
+    {
+      "epoch": 0.9419155596920213,
+      "grad_norm": 0.1664871722459793,
+      "learning_rate": 0.001,
+      "loss": 2.3184,
+      "step": 22265
+    },
+    {
+      "epoch": 0.9419578644555376,
+      "grad_norm": 0.47373244166374207,
+      "learning_rate": 0.001,
+      "loss": 2.0407,
+      "step": 22266
+    },
+    {
+      "epoch": 0.942000169219054,
+      "grad_norm": 0.5430980324745178,
+      "learning_rate": 0.001,
+      "loss": 1.881,
+      "step": 22267
+    },
+    {
+      "epoch": 0.9420424739825705,
+      "grad_norm": 0.17633113265037537,
+      "learning_rate": 0.001,
+      "loss": 1.8635,
+      "step": 22268
+    },
+    {
+      "epoch": 0.9420847787460868,
+      "grad_norm": 0.1332206279039383,
+      "learning_rate": 0.001,
+      "loss": 2.9,
+      "step": 22269
+    },
+    {
+      "epoch": 0.9421270835096032,
+      "grad_norm": 0.17408061027526855,
+      "learning_rate": 0.001,
+      "loss": 2.2226,
+      "step": 22270
+    },
+    {
+      "epoch": 0.9421693882731196,
+      "grad_norm": 0.19134634733200073,
+      "learning_rate": 0.001,
+      "loss": 3.0214,
+      "step": 22271
+    },
+    {
+      "epoch": 0.9422116930366359,
+      "grad_norm": 0.16139277815818787,
+      "learning_rate": 0.001,
+      "loss": 3.3642,
+      "step": 22272
+    },
+    {
+      "epoch": 0.9422539978001523,
+      "grad_norm": 0.14892078936100006,
+      "learning_rate": 0.001,
+      "loss": 2.0493,
+      "step": 22273
+    },
+    {
+      "epoch": 0.9422963025636687,
+      "grad_norm": 0.18064714968204498,
+      "learning_rate": 0.001,
+      "loss": 1.9058,
+      "step": 22274
+    },
+    {
+      "epoch": 0.942338607327185,
+      "grad_norm": 4.666594505310059,
+      "learning_rate": 0.001,
+      "loss": 1.9391,
+      "step": 22275
+    },
+    {
+      "epoch": 0.9423809120907014,
+      "grad_norm": 0.16249027848243713,
+      "learning_rate": 0.001,
+      "loss": 2.5574,
+      "step": 22276
+    },
+    {
+      "epoch": 0.9424232168542178,
+      "grad_norm": 0.14459872245788574,
+      "learning_rate": 0.001,
+      "loss": 1.8483,
+      "step": 22277
+    },
+    {
+      "epoch": 0.9424655216177341,
+      "grad_norm": 0.1476597934961319,
+      "learning_rate": 0.001,
+      "loss": 1.9867,
+      "step": 22278
+    },
+    {
+      "epoch": 0.9425078263812505,
+      "grad_norm": 0.4886634945869446,
+      "learning_rate": 0.001,
+      "loss": 2.0321,
+      "step": 22279
+    },
+    {
+      "epoch": 0.9425501311447669,
+      "grad_norm": 0.23067691922187805,
+      "learning_rate": 0.001,
+      "loss": 2.1118,
+      "step": 22280
+    },
+    {
+      "epoch": 0.9425924359082832,
+      "grad_norm": 0.14237605035305023,
+      "learning_rate": 0.001,
+      "loss": 1.9353,
+      "step": 22281
+    },
+    {
+      "epoch": 0.9426347406717996,
+      "grad_norm": 0.1730206310749054,
+      "learning_rate": 0.001,
+      "loss": 2.1826,
+      "step": 22282
+    },
+    {
+      "epoch": 0.942677045435316,
+      "grad_norm": 0.14405101537704468,
+      "learning_rate": 0.001,
+      "loss": 3.3043,
+      "step": 22283
+    },
+    {
+      "epoch": 0.9427193501988324,
+      "grad_norm": 0.2477824091911316,
+      "learning_rate": 0.001,
+      "loss": 3.2702,
+      "step": 22284
+    },
+    {
+      "epoch": 0.9427616549623488,
+      "grad_norm": 0.12695826590061188,
+      "learning_rate": 0.001,
+      "loss": 2.3018,
+      "step": 22285
+    },
+    {
+      "epoch": 0.9428039597258652,
+      "grad_norm": 2.1101129055023193,
+      "learning_rate": 0.001,
+      "loss": 2.0449,
+      "step": 22286
+    },
+    {
+      "epoch": 0.9428462644893815,
+      "grad_norm": 0.13695190846920013,
+      "learning_rate": 0.001,
+      "loss": 2.5038,
+      "step": 22287
+    },
+    {
+      "epoch": 0.9428885692528979,
+      "grad_norm": 0.1824672371149063,
+      "learning_rate": 0.001,
+      "loss": 2.4115,
+      "step": 22288
+    },
+    {
+      "epoch": 0.9429308740164143,
+      "grad_norm": 0.1646248996257782,
+      "learning_rate": 0.001,
+      "loss": 4.0476,
+      "step": 22289
+    },
+    {
+      "epoch": 0.9429731787799306,
+      "grad_norm": 0.13337703049182892,
+      "learning_rate": 0.001,
+      "loss": 2.0791,
+      "step": 22290
+    },
+    {
+      "epoch": 0.943015483543447,
+      "grad_norm": 0.13796573877334595,
+      "learning_rate": 0.001,
+      "loss": 1.7621,
+      "step": 22291
+    },
+    {
+      "epoch": 0.9430577883069634,
+      "grad_norm": 0.16016069054603577,
+      "learning_rate": 0.001,
+      "loss": 1.8191,
+      "step": 22292
+    },
+    {
+      "epoch": 0.9431000930704797,
+      "grad_norm": 0.15696018934249878,
+      "learning_rate": 0.001,
+      "loss": 1.7768,
+      "step": 22293
+    },
+    {
+      "epoch": 0.9431423978339961,
+      "grad_norm": 0.2677614092826843,
+      "learning_rate": 0.001,
+      "loss": 1.8438,
+      "step": 22294
+    },
+    {
+      "epoch": 0.9431847025975125,
+      "grad_norm": 0.17819730937480927,
+      "learning_rate": 0.001,
+      "loss": 1.4621,
+      "step": 22295
+    },
+    {
+      "epoch": 0.9432270073610288,
+      "grad_norm": 0.9130420088768005,
+      "learning_rate": 0.001,
+      "loss": 3.1412,
+      "step": 22296
+    },
+    {
+      "epoch": 0.9432693121245452,
+      "grad_norm": 0.1402437835931778,
+      "learning_rate": 0.001,
+      "loss": 2.0391,
+      "step": 22297
+    },
+    {
+      "epoch": 0.9433116168880616,
+      "grad_norm": 1.1969804763793945,
+      "learning_rate": 0.001,
+      "loss": 2.7918,
+      "step": 22298
+    },
+    {
+      "epoch": 0.9433539216515779,
+      "grad_norm": 0.7922569513320923,
+      "learning_rate": 0.001,
+      "loss": 1.9959,
+      "step": 22299
+    },
+    {
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.17550984025001526,
+      "learning_rate": 0.001,
+      "loss": 1.9096,
+      "step": 22300
+    },
+    {
+      "epoch": 0.9434385311786108,
+      "grad_norm": 0.14983303844928741,
+      "learning_rate": 0.001,
+      "loss": 1.9521,
+      "step": 22301
+    },
+    {
+      "epoch": 0.9434808359421271,
+      "grad_norm": 0.14537028968334198,
+      "learning_rate": 0.001,
+      "loss": 2.2255,
+      "step": 22302
+    },
+    {
+      "epoch": 0.9435231407056435,
+      "grad_norm": 0.18867170810699463,
+      "learning_rate": 0.001,
+      "loss": 1.7019,
+      "step": 22303
+    },
+    {
+      "epoch": 0.9435654454691599,
+      "grad_norm": 0.183610200881958,
+      "learning_rate": 0.001,
+      "loss": 1.4077,
+      "step": 22304
+    },
+    {
+      "epoch": 0.9436077502326762,
+      "grad_norm": 0.15308107435703278,
+      "learning_rate": 0.001,
+      "loss": 1.8193,
+      "step": 22305
+    },
+    {
+      "epoch": 0.9436500549961926,
+      "grad_norm": 0.15715880692005157,
+      "learning_rate": 0.001,
+      "loss": 1.7452,
+      "step": 22306
+    },
+    {
+      "epoch": 0.943692359759709,
+      "grad_norm": 0.15837711095809937,
+      "learning_rate": 0.001,
+      "loss": 1.8037,
+      "step": 22307
+    },
+    {
+      "epoch": 0.9437346645232253,
+      "grad_norm": 0.1577230989933014,
+      "learning_rate": 0.001,
+      "loss": 1.6906,
+      "step": 22308
+    },
+    {
+      "epoch": 0.9437769692867417,
+      "grad_norm": 0.1631125658750534,
+      "learning_rate": 0.001,
+      "loss": 1.8791,
+      "step": 22309
+    },
+    {
+      "epoch": 0.943819274050258,
+      "grad_norm": 0.15031826496124268,
+      "learning_rate": 0.001,
+      "loss": 3.3296,
+      "step": 22310
+    },
+    {
+      "epoch": 0.9438615788137744,
+      "grad_norm": 0.15790674090385437,
+      "learning_rate": 0.001,
+      "loss": 1.7894,
+      "step": 22311
+    },
+    {
+      "epoch": 0.9439038835772908,
+      "grad_norm": 0.16160251200199127,
+      "learning_rate": 0.001,
+      "loss": 2.2725,
+      "step": 22312
+    },
+    {
+      "epoch": 0.9439461883408071,
+      "grad_norm": 0.14644403755664825,
+      "learning_rate": 0.001,
+      "loss": 2.6085,
+      "step": 22313
+    },
+    {
+      "epoch": 0.9439884931043235,
+      "grad_norm": 0.14695337414741516,
+      "learning_rate": 0.001,
+      "loss": 1.806,
+      "step": 22314
+    },
+    {
+      "epoch": 0.9440307978678399,
+      "grad_norm": 0.22415880858898163,
+      "learning_rate": 0.001,
+      "loss": 1.6828,
+      "step": 22315
+    },
+    {
+      "epoch": 0.9440731026313562,
+      "grad_norm": 0.1559433937072754,
+      "learning_rate": 0.001,
+      "loss": 2.0603,
+      "step": 22316
+    },
+    {
+      "epoch": 0.9441154073948727,
+      "grad_norm": 0.14911769330501556,
+      "learning_rate": 0.001,
+      "loss": 3.1866,
+      "step": 22317
+    },
+    {
+      "epoch": 0.9441577121583891,
+      "grad_norm": 0.17388781905174255,
+      "learning_rate": 0.001,
+      "loss": 2.1254,
+      "step": 22318
+    },
+    {
+      "epoch": 0.9442000169219054,
+      "grad_norm": 1.0141180753707886,
+      "learning_rate": 0.001,
+      "loss": 2.1553,
+      "step": 22319
+    },
+    {
+      "epoch": 0.9442423216854218,
+      "grad_norm": 0.13785704970359802,
+      "learning_rate": 0.001,
+      "loss": 1.6176,
+      "step": 22320
+    },
+    {
+      "epoch": 0.9442846264489382,
+      "grad_norm": 1.1813472509384155,
+      "learning_rate": 0.001,
+      "loss": 1.9315,
+      "step": 22321
+    },
+    {
+      "epoch": 0.9443269312124545,
+      "grad_norm": 0.15482167899608612,
+      "learning_rate": 0.001,
+      "loss": 2.5054,
+      "step": 22322
+    },
+    {
+      "epoch": 0.9443692359759709,
+      "grad_norm": 0.164631187915802,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 22323
+    },
+    {
+      "epoch": 0.9444115407394873,
+      "grad_norm": 0.18228505551815033,
+      "learning_rate": 0.001,
+      "loss": 2.0032,
+      "step": 22324
+    },
+    {
+      "epoch": 0.9444538455030036,
+      "grad_norm": 0.2625649571418762,
+      "learning_rate": 0.001,
+      "loss": 2.6495,
+      "step": 22325
+    },
+    {
+      "epoch": 0.94449615026652,
+      "grad_norm": 0.2495313137769699,
+      "learning_rate": 0.001,
+      "loss": 1.887,
+      "step": 22326
+    },
+    {
+      "epoch": 0.9445384550300364,
+      "grad_norm": 0.16760574281215668,
+      "learning_rate": 0.001,
+      "loss": 1.9866,
+      "step": 22327
+    },
+    {
+      "epoch": 0.9445807597935527,
+      "grad_norm": 0.16984060406684875,
+      "learning_rate": 0.001,
+      "loss": 2.2137,
+      "step": 22328
+    },
+    {
+      "epoch": 0.9446230645570691,
+      "grad_norm": 0.20226795971393585,
+      "learning_rate": 0.001,
+      "loss": 2.0092,
+      "step": 22329
+    },
+    {
+      "epoch": 0.9446653693205855,
+      "grad_norm": 0.24898548424243927,
+      "learning_rate": 0.001,
+      "loss": 2.6886,
+      "step": 22330
+    },
+    {
+      "epoch": 0.9447076740841018,
+      "grad_norm": 0.40224114060401917,
+      "learning_rate": 0.001,
+      "loss": 3.2176,
+      "step": 22331
+    },
+    {
+      "epoch": 0.9447499788476182,
+      "grad_norm": 0.18034416437149048,
+      "learning_rate": 0.001,
+      "loss": 1.7549,
+      "step": 22332
+    },
+    {
+      "epoch": 0.9447922836111347,
+      "grad_norm": 0.14370644092559814,
+      "learning_rate": 0.001,
+      "loss": 1.5195,
+      "step": 22333
+    },
+    {
+      "epoch": 0.944834588374651,
+      "grad_norm": 0.7416594624519348,
+      "learning_rate": 0.001,
+      "loss": 1.7912,
+      "step": 22334
+    },
+    {
+      "epoch": 0.9448768931381674,
+      "grad_norm": 0.15781673789024353,
+      "learning_rate": 0.001,
+      "loss": 2.0478,
+      "step": 22335
+    },
+    {
+      "epoch": 0.9449191979016838,
+      "grad_norm": 0.17141476273536682,
+      "learning_rate": 0.001,
+      "loss": 2.7985,
+      "step": 22336
+    },
+    {
+      "epoch": 0.9449615026652001,
+      "grad_norm": 0.15994961559772491,
+      "learning_rate": 0.001,
+      "loss": 2.1938,
+      "step": 22337
+    },
+    {
+      "epoch": 0.9450038074287165,
+      "grad_norm": 0.3029901385307312,
+      "learning_rate": 0.001,
+      "loss": 1.5779,
+      "step": 22338
+    },
+    {
+      "epoch": 0.9450461121922329,
+      "grad_norm": 0.20497363805770874,
+      "learning_rate": 0.001,
+      "loss": 2.5866,
+      "step": 22339
+    },
+    {
+      "epoch": 0.9450884169557492,
+      "grad_norm": 0.1555112898349762,
+      "learning_rate": 0.001,
+      "loss": 3.3234,
+      "step": 22340
+    },
+    {
+      "epoch": 0.9451307217192656,
+      "grad_norm": 0.18291279673576355,
+      "learning_rate": 0.001,
+      "loss": 2.5366,
+      "step": 22341
+    },
+    {
+      "epoch": 0.945173026482782,
+      "grad_norm": 0.16539302468299866,
+      "learning_rate": 0.001,
+      "loss": 3.267,
+      "step": 22342
+    },
+    {
+      "epoch": 0.9452153312462983,
+      "grad_norm": 0.15976227819919586,
+      "learning_rate": 0.001,
+      "loss": 2.2329,
+      "step": 22343
+    },
+    {
+      "epoch": 0.9452576360098147,
+      "grad_norm": 0.262100487947464,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 22344
+    },
+    {
+      "epoch": 0.9452999407733311,
+      "grad_norm": 1.0867878198623657,
+      "learning_rate": 0.001,
+      "loss": 2.03,
+      "step": 22345
+    },
+    {
+      "epoch": 0.9453422455368474,
+      "grad_norm": 0.16407005488872528,
+      "learning_rate": 0.001,
+      "loss": 2.5212,
+      "step": 22346
+    },
+    {
+      "epoch": 0.9453845503003638,
+      "grad_norm": 0.1369040310382843,
+      "learning_rate": 0.001,
+      "loss": 2.0789,
+      "step": 22347
+    },
+    {
+      "epoch": 0.9454268550638802,
+      "grad_norm": 0.13893716037273407,
+      "learning_rate": 0.001,
+      "loss": 1.5555,
+      "step": 22348
+    },
+    {
+      "epoch": 0.9454691598273965,
+      "grad_norm": 0.5329959392547607,
+      "learning_rate": 0.001,
+      "loss": 2.0893,
+      "step": 22349
+    },
+    {
+      "epoch": 0.945511464590913,
+      "grad_norm": 0.17351019382476807,
+      "learning_rate": 0.001,
+      "loss": 1.7663,
+      "step": 22350
+    },
+    {
+      "epoch": 0.9455537693544294,
+      "grad_norm": 0.1668737381696701,
+      "learning_rate": 0.001,
+      "loss": 2.5798,
+      "step": 22351
+    },
+    {
+      "epoch": 0.9455960741179457,
+      "grad_norm": 1.1766389608383179,
+      "learning_rate": 0.001,
+      "loss": 1.6858,
+      "step": 22352
+    },
+    {
+      "epoch": 0.9456383788814621,
+      "grad_norm": 0.19534622132778168,
+      "learning_rate": 0.001,
+      "loss": 2.2868,
+      "step": 22353
+    },
+    {
+      "epoch": 0.9456806836449784,
+      "grad_norm": 0.1639569103717804,
+      "learning_rate": 0.001,
+      "loss": 1.7299,
+      "step": 22354
+    },
+    {
+      "epoch": 0.9457229884084948,
+      "grad_norm": 0.14215198159217834,
+      "learning_rate": 0.001,
+      "loss": 2.8559,
+      "step": 22355
+    },
+    {
+      "epoch": 0.9457652931720112,
+      "grad_norm": 0.1718878597021103,
+      "learning_rate": 0.001,
+      "loss": 2.5348,
+      "step": 22356
+    },
+    {
+      "epoch": 0.9458075979355275,
+      "grad_norm": 1.5104109048843384,
+      "learning_rate": 0.001,
+      "loss": 3.4204,
+      "step": 22357
+    },
+    {
+      "epoch": 0.9458499026990439,
+      "grad_norm": 0.18013618886470795,
+      "learning_rate": 0.001,
+      "loss": 3.1525,
+      "step": 22358
+    },
+    {
+      "epoch": 0.9458922074625603,
+      "grad_norm": 0.5852957963943481,
+      "learning_rate": 0.001,
+      "loss": 1.9951,
+      "step": 22359
+    },
+    {
+      "epoch": 0.9459345122260766,
+      "grad_norm": 0.3737950921058655,
+      "learning_rate": 0.001,
+      "loss": 2.5527,
+      "step": 22360
+    },
+    {
+      "epoch": 0.945976816989593,
+      "grad_norm": 0.20486804842948914,
+      "learning_rate": 0.001,
+      "loss": 2.8936,
+      "step": 22361
+    },
+    {
+      "epoch": 0.9460191217531094,
+      "grad_norm": 0.18885180354118347,
+      "learning_rate": 0.001,
+      "loss": 1.9923,
+      "step": 22362
+    },
+    {
+      "epoch": 0.9460614265166257,
+      "grad_norm": 0.1904320865869522,
+      "learning_rate": 0.001,
+      "loss": 1.5084,
+      "step": 22363
+    },
+    {
+      "epoch": 0.9461037312801421,
+      "grad_norm": 0.1865696907043457,
+      "learning_rate": 0.001,
+      "loss": 2.2982,
+      "step": 22364
+    },
+    {
+      "epoch": 0.9461460360436585,
+      "grad_norm": 0.14152313768863678,
+      "learning_rate": 0.001,
+      "loss": 1.7908,
+      "step": 22365
+    },
+    {
+      "epoch": 0.9461883408071748,
+      "grad_norm": 0.16738182306289673,
+      "learning_rate": 0.001,
+      "loss": 2.1825,
+      "step": 22366
+    },
+    {
+      "epoch": 0.9462306455706913,
+      "grad_norm": 0.4649767577648163,
+      "learning_rate": 0.001,
+      "loss": 2.4458,
+      "step": 22367
+    },
+    {
+      "epoch": 0.9462729503342077,
+      "grad_norm": 0.1782921850681305,
+      "learning_rate": 0.001,
+      "loss": 2.2509,
+      "step": 22368
+    },
+    {
+      "epoch": 0.946315255097724,
+      "grad_norm": 0.16340084373950958,
+      "learning_rate": 0.001,
+      "loss": 2.4144,
+      "step": 22369
+    },
+    {
+      "epoch": 0.9463575598612404,
+      "grad_norm": 0.19446922838687897,
+      "learning_rate": 0.001,
+      "loss": 2.3684,
+      "step": 22370
+    },
+    {
+      "epoch": 0.9463998646247568,
+      "grad_norm": 0.20399482548236847,
+      "learning_rate": 0.001,
+      "loss": 1.9716,
+      "step": 22371
+    },
+    {
+      "epoch": 0.9464421693882731,
+      "grad_norm": 0.14847496151924133,
+      "learning_rate": 0.001,
+      "loss": 2.6068,
+      "step": 22372
+    },
+    {
+      "epoch": 0.9464844741517895,
+      "grad_norm": 52.29764175415039,
+      "learning_rate": 0.001,
+      "loss": 2.1968,
+      "step": 22373
+    },
+    {
+      "epoch": 0.9465267789153059,
+      "grad_norm": 0.18324236571788788,
+      "learning_rate": 0.001,
+      "loss": 2.1071,
+      "step": 22374
+    },
+    {
+      "epoch": 0.9465690836788222,
+      "grad_norm": 0.1397266834974289,
+      "learning_rate": 0.001,
+      "loss": 1.4608,
+      "step": 22375
+    },
+    {
+      "epoch": 0.9466113884423386,
+      "grad_norm": 0.19050876796245575,
+      "learning_rate": 0.001,
+      "loss": 1.6181,
+      "step": 22376
+    },
+    {
+      "epoch": 0.946653693205855,
+      "grad_norm": 0.17032888531684875,
+      "learning_rate": 0.001,
+      "loss": 1.9212,
+      "step": 22377
+    },
+    {
+      "epoch": 0.9466959979693713,
+      "grad_norm": 0.1688220053911209,
+      "learning_rate": 0.001,
+      "loss": 1.6201,
+      "step": 22378
+    },
+    {
+      "epoch": 0.9467383027328877,
+      "grad_norm": 0.1576201170682907,
+      "learning_rate": 0.001,
+      "loss": 2.0552,
+      "step": 22379
+    },
+    {
+      "epoch": 0.9467806074964041,
+      "grad_norm": 0.17971685528755188,
+      "learning_rate": 0.001,
+      "loss": 1.8798,
+      "step": 22380
+    },
+    {
+      "epoch": 0.9468229122599204,
+      "grad_norm": 0.21118150651454926,
+      "learning_rate": 0.001,
+      "loss": 2.0006,
+      "step": 22381
+    },
+    {
+      "epoch": 0.9468652170234368,
+      "grad_norm": 0.17267964780330658,
+      "learning_rate": 0.001,
+      "loss": 1.6616,
+      "step": 22382
+    },
+    {
+      "epoch": 0.9469075217869533,
+      "grad_norm": 0.15508294105529785,
+      "learning_rate": 0.001,
+      "loss": 1.9034,
+      "step": 22383
+    },
+    {
+      "epoch": 0.9469498265504696,
+      "grad_norm": 0.2111867219209671,
+      "learning_rate": 0.001,
+      "loss": 1.9499,
+      "step": 22384
+    },
+    {
+      "epoch": 0.946992131313986,
+      "grad_norm": 0.1746092438697815,
+      "learning_rate": 0.001,
+      "loss": 1.4817,
+      "step": 22385
+    },
+    {
+      "epoch": 0.9470344360775024,
+      "grad_norm": 0.33903783559799194,
+      "learning_rate": 0.001,
+      "loss": 1.9695,
+      "step": 22386
+    },
+    {
+      "epoch": 0.9470767408410187,
+      "grad_norm": 0.362958699464798,
+      "learning_rate": 0.001,
+      "loss": 2.04,
+      "step": 22387
+    },
+    {
+      "epoch": 0.9471190456045351,
+      "grad_norm": 0.15616007149219513,
+      "learning_rate": 0.001,
+      "loss": 1.8717,
+      "step": 22388
+    },
+    {
+      "epoch": 0.9471613503680515,
+      "grad_norm": 0.19031497836112976,
+      "learning_rate": 0.001,
+      "loss": 2.02,
+      "step": 22389
+    },
+    {
+      "epoch": 0.9472036551315678,
+      "grad_norm": 0.1764826625585556,
+      "learning_rate": 0.001,
+      "loss": 1.3835,
+      "step": 22390
+    },
+    {
+      "epoch": 0.9472459598950842,
+      "grad_norm": 0.1563934087753296,
+      "learning_rate": 0.001,
+      "loss": 2.3536,
+      "step": 22391
+    },
+    {
+      "epoch": 0.9472882646586006,
+      "grad_norm": 0.15653590857982635,
+      "learning_rate": 0.001,
+      "loss": 2.4971,
+      "step": 22392
+    },
+    {
+      "epoch": 0.9473305694221169,
+      "grad_norm": 0.1690565049648285,
+      "learning_rate": 0.001,
+      "loss": 2.161,
+      "step": 22393
+    },
+    {
+      "epoch": 0.9473728741856333,
+      "grad_norm": 0.1902676671743393,
+      "learning_rate": 0.001,
+      "loss": 1.7483,
+      "step": 22394
+    },
+    {
+      "epoch": 0.9474151789491497,
+      "grad_norm": 0.14070867002010345,
+      "learning_rate": 0.001,
+      "loss": 1.9859,
+      "step": 22395
+    },
+    {
+      "epoch": 0.947457483712666,
+      "grad_norm": 0.1544153243303299,
+      "learning_rate": 0.001,
+      "loss": 1.7372,
+      "step": 22396
+    },
+    {
+      "epoch": 0.9474997884761824,
+      "grad_norm": 0.12571722269058228,
+      "learning_rate": 0.001,
+      "loss": 2.009,
+      "step": 22397
+    },
+    {
+      "epoch": 0.9475420932396987,
+      "grad_norm": 0.18831411004066467,
+      "learning_rate": 0.001,
+      "loss": 2.0311,
+      "step": 22398
+    },
+    {
+      "epoch": 0.9475843980032151,
+      "grad_norm": 0.19460955262184143,
+      "learning_rate": 0.001,
+      "loss": 1.6814,
+      "step": 22399
+    },
+    {
+      "epoch": 0.9476267027667316,
+      "grad_norm": 0.14858324825763702,
+      "learning_rate": 0.001,
+      "loss": 1.8082,
+      "step": 22400
+    },
+    {
+      "epoch": 0.9476690075302479,
+      "grad_norm": 0.18766085803508759,
+      "learning_rate": 0.001,
+      "loss": 2.5125,
+      "step": 22401
+    },
+    {
+      "epoch": 0.9477113122937643,
+      "grad_norm": 0.15538080036640167,
+      "learning_rate": 0.001,
+      "loss": 1.6178,
+      "step": 22402
+    },
+    {
+      "epoch": 0.9477536170572807,
+      "grad_norm": 0.21831777691841125,
+      "learning_rate": 0.001,
+      "loss": 3.1002,
+      "step": 22403
+    },
+    {
+      "epoch": 0.947795921820797,
+      "grad_norm": 0.1795506328344345,
+      "learning_rate": 0.001,
+      "loss": 1.9556,
+      "step": 22404
+    },
+    {
+      "epoch": 0.9478382265843134,
+      "grad_norm": 0.12825722992420197,
+      "learning_rate": 0.001,
+      "loss": 2.2971,
+      "step": 22405
+    },
+    {
+      "epoch": 0.9478805313478298,
+      "grad_norm": 0.16179698705673218,
+      "learning_rate": 0.001,
+      "loss": 2.6207,
+      "step": 22406
+    },
+    {
+      "epoch": 0.9479228361113461,
+      "grad_norm": 0.17438729107379913,
+      "learning_rate": 0.001,
+      "loss": 1.7021,
+      "step": 22407
+    },
+    {
+      "epoch": 0.9479651408748625,
+      "grad_norm": 0.14387691020965576,
+      "learning_rate": 0.001,
+      "loss": 2.1247,
+      "step": 22408
+    },
+    {
+      "epoch": 0.9480074456383789,
+      "grad_norm": 0.14491653442382812,
+      "learning_rate": 0.001,
+      "loss": 1.5185,
+      "step": 22409
+    },
+    {
+      "epoch": 0.9480497504018952,
+      "grad_norm": 0.15343250334262848,
+      "learning_rate": 0.001,
+      "loss": 1.7086,
+      "step": 22410
+    },
+    {
+      "epoch": 0.9480920551654116,
+      "grad_norm": 0.2698756456375122,
+      "learning_rate": 0.001,
+      "loss": 2.0098,
+      "step": 22411
+    },
+    {
+      "epoch": 0.948134359928928,
+      "grad_norm": 0.12875612080097198,
+      "learning_rate": 0.001,
+      "loss": 1.5075,
+      "step": 22412
+    },
+    {
+      "epoch": 0.9481766646924443,
+      "grad_norm": 0.2064325362443924,
+      "learning_rate": 0.001,
+      "loss": 1.9919,
+      "step": 22413
+    },
+    {
+      "epoch": 0.9482189694559607,
+      "grad_norm": 0.1798713207244873,
+      "learning_rate": 0.001,
+      "loss": 2.8559,
+      "step": 22414
+    },
+    {
+      "epoch": 0.9482612742194771,
+      "grad_norm": 4.14750337600708,
+      "learning_rate": 0.001,
+      "loss": 2.7561,
+      "step": 22415
+    },
+    {
+      "epoch": 0.9483035789829934,
+      "grad_norm": 0.15084944665431976,
+      "learning_rate": 0.001,
+      "loss": 2.2551,
+      "step": 22416
+    },
+    {
+      "epoch": 0.9483458837465099,
+      "grad_norm": 0.1535675972700119,
+      "learning_rate": 0.001,
+      "loss": 3.2881,
+      "step": 22417
+    },
+    {
+      "epoch": 0.9483881885100263,
+      "grad_norm": 0.1589122712612152,
+      "learning_rate": 0.001,
+      "loss": 1.8991,
+      "step": 22418
+    },
+    {
+      "epoch": 0.9484304932735426,
+      "grad_norm": 0.15410476922988892,
+      "learning_rate": 0.001,
+      "loss": 2.1234,
+      "step": 22419
+    },
+    {
+      "epoch": 0.948472798037059,
+      "grad_norm": 0.15095588564872742,
+      "learning_rate": 0.001,
+      "loss": 1.6618,
+      "step": 22420
+    },
+    {
+      "epoch": 0.9485151028005754,
+      "grad_norm": 0.14121654629707336,
+      "learning_rate": 0.001,
+      "loss": 1.6598,
+      "step": 22421
+    },
+    {
+      "epoch": 0.9485574075640917,
+      "grad_norm": 0.1701010763645172,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 22422
+    },
+    {
+      "epoch": 0.9485997123276081,
+      "grad_norm": 0.15959163010120392,
+      "learning_rate": 0.001,
+      "loss": 1.7565,
+      "step": 22423
+    },
+    {
+      "epoch": 0.9486420170911245,
+      "grad_norm": 0.1474931389093399,
+      "learning_rate": 0.001,
+      "loss": 2.6955,
+      "step": 22424
+    },
+    {
+      "epoch": 0.9486843218546408,
+      "grad_norm": 0.17197000980377197,
+      "learning_rate": 0.001,
+      "loss": 3.4178,
+      "step": 22425
+    },
+    {
+      "epoch": 0.9487266266181572,
+      "grad_norm": 0.27430710196495056,
+      "learning_rate": 0.001,
+      "loss": 2.2628,
+      "step": 22426
+    },
+    {
+      "epoch": 0.9487689313816736,
+      "grad_norm": 0.15919172763824463,
+      "learning_rate": 0.001,
+      "loss": 1.9106,
+      "step": 22427
+    },
+    {
+      "epoch": 0.9488112361451899,
+      "grad_norm": 11.105170249938965,
+      "learning_rate": 0.001,
+      "loss": 2.4885,
+      "step": 22428
+    },
+    {
+      "epoch": 0.9488535409087063,
+      "grad_norm": 0.18378356099128723,
+      "learning_rate": 0.001,
+      "loss": 1.9228,
+      "step": 22429
+    },
+    {
+      "epoch": 0.9488958456722227,
+      "grad_norm": 0.16684526205062866,
+      "learning_rate": 0.001,
+      "loss": 1.6823,
+      "step": 22430
+    },
+    {
+      "epoch": 0.948938150435739,
+      "grad_norm": 0.12937648594379425,
+      "learning_rate": 0.001,
+      "loss": 2.4837,
+      "step": 22431
+    },
+    {
+      "epoch": 0.9489804551992554,
+      "grad_norm": 6.135654449462891,
+      "learning_rate": 0.001,
+      "loss": 1.8389,
+      "step": 22432
+    },
+    {
+      "epoch": 0.9490227599627719,
+      "grad_norm": 0.15108750760555267,
+      "learning_rate": 0.001,
+      "loss": 1.4086,
+      "step": 22433
+    },
+    {
+      "epoch": 0.9490650647262882,
+      "grad_norm": 0.1532086580991745,
+      "learning_rate": 0.001,
+      "loss": 2.467,
+      "step": 22434
+    },
+    {
+      "epoch": 0.9491073694898046,
+      "grad_norm": 0.21752135455608368,
+      "learning_rate": 0.001,
+      "loss": 3.2817,
+      "step": 22435
+    },
+    {
+      "epoch": 0.949149674253321,
+      "grad_norm": 0.45808929204940796,
+      "learning_rate": 0.001,
+      "loss": 2.6698,
+      "step": 22436
+    },
+    {
+      "epoch": 0.9491919790168373,
