diff --git "a/checkpoint-1500/trainer_state.json" "b/checkpoint-1500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1500/trainer_state.json"
@@ -0,0 +1,10533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8317161075686166,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005544774050457444,
+      "grad_norm": 1.7043471336364746,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.8725,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011089548100914888,
+      "grad_norm": 1.2943507432937622,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.882,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016634322151372332,
+      "grad_norm": 1.2082455158233643,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.8304,
+      "step": 3
+    },
+    {
+      "epoch": 0.0022179096201829776,
+      "grad_norm": 0.9585644602775574,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.8958,
+      "step": 4
+    },
+    {
+      "epoch": 0.0027723870252287217,
+      "grad_norm": 1.1721417903900146,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.9696,
+      "step": 5
+    },
+    {
+      "epoch": 0.0033268644302744664,
+      "grad_norm": 0.9636610150337219,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.8921,
+      "step": 6
+    },
+    {
+      "epoch": 0.0038813418353202105,
+      "grad_norm": 1.1441367864608765,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.8852,
+      "step": 7
+    },
+    {
+      "epoch": 0.004435819240365955,
+      "grad_norm": 0.7332591414451599,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.8977,
+      "step": 8
+    },
+    {
+      "epoch": 0.0049902966454117,
+      "grad_norm": 0.736126184463501,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.8914,
+      "step": 9
+    },
+    {
+      "epoch": 0.0055447740504574435,
+      "grad_norm": 0.5888351798057556,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.959,
+      "step": 10
+    },
+    {
+      "epoch": 0.006099251455503188,
+      "grad_norm": 0.5506263971328735,
+      "learning_rate": 8.8e-06,
+      "loss": 1.8849,
+      "step": 11
+    },
+    {
+      "epoch": 0.006653728860548933,
+      "grad_norm": 0.5089067220687866,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.8318,
+      "step": 12
+    },
+    {
+      "epoch": 0.007208206265594677,
+      "grad_norm": 0.5495124459266663,
+      "learning_rate": 1.04e-05,
+      "loss": 1.8303,
+      "step": 13
+    },
+    {
+      "epoch": 0.007762683670640421,
+      "grad_norm": 0.5372406840324402,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.837,
+      "step": 14
+    },
+    {
+      "epoch": 0.008317161075686166,
+      "grad_norm": 0.4969247877597809,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8305,
+      "step": 15
+    },
+    {
+      "epoch": 0.00887163848073191,
+      "grad_norm": 0.43892425298690796,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.8299,
+      "step": 16
+    },
+    {
+      "epoch": 0.009426115885777655,
+      "grad_norm": 0.4186934530735016,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.8882,
+      "step": 17
+    },
+    {
+      "epoch": 0.0099805932908234,
+      "grad_norm": 0.36783891916275024,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.7997,
+      "step": 18
+    },
+    {
+      "epoch": 0.010535070695869144,
+      "grad_norm": 0.3592284023761749,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.7928,
+      "step": 19
+    },
+    {
+      "epoch": 0.011089548100914887,
+      "grad_norm": 0.4156821072101593,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.8382,
+      "step": 20
+    },
+    {
+      "epoch": 0.011644025505960632,
+      "grad_norm": 0.36838528513908386,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.765,
+      "step": 21
+    },
+    {
+      "epoch": 0.012198502911006376,
+      "grad_norm": 0.4337383508682251,
+      "learning_rate": 1.76e-05,
+      "loss": 1.7961,
+      "step": 22
+    },
+    {
+      "epoch": 0.01275298031605212,
+      "grad_norm": 0.34529271721839905,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.7795,
+      "step": 23
+    },
+    {
+      "epoch": 0.013307457721097865,
+      "grad_norm": 0.3445529341697693,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.8074,
+      "step": 24
+    },
+    {
+      "epoch": 0.01386193512614361,
+      "grad_norm": 0.32090720534324646,
+      "learning_rate": 2e-05,
+      "loss": 1.7261,
+      "step": 25
+    },
+    {
+      "epoch": 0.014416412531189355,
+      "grad_norm": 0.30814656615257263,
+      "learning_rate": 2.08e-05,
+      "loss": 1.7856,
+      "step": 26
+    },
+    {
+      "epoch": 0.0149708899362351,
+      "grad_norm": 0.3047591745853424,
+      "learning_rate": 2.1600000000000003e-05,
+      "loss": 1.7895,
+      "step": 27
+    },
+    {
+      "epoch": 0.015525367341280842,
+      "grad_norm": 0.2949264347553253,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.8077,
+      "step": 28
+    },
+    {
+      "epoch": 0.016079844746326587,
+      "grad_norm": 0.2805996239185333,
+      "learning_rate": 2.32e-05,
+      "loss": 1.7806,
+      "step": 29
+    },
+    {
+      "epoch": 0.01663432215137233,
+      "grad_norm": 0.2810855805873871,
+      "learning_rate": 2.4e-05,
+      "loss": 1.7969,
+      "step": 30
+    },
+    {
+      "epoch": 0.017188799556418076,
+      "grad_norm": 0.2863365411758423,
+      "learning_rate": 2.4800000000000003e-05,
+      "loss": 1.8053,
+      "step": 31
+    },
+    {
+      "epoch": 0.01774327696146382,
+      "grad_norm": 0.28471997380256653,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.7866,
+      "step": 32
+    },
+    {
+      "epoch": 0.018297754366509565,
+      "grad_norm": 0.2715449631214142,
+      "learning_rate": 2.6400000000000005e-05,
+      "loss": 1.7546,
+      "step": 33
+    },
+    {
+      "epoch": 0.01885223177155531,
+      "grad_norm": 0.256256639957428,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.7548,
+      "step": 34
+    },
+    {
+      "epoch": 0.019406709176601054,
+      "grad_norm": 0.26935723423957825,
+      "learning_rate": 2.8e-05,
+      "loss": 1.7742,
+      "step": 35
+    },
+    {
+      "epoch": 0.0199611865816468,
+      "grad_norm": 0.24071869254112244,
+      "learning_rate": 2.8800000000000002e-05,
+      "loss": 1.7903,
+      "step": 36
+    },
+    {
+      "epoch": 0.020515663986692544,
+      "grad_norm": 0.26816120743751526,
+      "learning_rate": 2.96e-05,
+      "loss": 1.83,
+      "step": 37
+    },
+    {
+      "epoch": 0.021070141391738288,
+      "grad_norm": 0.2523341774940491,
+      "learning_rate": 3.0400000000000004e-05,
+      "loss": 1.7478,
+      "step": 38
+    },
+    {
+      "epoch": 0.02162461879678403,
+      "grad_norm": 0.25324249267578125,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 1.763,
+      "step": 39
+    },
+    {
+      "epoch": 0.022179096201829774,
+      "grad_norm": 0.2640393376350403,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.7722,
+      "step": 40
+    },
+    {
+      "epoch": 0.02273357360687552,
+      "grad_norm": 0.2618652880191803,
+      "learning_rate": 3.28e-05,
+      "loss": 1.693,
+      "step": 41
+    },
+    {
+      "epoch": 0.023288051011921263,
+      "grad_norm": 0.2571040093898773,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.6989,
+      "step": 42
+    },
+    {
+      "epoch": 0.023842528416967008,
+      "grad_norm": 0.23714734613895416,
+      "learning_rate": 3.44e-05,
+      "loss": 1.769,
+      "step": 43
+    },
+    {
+      "epoch": 0.024397005822012752,
+      "grad_norm": 0.2353614717721939,
+      "learning_rate": 3.52e-05,
+      "loss": 1.7387,
+      "step": 44
+    },
+    {
+      "epoch": 0.024951483227058497,
+      "grad_norm": 0.2511078715324402,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7478,
+      "step": 45
+    },
+    {
+      "epoch": 0.02550596063210424,
+      "grad_norm": 0.2509211003780365,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.7231,
+      "step": 46
+    },
+    {
+      "epoch": 0.026060438037149986,
+      "grad_norm": 0.2628484070301056,
+      "learning_rate": 3.76e-05,
+      "loss": 1.8243,
+      "step": 47
+    },
+    {
+      "epoch": 0.02661491544219573,
+      "grad_norm": 0.23542599380016327,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.6862,
+      "step": 48
+    },
+    {
+      "epoch": 0.027169392847241475,
+      "grad_norm": 0.2473030835390091,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.6858,
+      "step": 49
+    },
+    {
+      "epoch": 0.02772387025228722,
+      "grad_norm": 0.24878697097301483,
+      "learning_rate": 4e-05,
+      "loss": 1.7847,
+      "step": 50
+    },
+    {
+      "epoch": 0.028278347657332965,
+      "grad_norm": 0.23445457220077515,
+      "learning_rate": 4.08e-05,
+      "loss": 1.7048,
+      "step": 51
+    },
+    {
+      "epoch": 0.02883282506237871,
+      "grad_norm": 0.23827125132083893,
+      "learning_rate": 4.16e-05,
+      "loss": 1.6579,
+      "step": 52
+    },
+    {
+      "epoch": 0.029387302467424454,
+      "grad_norm": 0.2579360902309418,
+      "learning_rate": 4.240000000000001e-05,
+      "loss": 1.7823,
+      "step": 53
+    },
+    {
+      "epoch": 0.0299417798724702,
+      "grad_norm": 0.2626456320285797,
+      "learning_rate": 4.3200000000000007e-05,
+      "loss": 1.7795,
+      "step": 54
+    },
+    {
+      "epoch": 0.03049625727751594,
+      "grad_norm": 0.24642065167427063,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.7575,
+      "step": 55
+    },
+    {
+      "epoch": 0.031050734682561684,
+      "grad_norm": 0.24688799679279327,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.7619,
+      "step": 56
+    },
+    {
+      "epoch": 0.03160521208760743,
+      "grad_norm": 0.24448993802070618,
+      "learning_rate": 4.56e-05,
+      "loss": 1.7114,
+      "step": 57
+    },
+    {
+      "epoch": 0.03215968949265317,
+      "grad_norm": 0.24135121703147888,
+      "learning_rate": 4.64e-05,
+      "loss": 1.7762,
+      "step": 58
+    },
+    {
+      "epoch": 0.03271416689769892,
+      "grad_norm": 0.24882449209690094,
+      "learning_rate": 4.72e-05,
+      "loss": 1.6934,
+      "step": 59
+    },
+    {
+      "epoch": 0.03326864430274466,
+      "grad_norm": 0.24142266809940338,
+      "learning_rate": 4.8e-05,
+      "loss": 1.6882,
+      "step": 60
+    },
+    {
+      "epoch": 0.03382312170779041,
+      "grad_norm": 0.2599814236164093,
+      "learning_rate": 4.88e-05,
+      "loss": 1.755,
+      "step": 61
+    },
+    {
+      "epoch": 0.03437759911283615,
+      "grad_norm": 0.2435712218284607,