+      "grad_norm": 0.1843804270029068,
+      "learning_rate": 0.001,
+      "loss": 1.7856,
+      "step": 22437
+    },
+    {
+      "epoch": 0.9492342837803537,
+      "grad_norm": 0.21448786556720734,
+      "learning_rate": 0.001,
+      "loss": 2.3707,
+      "step": 22438
+    },
+    {
+      "epoch": 0.9492765885438701,
+      "grad_norm": 0.13399538397789001,
+      "learning_rate": 0.001,
+      "loss": 1.7928,
+      "step": 22439
+    },
+    {
+      "epoch": 0.9493188933073864,
+      "grad_norm": 0.13753435015678406,
+      "learning_rate": 0.001,
+      "loss": 2.8811,
+      "step": 22440
+    },
+    {
+      "epoch": 0.9493611980709028,
+      "grad_norm": 0.14787648618221283,
+      "learning_rate": 0.001,
+      "loss": 1.6632,
+      "step": 22441
+    },
+    {
+      "epoch": 0.9494035028344192,
+      "grad_norm": 0.150163933634758,
+      "learning_rate": 0.001,
+      "loss": 1.6527,
+      "step": 22442
+    },
+    {
+      "epoch": 0.9494458075979355,
+      "grad_norm": 0.14982707798480988,
+      "learning_rate": 0.001,
+      "loss": 1.4879,
+      "step": 22443
+    },
+    {
+      "epoch": 0.9494881123614519,
+      "grad_norm": 0.9315142035484314,
+      "learning_rate": 0.001,
+      "loss": 1.8539,
+      "step": 22444
+    },
+    {
+      "epoch": 0.9495304171249682,
+      "grad_norm": 0.1498139351606369,
+      "learning_rate": 0.001,
+      "loss": 2.1401,
+      "step": 22445
+    },
+    {
+      "epoch": 0.9495727218884846,
+      "grad_norm": 0.1625237613916397,
+      "learning_rate": 0.001,
+      "loss": 2.2502,
+      "step": 22446
+    },
+    {
+      "epoch": 0.949615026652001,
+      "grad_norm": 1.6854742765426636,
+      "learning_rate": 0.001,
+      "loss": 1.7948,
+      "step": 22447
+    },
+    {
+      "epoch": 0.9496573314155173,
+      "grad_norm": 0.24923864006996155,
+      "learning_rate": 0.001,
+      "loss": 1.9564,
+      "step": 22448
+    },
+    {
+      "epoch": 0.9496996361790337,
+      "grad_norm": 0.16029444336891174,
+      "learning_rate": 0.001,
+      "loss": 1.8311,
+      "step": 22449
+    },
+    {
+      "epoch": 0.9497419409425502,
+      "grad_norm": 0.18315951526165009,
+      "learning_rate": 0.001,
+      "loss": 2.9322,
+      "step": 22450
+    },
+    {
+      "epoch": 0.9497842457060665,
+      "grad_norm": 0.14090849459171295,
+      "learning_rate": 0.001,
+      "loss": 1.3484,
+      "step": 22451
+    },
+    {
+      "epoch": 0.9498265504695829,
+      "grad_norm": 0.19691091775894165,
+      "learning_rate": 0.001,
+      "loss": 1.3982,
+      "step": 22452
+    },
+    {
+      "epoch": 0.9498688552330993,
+      "grad_norm": 0.19947366416454315,
+      "learning_rate": 0.001,
+      "loss": 2.8261,
+      "step": 22453
+    },
+    {
+      "epoch": 0.9499111599966156,
+      "grad_norm": 0.1901845932006836,
+      "learning_rate": 0.001,
+      "loss": 2.3973,
+      "step": 22454
+    },
+    {
+      "epoch": 0.949953464760132,
+      "grad_norm": 12.674116134643555,
+      "learning_rate": 0.001,
+      "loss": 2.4079,
+      "step": 22455
+    },
+    {
+      "epoch": 0.9499957695236484,
+      "grad_norm": 0.20128510892391205,
+      "learning_rate": 0.001,
+      "loss": 2.0499,
+      "step": 22456
+    },
+    {
+      "epoch": 0.9500380742871647,
+      "grad_norm": 1.169372320175171,
+      "learning_rate": 0.001,
+      "loss": 2.0313,
+      "step": 22457
+    },
+    {
+      "epoch": 0.9500803790506811,
+      "grad_norm": 0.7405765652656555,
+      "learning_rate": 0.001,
+      "loss": 2.3182,
+      "step": 22458
+    },
+    {
+      "epoch": 0.9501226838141975,
+      "grad_norm": 0.16553084552288055,
+      "learning_rate": 0.001,
+      "loss": 1.6338,
+      "step": 22459
+    },
+    {
+      "epoch": 0.9501649885777138,
+      "grad_norm": 0.34061887860298157,
+      "learning_rate": 0.001,
+      "loss": 1.7407,
+      "step": 22460
+    },
+    {
+      "epoch": 0.9502072933412302,
+      "grad_norm": 0.13990429043769836,
+      "learning_rate": 0.001,
+      "loss": 1.731,
+      "step": 22461
+    },
+    {
+      "epoch": 0.9502495981047466,
+      "grad_norm": 0.1741171032190323,
+      "learning_rate": 0.001,
+      "loss": 1.9323,
+      "step": 22462
+    },
+    {
+      "epoch": 0.9502919028682629,
+      "grad_norm": 0.1664520800113678,
+      "learning_rate": 0.001,
+      "loss": 2.2529,
+      "step": 22463
+    },
+    {
+      "epoch": 0.9503342076317793,
+      "grad_norm": 0.17836974561214447,
+      "learning_rate": 0.001,
+      "loss": 2.3956,
+      "step": 22464
+    },
+    {
+      "epoch": 0.9503765123952957,
+      "grad_norm": 0.16051559150218964,
+      "learning_rate": 0.001,
+      "loss": 2.4441,
+      "step": 22465
+    },
+    {
+      "epoch": 0.950418817158812,
+      "grad_norm": 0.14825597405433655,
+      "learning_rate": 0.001,
+      "loss": 2.1567,
+      "step": 22466
+    },
+    {
+      "epoch": 0.9504611219223285,
+      "grad_norm": 0.1824505478143692,
+      "learning_rate": 0.001,
+      "loss": 2.5797,
+      "step": 22467
+    },
+    {
+      "epoch": 0.9505034266858449,
+      "grad_norm": 0.21144576370716095,
+      "learning_rate": 0.001,
+      "loss": 2.8215,
+      "step": 22468
+    },
+    {
+      "epoch": 0.9505457314493612,
+      "grad_norm": 0.16636143624782562,
+      "learning_rate": 0.001,
+      "loss": 2.8168,
+      "step": 22469
+    },
+    {
+      "epoch": 0.9505880362128776,
+      "grad_norm": 0.3710878789424896,
+      "learning_rate": 0.001,
+      "loss": 3.2767,
+      "step": 22470
+    },
+    {
+      "epoch": 0.950630340976394,
+      "grad_norm": 0.1762601137161255,
+      "learning_rate": 0.001,
+      "loss": 2.8635,
+      "step": 22471
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 2.875742197036743,
+      "learning_rate": 0.001,
+      "loss": 1.802,
+      "step": 22472
+    },
+    {
+      "epoch": 0.9507149505034267,
+      "grad_norm": 0.16046804189682007,
+      "learning_rate": 0.001,
+      "loss": 1.5675,
+      "step": 22473
+    },
+    {
+      "epoch": 0.9507572552669431,
+      "grad_norm": 0.18604080379009247,
+      "learning_rate": 0.001,
+      "loss": 2.3506,
+      "step": 22474
+    },
+    {
+      "epoch": 0.9507995600304594,
+      "grad_norm": 0.1995047777891159,
+      "learning_rate": 0.001,
+      "loss": 1.4797,
+      "step": 22475
+    },
+    {
+      "epoch": 0.9508418647939758,
+      "grad_norm": 41.969749450683594,
+      "learning_rate": 0.001,
+      "loss": 1.6885,
+      "step": 22476
+    },
+    {
+      "epoch": 0.9508841695574922,
+      "grad_norm": 0.17194336652755737,
+      "learning_rate": 0.001,
+      "loss": 1.8758,
+      "step": 22477
+    },
+    {
+      "epoch": 0.9509264743210085,
+      "grad_norm": 0.1948256492614746,
+      "learning_rate": 0.001,
+      "loss": 2.4022,
+      "step": 22478
+    },
+    {
+      "epoch": 0.9509687790845249,
+      "grad_norm": 0.21217121183872223,
+      "learning_rate": 0.001,
+      "loss": 2.4836,
+      "step": 22479
+    },
+    {
+      "epoch": 0.9510110838480413,
+      "grad_norm": 0.1763080209493637,
+      "learning_rate": 0.001,
+      "loss": 2.3558,
+      "step": 22480
+    },
+    {
+      "epoch": 0.9510533886115576,
+      "grad_norm": 0.22325830161571503,
+      "learning_rate": 0.001,
+      "loss": 2.0567,
+      "step": 22481
+    },
+    {
+      "epoch": 0.951095693375074,
+      "grad_norm": 0.16838374733924866,
+      "learning_rate": 0.001,
+      "loss": 2.779,
+      "step": 22482
+    },
+    {
+      "epoch": 0.9511379981385905,
+      "grad_norm": 0.17531989514827728,
+      "learning_rate": 0.001,
+      "loss": 2.0818,
+      "step": 22483
+    },
+    {
+      "epoch": 0.9511803029021068,
+      "grad_norm": 0.17867770791053772,
+      "learning_rate": 0.001,
+      "loss": 2.5842,
+      "step": 22484
+    },
+    {
+      "epoch": 0.9512226076656232,
+      "grad_norm": 0.1932533085346222,
+      "learning_rate": 0.001,
+      "loss": 1.9442,
+      "step": 22485
+    },
+    {
+      "epoch": 0.9512649124291396,
+      "grad_norm": 0.1727113425731659,
+      "learning_rate": 0.001,
+      "loss": 1.5683,
+      "step": 22486
+    },
+    {
+      "epoch": 0.9513072171926559,
+      "grad_norm": 0.1291782408952713,
+      "learning_rate": 0.001,
+      "loss": 2.3555,
+      "step": 22487
+    },
+    {
+      "epoch": 0.9513495219561723,
+      "grad_norm": 0.7629174590110779,
+      "learning_rate": 0.001,
+      "loss": 2.651,
+      "step": 22488
+    },
+    {
+      "epoch": 0.9513918267196886,
+      "grad_norm": 0.12951643764972687,
+      "learning_rate": 0.001,
+      "loss": 1.9717,
+      "step": 22489
+    },
+    {
+      "epoch": 0.951434131483205,
+      "grad_norm": 0.6491950750350952,
+      "learning_rate": 0.001,
+      "loss": 1.9874,
+      "step": 22490
+    },
+    {
+      "epoch": 0.9514764362467214,
+      "grad_norm": 0.1988588273525238,
+      "learning_rate": 0.001,
+      "loss": 2.0231,
+      "step": 22491
+    },
+    {
+      "epoch": 0.9515187410102377,
+      "grad_norm": 0.18948745727539062,
+      "learning_rate": 0.001,
+      "loss": 2.0585,
+      "step": 22492
+    },
+    {
+      "epoch": 0.9515610457737541,
+      "grad_norm": 0.3615002930164337,
+      "learning_rate": 0.001,
+      "loss": 3.7874,
+      "step": 22493
+    },
+    {
+      "epoch": 0.9516033505372705,
+      "grad_norm": 0.13920491933822632,
+      "learning_rate": 0.001,
+      "loss": 1.6694,
+      "step": 22494
+    },
+    {
+      "epoch": 0.9516456553007868,
+      "grad_norm": 0.16164518892765045,
+      "learning_rate": 0.001,
+      "loss": 2.0692,
+      "step": 22495
+    },
+    {
+      "epoch": 0.9516879600643032,
+      "grad_norm": 0.47578608989715576,
+      "learning_rate": 0.001,
+      "loss": 2.1681,
+      "step": 22496
+    },
+    {
+      "epoch": 0.9517302648278196,
+      "grad_norm": 0.20885364711284637,
+      "learning_rate": 0.001,
+      "loss": 3.6518,
+      "step": 22497
+    },
+    {
+      "epoch": 0.9517725695913359,
+      "grad_norm": 0.16062334179878235,
+      "learning_rate": 0.001,
+      "loss": 1.9614,
+      "step": 22498
+    },
+    {
+      "epoch": 0.9518148743548523,
+      "grad_norm": 0.22703850269317627,
+      "learning_rate": 0.001,
+      "loss": 2.476,
+      "step": 22499
+    },
+    {
+      "epoch": 0.9518571791183688,
+      "grad_norm": 0.2395503968000412,
+      "learning_rate": 0.001,
+      "loss": 1.9883,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9518994838818851,
+      "grad_norm": 0.15861296653747559,
+      "learning_rate": 0.001,
+      "loss": 1.8161,
+      "step": 22501
+    },
+    {
+      "epoch": 0.9519417886454015,
+      "grad_norm": 0.15886007249355316,
+      "learning_rate": 0.001,
+      "loss": 1.8054,
+      "step": 22502
+    },
+    {
+      "epoch": 0.9519840934089179,
+      "grad_norm": 0.18226757645606995,
+      "learning_rate": 0.001,
+      "loss": 1.7382,
+      "step": 22503
+    },
+    {
+      "epoch": 0.9520263981724342,
+      "grad_norm": 0.17415927350521088,
+      "learning_rate": 0.001,
+      "loss": 1.6456,
+      "step": 22504
+    },
+    {
+      "epoch": 0.9520687029359506,
+      "grad_norm": 0.22157450020313263,
+      "learning_rate": 0.001,
+      "loss": 2.0265,
+      "step": 22505
+    },
+    {
+      "epoch": 0.952111007699467,
+      "grad_norm": 0.18525850772857666,
+      "learning_rate": 0.001,
+      "loss": 3.7023,
+      "step": 22506
+    },
+    {
+      "epoch": 0.9521533124629833,
+      "grad_norm": 0.2394370287656784,
+      "learning_rate": 0.001,
+      "loss": 3.2888,
+      "step": 22507
+    },
+    {
+      "epoch": 0.9521956172264997,
+      "grad_norm": 0.15804435312747955,
+      "learning_rate": 0.001,
+      "loss": 1.9221,
+      "step": 22508
+    },
+    {
+      "epoch": 0.9522379219900161,
+      "grad_norm": 0.18028190732002258,
+      "learning_rate": 0.001,
+      "loss": 2.1621,
+      "step": 22509
+    },
+    {
+      "epoch": 0.9522802267535324,
+      "grad_norm": 0.16329197585582733,
+      "learning_rate": 0.001,
+      "loss": 1.6036,
+      "step": 22510
+    },
+    {
+      "epoch": 0.9523225315170488,
+      "grad_norm": 0.15661050379276276,
+      "learning_rate": 0.001,
+      "loss": 1.8212,
+      "step": 22511
+    },
+    {
+      "epoch": 0.9523648362805652,
+      "grad_norm": 0.23924638330936432,
+      "learning_rate": 0.001,
+      "loss": 1.4197,
+      "step": 22512
+    },
+    {
+      "epoch": 0.9524071410440815,
+      "grad_norm": 0.17719987034797668,
+      "learning_rate": 0.001,
+      "loss": 1.427,
+      "step": 22513
+    },
+    {
+      "epoch": 0.9524494458075979,
+      "grad_norm": 0.1736730933189392,
+      "learning_rate": 0.001,
+      "loss": 1.5887,
+      "step": 22514
+    },
+    {
+      "epoch": 0.9524917505711143,
+      "grad_norm": 0.1556921899318695,
+      "learning_rate": 0.001,
+      "loss": 2.8394,
+      "step": 22515
+    },
+    {
+      "epoch": 0.9525340553346306,
+      "grad_norm": 0.15462371706962585,
+      "learning_rate": 0.001,
+      "loss": 1.9483,
+      "step": 22516
+    },
+    {
+      "epoch": 0.9525763600981471,
+      "grad_norm": 0.16534818708896637,
+      "learning_rate": 0.001,
+      "loss": 2.2795,
+      "step": 22517
+    },
+    {
+      "epoch": 0.9526186648616635,
+      "grad_norm": 0.276944100856781,
+      "learning_rate": 0.001,
+      "loss": 2.1088,
+      "step": 22518
+    },
+    {
+      "epoch": 0.9526609696251798,
+      "grad_norm": 0.13408029079437256,
+      "learning_rate": 0.001,
+      "loss": 1.9941,
+      "step": 22519
+    },
+    {
+      "epoch": 0.9527032743886962,
+      "grad_norm": 8.694868087768555,
+      "learning_rate": 0.001,
+      "loss": 1.924,
+      "step": 22520
+    },
+    {
+      "epoch": 0.9527455791522126,
+      "grad_norm": 0.24468626081943512,
+      "learning_rate": 0.001,
+      "loss": 2.0428,
+      "step": 22521
+    },
+    {
+      "epoch": 0.9527878839157289,
+      "grad_norm": 0.1722177416086197,
+      "learning_rate": 0.001,
+      "loss": 2.467,
+      "step": 22522
+    },
+    {
+      "epoch": 0.9528301886792453,
+      "grad_norm": 0.16793347895145416,
+      "learning_rate": 0.001,
+      "loss": 1.8234,
+      "step": 22523
+    },
+    {
+      "epoch": 0.9528724934427617,
+      "grad_norm": 0.6583682894706726,
+      "learning_rate": 0.001,
+      "loss": 1.7482,
+      "step": 22524
+    },
+    {
+      "epoch": 0.952914798206278,
+      "grad_norm": 0.17453870177268982,
+      "learning_rate": 0.001,
+      "loss": 1.8821,
+      "step": 22525
+    },
+    {
+      "epoch": 0.9529571029697944,
+      "grad_norm": 0.17266742885112762,
+      "learning_rate": 0.001,
+      "loss": 1.8525,
+      "step": 22526
+    },
+    {
+      "epoch": 0.9529994077333108,
+      "grad_norm": 0.1756075918674469,
+      "learning_rate": 0.001,
+      "loss": 2.13,
+      "step": 22527
+    },
+    {
+      "epoch": 0.9530417124968271,
+      "grad_norm": 0.18365883827209473,
+      "learning_rate": 0.001,
+      "loss": 1.8475,
+      "step": 22528
+    },
+    {
+      "epoch": 0.9530840172603435,
+      "grad_norm": 1.1204431056976318,
+      "learning_rate": 0.001,
+      "loss": 1.7542,
+      "step": 22529
+    },
+    {
+      "epoch": 0.9531263220238599,
+      "grad_norm": 0.16155286133289337,
+      "learning_rate": 0.001,
+      "loss": 2.4133,
+      "step": 22530
+    },
+    {
+      "epoch": 0.9531686267873762,
+      "grad_norm": 1.2708326578140259,
+      "learning_rate": 0.001,
+      "loss": 1.942,
+      "step": 22531
+    },
+    {
+      "epoch": 0.9532109315508926,
+      "grad_norm": 0.3815459609031677,
+      "learning_rate": 0.001,
+      "loss": 1.5244,
+      "step": 22532
+    },
+    {
+      "epoch": 0.953253236314409,
+      "grad_norm": 0.1591525673866272,
+      "learning_rate": 0.001,
+      "loss": 1.6586,
+      "step": 22533
+    },
+    {
+      "epoch": 0.9532955410779254,
+      "grad_norm": 0.16831563413143158,
+      "learning_rate": 0.001,
+      "loss": 1.9452,
+      "step": 22534
+    },
+    {
+      "epoch": 0.9533378458414418,
+      "grad_norm": 0.1781228631734848,
+      "learning_rate": 0.001,
+      "loss": 1.9083,
+      "step": 22535
+    },
+    {
+      "epoch": 0.9533801506049581,
+      "grad_norm": 0.196019247174263,
+      "learning_rate": 0.001,
+      "loss": 1.8808,
+      "step": 22536
+    },
+    {
+      "epoch": 0.9534224553684745,
+      "grad_norm": 0.16879303753376007,
+      "learning_rate": 0.001,
+      "loss": 2.5292,
+      "step": 22537
+    },
+    {
+      "epoch": 0.9534647601319909,
+      "grad_norm": 0.14344555139541626,
+      "learning_rate": 0.001,
+      "loss": 3.0776,
+      "step": 22538
+    },
+    {
+      "epoch": 0.9535070648955072,
+      "grad_norm": 0.13954542577266693,
+      "learning_rate": 0.001,
+      "loss": 1.8676,
+      "step": 22539
+    },
+    {
+      "epoch": 0.9535493696590236,
+      "grad_norm": 0.16517263650894165,
+      "learning_rate": 0.001,
+      "loss": 2.1363,
+      "step": 22540
+    },
+    {
+      "epoch": 0.95359167442254,
+      "grad_norm": 0.4212926924228668,
+      "learning_rate": 0.001,
+      "loss": 2.4361,
+      "step": 22541
+    },
+    {
+      "epoch": 0.9536339791860563,
+      "grad_norm": 0.19001418352127075,
+      "learning_rate": 0.001,
+      "loss": 2.4905,
+      "step": 22542
+    },
+    {
+      "epoch": 0.9536762839495727,
+      "grad_norm": 0.1455460637807846,
+      "learning_rate": 0.001,
+      "loss": 1.8486,
+      "step": 22543
+    },
+    {
+      "epoch": 0.9537185887130891,
+      "grad_norm": 0.13077104091644287,
+      "learning_rate": 0.001,
+      "loss": 1.5365,
+      "step": 22544
+    },
+    {
+      "epoch": 0.9537608934766054,
+      "grad_norm": 0.139632910490036,
+      "learning_rate": 0.001,
+      "loss": 1.5715,
+      "step": 22545
+    },
+    {
+      "epoch": 0.9538031982401218,
+      "grad_norm": 0.14451947808265686,
+      "learning_rate": 0.001,
+      "loss": 2.0676,
+      "step": 22546
+    },
+    {
+      "epoch": 0.9538455030036382,
+      "grad_norm": 0.2569085657596588,
+      "learning_rate": 0.001,
+      "loss": 2.4596,
+      "step": 22547
+    },
+    {
+      "epoch": 0.9538878077671545,
+      "grad_norm": 9.396105766296387,
+      "learning_rate": 0.001,
+      "loss": 1.7845,
+      "step": 22548
+    },
+    {
+      "epoch": 0.953930112530671,
+      "grad_norm": 0.17778228223323822,
+      "learning_rate": 0.001,
+      "loss": 2.6535,
+      "step": 22549
+    },
+    {
+      "epoch": 0.9539724172941874,
+      "grad_norm": 1.270423173904419,
+      "learning_rate": 0.001,
+      "loss": 1.7749,
+      "step": 22550
+    },
+    {
+      "epoch": 0.9540147220577037,
+      "grad_norm": 0.20643652975559235,
+      "learning_rate": 0.001,
+      "loss": 2.5151,
+      "step": 22551
+    },
+    {
+      "epoch": 0.9540570268212201,
+      "grad_norm": 0.5847330093383789,
+      "learning_rate": 0.001,
+      "loss": 2.6464,
+      "step": 22552
+    },
+    {
+      "epoch": 0.9540993315847365,
+      "grad_norm": 0.15910720825195312,
+      "learning_rate": 0.001,
+      "loss": 2.5165,
+      "step": 22553
+    },
+    {
+      "epoch": 0.9541416363482528,
+      "grad_norm": 0.2874158024787903,
+      "learning_rate": 0.001,
+      "loss": 2.2493,
+      "step": 22554
+    },
+    {
+      "epoch": 0.9541839411117692,
+      "grad_norm": 0.14962883293628693,
+      "learning_rate": 0.001,
+      "loss": 2.0772,
+      "step": 22555
+    },
+    {
+      "epoch": 0.9542262458752856,
+      "grad_norm": 0.16051600873470306,
+      "learning_rate": 0.001,
+      "loss": 1.4298,
+      "step": 22556
+    },
+    {
+      "epoch": 0.9542685506388019,
+      "grad_norm": 0.16125862300395966,
+      "learning_rate": 0.001,
+      "loss": 1.383,
+      "step": 22557
+    },
+    {
+      "epoch": 0.9543108554023183,
+      "grad_norm": 0.15643396973609924,
+      "learning_rate": 0.001,
+      "loss": 1.9494,
+      "step": 22558
+    },
+    {
+      "epoch": 0.9543531601658347,
+      "grad_norm": 0.2045285403728485,
+      "learning_rate": 0.001,
+      "loss": 2.7422,
+      "step": 22559
+    },
+    {
+      "epoch": 0.954395464929351,
+      "grad_norm": 0.2184513658285141,
+      "learning_rate": 0.001,
+      "loss": 2.085,
+      "step": 22560
+    },
+    {
+      "epoch": 0.9544377696928674,
+      "grad_norm": 0.16785749793052673,
+      "learning_rate": 0.001,
+      "loss": 2.3423,
+      "step": 22561
+    },
+    {
+      "epoch": 0.9544800744563838,
+      "grad_norm": 0.16938123106956482,
+      "learning_rate": 0.001,
+      "loss": 1.6398,
+      "step": 22562
+    },
+    {
+      "epoch": 0.9545223792199001,
+      "grad_norm": 0.15510663390159607,
+      "learning_rate": 0.001,
+      "loss": 1.739,
+      "step": 22563
+    },
+    {
+      "epoch": 0.9545646839834165,
+      "grad_norm": 0.16871055960655212,
+      "learning_rate": 0.001,
+      "loss": 1.6647,
+      "step": 22564
+    },
+    {
+      "epoch": 0.954606988746933,
+      "grad_norm": 0.2923224866390228,
+      "learning_rate": 0.001,
+      "loss": 3.0942,
+      "step": 22565
+    },
+    {
+      "epoch": 0.9546492935104492,
+      "grad_norm": 0.37078985571861267,
+      "learning_rate": 0.001,
+      "loss": 1.9319,
+      "step": 22566
+    },
+    {
+      "epoch": 0.9546915982739657,
+      "grad_norm": 0.16258271038532257,
+      "learning_rate": 0.001,
+      "loss": 3.1049,
+      "step": 22567
+    },
+    {
+      "epoch": 0.9547339030374821,
+      "grad_norm": 0.18606746196746826,
+      "learning_rate": 0.001,
+      "loss": 3.302,
+      "step": 22568
+    },
+    {
+      "epoch": 0.9547762078009984,
+      "grad_norm": 0.7223076224327087,
+      "learning_rate": 0.001,
+      "loss": 1.7057,
+      "step": 22569
+    },
+    {
+      "epoch": 0.9548185125645148,
+      "grad_norm": 0.1600056141614914,
+      "learning_rate": 0.001,
+      "loss": 2.7663,
+      "step": 22570
+    },
+    {
+      "epoch": 0.9548608173280312,
+      "grad_norm": 0.16260926425457,
+      "learning_rate": 0.001,
+      "loss": 3.0786,
+      "step": 22571
+    },
+    {
+      "epoch": 0.9549031220915475,
+      "grad_norm": 0.16478076577186584,
+      "learning_rate": 0.001,
+      "loss": 1.8152,
+      "step": 22572
+    },
+    {
+      "epoch": 0.9549454268550639,
+      "grad_norm": 0.14797167479991913,
+      "learning_rate": 0.001,
+      "loss": 1.6904,
+      "step": 22573
+    },
+    {
+      "epoch": 0.9549877316185803,
+      "grad_norm": 0.1376873105764389,
+      "learning_rate": 0.001,
+      "loss": 1.672,
+      "step": 22574
+    },
+    {
+      "epoch": 0.9550300363820966,
+      "grad_norm": 0.15510663390159607,
+      "learning_rate": 0.001,
+      "loss": 2.1794,
+      "step": 22575
+    },
+    {
+      "epoch": 0.955072341145613,
+      "grad_norm": 0.15933188796043396,
+      "learning_rate": 0.001,
+      "loss": 2.7918,
+      "step": 22576
+    },
+    {
+      "epoch": 0.9551146459091294,
+      "grad_norm": 0.15145090222358704,
+      "learning_rate": 0.001,
+      "loss": 1.8279,
+      "step": 22577
+    },
+    {
+      "epoch": 0.9551569506726457,
+      "grad_norm": 0.17468474805355072,
+      "learning_rate": 0.001,
+      "loss": 3.198,
+      "step": 22578
+    },
+    {
+      "epoch": 0.9551992554361621,
+      "grad_norm": 0.17872756719589233,
+      "learning_rate": 0.001,
+      "loss": 1.7174,
+      "step": 22579
+    },
+    {
+      "epoch": 0.9552415601996784,
+      "grad_norm": 0.17756347358226776,
+      "learning_rate": 0.001,
+      "loss": 2.8859,
+      "step": 22580
+    },
+    {
+      "epoch": 0.9552838649631948,
+      "grad_norm": 0.17524990439414978,
+      "learning_rate": 0.001,
+      "loss": 2.7646,
+      "step": 22581
+    },
+    {
+      "epoch": 0.9553261697267112,
+      "grad_norm": 0.1490987241268158,
+      "learning_rate": 0.001,
+      "loss": 2.0239,
+      "step": 22582
+    },
+    {
+      "epoch": 0.9553684744902275,
+      "grad_norm": 0.17126286029815674,
+      "learning_rate": 0.001,
+      "loss": 2.1528,
+      "step": 22583
+    },
+    {
+      "epoch": 0.955410779253744,
+      "grad_norm": 0.14318470656871796,
+      "learning_rate": 0.001,
+      "loss": 2.0535,
+      "step": 22584
+    },
+    {
+      "epoch": 0.9554530840172604,
+      "grad_norm": 0.1496138721704483,
+      "learning_rate": 0.001,
+      "loss": 2.2055,
+      "step": 22585
+    },
+    {
+      "epoch": 0.9554953887807767,
+      "grad_norm": 0.1591835618019104,
+      "learning_rate": 0.001,
+      "loss": 1.8175,
+      "step": 22586
+    },
+    {
+      "epoch": 0.9555376935442931,
+      "grad_norm": 0.18127931654453278,
+      "learning_rate": 0.001,
+      "loss": 3.1087,
+      "step": 22587
+    },
+    {
+      "epoch": 0.9555799983078095,
+      "grad_norm": 0.12975792586803436,
+      "learning_rate": 0.001,
+      "loss": 2.2168,
+      "step": 22588
+    },
+    {
+      "epoch": 0.9556223030713258,
+      "grad_norm": 0.1456218957901001,
+      "learning_rate": 0.001,
+      "loss": 1.4985,
+      "step": 22589
+    },
+    {
+      "epoch": 0.9556646078348422,
+      "grad_norm": 0.1666790097951889,
+      "learning_rate": 0.001,
+      "loss": 1.8393,
+      "step": 22590
+    },
+    {
+      "epoch": 0.9557069125983586,
+      "grad_norm": 0.1480528861284256,
+      "learning_rate": 0.001,
+      "loss": 1.7311,
+      "step": 22591
+    },
+    {
+      "epoch": 0.9557492173618749,
+      "grad_norm": 0.1545533984899521,
+      "learning_rate": 0.001,
+      "loss": 2.0204,
+      "step": 22592
+    },
+    {
+      "epoch": 0.9557915221253913,
+      "grad_norm": 0.17958112061023712,
+      "learning_rate": 0.001,
+      "loss": 2.5396,
+      "step": 22593
+    },
+    {
+      "epoch": 0.9558338268889077,
+      "grad_norm": 0.20622247457504272,
+      "learning_rate": 0.001,
+      "loss": 2.4856,
+      "step": 22594
+    },
+    {
+      "epoch": 0.955876131652424,
+      "grad_norm": 0.1637284755706787,
+      "learning_rate": 0.001,
+      "loss": 1.561,
+      "step": 22595
+    },
+    {
+      "epoch": 0.9559184364159404,
+      "grad_norm": 0.13965626060962677,
+      "learning_rate": 0.001,
+      "loss": 1.7077,
+      "step": 22596
+    },
+    {
+      "epoch": 0.9559607411794568,
+      "grad_norm": 0.15451842546463013,
+      "learning_rate": 0.001,
+      "loss": 2.9797,
+      "step": 22597
+    },
+    {
+      "epoch": 0.9560030459429731,
+      "grad_norm": 0.1734355241060257,
+      "learning_rate": 0.001,
+      "loss": 1.9119,
+      "step": 22598
+    },
+    {
+      "epoch": 0.9560453507064895,
+      "grad_norm": 0.16818620264530182,
+      "learning_rate": 0.001,
+      "loss": 2.4106,
+      "step": 22599
+    },
+    {
+      "epoch": 0.956087655470006,
+      "grad_norm": 0.4067355990409851,
+      "learning_rate": 0.001,
+      "loss": 2.4487,
+      "step": 22600
+    },
+    {
+      "epoch": 0.9561299602335223,
+      "grad_norm": 0.16298459470272064,
+      "learning_rate": 0.001,
+      "loss": 2.2041,
+      "step": 22601
+    },
+    {
+      "epoch": 0.9561722649970387,
+      "grad_norm": 0.7233208417892456,
+      "learning_rate": 0.001,
+      "loss": 1.9019,
+      "step": 22602
+    },
+    {
+      "epoch": 0.9562145697605551,
+      "grad_norm": 0.15039215981960297,
+      "learning_rate": 0.001,
+      "loss": 2.3582,
+      "step": 22603
+    },
+    {
+      "epoch": 0.9562568745240714,
+      "grad_norm": 0.25849875807762146,
+      "learning_rate": 0.001,
+      "loss": 2.4879,
+      "step": 22604
+    },
+    {
+      "epoch": 0.9562991792875878,
+      "grad_norm": 0.14870387315750122,
+      "learning_rate": 0.001,
+      "loss": 2.3015,
+      "step": 22605
+    },
+    {
+      "epoch": 0.9563414840511042,
+      "grad_norm": 0.16163372993469238,
+      "learning_rate": 0.001,
+      "loss": 2.9453,
+      "step": 22606
+    },
+    {
+      "epoch": 0.9563837888146205,
+      "grad_norm": 0.17681002616882324,
+      "learning_rate": 0.001,
+      "loss": 2.3321,
+      "step": 22607
+    },
+    {
+      "epoch": 0.9564260935781369,
+      "grad_norm": 0.16098996996879578,
+      "learning_rate": 0.001,
+      "loss": 1.9478,
+      "step": 22608
+    },
+    {
+      "epoch": 0.9564683983416533,