+      "learning_rate": 4.9600000000000006e-05,
+      "loss": 1.7064,
+      "step": 62
+    },
+    {
+      "epoch": 0.034932076517881896,
+      "grad_norm": 0.25077876448631287,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 1.6786,
+      "step": 63
+    },
+    {
+      "epoch": 0.03548655392292764,
+      "grad_norm": 0.25001490116119385,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 1.6983,
+      "step": 64
+    },
+    {
+      "epoch": 0.036041031327973386,
+      "grad_norm": 0.24909815192222595,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.6403,
+      "step": 65
+    },
+    {
+      "epoch": 0.03659550873301913,
+      "grad_norm": 0.2543540298938751,
+      "learning_rate": 5.280000000000001e-05,
+      "loss": 1.6812,
+      "step": 66
+    },
+    {
+      "epoch": 0.037149986138064875,
+      "grad_norm": 0.26552706956863403,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 1.7326,
+      "step": 67
+    },
+    {
+      "epoch": 0.03770446354311062,
+      "grad_norm": 0.24254681169986725,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 1.7514,
+      "step": 68
+    },
+    {
+      "epoch": 0.038258940948156364,
+      "grad_norm": 0.2548975944519043,
+      "learning_rate": 5.52e-05,
+      "loss": 1.6883,
+      "step": 69
+    },
+    {
+      "epoch": 0.03881341835320211,
+      "grad_norm": 0.25074952840805054,
+      "learning_rate": 5.6e-05,
+      "loss": 1.7236,
+      "step": 70
+    },
+    {
+      "epoch": 0.03936789575824785,
+      "grad_norm": 0.26199671626091003,
+      "learning_rate": 5.6800000000000005e-05,
+      "loss": 1.6943,
+      "step": 71
+    },
+    {
+      "epoch": 0.0399223731632936,
+      "grad_norm": 0.27566343545913696,
+      "learning_rate": 5.7600000000000004e-05,
+      "loss": 1.6859,
+      "step": 72
+    },
+    {
+      "epoch": 0.04047685056833934,
+      "grad_norm": 0.26068150997161865,
+      "learning_rate": 5.84e-05,
+      "loss": 1.7378,
+      "step": 73
+    },
+    {
+      "epoch": 0.04103132797338509,
+      "grad_norm": 0.2568175196647644,
+      "learning_rate": 5.92e-05,
+      "loss": 1.6894,
+      "step": 74
+    },
+    {
+      "epoch": 0.04158580537843083,
+      "grad_norm": 0.2596546709537506,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 1.7519,
+      "step": 75
+    },
+    {
+      "epoch": 0.042140282783476576,
+      "grad_norm": 0.2714129686355591,
+      "learning_rate": 6.080000000000001e-05,
+      "loss": 1.7159,
+      "step": 76
+    },
+    {
+      "epoch": 0.04269476018852232,
+      "grad_norm": 0.2649083733558655,
+      "learning_rate": 6.16e-05,
+      "loss": 1.7031,
+      "step": 77
+    },
+    {
+      "epoch": 0.04324923759356806,
+      "grad_norm": 0.30052971839904785,
+      "learning_rate": 6.240000000000001e-05,
+      "loss": 1.7358,
+      "step": 78
+    },
+    {
+      "epoch": 0.0438037149986138,
+      "grad_norm": 0.30009427666664124,
+      "learning_rate": 6.32e-05,
+      "loss": 1.6739,
+      "step": 79
+    },
+    {
+      "epoch": 0.04435819240365955,
+      "grad_norm": 0.27122125029563904,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.721,
+      "step": 80
+    },
+    {
+      "epoch": 0.04491266980870529,
+      "grad_norm": 0.30771657824516296,
+      "learning_rate": 6.48e-05,
+      "loss": 1.8069,
+      "step": 81
+    },
+    {
+      "epoch": 0.04546714721375104,
+      "grad_norm": 0.24620802700519562,
+      "learning_rate": 6.56e-05,
+      "loss": 1.6847,
+      "step": 82
+    },
+    {
+      "epoch": 0.04602162461879678,
+      "grad_norm": 0.2720058262348175,
+      "learning_rate": 6.64e-05,
+      "loss": 1.7179,
+      "step": 83
+    },
+    {
+      "epoch": 0.046576102023842526,
+      "grad_norm": 0.24807985126972198,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 1.6593,
+      "step": 84
+    },
+    {
+      "epoch": 0.04713057942888827,
+      "grad_norm": 0.2534205913543701,
+      "learning_rate": 6.8e-05,
+      "loss": 1.7475,
+      "step": 85
+    },
+    {
+      "epoch": 0.047685056833934016,
+      "grad_norm": 0.28452959656715393,
+      "learning_rate": 6.88e-05,
+      "loss": 1.6402,
+      "step": 86
+    },
+    {
+      "epoch": 0.04823953423897976,
+      "grad_norm": 0.2749129831790924,
+      "learning_rate": 6.960000000000001e-05,
+      "loss": 1.7025,
+      "step": 87
+    },
+    {
+      "epoch": 0.048794011644025505,
+      "grad_norm": 0.26466885209083557,
+      "learning_rate": 7.04e-05,
+      "loss": 1.7151,
+      "step": 88
+    },
+    {
+      "epoch": 0.04934848904907125,
+      "grad_norm": 0.32394999265670776,
+      "learning_rate": 7.120000000000001e-05,
+      "loss": 1.6569,
+      "step": 89
+    },
+    {
+      "epoch": 0.049902966454116994,
+      "grad_norm": 0.23801451921463013,
+      "learning_rate": 7.2e-05,
+      "loss": 1.7009,
+      "step": 90
+    },
+    {
+      "epoch": 0.05045744385916274,
+      "grad_norm": 0.33740007877349854,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 1.6932,
+      "step": 91
+    },
+    {
+      "epoch": 0.05101192126420848,
+      "grad_norm": 0.2623404562473297,
+      "learning_rate": 7.360000000000001e-05,
+      "loss": 1.7311,
+      "step": 92
+    },
+    {
+      "epoch": 0.05156639866925423,
+      "grad_norm": 0.292054146528244,
+      "learning_rate": 7.44e-05,
+      "loss": 1.6297,
+      "step": 93
+    },
+    {
+      "epoch": 0.05212087607429997,
+      "grad_norm": 0.3140726387500763,
+      "learning_rate": 7.52e-05,
+      "loss": 1.7442,
+      "step": 94
+    },
+    {
+      "epoch": 0.05267535347934572,
+      "grad_norm": 0.2532927989959717,
+      "learning_rate": 7.6e-05,
+      "loss": 1.736,
+      "step": 95
+    },
+    {
+      "epoch": 0.05322983088439146,
+      "grad_norm": 0.31100720167160034,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 1.6931,
+      "step": 96
+    },
+    {
+      "epoch": 0.053784308289437206,
+      "grad_norm": 0.275055855512619,
+      "learning_rate": 7.76e-05,
+      "loss": 1.6661,
+      "step": 97
+    },
+    {
+      "epoch": 0.05433878569448295,
+      "grad_norm": 0.3106204569339752,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 1.7421,
+      "step": 98
+    },
+    {
+      "epoch": 0.054893263099528695,
+      "grad_norm": 0.30006468296051025,
+      "learning_rate": 7.92e-05,
+      "loss": 1.686,
+      "step": 99
+    },
+    {
+      "epoch": 0.05544774050457444,
+      "grad_norm": 0.2701585292816162,
+      "learning_rate": 8e-05,
+      "loss": 1.6339,
+      "step": 100
+    },
+    {
+      "epoch": 0.056002217909620185,
+      "grad_norm": 0.31820419430732727,
+      "learning_rate": 7.999993193868717e-05,
+      "loss": 1.772,
+      "step": 101
+    },
+    {
+      "epoch": 0.05655669531466593,
+      "grad_norm": 0.25506505370140076,
+      "learning_rate": 7.999972775498027e-05,
+      "loss": 1.6857,
+      "step": 102
+    },
+    {
+      "epoch": 0.057111172719711674,
+      "grad_norm": 0.27725860476493835,
+      "learning_rate": 7.999938744957418e-05,
+      "loss": 1.756,
+      "step": 103
+    },
+    {
+      "epoch": 0.05766565012475742,
+      "grad_norm": 0.237404003739357,
+      "learning_rate": 7.999891102362694e-05,
+      "loss": 1.6907,
+      "step": 104
+    },
+    {
+      "epoch": 0.05822012752980316,
+      "grad_norm": 0.26032620668411255,
+      "learning_rate": 7.999829847875989e-05,
+      "loss": 1.6277,
+      "step": 105
+    },
+    {
+      "epoch": 0.05877460493484891,
+      "grad_norm": 0.23683027923107147,
+      "learning_rate": 7.999754981705756e-05,
+      "loss": 1.6519,
+      "step": 106
+    },
+    {
+      "epoch": 0.05932908233989465,
+      "grad_norm": 0.2389203906059265,
+      "learning_rate": 7.999666504106769e-05,
+      "loss": 1.6287,
+      "step": 107
+    },
+    {
+      "epoch": 0.0598835597449404,
+      "grad_norm": 0.25654640793800354,
+      "learning_rate": 7.999564415380122e-05,
+      "loss": 1.7226,
+      "step": 108
+    },
+    {
+      "epoch": 0.060438037149986135,
+      "grad_norm": 0.25321975350379944,
+      "learning_rate": 7.99944871587323e-05,
+      "loss": 1.6724,
+      "step": 109
+    },
+    {
+      "epoch": 0.06099251455503188,
+      "grad_norm": 0.25723859667778015,
+      "learning_rate": 7.999319405979828e-05,
+      "loss": 1.7587,
+      "step": 110
+    },
+    {
+      "epoch": 0.061546991960077624,
+      "grad_norm": 0.23742374777793884,
+      "learning_rate": 7.999176486139964e-05,
+      "loss": 1.7245,
+      "step": 111
+    },
+    {
+      "epoch": 0.06210146936512337,
+      "grad_norm": 0.23199070990085602,
+      "learning_rate": 7.999019956840004e-05,
+      "loss": 1.7447,
+      "step": 112
+    },
+    {
+      "epoch": 0.06265594677016911,
+      "grad_norm": 0.2442471832036972,
+      "learning_rate": 7.998849818612628e-05,
+      "loss": 1.8053,
+      "step": 113
+    },
+    {
+      "epoch": 0.06321042417521486,
+      "grad_norm": 0.22759103775024414,
+      "learning_rate": 7.998666072036827e-05,
+      "loss": 1.6593,
+      "step": 114
+    },
+    {
+      "epoch": 0.0637649015802606,
+      "grad_norm": 0.23776134848594666,
+      "learning_rate": 7.998468717737903e-05,
+      "loss": 1.7303,
+      "step": 115
+    },
+    {
+      "epoch": 0.06431937898530635,
+      "grad_norm": 0.2140997052192688,
+      "learning_rate": 7.998257756387466e-05,
+      "loss": 1.6146,
+      "step": 116
+    },
+    {
+      "epoch": 0.06487385639035209,
+      "grad_norm": 0.2312529981136322,
+      "learning_rate": 7.99803318870343e-05,
+      "loss": 1.691,
+      "step": 117
+    },
+    {
+      "epoch": 0.06542833379539784,
+      "grad_norm": 0.23457646369934082,
+      "learning_rate": 7.997795015450015e-05,
+      "loss": 1.6844,
+      "step": 118
+    },
+    {
+      "epoch": 0.06598281120044358,
+      "grad_norm": 0.25258150696754456,
+      "learning_rate": 7.997543237437738e-05,
+      "loss": 1.7305,
+      "step": 119
+    },
+    {
+      "epoch": 0.06653728860548933,
+      "grad_norm": 0.21662160754203796,
+      "learning_rate": 7.99727785552342e-05,
+      "loss": 1.6637,
+      "step": 120
+    },
+    {
+      "epoch": 0.06709176601053507,
+      "grad_norm": 0.23025107383728027,
+      "learning_rate": 7.99699887061017e-05,
+      "loss": 1.7312,
+      "step": 121
+    },
+    {
+      "epoch": 0.06764624341558081,
+      "grad_norm": 0.2787688374519348,
+      "learning_rate": 7.996706283647393e-05,
+      "loss": 1.7159,
+      "step": 122
+    },
+    {
+      "epoch": 0.06820072082062656,
+      "grad_norm": 0.26639994978904724,
+      "learning_rate": 7.996400095630781e-05,
+      "loss": 1.7311,
+      "step": 123
+    },
+    {
+      "epoch": 0.0687551982256723,
+      "grad_norm": 0.28086450695991516,
+      "learning_rate": 7.996080307602312e-05,
+      "loss": 1.6401,
+      "step": 124
+    },
+    {
+      "epoch": 0.06930967563071805,
+      "grad_norm": 0.24936339259147644,
+      "learning_rate": 7.995746920650248e-05,
+      "loss": 1.6197,
+      "step": 125
+    },
+    {
+      "epoch": 0.06986415303576379,
+      "grad_norm": 0.263729065656662,
+      "learning_rate": 7.995399935909122e-05,
+      "loss": 1.6693,
+      "step": 126
+    },
+    {
+      "epoch": 0.07041863044080954,
+      "grad_norm": 0.3544696271419525,
+      "learning_rate": 7.99503935455975e-05,
+      "loss": 1.6949,
+      "step": 127
+    },
+    {
+      "epoch": 0.07097310784585528,
+      "grad_norm": 0.23669736087322235,
+      "learning_rate": 7.994665177829211e-05,
+      "loss": 1.6262,
+      "step": 128
+    },
+    {
+      "epoch": 0.07152758525090103,
+      "grad_norm": 0.3165992498397827,
+      "learning_rate": 7.994277406990857e-05,
+      "loss": 1.6604,
+      "step": 129
+    },
+    {
+      "epoch": 0.07208206265594677,
+      "grad_norm": 0.29037636518478394,
+      "learning_rate": 7.993876043364294e-05,
+      "loss": 1.5977,
+      "step": 130
+    },
+    {
+      "epoch": 0.07263654006099252,
+      "grad_norm": 0.26414918899536133,
+      "learning_rate": 7.993461088315389e-05,
+      "loss": 1.7664,
+      "step": 131
+    },
+    {
+      "epoch": 0.07319101746603826,
+      "grad_norm": 0.3492623567581177,
+      "learning_rate": 7.993032543256263e-05,
+      "loss": 1.7055,
+      "step": 132
+    },
+    {
+      "epoch": 0.073745494871084,
+      "grad_norm": 0.2867588698863983,
+      "learning_rate": 7.992590409645282e-05,
+      "loss": 1.8092,
+      "step": 133
+    },
+    {
+      "epoch": 0.07429997227612975,
+      "grad_norm": 0.248403400182724,
+      "learning_rate": 7.992134688987056e-05,
+      "loss": 1.6885,
+      "step": 134
+    },
+    {
+      "epoch": 0.0748544496811755,
+      "grad_norm": 0.306547075510025,
+      "learning_rate": 7.991665382832433e-05,
+      "loss": 1.7218,
+      "step": 135
+    },
+    {
+      "epoch": 0.07540892708622124,
+      "grad_norm": 0.25006142258644104,
+      "learning_rate": 7.99118249277849e-05,
+      "loss": 1.6728,
+      "step": 136
+    },
+    {
+      "epoch": 0.07596340449126698,
+      "grad_norm": 0.29479724168777466,
+      "learning_rate": 7.990686020468536e-05,
+      "loss": 1.6751,
+      "step": 137
+    },
+    {
+      "epoch": 0.07651788189631273,
+      "grad_norm": 0.23760248720645905,
+      "learning_rate": 7.990175967592098e-05,
+      "loss": 1.625,
+      "step": 138
+    },
+    {
+      "epoch": 0.07707235930135847,
+      "grad_norm": 0.30315011739730835,
+      "learning_rate": 7.98965233588492e-05,
+      "loss": 1.6881,
+      "step": 139
+    },
+    {
+      "epoch": 0.07762683670640422,
+      "grad_norm": 0.22332574427127838,
+      "learning_rate": 7.989115127128955e-05,
+      "loss": 1.6179,
+      "step": 140
+    },
+    {
+      "epoch": 0.07818131411144996,
+      "grad_norm": 0.28878095746040344,
+      "learning_rate": 7.98856434315236e-05,
+      "loss": 1.7134,
+      "step": 141
+    },
+    {
+      "epoch": 0.0787357915164957,
+      "grad_norm": 0.231205016374588,
+      "learning_rate": 7.987999985829486e-05,
+      "loss": 1.6805,
+      "step": 142
+    },
+    {
+      "epoch": 0.07929026892154145,
+      "grad_norm": 0.29243049025535583,
+      "learning_rate": 7.987422057080881e-05,
+      "loss": 1.7045,
+      "step": 143
+    },
+    {
+      "epoch": 0.0798447463265872,
+      "grad_norm": 0.23502500355243683,
+      "learning_rate": 7.986830558873275e-05,
+      "loss": 1.6984,
+      "step": 144
+    },
+    {
+      "epoch": 0.08039922373163294,
+      "grad_norm": 0.23500199615955353,
+      "learning_rate": 7.986225493219573e-05,
+      "loss": 1.5982,
+      "step": 145
+    },
+    {
+      "epoch": 0.08095370113667869,
+      "grad_norm": 0.2353154569864273,
+      "learning_rate": 7.985606862178855e-05,
+      "loss": 1.7826,
+      "step": 146
+    },
+    {
+      "epoch": 0.08150817854172443,
+      "grad_norm": 0.2630767524242401,
+      "learning_rate": 7.984974667856362e-05,
+      "loss": 1.7355,
+      "step": 147
+    },
+    {
+      "epoch": 0.08206265594677017,
+      "grad_norm": 0.221221461892128,
+      "learning_rate": 7.984328912403494e-05,
+      "loss": 1.7203,
+      "step": 148
+    },
+    {
+      "epoch": 0.08261713335181592,
+      "grad_norm": 0.23368635773658752,
+      "learning_rate": 7.983669598017798e-05,
+      "loss": 1.6725,
+      "step": 149
+    },
+    {
+      "epoch": 0.08317161075686166,
+      "grad_norm": 0.22409488260746002,
+      "learning_rate": 7.982996726942963e-05,
+      "loss": 1.685,
+      "step": 150
+    },
+    {
+      "epoch": 0.08372608816190741,
+      "grad_norm": 0.2248145341873169,
+      "learning_rate": 7.982310301468815e-05,
+      "loss": 1.6743,
+      "step": 151
+    },
+    {
+      "epoch": 0.08428056556695315,
+      "grad_norm": 0.2202092409133911,
+      "learning_rate": 7.981610323931306e-05,
+      "loss": 1.6568,
+      "step": 152
+    },
+    {
+      "epoch": 0.0848350429719989,
+      "grad_norm": 0.2243170440196991,
+      "learning_rate": 7.980896796712504e-05,
+      "loss": 1.7107,
+      "step": 153
+    },
+    {
+      "epoch": 0.08538952037704464,
+      "grad_norm": 0.23059862852096558,
+      "learning_rate": 7.980169722240589e-05,
+      "loss": 1.6998,
+      "step": 154
+    },
+    {
+      "epoch": 0.08594399778209039,
+      "grad_norm": 0.2216973900794983,
+      "learning_rate": 7.979429102989842e-05,
+      "loss": 1.6527,
+      "step": 155
+    },
+    {
+      "epoch": 0.08649847518713612,
+      "grad_norm": 0.23553825914859772,
+      "learning_rate": 7.978674941480643e-05,
+      "loss": 1.6883,
+      "step": 156
+    },
+    {
+      "epoch": 0.08705295259218186,
+      "grad_norm": 0.2300165742635727,
+      "learning_rate": 7.977907240279449e-05,
+      "loss": 1.6638,
+      "step": 157
+    },
+    {
+      "epoch": 0.0876074299972276,
+      "grad_norm": 0.23822426795959473,
+      "learning_rate": 7.977126001998798e-05,
+      "loss": 1.6435,
+      "step": 158
+    },
+    {
+      "epoch": 0.08816190740227335,
+      "grad_norm": 0.2425059974193573,
+      "learning_rate": 7.976331229297298e-05,
+      "loss": 1.6745,
+      "step": 159
+    },
+    {
+      "epoch": 0.0887163848073191,
+      "grad_norm": 0.26218435168266296,
+      "learning_rate": 7.975522924879609e-05,
+      "loss": 1.6546,
+      "step": 160
+    },
+    {
+      "epoch": 0.08927086221236484,
+      "grad_norm": 0.23087595403194427,
+      "learning_rate": 7.974701091496448e-05,
+      "loss": 1.6581,
+      "step": 161
+    },
+    {
+      "epoch": 0.08982533961741059,
+      "grad_norm": 0.24700067937374115,
+      "learning_rate": 7.973865731944565e-05,
+      "loss": 1.7501,
+      "step": 162
+    },
+    {
+      "epoch": 0.09037981702245633,
+      "grad_norm": 0.22620323300361633,
+      "learning_rate": 7.973016849066742e-05,
+      "loss": 1.698,
+      "step": 163
+    },
+    {
+      "epoch": 0.09093429442750207,
+      "grad_norm": 0.21732601523399353,
+      "learning_rate": 7.972154445751788e-05,
+      "loss": 1.622,
+      "step": 164
+    },
+    {
+      "epoch": 0.09148877183254782,
+      "grad_norm": 0.22213229537010193,
+      "learning_rate": 7.971278524934515e-05,
+      "loss": 1.7224,
+      "step": 165
+    },
+    {
+      "epoch": 0.09204324923759356,
+      "grad_norm": 0.27407029271125793,
+      "learning_rate": 7.970389089595738e-05,
+      "loss": 1.679,
+      "step": 166
+    },
+    {
+      "epoch": 0.09259772664263931,
+      "grad_norm": 0.24043461680412292,
+      "learning_rate": 7.969486142762266e-05,
+      "loss": 1.6364,
+      "step": 167
+    },
+    {
+      "epoch": 0.09315220404768505,
+      "grad_norm": 0.22668756544589996,
+      "learning_rate": 7.968569687506886e-05,
+      "loss": 1.6504,
+      "step": 168
+    },
+    {
+      "epoch": 0.0937066814527308,
+      "grad_norm": 0.2548286020755768,
+      "learning_rate": 7.967639726948355e-05,
+      "loss": 1.6703,
+      "step": 169
+    },
+    {
+      "epoch": 0.09426115885777654,
+      "grad_norm": 0.2330275923013687,
+      "learning_rate": 7.96669626425139e-05,
+      "loss": 1.691,
+      "step": 170
+    },
+    {
+      "epoch": 0.09481563626282229,
+      "grad_norm": 0.2295777052640915,
+      "learning_rate": 7.965739302626656e-05,
+      "loss": 1.66,
+      "step": 171
+    },
+    {
+      "epoch": 0.09537011366786803,
+      "grad_norm": 0.23224836587905884,
+      "learning_rate": 7.964768845330756e-05,
+      "loss": 1.6317,
+      "step": 172
+    },
+    {
+      "epoch": 0.09592459107291378,
+      "grad_norm": 0.21784305572509766,
+      "learning_rate": 7.963784895666221e-05,
+      "loss": 1.674,
+      "step": 173
+    },
+    {
+      "epoch": 0.09647906847795952,
+      "grad_norm": 0.25221794843673706,
+      "learning_rate": 7.962787456981498e-05,
+      "loss": 1.6848,
+      "step": 174
+    },
+    {
+      "epoch": 0.09703354588300526,
+      "grad_norm": 0.2345496118068695,
+      "learning_rate": 7.961776532670931e-05,
+      "loss": 1.6349,
+      "step": 175
+    },
+    {
+      "epoch": 0.09758802328805101,
+      "grad_norm": 0.23866072297096252,
+      "learning_rate": 7.960752126174765e-05,
+      "loss": 1.6078,
+      "step": 176
+    },
+    {
+      "epoch": 0.09814250069309675,
+      "grad_norm": 0.2709258794784546,
+      "learning_rate": 7.959714240979124e-05,
+      "loss": 1.6811,
+      "step": 177
+    },
+    {
+      "epoch": 0.0986969780981425,
+      "grad_norm": 0.22360965609550476,
+      "learning_rate": 7.958662880615997e-05,
+      "loss": 1.633,
+      "step": 178
+    },
+    {
+      "epoch": 0.09925145550318824,
+      "grad_norm": 0.25358742475509644,
+      "learning_rate": 7.957598048663234e-05,
+      "loss": 1.6449,
+      "step": 179
+    },
+    {
+      "epoch": 0.09980593290823399,
+      "grad_norm": 0.2268538922071457,
+      "learning_rate": 7.956519748744525e-05,
+      "loss": 1.6675,
+      "step": 180
+    },
+    {
+      "epoch": 0.10036041031327973,
+      "grad_norm": 0.26430633664131165,
+      "learning_rate": 7.9554279845294e-05,
+      "loss": 1.7128,
+      "step": 181
+    },
+    {
+      "epoch": 0.10091488771832548,
+      "grad_norm": 0.22235046327114105,
+      "learning_rate": 7.9543227597332e-05,
+      "loss": 1.6033,
+      "step": 182
+    },