+      "grad_norm": 0.14798353612422943,
+      "learning_rate": 0.001,
+      "loss": 1.9331,
+      "step": 22609
+    },
+    {
+      "epoch": 0.9565107031051696,
+      "grad_norm": 0.31577110290527344,
+      "learning_rate": 0.001,
+      "loss": 3.3183,
+      "step": 22610
+    },
+    {
+      "epoch": 0.956553007868686,
+      "grad_norm": 0.1621190905570984,
+      "learning_rate": 0.001,
+      "loss": 1.9568,
+      "step": 22611
+    },
+    {
+      "epoch": 0.9565953126322024,
+      "grad_norm": 0.17691479623317719,
+      "learning_rate": 0.001,
+      "loss": 2.06,
+      "step": 22612
+    },
+    {
+      "epoch": 0.9566376173957187,
+      "grad_norm": 0.15779487788677216,
+      "learning_rate": 0.001,
+      "loss": 1.6084,
+      "step": 22613
+    },
+    {
+      "epoch": 0.9566799221592351,
+      "grad_norm": 0.15047040581703186,
+      "learning_rate": 0.001,
+      "loss": 2.1398,
+      "step": 22614
+    },
+    {
+      "epoch": 0.9567222269227516,
+      "grad_norm": 0.17002348601818085,
+      "learning_rate": 0.001,
+      "loss": 1.7664,
+      "step": 22615
+    },
+    {
+      "epoch": 0.9567645316862678,
+      "grad_norm": 0.14370211958885193,
+      "learning_rate": 0.001,
+      "loss": 1.598,
+      "step": 22616
+    },
+    {
+      "epoch": 0.9568068364497843,
+      "grad_norm": 1.530818223953247,
+      "learning_rate": 0.001,
+      "loss": 3.326,
+      "step": 22617
+    },
+    {
+      "epoch": 0.9568491412133007,
+      "grad_norm": 0.1687633991241455,
+      "learning_rate": 0.001,
+      "loss": 1.355,
+      "step": 22618
+    },
+    {
+      "epoch": 0.956891445976817,
+      "grad_norm": 0.13722075521945953,
+      "learning_rate": 0.001,
+      "loss": 3.3707,
+      "step": 22619
+    },
+    {
+      "epoch": 0.9569337507403334,
+      "grad_norm": 0.16178375482559204,
+      "learning_rate": 0.001,
+      "loss": 2.5891,
+      "step": 22620
+    },
+    {
+      "epoch": 0.9569760555038498,
+      "grad_norm": 0.7043463587760925,
+      "learning_rate": 0.001,
+      "loss": 2.3528,
+      "step": 22621
+    },
+    {
+      "epoch": 0.9570183602673661,
+      "grad_norm": 0.14952924847602844,
+      "learning_rate": 0.001,
+      "loss": 1.4348,
+      "step": 22622
+    },
+    {
+      "epoch": 0.9570606650308825,
+      "grad_norm": 0.14588811993598938,
+      "learning_rate": 0.001,
+      "loss": 2.168,
+      "step": 22623
+    },
+    {
+      "epoch": 0.9571029697943988,
+      "grad_norm": 0.136274516582489,
+      "learning_rate": 0.001,
+      "loss": 3.4291,
+      "step": 22624
+    },
+    {
+      "epoch": 0.9571452745579152,
+      "grad_norm": 0.15219195187091827,
+      "learning_rate": 0.001,
+      "loss": 2.353,
+      "step": 22625
+    },
+    {
+      "epoch": 0.9571875793214316,
+      "grad_norm": 0.192640021443367,
+      "learning_rate": 0.001,
+      "loss": 2.5704,
+      "step": 22626
+    },
+    {
+      "epoch": 0.9572298840849479,
+      "grad_norm": 2.2798144817352295,
+      "learning_rate": 0.001,
+      "loss": 2.4483,
+      "step": 22627
+    },
+    {
+      "epoch": 0.9572721888484643,
+      "grad_norm": 0.12433091551065445,
+      "learning_rate": 0.001,
+      "loss": 2.9738,
+      "step": 22628
+    },
+    {
+      "epoch": 0.9573144936119807,
+      "grad_norm": 0.14844204485416412,
+      "learning_rate": 0.001,
+      "loss": 2.3593,
+      "step": 22629
+    },
+    {
+      "epoch": 0.957356798375497,
+      "grad_norm": 0.18237383663654327,
+      "learning_rate": 0.001,
+      "loss": 2.4718,
+      "step": 22630
+    },
+    {
+      "epoch": 0.9573991031390134,
+      "grad_norm": 0.15633726119995117,
+      "learning_rate": 0.001,
+      "loss": 2.9718,
+      "step": 22631
+    },
+    {
+      "epoch": 0.9574414079025299,
+      "grad_norm": 0.1357184201478958,
+      "learning_rate": 0.001,
+      "loss": 1.8955,
+      "step": 22632
+    },
+    {
+      "epoch": 0.9574837126660461,
+      "grad_norm": 0.15892942249774933,
+      "learning_rate": 0.001,
+      "loss": 1.6971,
+      "step": 22633
+    },
+    {
+      "epoch": 0.9575260174295626,
+      "grad_norm": 0.1862511783838272,
+      "learning_rate": 0.001,
+      "loss": 2.416,
+      "step": 22634
+    },
+    {
+      "epoch": 0.957568322193079,
+      "grad_norm": 0.18506558239459991,
+      "learning_rate": 0.001,
+      "loss": 2.481,
+      "step": 22635
+    },
+    {
+      "epoch": 0.9576106269565953,
+      "grad_norm": 2.206505298614502,
+      "learning_rate": 0.001,
+      "loss": 2.0285,
+      "step": 22636
+    },
+    {
+      "epoch": 0.9576529317201117,
+      "grad_norm": 0.19020433723926544,
+      "learning_rate": 0.001,
+      "loss": 1.833,
+      "step": 22637
+    },
+    {
+      "epoch": 0.9576952364836281,
+      "grad_norm": 0.22142185270786285,
+      "learning_rate": 0.001,
+      "loss": 2.196,
+      "step": 22638
+    },
+    {
+      "epoch": 0.9577375412471444,
+      "grad_norm": 0.1741228997707367,
+      "learning_rate": 0.001,
+      "loss": 2.101,
+      "step": 22639
+    },
+    {
+      "epoch": 0.9577798460106608,
+      "grad_norm": 0.1707838922739029,
+      "learning_rate": 0.001,
+      "loss": 2.0027,
+      "step": 22640
+    },
+    {
+      "epoch": 0.9578221507741772,
+      "grad_norm": 0.16270555555820465,
+      "learning_rate": 0.001,
+      "loss": 2.9908,
+      "step": 22641
+    },
+    {
+      "epoch": 0.9578644555376935,
+      "grad_norm": 0.1549161970615387,
+      "learning_rate": 0.001,
+      "loss": 1.7517,
+      "step": 22642
+    },
+    {
+      "epoch": 0.9579067603012099,
+      "grad_norm": 0.16210713982582092,
+      "learning_rate": 0.001,
+      "loss": 2.5452,
+      "step": 22643
+    },
+    {
+      "epoch": 0.9579490650647263,
+      "grad_norm": 0.14848294854164124,
+      "learning_rate": 0.001,
+      "loss": 2.117,
+      "step": 22644
+    },
+    {
+      "epoch": 0.9579913698282426,
+      "grad_norm": 0.7370397448539734,
+      "learning_rate": 0.001,
+      "loss": 1.9066,
+      "step": 22645
+    },
+    {
+      "epoch": 0.958033674591759,
+      "grad_norm": 0.15546540915966034,
+      "learning_rate": 0.001,
+      "loss": 1.5374,
+      "step": 22646
+    },
+    {
+      "epoch": 0.9580759793552754,
+      "grad_norm": 0.1666499376296997,
+      "learning_rate": 0.001,
+      "loss": 1.7665,
+      "step": 22647
+    },
+    {
+      "epoch": 0.9581182841187917,
+      "grad_norm": 0.16058315336704254,
+      "learning_rate": 0.001,
+      "loss": 1.7836,
+      "step": 22648
+    },
+    {
+      "epoch": 0.9581605888823082,
+      "grad_norm": 0.1704726219177246,
+      "learning_rate": 0.001,
+      "loss": 2.0733,
+      "step": 22649
+    },
+    {
+      "epoch": 0.9582028936458246,
+      "grad_norm": 1.3915730714797974,
+      "learning_rate": 0.001,
+      "loss": 2.1041,
+      "step": 22650
+    },
+    {
+      "epoch": 0.9582451984093409,
+      "grad_norm": 0.14765958487987518,
+      "learning_rate": 0.001,
+      "loss": 2.1703,
+      "step": 22651
+    },
+    {
+      "epoch": 0.9582875031728573,
+      "grad_norm": 0.136846661567688,
+      "learning_rate": 0.001,
+      "loss": 2.0503,
+      "step": 22652
+    },
+    {
+      "epoch": 0.9583298079363737,
+      "grad_norm": 2.0541558265686035,
+      "learning_rate": 0.001,
+      "loss": 1.9055,
+      "step": 22653
+    },
+    {
+      "epoch": 0.95837211269989,
+      "grad_norm": 0.168845996260643,
+      "learning_rate": 0.001,
+      "loss": 2.0872,
+      "step": 22654
+    },
+    {
+      "epoch": 0.9584144174634064,
+      "grad_norm": 0.18082398176193237,
+      "learning_rate": 0.001,
+      "loss": 2.6359,
+      "step": 22655
+    },
+    {
+      "epoch": 0.9584567222269228,
+      "grad_norm": 0.2050907164812088,
+      "learning_rate": 0.001,
+      "loss": 2.1573,
+      "step": 22656
+    },
+    {
+      "epoch": 0.9584990269904391,
+      "grad_norm": 0.23154161870479584,
+      "learning_rate": 0.001,
+      "loss": 2.4487,
+      "step": 22657
+    },
+    {
+      "epoch": 0.9585413317539555,
+      "grad_norm": 0.228712260723114,
+      "learning_rate": 0.001,
+      "loss": 2.5394,
+      "step": 22658
+    },
+    {
+      "epoch": 0.9585836365174719,
+      "grad_norm": 0.18508756160736084,
+      "learning_rate": 0.001,
+      "loss": 1.982,
+      "step": 22659
+    },
+    {
+      "epoch": 0.9586259412809882,
+      "grad_norm": 0.2040599286556244,
+      "learning_rate": 0.001,
+      "loss": 1.9579,
+      "step": 22660
+    },
+    {
+      "epoch": 0.9586682460445046,
+      "grad_norm": 0.2641620337963104,
+      "learning_rate": 0.001,
+      "loss": 2.3784,
+      "step": 22661
+    },
+    {
+      "epoch": 0.958710550808021,
+      "grad_norm": 0.17128469049930573,
+      "learning_rate": 0.001,
+      "loss": 2.367,
+      "step": 22662
+    },
+    {
+      "epoch": 0.9587528555715373,
+      "grad_norm": 0.5706077814102173,
+      "learning_rate": 0.001,
+      "loss": 2.7794,
+      "step": 22663
+    },
+    {
+      "epoch": 0.9587951603350537,
+      "grad_norm": 0.15378667414188385,
+      "learning_rate": 0.001,
+      "loss": 1.2554,
+      "step": 22664
+    },
+    {
+      "epoch": 0.9588374650985702,
+      "grad_norm": 0.14689679443836212,
+      "learning_rate": 0.001,
+      "loss": 2.1687,
+      "step": 22665
+    },
+    {
+      "epoch": 0.9588797698620865,
+      "grad_norm": 0.19567295908927917,
+      "learning_rate": 0.001,
+      "loss": 2.2162,
+      "step": 22666
+    },
+    {
+      "epoch": 0.9589220746256029,
+      "grad_norm": 0.3699814975261688,
+      "learning_rate": 0.001,
+      "loss": 2.1986,
+      "step": 22667
+    },
+    {
+      "epoch": 0.9589643793891193,
+      "grad_norm": 0.19808320701122284,
+      "learning_rate": 0.001,
+      "loss": 2.4602,
+      "step": 22668
+    },
+    {
+      "epoch": 0.9590066841526356,
+      "grad_norm": 0.17543701827526093,
+      "learning_rate": 0.001,
+      "loss": 2.4826,
+      "step": 22669
+    },
+    {
+      "epoch": 0.959048988916152,
+      "grad_norm": 0.16463440656661987,
+      "learning_rate": 0.001,
+      "loss": 2.3312,
+      "step": 22670
+    },
+    {
+      "epoch": 0.9590912936796683,
+      "grad_norm": 15.79118537902832,
+      "learning_rate": 0.001,
+      "loss": 2.4984,
+      "step": 22671
+    },
+    {
+      "epoch": 0.9591335984431847,
+      "grad_norm": 0.457317054271698,
+      "learning_rate": 0.001,
+      "loss": 1.6768,
+      "step": 22672
+    },
+    {
+      "epoch": 0.9591759032067011,
+      "grad_norm": 0.3178519904613495,
+      "learning_rate": 0.001,
+      "loss": 1.9121,
+      "step": 22673
+    },
+    {
+      "epoch": 0.9592182079702174,
+      "grad_norm": 0.1642330437898636,
+      "learning_rate": 0.001,
+      "loss": 1.1973,
+      "step": 22674
+    },
+    {
+      "epoch": 0.9592605127337338,
+      "grad_norm": 0.9215223789215088,
+      "learning_rate": 0.001,
+      "loss": 1.7921,
+      "step": 22675
+    },
+    {
+      "epoch": 0.9593028174972502,
+      "grad_norm": 0.15992654860019684,
+      "learning_rate": 0.001,
+      "loss": 1.6263,
+      "step": 22676
+    },
+    {
+      "epoch": 0.9593451222607665,
+      "grad_norm": 0.18335092067718506,
+      "learning_rate": 0.001,
+      "loss": 1.7375,
+      "step": 22677
+    },
+    {
+      "epoch": 0.9593874270242829,
+      "grad_norm": 0.19411593675613403,
+      "learning_rate": 0.001,
+      "loss": 1.8051,
+      "step": 22678
+    },
+    {
+      "epoch": 0.9594297317877993,
+      "grad_norm": 0.16195566952228546,
+      "learning_rate": 0.001,
+      "loss": 3.0332,
+      "step": 22679
+    },
+    {
+      "epoch": 0.9594720365513156,
+      "grad_norm": 0.5709326863288879,
+      "learning_rate": 0.001,
+      "loss": 2.7656,
+      "step": 22680
+    },
+    {
+      "epoch": 0.959514341314832,
+      "grad_norm": 0.145700141787529,
+      "learning_rate": 0.001,
+      "loss": 1.9661,
+      "step": 22681
+    },
+    {
+      "epoch": 0.9595566460783485,
+      "grad_norm": 0.16970385611057281,
+      "learning_rate": 0.001,
+      "loss": 1.9235,
+      "step": 22682
+    },
+    {
+      "epoch": 0.9595989508418648,
+      "grad_norm": 0.17255772650241852,
+      "learning_rate": 0.001,
+      "loss": 2.1868,
+      "step": 22683
+    },
+    {
+      "epoch": 0.9596412556053812,
+      "grad_norm": 0.16571679711341858,
+      "learning_rate": 0.001,
+      "loss": 1.8325,
+      "step": 22684
+    },
+    {
+      "epoch": 0.9596835603688976,
+      "grad_norm": 0.12924878299236298,
+      "learning_rate": 0.001,
+      "loss": 1.8835,
+      "step": 22685
+    },
+    {
+      "epoch": 0.9597258651324139,
+      "grad_norm": 0.20774781703948975,
+      "learning_rate": 0.001,
+      "loss": 2.1721,
+      "step": 22686
+    },
+    {
+      "epoch": 0.9597681698959303,
+      "grad_norm": 0.35853028297424316,
+      "learning_rate": 0.001,
+      "loss": 2.2272,
+      "step": 22687
+    },
+    {
+      "epoch": 0.9598104746594467,
+      "grad_norm": 0.5301867127418518,
+      "learning_rate": 0.001,
+      "loss": 1.568,
+      "step": 22688
+    },
+    {
+      "epoch": 0.959852779422963,
+      "grad_norm": 0.19120998680591583,
+      "learning_rate": 0.001,
+      "loss": 2.1802,
+      "step": 22689
+    },
+    {
+      "epoch": 0.9598950841864794,
+      "grad_norm": 0.16802310943603516,
+      "learning_rate": 0.001,
+      "loss": 1.6033,
+      "step": 22690
+    },
+    {
+      "epoch": 0.9599373889499958,
+      "grad_norm": 0.17627355456352234,
+      "learning_rate": 0.001,
+      "loss": 2.273,
+      "step": 22691
+    },
+    {
+      "epoch": 0.9599796937135121,
+      "grad_norm": 0.22242601215839386,
+      "learning_rate": 0.001,
+      "loss": 1.9304,
+      "step": 22692
+    },
+    {
+      "epoch": 0.9600219984770285,
+      "grad_norm": 0.1294320523738861,
+      "learning_rate": 0.001,
+      "loss": 2.0693,
+      "step": 22693
+    },
+    {
+      "epoch": 0.9600643032405449,
+      "grad_norm": 0.17644540965557098,
+      "learning_rate": 0.001,
+      "loss": 2.0387,
+      "step": 22694
+    },
+    {
+      "epoch": 0.9601066080040612,
+      "grad_norm": 1.0202915668487549,
+      "learning_rate": 0.001,
+      "loss": 1.9817,
+      "step": 22695
+    },
+    {
+      "epoch": 0.9601489127675776,
+      "grad_norm": 0.15253128111362457,
+      "learning_rate": 0.001,
+      "loss": 2.055,
+      "step": 22696
+    },
+    {
+      "epoch": 0.960191217531094,
+      "grad_norm": 0.15679995715618134,
+      "learning_rate": 0.001,
+      "loss": 2.1456,
+      "step": 22697
+    },
+    {
+      "epoch": 0.9602335222946103,
+      "grad_norm": 0.15941359102725983,
+      "learning_rate": 0.001,
+      "loss": 2.0955,
+      "step": 22698
+    },
+    {
+      "epoch": 0.9602758270581268,
+      "grad_norm": 0.17292888462543488,
+      "learning_rate": 0.001,
+      "loss": 1.624,
+      "step": 22699
+    },
+    {
+      "epoch": 0.9603181318216432,
+      "grad_norm": 0.1328788846731186,
+      "learning_rate": 0.001,
+      "loss": 1.4056,
+      "step": 22700
+    },
+    {
+      "epoch": 0.9603604365851595,
+      "grad_norm": 0.18797191977500916,
+      "learning_rate": 0.001,
+      "loss": 2.5867,
+      "step": 22701
+    },
+    {
+      "epoch": 0.9604027413486759,
+      "grad_norm": 0.7679688930511475,
+      "learning_rate": 0.001,
+      "loss": 3.1701,
+      "step": 22702
+    },
+    {
+      "epoch": 0.9604450461121923,
+      "grad_norm": 0.6459025740623474,
+      "learning_rate": 0.001,
+      "loss": 4.0775,
+      "step": 22703
+    },
+    {
+      "epoch": 0.9604873508757086,
+      "grad_norm": 0.13661043345928192,
+      "learning_rate": 0.001,
+      "loss": 2.6037,
+      "step": 22704
+    },
+    {
+      "epoch": 0.960529655639225,
+      "grad_norm": 0.15741907060146332,
+      "learning_rate": 0.001,
+      "loss": 2.0212,
+      "step": 22705
+    },
+    {
+      "epoch": 0.9605719604027414,
+      "grad_norm": 0.9307416677474976,
+      "learning_rate": 0.001,
+      "loss": 2.2273,
+      "step": 22706
+    },
+    {
+      "epoch": 0.9606142651662577,
+      "grad_norm": 0.19062964618206024,
+      "learning_rate": 0.001,
+      "loss": 3.1016,
+      "step": 22707
+    },
+    {
+      "epoch": 0.9606565699297741,
+      "grad_norm": 0.17339986562728882,
+      "learning_rate": 0.001,
+      "loss": 1.9746,
+      "step": 22708
+    },
+    {
+      "epoch": 0.9606988746932905,
+      "grad_norm": 0.17814798653125763,
+      "learning_rate": 0.001,
+      "loss": 2.9384,
+      "step": 22709
+    },
+    {
+      "epoch": 0.9607411794568068,
+      "grad_norm": 0.19734229147434235,
+      "learning_rate": 0.001,
+      "loss": 2.3626,
+      "step": 22710
+    },
+    {
+      "epoch": 0.9607834842203232,
+      "grad_norm": 0.2042291760444641,
+      "learning_rate": 0.001,
+      "loss": 1.851,
+      "step": 22711
+    },
+    {
+      "epoch": 0.9608257889838396,
+      "grad_norm": 1.1738054752349854,
+      "learning_rate": 0.001,
+      "loss": 3.5114,
+      "step": 22712
+    },
+    {
+      "epoch": 0.9608680937473559,
+      "grad_norm": 3.5392398834228516,
+      "learning_rate": 0.001,
+      "loss": 2.1209,
+      "step": 22713
+    },
+    {
+      "epoch": 0.9609103985108723,
+      "grad_norm": 0.2352452278137207,
+      "learning_rate": 0.001,
+      "loss": 2.2434,
+      "step": 22714
+    },
+    {
+      "epoch": 0.9609527032743886,
+      "grad_norm": 0.696122407913208,
+      "learning_rate": 0.001,
+      "loss": 2.0429,
+      "step": 22715
+    },
+    {
+      "epoch": 0.960995008037905,
+      "grad_norm": 0.24021172523498535,
+      "learning_rate": 0.001,
+      "loss": 1.927,
+      "step": 22716
+    },
+    {
+      "epoch": 0.9610373128014215,
+      "grad_norm": 0.1911345273256302,
+      "learning_rate": 0.001,
+      "loss": 2.3258,
+      "step": 22717
+    },
+    {
+      "epoch": 0.9610796175649378,
+      "grad_norm": 0.1680738478899002,
+      "learning_rate": 0.001,
+      "loss": 1.7179,
+      "step": 22718
+    },
+    {
+      "epoch": 0.9611219223284542,
+      "grad_norm": 0.20426170527935028,
+      "learning_rate": 0.001,
+      "loss": 1.8262,
+      "step": 22719
+    },
+    {
+      "epoch": 0.9611642270919706,
+      "grad_norm": 0.9206085205078125,
+      "learning_rate": 0.001,
+      "loss": 1.7566,
+      "step": 22720
+    },
+    {
+      "epoch": 0.9612065318554869,
+      "grad_norm": 0.17064528167247772,
+      "learning_rate": 0.001,
+      "loss": 3.1635,
+      "step": 22721
+    },
+    {
+      "epoch": 0.9612488366190033,
+      "grad_norm": 0.1871117353439331,
+      "learning_rate": 0.001,
+      "loss": 3.1601,
+      "step": 22722
+    },
+    {
+      "epoch": 0.9612911413825197,
+      "grad_norm": 0.14905090630054474,
+      "learning_rate": 0.001,
+      "loss": 2.5749,
+      "step": 22723
+    },
+    {
+      "epoch": 0.961333446146036,
+      "grad_norm": 0.3839198350906372,
+      "learning_rate": 0.001,
+      "loss": 3.2174,
+      "step": 22724
+    },
+    {
+      "epoch": 0.9613757509095524,
+      "grad_norm": 0.15603891015052795,
+      "learning_rate": 0.001,
+      "loss": 1.784,
+      "step": 22725
+    },
+    {
+      "epoch": 0.9614180556730688,
+      "grad_norm": 0.16384871304035187,
+      "learning_rate": 0.001,
+      "loss": 2.1131,
+      "step": 22726
+    },
+    {
+      "epoch": 0.9614603604365851,
+      "grad_norm": 0.1653030961751938,
+      "learning_rate": 0.001,
+      "loss": 2.8662,
+      "step": 22727
+    },
+    {
+      "epoch": 0.9615026652001015,
+      "grad_norm": 0.15746630728244781,
+      "learning_rate": 0.001,
+      "loss": 2.3647,
+      "step": 22728
+    },
+    {
+      "epoch": 0.9615449699636179,
+      "grad_norm": 0.1539701670408249,
+      "learning_rate": 0.001,
+      "loss": 1.9527,
+      "step": 22729
+    },
+    {
+      "epoch": 0.9615872747271342,
+      "grad_norm": 0.15698012709617615,
+      "learning_rate": 0.001,
+      "loss": 1.9104,
+      "step": 22730
+    },
+    {
+      "epoch": 0.9616295794906506,
+      "grad_norm": 0.17898909747600555,
+      "learning_rate": 0.001,
+      "loss": 2.0614,
+      "step": 22731
+    },
+    {
+      "epoch": 0.961671884254167,
+      "grad_norm": 0.1490541249513626,
+      "learning_rate": 0.001,
+      "loss": 1.9078,
+      "step": 22732
+    },
+    {
+      "epoch": 0.9617141890176834,
+      "grad_norm": 0.17301572859287262,
+      "learning_rate": 0.001,
+      "loss": 1.5119,
+      "step": 22733
+    },
+    {
+      "epoch": 0.9617564937811998,
+      "grad_norm": 1.670694351196289,
+      "learning_rate": 0.001,
+      "loss": 1.6093,
+      "step": 22734
+    },
+    {
+      "epoch": 0.9617987985447162,
+      "grad_norm": 0.16678212583065033,
+      "learning_rate": 0.001,
+      "loss": 1.5122,
+      "step": 22735
+    },
+    {
+      "epoch": 0.9618411033082325,
+      "grad_norm": 0.15618404746055603,
+      "learning_rate": 0.001,
+      "loss": 1.4957,
+      "step": 22736
+    },
+    {
+      "epoch": 0.9618834080717489,
+      "grad_norm": 0.15901844203472137,
+      "learning_rate": 0.001,
+      "loss": 2.9647,
+      "step": 22737
+    },
+    {
+      "epoch": 0.9619257128352653,
+      "grad_norm": 0.162552148103714,
+      "learning_rate": 0.001,
+      "loss": 2.2336,
+      "step": 22738
+    },
+    {
+      "epoch": 0.9619680175987816,
+      "grad_norm": 0.6180990934371948,
+      "learning_rate": 0.001,
+      "loss": 2.6319,
+      "step": 22739
+    },
+    {
+      "epoch": 0.962010322362298,
+      "grad_norm": 0.16461017727851868,
+      "learning_rate": 0.001,
+      "loss": 2.8503,
+      "step": 22740
+    },
+    {
+      "epoch": 0.9620526271258144,
+      "grad_norm": 0.15487508475780487,
+      "learning_rate": 0.001,
+      "loss": 1.7637,
+      "step": 22741
+    },
+    {
+      "epoch": 0.9620949318893307,
+      "grad_norm": 0.16154402494430542,
+      "learning_rate": 0.001,
+      "loss": 1.7477,
+      "step": 22742
+    },
+    {
+      "epoch": 0.9621372366528471,
+      "grad_norm": 0.162453293800354,
+      "learning_rate": 0.001,
+      "loss": 1.1528,
+      "step": 22743
+    },
+    {
+      "epoch": 0.9621795414163635,
+      "grad_norm": 0.15126554667949677,
+      "learning_rate": 0.001,
+      "loss": 2.702,
+      "step": 22744
+    },
+    {
+      "epoch": 0.9622218461798798,
+      "grad_norm": 0.13284894824028015,
+      "learning_rate": 0.001,
+      "loss": 3.2294,
+      "step": 22745
+    },
+    {
+      "epoch": 0.9622641509433962,
+      "grad_norm": 1.1502666473388672,
+      "learning_rate": 0.001,
+      "loss": 2.4184,
+      "step": 22746
+    },
+    {
+      "epoch": 0.9623064557069126,
+      "grad_norm": 0.1922639161348343,
+      "learning_rate": 0.001,
+      "loss": 2.0937,
+      "step": 22747
+    },
+    {
+      "epoch": 0.9623487604704289,
+      "grad_norm": 0.16614429652690887,
+      "learning_rate": 0.001,
+      "loss": 1.959,
+      "step": 22748
+    },
+    {
+      "epoch": 0.9623910652339454,
+      "grad_norm": 0.1562388688325882,
+      "learning_rate": 0.001,
+      "loss": 2.9115,
+      "step": 22749
+    },
+    {
+      "epoch": 0.9624333699974618,
+      "grad_norm": 0.15959501266479492,
+      "learning_rate": 0.001,
+      "loss": 1.7341,
+      "step": 22750
+    },
+    {
+      "epoch": 0.9624756747609781,
+      "grad_norm": 0.5524002909660339,
+      "learning_rate": 0.001,
+      "loss": 1.8583,
+      "step": 22751
+    },
+    {
+      "epoch": 0.9625179795244945,
+      "grad_norm": 0.16397516429424286,
+      "learning_rate": 0.001,
+      "loss": 1.9359,
+      "step": 22752
+    },
+    {
+      "epoch": 0.9625602842880109,
+      "grad_norm": 0.19087645411491394,
+      "learning_rate": 0.001,
+      "loss": 2.3213,
+      "step": 22753
+    },
+    {
+      "epoch": 0.9626025890515272,
+      "grad_norm": 0.18343958258628845,
+      "learning_rate": 0.001,
+      "loss": 2.2451,
+      "step": 22754
+    },
+    {
+      "epoch": 0.9626448938150436,
+      "grad_norm": 0.17732329666614532,
+      "learning_rate": 0.001,
+      "loss": 2.5962,
+      "step": 22755
+    },
+    {
+      "epoch": 0.96268719857856,
+      "grad_norm": 0.1700247973203659,
+      "learning_rate": 0.001,
+      "loss": 1.6104,
+      "step": 22756
+    },
+    {
+      "epoch": 0.9627295033420763,
+      "grad_norm": 18.20061492919922,
+      "learning_rate": 0.001,
+      "loss": 2.0838,
+      "step": 22757
+    },
+    {
+      "epoch": 0.9627718081055927,
+      "grad_norm": 0.8646219372749329,
+      "learning_rate": 0.001,
+      "loss": 3.463,
+      "step": 22758
+    },
+    {
+      "epoch": 0.962814112869109,
+      "grad_norm": 0.6529208421707153,
+      "learning_rate": 0.001,
+      "loss": 3.052,
+      "step": 22759
+    },
+    {
+      "epoch": 0.9628564176326254,
+      "grad_norm": 0.16437289118766785,
+      "learning_rate": 0.001,
+      "loss": 1.9609,
+      "step": 22760
+    },
+    {
+      "epoch": 0.9628987223961418,
+      "grad_norm": 0.2545870840549469,
+      "learning_rate": 0.001,
+      "loss": 2.0325,
+      "step": 22761
+    },
+    {
+      "epoch": 0.9629410271596581,
+      "grad_norm": 0.4441566467285156,
+      "learning_rate": 0.001,
+      "loss": 2.3939,
+      "step": 22762
+    },
+    {
+      "epoch": 0.9629833319231745,
+      "grad_norm": 0.17705874145030975,
+      "learning_rate": 0.001,
+      "loss": 2.0372,
+      "step": 22763
+    },
+    {
+      "epoch": 0.963025636686691,
+      "grad_norm": 0.14965033531188965,
+      "learning_rate": 0.001,
+      "loss": 1.515,
+      "step": 22764
+    },
+    {
+      "epoch": 0.9630679414502072,
+      "grad_norm": 0.1866179257631302,
+      "learning_rate": 0.001,
+      "loss": 2.3098,
+      "step": 22765
+    },
+    {
+      "epoch": 0.9631102462137237,
+      "grad_norm": 2.7773499488830566,
+      "learning_rate": 0.001,
+      "loss": 2.5506,
+      "step": 22766
+    },
+    {
+      "epoch": 0.9631525509772401,
+      "grad_norm": 0.18301425874233246,
+      "learning_rate": 0.001,
+      "loss": 1.7283,
+      "step": 22767
+    },
+    {
+      "epoch": 0.9631948557407564,
+      "grad_norm": 0.15599261224269867,
+      "learning_rate": 0.001,
+      "loss": 2.4437,
+      "step": 22768
+    },
+    {
+      "epoch": 0.9632371605042728,
+      "grad_norm": 0.35363784432411194,
+      "learning_rate": 0.001,
+      "loss": 2.3903,
+      "step": 22769
+    },
+    {
+      "epoch": 0.9632794652677892,
+      "grad_norm": 5.986473560333252,
+      "learning_rate": 0.001,
+      "loss": 1.5687,
+      "step": 22770
+    },
+    {
+      "epoch": 0.9633217700313055,
+      "grad_norm": 0.20316632091999054,
+      "learning_rate": 0.001,
+      "loss": 2.2692,
+      "step": 22771
+    },
+    {
+      "epoch": 0.9633640747948219,
+      "grad_norm": 0.16965316236019135,
+      "learning_rate": 0.001,
+      "loss": 2.4032,
+      "step": 22772
+    },
+    {
+      "epoch": 0.9634063795583383,
+      "grad_norm": 0.18180473148822784,
+      "learning_rate": 0.001,
+      "loss": 2.0577,
+      "step": 22773
+    },
+    {
+      "epoch": 0.9634486843218546,
+      "grad_norm": 0.16141119599342346,
+      "learning_rate": 0.001,
+      "loss": 1.8242,
+      "step": 22774
+    },
+    {
+      "epoch": 0.963490989085371,
+      "grad_norm": 0.18191201984882355,
+      "learning_rate": 0.001,
+      "loss": 2.1905,
+      "step": 22775
+    },
+    {
+      "epoch": 0.9635332938488874,
+      "grad_norm": 0.783811092376709,
+      "learning_rate": 0.001,
+      "loss": 3.1841,
+      "step": 22776
+    },
+    {
+      "epoch": 0.9635755986124037,
+      "grad_norm": 0.3125717043876648,
+      "learning_rate": 0.001,
+      "loss": 1.7756,
+      "step": 22777
+    },
+    {
+      "epoch": 0.9636179033759201,
+      "grad_norm": 0.25806742906570435,
+      "learning_rate": 0.001,
+      "loss": 1.7628,
+      "step": 22778
+    },
+    {
+      "epoch": 0.9636602081394365,
+      "grad_norm": 0.16139912605285645,
+      "learning_rate": 0.001,
+      "loss": 1.7258,
+      "step": 22779
+    },
+    {
+      "epoch": 0.9637025129029528,
+      "grad_norm": 0.15739895403385162,
+      "learning_rate": 0.001,
+      "loss": 1.9454,
+      "step": 22780
+    },
+    {
+      "epoch": 0.9637448176664692,