+    {
+      "epoch": 0.10146936512337122,
+      "grad_norm": 0.26423388719558716,
+      "learning_rate": 7.953204078117081e-05,
+      "loss": 1.6274,
+      "step": 183
+    },
+    {
+      "epoch": 0.10202384252841697,
+      "grad_norm": 0.22570672631263733,
+      "learning_rate": 7.952071943487987e-05,
+      "loss": 1.6938,
+      "step": 184
+    },
+    {
+      "epoch": 0.10257831993346271,
+      "grad_norm": 0.25327280163764954,
+      "learning_rate": 7.95092635969865e-05,
+      "loss": 1.7074,
+      "step": 185
+    },
+    {
+      "epoch": 0.10313279733850846,
+      "grad_norm": 0.21103128790855408,
+      "learning_rate": 7.949767330647562e-05,
+      "loss": 1.6202,
+      "step": 186
+    },
+    {
+      "epoch": 0.1036872747435542,
+      "grad_norm": 0.24825690686702728,
+      "learning_rate": 7.94859486027898e-05,
+      "loss": 1.6757,
+      "step": 187
+    },
+    {
+      "epoch": 0.10424175214859994,
+      "grad_norm": 0.2204912304878235,
+      "learning_rate": 7.947408952582892e-05,
+      "loss": 1.6779,
+      "step": 188
+    },
+    {
+      "epoch": 0.10479622955364569,
+      "grad_norm": 0.26558375358581543,
+      "learning_rate": 7.946209611595026e-05,
+      "loss": 1.6829,
+      "step": 189
+    },
+    {
+      "epoch": 0.10535070695869143,
+      "grad_norm": 0.2372429221868515,
+      "learning_rate": 7.944996841396815e-05,
+      "loss": 1.6825,
+      "step": 190
+    },
+    {
+      "epoch": 0.10590518436373718,
+      "grad_norm": 0.21156929433345795,
+      "learning_rate": 7.943770646115396e-05,
+      "loss": 1.6063,
+      "step": 191
+    },
+    {
+      "epoch": 0.10645966176878292,
+      "grad_norm": 0.21675823628902435,
+      "learning_rate": 7.94253102992359e-05,
+      "loss": 1.612,
+      "step": 192
+    },
+    {
+      "epoch": 0.10701413917382867,
+      "grad_norm": 0.2156941145658493,
+      "learning_rate": 7.941277997039894e-05,
+      "loss": 1.6382,
+      "step": 193
+    },
+    {
+      "epoch": 0.10756861657887441,
+      "grad_norm": 0.20940068364143372,
+      "learning_rate": 7.940011551728463e-05,
+      "loss": 1.6607,
+      "step": 194
+    },
+    {
+      "epoch": 0.10812309398392016,
+      "grad_norm": 0.21804174780845642,
+      "learning_rate": 7.93873169829909e-05,
+      "loss": 1.6386,
+      "step": 195
+    },
+    {
+      "epoch": 0.1086775713889659,
+      "grad_norm": 0.21801279485225677,
+      "learning_rate": 7.937438441107203e-05,
+      "loss": 1.6612,
+      "step": 196
+    },
+    {
+      "epoch": 0.10923204879401165,
+      "grad_norm": 0.254651814699173,
+      "learning_rate": 7.93613178455384e-05,
+      "loss": 1.6548,
+      "step": 197
+    },
+    {
+      "epoch": 0.10978652619905739,
+      "grad_norm": 0.20725683867931366,
+      "learning_rate": 7.934811733085637e-05,
+      "loss": 1.6508,
+      "step": 198
+    },
+    {
+      "epoch": 0.11034100360410314,
+      "grad_norm": 0.25397831201553345,
+      "learning_rate": 7.933478291194821e-05,
+      "loss": 1.6434,
+      "step": 199
+    },
+    {
+      "epoch": 0.11089548100914888,
+      "grad_norm": 0.21976791322231293,
+      "learning_rate": 7.932131463419177e-05,
+      "loss": 1.6073,
+      "step": 200
+    },
+    {
+      "epoch": 0.11144995841419462,
+      "grad_norm": 0.23310157656669617,
+      "learning_rate": 7.930771254342051e-05,
+      "loss": 1.6755,
+      "step": 201
+    },
+    {
+      "epoch": 0.11200443581924037,
+      "grad_norm": 0.24408988654613495,
+      "learning_rate": 7.929397668592323e-05,
+      "loss": 1.6681,
+      "step": 202
+    },
+    {
+      "epoch": 0.11255891322428611,
+      "grad_norm": 0.22306199371814728,
+      "learning_rate": 7.928010710844397e-05,
+      "loss": 1.6328,
+      "step": 203
+    },
+    {
+      "epoch": 0.11311339062933186,
+      "grad_norm": 0.2562360465526581,
+      "learning_rate": 7.926610385818179e-05,
+      "loss": 1.639,
+      "step": 204
+    },
+    {
+      "epoch": 0.1136678680343776,
+      "grad_norm": 0.22006537020206451,
+      "learning_rate": 7.925196698279068e-05,
+      "loss": 1.6636,
+      "step": 205
+    },
+    {
+      "epoch": 0.11422234543942335,
+      "grad_norm": 0.23495469987392426,
+      "learning_rate": 7.923769653037935e-05,
+      "loss": 1.6408,
+      "step": 206
+    },
+    {
+      "epoch": 0.11477682284446909,
+      "grad_norm": 0.215364009141922,
+      "learning_rate": 7.92232925495111e-05,
+      "loss": 1.6994,
+      "step": 207
+    },
+    {
+      "epoch": 0.11533130024951484,
+      "grad_norm": 0.2487768530845642,
+      "learning_rate": 7.920875508920361e-05,
+      "loss": 1.6307,
+      "step": 208
+    },
+    {
+      "epoch": 0.11588577765456058,
+      "grad_norm": 0.2132566124200821,
+      "learning_rate": 7.919408419892881e-05,
+      "loss": 1.6538,
+      "step": 209
+    },
+    {
+      "epoch": 0.11644025505960633,
+      "grad_norm": 0.22761502861976624,
+      "learning_rate": 7.917927992861272e-05,
+      "loss": 1.6135,
+      "step": 210
+    },
+    {
+      "epoch": 0.11699473246465207,
+      "grad_norm": 0.22848926484584808,
+      "learning_rate": 7.916434232863522e-05,
+      "loss": 1.7293,
+      "step": 211
+    },
+    {
+      "epoch": 0.11754920986969782,
+      "grad_norm": 0.21311335265636444,
+      "learning_rate": 7.914927144982995e-05,
+      "loss": 1.6792,
+      "step": 212
+    },
+    {
+      "epoch": 0.11810368727474356,
+      "grad_norm": 0.23164795339107513,
+      "learning_rate": 7.913406734348412e-05,
+      "loss": 1.5472,
+      "step": 213
+    },
+    {
+      "epoch": 0.1186581646797893,
+      "grad_norm": 0.21141847968101501,
+      "learning_rate": 7.911873006133827e-05,
+      "loss": 1.6259,
+      "step": 214
+    },
+    {
+      "epoch": 0.11921264208483505,
+      "grad_norm": 0.2537563443183899,
+      "learning_rate": 7.910325965558621e-05,
+      "loss": 1.7058,
+      "step": 215
+    },
+    {
+      "epoch": 0.1197671194898808,
+      "grad_norm": 0.2353929728269577,
+      "learning_rate": 7.908765617887473e-05,
+      "loss": 1.6479,
+      "step": 216
+    },
+    {
+      "epoch": 0.12032159689492654,
+      "grad_norm": 0.20736263692378998,
+      "learning_rate": 7.907191968430347e-05,
+      "loss": 1.603,
+      "step": 217
+    },
+    {
+      "epoch": 0.12087607429997227,
+      "grad_norm": 0.23365424573421478,
+      "learning_rate": 7.905605022542478e-05,
+      "loss": 1.6792,
+      "step": 218
+    },
+    {
+      "epoch": 0.12143055170501801,
+      "grad_norm": 0.22553478181362152,
+      "learning_rate": 7.904004785624345e-05,
+      "loss": 1.7089,
+      "step": 219
+    },
+    {
+      "epoch": 0.12198502911006376,
+      "grad_norm": 0.2107611447572708,
+      "learning_rate": 7.902391263121662e-05,
+      "loss": 1.6541,
+      "step": 220
+    },
+    {
+      "epoch": 0.1225395065151095,
+      "grad_norm": 0.21160966157913208,
+      "learning_rate": 7.900764460525349e-05,
+      "loss": 1.6459,
+      "step": 221
+    },
+    {
+      "epoch": 0.12309398392015525,
+      "grad_norm": 0.21517297625541687,
+      "learning_rate": 7.899124383371524e-05,
+      "loss": 1.6354,
+      "step": 222
+    },
+    {
+      "epoch": 0.12364846132520099,
+      "grad_norm": 0.21948044002056122,
+      "learning_rate": 7.897471037241476e-05,
+      "loss": 1.7011,
+      "step": 223
+    },
+    {
+      "epoch": 0.12420293873024674,
+      "grad_norm": 0.21453773975372314,
+      "learning_rate": 7.895804427761651e-05,
+      "loss": 1.6508,
+      "step": 224
+    },
+    {
+      "epoch": 0.12475741613529248,
+      "grad_norm": 0.23449410498142242,
+      "learning_rate": 7.894124560603631e-05,
+      "loss": 1.6891,
+      "step": 225
+    },
+    {
+      "epoch": 0.12531189354033823,
+      "grad_norm": 0.22934886813163757,
+      "learning_rate": 7.892431441484113e-05,
+      "loss": 1.6989,
+      "step": 226
+    },
+    {
+      "epoch": 0.12586637094538397,
+      "grad_norm": 0.24871671199798584,
+      "learning_rate": 7.890725076164894e-05,
+      "loss": 1.6596,
+      "step": 227
+    },
+    {
+      "epoch": 0.12642084835042972,
+      "grad_norm": 0.21111659705638885,
+      "learning_rate": 7.889005470452845e-05,
+      "loss": 1.6158,
+      "step": 228
+    },
+    {
+      "epoch": 0.12697532575547546,
+      "grad_norm": 0.24391771852970123,
+      "learning_rate": 7.8872726301999e-05,
+      "loss": 1.6943,
+      "step": 229
+    },
+    {
+      "epoch": 0.1275298031605212,
+      "grad_norm": 0.2066636085510254,
+      "learning_rate": 7.885526561303024e-05,
+      "loss": 1.6627,
+      "step": 230
+    },
+    {
+      "epoch": 0.12808428056556695,
+      "grad_norm": 0.2221907377243042,
+      "learning_rate": 7.883767269704209e-05,
+      "loss": 1.6358,
+      "step": 231
+    },
+    {
+      "epoch": 0.1286387579706127,
+      "grad_norm": 0.21195903420448303,
+      "learning_rate": 7.881994761390437e-05,
+      "loss": 1.6259,
+      "step": 232
+    },
+    {
+      "epoch": 0.12919323537565844,
+      "grad_norm": 0.22241343557834625,
+      "learning_rate": 7.88020904239367e-05,
+      "loss": 1.7442,
+      "step": 233
+    },
+    {
+      "epoch": 0.12974771278070418,
+      "grad_norm": 0.20912249386310577,
+      "learning_rate": 7.878410118790827e-05,
+      "loss": 1.6546,
+      "step": 234
+    },
+    {
+      "epoch": 0.13030219018574993,
+      "grad_norm": 0.21954858303070068,
+      "learning_rate": 7.876597996703763e-05,
+      "loss": 1.6339,
+      "step": 235
+    },
+    {
+      "epoch": 0.13085666759079567,
+      "grad_norm": 0.23016227781772614,
+      "learning_rate": 7.874772682299251e-05,
+      "loss": 1.6333,
+      "step": 236
+    },
+    {
+      "epoch": 0.13141114499584142,
+      "grad_norm": 0.22401848435401917,
+      "learning_rate": 7.872934181788953e-05,
+      "loss": 1.7135,
+      "step": 237
+    },
+    {
+      "epoch": 0.13196562240088716,
+      "grad_norm": 0.23134422302246094,
+      "learning_rate": 7.871082501429409e-05,
+      "loss": 1.704,
+      "step": 238
+    },
+    {
+      "epoch": 0.1325200998059329,
+      "grad_norm": 0.24263045191764832,
+      "learning_rate": 7.869217647522006e-05,
+      "loss": 1.7007,
+      "step": 239
+    },
+    {
+      "epoch": 0.13307457721097865,
+      "grad_norm": 0.21756958961486816,
+      "learning_rate": 7.867339626412965e-05,
+      "loss": 1.6961,
+      "step": 240
+    },
+    {
+      "epoch": 0.1336290546160244,
+      "grad_norm": 0.23123596608638763,
+      "learning_rate": 7.865448444493317e-05,
+      "loss": 1.6698,
+      "step": 241
+    },
+    {
+      "epoch": 0.13418353202107014,