+      "grad_norm": 0.15625838935375214,
+      "learning_rate": 0.001,
+      "loss": 2.3343,
+      "step": 22781
+    },
+    {
+      "epoch": 0.9637871224299857,
+      "grad_norm": 0.15970557928085327,
+      "learning_rate": 0.001,
+      "loss": 2.6073,
+      "step": 22782
+    },
+    {
+      "epoch": 0.963829427193502,
+      "grad_norm": 2.783663272857666,
+      "learning_rate": 0.001,
+      "loss": 2.5539,
+      "step": 22783
+    },
+    {
+      "epoch": 0.9638717319570184,
+      "grad_norm": 0.15198123455047607,
+      "learning_rate": 0.001,
+      "loss": 1.7133,
+      "step": 22784
+    },
+    {
+      "epoch": 0.9639140367205348,
+      "grad_norm": 0.25221678614616394,
+      "learning_rate": 0.001,
+      "loss": 3.6873,
+      "step": 22785
+    },
+    {
+      "epoch": 0.9639563414840511,
+      "grad_norm": 0.1301819235086441,
+      "learning_rate": 0.001,
+      "loss": 2.8356,
+      "step": 22786
+    },
+    {
+      "epoch": 0.9639986462475675,
+      "grad_norm": 0.2297348827123642,
+      "learning_rate": 0.001,
+      "loss": 2.3075,
+      "step": 22787
+    },
+    {
+      "epoch": 0.9640409510110839,
+      "grad_norm": 0.15260782837867737,
+      "learning_rate": 0.001,
+      "loss": 1.8023,
+      "step": 22788
+    },
+    {
+      "epoch": 0.9640832557746002,
+      "grad_norm": 0.18781907856464386,
+      "learning_rate": 0.001,
+      "loss": 2.7565,
+      "step": 22789
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 0.38502037525177,
+      "learning_rate": 0.001,
+      "loss": 2.6757,
+      "step": 22790
+    },
+    {
+      "epoch": 0.964167865301633,
+      "grad_norm": 0.1971847265958786,
+      "learning_rate": 0.001,
+      "loss": 1.6928,
+      "step": 22791
+    },
+    {
+      "epoch": 0.9642101700651493,
+      "grad_norm": 0.1461469531059265,
+      "learning_rate": 0.001,
+      "loss": 2.707,
+      "step": 22792
+    },
+    {
+      "epoch": 0.9642524748286657,
+      "grad_norm": 0.15411047637462616,
+      "learning_rate": 0.001,
+      "loss": 1.8741,
+      "step": 22793
+    },
+    {
+      "epoch": 0.9642947795921821,
+      "grad_norm": 0.15573959052562714,
+      "learning_rate": 0.001,
+      "loss": 2.3374,
+      "step": 22794
+    },
+    {
+      "epoch": 0.9643370843556984,
+      "grad_norm": 0.1828451305627823,
+      "learning_rate": 0.001,
+      "loss": 2.5809,
+      "step": 22795
+    },
+    {
+      "epoch": 0.9643793891192148,
+      "grad_norm": 2.2866430282592773,
+      "learning_rate": 0.001,
+      "loss": 2.0316,
+      "step": 22796
+    },
+    {
+      "epoch": 0.9644216938827312,
+      "grad_norm": 5.461979866027832,
+      "learning_rate": 0.001,
+      "loss": 1.6355,
+      "step": 22797
+    },
+    {
+      "epoch": 0.9644639986462475,
+      "grad_norm": 0.24063386023044586,
+      "learning_rate": 0.001,
+      "loss": 1.7713,
+      "step": 22798
+    },
+    {
+      "epoch": 0.964506303409764,
+      "grad_norm": 0.15354004502296448,
+      "learning_rate": 0.001,
+      "loss": 1.7979,
+      "step": 22799
+    },
+    {
+      "epoch": 0.9645486081732804,
+      "grad_norm": 0.20435552299022675,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 22800
+    },
+    {
+      "epoch": 0.9645909129367967,
+      "grad_norm": 0.17988991737365723,
+      "learning_rate": 0.001,
+      "loss": 3.3384,
+      "step": 22801
+    },
+    {
+      "epoch": 0.9646332177003131,
+      "grad_norm": 0.19008944928646088,
+      "learning_rate": 0.001,
+      "loss": 1.8226,
+      "step": 22802
+    },
+    {
+      "epoch": 0.9646755224638295,
+      "grad_norm": 0.20340676605701447,
+      "learning_rate": 0.001,
+      "loss": 3.0798,
+      "step": 22803
+    },
+    {
+      "epoch": 0.9647178272273458,
+      "grad_norm": 0.18513543903827667,
+      "learning_rate": 0.001,
+      "loss": 1.7459,
+      "step": 22804
+    },
+    {
+      "epoch": 0.9647601319908622,
+      "grad_norm": 0.18096236884593964,
+      "learning_rate": 0.001,
+      "loss": 2.0214,
+      "step": 22805
+    },
+    {
+      "epoch": 0.9648024367543785,
+      "grad_norm": 0.20372340083122253,
+      "learning_rate": 0.001,
+      "loss": 2.1574,
+      "step": 22806
+    },
+    {
+      "epoch": 0.9648447415178949,
+      "grad_norm": 0.1942087858915329,
+      "learning_rate": 0.001,
+      "loss": 1.8102,
+      "step": 22807
+    },
+    {
+      "epoch": 0.9648870462814113,
+      "grad_norm": 0.16786138713359833,
+      "learning_rate": 0.001,
+      "loss": 1.7745,
+      "step": 22808
+    },
+    {
+      "epoch": 0.9649293510449276,
+      "grad_norm": 21.640262603759766,
+      "learning_rate": 0.001,
+      "loss": 2.6787,
+      "step": 22809
+    },
+    {
+      "epoch": 0.964971655808444,
+      "grad_norm": 0.17724758386611938,
+      "learning_rate": 0.001,
+      "loss": 2.0269,
+      "step": 22810
+    },
+    {
+      "epoch": 0.9650139605719604,
+      "grad_norm": 0.17593099176883698,
+      "learning_rate": 0.001,
+      "loss": 2.6293,
+      "step": 22811
+    },
+    {
+      "epoch": 0.9650562653354767,
+      "grad_norm": 0.1671978384256363,
+      "learning_rate": 0.001,
+      "loss": 2.3804,
+      "step": 22812
+    },
+    {
+      "epoch": 0.9650985700989931,
+      "grad_norm": 0.19632820785045624,
+      "learning_rate": 0.001,
+      "loss": 1.808,
+      "step": 22813
+    },
+    {
+      "epoch": 0.9651408748625095,
+      "grad_norm": 0.35916808247566223,
+      "learning_rate": 0.001,
+      "loss": 1.8704,
+      "step": 22814
+    },
+    {
+      "epoch": 0.9651831796260258,
+      "grad_norm": 0.16983725130558014,
+      "learning_rate": 0.001,
+      "loss": 3.2959,
+      "step": 22815
+    },
+    {
+      "epoch": 0.9652254843895423,
+      "grad_norm": 0.6574493050575256,
+      "learning_rate": 0.001,
+      "loss": 3.0752,
+      "step": 22816
+    },
+    {
+      "epoch": 0.9652677891530587,
+      "grad_norm": 0.14061124622821808,
+      "learning_rate": 0.001,
+      "loss": 2.3643,
+      "step": 22817
+    },
+    {
+      "epoch": 0.965310093916575,
+      "grad_norm": 0.12727145850658417,
+      "learning_rate": 0.001,
+      "loss": 2.3003,
+      "step": 22818
+    },
+    {
+      "epoch": 0.9653523986800914,
+      "grad_norm": 0.15103477239608765,
+      "learning_rate": 0.001,
+      "loss": 2.3464,
+      "step": 22819
+    },
+    {
+      "epoch": 0.9653947034436078,
+      "grad_norm": 0.15163427591323853,
+      "learning_rate": 0.001,
+      "loss": 1.8442,
+      "step": 22820
+    },
+    {
+      "epoch": 0.9654370082071241,
+      "grad_norm": 0.16377829015254974,
+      "learning_rate": 0.001,
+      "loss": 2.1828,
+      "step": 22821
+    },
+    {
+      "epoch": 0.9654793129706405,
+      "grad_norm": 0.44327592849731445,
+      "learning_rate": 0.001,
+      "loss": 2.5437,
+      "step": 22822
+    },
+    {
+      "epoch": 0.9655216177341569,
+      "grad_norm": 0.2196311205625534,
+      "learning_rate": 0.001,
+      "loss": 1.9828,
+      "step": 22823
+    },
+    {
+      "epoch": 0.9655639224976732,
+      "grad_norm": 0.13134104013442993,
+      "learning_rate": 0.001,
+      "loss": 1.5855,
+      "step": 22824
+    },
+    {
+      "epoch": 0.9656062272611896,
+      "grad_norm": 0.4789327085018158,
+      "learning_rate": 0.001,
+      "loss": 2.3847,
+      "step": 22825
+    },
+    {
+      "epoch": 0.965648532024706,
+      "grad_norm": 0.14557301998138428,
+      "learning_rate": 0.001,
+      "loss": 2.6556,
+      "step": 22826
+    },
+    {
+      "epoch": 0.9656908367882223,
+      "grad_norm": 0.1414880007505417,
+      "learning_rate": 0.001,
+      "loss": 1.5357,
+      "step": 22827
+    },
+    {
+      "epoch": 0.9657331415517387,
+      "grad_norm": 0.14299647510051727,
+      "learning_rate": 0.001,
+      "loss": 2.0178,
+      "step": 22828
+    },
+    {
+      "epoch": 0.9657754463152551,
+      "grad_norm": 1.3442524671554565,
+      "learning_rate": 0.001,
+      "loss": 1.8132,
+      "step": 22829
+    },
+    {
+      "epoch": 0.9658177510787714,
+      "grad_norm": 0.14745590090751648,
+      "learning_rate": 0.001,
+      "loss": 2.4792,
+      "step": 22830
+    },
+    {
+      "epoch": 0.9658600558422878,
+      "grad_norm": 0.14321614801883698,
+      "learning_rate": 0.001,
+      "loss": 2.3954,
+      "step": 22831
+    },
+    {
+      "epoch": 0.9659023606058043,
+      "grad_norm": 0.15746840834617615,
+      "learning_rate": 0.001,
+      "loss": 1.694,
+      "step": 22832
+    },
+    {
+      "epoch": 0.9659446653693206,
+      "grad_norm": 0.15602366626262665,
+      "learning_rate": 0.001,
+      "loss": 2.3334,
+      "step": 22833
+    },
+    {
+      "epoch": 0.965986970132837,
+      "grad_norm": 0.1474500298500061,
+      "learning_rate": 0.001,
+      "loss": 2.1187,
+      "step": 22834
+    },
+    {
+      "epoch": 0.9660292748963534,
+      "grad_norm": 1.2670032978057861,
+      "learning_rate": 0.001,
+      "loss": 1.8418,
+      "step": 22835
+    },
+    {
+      "epoch": 0.9660715796598697,
+      "grad_norm": 0.20300118625164032,
+      "learning_rate": 0.001,
+      "loss": 1.9915,
+      "step": 22836
+    },
+    {
+      "epoch": 0.9661138844233861,
+      "grad_norm": 0.20837882161140442,
+      "learning_rate": 0.001,
+      "loss": 2.6805,
+      "step": 22837
+    },
+    {
+      "epoch": 0.9661561891869025,
+      "grad_norm": 0.1467253714799881,
+      "learning_rate": 0.001,
+      "loss": 1.7238,
+      "step": 22838
+    },
+    {
+      "epoch": 0.9661984939504188,
+      "grad_norm": 8.646101951599121,
+      "learning_rate": 0.001,
+      "loss": 2.1153,
+      "step": 22839
+    },
+    {
+      "epoch": 0.9662407987139352,
+      "grad_norm": 0.12623724341392517,
+      "learning_rate": 0.001,
+      "loss": 1.9916,
+      "step": 22840
+    },
+    {
+      "epoch": 0.9662831034774516,
+      "grad_norm": 0.1795465052127838,
+      "learning_rate": 0.001,
+      "loss": 2.2702,
+      "step": 22841
+    },
+    {
+      "epoch": 0.9663254082409679,
+      "grad_norm": 0.14602890610694885,
+      "learning_rate": 0.001,
+      "loss": 2.0795,
+      "step": 22842
+    },
+    {
+      "epoch": 0.9663677130044843,
+      "grad_norm": 0.18272976577281952,
+      "learning_rate": 0.001,
+      "loss": 2.061,
+      "step": 22843
+    },
+    {
+      "epoch": 0.9664100177680007,
+      "grad_norm": 0.19199426472187042,
+      "learning_rate": 0.001,
+      "loss": 2.0847,
+      "step": 22844
+    },
+    {
+      "epoch": 0.966452322531517,
+      "grad_norm": 0.19457626342773438,
+      "learning_rate": 0.001,
+      "loss": 1.9797,
+      "step": 22845
+    },
+    {
+      "epoch": 0.9664946272950334,
+      "grad_norm": 0.29736241698265076,
+      "learning_rate": 0.001,
+      "loss": 3.3281,
+      "step": 22846
+    },
+    {
+      "epoch": 0.9665369320585498,
+      "grad_norm": 0.29135966300964355,
+      "learning_rate": 0.001,
+      "loss": 5.947,
+      "step": 22847
+    },
+    {
+      "epoch": 0.9665792368220661,
+      "grad_norm": 3.7206881046295166,
+      "learning_rate": 0.001,
+      "loss": 1.9796,
+      "step": 22848
+    },
+    {
+      "epoch": 0.9666215415855826,
+      "grad_norm": 0.14259220659732819,
+      "learning_rate": 0.001,
+      "loss": 3.0848,
+      "step": 22849
+    },
+    {
+      "epoch": 0.9666638463490989,
+      "grad_norm": 2.7031147480010986,
+      "learning_rate": 0.001,
+      "loss": 2.8408,
+      "step": 22850
+    },
+    {
+      "epoch": 0.9667061511126153,
+      "grad_norm": 0.1781085729598999,
+      "learning_rate": 0.001,
+      "loss": 2.7997,
+      "step": 22851
+    },
+    {
+      "epoch": 0.9667484558761317,
+      "grad_norm": 0.1732090413570404,
+      "learning_rate": 0.001,
+      "loss": 1.9524,
+      "step": 22852
+    },
+    {
+      "epoch": 0.966790760639648,
+      "grad_norm": 0.13723929226398468,
+      "learning_rate": 0.001,
+      "loss": 1.7664,
+      "step": 22853
+    },
+    {
+      "epoch": 0.9668330654031644,
+      "grad_norm": 0.1692230999469757,
+      "learning_rate": 0.001,
+      "loss": 2.081,
+      "step": 22854
+    },
+    {
+      "epoch": 0.9668753701666808,
+      "grad_norm": 0.1593841016292572,
+      "learning_rate": 0.001,
+      "loss": 1.6948,
+      "step": 22855
+    },
+    {
+      "epoch": 0.9669176749301971,
+      "grad_norm": 0.13628552854061127,
+      "learning_rate": 0.001,
+      "loss": 2.2481,
+      "step": 22856
+    },
+    {
+      "epoch": 0.9669599796937135,
+      "grad_norm": 0.1631813496351242,
+      "learning_rate": 0.001,
+      "loss": 1.9093,
+      "step": 22857
+    },
+    {
+      "epoch": 0.9670022844572299,
+      "grad_norm": 0.16931277513504028,
+      "learning_rate": 0.001,
+      "loss": 3.1397,
+      "step": 22858
+    },
+    {
+      "epoch": 0.9670445892207462,
+      "grad_norm": 0.14777718484401703,
+      "learning_rate": 0.001,
+      "loss": 1.9937,
+      "step": 22859
+    },
+    {
+      "epoch": 0.9670868939842626,
+      "grad_norm": 0.23261679708957672,
+      "learning_rate": 0.001,
+      "loss": 1.5095,
+      "step": 22860
+    },
+    {
+      "epoch": 0.967129198747779,
+      "grad_norm": 0.23100370168685913,
+      "learning_rate": 0.001,
+      "loss": 3.1784,
+      "step": 22861
+    },
+    {
+      "epoch": 0.9671715035112953,
+      "grad_norm": 0.17123155295848846,
+      "learning_rate": 0.001,
+      "loss": 2.0223,
+      "step": 22862
+    },
+    {
+      "epoch": 0.9672138082748117,
+      "grad_norm": 0.16940762102603912,
+      "learning_rate": 0.001,
+      "loss": 2.0325,
+      "step": 22863
+    },
+    {
+      "epoch": 0.9672561130383281,
+      "grad_norm": 0.15157361328601837,
+      "learning_rate": 0.001,
+      "loss": 2.8099,
+      "step": 22864
+    },
+    {
+      "epoch": 0.9672984178018444,
+      "grad_norm": 0.1540924608707428,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 22865
+    },
+    {
+      "epoch": 0.9673407225653609,
+      "grad_norm": 0.17207752168178558,
+      "learning_rate": 0.001,
+      "loss": 1.7778,
+      "step": 22866
+    },
+    {
+      "epoch": 0.9673830273288773,
+      "grad_norm": 0.2308436930179596,
+      "learning_rate": 0.001,
+      "loss": 2.6066,
+      "step": 22867
+    },
+    {
+      "epoch": 0.9674253320923936,
+      "grad_norm": 0.15052172541618347,
+      "learning_rate": 0.001,
+      "loss": 1.6072,
+      "step": 22868
+    },
+    {
+      "epoch": 0.96746763685591,
+      "grad_norm": 0.13643066585063934,
+      "learning_rate": 0.001,
+      "loss": 1.8093,
+      "step": 22869
+    },
+    {
+      "epoch": 0.9675099416194264,
+      "grad_norm": 0.16846615076065063,
+      "learning_rate": 0.001,
+      "loss": 1.918,
+      "step": 22870
+    },
+    {
+      "epoch": 0.9675522463829427,
+      "grad_norm": 1.7029542922973633,
+      "learning_rate": 0.001,
+      "loss": 1.7533,
+      "step": 22871
+    },
+    {
+      "epoch": 0.9675945511464591,
+      "grad_norm": 0.16174376010894775,
+      "learning_rate": 0.001,
+      "loss": 2.0584,
+      "step": 22872
+    },
+    {
+      "epoch": 0.9676368559099755,
+      "grad_norm": 0.17282286286354065,
+      "learning_rate": 0.001,
+      "loss": 1.9735,
+      "step": 22873
+    },
+    {
+      "epoch": 0.9676791606734918,
+      "grad_norm": 3.2396557331085205,
+      "learning_rate": 0.001,
+      "loss": 2.9205,
+      "step": 22874
+    },
+    {
+      "epoch": 0.9677214654370082,
+      "grad_norm": 0.18239837884902954,
+      "learning_rate": 0.001,
+      "loss": 1.9157,
+      "step": 22875
+    },
+    {
+      "epoch": 0.9677637702005246,
+      "grad_norm": 0.16040238738059998,
+      "learning_rate": 0.001,
+      "loss": 2.5035,
+      "step": 22876
+    },
+    {
+      "epoch": 0.9678060749640409,
+      "grad_norm": 0.31164249777793884,
+      "learning_rate": 0.001,
+      "loss": 2.2374,
+      "step": 22877
+    },
+    {
+      "epoch": 0.9678483797275573,
+      "grad_norm": 0.1581432968378067,
+      "learning_rate": 0.001,
+      "loss": 1.8085,
+      "step": 22878
+    },
+    {
+      "epoch": 0.9678906844910737,
+      "grad_norm": 0.17101150751113892,
+      "learning_rate": 0.001,
+      "loss": 1.9065,
+      "step": 22879
+    },
+    {
+      "epoch": 0.96793298925459,
+      "grad_norm": 0.1634765863418579,
+      "learning_rate": 0.001,
+      "loss": 1.3569,
+      "step": 22880
+    },
+    {
+      "epoch": 0.9679752940181064,
+      "grad_norm": 0.1971825212240219,
+      "learning_rate": 0.001,
+      "loss": 1.9538,
+      "step": 22881
+    },
+    {
+      "epoch": 0.9680175987816229,
+      "grad_norm": 0.17837342619895935,
+      "learning_rate": 0.001,
+      "loss": 1.7106,
+      "step": 22882
+    },
+    {
+      "epoch": 0.9680599035451392,
+      "grad_norm": 0.18054227530956268,
+      "learning_rate": 0.001,
+      "loss": 2.0404,
+      "step": 22883
+    },
+    {
+      "epoch": 0.9681022083086556,
+      "grad_norm": 0.16137461364269257,
+      "learning_rate": 0.001,
+      "loss": 1.4786,
+      "step": 22884
+    },
+    {
+      "epoch": 0.968144513072172,
+      "grad_norm": 0.1961548924446106,
+      "learning_rate": 0.001,
+      "loss": 2.127,
+      "step": 22885
+    },
+    {
+      "epoch": 0.9681868178356883,
+      "grad_norm": 0.1829284280538559,
+      "learning_rate": 0.001,
+      "loss": 1.826,
+      "step": 22886
+    },
+    {
+      "epoch": 0.9682291225992047,
+      "grad_norm": 0.16938935220241547,
+      "learning_rate": 0.001,
+      "loss": 2.0336,
+      "step": 22887
+    },
+    {
+      "epoch": 0.9682714273627211,
+      "grad_norm": 0.1795877367258072,
+      "learning_rate": 0.001,
+      "loss": 2.4894,
+      "step": 22888
+    },
+    {
+      "epoch": 0.9683137321262374,
+      "grad_norm": 0.16748693585395813,
+      "learning_rate": 0.001,
+      "loss": 2.4214,
+      "step": 22889
+    },
+    {
+      "epoch": 0.9683560368897538,
+      "grad_norm": 0.15759611129760742,
+      "learning_rate": 0.001,
+      "loss": 1.9249,
+      "step": 22890
+    },
+    {
+      "epoch": 0.9683983416532702,
+      "grad_norm": 0.16091367602348328,
+      "learning_rate": 0.001,
+      "loss": 1.8189,
+      "step": 22891
+    },
+    {
+      "epoch": 0.9684406464167865,
+      "grad_norm": 0.1906583309173584,
+      "learning_rate": 0.001,
+      "loss": 2.3985,
+      "step": 22892
+    },
+    {
+      "epoch": 0.9684829511803029,
+      "grad_norm": 0.15400663018226624,
+      "learning_rate": 0.001,
+      "loss": 2.1882,
+      "step": 22893
+    },
+    {
+      "epoch": 0.9685252559438192,
+      "grad_norm": 0.1885334551334381,
+      "learning_rate": 0.001,
+      "loss": 1.759,
+      "step": 22894
+    },
+    {
+      "epoch": 0.9685675607073356,
+      "grad_norm": 0.2451082170009613,
+      "learning_rate": 0.001,
+      "loss": 2.4656,
+      "step": 22895
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.25269943475723267,
+      "learning_rate": 0.001,
+      "loss": 2.1214,
+      "step": 22896
+    },
+    {
+      "epoch": 0.9686521702343683,
+      "grad_norm": 0.21075433492660522,
+      "learning_rate": 0.001,
+      "loss": 1.8323,
+      "step": 22897
+    },
+    {
+      "epoch": 0.9686944749978847,
+      "grad_norm": 0.168150395154953,
+      "learning_rate": 0.001,
+      "loss": 1.8325,
+      "step": 22898
+    },
+    {
+      "epoch": 0.9687367797614012,
+      "grad_norm": 0.1418176144361496,
+      "learning_rate": 0.001,
+      "loss": 2.0136,
+      "step": 22899
+    },
+    {
+      "epoch": 0.9687790845249175,
+      "grad_norm": 0.21564440429210663,
+      "learning_rate": 0.001,
+      "loss": 2.308,
+      "step": 22900
+    },
+    {
+      "epoch": 0.9688213892884339,
+      "grad_norm": 0.18524868786334991,
+      "learning_rate": 0.001,
+      "loss": 1.8404,
+      "step": 22901
+    },
+    {
+      "epoch": 0.9688636940519503,
+      "grad_norm": 0.13573579490184784,
+      "learning_rate": 0.001,
+      "loss": 1.5914,
+      "step": 22902
+    },
+    {
+      "epoch": 0.9689059988154666,
+      "grad_norm": 0.7693604826927185,
+      "learning_rate": 0.001,
+      "loss": 1.7203,
+      "step": 22903
+    },
+    {
+      "epoch": 0.968948303578983,
+      "grad_norm": 0.2601660490036011,
+      "learning_rate": 0.001,
+      "loss": 2.7439,
+      "step": 22904
+    },
+    {
+      "epoch": 0.9689906083424994,
+      "grad_norm": 0.16246432065963745,
+      "learning_rate": 0.001,
+      "loss": 2.6348,
+      "step": 22905
+    },
+    {
+      "epoch": 0.9690329131060157,
+      "grad_norm": 0.2047150582075119,
+      "learning_rate": 0.001,
+      "loss": 2.1811,
+      "step": 22906
+    },
+    {
+      "epoch": 0.9690752178695321,
+      "grad_norm": 0.19627639651298523,
+      "learning_rate": 0.001,
+      "loss": 1.8717,
+      "step": 22907
+    },
+    {
+      "epoch": 0.9691175226330485,
+      "grad_norm": 0.2079043984413147,
+      "learning_rate": 0.001,
+      "loss": 3.3459,
+      "step": 22908
+    },
+    {
+      "epoch": 0.9691598273965648,
+      "grad_norm": 0.16984863579273224,
+      "learning_rate": 0.001,
+      "loss": 1.721,
+      "step": 22909
+    },
+    {
+      "epoch": 0.9692021321600812,
+      "grad_norm": 0.23695328831672668,
+      "learning_rate": 0.001,
+      "loss": 1.92,
+      "step": 22910
+    },
+    {
+      "epoch": 0.9692444369235976,
+      "grad_norm": 0.13472682237625122,
+      "learning_rate": 0.001,
+      "loss": 1.5671,
+      "step": 22911
+    },
+    {
+      "epoch": 0.9692867416871139,
+      "grad_norm": 0.1398351788520813,
+      "learning_rate": 0.001,
+      "loss": 2.6105,
+      "step": 22912
+    },
+    {
+      "epoch": 0.9693290464506303,
+      "grad_norm": 0.16584454476833344,
+      "learning_rate": 0.001,
+      "loss": 2.4794,
+      "step": 22913
+    },
+    {
+      "epoch": 0.9693713512141467,
+      "grad_norm": 0.20908065140247345,
+      "learning_rate": 0.001,
+      "loss": 4.0282,
+      "step": 22914
+    },
+    {
+      "epoch": 0.969413655977663,
+      "grad_norm": 0.14626561105251312,
+      "learning_rate": 0.001,
+      "loss": 3.0325,
+      "step": 22915
+    },
+    {
+      "epoch": 0.9694559607411795,
+      "grad_norm": 0.190009206533432,
+      "learning_rate": 0.001,
+      "loss": 1.9914,
+      "step": 22916
+    },
+    {
+      "epoch": 0.9694982655046959,
+      "grad_norm": 0.30894187092781067,
+      "learning_rate": 0.001,
+      "loss": 1.9338,
+      "step": 22917
+    },
+    {
+      "epoch": 0.9695405702682122,
+      "grad_norm": 0.1590581089258194,
+      "learning_rate": 0.001,
+      "loss": 2.0382,
+      "step": 22918
+    },
+    {
+      "epoch": 0.9695828750317286,
+      "grad_norm": 0.1639058142900467,
+      "learning_rate": 0.001,
+      "loss": 1.5866,
+      "step": 22919
+    },
+    {
+      "epoch": 0.969625179795245,
+      "grad_norm": 0.1441287249326706,
+      "learning_rate": 0.001,
+      "loss": 1.7956,
+      "step": 22920
+    },
+    {
+      "epoch": 0.9696674845587613,
+      "grad_norm": 0.14907407760620117,
+      "learning_rate": 0.001,
+      "loss": 2.4144,
+      "step": 22921
+    },
+    {
+      "epoch": 0.9697097893222777,
+      "grad_norm": 0.17159616947174072,
+      "learning_rate": 0.001,
+      "loss": 1.4419,
+      "step": 22922
+    },
+    {
+      "epoch": 0.9697520940857941,
+      "grad_norm": 0.1924789398908615,
+      "learning_rate": 0.001,
+      "loss": 2.0441,
+      "step": 22923
+    },
+    {
+      "epoch": 0.9697943988493104,
+      "grad_norm": 0.8495581150054932,
+      "learning_rate": 0.001,
+      "loss": 2.2902,
+      "step": 22924
+    },
+    {
+      "epoch": 0.9698367036128268,
+      "grad_norm": 0.19208434224128723,
+      "learning_rate": 0.001,
+      "loss": 3.4082,
+      "step": 22925
+    },
+    {
+      "epoch": 0.9698790083763432,
+      "grad_norm": 0.16336600482463837,
+      "learning_rate": 0.001,
+      "loss": 1.8806,
+      "step": 22926
+    },
+    {
+      "epoch": 0.9699213131398595,
+      "grad_norm": 0.3081340491771698,
+      "learning_rate": 0.001,
+      "loss": 2.5719,
+      "step": 22927
+    },
+    {
+      "epoch": 0.9699636179033759,
+      "grad_norm": 0.743816614151001,
+      "learning_rate": 0.001,
+      "loss": 1.715,
+      "step": 22928
+    },
+    {
+      "epoch": 0.9700059226668923,
+      "grad_norm": 0.14184142649173737,
+      "learning_rate": 0.001,
+      "loss": 2.3673,
+      "step": 22929
+    },
+    {
+      "epoch": 0.9700482274304086,
+      "grad_norm": 0.14244288206100464,
+      "learning_rate": 0.001,
+      "loss": 2.542,
+      "step": 22930
+    },
+    {
+      "epoch": 0.970090532193925,
+      "grad_norm": 0.15329469740390778,
+      "learning_rate": 0.001,
+      "loss": 2.1073,
+      "step": 22931
+    },
+    {
+      "epoch": 0.9701328369574415,
+      "grad_norm": 7.721384048461914,
+      "learning_rate": 0.001,
+      "loss": 2.1494,
+      "step": 22932
+    },
+    {
+      "epoch": 0.9701751417209578,
+      "grad_norm": 0.22561503946781158,
+      "learning_rate": 0.001,
+      "loss": 1.9826,
+      "step": 22933
+    },
+    {
+      "epoch": 0.9702174464844742,
+      "grad_norm": 0.3391435444355011,
+      "learning_rate": 0.001,
+      "loss": 1.8618,
+      "step": 22934
+    },
+    {
+      "epoch": 0.9702597512479906,
+      "grad_norm": 0.17945805191993713,
+      "learning_rate": 0.001,
+      "loss": 1.4256,
+      "step": 22935
+    },
+    {
+      "epoch": 0.9703020560115069,
+      "grad_norm": 0.1751527637243271,
+      "learning_rate": 0.001,
+      "loss": 2.4109,
+      "step": 22936
+    },
+    {
+      "epoch": 0.9703443607750233,
+      "grad_norm": 0.15709732472896576,
+      "learning_rate": 0.001,
+      "loss": 1.6146,
+      "step": 22937
+    },
+    {
+      "epoch": 0.9703866655385397,
+      "grad_norm": 0.16432954370975494,
+      "learning_rate": 0.001,
+      "loss": 2.0583,
+      "step": 22938
+    },
+    {
+      "epoch": 0.970428970302056,
+      "grad_norm": 0.44643503427505493,
+      "learning_rate": 0.001,
+      "loss": 2.3525,
+      "step": 22939
+    },
+    {
+      "epoch": 0.9704712750655724,
+      "grad_norm": 0.1962294727563858,
+      "learning_rate": 0.001,
+      "loss": 1.874,
+      "step": 22940
+    },
+    {
+      "epoch": 0.9705135798290887,
+      "grad_norm": 0.1391528844833374,
+      "learning_rate": 0.001,
+      "loss": 2.7016,
+      "step": 22941
+    },
+    {
+      "epoch": 0.9705558845926051,
+      "grad_norm": 0.15284281969070435,
+      "learning_rate": 0.001,
+      "loss": 2.5978,
+      "step": 22942
+    },
+    {
+      "epoch": 0.9705981893561215,
+      "grad_norm": 0.19085395336151123,
+      "learning_rate": 0.001,
+      "loss": 1.2241,
+      "step": 22943
+    },
+    {
+      "epoch": 0.9706404941196378,
+      "grad_norm": 0.17279383540153503,
+      "learning_rate": 0.001,
+      "loss": 2.0888,
+      "step": 22944
+    },
+    {
+      "epoch": 0.9706827988831542,
+      "grad_norm": 0.15863092243671417,
+      "learning_rate": 0.001,
+      "loss": 1.6636,
+      "step": 22945
+    },
+    {
+      "epoch": 0.9707251036466706,
+      "grad_norm": 0.1498878002166748,
+      "learning_rate": 0.001,
+      "loss": 1.7367,
+      "step": 22946
+    },
+    {
+      "epoch": 0.9707674084101869,
+      "grad_norm": 0.14784745872020721,
+      "learning_rate": 0.001,
+      "loss": 1.7366,
+      "step": 22947
+    },
+    {
+      "epoch": 0.9708097131737033,
+      "grad_norm": 0.14569970965385437,
+      "learning_rate": 0.001,
+      "loss": 1.6962,
+      "step": 22948
+    },
+    {
+      "epoch": 0.9708520179372198,
+      "grad_norm": 0.13814716041088104,
+      "learning_rate": 0.001,
+      "loss": 1.8074,
+      "step": 22949
+    },
+    {
+      "epoch": 0.9708943227007361,
+      "grad_norm": 0.18621012568473816,
+      "learning_rate": 0.001,
+      "loss": 2.2638,
+      "step": 22950
+    },
+    {
+      "epoch": 0.9709366274642525,
+      "grad_norm": 0.1415053755044937,
+      "learning_rate": 0.001,
+      "loss": 1.6322,
+      "step": 22951
+    },
+    {
+      "epoch": 0.9709789322277689,
+      "grad_norm": 0.1956765204668045,
+      "learning_rate": 0.001,
+      "loss": 2.0337,
+      "step": 22952
+    },
+    {
+      "epoch": 0.9710212369912852,
+      "grad_norm": 0.16318294405937195,
+      "learning_rate": 0.001,