+      "grad_norm": 0.21551208198070526,
+      "learning_rate": 7.863544108198877e-05,
+      "loss": 1.6982,
+      "step": 242
+    },
+    {
+      "epoch": 0.13473800942611588,
+      "grad_norm": 0.22842957079410553,
+      "learning_rate": 7.861626624010226e-05,
+      "loss": 1.5909,
+      "step": 243
+    },
+    {
+      "epoch": 0.13529248683116163,
+      "grad_norm": 0.22002427279949188,
+      "learning_rate": 7.85969599845269e-05,
+      "loss": 1.649,
+      "step": 244
+    },
+    {
+      "epoch": 0.13584696423620737,
+      "grad_norm": 0.23170796036720276,
+      "learning_rate": 7.857752238096313e-05,
+      "loss": 1.685,
+      "step": 245
+    },
+    {
+      "epoch": 0.13640144164125312,
+      "grad_norm": 0.21300175786018372,
+      "learning_rate": 7.855795349555839e-05,
+      "loss": 1.6501,
+      "step": 246
+    },
+    {
+      "epoch": 0.13695591904629886,
+      "grad_norm": 0.2336280345916748,
+      "learning_rate": 7.853825339490689e-05,
+      "loss": 1.6482,
+      "step": 247
+    },
+    {
+      "epoch": 0.1375103964513446,
+      "grad_norm": 0.23497241735458374,
+      "learning_rate": 7.851842214604937e-05,
+      "loss": 1.5985,
+      "step": 248
+    },
+    {
+      "epoch": 0.13806487385639035,
+      "grad_norm": 0.21944530308246613,
+      "learning_rate": 7.849845981647285e-05,
+      "loss": 1.6658,
+      "step": 249
+    },
+    {
+      "epoch": 0.1386193512614361,
+      "grad_norm": 0.21688306331634521,
+      "learning_rate": 7.847836647411049e-05,
+      "loss": 1.6014,
+      "step": 250
+    },
+    {
+      "epoch": 0.13917382866648184,
+      "grad_norm": 0.21929487586021423,
+      "learning_rate": 7.84581421873412e-05,
+      "loss": 1.6299,
+      "step": 251
+    },
+    {
+      "epoch": 0.13972830607152759,
+      "grad_norm": 0.21958981454372406,
+      "learning_rate": 7.843778702498961e-05,
+      "loss": 1.6961,
+      "step": 252
+    },
+    {
+      "epoch": 0.14028278347657333,
+      "grad_norm": 0.21822874248027802,
+      "learning_rate": 7.841730105632563e-05,
+      "loss": 1.6779,
+      "step": 253
+    },
+    {
+      "epoch": 0.14083726088161908,
+      "grad_norm": 0.21482256054878235,
+      "learning_rate": 7.839668435106437e-05,
+      "loss": 1.6304,
+      "step": 254
+    },
+    {
+      "epoch": 0.14139173828666482,
+      "grad_norm": 0.22191710770130157,
+      "learning_rate": 7.837593697936582e-05,
+      "loss": 1.6461,
+      "step": 255
+    },
+    {
+      "epoch": 0.14194621569171056,
+      "grad_norm": 0.227654829621315,
+      "learning_rate": 7.835505901183468e-05,
+      "loss": 1.7217,
+      "step": 256
+    },
+    {
+      "epoch": 0.1425006930967563,
+      "grad_norm": 0.2576013505458832,
+      "learning_rate": 7.833405051952002e-05,
+      "loss": 1.7567,
+      "step": 257
+    },
+    {
+      "epoch": 0.14305517050180205,
+      "grad_norm": 0.20585085451602936,
+      "learning_rate": 7.831291157391513e-05,
+      "loss": 1.6422,
+      "step": 258
+    },
+    {
+      "epoch": 0.1436096479068478,
+      "grad_norm": 0.22537145018577576,
+      "learning_rate": 7.82916422469572e-05,
+      "loss": 1.6604,
+      "step": 259
+    },
+    {
+      "epoch": 0.14416412531189354,
+      "grad_norm": 0.23208315670490265,
+      "learning_rate": 7.827024261102718e-05,
+      "loss": 1.615,
+      "step": 260
+    },
+    {
+      "epoch": 0.1447186027169393,
+      "grad_norm": 0.2133701741695404,
+      "learning_rate": 7.824871273894943e-05,
+      "loss": 1.6413,
+      "step": 261
+    },
+    {
+      "epoch": 0.14527308012198503,
+      "grad_norm": 0.2303103655576706,
+      "learning_rate": 7.82270527039915e-05,
+      "loss": 1.661,
+      "step": 262
+    },
+    {
+      "epoch": 0.14582755752703078,
+      "grad_norm": 0.22526273131370544,
+      "learning_rate": 7.820526257986393e-05,
+      "loss": 1.6806,
+      "step": 263
+    },
+    {
+      "epoch": 0.14638203493207652,
+      "grad_norm": 0.22275319695472717,
+      "learning_rate": 7.818334244071994e-05,
+      "loss": 1.7037,
+      "step": 264
+    },
+    {
+      "epoch": 0.14693651233712227,
+      "grad_norm": 0.22790545225143433,
+      "learning_rate": 7.81612923611552e-05,
+      "loss": 1.654,
+      "step": 265
+    },
+    {
+      "epoch": 0.147490989742168,
+      "grad_norm": 0.22253917157649994,
+      "learning_rate": 7.813911241620755e-05,
+      "loss": 1.6492,
+      "step": 266
+    },
+    {
+      "epoch": 0.14804546714721376,
+      "grad_norm": 0.22290851175785065,
+      "learning_rate": 7.811680268135684e-05,
+      "loss": 1.6355,
+      "step": 267
+    },
+    {
+      "epoch": 0.1485999445522595,
+      "grad_norm": 0.22694170475006104,
+      "learning_rate": 7.809436323252456e-05,
+      "loss": 1.622,
+      "step": 268
+    },
+    {
+      "epoch": 0.14915442195730524,
+      "grad_norm": 0.21007151901721954,
+      "learning_rate": 7.80717941460736e-05,
+      "loss": 1.6312,
+      "step": 269
+    },
+    {
+      "epoch": 0.149708899362351,
+      "grad_norm": 0.2363370954990387,
+      "learning_rate": 7.804909549880806e-05,
+      "loss": 1.6222,
+      "step": 270
+    },
+    {
+      "epoch": 0.15026337676739673,
+      "grad_norm": 0.26099053025245667,
+      "learning_rate": 7.802626736797292e-05,
+      "loss": 1.6424,
+      "step": 271
+    },
+    {
+      "epoch": 0.15081785417244248,
+      "grad_norm": 0.21376170217990875,
+      "learning_rate": 7.800330983125381e-05,
+      "loss": 1.6055,
+      "step": 272
+    },
+    {
+      "epoch": 0.15137233157748822,
+      "grad_norm": 0.26935890316963196,
+      "learning_rate": 7.798022296677675e-05,
+      "loss": 1.6858,
+      "step": 273
+    },
+    {
+      "epoch": 0.15192680898253397,
+      "grad_norm": 0.22003325819969177,
+      "learning_rate": 7.795700685310783e-05,
+      "loss": 1.6803,
+      "step": 274
+    },
+    {
+      "epoch": 0.1524812863875797,
+      "grad_norm": 0.25127747654914856,
+      "learning_rate": 7.793366156925302e-05,
+      "loss": 1.5456,
+      "step": 275
+    },
+    {
+      "epoch": 0.15303576379262546,
+      "grad_norm": 0.21749310195446014,
+      "learning_rate": 7.791018719465785e-05,
+      "loss": 1.6904,
+      "step": 276
+    },
+    {
+      "epoch": 0.1535902411976712,
+      "grad_norm": 0.2340991199016571,
+      "learning_rate": 7.788658380920716e-05,
+      "loss": 1.6946,
+      "step": 277
+    },
+    {
+      "epoch": 0.15414471860271695,
+      "grad_norm": 0.21612657606601715,
+      "learning_rate": 7.786285149322483e-05,
+      "loss": 1.5836,
+      "step": 278
+    },
+    {
+      "epoch": 0.1546991960077627,
+      "grad_norm": 0.2601798474788666,
+      "learning_rate": 7.783899032747346e-05,
+      "loss": 1.679,
+      "step": 279
+    },
+    {
+      "epoch": 0.15525367341280844,
+      "grad_norm": 0.24051488935947418,
+      "learning_rate": 7.78150003931542e-05,
+      "loss": 1.6504,
+      "step": 280
+    },
+    {
+      "epoch": 0.15580815081785418,
+      "grad_norm": 0.22417373955249786,
+      "learning_rate": 7.779088177190636e-05,
+      "loss": 1.6767,
+      "step": 281
+    },
+    {
+      "epoch": 0.15636262822289992,
+      "grad_norm": 0.23958833515644073,
+      "learning_rate": 7.776663454580718e-05,
+      "loss": 1.6636,
+      "step": 282
+    },
+    {
+      "epoch": 0.15691710562794567,
+      "grad_norm": 0.22947099804878235,
+      "learning_rate": 7.774225879737156e-05,
+      "loss": 1.6653,
+      "step": 283
+    },
+    {
+      "epoch": 0.1574715830329914,
+      "grad_norm": 0.23993003368377686,
+      "learning_rate": 7.771775460955178e-05,
+      "loss": 1.6637,
+      "step": 284
+    },
+    {
+      "epoch": 0.15802606043803716,
+      "grad_norm": 0.2175922393798828,
+      "learning_rate": 7.76931220657372e-05,
+      "loss": 1.6086,
+      "step": 285
+    },
+    {
+      "epoch": 0.1585805378430829,
+      "grad_norm": 0.22795039415359497,
+      "learning_rate": 7.766836124975399e-05,
+      "loss": 1.6274,
+      "step": 286
+    },
+    {
+      "epoch": 0.15913501524812865,
+      "grad_norm": 0.2554066479206085,
+      "learning_rate": 7.764347224586482e-05,
+      "loss": 1.6059,
+      "step": 287
+    },
+    {
+      "epoch": 0.1596894926531744,
+      "grad_norm": 0.2622605860233307,
+      "learning_rate": 7.761845513876861e-05,
+      "loss": 1.6815,
+      "step": 288
+    },
+    {
+      "epoch": 0.16024397005822014,
+      "grad_norm": 0.25482630729675293,
+      "learning_rate": 7.759331001360021e-05,
+      "loss": 1.7245,
+      "step": 289
+    },
+    {
+      "epoch": 0.16079844746326588,
+      "grad_norm": 0.23500695824623108,
+      "learning_rate": 7.756803695593015e-05,
+      "loss": 1.6039,
+      "step": 290
+    },
+    {
+      "epoch": 0.16135292486831163,
+      "grad_norm": 0.24643713235855103,
+      "learning_rate": 7.754263605176429e-05,
+      "loss": 1.6804,
+      "step": 291
+    },
+    {
+      "epoch": 0.16190740227335737,
+      "grad_norm": 0.301843523979187,
+      "learning_rate": 7.751710738754357e-05,
+      "loss": 1.6837,
+      "step": 292
+    },
+    {
+      "epoch": 0.16246187967840311,
+      "grad_norm": 0.22402901947498322,
+      "learning_rate": 7.749145105014372e-05,
+      "loss": 1.7017,
+      "step": 293
+    },
+    {
+      "epoch": 0.16301635708344886,
+      "grad_norm": 0.30275362730026245,
+      "learning_rate": 7.746566712687493e-05,
+      "loss": 1.7249,
+      "step": 294
+    },
+    {
+      "epoch": 0.1635708344884946,
+      "grad_norm": 0.20885106921195984,
+      "learning_rate": 7.74397557054816e-05,
+      "loss": 1.6964,
+      "step": 295
+    },
+    {
+      "epoch": 0.16412531189354035,
+      "grad_norm": 0.28100892901420593,
+      "learning_rate": 7.741371687414198e-05,
+      "loss": 1.6428,
+      "step": 296
+    },
+    {
+      "epoch": 0.1646797892985861,
+      "grad_norm": 0.24655266106128693,
+      "learning_rate": 7.738755072146794e-05,
+      "loss": 1.638,
+      "step": 297
+    },
+    {
+      "epoch": 0.16523426670363184,
+      "grad_norm": 0.2542823255062103,
+      "learning_rate": 7.736125733650461e-05,
+      "loss": 1.6958,
+      "step": 298
+    },
+    {
+      "epoch": 0.16578874410867758,
+      "grad_norm": 0.24665483832359314,
+      "learning_rate": 7.733483680873009e-05,
+      "loss": 1.6095,
+      "step": 299
+    },
+    {
+      "epoch": 0.16634322151372333,
+      "grad_norm": 0.25384095311164856,
+      "learning_rate": 7.730828922805517e-05,
+      "loss": 1.6943,
+      "step": 300
+    },
+    {
+      "epoch": 0.16689769891876907,
+      "grad_norm": 0.26568636298179626,