+      "loss": 2.2628,
+      "step": 22953
+    },
+    {
+      "epoch": 0.9710635417548016,
+      "grad_norm": 6.47411584854126,
+      "learning_rate": 0.001,
+      "loss": 2.0023,
+      "step": 22954
+    },
+    {
+      "epoch": 0.971105846518318,
+      "grad_norm": 0.14443252980709076,
+      "learning_rate": 0.001,
+      "loss": 2.8807,
+      "step": 22955
+    },
+    {
+      "epoch": 0.9711481512818343,
+      "grad_norm": 0.12970362603664398,
+      "learning_rate": 0.001,
+      "loss": 3.2348,
+      "step": 22956
+    },
+    {
+      "epoch": 0.9711904560453507,
+      "grad_norm": 0.21394479274749756,
+      "learning_rate": 0.001,
+      "loss": 1.8296,
+      "step": 22957
+    },
+    {
+      "epoch": 0.9712327608088671,
+      "grad_norm": 0.15435615181922913,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 22958
+    },
+    {
+      "epoch": 0.9712750655723834,
+      "grad_norm": 0.21590697765350342,
+      "learning_rate": 0.001,
+      "loss": 2.2599,
+      "step": 22959
+    },
+    {
+      "epoch": 0.9713173703358998,
+      "grad_norm": 0.14091353118419647,
+      "learning_rate": 0.001,
+      "loss": 2.0747,
+      "step": 22960
+    },
+    {
+      "epoch": 0.9713596750994162,
+      "grad_norm": 0.1489262878894806,
+      "learning_rate": 0.001,
+      "loss": 1.9124,
+      "step": 22961
+    },
+    {
+      "epoch": 0.9714019798629325,
+      "grad_norm": 0.14029927551746368,
+      "learning_rate": 0.001,
+      "loss": 2.5653,
+      "step": 22962
+    },
+    {
+      "epoch": 0.9714442846264489,
+      "grad_norm": 0.1342879831790924,
+      "learning_rate": 0.001,
+      "loss": 1.4259,
+      "step": 22963
+    },
+    {
+      "epoch": 0.9714865893899653,
+      "grad_norm": 0.13123171031475067,
+      "learning_rate": 0.001,
+      "loss": 2.2616,
+      "step": 22964
+    },
+    {
+      "epoch": 0.9715288941534816,
+      "grad_norm": 0.13907508552074432,
+      "learning_rate": 0.001,
+      "loss": 1.5811,
+      "step": 22965
+    },
+    {
+      "epoch": 0.9715711989169981,
+      "grad_norm": 0.14411531388759613,
+      "learning_rate": 0.001,
+      "loss": 2.1203,
+      "step": 22966
+    },
+    {
+      "epoch": 0.9716135036805145,
+      "grad_norm": 0.15099336206912994,
+      "learning_rate": 0.001,
+      "loss": 2.7498,
+      "step": 22967
+    },
+    {
+      "epoch": 0.9716558084440308,
+      "grad_norm": 0.14100094139575958,
+      "learning_rate": 0.001,
+      "loss": 1.9042,
+      "step": 22968
+    },
+    {
+      "epoch": 0.9716981132075472,
+      "grad_norm": 0.16299395263195038,
+      "learning_rate": 0.001,
+      "loss": 3.2449,
+      "step": 22969
+    },
+    {
+      "epoch": 0.9717404179710636,
+      "grad_norm": 0.1855599284172058,
+      "learning_rate": 0.001,
+      "loss": 1.2989,
+      "step": 22970
+    },
+    {
+      "epoch": 0.9717827227345799,
+      "grad_norm": 0.13893848657608032,
+      "learning_rate": 0.001,
+      "loss": 1.8379,
+      "step": 22971
+    },
+    {
+      "epoch": 0.9718250274980963,
+      "grad_norm": 0.22873970866203308,
+      "learning_rate": 0.001,
+      "loss": 2.7587,
+      "step": 22972
+    },
+    {
+      "epoch": 0.9718673322616127,
+      "grad_norm": 0.21288566291332245,
+      "learning_rate": 0.001,
+      "loss": 1.7145,
+      "step": 22973
+    },
+    {
+      "epoch": 0.971909637025129,
+      "grad_norm": 0.16666394472122192,
+      "learning_rate": 0.001,
+      "loss": 2.0038,
+      "step": 22974
+    },
+    {
+      "epoch": 0.9719519417886454,
+      "grad_norm": 0.1581132858991623,
+      "learning_rate": 0.001,
+      "loss": 1.5963,
+      "step": 22975
+    },
+    {
+      "epoch": 0.9719942465521618,
+      "grad_norm": 0.1787661761045456,
+      "learning_rate": 0.001,
+      "loss": 2.5239,
+      "step": 22976
+    },
+    {
+      "epoch": 0.9720365513156781,
+      "grad_norm": 0.1651124209165573,
+      "learning_rate": 0.001,
+      "loss": 2.6818,
+      "step": 22977
+    },
+    {
+      "epoch": 0.9720788560791945,
+      "grad_norm": 0.14961262047290802,
+      "learning_rate": 0.001,
+      "loss": 2.9672,
+      "step": 22978
+    },
+    {
+      "epoch": 0.9721211608427109,
+      "grad_norm": 0.15636776387691498,
+      "learning_rate": 0.001,
+      "loss": 2.4673,
+      "step": 22979
+    },
+    {
+      "epoch": 0.9721634656062272,
+      "grad_norm": 1.0559109449386597,
+      "learning_rate": 0.001,
+      "loss": 2.7908,
+      "step": 22980
+    },
+    {
+      "epoch": 0.9722057703697436,
+      "grad_norm": 0.15720365941524506,
+      "learning_rate": 0.001,
+      "loss": 2.1491,
+      "step": 22981
+    },
+    {
+      "epoch": 0.9722480751332601,
+      "grad_norm": 0.14667530357837677,
+      "learning_rate": 0.001,
+      "loss": 1.7503,
+      "step": 22982
+    },
+    {
+      "epoch": 0.9722903798967764,
+      "grad_norm": 0.15242712199687958,
+      "learning_rate": 0.001,
+      "loss": 2.048,
+      "step": 22983
+    },
+    {
+      "epoch": 0.9723326846602928,
+      "grad_norm": 7.654297828674316,
+      "learning_rate": 0.001,
+      "loss": 1.64,
+      "step": 22984
+    },
+    {
+      "epoch": 0.9723749894238091,
+      "grad_norm": 3.1997509002685547,
+      "learning_rate": 0.001,
+      "loss": 1.8606,
+      "step": 22985
+    },
+    {
+      "epoch": 0.9724172941873255,
+      "grad_norm": 0.3333907723426819,
+      "learning_rate": 0.001,
+      "loss": 1.8566,
+      "step": 22986
+    },
+    {
+      "epoch": 0.9724595989508419,
+      "grad_norm": 0.2523484528064728,
+      "learning_rate": 0.001,
+      "loss": 2.3548,
+      "step": 22987
+    },
+    {
+      "epoch": 0.9725019037143582,
+      "grad_norm": 0.1559314876794815,
+      "learning_rate": 0.001,
+      "loss": 1.3173,
+      "step": 22988
+    },
+    {
+      "epoch": 0.9725442084778746,
+      "grad_norm": 0.13869056105613708,
+      "learning_rate": 0.001,
+      "loss": 2.4087,
+      "step": 22989
+    },
+    {
+      "epoch": 0.972586513241391,
+      "grad_norm": 0.13740174472332,
+      "learning_rate": 0.001,
+      "loss": 1.849,
+      "step": 22990
+    },
+    {
+      "epoch": 0.9726288180049073,
+      "grad_norm": 0.1633240431547165,
+      "learning_rate": 0.001,
+      "loss": 2.104,
+      "step": 22991
+    },
+    {
+      "epoch": 0.9726711227684237,
+      "grad_norm": 0.14043399691581726,
+      "learning_rate": 0.001,
+      "loss": 2.5682,
+      "step": 22992
+    },
+    {
+      "epoch": 0.9727134275319401,
+      "grad_norm": 0.16599078476428986,
+      "learning_rate": 0.001,
+      "loss": 3.1016,
+      "step": 22993
+    },
+    {
+      "epoch": 0.9727557322954564,
+      "grad_norm": 0.6337260007858276,
+      "learning_rate": 0.001,
+      "loss": 2.6611,
+      "step": 22994
+    },
+    {
+      "epoch": 0.9727980370589728,
+      "grad_norm": 0.21094225347042084,
+      "learning_rate": 0.001,
+      "loss": 3.6387,
+      "step": 22995
+    },
+    {
+      "epoch": 0.9728403418224892,
+      "grad_norm": 0.3254936635494232,
+      "learning_rate": 0.001,
+      "loss": 3.3134,
+      "step": 22996
+    },
+    {
+      "epoch": 0.9728826465860055,
+      "grad_norm": 0.16493062674999237,
+      "learning_rate": 0.001,
+      "loss": 2.4989,
+      "step": 22997
+    },
+    {
+      "epoch": 0.972924951349522,
+      "grad_norm": 0.18628916144371033,
+      "learning_rate": 0.001,
+      "loss": 2.0522,
+      "step": 22998
+    },
+    {
+      "epoch": 0.9729672561130384,
+      "grad_norm": 0.20497991144657135,
+      "learning_rate": 0.001,
+      "loss": 2.1398,
+      "step": 22999
+    },
+    {
+      "epoch": 0.9730095608765547,
+      "grad_norm": 0.13439905643463135,
+      "learning_rate": 0.001,
+      "loss": 1.7492,
+      "step": 23000
+    },
+    {
+      "epoch": 0.9730518656400711,
+      "grad_norm": 0.20762348175048828,
+      "learning_rate": 0.001,
+      "loss": 2.6112,
+      "step": 23001
+    },
+    {
+      "epoch": 0.9730941704035875,
+      "grad_norm": 0.17342343926429749,
+      "learning_rate": 0.001,
+      "loss": 1.9358,
+      "step": 23002
+    },
+    {
+      "epoch": 0.9731364751671038,
+      "grad_norm": 0.15936243534088135,
+      "learning_rate": 0.001,
+      "loss": 2.0159,
+      "step": 23003
+    },
+    {
+      "epoch": 0.9731787799306202,
+      "grad_norm": 0.15786008536815643,
+      "learning_rate": 0.001,
+      "loss": 1.3607,
+      "step": 23004
+    },
+    {
+      "epoch": 0.9732210846941366,
+      "grad_norm": 0.18366417288780212,
+      "learning_rate": 0.001,
+      "loss": 1.746,
+      "step": 23005
+    },
+    {
+      "epoch": 0.9732633894576529,
+      "grad_norm": 0.2870394289493561,
+      "learning_rate": 0.001,
+      "loss": 1.9022,
+      "step": 23006
+    },
+    {
+      "epoch": 0.9733056942211693,
+      "grad_norm": 0.18606680631637573,
+      "learning_rate": 0.001,
+      "loss": 2.4152,
+      "step": 23007
+    },
+    {
+      "epoch": 0.9733479989846857,
+      "grad_norm": 0.15611374378204346,
+      "learning_rate": 0.001,
+      "loss": 2.4503,
+      "step": 23008
+    },
+    {
+      "epoch": 0.973390303748202,
+      "grad_norm": 0.1782141774892807,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 23009
+    },
+    {
+      "epoch": 0.9734326085117184,
+      "grad_norm": 0.817843496799469,
+      "learning_rate": 0.001,
+      "loss": 2.758,
+      "step": 23010
+    },
+    {
+      "epoch": 0.9734749132752348,
+      "grad_norm": 0.13205979764461517,
+      "learning_rate": 0.001,
+      "loss": 1.4987,
+      "step": 23011
+    },
+    {
+      "epoch": 0.9735172180387511,
+      "grad_norm": 3.9600741863250732,
+      "learning_rate": 0.001,
+      "loss": 1.5233,
+      "step": 23012
+    },
+    {
+      "epoch": 0.9735595228022675,
+      "grad_norm": 0.17472542822360992,
+      "learning_rate": 0.001,
+      "loss": 1.5872,
+      "step": 23013
+    },
+    {
+      "epoch": 0.973601827565784,
+      "grad_norm": 0.17888934910297394,
+      "learning_rate": 0.001,
+      "loss": 1.989,
+      "step": 23014
+    },
+    {
+      "epoch": 0.9736441323293002,
+      "grad_norm": 0.162007674574852,
+      "learning_rate": 0.001,
+      "loss": 2.0518,
+      "step": 23015
+    },
+    {
+      "epoch": 0.9736864370928167,
+      "grad_norm": 0.17636260390281677,
+      "learning_rate": 0.001,
+      "loss": 1.9944,
+      "step": 23016
+    },
+    {
+      "epoch": 0.9737287418563331,
+      "grad_norm": 0.19297371804714203,
+      "learning_rate": 0.001,
+      "loss": 1.6305,
+      "step": 23017
+    },
+    {
+      "epoch": 0.9737710466198494,
+      "grad_norm": 0.16038502752780914,
+      "learning_rate": 0.001,
+      "loss": 2.8023,
+      "step": 23018
+    },
+    {
+      "epoch": 0.9738133513833658,
+      "grad_norm": 0.24749846756458282,
+      "learning_rate": 0.001,
+      "loss": 1.8999,
+      "step": 23019
+    },
+    {
+      "epoch": 0.9738556561468822,
+      "grad_norm": 0.16544784605503082,
+      "learning_rate": 0.001,
+      "loss": 1.9664,
+      "step": 23020
+    },
+    {
+      "epoch": 0.9738979609103985,
+      "grad_norm": 0.16566026210784912,
+      "learning_rate": 0.001,
+      "loss": 1.6773,
+      "step": 23021
+    },
+    {
+      "epoch": 0.9739402656739149,
+      "grad_norm": 0.16837327182292938,
+      "learning_rate": 0.001,
+      "loss": 2.1888,
+      "step": 23022
+    },
+    {
+      "epoch": 0.9739825704374313,
+      "grad_norm": 0.16060344874858856,
+      "learning_rate": 0.001,
+      "loss": 1.9338,
+      "step": 23023
+    },
+    {
+      "epoch": 0.9740248752009476,
+      "grad_norm": 0.21505653858184814,
+      "learning_rate": 0.001,
+      "loss": 2.1727,
+      "step": 23024
+    },
+    {
+      "epoch": 0.974067179964464,
+      "grad_norm": 0.17655415832996368,
+      "learning_rate": 0.001,
+      "loss": 1.7649,
+      "step": 23025
+    },
+    {
+      "epoch": 0.9741094847279804,
+      "grad_norm": 0.1724175214767456,
+      "learning_rate": 0.001,
+      "loss": 1.7799,
+      "step": 23026
+    },
+    {
+      "epoch": 0.9741517894914967,
+      "grad_norm": 0.21680817008018494,
+      "learning_rate": 0.001,
+      "loss": 1.9723,
+      "step": 23027
+    },
+    {
+      "epoch": 0.9741940942550131,
+      "grad_norm": 5.8497748374938965,
+      "learning_rate": 0.001,
+      "loss": 2.4025,
+      "step": 23028
+    },
+    {
+      "epoch": 0.9742363990185295,
+      "grad_norm": 0.18116092681884766,
+      "learning_rate": 0.001,
+      "loss": 2.083,
+      "step": 23029
+    },
+    {
+      "epoch": 0.9742787037820458,
+      "grad_norm": 0.1881796270608902,
+      "learning_rate": 0.001,
+      "loss": 1.8736,
+      "step": 23030
+    },
+    {
+      "epoch": 0.9743210085455623,
+      "grad_norm": 0.2181680053472519,
+      "learning_rate": 0.001,
+      "loss": 1.9318,
+      "step": 23031
+    },
+    {
+      "epoch": 0.9743633133090785,
+      "grad_norm": 0.2097787708044052,
+      "learning_rate": 0.001,
+      "loss": 1.819,
+      "step": 23032
+    },
+    {
+      "epoch": 0.974405618072595,
+      "grad_norm": 0.16672278940677643,
+      "learning_rate": 0.001,
+      "loss": 1.8597,
+      "step": 23033
+    },
+    {
+      "epoch": 0.9744479228361114,
+      "grad_norm": 0.1386634260416031,
+      "learning_rate": 0.001,
+      "loss": 1.5876,
+      "step": 23034
+    },
+    {
+      "epoch": 0.9744902275996277,
+      "grad_norm": 0.1755525916814804,
+      "learning_rate": 0.001,
+      "loss": 1.8865,
+      "step": 23035
+    },
+    {
+      "epoch": 0.9745325323631441,
+      "grad_norm": 0.17198018729686737,
+      "learning_rate": 0.001,
+      "loss": 2.1409,
+      "step": 23036
+    },
+    {
+      "epoch": 0.9745748371266605,
+      "grad_norm": 0.20147661864757538,
+      "learning_rate": 0.001,
+      "loss": 1.6044,
+      "step": 23037
+    },
+    {
+      "epoch": 0.9746171418901768,
+      "grad_norm": 0.292715460062027,
+      "learning_rate": 0.001,
+      "loss": 1.906,
+      "step": 23038
+    },
+    {
+      "epoch": 0.9746594466536932,
+      "grad_norm": 3.269289493560791,
+      "learning_rate": 0.001,
+      "loss": 2.3023,
+      "step": 23039
+    },
+    {
+      "epoch": 0.9747017514172096,
+      "grad_norm": 0.30846601724624634,
+      "learning_rate": 0.001,
+      "loss": 2.5841,
+      "step": 23040
+    },
+    {
+      "epoch": 0.9747440561807259,
+      "grad_norm": 0.2580687701702118,
+      "learning_rate": 0.001,
+      "loss": 2.0838,
+      "step": 23041
+    },
+    {
+      "epoch": 0.9747863609442423,
+      "grad_norm": 0.4336620271205902,
+      "learning_rate": 0.001,
+      "loss": 2.2656,
+      "step": 23042
+    },
+    {
+      "epoch": 0.9748286657077587,
+      "grad_norm": 0.9304342865943909,
+      "learning_rate": 0.001,
+      "loss": 2.4475,
+      "step": 23043
+    },
+    {
+      "epoch": 0.974870970471275,
+      "grad_norm": 5.938025951385498,
+      "learning_rate": 0.001,
+      "loss": 2.5208,
+      "step": 23044
+    },
+    {
+      "epoch": 0.9749132752347914,
+      "grad_norm": 24.187389373779297,
+      "learning_rate": 0.001,
+      "loss": 1.5931,
+      "step": 23045
+    },
+    {
+      "epoch": 0.9749555799983078,
+      "grad_norm": 0.4992624819278717,
+      "learning_rate": 0.001,
+      "loss": 2.4145,
+      "step": 23046
+    },
+    {
+      "epoch": 0.9749978847618241,
+      "grad_norm": 2.016716718673706,
+      "learning_rate": 0.001,
+      "loss": 2.8388,
+      "step": 23047
+    },
+    {
+      "epoch": 0.9750401895253406,
+      "grad_norm": 1.3812201023101807,
+      "learning_rate": 0.001,
+      "loss": 2.0137,
+      "step": 23048
+    },
+    {
+      "epoch": 0.975082494288857,
+      "grad_norm": 0.2711379826068878,
+      "learning_rate": 0.001,
+      "loss": 1.798,
+      "step": 23049
+    },
+    {
+      "epoch": 0.9751247990523733,
+      "grad_norm": 0.3764013350009918,
+      "learning_rate": 0.001,
+      "loss": 2.0769,
+      "step": 23050
+    },
+    {
+      "epoch": 0.9751671038158897,
+      "grad_norm": 2.7994561195373535,
+      "learning_rate": 0.001,
+      "loss": 1.6256,
+      "step": 23051
+    },
+    {
+      "epoch": 0.9752094085794061,
+      "grad_norm": 0.26443973183631897,
+      "learning_rate": 0.001,
+      "loss": 2.8268,
+      "step": 23052
+    },
+    {
+      "epoch": 0.9752517133429224,
+      "grad_norm": 0.17250576615333557,
+      "learning_rate": 0.001,
+      "loss": 2.3258,
+      "step": 23053
+    },
+    {
+      "epoch": 0.9752940181064388,
+      "grad_norm": 0.23773407936096191,
+      "learning_rate": 0.001,
+      "loss": 2.3701,
+      "step": 23054
+    },
+    {
+      "epoch": 0.9753363228699552,
+      "grad_norm": 0.17003774642944336,
+      "learning_rate": 0.001,
+      "loss": 1.8557,
+      "step": 23055
+    },
+    {
+      "epoch": 0.9753786276334715,
+      "grad_norm": 0.6521497368812561,
+      "learning_rate": 0.001,
+      "loss": 2.9078,
+      "step": 23056
+    },
+    {
+      "epoch": 0.9754209323969879,
+      "grad_norm": 0.1700069010257721,
+      "learning_rate": 0.001,
+      "loss": 2.3583,
+      "step": 23057
+    },
+    {
+      "epoch": 0.9754632371605043,
+      "grad_norm": 1.8804014921188354,
+      "learning_rate": 0.001,
+      "loss": 1.5116,
+      "step": 23058
+    },
+    {
+      "epoch": 0.9755055419240206,
+      "grad_norm": 0.16962087154388428,
+      "learning_rate": 0.001,
+      "loss": 2.3743,
+      "step": 23059
+    },
+    {
+      "epoch": 0.975547846687537,
+      "grad_norm": 0.1760241985321045,
+      "learning_rate": 0.001,
+      "loss": 1.7771,
+      "step": 23060
+    },
+    {
+      "epoch": 0.9755901514510534,
+      "grad_norm": 0.3765549063682556,
+      "learning_rate": 0.001,
+      "loss": 2.3506,
+      "step": 23061
+    },
+    {
+      "epoch": 0.9756324562145697,
+      "grad_norm": 0.26375001668930054,
+      "learning_rate": 0.001,
+      "loss": 2.108,
+      "step": 23062
+    },
+    {
+      "epoch": 0.9756747609780861,
+      "grad_norm": 0.17104345560073853,
+      "learning_rate": 0.001,
+      "loss": 1.2225,
+      "step": 23063
+    },
+    {
+      "epoch": 0.9757170657416026,
+      "grad_norm": 0.15106819570064545,
+      "learning_rate": 0.001,
+      "loss": 2.0871,
+      "step": 23064
+    },
+    {
+      "epoch": 0.9757593705051189,
+      "grad_norm": 0.16280482709407806,
+      "learning_rate": 0.001,
+      "loss": 1.6354,
+      "step": 23065
+    },
+    {
+      "epoch": 0.9758016752686353,
+      "grad_norm": 0.19860334694385529,
+      "learning_rate": 0.001,
+      "loss": 1.9093,
+      "step": 23066
+    },
+    {
+      "epoch": 0.9758439800321517,
+      "grad_norm": 0.14197562634944916,
+      "learning_rate": 0.001,
+      "loss": 1.5894,
+      "step": 23067
+    },
+    {
+      "epoch": 0.975886284795668,
+      "grad_norm": 0.16649185121059418,
+      "learning_rate": 0.001,
+      "loss": 2.0655,
+      "step": 23068
+    },
+    {
+      "epoch": 0.9759285895591844,
+      "grad_norm": 0.15236613154411316,
+      "learning_rate": 0.001,
+      "loss": 1.4908,
+      "step": 23069
+    },
+    {
+      "epoch": 0.9759708943227008,
+      "grad_norm": 0.7669524550437927,
+      "learning_rate": 0.001,
+      "loss": 1.9742,
+      "step": 23070
+    },
+    {
+      "epoch": 0.9760131990862171,
+      "grad_norm": 0.2197335958480835,
+      "learning_rate": 0.001,
+      "loss": 1.8815,
+      "step": 23071
+    },
+    {
+      "epoch": 0.9760555038497335,
+      "grad_norm": 0.18375352025032043,
+      "learning_rate": 0.001,
+      "loss": 2.4837,
+      "step": 23072
+    },
+    {
+      "epoch": 0.9760978086132499,
+      "grad_norm": 0.31479236483573914,
+      "learning_rate": 0.001,
+      "loss": 1.4933,
+      "step": 23073
+    },
+    {
+      "epoch": 0.9761401133767662,
+      "grad_norm": 6.906200408935547,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 23074
+    },
+    {
+      "epoch": 0.9761824181402826,
+      "grad_norm": 0.17486466467380524,
+      "learning_rate": 0.001,
+      "loss": 3.0518,
+      "step": 23075
+    },
+    {
+      "epoch": 0.9762247229037989,
+      "grad_norm": 0.16075573861598969,
+      "learning_rate": 0.001,
+      "loss": 2.0813,
+      "step": 23076
+    },
+    {
+      "epoch": 0.9762670276673153,
+      "grad_norm": 0.16066281497478485,
+      "learning_rate": 0.001,
+      "loss": 3.0841,
+      "step": 23077
+    },
+    {
+      "epoch": 0.9763093324308317,
+      "grad_norm": 0.20229189097881317,
+      "learning_rate": 0.001,
+      "loss": 2.0031,
+      "step": 23078
+    },
+    {
+      "epoch": 0.976351637194348,
+      "grad_norm": 0.1783391684293747,
+      "learning_rate": 0.001,
+      "loss": 2.0519,
+      "step": 23079
+    },
+    {
+      "epoch": 0.9763939419578644,
+      "grad_norm": 0.16642500460147858,
+      "learning_rate": 0.001,
+      "loss": 1.9304,
+      "step": 23080
+    },
+    {
+      "epoch": 0.9764362467213809,
+      "grad_norm": 1.8268184661865234,
+      "learning_rate": 0.001,
+      "loss": 1.4835,
+      "step": 23081
+    },
+    {
+      "epoch": 0.9764785514848972,
+      "grad_norm": 0.14119704067707062,
+      "learning_rate": 0.001,
+      "loss": 2.4155,
+      "step": 23082
+    },
+    {
+      "epoch": 0.9765208562484136,
+      "grad_norm": 0.3065393567085266,
+      "learning_rate": 0.001,
+      "loss": 3.0736,
+      "step": 23083
+    },
+    {
+      "epoch": 0.97656316101193,
+      "grad_norm": 1.7388895750045776,
+      "learning_rate": 0.001,
+      "loss": 1.6748,
+      "step": 23084
+    },
+    {
+      "epoch": 0.9766054657754463,
+      "grad_norm": 0.18974030017852783,
+      "learning_rate": 0.001,
+      "loss": 2.2323,
+      "step": 23085
+    },
+    {
+      "epoch": 0.9766477705389627,
+      "grad_norm": 0.20437590777873993,
+      "learning_rate": 0.001,
+      "loss": 2.0711,
+      "step": 23086
+    },
+    {
+      "epoch": 0.9766900753024791,
+      "grad_norm": 0.144393190741539,
+      "learning_rate": 0.001,
+      "loss": 1.9763,
+      "step": 23087
+    },
+    {
+      "epoch": 0.9767323800659954,
+      "grad_norm": 0.1816982924938202,
+      "learning_rate": 0.001,
+      "loss": 2.3921,
+      "step": 23088
+    },
+    {
+      "epoch": 0.9767746848295118,
+      "grad_norm": 0.24456720054149628,
+      "learning_rate": 0.001,
+      "loss": 2.6854,
+      "step": 23089
+    },
+    {
+      "epoch": 0.9768169895930282,
+      "grad_norm": 0.17447160184383392,
+      "learning_rate": 0.001,
+      "loss": 1.757,
+      "step": 23090
+    },
+    {
+      "epoch": 0.9768592943565445,
+      "grad_norm": 25.030933380126953,
+      "learning_rate": 0.001,
+      "loss": 2.1183,
+      "step": 23091
+    },
+    {
+      "epoch": 0.9769015991200609,
+      "grad_norm": 0.20335282385349274,
+      "learning_rate": 0.001,
+      "loss": 2.738,
+      "step": 23092
+    },
+    {
+      "epoch": 0.9769439038835773,
+      "grad_norm": 0.4485716223716736,
+      "learning_rate": 0.001,
+      "loss": 1.7982,
+      "step": 23093
+    },
+    {
+      "epoch": 0.9769862086470936,
+      "grad_norm": 0.2497912049293518,
+      "learning_rate": 0.001,
+      "loss": 2.0586,
+      "step": 23094
+    },
+    {
+      "epoch": 0.97702851341061,
+      "grad_norm": 0.19284145534038544,
+      "learning_rate": 0.001,
+      "loss": 1.5553,
+      "step": 23095
+    },
+    {
+      "epoch": 0.9770708181741264,
+      "grad_norm": 0.14969174563884735,
+      "learning_rate": 0.001,
+      "loss": 1.8045,
+      "step": 23096
+    },
+    {
+      "epoch": 0.9771131229376427,
+      "grad_norm": 0.18470831215381622,
+      "learning_rate": 0.001,
+      "loss": 2.3824,
+      "step": 23097
+    },
+    {
+      "epoch": 0.9771554277011592,
+      "grad_norm": 0.1780456155538559,
+      "learning_rate": 0.001,
+      "loss": 2.5985,
+      "step": 23098
+    },
+    {
+      "epoch": 0.9771977324646756,
+      "grad_norm": 0.18881769478321075,
+      "learning_rate": 0.001,
+      "loss": 2.6814,
+      "step": 23099
+    },
+    {
+      "epoch": 0.9772400372281919,
+      "grad_norm": 0.160416379570961,
+      "learning_rate": 0.001,
+      "loss": 2.7233,
+      "step": 23100
+    },
+    {
+      "epoch": 0.9772823419917083,
+      "grad_norm": 0.1704859733581543,
+      "learning_rate": 0.001,
+      "loss": 2.3356,
+      "step": 23101
+    },
+    {
+      "epoch": 0.9773246467552247,
+      "grad_norm": 0.15877144038677216,
+      "learning_rate": 0.001,
+      "loss": 1.8551,
+      "step": 23102
+    },
+    {
+      "epoch": 0.977366951518741,
+      "grad_norm": 0.15213461220264435,
+      "learning_rate": 0.001,
+      "loss": 2.2297,
+      "step": 23103
+    },
+    {
+      "epoch": 0.9774092562822574,
+      "grad_norm": 0.23088282346725464,
+      "learning_rate": 0.001,
+      "loss": 2.2761,
+      "step": 23104
+    },
+    {
+      "epoch": 0.9774515610457738,
+      "grad_norm": 3.229534864425659,
+      "learning_rate": 0.001,
+      "loss": 3.1849,
+      "step": 23105
+    },
+    {
+      "epoch": 0.9774938658092901,
+      "grad_norm": 0.14823846518993378,
+      "learning_rate": 0.001,
+      "loss": 2.04,
+      "step": 23106
+    },
+    {
+      "epoch": 0.9775361705728065,
+      "grad_norm": 0.13392066955566406,
+      "learning_rate": 0.001,
+      "loss": 2.1271,
+      "step": 23107
+    },
+    {
+      "epoch": 0.9775784753363229,
+      "grad_norm": 0.251636266708374,
+      "learning_rate": 0.001,
+      "loss": 2.1417,
+      "step": 23108
+    },
+    {
+      "epoch": 0.9776207800998392,
+      "grad_norm": 0.12352551519870758,
+      "learning_rate": 0.001,
+      "loss": 1.8046,
+      "step": 23109
+    },
+    {
+      "epoch": 0.9776630848633556,
+      "grad_norm": 0.44623270630836487,
+      "learning_rate": 0.001,
+      "loss": 3.5603,
+      "step": 23110
+    },
+    {
+      "epoch": 0.977705389626872,
+      "grad_norm": 0.46664145588874817,
+      "learning_rate": 0.001,
+      "loss": 1.9793,
+      "step": 23111
+    },
+    {
+      "epoch": 0.9777476943903883,
+      "grad_norm": 0.19913674890995026,
+      "learning_rate": 0.001,
+      "loss": 2.2404,
+      "step": 23112
+    },
+    {
+      "epoch": 0.9777899991539047,
+      "grad_norm": 0.14924490451812744,
+      "learning_rate": 0.001,
+      "loss": 1.7245,
+      "step": 23113
+    },
+    {
+      "epoch": 0.9778323039174212,
+      "grad_norm": 0.17213156819343567,
+      "learning_rate": 0.001,
+      "loss": 1.6626,
+      "step": 23114
+    },
+    {
+      "epoch": 0.9778746086809375,
+      "grad_norm": 0.16330088675022125,
+      "learning_rate": 0.001,
+      "loss": 1.6231,
+      "step": 23115
+    },
+    {
+      "epoch": 0.9779169134444539,
+      "grad_norm": 0.5413417816162109,
+      "learning_rate": 0.001,
+      "loss": 2.2929,
+      "step": 23116
+    },
+    {
+      "epoch": 0.9779592182079703,
+      "grad_norm": 0.8194732069969177,
+      "learning_rate": 0.001,
+      "loss": 2.0811,
+      "step": 23117
+    },
+    {
+      "epoch": 0.9780015229714866,
+      "grad_norm": 0.3619847297668457,
+      "learning_rate": 0.001,
+      "loss": 2.1488,
+      "step": 23118
+    },
+    {
+      "epoch": 0.978043827735003,
+      "grad_norm": 0.2249876707792282,
+      "learning_rate": 0.001,
+      "loss": 2.3759,
+      "step": 23119
+    },
+    {
+      "epoch": 0.9780861324985193,
+      "grad_norm": 2.6536123752593994,
+      "learning_rate": 0.001,
+      "loss": 2.256,
+      "step": 23120
+    },
+    {
+      "epoch": 0.9781284372620357,
+      "grad_norm": 0.1619919091463089,
+      "learning_rate": 0.001,
+      "loss": 2.3861,
+      "step": 23121
+    },
+    {
+      "epoch": 0.9781707420255521,
+      "grad_norm": 0.16015075147151947,
+      "learning_rate": 0.001,
+      "loss": 1.9889,
+      "step": 23122
+    },
+    {
+      "epoch": 0.9782130467890684,
+      "grad_norm": 0.14887668192386627,
+      "learning_rate": 0.001,
+      "loss": 2.1329,
+      "step": 23123
+    },
+    {
+      "epoch": 0.9782553515525848,
+      "grad_norm": 0.14117036759853363,
+      "learning_rate": 0.001,
+      "loss": 2.0059,
+      "step": 23124
+    },
+    {
+      "epoch": 0.9782976563161012,
+      "grad_norm": 0.8189803957939148,
+      "learning_rate": 0.001,
+      "loss": 2.0877,
+      "step": 23125
+    },
+    {
+      "epoch": 0.9783399610796175,
+      "grad_norm": 0.93526691198349,
+      "learning_rate": 0.001,
+      "loss": 1.3733,
+      "step": 23126
+    },
+    {
+      "epoch": 0.9783822658431339,
+      "grad_norm": 0.15379607677459717,
+      "learning_rate": 0.001,
+      "loss": 2.3471,
+      "step": 23127
+    },
+    {
+      "epoch": 0.9784245706066503,
+      "grad_norm": 0.1474200189113617,
+      "learning_rate": 0.001,
+      "loss": 2.0631,
+      "step": 23128
+    },
+    {
+      "epoch": 0.9784668753701666,
+      "grad_norm": 0.14249064028263092,
+      "learning_rate": 0.001,
+      "loss": 1.8189,
+      "step": 23129
+    },
+    {
+      "epoch": 0.978509180133683,
+      "grad_norm": 0.17934098839759827,
+      "learning_rate": 0.001,
+      "loss": 1.7359,
+      "step": 23130
+    },
+    {
+      "epoch": 0.9785514848971995,
+      "grad_norm": 0.16001644730567932,
+      "learning_rate": 0.001,
+      "loss": 1.8006,
+      "step": 23131
+    },
+    {
+      "epoch": 0.9785937896607158,
+      "grad_norm": 0.1634327471256256,
+      "learning_rate": 0.001,
+      "loss": 1.937,
+      "step": 23132
+    },
+    {
+      "epoch": 0.9786360944242322,
+      "grad_norm": 0.1568276584148407,
+      "learning_rate": 0.001,
+      "loss": 1.8826,
+      "step": 23133
+    },
+    {
+      "epoch": 0.9786783991877486,
+      "grad_norm": 0.17905938625335693,
+      "learning_rate": 0.001,
+      "loss": 1.6522,
+      "step": 23134
+    },
+    {
+      "epoch": 0.9787207039512649,
+      "grad_norm": 0.2984275817871094,
+      "learning_rate": 0.001,
+      "loss": 1.7585,
+      "step": 23135
+    },
+    {
+      "epoch": 0.9787630087147813,
+      "grad_norm": 0.16698578000068665,
+      "learning_rate": 0.001,
+      "loss": 2.5787,
+      "step": 23136
+    },
+    {
+      "epoch": 0.9788053134782977,
+      "grad_norm": 0.13776206970214844,
+      "learning_rate": 0.001,
+      "loss": 1.9259,
+      "step": 23137
+    },
+    {
+      "epoch": 0.978847618241814,
+      "grad_norm": 0.1826232522726059,
+      "learning_rate": 0.001,
+      "loss": 2.3701,
+      "step": 23138
+    },
+    {
+      "epoch": 0.9788899230053304,
+      "grad_norm": 0.14153793454170227,
+      "learning_rate": 0.001,
+      "loss": 1.9833,
+      "step": 23139
+    },
+    {
+      "epoch": 0.9789322277688468,
+      "grad_norm": 0.19493409991264343,
+      "learning_rate": 0.001,
+      "loss": 3.4682,
+      "step": 23140
+    },
+    {
+      "epoch": 0.9789745325323631,
+      "grad_norm": 0.13115043938159943,
+      "learning_rate": 0.001,
+      "loss": 1.2842,
+      "step": 23141
+    },
+    {
+      "epoch": 0.9790168372958795,
+      "grad_norm": 0.13824576139450073,
+      "learning_rate": 0.001,
+      "loss": 2.2034,
+      "step": 23142
+    },
+    {
+      "epoch": 0.9790591420593959,
+      "grad_norm": 0.17789553105831146,
+      "learning_rate": 0.001,
+      "loss": 3.4185,
+      "step": 23143
+    },
+    {
+      "epoch": 0.9791014468229122,
+      "grad_norm": 0.1795099377632141,
+      "learning_rate": 0.001,
+      "loss": 2.0876,
+      "step": 23144
+    },
+    {
+      "epoch": 0.9791437515864286,
+      "grad_norm": 0.15983949601650238,
+      "learning_rate": 0.001,
+      "loss": 1.6265,
+      "step": 23145
+    },
+    {
+      "epoch": 0.979186056349945,
+      "grad_norm": 0.1699821501970291,
+      "learning_rate": 0.001,
+      "loss": 2.6895,
+      "step": 23146
+    },
+    {
+      "epoch": 0.9792283611134613,
+      "grad_norm": 2.1327931880950928,
+      "learning_rate": 0.001,
+      "loss": 1.7483,
+      "step": 23147
+    },
+    {
+      "epoch": 0.9792706658769778,
+      "grad_norm": 0.14410525560379028,
+      "learning_rate": 0.001,
+      "loss": 2.2787,
+      "step": 23148
+    },
+    {
+      "epoch": 0.9793129706404942,
+      "grad_norm": 0.14737246930599213,
+      "learning_rate": 0.001,
+      "loss": 2.1674,
+      "step": 23149
+    },
+    {
+      "epoch": 0.9793552754040105,
+      "grad_norm": 0.17130975425243378,
+      "learning_rate": 0.001,
+      "loss": 2.8032,
+      "step": 23150
+    },
+    {
+      "epoch": 0.9793975801675269,
+      "grad_norm": 0.1790180206298828,
+      "learning_rate": 0.001,
+      "loss": 2.6401,
+      "step": 23151
+    },
+    {
+      "epoch": 0.9794398849310433,
+      "grad_norm": 0.1794183999300003,
+      "learning_rate": 0.001,
+      "loss": 1.896,
+      "step": 23152
+    },
+    {
+      "epoch": 0.9794821896945596,
+      "grad_norm": 0.12838570773601532,
+      "learning_rate": 0.001,
+      "loss": 1.8266,
+      "step": 23153
+    },
+    {
+      "epoch": 0.979524494458076,
+      "grad_norm": 0.18338772654533386,
+      "learning_rate": 0.001,
+      "loss": 1.7244,
+      "step": 23154
+    },
+    {
+      "epoch": 0.9795667992215924,
+      "grad_norm": 0.17173555493354797,
+      "learning_rate": 0.001,
+      "loss": 1.6205,
+      "step": 23155
+    },
+    {
+      "epoch": 0.9796091039851087,
+      "grad_norm": 0.18049339950084686,
+      "learning_rate": 0.001,
+      "loss": 2.2256,
+      "step": 23156
+    },
+    {
+      "epoch": 0.9796514087486251,
+      "grad_norm": 0.13144956529140472,
+      "learning_rate": 0.001,
+      "loss": 2.6915,
+      "step": 23157
+    },
+    {
+      "epoch": 0.9796937135121415,
+      "grad_norm": 0.1813870072364807,
+      "learning_rate": 0.001,
+      "loss": 1.9047,
+      "step": 23158
+    },
+    {
+      "epoch": 0.9797360182756578,
+      "grad_norm": 1.0048320293426514,
+      "learning_rate": 0.001,
+      "loss": 2.8726,
+      "step": 23159
+    },
+    {
+      "epoch": 0.9797783230391742,
+      "grad_norm": 0.1345098316669464,
+      "learning_rate": 0.001,
+      "loss": 1.4787,
+      "step": 23160
+    },
+    {
+      "epoch": 0.9798206278026906,
+      "grad_norm": 0.1486302614212036,
+      "learning_rate": 0.001,
+      "loss": 2.3496,
+      "step": 23161
+    },
+    {
+      "epoch": 0.9798629325662069,
+      "grad_norm": 0.198981374502182,
+      "learning_rate": 0.001,
+      "loss": 1.918,
+      "step": 23162
+    },
+    {
+      "epoch": 0.9799052373297233,
+      "grad_norm": 0.37297117710113525,
+      "learning_rate": 0.001,
+      "loss": 3.412,
+      "step": 23163
+    },
+    {
+      "epoch": 0.9799475420932398,
+      "grad_norm": 0.18499860167503357,
+      "learning_rate": 0.001,
+      "loss": 2.2248,
+      "step": 23164
+    },
+    {
+      "epoch": 0.979989846856756,
+      "grad_norm": 0.16248784959316254,
+      "learning_rate": 0.001,
+      "loss": 2.2262,
+      "step": 23165
+    },
+    {
+      "epoch": 0.9800321516202725,
+      "grad_norm": 0.13412149250507355,
+      "learning_rate": 0.001,
+      "loss": 1.9383,
+      "step": 23166
+    },
+    {
+      "epoch": 0.9800744563837888,
+      "grad_norm": 0.14490722119808197,
+      "learning_rate": 0.001,
+      "loss": 1.7336,
+      "step": 23167
+    },
+    {
+      "epoch": 0.9801167611473052,
+      "grad_norm": 0.16167306900024414,
+      "learning_rate": 0.001,
+      "loss": 3.7023,
+      "step": 23168
+    },
+    {
+      "epoch": 0.9801590659108216,
+      "grad_norm": 0.2081134021282196,
+      "learning_rate": 0.001,
+      "loss": 3.0097,
+      "step": 23169
+    },
+    {
+      "epoch": 0.9802013706743379,
+      "grad_norm": 0.15773500502109528,
+      "learning_rate": 0.001,
+      "loss": 1.7977,
+      "step": 23170
+    },
+    {
+      "epoch": 0.9802436754378543,
+      "grad_norm": 0.2358037680387497,
+      "learning_rate": 0.001,
+      "loss": 2.6206,
+      "step": 23171
+    },
+    {
+      "epoch": 0.9802859802013707,
+      "grad_norm": 0.17311686277389526,
+      "learning_rate": 0.001,
+      "loss": 2.619,
+      "step": 23172
+    },
+    {
+      "epoch": 0.980328284964887,
+      "grad_norm": 15.802042961120605,
+      "learning_rate": 0.001,
+      "loss": 1.6373,
+      "step": 23173
+    },
+    {
+      "epoch": 0.9803705897284034,
+      "grad_norm": 0.14769822359085083,
+      "learning_rate": 0.001,
+      "loss": 1.7526,
+      "step": 23174
+    },
+    {
+      "epoch": 0.9804128944919198,
+      "grad_norm": 2.210271120071411,
+      "learning_rate": 0.001,
+      "loss": 1.7164,
+      "step": 23175
+    },
+    {
+      "epoch": 0.9804551992554361,
+      "grad_norm": 0.15342247486114502,
+      "learning_rate": 0.001,
+      "loss": 3.1296,
+      "step": 23176
+    },
+    {
+      "epoch": 0.9804975040189525,
+      "grad_norm": 0.1497957855463028,
+      "learning_rate": 0.001,
+      "loss": 2.6274,
+      "step": 23177
+    },
+    {
+      "epoch": 0.9805398087824689,
+      "grad_norm": 0.1614879071712494,
+      "learning_rate": 0.001,
+      "loss": 1.4673,
+      "step": 23178
+    },
+    {
+      "epoch": 0.9805821135459852,
+      "grad_norm": 0.1438165307044983,
+      "learning_rate": 0.001,
+      "loss": 3.2337,
+      "step": 23179
+    },
+    {
+      "epoch": 0.9806244183095016,
+      "grad_norm": 0.1747131198644638,
+      "learning_rate": 0.001,
+      "loss": 2.1779,
+      "step": 23180
+    },
+    {
+      "epoch": 0.980666723073018,
+      "grad_norm": 0.15409283339977264,
+      "learning_rate": 0.001,
+      "loss": 2.4793,
+      "step": 23181
+    },
+    {
+      "epoch": 0.9807090278365344,
+      "grad_norm": 0.1740875393152237,
+      "learning_rate": 0.001,
+      "loss": 2.0542,
+      "step": 23182
+    },
+    {
+      "epoch": 0.9807513326000508,
+      "grad_norm": 0.13922379910945892,
+      "learning_rate": 0.001,
+      "loss": 2.1908,
+      "step": 23183
+    },
+    {
+      "epoch": 0.9807936373635672,
+      "grad_norm": 0.20002827048301697,
+      "learning_rate": 0.001,
+      "loss": 1.9136,
+      "step": 23184
+    },
+    {
+      "epoch": 0.9808359421270835,
+      "grad_norm": 0.13429036736488342,
+      "learning_rate": 0.001,
+      "loss": 1.3618,
+      "step": 23185
+    },
+    {
+      "epoch": 0.9808782468905999,
+      "grad_norm": 0.16126492619514465,
+      "learning_rate": 0.001,
+      "loss": 2.7224,
+      "step": 23186
+    },
+    {
+      "epoch": 0.9809205516541163,
+      "grad_norm": 0.14296741783618927,
+      "learning_rate": 0.001,
+      "loss": 3.2186,
+      "step": 23187
+    },
+    {
+      "epoch": 0.9809628564176326,
+      "grad_norm": 5.455838203430176,
+      "learning_rate": 0.001,
+      "loss": 2.5349,
+      "step": 23188
+    },
+    {
+      "epoch": 0.981005161181149,
+      "grad_norm": 0.13420797884464264,
+      "learning_rate": 0.001,
+      "loss": 1.4715,
+      "step": 23189
+    },
+    {
+      "epoch": 0.9810474659446654,
+      "grad_norm": 0.2915991544723511,
+      "learning_rate": 0.001,
+      "loss": 1.9732,
+      "step": 23190
+    },
+    {
+      "epoch": 0.9810897707081817,
+      "grad_norm": 0.12874186038970947,
+      "learning_rate": 0.001,
+      "loss": 1.8352,
+      "step": 23191
+    },
+    {
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.14853565394878387,
+      "learning_rate": 0.001,
+      "loss": 2.8209,
+      "step": 23192
+    },
+    {
+      "epoch": 0.9811743802352145,
+      "grad_norm": 0.14589852094650269,
+      "learning_rate": 0.001,
+      "loss": 2.0377,
+      "step": 23193
+    },
+    {
+      "epoch": 0.9812166849987308,
+      "grad_norm": 0.15394212305545807,
+      "learning_rate": 0.001,
+      "loss": 1.9553,
+      "step": 23194
+    },
+    {
+      "epoch": 0.9812589897622472,
+      "grad_norm": 0.1757000833749771,
+      "learning_rate": 0.001,
+      "loss": 2.2428,
+      "step": 23195
+    },
+    {
+      "epoch": 0.9813012945257636,
+      "grad_norm": 0.15882089734077454,
+      "learning_rate": 0.001,
+      "loss": 2.0133,
+      "step": 23196
+    },
+    {
+      "epoch": 0.9813435992892799,
+      "grad_norm": 0.14844751358032227,
+      "learning_rate": 0.001,
+      "loss": 1.7509,
+      "step": 23197
+    },
+    {
+      "epoch": 0.9813859040527964,
+      "grad_norm": 0.16744114458560944,
+      "learning_rate": 0.001,
+      "loss": 1.7146,
+      "step": 23198
+    },
+    {
+      "epoch": 0.9814282088163128,
+      "grad_norm": 0.1688256859779358,
+      "learning_rate": 0.001,
+      "loss": 3.276,
+      "step": 23199
+    },
+    {
+      "epoch": 0.9814705135798291,
+      "grad_norm": 4.072800159454346,
+      "learning_rate": 0.001,
+      "loss": 2.5557,
+      "step": 23200
+    },
+    {
+      "epoch": 0.9815128183433455,
+      "grad_norm": 0.1719663292169571,
+      "learning_rate": 0.001,
+      "loss": 1.8419,
+      "step": 23201
+    },
+    {
+      "epoch": 0.9815551231068619,
+      "grad_norm": 0.15536484122276306,
+      "learning_rate": 0.001,
+      "loss": 1.8884,
+      "step": 23202
+    },
+    {
+      "epoch": 0.9815974278703782,
+      "grad_norm": 1.0643222332000732,
+      "learning_rate": 0.001,
+      "loss": 1.9732,
+      "step": 23203
+    },
+    {
+      "epoch": 0.9816397326338946,
+      "grad_norm": 0.19347727298736572,
+      "learning_rate": 0.001,
+      "loss": 2.0092,
+      "step": 23204
+    },
+    {
+      "epoch": 0.981682037397411,
+      "grad_norm": 0.14213740825653076,
+      "learning_rate": 0.001,
+      "loss": 2.3664,
+      "step": 23205
+    },
+    {
+      "epoch": 0.9817243421609273,
+      "grad_norm": 0.5251598954200745,
+      "learning_rate": 0.001,
+      "loss": 2.6287,
+      "step": 23206
+    },
+    {
+      "epoch": 0.9817666469244437,
+      "grad_norm": 0.19425900280475616,
+      "learning_rate": 0.001,
+      "loss": 2.4588,
+      "step": 23207
+    },
+    {
+      "epoch": 0.9818089516879601,
+      "grad_norm": 1.0385212898254395,
+      "learning_rate": 0.001,
+      "loss": 1.961,
+      "step": 23208
+    },
+    {
+      "epoch": 0.9818512564514764,
+      "grad_norm": 3.3503012657165527,
+      "learning_rate": 0.001,
+      "loss": 2.6355,
+      "step": 23209
+    },
+    {
+      "epoch": 0.9818935612149928,
+      "grad_norm": 0.344450980424881,
+      "learning_rate": 0.001,
+      "loss": 3.4627,
+      "step": 23210
+    },
+    {
+      "epoch": 0.9819358659785091,
+      "grad_norm": 0.3758525252342224,
+      "learning_rate": 0.001,
+      "loss": 2.3434,
+      "step": 23211
+    },
+    {
+      "epoch": 0.9819781707420255,
+      "grad_norm": 0.23000246286392212,
+      "learning_rate": 0.001,
+      "loss": 2.1027,
+      "step": 23212
+    },
+    {
+      "epoch": 0.982020475505542,
+      "grad_norm": 0.21098732948303223,
+      "learning_rate": 0.001,
+      "loss": 1.8135,
+      "step": 23213
+    },
+    {
+      "epoch": 0.9820627802690582,
+      "grad_norm": 0.3435835540294647,
+      "learning_rate": 0.001,
+      "loss": 2.3019,
+      "step": 23214
+    },
+    {
+      "epoch": 0.9821050850325747,
+      "grad_norm": 0.1896451711654663,
+      "learning_rate": 0.001,
+      "loss": 3.6827,
+      "step": 23215
+    },
+    {
+      "epoch": 0.9821473897960911,
+      "grad_norm": 0.9407826066017151,
+      "learning_rate": 0.001,
+      "loss": 2.207,
+      "step": 23216
+    },
+    {
+      "epoch": 0.9821896945596074,
+      "grad_norm": 0.17004285752773285,
+      "learning_rate": 0.001,
+      "loss": 1.6758,
+      "step": 23217
+    },
+    {
+      "epoch": 0.9822319993231238,
+      "grad_norm": 0.21439626812934875,
+      "learning_rate": 0.001,
+      "loss": 2.7161,
+      "step": 23218
+    },
+    {
+      "epoch": 0.9822743040866402,
+      "grad_norm": 0.6642925143241882,
+      "learning_rate": 0.001,
+      "loss": 2.7868,
+      "step": 23219
+    },
+    {
+      "epoch": 0.9823166088501565,
+      "grad_norm": 0.20337320864200592,
+      "learning_rate": 0.001,
+      "loss": 1.9839,
+      "step": 23220
+    },
+    {
+      "epoch": 0.9823589136136729,
+      "grad_norm": 0.15737198293209076,
+      "learning_rate": 0.001,
+      "loss": 2.3604,
+      "step": 23221
+    },
+    {
+      "epoch": 0.9824012183771893,
+      "grad_norm": 0.6608482599258423,
+      "learning_rate": 0.001,
+      "loss": 1.6606,
+      "step": 23222
+    },
+    {
+      "epoch": 0.9824435231407056,
+      "grad_norm": 0.11500653624534607,
+      "learning_rate": 0.001,
+      "loss": 2.2689,
+      "step": 23223
+    },
+    {
+      "epoch": 0.982485827904222,
+      "grad_norm": 0.2476760298013687,
+      "learning_rate": 0.001,
+      "loss": 2.4611,
+      "step": 23224
+    },
+    {
+      "epoch": 0.9825281326677384,
+      "grad_norm": 0.20634648203849792,
+      "learning_rate": 0.001,
+      "loss": 3.2298,
+      "step": 23225
+    },
+    {
+      "epoch": 0.9825704374312547,
+      "grad_norm": 0.174639493227005,
+      "learning_rate": 0.001,
+      "loss": 2.3492,
+      "step": 23226
+    },
+    {
+      "epoch": 0.9826127421947711,
+      "grad_norm": 0.1685517132282257,
+      "learning_rate": 0.001,
+      "loss": 3.3922,
+      "step": 23227
+    },
+    {
+      "epoch": 0.9826550469582875,
+      "grad_norm": 0.17105820775032043,
+      "learning_rate": 0.001,
+      "loss": 1.8573,
+      "step": 23228
+    },
+    {
+      "epoch": 0.9826973517218038,
+      "grad_norm": 0.1790645271539688,
+      "learning_rate": 0.001,
+      "loss": 3.3843,
+      "step": 23229
+    },
+    {
+      "epoch": 0.9827396564853202,
+      "grad_norm": 0.14651429653167725,
+      "learning_rate": 0.001,
+      "loss": 2.4584,
+      "step": 23230
+    },
+    {
+      "epoch": 0.9827819612488367,
+      "grad_norm": 0.15420038998126984,
+      "learning_rate": 0.001,
+      "loss": 1.8876,
+      "step": 23231
+    },
+    {
+      "epoch": 0.982824266012353,
+      "grad_norm": 0.13720165193080902,
+      "learning_rate": 0.001,
+      "loss": 1.6773,
+      "step": 23232
+    },
+    {
+      "epoch": 0.9828665707758694,
+      "grad_norm": 0.3285228908061981,
+      "learning_rate": 0.001,
+      "loss": 2.1622,
+      "step": 23233
+    },
+    {
+      "epoch": 0.9829088755393858,
+      "grad_norm": 0.22455190122127533,
+      "learning_rate": 0.001,
+      "loss": 3.7145,
+      "step": 23234
+    },
+    {
+      "epoch": 0.9829511803029021,
+      "grad_norm": 0.14361530542373657,
+      "learning_rate": 0.001,
+      "loss": 2.551,
+      "step": 23235
+    },
+    {
+      "epoch": 0.9829934850664185,
+      "grad_norm": 0.20710772275924683,
+      "learning_rate": 0.001,
+      "loss": 1.8731,
+      "step": 23236
+    },
+    {
+      "epoch": 0.9830357898299349,
+      "grad_norm": 0.15071751177310944,
+      "learning_rate": 0.001,
+      "loss": 1.4985,
+      "step": 23237
+    },
+    {
+      "epoch": 0.9830780945934512,
+      "grad_norm": 0.15085220336914062,
+      "learning_rate": 0.001,
+      "loss": 2.0827,
+      "step": 23238
+    },
+    {
+      "epoch": 0.9831203993569676,
+      "grad_norm": 0.1446499526500702,
+      "learning_rate": 0.001,
+      "loss": 2.4606,
+      "step": 23239
+    },
+    {
+      "epoch": 0.983162704120484,
+      "grad_norm": 0.1713118702173233,
+      "learning_rate": 0.001,
+      "loss": 1.947,
+      "step": 23240
+    },
+    {
+      "epoch": 0.9832050088840003,
+      "grad_norm": 0.23264853656291962,
+      "learning_rate": 0.001,
+      "loss": 1.7186,
+      "step": 23241
+    },
+    {
+      "epoch": 0.9832473136475167,
+      "grad_norm": 0.15223060548305511,
+      "learning_rate": 0.001,
+      "loss": 2.3135,
+      "step": 23242
+    },
+    {
+      "epoch": 0.9832896184110331,
+      "grad_norm": 2.1504344940185547,
+      "learning_rate": 0.001,
+      "loss": 2.345,
+      "step": 23243
+    },
+    {
+      "epoch": 0.9833319231745494,
+      "grad_norm": 0.2231014370918274,
+      "learning_rate": 0.001,
+      "loss": 2.4845,
+      "step": 23244
+    },
+    {
+      "epoch": 0.9833742279380658,
+      "grad_norm": 30.52190399169922,
+      "learning_rate": 0.001,
+      "loss": 1.5587,
+      "step": 23245
+    },
+    {
+      "epoch": 0.9834165327015822,
+      "grad_norm": 0.1924566775560379,
+      "learning_rate": 0.001,
+      "loss": 1.8833,
+      "step": 23246
+    },
+    {
+      "epoch": 0.9834588374650985,
+      "grad_norm": 0.17296500504016876,
+      "learning_rate": 0.001,
+      "loss": 1.9576,
+      "step": 23247
+    },
+    {
+      "epoch": 0.983501142228615,
+      "grad_norm": 0.17767272889614105,
+      "learning_rate": 0.001,
+      "loss": 2.5095,
+      "step": 23248
+    },
+    {
+      "epoch": 0.9835434469921314,
+      "grad_norm": 16.753442764282227,
+      "learning_rate": 0.001,
+      "loss": 2.5697,
+      "step": 23249
+    },
+    {
+      "epoch": 0.9835857517556477,
+      "grad_norm": 0.19115328788757324,
+      "learning_rate": 0.001,
+      "loss": 1.6478,
+      "step": 23250
+    },
+    {
+      "epoch": 0.9836280565191641,
+      "grad_norm": 22.24527359008789,
+      "learning_rate": 0.001,
+      "loss": 2.6725,
+      "step": 23251
+    },
+    {
+      "epoch": 0.9836703612826805,
+      "grad_norm": 2.2926394939422607,
+      "learning_rate": 0.001,
+      "loss": 1.8396,
+      "step": 23252
+    },
+    {
+      "epoch": 0.9837126660461968,
+      "grad_norm": 0.22507277131080627,
+      "learning_rate": 0.001,
+      "loss": 2.7259,
+      "step": 23253
+    },
+    {
+      "epoch": 0.9837549708097132,
+      "grad_norm": 0.20244504511356354,
+      "learning_rate": 0.001,
+      "loss": 1.8665,
+      "step": 23254
+    },
+    {
+      "epoch": 0.9837972755732295,
+      "grad_norm": 0.17822127044200897,
+      "learning_rate": 0.001,
+      "loss": 2.7145,
+      "step": 23255
+    },
+    {
+      "epoch": 0.9838395803367459,
+      "grad_norm": 0.21497797966003418,
+      "learning_rate": 0.001,
+      "loss": 1.9941,
+      "step": 23256
+    },
+    {
+      "epoch": 0.9838818851002623,
+      "grad_norm": 0.2271234095096588,
+      "learning_rate": 0.001,
+      "loss": 2.2163,
+      "step": 23257
+    },
+    {
+      "epoch": 0.9839241898637786,
+      "grad_norm": 1.1759521961212158,
+      "learning_rate": 0.001,
+      "loss": 2.7644,
+      "step": 23258
+    },
+    {
+      "epoch": 0.983966494627295,
+      "grad_norm": 0.19246691465377808,
+      "learning_rate": 0.001,
+      "loss": 1.9231,
+      "step": 23259
+    },
+    {
+      "epoch": 0.9840087993908114,
+      "grad_norm": 4.506860733032227,
+      "learning_rate": 0.001,
+      "loss": 2.4722,
+      "step": 23260
+    },
+    {
+      "epoch": 0.9840511041543277,
+      "grad_norm": 0.20085294544696808,
+      "learning_rate": 0.001,
+      "loss": 3.0157,
+      "step": 23261
+    },
+    {
+      "epoch": 0.9840934089178441,
+      "grad_norm": 1.484424352645874,
+      "learning_rate": 0.001,
+      "loss": 2.2361,
+      "step": 23262
+    },
+    {
+      "epoch": 0.9841357136813605,
+      "grad_norm": 0.21662937104701996,
+      "learning_rate": 0.001,
+      "loss": 2.2051,
+      "step": 23263
+    },
+    {
+      "epoch": 0.9841780184448768,
+      "grad_norm": 0.20913396775722504,
+      "learning_rate": 0.001,
+      "loss": 1.697,
+      "step": 23264
+    },
+    {
+      "epoch": 0.9842203232083933,
+      "grad_norm": 0.19714033603668213,
+      "learning_rate": 0.001,
+      "loss": 2.0469,
+      "step": 23265
+    },
+    {
+      "epoch": 0.9842626279719097,
+      "grad_norm": 0.18258053064346313,
+      "learning_rate": 0.001,
+      "loss": 2.446,
+      "step": 23266
+    },
+    {
+      "epoch": 0.984304932735426,
+      "grad_norm": 0.43898874521255493,
+      "learning_rate": 0.001,
+      "loss": 1.9534,
+      "step": 23267
+    },
+    {
+      "epoch": 0.9843472374989424,
+      "grad_norm": 0.16025614738464355,
+      "learning_rate": 0.001,
+      "loss": 2.2371,
+      "step": 23268
+    },
+    {
+      "epoch": 0.9843895422624588,
+      "grad_norm": 0.4538215100765228,
+      "learning_rate": 0.001,
+      "loss": 1.9265,
+      "step": 23269
+    },
+    {
+      "epoch": 0.9844318470259751,
+      "grad_norm": 0.18556450307369232,
+      "learning_rate": 0.001,
+      "loss": 1.5331,
+      "step": 23270
+    },
+    {
+      "epoch": 0.9844741517894915,
+      "grad_norm": 0.15075351297855377,
+      "learning_rate": 0.001,
+      "loss": 3.1292,
+      "step": 23271
+    },
+    {
+      "epoch": 0.9845164565530079,
+      "grad_norm": 19.696006774902344,
+      "learning_rate": 0.001,
+      "loss": 2.7711,
+      "step": 23272
+    },
+    {
+      "epoch": 0.9845587613165242,
+      "grad_norm": 0.22151979804039001,
+      "learning_rate": 0.001,
+      "loss": 1.6431,
+      "step": 23273
+    },
+    {
+      "epoch": 0.9846010660800406,
+      "grad_norm": 0.27550455927848816,
+      "learning_rate": 0.001,
+      "loss": 2.4787,
+      "step": 23274
+    },
+    {
+      "epoch": 0.984643370843557,
+      "grad_norm": 0.22198323905467987,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 23275
+    },
+    {
+      "epoch": 0.9846856756070733,
+      "grad_norm": 0.20321761071681976,
+      "learning_rate": 0.001,
+      "loss": 2.4743,
+      "step": 23276
+    },
+    {
+      "epoch": 0.9847279803705897,
+      "grad_norm": 0.19668048620224,
+      "learning_rate": 0.001,
+      "loss": 1.825,
+      "step": 23277
+    },
+    {
+      "epoch": 0.9847702851341061,
+      "grad_norm": 0.7689309120178223,
+      "learning_rate": 0.001,
+      "loss": 3.5899,
+      "step": 23278
+    },
+    {
+      "epoch": 0.9848125898976224,
+      "grad_norm": 0.18673300743103027,
+      "learning_rate": 0.001,
+      "loss": 2.8243,
+      "step": 23279
+    },
+    {
+      "epoch": 0.9848548946611388,
+      "grad_norm": 1.7281594276428223,
+      "learning_rate": 0.001,
+      "loss": 1.5403,
+      "step": 23280
+    },
+    {
+      "epoch": 0.9848971994246553,
+      "grad_norm": 0.15205202996730804,
+      "learning_rate": 0.001,
+      "loss": 3.1236,
+      "step": 23281
+    },
+    {
+      "epoch": 0.9849395041881716,
+      "grad_norm": 0.14774078130722046,
+      "learning_rate": 0.001,
+      "loss": 2.1822,
+      "step": 23282
+    },
+    {
+      "epoch": 0.984981808951688,
+      "grad_norm": 0.15019498765468597,
+      "learning_rate": 0.001,
+      "loss": 1.8555,
+      "step": 23283
+    },
+    {
+      "epoch": 0.9850241137152044,
+      "grad_norm": 0.17441150546073914,
+      "learning_rate": 0.001,
+      "loss": 1.8417,
+      "step": 23284
+    },
+    {
+      "epoch": 0.9850664184787207,
+      "grad_norm": 0.16328319907188416,
+      "learning_rate": 0.001,
+      "loss": 1.957,
+      "step": 23285
+    },
+    {
+      "epoch": 0.9851087232422371,
+      "grad_norm": 1.1682124137878418,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 23286
+    },
+    {
+      "epoch": 0.9851510280057535,
+      "grad_norm": 0.8976200819015503,
+      "learning_rate": 0.001,
+      "loss": 2.9577,
+      "step": 23287
+    },
+    {
+      "epoch": 0.9851933327692698,
+      "grad_norm": 0.18787872791290283,
+      "learning_rate": 0.001,
+      "loss": 2.205,
+      "step": 23288
+    },
+    {
+      "epoch": 0.9852356375327862,
+      "grad_norm": 0.17400221526622772,
+      "learning_rate": 0.001,
+      "loss": 1.8947,
+      "step": 23289
+    },
+    {
+      "epoch": 0.9852779422963026,
+      "grad_norm": 0.16904416680335999,
+      "learning_rate": 0.001,
+      "loss": 1.7336,
+      "step": 23290
+    },
+    {
+      "epoch": 0.9853202470598189,
+      "grad_norm": 0.2862572968006134,
+      "learning_rate": 0.001,
+      "loss": 2.0924,
+      "step": 23291
+    },
+    {
+      "epoch": 0.9853625518233353,
+      "grad_norm": 0.18694454431533813,
+      "learning_rate": 0.001,
+      "loss": 2.4069,
+      "step": 23292
+    },
+    {
+      "epoch": 0.9854048565868517,
+      "grad_norm": 0.17351536452770233,
+      "learning_rate": 0.001,
+      "loss": 2.3785,
+      "step": 23293
+    },
+    {
+      "epoch": 0.985447161350368,
+      "grad_norm": 0.16305550932884216,
+      "learning_rate": 0.001,
+      "loss": 2.5295,
+      "step": 23294
+    },
+    {
+      "epoch": 0.9854894661138844,
+      "grad_norm": 0.28731977939605713,
+      "learning_rate": 0.001,
+      "loss": 1.5698,
+      "step": 23295
+    },
+    {
+      "epoch": 0.9855317708774008,
+      "grad_norm": 0.19942089915275574,
+      "learning_rate": 0.001,
+      "loss": 2.6501,
+      "step": 23296
+    },
+    {
+      "epoch": 0.9855740756409171,
+      "grad_norm": 0.15859198570251465,
+      "learning_rate": 0.001,
+      "loss": 1.9334,
+      "step": 23297
+    },