+      "learning_rate": 7.728161468482304e-05,
+      "loss": 1.693,
+      "step": 301
+    },
+    {
+      "epoch": 0.16745217632381482,
+      "grad_norm": 0.2455059140920639,
+      "learning_rate": 7.72548132698089e-05,
+      "loss": 1.6415,
+      "step": 302
+    },
+    {
+      "epoch": 0.16800665372886056,
+      "grad_norm": 0.24293363094329834,
+      "learning_rate": 7.722788507421971e-05,
+      "loss": 1.6402,
+      "step": 303
+    },
+    {
+      "epoch": 0.1685611311339063,
+      "grad_norm": 0.22879062592983246,
+      "learning_rate": 7.720083018969393e-05,
+      "loss": 1.6153,
+      "step": 304
+    },
+    {
+      "epoch": 0.16911560853895205,
+      "grad_norm": 0.21951867640018463,
+      "learning_rate": 7.717364870830107e-05,
+      "loss": 1.6482,
+      "step": 305
+    },
+    {
+      "epoch": 0.1696700859439978,
+      "grad_norm": 0.2079135924577713,
+      "learning_rate": 7.71463407225415e-05,
+      "loss": 1.5899,
+      "step": 306
+    },
+    {
+      "epoch": 0.17022456334904354,
+      "grad_norm": 0.20707400143146515,
+      "learning_rate": 7.71189063253461e-05,
+      "loss": 1.6667,
+      "step": 307
+    },
+    {
+      "epoch": 0.17077904075408928,
+      "grad_norm": 0.22458939254283905,
+      "learning_rate": 7.709134561007592e-05,
+      "loss": 1.5537,
+      "step": 308
+    },
+    {
+      "epoch": 0.17133351815913503,
+      "grad_norm": 0.21969960629940033,
+      "learning_rate": 7.706365867052188e-05,
+      "loss": 1.6693,
+      "step": 309
+    },
+    {
+      "epoch": 0.17188799556418077,
+      "grad_norm": 0.22047635912895203,
+      "learning_rate": 7.703584560090447e-05,
+      "loss": 1.5548,
+      "step": 310
+    },
+    {
+      "epoch": 0.1724424729692265,
+      "grad_norm": 0.22892658412456512,
+      "learning_rate": 7.700790649587336e-05,
+      "loss": 1.6342,
+      "step": 311
+    },
+    {
+      "epoch": 0.17299695037427223,
+      "grad_norm": 0.23613335192203522,
+      "learning_rate": 7.697984145050718e-05,
+      "loss": 1.6324,
+      "step": 312
+    },
+    {
+      "epoch": 0.17355142777931798,
+      "grad_norm": 0.2484220266342163,
+      "learning_rate": 7.695165056031313e-05,
+      "loss": 1.6467,
+      "step": 313
+    },
+    {
+      "epoch": 0.17410590518436372,
+      "grad_norm": 0.2812964618206024,
+      "learning_rate": 7.692333392122663e-05,
+      "loss": 1.6744,
+      "step": 314
+    },
+    {
+      "epoch": 0.17466038258940947,
+      "grad_norm": 0.2211124300956726,
+      "learning_rate": 7.689489162961109e-05,
+      "loss": 1.6213,
+      "step": 315
+    },
+    {
+      "epoch": 0.1752148599944552,
+      "grad_norm": 0.22334423661231995,
+      "learning_rate": 7.686632378225748e-05,
+      "loss": 1.6909,
+      "step": 316
+    },
+    {
+      "epoch": 0.17576933739950096,
+      "grad_norm": 0.2386290282011032,
+      "learning_rate": 7.683763047638407e-05,
+      "loss": 1.6193,
+      "step": 317
+    },
+    {
+      "epoch": 0.1763238148045467,
+      "grad_norm": 0.20850642025470734,
+      "learning_rate": 7.680881180963605e-05,
+      "loss": 1.6541,
+      "step": 318
+    },
+    {
+      "epoch": 0.17687829220959245,
+      "grad_norm": 0.22296078503131866,
+      "learning_rate": 7.677986788008524e-05,
+      "loss": 1.6285,
+      "step": 319
+    },
+    {
+      "epoch": 0.1774327696146382,
+      "grad_norm": 0.20673969388008118,
+      "learning_rate": 7.675079878622974e-05,
+      "loss": 1.6835,
+      "step": 320
+    },
+    {
+      "epoch": 0.17798724701968394,
+      "grad_norm": 0.21489155292510986,
+      "learning_rate": 7.672160462699359e-05,
+      "loss": 1.6354,
+      "step": 321
+    },
+    {
+      "epoch": 0.17854172442472968,
+      "grad_norm": 0.21602220833301544,
+      "learning_rate": 7.669228550172639e-05,
+      "loss": 1.5907,
+      "step": 322
+    },
+    {
+      "epoch": 0.17909620182977543,
+      "grad_norm": 0.2056579738855362,
+      "learning_rate": 7.666284151020309e-05,
+      "loss": 1.6304,
+      "step": 323
+    },
+    {
+      "epoch": 0.17965067923482117,
+      "grad_norm": 0.22408543527126312,
+      "learning_rate": 7.663327275262353e-05,
+      "loss": 1.636,
+      "step": 324
+    },
+    {
+      "epoch": 0.18020515663986691,
+      "grad_norm": 0.20583994686603546,
+      "learning_rate": 7.66035793296121e-05,
+      "loss": 1.6401,
+      "step": 325
+    },
+    {
+      "epoch": 0.18075963404491266,
+      "grad_norm": 0.21680684387683868,
+      "learning_rate": 7.657376134221749e-05,
+      "loss": 1.6827,
+      "step": 326
+    },
+    {
+      "epoch": 0.1813141114499584,
+      "grad_norm": 0.218284010887146,
+      "learning_rate": 7.654381889191225e-05,
+      "loss": 1.585,
+      "step": 327
+    },
+    {
+      "epoch": 0.18186858885500415,
+      "grad_norm": 0.20787712931632996,
+      "learning_rate": 7.651375208059252e-05,
+      "loss": 1.6359,
+      "step": 328
+    },
+    {
+      "epoch": 0.1824230662600499,
+      "grad_norm": 0.216603085398674,
+      "learning_rate": 7.648356101057764e-05,
+      "loss": 1.6884,
+      "step": 329
+    },
+    {
+      "epoch": 0.18297754366509564,
+      "grad_norm": 0.2224433422088623,
+      "learning_rate": 7.645324578460978e-05,
+      "loss": 1.6869,
+      "step": 330
+    },
+    {
+      "epoch": 0.18353202107014138,
+      "grad_norm": 0.21461333334445953,
+      "learning_rate": 7.642280650585366e-05,
+      "loss": 1.5748,
+      "step": 331
+    },
+    {
+      "epoch": 0.18408649847518713,
+      "grad_norm": 0.20232297480106354,
+      "learning_rate": 7.639224327789613e-05,
+      "loss": 1.6181,
+      "step": 332
+    },
+    {
+      "epoch": 0.18464097588023287,
+      "grad_norm": 0.2159082591533661,
+      "learning_rate": 7.636155620474589e-05,
+      "loss": 1.6346,
+      "step": 333
+    },
+    {
+      "epoch": 0.18519545328527862,
+      "grad_norm": 0.21551144123077393,
+      "learning_rate": 7.633074539083302e-05,
+      "loss": 1.6951,
+      "step": 334
+    },
+    {
+      "epoch": 0.18574993069032436,
+      "grad_norm": 0.2233298122882843,
+      "learning_rate": 7.629981094100878e-05,
+      "loss": 1.6491,
+      "step": 335
+    },
+    {
+      "epoch": 0.1863044080953701,
+      "grad_norm": 0.21482495963573456,
+      "learning_rate": 7.626875296054512e-05,
+      "loss": 1.6596,
+      "step": 336
+    },
+    {
+      "epoch": 0.18685888550041585,
+      "grad_norm": 0.2133273333311081,
+      "learning_rate": 7.623757155513439e-05,
+      "loss": 1.6508,
+      "step": 337
+    },
+    {
+      "epoch": 0.1874133629054616,
+      "grad_norm": 0.20683740079402924,
+      "learning_rate": 7.620626683088894e-05,
+      "loss": 1.6145,
+      "step": 338
+    },
+    {
+      "epoch": 0.18796784031050734,
+      "grad_norm": 0.20707263052463531,
+      "learning_rate": 7.617483889434083e-05,
+      "loss": 1.6415,
+      "step": 339
+    },
+    {
+      "epoch": 0.18852231771555308,
+      "grad_norm": 0.2513361871242523,
+      "learning_rate": 7.614328785244138e-05,
+      "loss": 1.7166,
+      "step": 340
+    },
+    {
+      "epoch": 0.18907679512059883,
+      "grad_norm": 0.21265187859535217,
+      "learning_rate": 7.611161381256084e-05,
+      "loss": 1.6454,
+      "step": 341
+    },
+    {
+      "epoch": 0.18963127252564457,
+      "grad_norm": 0.28207194805145264,
+      "learning_rate": 7.607981688248807e-05,
+      "loss": 1.6473,
+      "step": 342
+    },
+    {
+      "epoch": 0.19018574993069032,
+      "grad_norm": 0.256237268447876,
+      "learning_rate": 7.604789717043011e-05,
+      "loss": 1.7602,
+      "step": 343
+    },
+    {
+      "epoch": 0.19074022733573606,
+      "grad_norm": 0.2375311702489853,
+      "learning_rate": 7.601585478501181e-05,
+      "loss": 1.6399,
+      "step": 344
+    },
+    {
+      "epoch": 0.1912947047407818,
+      "grad_norm": 0.23850996792316437,
+      "learning_rate": 7.598368983527554e-05,
+      "loss": 1.6538,
+      "step": 345
+    },
+    {
+      "epoch": 0.19184918214582755,
+      "grad_norm": 0.2116571068763733,
+      "learning_rate": 7.595140243068072e-05,
+      "loss": 1.6129,
+      "step": 346
+    },
+    {
+      "epoch": 0.1924036595508733,
+      "grad_norm": 0.2089197337627411,
+      "learning_rate": 7.591899268110352e-05,
+      "loss": 1.6216,
+      "step": 347
+    },
+    {
+      "epoch": 0.19295813695591904,
+      "grad_norm": 0.20319993793964386,
+      "learning_rate": 7.588646069683642e-05,
+      "loss": 1.6086,
+      "step": 348
+    },
+    {
+      "epoch": 0.19351261436096479,
+      "grad_norm": 0.2238406538963318,
+      "learning_rate": 7.585380658858793e-05,
+      "loss": 1.6992,
+      "step": 349
+    },
+    {
+      "epoch": 0.19406709176601053,
+      "grad_norm": 0.21192164719104767,
+      "learning_rate": 7.58210304674821e-05,
+      "loss": 1.6756,
+      "step": 350
+    },
+    {
+      "epoch": 0.19462156917105627,
+      "grad_norm": 0.21115197241306305,
+      "learning_rate": 7.578813244505823e-05,
+      "loss": 1.5859,
+      "step": 351
+    },
+    {
+      "epoch": 0.19517604657610202,
+      "grad_norm": 0.20488910377025604,
+      "learning_rate": 7.575511263327044e-05,
+      "loss": 1.6174,
+      "step": 352
+    },
+    {
+      "epoch": 0.19573052398114776,
+      "grad_norm": 0.21583472192287445,
+      "learning_rate": 7.572197114448733e-05,
+      "loss": 1.639,
+      "step": 353
+    },
+    {
+      "epoch": 0.1962850013861935,
+      "grad_norm": 0.22085356712341309,
+      "learning_rate": 7.568870809149155e-05,
+      "loss": 1.7244,
+      "step": 354
+    },
+    {
+      "epoch": 0.19683947879123925,
+      "grad_norm": 0.20893336832523346,
+      "learning_rate": 7.565532358747949e-05,
+      "loss": 1.6678,
+      "step": 355
+    },
+    {
+      "epoch": 0.197393956196285,
+      "grad_norm": 0.20535239577293396,
+      "learning_rate": 7.562181774606075e-05,
+      "loss": 1.5594,
+      "step": 356
+    },
+    {
+      "epoch": 0.19794843360133074,
+      "grad_norm": 0.23027966916561127,
+      "learning_rate": 7.558819068125796e-05,
+      "loss": 1.7349,
+      "step": 357
+    },
+    {
+      "epoch": 0.1985029110063765,
+      "grad_norm": 0.20420025289058685,
+      "learning_rate": 7.555444250750618e-05,
+      "loss": 1.6259,
+      "step": 358
+    },
+    {
+      "epoch": 0.19905738841142223,
+      "grad_norm": 0.21578669548034668,
+      "learning_rate": 7.552057333965271e-05,
+      "loss": 1.652,
+      "step": 359
+    },
+    {
+      "epoch": 0.19961186581646798,
+      "grad_norm": 0.21920384466648102,
+      "learning_rate": 7.548658329295651e-05,