+    {
+      "epoch": 0.9856163804044336,
+      "grad_norm": 0.17347994446754456,
+      "learning_rate": 0.001,
+      "loss": 2.281,
+      "step": 23298
+    },
+    {
+      "epoch": 0.98565868516795,
+      "grad_norm": 0.1470501720905304,
+      "learning_rate": 0.001,
+      "loss": 2.8596,
+      "step": 23299
+    },
+    {
+      "epoch": 0.9857009899314663,
+      "grad_norm": 0.17378562688827515,
+      "learning_rate": 0.001,
+      "loss": 2.6134,
+      "step": 23300
+    },
+    {
+      "epoch": 0.9857432946949827,
+      "grad_norm": 2.3224494457244873,
+      "learning_rate": 0.001,
+      "loss": 1.8163,
+      "step": 23301
+    },
+    {
+      "epoch": 0.985785599458499,
+      "grad_norm": 0.15434542298316956,
+      "learning_rate": 0.001,
+      "loss": 2.0795,
+      "step": 23302
+    },
+    {
+      "epoch": 0.9858279042220154,
+      "grad_norm": 0.17513182759284973,
+      "learning_rate": 0.001,
+      "loss": 1.9282,
+      "step": 23303
+    },
+    {
+      "epoch": 0.9858702089855318,
+      "grad_norm": 0.18898020684719086,
+      "learning_rate": 0.001,
+      "loss": 2.8024,
+      "step": 23304
+    },
+    {
+      "epoch": 0.9859125137490481,
+      "grad_norm": 0.5918177366256714,
+      "learning_rate": 0.001,
+      "loss": 2.3883,
+      "step": 23305
+    },
+    {
+      "epoch": 0.9859548185125645,
+      "grad_norm": 0.13751918077468872,
+      "learning_rate": 0.001,
+      "loss": 1.7054,
+      "step": 23306
+    },
+    {
+      "epoch": 0.9859971232760809,
+      "grad_norm": 0.1347641944885254,
+      "learning_rate": 0.001,
+      "loss": 1.615,
+      "step": 23307
+    },
+    {
+      "epoch": 0.9860394280395972,
+      "grad_norm": 0.21464622020721436,
+      "learning_rate": 0.001,
+      "loss": 2.3099,
+      "step": 23308
+    },
+    {
+      "epoch": 0.9860817328031136,
+      "grad_norm": 0.15518558025360107,
+      "learning_rate": 0.001,
+      "loss": 2.1591,
+      "step": 23309
+    },
+    {
+      "epoch": 0.98612403756663,
+      "grad_norm": 0.1407310962677002,
+      "learning_rate": 0.001,
+      "loss": 1.9204,
+      "step": 23310
+    },
+    {
+      "epoch": 0.9861663423301463,
+      "grad_norm": 0.11285633593797684,
+      "learning_rate": 0.001,
+      "loss": 1.7241,
+      "step": 23311
+    },
+    {
+      "epoch": 0.9862086470936627,
+      "grad_norm": 8.615754127502441,
+      "learning_rate": 0.001,
+      "loss": 2.5709,
+      "step": 23312
+    },
+    {
+      "epoch": 0.9862509518571791,
+      "grad_norm": 0.13533753156661987,
+      "learning_rate": 0.001,
+      "loss": 1.6491,
+      "step": 23313
+    },
+    {
+      "epoch": 0.9862932566206954,
+      "grad_norm": 0.16837237775325775,
+      "learning_rate": 0.001,
+      "loss": 2.4037,
+      "step": 23314
+    },
+    {
+      "epoch": 0.9863355613842119,
+      "grad_norm": 0.14844800531864166,
+      "learning_rate": 0.001,
+      "loss": 1.6873,
+      "step": 23315
+    },
+    {
+      "epoch": 0.9863778661477283,
+      "grad_norm": 0.13803304731845856,
+      "learning_rate": 0.001,
+      "loss": 2.2417,
+      "step": 23316
+    },
+    {
+      "epoch": 0.9864201709112446,
+      "grad_norm": 0.19187100231647491,
+      "learning_rate": 0.001,
+      "loss": 1.9479,
+      "step": 23317
+    },
+    {
+      "epoch": 0.986462475674761,
+      "grad_norm": 1.3388183116912842,
+      "learning_rate": 0.001,
+      "loss": 1.8091,
+      "step": 23318
+    },
+    {
+      "epoch": 0.9865047804382774,
+      "grad_norm": 0.18258285522460938,
+      "learning_rate": 0.001,
+      "loss": 2.8915,
+      "step": 23319
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 0.1443546712398529,
+      "learning_rate": 0.001,
+      "loss": 2.0421,
+      "step": 23320
+    },
+    {
+      "epoch": 0.9865893899653101,
+      "grad_norm": 0.16136842966079712,
+      "learning_rate": 0.001,
+      "loss": 2.1917,
+      "step": 23321
+    },
+    {
+      "epoch": 0.9866316947288265,
+      "grad_norm": 0.13147707283496857,
+      "learning_rate": 0.001,
+      "loss": 1.6761,
+      "step": 23322
+    },
+    {
+      "epoch": 0.9866739994923428,
+      "grad_norm": 0.18694503605365753,
+      "learning_rate": 0.001,
+      "loss": 2.0926,
+      "step": 23323
+    },
+    {
+      "epoch": 0.9867163042558592,
+      "grad_norm": 0.1893046796321869,
+      "learning_rate": 0.001,
+      "loss": 2.8645,
+      "step": 23324
+    },
+    {
+      "epoch": 0.9867586090193756,
+      "grad_norm": 0.2780762314796448,
+      "learning_rate": 0.001,
+      "loss": 2.0847,
+      "step": 23325
+    },
+    {
+      "epoch": 0.9868009137828919,
+      "grad_norm": 0.19630302488803864,
+      "learning_rate": 0.001,
+      "loss": 2.0885,
+      "step": 23326
+    },
+    {
+      "epoch": 0.9868432185464083,
+      "grad_norm": 0.1403011828660965,
+      "learning_rate": 0.001,
+      "loss": 1.564,
+      "step": 23327
+    },
+    {
+      "epoch": 0.9868855233099247,
+      "grad_norm": 0.8801196813583374,
+      "learning_rate": 0.001,
+      "loss": 2.5189,
+      "step": 23328
+    },
+    {
+      "epoch": 0.986927828073441,
+      "grad_norm": 0.13722547888755798,
+      "learning_rate": 0.001,
+      "loss": 2.0738,
+      "step": 23329
+    },
+    {
+      "epoch": 0.9869701328369574,
+      "grad_norm": 0.42092952132225037,
+      "learning_rate": 0.001,
+      "loss": 1.6655,
+      "step": 23330
+    },
+    {
+      "epoch": 0.9870124376004739,
+      "grad_norm": 0.19227460026741028,
+      "learning_rate": 0.001,
+      "loss": 2.1327,
+      "step": 23331
+    },
+    {
+      "epoch": 0.9870547423639902,
+      "grad_norm": 0.14953431487083435,
+      "learning_rate": 0.001,
+      "loss": 1.5976,
+      "step": 23332
+    },
+    {
+      "epoch": 0.9870970471275066,
+      "grad_norm": 1.6279288530349731,
+      "learning_rate": 0.001,
+      "loss": 2.4999,
+      "step": 23333
+    },
+    {
+      "epoch": 0.987139351891023,
+      "grad_norm": 0.15288491547107697,
+      "learning_rate": 0.001,
+      "loss": 1.9023,
+      "step": 23334
+    },
+    {
+      "epoch": 0.9871816566545393,
+      "grad_norm": 0.16339105367660522,
+      "learning_rate": 0.001,
+      "loss": 1.5201,
+      "step": 23335
+    },
+    {
+      "epoch": 0.9872239614180557,
+      "grad_norm": 0.19303716719150543,
+      "learning_rate": 0.001,
+      "loss": 2.255,
+      "step": 23336
+    },
+    {
+      "epoch": 0.9872662661815721,
+      "grad_norm": 0.16261610388755798,
+      "learning_rate": 0.001,
+      "loss": 1.7459,
+      "step": 23337
+    },
+    {
+      "epoch": 0.9873085709450884,
+      "grad_norm": 0.2173750400543213,
+      "learning_rate": 0.001,
+      "loss": 2.1332,
+      "step": 23338
+    },
+    {
+      "epoch": 0.9873508757086048,
+      "grad_norm": 0.5034084320068359,
+      "learning_rate": 0.001,
+      "loss": 2.0107,
+      "step": 23339
+    },
+    {
+      "epoch": 0.9873931804721212,
+      "grad_norm": 0.1737852692604065,
+      "learning_rate": 0.001,
+      "loss": 1.8748,
+      "step": 23340
+    },
+    {
+      "epoch": 0.9874354852356375,
+      "grad_norm": 0.19651900231838226,
+      "learning_rate": 0.001,
+      "loss": 2.7285,
+      "step": 23341
+    },
+    {
+      "epoch": 0.9874777899991539,
+      "grad_norm": 0.14730535447597504,
+      "learning_rate": 0.001,
+      "loss": 2.8565,
+      "step": 23342
+    },
+    {
+      "epoch": 0.9875200947626703,
+      "grad_norm": 0.15681281685829163,
+      "learning_rate": 0.001,
+      "loss": 2.0837,
+      "step": 23343
+    },
+    {
+      "epoch": 0.9875623995261866,
+      "grad_norm": 0.17979633808135986,
+      "learning_rate": 0.001,
+      "loss": 2.1084,
+      "step": 23344
+    },
+    {
+      "epoch": 0.987604704289703,
+      "grad_norm": 0.29083535075187683,
+      "learning_rate": 0.001,
+      "loss": 2.7915,
+      "step": 23345
+    },
+    {
+      "epoch": 0.9876470090532193,
+      "grad_norm": 0.16025808453559875,
+      "learning_rate": 0.001,
+      "loss": 2.1277,
+      "step": 23346
+    },
+    {
+      "epoch": 0.9876893138167357,
+      "grad_norm": 0.187299445271492,
+      "learning_rate": 0.001,
+      "loss": 2.0213,
+      "step": 23347
+    },
+    {
+      "epoch": 0.9877316185802522,
+      "grad_norm": 0.1659306287765503,
+      "learning_rate": 0.001,
+      "loss": 1.4863,
+      "step": 23348
+    },
+    {
+      "epoch": 0.9877739233437685,
+      "grad_norm": 0.13653291761875153,
+      "learning_rate": 0.001,
+      "loss": 2.0724,
+      "step": 23349
+    },
+    {
+      "epoch": 0.9878162281072849,
+      "grad_norm": 0.19335125386714935,
+      "learning_rate": 0.001,
+      "loss": 2.5707,
+      "step": 23350
+    },
+    {
+      "epoch": 0.9878585328708013,
+      "grad_norm": 0.17489448189735413,
+      "learning_rate": 0.001,
+      "loss": 2.4171,
+      "step": 23351
+    },
+    {
+      "epoch": 0.9879008376343176,
+      "grad_norm": 0.15927927196025848,
+      "learning_rate": 0.001,
+      "loss": 1.9707,
+      "step": 23352
+    },
+    {
+      "epoch": 0.987943142397834,
+      "grad_norm": 0.13132639229297638,
+      "learning_rate": 0.001,
+      "loss": 1.7196,
+      "step": 23353
+    },
+    {
+      "epoch": 0.9879854471613504,
+      "grad_norm": 0.7293382883071899,
+      "learning_rate": 0.001,
+      "loss": 2.0281,
+      "step": 23354
+    },
+    {
+      "epoch": 0.9880277519248667,
+      "grad_norm": 0.151032954454422,
+      "learning_rate": 0.001,
+      "loss": 1.6911,
+      "step": 23355
+    },
+    {
+      "epoch": 0.9880700566883831,
+      "grad_norm": 0.13139599561691284,
+      "learning_rate": 0.001,
+      "loss": 1.4657,
+      "step": 23356
+    },
+    {
+      "epoch": 0.9881123614518995,
+      "grad_norm": 0.14863912761211395,
+      "learning_rate": 0.001,
+      "loss": 1.7769,
+      "step": 23357
+    },
+    {
+      "epoch": 0.9881546662154158,
+      "grad_norm": 0.14839519560337067,
+      "learning_rate": 0.001,
+      "loss": 2.1032,
+      "step": 23358
+    },
+    {
+      "epoch": 0.9881969709789322,
+      "grad_norm": 0.12207513302564621,
+      "learning_rate": 0.001,
+      "loss": 1.7226,
+      "step": 23359
+    },
+    {
+      "epoch": 0.9882392757424486,
+      "grad_norm": 0.18079492449760437,
+      "learning_rate": 0.001,
+      "loss": 1.8141,
+      "step": 23360
+    },
+    {
+      "epoch": 0.9882815805059649,
+      "grad_norm": 0.1407337784767151,
+      "learning_rate": 0.001,
+      "loss": 2.2357,
+      "step": 23361
+    },
+    {
+      "epoch": 0.9883238852694813,
+      "grad_norm": 0.1415151059627533,
+      "learning_rate": 0.001,
+      "loss": 1.8046,
+      "step": 23362
+    },
+    {
+      "epoch": 0.9883661900329977,
+      "grad_norm": 2.5134902000427246,
+      "learning_rate": 0.001,
+      "loss": 2.7073,
+      "step": 23363
+    },
+    {
+      "epoch": 0.988408494796514,
+      "grad_norm": 0.16898906230926514,
+      "learning_rate": 0.001,
+      "loss": 3.0829,
+      "step": 23364
+    },
+    {
+      "epoch": 0.9884507995600305,
+      "grad_norm": 0.13951623439788818,
+      "learning_rate": 0.001,
+      "loss": 2.0555,
+      "step": 23365
+    },
+    {
+      "epoch": 0.9884931043235469,
+      "grad_norm": 0.14504913985729218,
+      "learning_rate": 0.001,
+      "loss": 2.0594,
+      "step": 23366
+    },
+    {
+      "epoch": 0.9885354090870632,
+      "grad_norm": 0.5770459771156311,
+      "learning_rate": 0.001,
+      "loss": 2.0616,
+      "step": 23367
+    },
+    {
+      "epoch": 0.9885777138505796,
+      "grad_norm": 0.16317375004291534,
+      "learning_rate": 0.001,
+      "loss": 2.1634,
+      "step": 23368
+    },
+    {
+      "epoch": 0.988620018614096,
+      "grad_norm": 0.1316281110048294,
+      "learning_rate": 0.001,
+      "loss": 2.3575,
+      "step": 23369
+    },
+    {
+      "epoch": 0.9886623233776123,
+      "grad_norm": 0.14289572834968567,
+      "learning_rate": 0.001,
+      "loss": 1.5892,
+      "step": 23370
+    },
+    {
+      "epoch": 0.9887046281411287,
+      "grad_norm": 0.16080570220947266,
+      "learning_rate": 0.001,
+      "loss": 2.1795,
+      "step": 23371
+    },
+    {
+      "epoch": 0.9887469329046451,
+      "grad_norm": 0.185518279671669,
+      "learning_rate": 0.001,
+      "loss": 3.0211,
+      "step": 23372
+    },
+    {
+      "epoch": 0.9887892376681614,
+      "grad_norm": 0.13568751513957977,
+      "learning_rate": 0.001,
+      "loss": 2.4641,
+      "step": 23373
+    },
+    {
+      "epoch": 0.9888315424316778,
+      "grad_norm": 0.17552120983600616,
+      "learning_rate": 0.001,
+      "loss": 2.334,
+      "step": 23374
+    },
+    {
+      "epoch": 0.9888738471951942,
+      "grad_norm": 0.15184400975704193,
+      "learning_rate": 0.001,
+      "loss": 2.95,
+      "step": 23375
+    },
+    {
+      "epoch": 0.9889161519587105,
+      "grad_norm": 0.14975741505622864,
+      "learning_rate": 0.001,
+      "loss": 2.5531,
+      "step": 23376
+    },
+    {
+      "epoch": 0.9889584567222269,
+      "grad_norm": 0.12690399587154388,
+      "learning_rate": 0.001,
+      "loss": 1.8152,
+      "step": 23377
+    },
+    {
+      "epoch": 0.9890007614857433,
+      "grad_norm": 0.15675854682922363,
+      "learning_rate": 0.001,
+      "loss": 2.5597,
+      "step": 23378
+    },
+    {
+      "epoch": 0.9890430662492596,
+      "grad_norm": 0.15196675062179565,
+      "learning_rate": 0.001,
+      "loss": 2.6782,
+      "step": 23379
+    },
+    {
+      "epoch": 0.989085371012776,
+      "grad_norm": 0.7298043966293335,
+      "learning_rate": 0.001,
+      "loss": 2.3701,
+      "step": 23380
+    },
+    {
+      "epoch": 0.9891276757762925,
+      "grad_norm": 0.14933039247989655,
+      "learning_rate": 0.001,
+      "loss": 1.7975,
+      "step": 23381
+    },
+    {
+      "epoch": 0.9891699805398088,
+      "grad_norm": 0.15482641756534576,
+      "learning_rate": 0.001,
+      "loss": 2.4515,
+      "step": 23382
+    },
+    {
+      "epoch": 0.9892122853033252,
+      "grad_norm": 1.0311862230300903,
+      "learning_rate": 0.001,
+      "loss": 2.0674,
+      "step": 23383
+    },
+    {
+      "epoch": 0.9892545900668416,
+      "grad_norm": 0.13115628063678741,
+      "learning_rate": 0.001,
+      "loss": 2.2563,
+      "step": 23384
+    },
+    {
+      "epoch": 0.9892968948303579,
+      "grad_norm": 0.15161246061325073,
+      "learning_rate": 0.001,
+      "loss": 1.3509,
+      "step": 23385
+    },
+    {
+      "epoch": 0.9893391995938743,
+      "grad_norm": 0.14778055250644684,
+      "learning_rate": 0.001,
+      "loss": 2.6554,
+      "step": 23386
+    },
+    {
+      "epoch": 0.9893815043573907,
+      "grad_norm": 0.1839752495288849,
+      "learning_rate": 0.001,
+      "loss": 2.5877,
+      "step": 23387
+    },
+    {
+      "epoch": 0.989423809120907,
+      "grad_norm": 0.5433216094970703,
+      "learning_rate": 0.001,
+      "loss": 2.4707,
+      "step": 23388
+    },
+    {
+      "epoch": 0.9894661138844234,
+      "grad_norm": 0.19804272055625916,
+      "learning_rate": 0.001,
+      "loss": 1.9891,
+      "step": 23389
+    },
+    {
+      "epoch": 0.9895084186479397,
+      "grad_norm": 0.21847547590732574,
+      "learning_rate": 0.001,
+      "loss": 5.1899,
+      "step": 23390
+    },
+    {
+      "epoch": 0.9895507234114561,
+      "grad_norm": 0.8081578612327576,
+      "learning_rate": 0.001,
+      "loss": 2.3353,
+      "step": 23391
+    },
+    {
+      "epoch": 0.9895930281749725,
+      "grad_norm": 0.13023246824741364,
+      "learning_rate": 0.001,
+      "loss": 2.8709,
+      "step": 23392
+    },
+    {
+      "epoch": 0.9896353329384888,
+      "grad_norm": 0.16377229988574982,
+      "learning_rate": 0.001,
+      "loss": 1.5964,
+      "step": 23393
+    },
+    {
+      "epoch": 0.9896776377020052,
+      "grad_norm": 0.17955902218818665,
+      "learning_rate": 0.001,
+      "loss": 1.7785,
+      "step": 23394
+    },
+    {
+      "epoch": 0.9897199424655216,
+      "grad_norm": 0.18906940519809723,
+      "learning_rate": 0.001,
+      "loss": 2.842,
+      "step": 23395
+    },
+    {
+      "epoch": 0.9897622472290379,
+      "grad_norm": 0.20560045540332794,
+      "learning_rate": 0.001,
+      "loss": 1.9539,
+      "step": 23396
+    },
+    {
+      "epoch": 0.9898045519925543,
+      "grad_norm": 0.1924160122871399,
+      "learning_rate": 0.001,
+      "loss": 1.8572,
+      "step": 23397
+    },
+    {
+      "epoch": 0.9898468567560708,
+      "grad_norm": 0.1467902511358261,
+      "learning_rate": 0.001,
+      "loss": 1.7799,
+      "step": 23398
+    },
+    {
+      "epoch": 0.9898891615195871,
+      "grad_norm": 0.15988442301750183,
+      "learning_rate": 0.001,
+      "loss": 1.7468,
+      "step": 23399
+    },
+    {
+      "epoch": 0.9899314662831035,
+      "grad_norm": 0.16045702993869781,
+      "learning_rate": 0.001,
+      "loss": 2.8367,
+      "step": 23400
+    },
+    {
+      "epoch": 0.9899737710466199,
+      "grad_norm": 0.16333599388599396,
+      "learning_rate": 0.001,
+      "loss": 1.6981,
+      "step": 23401
+    },
+    {
+      "epoch": 0.9900160758101362,
+      "grad_norm": 0.1779310405254364,
+      "learning_rate": 0.001,
+      "loss": 1.8536,
+      "step": 23402
+    },
+    {
+      "epoch": 0.9900583805736526,
+      "grad_norm": 0.8571863770484924,
+      "learning_rate": 0.001,
+      "loss": 2.8216,
+      "step": 23403
+    },
+    {
+      "epoch": 0.990100685337169,
+      "grad_norm": 0.18227769434452057,
+      "learning_rate": 0.001,
+      "loss": 1.8207,
+      "step": 23404
+    },
+    {
+      "epoch": 0.9901429901006853,
+      "grad_norm": 0.1805211752653122,
+      "learning_rate": 0.001,
+      "loss": 2.2027,
+      "step": 23405
+    },
+    {
+      "epoch": 0.9901852948642017,
+      "grad_norm": 1.146898865699768,
+      "learning_rate": 0.001,
+      "loss": 2.0096,
+      "step": 23406
+    },
+    {
+      "epoch": 0.9902275996277181,
+      "grad_norm": 0.14902471005916595,
+      "learning_rate": 0.001,
+      "loss": 1.337,
+      "step": 23407
+    },
+    {
+      "epoch": 0.9902699043912344,
+      "grad_norm": 0.17255893349647522,
+      "learning_rate": 0.001,
+      "loss": 2.4503,
+      "step": 23408
+    },
+    {
+      "epoch": 0.9903122091547508,
+      "grad_norm": 0.18396355211734772,
+      "learning_rate": 0.001,
+      "loss": 2.3969,
+      "step": 23409
+    },
+    {
+      "epoch": 0.9903545139182672,
+      "grad_norm": 0.17676624655723572,
+      "learning_rate": 0.001,
+      "loss": 1.7923,
+      "step": 23410
+    },
+    {
+      "epoch": 0.9903968186817835,
+      "grad_norm": 0.19076810777187347,
+      "learning_rate": 0.001,
+      "loss": 1.9227,
+      "step": 23411
+    },
+    {
+      "epoch": 0.9904391234452999,
+      "grad_norm": 0.15760193765163422,
+      "learning_rate": 0.001,
+      "loss": 2.9747,
+      "step": 23412
+    },
+    {
+      "epoch": 0.9904814282088164,
+      "grad_norm": 0.20325231552124023,
+      "learning_rate": 0.001,
+      "loss": 1.7755,
+      "step": 23413
+    },
+    {
+      "epoch": 0.9905237329723326,
+      "grad_norm": 0.17613425850868225,
+      "learning_rate": 0.001,
+      "loss": 2.0028,
+      "step": 23414
+    },
+    {
+      "epoch": 0.9905660377358491,
+      "grad_norm": 0.20244133472442627,
+      "learning_rate": 0.001,
+      "loss": 2.5725,
+      "step": 23415
+    },
+    {
+      "epoch": 0.9906083424993655,
+      "grad_norm": 0.18409040570259094,
+      "learning_rate": 0.001,
+      "loss": 3.6709,
+      "step": 23416
+    },
+    {
+      "epoch": 0.9906506472628818,
+      "grad_norm": 0.15753605961799622,
+      "learning_rate": 0.001,
+      "loss": 2.0971,
+      "step": 23417
+    },
+    {
+      "epoch": 0.9906929520263982,
+      "grad_norm": 0.14398153126239777,
+      "learning_rate": 0.001,
+      "loss": 2.2371,
+      "step": 23418
+    },
+    {
+      "epoch": 0.9907352567899146,
+      "grad_norm": 0.1526634693145752,
+      "learning_rate": 0.001,
+      "loss": 2.2506,
+      "step": 23419
+    },
+    {
+      "epoch": 0.9907775615534309,
+      "grad_norm": 0.12150632590055466,
+      "learning_rate": 0.001,
+      "loss": 2.0968,
+      "step": 23420
+    },
+    {
+      "epoch": 0.9908198663169473,
+      "grad_norm": 0.7939785122871399,
+      "learning_rate": 0.001,
+      "loss": 2.479,
+      "step": 23421
+    },
+    {
+      "epoch": 0.9908621710804637,
+      "grad_norm": 0.16594654321670532,
+      "learning_rate": 0.001,
+      "loss": 1.6048,
+      "step": 23422
+    },
+    {
+      "epoch": 0.99090447584398,
+      "grad_norm": 0.1510103940963745,
+      "learning_rate": 0.001,
+      "loss": 2.7858,
+      "step": 23423
+    },
+    {
+      "epoch": 0.9909467806074964,
+      "grad_norm": 0.23928913474082947,
+      "learning_rate": 0.001,
+      "loss": 1.7646,
+      "step": 23424
+    },
+    {
+      "epoch": 0.9909890853710128,
+      "grad_norm": 0.14759401977062225,
+      "learning_rate": 0.001,
+      "loss": 1.6538,
+      "step": 23425
+    },
+    {
+      "epoch": 0.9910313901345291,
+      "grad_norm": 0.19283150136470795,
+      "learning_rate": 0.001,
+      "loss": 2.6298,
+      "step": 23426
+    },
+    {
+      "epoch": 0.9910736948980455,
+      "grad_norm": 3.1644039154052734,
+      "learning_rate": 0.001,
+      "loss": 2.4429,
+      "step": 23427
+    },
+    {
+      "epoch": 0.9911159996615619,
+      "grad_norm": 0.15195727348327637,
+      "learning_rate": 0.001,
+      "loss": 2.3545,
+      "step": 23428
+    },
+    {
+      "epoch": 0.9911583044250782,
+      "grad_norm": 0.13520923256874084,
+      "learning_rate": 0.001,
+      "loss": 2.0637,
+      "step": 23429
+    },
+    {
+      "epoch": 0.9912006091885947,
+      "grad_norm": 0.1306263655424118,
+      "learning_rate": 0.001,
+      "loss": 1.5021,
+      "step": 23430
+    },
+    {
+      "epoch": 0.9912429139521111,
+      "grad_norm": 0.16827377676963806,
+      "learning_rate": 0.001,
+      "loss": 3.3265,
+      "step": 23431
+    },
+    {
+      "epoch": 0.9912852187156274,
+      "grad_norm": 0.14732570946216583,
+      "learning_rate": 0.001,
+      "loss": 2.0172,
+      "step": 23432
+    },
+    {
+      "epoch": 0.9913275234791438,
+      "grad_norm": 3.2398881912231445,
+      "learning_rate": 0.001,
+      "loss": 2.523,
+      "step": 23433
+    },
+    {
+      "epoch": 0.9913698282426602,
+      "grad_norm": 0.14063461124897003,
+      "learning_rate": 0.001,
+      "loss": 1.9463,
+      "step": 23434
+    },
+    {
+      "epoch": 0.9914121330061765,
+      "grad_norm": 0.2017151415348053,
+      "learning_rate": 0.001,
+      "loss": 1.9935,
+      "step": 23435
+    },
+    {
+      "epoch": 0.9914544377696929,
+      "grad_norm": 0.1547040343284607,
+      "learning_rate": 0.001,
+      "loss": 1.7171,
+      "step": 23436
+    },
+    {
+      "epoch": 0.9914967425332092,
+      "grad_norm": 0.1772545576095581,
+      "learning_rate": 0.001,
+      "loss": 2.6763,
+      "step": 23437
+    },
+    {
+      "epoch": 0.9915390472967256,
+      "grad_norm": 0.158152773976326,
+      "learning_rate": 0.001,
+      "loss": 1.7772,
+      "step": 23438
+    },
+    {
+      "epoch": 0.991581352060242,
+      "grad_norm": 0.17712469398975372,
+      "learning_rate": 0.001,
+      "loss": 1.6187,
+      "step": 23439
+    },
+    {
+      "epoch": 0.9916236568237583,
+      "grad_norm": 0.16980746388435364,
+      "learning_rate": 0.001,
+      "loss": 2.7735,
+      "step": 23440
+    },
+    {
+      "epoch": 0.9916659615872747,
+      "grad_norm": 1.1700719594955444,
+      "learning_rate": 0.001,
+      "loss": 2.1555,
+      "step": 23441
+    },
+    {
+      "epoch": 0.9917082663507911,
+      "grad_norm": 0.1690703183412552,
+      "learning_rate": 0.001,
+      "loss": 1.6741,
+      "step": 23442
+    },
+    {
+      "epoch": 0.9917505711143074,
+      "grad_norm": 0.15942858159542084,
+      "learning_rate": 0.001,
+      "loss": 1.7488,
+      "step": 23443
+    },
+    {
+      "epoch": 0.9917928758778238,
+      "grad_norm": 0.15770910680294037,
+      "learning_rate": 0.001,
+      "loss": 3.3523,
+      "step": 23444
+    },
+    {
+      "epoch": 0.9918351806413402,
+      "grad_norm": 11.168841361999512,
+      "learning_rate": 0.001,
+      "loss": 2.2278,
+      "step": 23445
+    },
+    {
+      "epoch": 0.9918774854048565,
+      "grad_norm": 0.173135906457901,
+      "learning_rate": 0.001,
+      "loss": 2.5919,
+      "step": 23446
+    },
+    {
+      "epoch": 0.991919790168373,
+      "grad_norm": 0.13830503821372986,
+      "learning_rate": 0.001,
+      "loss": 2.0946,
+      "step": 23447
+    },
+    {
+      "epoch": 0.9919620949318894,
+      "grad_norm": 0.1732984036207199,
+      "learning_rate": 0.001,
+      "loss": 1.8609,
+      "step": 23448
+    },
+    {
+      "epoch": 0.9920043996954057,
+      "grad_norm": 0.1577487289905548,
+      "learning_rate": 0.001,
+      "loss": 2.0619,
+      "step": 23449
+    },
+    {
+      "epoch": 0.9920467044589221,
+      "grad_norm": 0.9779529571533203,
+      "learning_rate": 0.001,
+      "loss": 1.9558,
+      "step": 23450
+    },
+    {
+      "epoch": 0.9920890092224385,
+      "grad_norm": 0.21312867105007172,
+      "learning_rate": 0.001,
+      "loss": 2.2838,
+      "step": 23451
+    },
+    {
+      "epoch": 0.9921313139859548,
+      "grad_norm": 0.18126481771469116,
+      "learning_rate": 0.001,
+      "loss": 1.9558,
+      "step": 23452
+    },
+    {
+      "epoch": 0.9921736187494712,
+      "grad_norm": 0.9032267928123474,
+      "learning_rate": 0.001,
+      "loss": 2.3783,
+      "step": 23453
+    },
+    {
+      "epoch": 0.9922159235129876,
+      "grad_norm": 0.13526488840579987,
+      "learning_rate": 0.001,
+      "loss": 2.0193,
+      "step": 23454
+    },
+    {
+      "epoch": 0.9922582282765039,
+      "grad_norm": 0.16340215504169464,
+      "learning_rate": 0.001,
+      "loss": 2.604,
+      "step": 23455
+    },
+    {
+      "epoch": 0.9923005330400203,
+      "grad_norm": 0.14608405530452728,
+      "learning_rate": 0.001,
+      "loss": 3.0248,
+      "step": 23456
+    },
+    {
+      "epoch": 0.9923428378035367,
+      "grad_norm": 0.15734529495239258,
+      "learning_rate": 0.001,
+      "loss": 1.6765,
+      "step": 23457
+    },
+    {
+      "epoch": 0.992385142567053,
+      "grad_norm": 0.14332926273345947,
+      "learning_rate": 0.001,
+      "loss": 2.2598,
+      "step": 23458
+    },
+    {
+      "epoch": 0.9924274473305694,
+      "grad_norm": 1.3368622064590454,
+      "learning_rate": 0.001,
+      "loss": 3.2208,
+      "step": 23459
+    },
+    {
+      "epoch": 0.9924697520940858,
+      "grad_norm": 0.16903036832809448,
+      "learning_rate": 0.001,
+      "loss": 1.5743,
+      "step": 23460
+    },
+    {
+      "epoch": 0.9925120568576021,
+      "grad_norm": 0.1450778692960739,
+      "learning_rate": 0.001,
+      "loss": 2.1063,
+      "step": 23461
+    },
+    {
+      "epoch": 0.9925543616211185,
+      "grad_norm": 0.24200314283370972,
+      "learning_rate": 0.001,
+      "loss": 2.7203,
+      "step": 23462
+    },
+    {
+      "epoch": 0.992596666384635,
+      "grad_norm": 0.1364940106868744,
+      "learning_rate": 0.001,
+      "loss": 2.2594,
+      "step": 23463
+    },
+    {
+      "epoch": 0.9926389711481513,
+      "grad_norm": 0.15272387862205505,
+      "learning_rate": 0.001,
+      "loss": 1.5758,
+      "step": 23464
+    },
+    {
+      "epoch": 0.9926812759116677,
+      "grad_norm": 0.1937706619501114,
+      "learning_rate": 0.001,
+      "loss": 2.4084,
+      "step": 23465
+    },
+    {
+      "epoch": 0.9927235806751841,
+      "grad_norm": 0.1765604466199875,
+      "learning_rate": 0.001,
+      "loss": 1.7184,
+      "step": 23466
+    },
+    {
+      "epoch": 0.9927658854387004,
+      "grad_norm": 0.17739205062389374,
+      "learning_rate": 0.001,
+      "loss": 1.9615,
+      "step": 23467
+    },
+    {
+      "epoch": 0.9928081902022168,
+      "grad_norm": 1.2467105388641357,
+      "learning_rate": 0.001,
+      "loss": 1.6635,
+      "step": 23468
+    },
+    {
+      "epoch": 0.9928504949657332,
+      "grad_norm": 1.0034493207931519,
+      "learning_rate": 0.001,
+      "loss": 2.3672,
+      "step": 23469
+    },
+    {
+      "epoch": 0.9928927997292495,
+      "grad_norm": 0.19060944020748138,
+      "learning_rate": 0.001,
+      "loss": 2.5745,
+      "step": 23470
+    },
+    {
+      "epoch": 0.9929351044927659,
+      "grad_norm": 0.18545667827129364,
+      "learning_rate": 0.001,
+      "loss": 2.2075,
+      "step": 23471
+    },
+    {
+      "epoch": 0.9929774092562823,
+      "grad_norm": 0.1679558902978897,
+      "learning_rate": 0.001,
+      "loss": 1.7822,
+      "step": 23472
+    },
+    {
+      "epoch": 0.9930197140197986,
+      "grad_norm": 0.21493469178676605,
+      "learning_rate": 0.001,
+      "loss": 3.6993,
+      "step": 23473
+    },
+    {
+      "epoch": 0.993062018783315,
+      "grad_norm": 0.21583551168441772,
+      "learning_rate": 0.001,
+      "loss": 2.2144,
+      "step": 23474
+    },
+    {
+      "epoch": 0.9931043235468314,
+      "grad_norm": 0.1545204520225525,
+      "learning_rate": 0.001,
+      "loss": 1.8569,
+      "step": 23475
+    },
+    {
+      "epoch": 0.9931466283103477,
+      "grad_norm": 0.17547298967838287,
+      "learning_rate": 0.001,
+      "loss": 1.6119,
+      "step": 23476
+    },
+    {
+      "epoch": 0.9931889330738641,
+      "grad_norm": 0.18335486948490143,
+      "learning_rate": 0.001,
+      "loss": 1.8375,
+      "step": 23477
+    },
+    {
+      "epoch": 0.9932312378373805,
+      "grad_norm": 0.17604441940784454,
+      "learning_rate": 0.001,
+      "loss": 1.7185,
+      "step": 23478
+    },
+    {
+      "epoch": 0.9932735426008968,
+      "grad_norm": 0.1612115502357483,
+      "learning_rate": 0.001,
+      "loss": 1.7173,
+      "step": 23479
+    },
+    {
+      "epoch": 0.9933158473644133,
+      "grad_norm": 0.17903564870357513,
+      "learning_rate": 0.001,
+      "loss": 2.2194,
+      "step": 23480
+    },
+    {
+      "epoch": 0.9933581521279296,
+      "grad_norm": 0.15683114528656006,
+      "learning_rate": 0.001,
+      "loss": 1.9522,
+      "step": 23481
+    },
+    {
+      "epoch": 0.993400456891446,
+      "grad_norm": 0.19979719817638397,
+      "learning_rate": 0.001,
+      "loss": 2.8421,
+      "step": 23482
+    },
+    {
+      "epoch": 0.9934427616549624,
+      "grad_norm": 0.16343528032302856,
+      "learning_rate": 0.001,
+      "loss": 2.3331,
+      "step": 23483
+    },
+    {
+      "epoch": 0.9934850664184787,
+      "grad_norm": 0.14997512102127075,
+      "learning_rate": 0.001,
+      "loss": 1.9094,
+      "step": 23484
+    },
+    {
+      "epoch": 0.9935273711819951,
+      "grad_norm": 0.14654596149921417,
+      "learning_rate": 0.001,
+      "loss": 2.325,
+      "step": 23485
+    },
+    {
+      "epoch": 0.9935696759455115,
+      "grad_norm": 0.15524841845035553,
+      "learning_rate": 0.001,
+      "loss": 2.0226,
+      "step": 23486
+    },
+    {
+      "epoch": 0.9936119807090278,
+      "grad_norm": 0.1682468056678772,
+      "learning_rate": 0.001,
+      "loss": 1.9113,
+      "step": 23487
+    },
+    {
+      "epoch": 0.9936542854725442,
+      "grad_norm": 0.15577398240566254,
+      "learning_rate": 0.001,
+      "loss": 2.6625,
+      "step": 23488
+    },
+    {
+      "epoch": 0.9936965902360606,
+      "grad_norm": 0.1386529803276062,
+      "learning_rate": 0.001,
+      "loss": 1.7642,
+      "step": 23489
+    },
+    {
+      "epoch": 0.9937388949995769,
+      "grad_norm": 0.1561625450849533,
+      "learning_rate": 0.001,
+      "loss": 2.7162,
+      "step": 23490
+    },
+    {
+      "epoch": 0.9937811997630933,
+      "grad_norm": 0.22317662835121155,
+      "learning_rate": 0.001,
+      "loss": 2.0305,
+      "step": 23491
+    },
+    {
+      "epoch": 0.9938235045266097,
+      "grad_norm": 0.1401653289794922,
+      "learning_rate": 0.001,
+      "loss": 2.2692,
+      "step": 23492
+    },
+    {
+      "epoch": 0.993865809290126,
+      "grad_norm": 0.14323166012763977,
+      "learning_rate": 0.001,
+      "loss": 1.724,
+      "step": 23493
+    },
+    {
+      "epoch": 0.9939081140536424,
+      "grad_norm": 0.1282302439212799,
+      "learning_rate": 0.001,
+      "loss": 2.036,
+      "step": 23494
+    },
+    {
+      "epoch": 0.9939504188171588,
+      "grad_norm": 0.1263008862733841,
+      "learning_rate": 0.001,
+      "loss": 1.3955,
+      "step": 23495
+    },
+    {
+      "epoch": 0.9939927235806751,
+      "grad_norm": 0.13142234086990356,
+      "learning_rate": 0.001,
+      "loss": 2.1095,
+      "step": 23496
+    },
+    {
+      "epoch": 0.9940350283441916,
+      "grad_norm": 0.17697520554065704,
+      "learning_rate": 0.001,
+      "loss": 2.3401,
+      "step": 23497
+    },
+    {
+      "epoch": 0.994077333107708,
+      "grad_norm": 1.2804735898971558,
+      "learning_rate": 0.001,
+      "loss": 2.5458,
+      "step": 23498
+    },
+    {
+      "epoch": 0.9941196378712243,
+      "grad_norm": 0.24874600768089294,
+      "learning_rate": 0.001,
+      "loss": 2.6729,
+      "step": 23499
+    },
+    {
+      "epoch": 0.9941619426347407,
+      "grad_norm": 0.2032213658094406,
+      "learning_rate": 0.001,
+      "loss": 2.7131,
+      "step": 23500
+    },
+    {
+      "epoch": 0.9942042473982571,
+      "grad_norm": 0.267898827791214,
+      "learning_rate": 0.001,
+      "loss": 2.2249,
+      "step": 23501
+    },
+    {
+      "epoch": 0.9942465521617734,
+      "grad_norm": 0.1407955288887024,
+      "learning_rate": 0.001,
+      "loss": 2.842,
+      "step": 23502
+    },
+    {
+      "epoch": 0.9942888569252898,
+      "grad_norm": 10.123157501220703,
+      "learning_rate": 0.001,
+      "loss": 1.9959,
+      "step": 23503
+    },
+    {
+      "epoch": 0.9943311616888062,
+      "grad_norm": 0.1459282636642456,
+      "learning_rate": 0.001,
+      "loss": 2.1619,
+      "step": 23504
+    },
+    {
+      "epoch": 0.9943734664523225,
+      "grad_norm": 0.17329807579517365,
+      "learning_rate": 0.001,
+      "loss": 2.1785,
+      "step": 23505
+    },
+    {
+      "epoch": 0.9944157712158389,
+      "grad_norm": 0.11436944454908371,
+      "learning_rate": 0.001,
+      "loss": 2.0212,
+      "step": 23506
+    },
+    {
+      "epoch": 0.9944580759793553,
+      "grad_norm": 0.15068282186985016,
+      "learning_rate": 0.001,
+      "loss": 1.7319,
+      "step": 23507
+    },
+    {
+      "epoch": 0.9945003807428716,
+      "grad_norm": 0.14463482797145844,
+      "learning_rate": 0.001,
+      "loss": 1.932,
+      "step": 23508
+    },
+    {
+      "epoch": 0.994542685506388,
+      "grad_norm": 0.1417880356311798,
+      "learning_rate": 0.001,
+      "loss": 2.3686,
+      "step": 23509
+    },
+    {
+      "epoch": 0.9945849902699044,
+      "grad_norm": 2.411515951156616,
+      "learning_rate": 0.001,
+      "loss": 1.9726,
+      "step": 23510
+    },
+    {
+      "epoch": 0.9946272950334207,
+      "grad_norm": 0.15016627311706543,
+      "learning_rate": 0.001,
+      "loss": 2.7317,
+      "step": 23511
+    },
+    {
+      "epoch": 0.9946695997969371,
+      "grad_norm": 0.16307124495506287,
+      "learning_rate": 0.001,
+      "loss": 2.3945,
+      "step": 23512
+    },
+    {
+      "epoch": 0.9947119045604536,
+      "grad_norm": 0.2472067028284073,
+      "learning_rate": 0.001,
+      "loss": 1.323,
+      "step": 23513
+    },
+    {
+      "epoch": 0.9947542093239699,
+      "grad_norm": 0.14261025190353394,
+      "learning_rate": 0.001,
+      "loss": 1.6545,
+      "step": 23514
+    },
+    {
+      "epoch": 0.9947965140874863,
+      "grad_norm": 0.21985666453838348,
+      "learning_rate": 0.001,
+      "loss": 1.6186,
+      "step": 23515
+    },
+    {
+      "epoch": 0.9948388188510027,
+      "grad_norm": 0.18958121538162231,
+      "learning_rate": 0.001,
+      "loss": 3.1224,
+      "step": 23516
+    },
+    {
+      "epoch": 0.994881123614519,
+      "grad_norm": 0.18746091425418854,
+      "learning_rate": 0.001,
+      "loss": 1.7352,
+      "step": 23517
+    },
+    {
+      "epoch": 0.9949234283780354,
+      "grad_norm": 0.3156006932258606,
+      "learning_rate": 0.001,
+      "loss": 3.373,
+      "step": 23518
+    },
+    {
+      "epoch": 0.9949657331415518,
+      "grad_norm": 0.1298127919435501,
+      "learning_rate": 0.001,
+      "loss": 2.0672,
+      "step": 23519
+    },
+    {
+      "epoch": 0.9950080379050681,
+      "grad_norm": 0.12908710539340973,
+      "learning_rate": 0.001,
+      "loss": 1.835,
+      "step": 23520
+    },
+    {
+      "epoch": 0.9950503426685845,
+      "grad_norm": 0.2133929282426834,
+      "learning_rate": 0.001,
+      "loss": 3.1512,
+      "step": 23521
+    },
+    {
+      "epoch": 0.9950926474321009,
+      "grad_norm": 0.145905002951622,
+      "learning_rate": 0.001,
+      "loss": 1.3302,
+      "step": 23522
+    },
+    {
+      "epoch": 0.9951349521956172,
+      "grad_norm": 0.19927392899990082,
+      "learning_rate": 0.001,
+      "loss": 3.0666,
+      "step": 23523
+    },
+    {
+      "epoch": 0.9951772569591336,
+      "grad_norm": 0.1459832340478897,
+      "learning_rate": 0.001,
+      "loss": 2.3249,
+      "step": 23524
+    },
+    {
+      "epoch": 0.99521956172265,
+      "grad_norm": 0.19735519587993622,
+      "learning_rate": 0.001,
+      "loss": 1.8252,
+      "step": 23525
+    },
+    {
+      "epoch": 0.9952618664861663,
+      "grad_norm": 0.15133242309093475,
+      "learning_rate": 0.001,
+      "loss": 1.5066,
+      "step": 23526
+    },
+    {
+      "epoch": 0.9953041712496827,
+      "grad_norm": 7.522138595581055,
+      "learning_rate": 0.001,
+      "loss": 2.1434,
+      "step": 23527
+    },
+    {
+      "epoch": 0.995346476013199,
+      "grad_norm": 0.1376638412475586,
+      "learning_rate": 0.001,
+      "loss": 1.7716,
+      "step": 23528
+    },
+    {
+      "epoch": 0.9953887807767154,
+      "grad_norm": 1.5856484174728394,
+      "learning_rate": 0.001,
+      "loss": 1.7964,
+      "step": 23529
+    },
+    {
+      "epoch": 0.9954310855402319,
+      "grad_norm": 0.16695480048656464,
+      "learning_rate": 0.001,
+      "loss": 1.9568,
+      "step": 23530
+    },
+    {
+      "epoch": 0.9954733903037482,
+      "grad_norm": 0.6410706043243408,
+      "learning_rate": 0.001,
+      "loss": 2.1194,
+      "step": 23531
+    },
+    {
+      "epoch": 0.9955156950672646,
+      "grad_norm": 0.31505700945854187,
+      "learning_rate": 0.001,
+      "loss": 1.8234,
+      "step": 23532
+    },
+    {
+      "epoch": 0.995557999830781,
+      "grad_norm": 0.1771719753742218,
+      "learning_rate": 0.001,
+      "loss": 2.5342,
+      "step": 23533
+    },
+    {
+      "epoch": 0.9956003045942973,
+      "grad_norm": 0.18417403101921082,
+      "learning_rate": 0.001,
+      "loss": 2.5644,
+      "step": 23534
+    },
+    {
+      "epoch": 0.9956426093578137,
+      "grad_norm": 0.15871447324752808,
+      "learning_rate": 0.001,
+      "loss": 2.7864,
+      "step": 23535
+    },
+    {
+      "epoch": 0.9956849141213301,
+      "grad_norm": 0.2665135860443115,
+      "learning_rate": 0.001,
+      "loss": 2.847,
+      "step": 23536
+    },
+    {
+      "epoch": 0.9957272188848464,
+      "grad_norm": 0.7563045620918274,
+      "learning_rate": 0.001,
+      "loss": 3.1899,
+      "step": 23537
+    },
+    {
+      "epoch": 0.9957695236483628,
+      "grad_norm": 0.1422766149044037,
+      "learning_rate": 0.001,
+      "loss": 2.0713,
+      "step": 23538
+    },
+    {
+      "epoch": 0.9958118284118792,
+      "grad_norm": 0.14153209328651428,
+      "learning_rate": 0.001,
+      "loss": 1.7366,
+      "step": 23539
+    },
+    {
+      "epoch": 0.9958541331753955,
+      "grad_norm": 0.1573881059885025,
+      "learning_rate": 0.001,
+      "loss": 1.9069,
+      "step": 23540
+    },
+    {
+      "epoch": 0.9958964379389119,
+      "grad_norm": 39.94224166870117,
+      "learning_rate": 0.001,
+      "loss": 2.0561,
+      "step": 23541
+    },
+    {
+      "epoch": 0.9959387427024283,
+      "grad_norm": 0.1591683328151703,
+      "learning_rate": 0.001,
+      "loss": 3.2608,
+      "step": 23542
+    },
+    {
+      "epoch": 0.9959810474659446,
+      "grad_norm": 0.19343800842761993,
+      "learning_rate": 0.001,
+      "loss": 3.1865,
+      "step": 23543
+    },
+    {
+      "epoch": 0.996023352229461,
+      "grad_norm": 0.15812444686889648,
+      "learning_rate": 0.001,
+      "loss": 2.2566,
+      "step": 23544
+    },
+    {
+      "epoch": 0.9960656569929774,
+      "grad_norm": 0.3023439049720764,
+      "learning_rate": 0.001,
+      "loss": 2.1273,
+      "step": 23545
+    },
+    {
+      "epoch": 0.9961079617564937,
+      "grad_norm": 0.14426670968532562,
+      "learning_rate": 0.001,
+      "loss": 1.5217,
+      "step": 23546
+    },
+    {
+      "epoch": 0.9961502665200102,
+      "grad_norm": 0.5261450409889221,
+      "learning_rate": 0.001,
+      "loss": 1.8896,
+      "step": 23547
+    },
+    {
+      "epoch": 0.9961925712835266,
+      "grad_norm": 0.1593768447637558,
+      "learning_rate": 0.001,
+      "loss": 1.302,
+      "step": 23548
+    },
+    {
+      "epoch": 0.9962348760470429,
+      "grad_norm": 0.19295907020568848,
+      "learning_rate": 0.001,
+      "loss": 2.4438,
+      "step": 23549
+    },
+    {
+      "epoch": 0.9962771808105593,
+      "grad_norm": 0.1521875560283661,
+      "learning_rate": 0.001,
+      "loss": 1.3707,
+      "step": 23550
+    },
+    {
+      "epoch": 0.9963194855740757,
+      "grad_norm": 0.1482807993888855,
+      "learning_rate": 0.001,
+      "loss": 2.3517,
+      "step": 23551
+    },
+    {
+      "epoch": 0.996361790337592,
+      "grad_norm": 0.1328219175338745,
+      "learning_rate": 0.001,
+      "loss": 1.6607,
+      "step": 23552
+    },
+    {
+      "epoch": 0.9964040951011084,
+      "grad_norm": 0.13568271696567535,
+      "learning_rate": 0.001,
+      "loss": 2.0808,
+      "step": 23553
+    },
+    {
+      "epoch": 0.9964463998646248,
+      "grad_norm": 0.17867201566696167,
+      "learning_rate": 0.001,
+      "loss": 1.8838,
+      "step": 23554
+    },
+    {
+      "epoch": 0.9964887046281411,
+      "grad_norm": 0.14420682191848755,
+      "learning_rate": 0.001,
+      "loss": 2.1963,
+      "step": 23555
+    },
+    {
+      "epoch": 0.9965310093916575,
+      "grad_norm": 0.19156771898269653,
+      "learning_rate": 0.001,
+      "loss": 2.6953,
+      "step": 23556
+    },
+    {
+      "epoch": 0.9965733141551739,
+      "grad_norm": 0.13624778389930725,
+      "learning_rate": 0.001,
+      "loss": 2.076,
+      "step": 23557
+    },
+    {
+      "epoch": 0.9966156189186902,
+      "grad_norm": 0.15911436080932617,
+      "learning_rate": 0.001,
+      "loss": 2.2559,
+      "step": 23558
+    },
+    {
+      "epoch": 0.9966579236822066,
+      "grad_norm": 0.16328518092632294,
+      "learning_rate": 0.001,
+      "loss": 1.7107,
+      "step": 23559
+    },
+    {
+      "epoch": 0.996700228445723,
+      "grad_norm": 0.20304375886917114,
+      "learning_rate": 0.001,
+      "loss": 2.0468,
+      "step": 23560
+    },
+    {
+      "epoch": 0.9967425332092393,
+      "grad_norm": 0.16803617775440216,
+      "learning_rate": 0.001,
+      "loss": 2.2345,
+      "step": 23561
+    },
+    {
+      "epoch": 0.9967848379727557,
+      "grad_norm": 0.13952644169330597,
+      "learning_rate": 0.001,
+      "loss": 2.2767,
+      "step": 23562
+    },
+    {
+      "epoch": 0.9968271427362722,
+      "grad_norm": 0.15917642414569855,
+      "learning_rate": 0.001,
+      "loss": 1.8884,
+      "step": 23563
+    },
+    {
+      "epoch": 0.9968694474997885,
+      "grad_norm": 0.1422685831785202,
+      "learning_rate": 0.001,
+      "loss": 1.6789,
+      "step": 23564
+    },
+    {
+      "epoch": 0.9969117522633049,
+      "grad_norm": 0.9447866082191467,
+      "learning_rate": 0.001,
+      "loss": 2.6736,
+      "step": 23565
+    },
+    {
+      "epoch": 0.9969540570268213,
+      "grad_norm": 0.15833914279937744,
+      "learning_rate": 0.001,
+      "loss": 2.3225,
+      "step": 23566
+    },
+    {
+      "epoch": 0.9969963617903376,
+      "grad_norm": 0.15113773941993713,
+      "learning_rate": 0.001,
+      "loss": 2.1877,
+      "step": 23567
+    },
+    {
+      "epoch": 0.997038666553854,
+      "grad_norm": 0.17661526799201965,
+      "learning_rate": 0.001,
+      "loss": 2.0605,
+      "step": 23568
+    },
+    {
+      "epoch": 0.9970809713173704,
+      "grad_norm": 0.2037663608789444,
+      "learning_rate": 0.001,
+      "loss": 2.095,
+      "step": 23569
+    },
+    {
+      "epoch": 0.9971232760808867,
+      "grad_norm": 0.40272435545921326,
+      "learning_rate": 0.001,
+      "loss": 1.8401,
+      "step": 23570
+    },
+    {
+      "epoch": 0.9971655808444031,
+      "grad_norm": 1.8662751913070679,
+      "learning_rate": 0.001,
+      "loss": 3.3036,
+      "step": 23571
+    },
+    {
+      "epoch": 0.9972078856079194,
+      "grad_norm": 0.14151303470134735,
+      "learning_rate": 0.001,
+      "loss": 2.6741,
+      "step": 23572
+    },
+    {
+      "epoch": 0.9972501903714358,
+      "grad_norm": 0.17152297496795654,
+      "learning_rate": 0.001,
+      "loss": 2.5857,
+      "step": 23573
+    },
+    {
+      "epoch": 0.9972924951349522,
+      "grad_norm": 1.016983151435852,
+      "learning_rate": 0.001,
+      "loss": 1.5302,
+      "step": 23574
+    },
+    {
+      "epoch": 0.9973347998984685,
+      "grad_norm": 0.16691197454929352,
+      "learning_rate": 0.001,
+      "loss": 2.3803,
+      "step": 23575
+    },
+    {
+      "epoch": 0.9973771046619849,
+      "grad_norm": 0.16801585257053375,
+      "learning_rate": 0.001,
+      "loss": 1.581,
+      "step": 23576
+    },
+    {
+      "epoch": 0.9974194094255013,
+      "grad_norm": 0.20410461723804474,
+      "learning_rate": 0.001,
+      "loss": 1.7048,
+      "step": 23577
+    },
+    {
+      "epoch": 0.9974617141890176,
+      "grad_norm": 0.17868375778198242,
+      "learning_rate": 0.001,
+      "loss": 2.3364,
+      "step": 23578
+    },
+    {
+      "epoch": 0.997504018952534,
+      "grad_norm": 0.1568222939968109,
+      "learning_rate": 0.001,
+      "loss": 1.8333,
+      "step": 23579
+    },
+    {
+      "epoch": 0.9975463237160505,
+      "grad_norm": 0.1552361100912094,
+      "learning_rate": 0.001,
+      "loss": 1.9154,
+      "step": 23580
+    },
+    {
+      "epoch": 0.9975886284795668,
+      "grad_norm": 0.15056028962135315,
+      "learning_rate": 0.001,
+      "loss": 2.2219,
+      "step": 23581
+    },
+    {
+      "epoch": 0.9976309332430832,
+      "grad_norm": 0.17635402083396912,
+      "learning_rate": 0.001,
+      "loss": 1.6218,
+      "step": 23582
+    },
+    {
+      "epoch": 0.9976732380065996,
+      "grad_norm": 0.273888498544693,
+      "learning_rate": 0.001,
+      "loss": 3.4827,
+      "step": 23583
+    },
+    {
+      "epoch": 0.9977155427701159,
+      "grad_norm": 0.14739017188549042,
+      "learning_rate": 0.001,
+      "loss": 2.1994,
+      "step": 23584
+    },
+    {
+      "epoch": 0.9977578475336323,
+      "grad_norm": 0.18939383327960968,
+      "learning_rate": 0.001,
+      "loss": 2.3165,
+      "step": 23585
+    },
+    {
+      "epoch": 0.9978001522971487,
+      "grad_norm": 0.15688124299049377,
+      "learning_rate": 0.001,
+      "loss": 1.771,
+      "step": 23586
+    },
+    {
+      "epoch": 0.997842457060665,
+      "grad_norm": 0.15662074089050293,
+      "learning_rate": 0.001,
+      "loss": 2.1358,
+      "step": 23587
+    },
+    {
+      "epoch": 0.9978847618241814,
+      "grad_norm": 0.15011928975582123,
+      "learning_rate": 0.001,
+      "loss": 2.084,
+      "step": 23588
+    },
+    {
+      "epoch": 0.9979270665876978,
+      "grad_norm": 0.1399097889661789,
+      "learning_rate": 0.001,
+      "loss": 3.6885,
+      "step": 23589
+    },
+    {
+      "epoch": 0.9979693713512141,
+      "grad_norm": 0.1603403389453888,
+      "learning_rate": 0.001,
+      "loss": 2.0335,
+      "step": 23590
+    },
+    {
+      "epoch": 0.9980116761147305,
+      "grad_norm": 0.14453408122062683,
+      "learning_rate": 0.001,
+      "loss": 2.7993,
+      "step": 23591
+    },
+    {
+      "epoch": 0.9980539808782469,
+      "grad_norm": 0.2429407685995102,
+      "learning_rate": 0.001,
+      "loss": 2.0416,
+      "step": 23592
+    },
+    {
+      "epoch": 0.9980962856417632,
+      "grad_norm": 0.15473546087741852,
+      "learning_rate": 0.001,
+      "loss": 1.5858,
+      "step": 23593
+    },
+    {
+      "epoch": 0.9981385904052796,
+      "grad_norm": 0.13461685180664062,
+      "learning_rate": 0.001,
+      "loss": 2.3416,
+      "step": 23594
+    },
+    {
+      "epoch": 0.998180895168796,
+      "grad_norm": 0.8460047841072083,
+      "learning_rate": 0.001,
+      "loss": 2.1287,
+      "step": 23595
+    },
+    {
+      "epoch": 0.9982231999323123,
+      "grad_norm": 0.15388406813144684,
+      "learning_rate": 0.001,
+      "loss": 1.5546,
+      "step": 23596
+    },
+    {
+      "epoch": 0.9982655046958288,
+      "grad_norm": 0.15559299290180206,
+      "learning_rate": 0.001,
+      "loss": 1.6043,
+      "step": 23597
+    },
+    {
+      "epoch": 0.9983078094593452,
+      "grad_norm": 0.14206543564796448,
+      "learning_rate": 0.001,
+      "loss": 1.5113,
+      "step": 23598
+    },
+    {
+      "epoch": 0.9983501142228615,
+      "grad_norm": 0.20963889360427856,
+      "learning_rate": 0.001,
+      "loss": 1.5628,
+      "step": 23599
+    },
+    {
+      "epoch": 0.9983924189863779,
+      "grad_norm": 0.20047658681869507,
+      "learning_rate": 0.001,
+      "loss": 1.5766,
+      "step": 23600
+    },
+    {
+      "epoch": 0.9984347237498943,
+      "grad_norm": 0.1295713633298874,
+      "learning_rate": 0.001,
+      "loss": 1.6966,
+      "step": 23601
+    },
+    {
+      "epoch": 0.9984770285134106,
+      "grad_norm": 0.20154249668121338,
+      "learning_rate": 0.001,
+      "loss": 1.6119,
+      "step": 23602
+    },
+    {
+      "epoch": 0.998519333276927,
+      "grad_norm": 0.14119143784046173,
+      "learning_rate": 0.001,
+      "loss": 1.534,
+      "step": 23603
+    },
+    {
+      "epoch": 0.9985616380404434,
+      "grad_norm": 0.15057438611984253,
+      "learning_rate": 0.001,
+      "loss": 2.1838,
+      "step": 23604
+    },
+    {
+      "epoch": 0.9986039428039597,
+      "grad_norm": 0.1763651818037033,
+      "learning_rate": 0.001,
+      "loss": 2.8296,
+      "step": 23605
+    },
+    {
+      "epoch": 0.9986462475674761,
+      "grad_norm": 0.20839767158031464,
+      "learning_rate": 0.001,
+      "loss": 2.2355,
+      "step": 23606
+    },
+    {
+      "epoch": 0.9986885523309925,
+      "grad_norm": 0.20311135053634644,
+      "learning_rate": 0.001,
+      "loss": 2.5101,
+      "step": 23607
+    },
+    {
+      "epoch": 0.9987308570945088,
+      "grad_norm": 0.17892763018608093,
+      "learning_rate": 0.001,
+      "loss": 2.4245,
+      "step": 23608
+    },
+    {
+      "epoch": 0.9987731618580252,
+      "grad_norm": 0.15312178432941437,
+      "learning_rate": 0.001,
+      "loss": 2.0084,
+      "step": 23609
+    },
+    {
+      "epoch": 0.9988154666215416,
+      "grad_norm": 0.15701162815093994,
+      "learning_rate": 0.001,
+      "loss": 1.6143,
+      "step": 23610
+    },
+    {
+      "epoch": 0.9988577713850579,
+      "grad_norm": 0.16364918649196625,
+      "learning_rate": 0.001,
+      "loss": 1.4365,
+      "step": 23611
+    },
+    {
+      "epoch": 0.9989000761485743,
+      "grad_norm": 0.4083903729915619,
+      "learning_rate": 0.001,
+      "loss": 1.6857,
+      "step": 23612
+    },
+    {
+      "epoch": 0.9989423809120908,
+      "grad_norm": 0.13597403466701508,
+      "learning_rate": 0.001,
+      "loss": 2.4054,
+      "step": 23613
+    },
+    {
+      "epoch": 0.998984685675607,
+      "grad_norm": 0.12076670676469803,
+      "learning_rate": 0.001,
+      "loss": 1.7387,
+      "step": 23614
+    },
+    {
+      "epoch": 0.9990269904391235,
+      "grad_norm": 0.20430530607700348,
+      "learning_rate": 0.001,
+      "loss": 2.5614,
+      "step": 23615
+    },
+    {
+      "epoch": 0.9990692952026398,
+      "grad_norm": 0.1506972312927246,
+      "learning_rate": 0.001,
+      "loss": 2.3331,
+      "step": 23616
+    },
+    {
+      "epoch": 0.9991115999661562,
+      "grad_norm": 0.14669986069202423,
+      "learning_rate": 0.001,
+      "loss": 1.5766,
+      "step": 23617
+    },
+    {
+      "epoch": 0.9991539047296726,
+      "grad_norm": 0.15908612310886383,
+      "learning_rate": 0.001,
+      "loss": 2.0938,
+      "step": 23618
+    },
+    {
+      "epoch": 0.9991962094931889,
+      "grad_norm": 0.1240277886390686,
+      "learning_rate": 0.001,
+      "loss": 2.1356,
+      "step": 23619
+    },
+    {
+      "epoch": 0.9992385142567053,
+      "grad_norm": 0.14557062089443207,
+      "learning_rate": 0.001,
+      "loss": 2.1815,
+      "step": 23620
+    },
+    {
+      "epoch": 0.9992808190202217,
+      "grad_norm": 0.16171297430992126,
+      "learning_rate": 0.001,
+      "loss": 1.7981,
+      "step": 23621
+    },
+    {
+      "epoch": 0.999323123783738,
+      "grad_norm": 0.157927006483078,
+      "learning_rate": 0.001,
+      "loss": 1.6165,
+      "step": 23622
+    },
+    {
+      "epoch": 0.9993654285472544,
+      "grad_norm": 0.15406683087348938,
+      "learning_rate": 0.001,
+      "loss": 2.1801,
+      "step": 23623
+    },
+    {
+      "epoch": 0.9994077333107708,
+      "grad_norm": 0.19311675429344177,
+      "learning_rate": 0.001,
+      "loss": 1.6993,
+      "step": 23624
+    },
+    {
+      "epoch": 0.9994500380742871,
+      "grad_norm": 0.24103492498397827,
+      "learning_rate": 0.001,
+      "loss": 2.6117,
+      "step": 23625
+    },
+    {
+      "epoch": 0.9994923428378035,
+      "grad_norm": 0.20797502994537354,
+      "learning_rate": 0.001,
+      "loss": 2.2529,
+      "step": 23626
+    },
+    {
+      "epoch": 0.9995346476013199,
+      "grad_norm": 0.3445920944213867,
+      "learning_rate": 0.001,
+      "loss": 2.4338,
+      "step": 23627
+    },
+    {
+      "epoch": 0.9995769523648362,
+      "grad_norm": 0.1522158980369568,
+      "learning_rate": 0.001,
+      "loss": 1.6047,
+      "step": 23628
+    },
+    {
+      "epoch": 0.9996192571283526,
+      "grad_norm": 1.6727036237716675,
+      "learning_rate": 0.001,
+      "loss": 3.4602,
+      "step": 23629
+    },
+    {
+      "epoch": 0.999661561891869,
+      "grad_norm": 0.14390981197357178,
+      "learning_rate": 0.001,
+      "loss": 1.7825,
+      "step": 23630
+    },
+    {
+      "epoch": 0.9997038666553854,
+      "grad_norm": 0.16790489852428436,
+      "learning_rate": 0.001,
+      "loss": 2.0287,
+      "step": 23631
+    },
+    {
+      "epoch": 0.9997461714189018,
+      "grad_norm": 1.9648386240005493,
+      "learning_rate": 0.001,
+      "loss": 1.9811,
+      "step": 23632
+    },
+    {
+      "epoch": 0.9997884761824182,
+      "grad_norm": 0.17755821347236633,
+      "learning_rate": 0.001,
+      "loss": 1.6772,
+      "step": 23633
+    },
+    {
+      "epoch": 0.9998307809459345,
+      "grad_norm": 0.1543680876493454,
+      "learning_rate": 0.001,
+      "loss": 2.2835,
+      "step": 23634
+    },
+    {
+      "epoch": 0.9998730857094509,
+      "grad_norm": 0.1278516948223114,
+      "learning_rate": 0.001,
+      "loss": 2.8439,
+      "step": 23635
+    },
+    {
+      "epoch": 0.9999153904729673,
+      "grad_norm": 0.15573585033416748,
+      "learning_rate": 0.001,
+      "loss": 2.505,
+      "step": 23636
+    },
+    {
+      "epoch": 0.9999576952364836,
+      "grad_norm": 0.17682018876075745,
+      "learning_rate": 0.001,
+      "loss": 1.2282,
+      "step": 23637
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19988803565502167,
+      "learning_rate": 0.001,
+      "loss": 1.9946,
+      "step": 23638
+    },
+    {
+      "epoch": 1.0,
+      "step": 23638,
+      "total_flos": 1.3698602096404378e+17,
+      "train_loss": 2.3282218011757556,
+      "train_runtime": 26726.0005,
+      "train_samples_per_second": 7.076,
+      "train_steps_per_second": 0.884
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 23638,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5910,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3698602096404378e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}