+      "loss": 1.6564,
+      "step": 360
+    },
+    {
+      "epoch": 0.20016634322151372,
+      "grad_norm": 0.23153354227542877,
+      "learning_rate": 7.545247248308798e-05,
+      "loss": 1.6196,
+      "step": 361
+    },
+    {
+      "epoch": 0.20072082062655947,
+      "grad_norm": 0.22002752125263214,
+      "learning_rate": 7.541824102612839e-05,
+      "loss": 1.6524,
+      "step": 362
+    },
+    {
+      "epoch": 0.2012752980316052,
+      "grad_norm": 0.21228551864624023,
+      "learning_rate": 7.53838890385697e-05,
+      "loss": 1.7047,
+      "step": 363
+    },
+    {
+      "epoch": 0.20182977543665095,
+      "grad_norm": 0.21426419913768768,
+      "learning_rate": 7.534941663731394e-05,
+      "loss": 1.6433,
+      "step": 364
+    },
+    {
+      "epoch": 0.2023842528416967,
+      "grad_norm": 0.20719939470291138,
+      "learning_rate": 7.531482393967295e-05,
+      "loss": 1.6065,
+      "step": 365
+    },
+    {
+      "epoch": 0.20293873024674244,
+      "grad_norm": 0.21919012069702148,
+      "learning_rate": 7.528011106336797e-05,
+      "loss": 1.6418,
+      "step": 366
+    },
+    {
+      "epoch": 0.2034932076517882,
+      "grad_norm": 0.201614648103714,
+      "learning_rate": 7.524527812652917e-05,
+      "loss": 1.5683,
+      "step": 367
+    },
+    {
+      "epoch": 0.20404768505683393,
+      "grad_norm": 0.20898142457008362,
+      "learning_rate": 7.521032524769537e-05,
+      "loss": 1.6671,
+      "step": 368
+    },
+    {
+      "epoch": 0.20460216246187968,
+      "grad_norm": 0.2096984088420868,
+      "learning_rate": 7.517525254581346e-05,
+      "loss": 1.5993,
+      "step": 369
+    },
+    {
+      "epoch": 0.20515663986692542,
+      "grad_norm": 0.21319280564785004,
+      "learning_rate": 7.514006014023817e-05,
+      "loss": 1.5941,
+      "step": 370
+    },
+    {
+      "epoch": 0.20571111727197117,
+      "grad_norm": 0.2070310115814209,
+      "learning_rate": 7.510474815073157e-05,
+      "loss": 1.6734,
+      "step": 371
+    },
+    {
+      "epoch": 0.2062655946770169,
+      "grad_norm": 0.20351865887641907,
+      "learning_rate": 7.506931669746266e-05,
+      "loss": 1.6756,
+      "step": 372
+    },
+    {
+      "epoch": 0.20682007208206266,
+      "grad_norm": 0.20868165791034698,
+      "learning_rate": 7.503376590100702e-05,
+      "loss": 1.6391,
+      "step": 373
+    },
+    {
+      "epoch": 0.2073745494871084,
+      "grad_norm": 0.24216507375240326,
+      "learning_rate": 7.499809588234634e-05,
+      "loss": 1.7327,
+      "step": 374
+    },
+    {
+      "epoch": 0.20792902689215415,
+      "grad_norm": 0.21316882967948914,
+      "learning_rate": 7.496230676286802e-05,
+      "loss": 1.6154,
+      "step": 375
+    },
+    {
+      "epoch": 0.2084835042971999,
+      "grad_norm": 0.23029237985610962,
+      "learning_rate": 7.492639866436479e-05,
+      "loss": 1.6068,
+      "step": 376
+    },
+    {
+      "epoch": 0.20903798170224563,
+      "grad_norm": 0.21493342518806458,
+      "learning_rate": 7.489037170903429e-05,
+      "loss": 1.6474,
+      "step": 377
+    },
+    {
+      "epoch": 0.20959245910729138,
+      "grad_norm": 0.2048824578523636,
+      "learning_rate": 7.485422601947858e-05,
+      "loss": 1.5741,
+      "step": 378
+    },
+    {
+      "epoch": 0.21014693651233712,
+      "grad_norm": 0.20953327417373657,
+      "learning_rate": 7.481796171870383e-05,
+      "loss": 1.609,
+      "step": 379
+    },
+    {
+      "epoch": 0.21070141391738287,
+      "grad_norm": 0.21259397268295288,
+      "learning_rate": 7.478157893011984e-05,
+      "loss": 1.681,
+      "step": 380
+    },
+    {
+      "epoch": 0.2112558913224286,
+      "grad_norm": 0.2035396844148636,
+      "learning_rate": 7.474507777753962e-05,
+      "loss": 1.5764,
+      "step": 381
+    },
+    {
+      "epoch": 0.21181036872747436,
+      "grad_norm": 0.21311511099338531,
+      "learning_rate": 7.470845838517899e-05,
+      "loss": 1.6694,
+      "step": 382
+    },
+    {
+      "epoch": 0.2123648461325201,
+      "grad_norm": 0.21380078792572021,
+      "learning_rate": 7.467172087765616e-05,
+      "loss": 1.6693,
+      "step": 383
+    },
+    {
+      "epoch": 0.21291932353756585,
+      "grad_norm": 0.20704282820224762,
+      "learning_rate": 7.463486537999125e-05,
+      "loss": 1.6214,
+      "step": 384
+    },
+    {
+      "epoch": 0.2134738009426116,
+      "grad_norm": 0.20259855687618256,
+      "learning_rate": 7.459789201760596e-05,
+      "loss": 1.6286,
+      "step": 385
+    },
+    {
+      "epoch": 0.21402827834765734,
+      "grad_norm": 0.2159855216741562,
+      "learning_rate": 7.456080091632305e-05,
+      "loss": 1.6627,
+      "step": 386
+    },
+    {
+      "epoch": 0.21458275575270308,
+      "grad_norm": 0.21545730531215668,
+      "learning_rate": 7.452359220236601e-05,
+      "loss": 1.6175,
+      "step": 387
+    },
+    {
+      "epoch": 0.21513723315774883,
+      "grad_norm": 0.21161332726478577,
+      "learning_rate": 7.44862660023585e-05,
+      "loss": 1.6789,
+      "step": 388
+    },
+    {
+      "epoch": 0.21569171056279457,
+      "grad_norm": 0.2151462882757187,
+      "learning_rate": 7.444882244332403e-05,
+      "loss": 1.6574,
+      "step": 389
+    },
+    {
+      "epoch": 0.21624618796784031,
+      "grad_norm": 0.23342940211296082,
+      "learning_rate": 7.441126165268552e-05,
+      "loss": 1.6155,
+      "step": 390
+    },
+    {
+      "epoch": 0.21680066537288606,
+      "grad_norm": 0.2156396359205246,
+      "learning_rate": 7.437358375826476e-05,
+      "loss": 1.6173,
+      "step": 391
+    },
+    {
+      "epoch": 0.2173551427779318,
+      "grad_norm": 0.2337697595357895,
+      "learning_rate": 7.433578888828215e-05,
+      "loss": 1.6365,
+      "step": 392
+    },
+    {
+      "epoch": 0.21790962018297755,
+      "grad_norm": 0.21427282691001892,
+      "learning_rate": 7.429787717135608e-05,
+      "loss": 1.6423,
+      "step": 393
+    },
+    {
+      "epoch": 0.2184640975880233,
+      "grad_norm": 0.23731502890586853,
+      "learning_rate": 7.425984873650262e-05,
+      "loss": 1.6152,
+      "step": 394
+    },
+    {
+      "epoch": 0.21901857499306904,
+      "grad_norm": 0.21108436584472656,
+      "learning_rate": 7.422170371313501e-05,
+      "loss": 1.6023,
+      "step": 395
+    },
+    {
+      "epoch": 0.21957305239811478,
+      "grad_norm": 0.24385766685009003,
+      "learning_rate": 7.418344223106331e-05,
+      "loss": 1.6968,
+      "step": 396
+    },
+    {
+      "epoch": 0.22012752980316053,
+      "grad_norm": 0.22708410024642944,
+      "learning_rate": 7.414506442049382e-05,
+      "loss": 1.6314,
+      "step": 397
+    },
+    {
+      "epoch": 0.22068200720820627,
+      "grad_norm": 0.2248944193124771,
+      "learning_rate": 7.410657041202877e-05,
+      "loss": 1.6085,
+      "step": 398
+    },
+    {
+      "epoch": 0.22123648461325202,
+      "grad_norm": 0.2051655650138855,
+      "learning_rate": 7.406796033666577e-05,
+      "loss": 1.6075,
+      "step": 399
+    },
+    {
+      "epoch": 0.22179096201829776,
+      "grad_norm": 0.24941745400428772,
+      "learning_rate": 7.402923432579749e-05,
+      "loss": 1.7027,
+      "step": 400
+    },
+    {
+      "epoch": 0.2223454394233435,
+      "grad_norm": 0.20685605704784393,
+      "learning_rate": 7.399039251121104e-05,
+      "loss": 1.5774,
+      "step": 401
+    },
+    {
+      "epoch": 0.22289991682838925,
+      "grad_norm": 0.25279372930526733,
+      "learning_rate": 7.395143502508767e-05,
+      "loss": 1.659,
+      "step": 402
+    },
+    {
+      "epoch": 0.223454394233435,
+      "grad_norm": 0.21988433599472046,
+      "learning_rate": 7.391236200000227e-05,
+      "loss": 1.6034,
+      "step": 403
+    },
+    {
+      "epoch": 0.22400887163848074,
+      "grad_norm": 0.25263452529907227,
+      "learning_rate": 7.387317356892294e-05,
+      "loss": 1.6467,
+      "step": 404
+    },
+    {
+      "epoch": 0.22456334904352648,
+      "grad_norm": 0.2200496941804886,
+      "learning_rate": 7.383386986521044e-05,
+      "loss": 1.5981,
+      "step": 405
+    },
+    {
+      "epoch": 0.22511782644857223,
+      "grad_norm": 0.22581440210342407,
+      "learning_rate": 7.379445102261787e-05,
+      "loss": 1.6179,
+      "step": 406
+    },
+    {
+      "epoch": 0.22567230385361797,
+      "grad_norm": 0.2082045078277588,
+      "learning_rate": 7.375491717529014e-05,
+      "loss": 1.6553,
+      "step": 407
+    },
+    {
+      "epoch": 0.22622678125866372,
+      "grad_norm": 0.21761249005794525,
+      "learning_rate": 7.371526845776351e-05,
+      "loss": 1.5159,
+      "step": 408
+    },
+    {
+      "epoch": 0.22678125866370946,
+      "grad_norm": 0.22059603035449982,
+      "learning_rate": 7.36755050049652e-05,
+      "loss": 1.6327,
+      "step": 409
+    },
+    {
+      "epoch": 0.2273357360687552,
+      "grad_norm": 0.21581657230854034,
+      "learning_rate": 7.363562695221285e-05,
+      "loss": 1.6646,
+      "step": 410
+    },
+    {
+      "epoch": 0.22789021347380095,
+      "grad_norm": 0.21198253333568573,
+      "learning_rate": 7.359563443521407e-05,
+      "loss": 1.6374,
+      "step": 411
+    },
+    {
+      "epoch": 0.2284446908788467,
+      "grad_norm": 0.21386800706386566,
+      "learning_rate": 7.3555527590066e-05,
+      "loss": 1.6085,
+      "step": 412
+    },
+    {
+      "epoch": 0.22899916828389244,
+      "grad_norm": 0.21344758570194244,
+      "learning_rate": 7.351530655325492e-05,
+      "loss": 1.6691,
+      "step": 413
+    },
+    {
+      "epoch": 0.22955364568893818,
+      "grad_norm": 0.21234236657619476,
+      "learning_rate": 7.347497146165562e-05,
+      "loss": 1.5652,
+      "step": 414
+    },
+    {
+      "epoch": 0.23010812309398393,
+      "grad_norm": 0.21020184457302094,
+      "learning_rate": 7.343452245253108e-05,
+      "loss": 1.6386,
+      "step": 415
+    },
+    {
+      "epoch": 0.23066260049902967,
+      "grad_norm": 0.22808673977851868,
+      "learning_rate": 7.339395966353193e-05,
+      "loss": 1.6169,
+      "step": 416
+    },
+    {
+      "epoch": 0.23121707790407542,
+      "grad_norm": 0.21194703876972198,
+      "learning_rate": 7.335328323269599e-05,
+      "loss": 1.6279,
+      "step": 417
+    },
+    {
+      "epoch": 0.23177155530912116,
+      "grad_norm": 0.2219015657901764,
+      "learning_rate": 7.331249329844784e-05,
+      "loss": 1.6187,
+      "step": 418
+    },
+    {
+      "epoch": 0.2323260327141669,
+      "grad_norm": 0.2099056988954544,
+      "learning_rate": 7.327158999959831e-05,
+      "loss": 1.6131,
+      "step": 419
+    },
+    {
+      "epoch": 0.23288051011921265,
+      "grad_norm": 0.2129475176334381,
+      "learning_rate": 7.323057347534401e-05,
+      "loss": 1.7079,
+      "step": 420
+    },
+    {
+      "epoch": 0.2334349875242584,
+      "grad_norm": 0.20943042635917664,
+      "learning_rate": 7.318944386526683e-05,
+      "loss": 1.6559,
+      "step": 421
+    },
+    {
+      "epoch": 0.23398946492930414,
+      "grad_norm": 0.21280522644519806,
+      "learning_rate": 7.314820130933358e-05,
+      "loss": 1.6158,
+      "step": 422
+    },
+    {
+      "epoch": 0.2345439423343499,
+      "grad_norm": 0.21108831465244293,
+      "learning_rate": 7.310684594789535e-05,
+      "loss": 1.5913,
+      "step": 423
+    },
+    {
+      "epoch": 0.23509841973939563,
+      "grad_norm": 0.21344444155693054,
+      "learning_rate": 7.306537792168717e-05,
+      "loss": 1.6157,
+      "step": 424
+    },
+    {
+      "epoch": 0.23565289714444138,
+      "grad_norm": 0.20498408377170563,
+      "learning_rate": 7.302379737182746e-05,
+      "loss": 1.6026,
+      "step": 425
+    },
+    {
+      "epoch": 0.23620737454948712,
+      "grad_norm": 0.21187523007392883,
+      "learning_rate": 7.298210443981754e-05,
+      "loss": 1.6558,
+      "step": 426
+    },
+    {
+      "epoch": 0.23676185195453286,
+      "grad_norm": 0.2079460322856903,
+      "learning_rate": 7.29402992675412e-05,
+      "loss": 1.7147,
+      "step": 427
+    },
+    {
+      "epoch": 0.2373163293595786,
+      "grad_norm": 0.22233690321445465,
+      "learning_rate": 7.289838199726419e-05,
+      "loss": 1.681,
+      "step": 428
+    },
+    {
+      "epoch": 0.23787080676462435,
+      "grad_norm": 0.21120913326740265,
+      "learning_rate": 7.285635277163373e-05,
+      "loss": 1.6054,
+      "step": 429
+    },
+    {
+      "epoch": 0.2384252841696701,
+      "grad_norm": 0.22804483771324158,
+      "learning_rate": 7.281421173367805e-05,
+      "loss": 1.716,
+      "step": 430
+    },
+    {
+      "epoch": 0.23897976157471584,
+      "grad_norm": 0.21392664313316345,
+      "learning_rate": 7.277195902680584e-05,
+      "loss": 1.6429,
+      "step": 431
+    },
+    {
+      "epoch": 0.2395342389797616,
+      "grad_norm": 0.21854659914970398,
+      "learning_rate": 7.272959479480584e-05,
+      "loss": 1.6437,
+      "step": 432
+    },
+    {
+      "epoch": 0.24008871638480733,
+      "grad_norm": 0.21218635141849518,
+      "learning_rate": 7.268711918184635e-05,
+      "loss": 1.7084,
+      "step": 433
+    },
+    {
+      "epoch": 0.24064319378985308,
+      "grad_norm": 0.21876314282417297,
+      "learning_rate": 7.26445323324746e-05,
+      "loss": 1.6482,
+      "step": 434
+    },
+    {
+      "epoch": 0.24119767119489882,
+      "grad_norm": 0.2061976045370102,
+      "learning_rate": 7.260183439161651e-05,
+      "loss": 1.6213,
+      "step": 435
+    },
+    {
+      "epoch": 0.24175214859994454,
+      "grad_norm": 0.22555728256702423,
+      "learning_rate": 7.255902550457592e-05,
+      "loss": 1.7004,
+      "step": 436
+    },
+    {
+      "epoch": 0.24230662600499028,
+      "grad_norm": 0.20105816423892975,
+      "learning_rate": 7.251610581703432e-05,
+      "loss": 1.5968,
+      "step": 437
+    },
+    {
+      "epoch": 0.24286110341003603,
+      "grad_norm": 0.22831952571868896,
+      "learning_rate": 7.24730754750502e-05,
+      "loss": 1.6453,
+      "step": 438
+    },
+    {
+      "epoch": 0.24341558081508177,
+      "grad_norm": 0.20313981175422668,
+      "learning_rate": 7.242993462505861e-05,
+      "loss": 1.6452,
+      "step": 439
+    },
+    {
+      "epoch": 0.24397005822012752,
+      "grad_norm": 0.21168267726898193,
+      "learning_rate": 7.238668341387078e-05,
+      "loss": 1.6479,
+      "step": 440
+    },
+    {
+      "epoch": 0.24452453562517326,
+      "grad_norm": 0.20957493782043457,
+      "learning_rate": 7.234332198867334e-05,
+      "loss": 1.63,
+      "step": 441
+    },
+    {
+      "epoch": 0.245079013030219,
+      "grad_norm": 0.2083057314157486,
+      "learning_rate": 7.22998504970281e-05,
+      "loss": 1.6362,
+      "step": 442
+    },
+    {
+      "epoch": 0.24563349043526475,
+      "grad_norm": 0.23519518971443176,
+      "learning_rate": 7.22562690868714e-05,
+      "loss": 1.7074,
+      "step": 443
+    },
+    {
+      "epoch": 0.2461879678403105,
+      "grad_norm": 0.21451757848262787,
+      "learning_rate": 7.221257790651364e-05,
+      "loss": 1.5854,
+      "step": 444
+    },
+    {
+      "epoch": 0.24674244524535624,
+      "grad_norm": 0.21790243685245514,
+      "learning_rate": 7.216877710463877e-05,
+      "loss": 1.5891,
+      "step": 445
+    },
+    {
+      "epoch": 0.24729692265040198,
+      "grad_norm": 0.2403048574924469,
+      "learning_rate": 7.212486683030377e-05,
+      "loss": 1.6683,
+      "step": 446
+    },
+    {
+      "epoch": 0.24785140005544773,
+      "grad_norm": 0.252909779548645,
+      "learning_rate": 7.208084723293823e-05,
+      "loss": 1.5939,
+      "step": 447
+    },
+    {
+      "epoch": 0.24840587746049347,
+      "grad_norm": 0.22861145436763763,
+      "learning_rate": 7.203671846234371e-05,
+      "loss": 1.6284,
+      "step": 448
+    },
+    {
+      "epoch": 0.24896035486553922,
+      "grad_norm": 0.2679903209209442,
+      "learning_rate": 7.199248066869331e-05,
+      "loss": 1.6749,
+      "step": 449
+    },
+    {
+      "epoch": 0.24951483227058496,
+      "grad_norm": 0.19942152500152588,
+      "learning_rate": 7.194813400253114e-05,
+      "loss": 1.5894,
+      "step": 450
+    },
+    {
+      "epoch": 0.25006930967563074,
+      "grad_norm": 0.23093563318252563,
+      "learning_rate": 7.190367861477183e-05,
+      "loss": 1.5947,
+      "step": 451
+    },
+    {
+      "epoch": 0.25062378708067645,
+      "grad_norm": 0.22997967898845673,
+      "learning_rate": 7.185911465669998e-05,
+      "loss": 1.6138,
+      "step": 452
+    },
+    {
+      "epoch": 0.2511782644857222,
+      "grad_norm": 0.2587045729160309,
+      "learning_rate": 7.181444227996966e-05,
+      "loss": 1.6803,
+      "step": 453
+    },
+    {
+      "epoch": 0.25173274189076794,
+      "grad_norm": 0.2321499139070511,
+      "learning_rate": 7.17696616366039e-05,
+      "loss": 1.7357,
+      "step": 454
+    },
+    {
+      "epoch": 0.2522872192958137,
+      "grad_norm": 0.24048465490341187,
+      "learning_rate": 7.172477287899418e-05,
+      "loss": 1.6514,
+      "step": 455
+    },
+    {
+      "epoch": 0.25284169670085943,
+      "grad_norm": 0.22340674698352814,
+      "learning_rate": 7.167977615989985e-05,
+      "loss": 1.6029,
+      "step": 456
+    },
+    {
+      "epoch": 0.2533961741059052,
+      "grad_norm": 0.21953530609607697,
+      "learning_rate": 7.163467163244775e-05,
+      "loss": 1.5879,
+      "step": 457
+    },
+    {
+      "epoch": 0.2539506515109509,
+      "grad_norm": 0.22932195663452148,
+      "learning_rate": 7.158945945013151e-05,
+      "loss": 1.6812,
+      "step": 458
+    },
+    {
+      "epoch": 0.2545051289159967,
+      "grad_norm": 0.21497461199760437,
+      "learning_rate": 7.15441397668112e-05,
+      "loss": 1.6981,
+      "step": 459
+    },
+    {
+      "epoch": 0.2550596063210424,
+      "grad_norm": 0.22530704736709595,
+      "learning_rate": 7.149871273671262e-05,
+      "loss": 1.6335,
+      "step": 460
+    },
+    {
+      "epoch": 0.2556140837260882,
+      "grad_norm": 0.20990800857543945,
+      "learning_rate": 7.145317851442696e-05,
+      "loss": 1.6352,
+      "step": 461
+    },
+    {
+      "epoch": 0.2561685611311339,
+      "grad_norm": 0.2138289213180542,
+      "learning_rate": 7.140753725491019e-05,
+      "loss": 1.6216,
+      "step": 462
+    },
+    {
+      "epoch": 0.25672303853617967,
+      "grad_norm": 0.2041904479265213,
+      "learning_rate": 7.136178911348248e-05,
+      "loss": 1.6162,
+      "step": 463
+    },
+    {
+      "epoch": 0.2572775159412254,
+      "grad_norm": 0.22262172400951385,
+      "learning_rate": 7.131593424582777e-05,
+      "loss": 1.633,
+      "step": 464
+    },
+    {
+      "epoch": 0.25783199334627116,
+      "grad_norm": 0.20571322739124298,
+      "learning_rate": 7.12699728079932e-05,
+      "loss": 1.6332,
+      "step": 465
+    },
+    {
+      "epoch": 0.2583864707513169,
+      "grad_norm": 0.22316017746925354,
+      "learning_rate": 7.122390495638853e-05,
+      "loss": 1.6925,
+      "step": 466
+    },
+    {
+      "epoch": 0.25894094815636265,
+      "grad_norm": 0.20561595261096954,
+      "learning_rate": 7.117773084778568e-05,
+      "loss": 1.5525,
+      "step": 467
+    },
+    {
+      "epoch": 0.25949542556140837,
+      "grad_norm": 0.2053271234035492,
+      "learning_rate": 7.113145063931821e-05,
+      "loss": 1.6606,
+      "step": 468
+    },
+    {
+      "epoch": 0.26004990296645414,
+      "grad_norm": 0.2116890847682953,
+      "learning_rate": 7.108506448848069e-05,
+      "loss": 1.5761,
+      "step": 469
+    },
+    {
+      "epoch": 0.26060438037149986,
+      "grad_norm": 0.20428834855556488,
+      "learning_rate": 7.103857255312823e-05,
+      "loss": 1.6077,
+      "step": 470
+    },
+    {
+      "epoch": 0.2611588577765456,
+      "grad_norm": 0.20274734497070312,
+      "learning_rate": 7.099197499147594e-05,
+      "loss": 1.5833,
+      "step": 471
+    },
+    {
+      "epoch": 0.26171333518159134,
+      "grad_norm": 0.20062290132045746,
+      "learning_rate": 7.094527196209838e-05,
+      "loss": 1.5884,
+      "step": 472
+    },
+    {
+      "epoch": 0.2622678125866371,
+      "grad_norm": 0.2172449678182602,
+      "learning_rate": 7.089846362392904e-05,
+      "loss": 1.7173,
+      "step": 473
+    },
+    {
+      "epoch": 0.26282228999168283,
+      "grad_norm": 0.21342062950134277,
+      "learning_rate": 7.085155013625974e-05,
+      "loss": 1.6288,
+      "step": 474
+    },
+    {
+      "epoch": 0.2633767673967286,
+      "grad_norm": 0.2020542323589325,
+      "learning_rate": 7.080453165874018e-05,
+      "loss": 1.5916,
+      "step": 475
+    },
+    {
+      "epoch": 0.2639312448017743,
+      "grad_norm": 0.20829129219055176,
+      "learning_rate": 7.07574083513773e-05,
+      "loss": 1.6227,
+      "step": 476
+    },
+    {
+      "epoch": 0.2644857222068201,
+      "grad_norm": 0.20422028005123138,
+      "learning_rate": 7.071018037453485e-05,
+      "loss": 1.63,
+      "step": 477
+    },
+    {
+      "epoch": 0.2650401996118658,
+      "grad_norm": 0.19959598779678345,
+      "learning_rate": 7.066284788893268e-05,
+      "loss": 1.6,
+      "step": 478
+    },
+    {
+      "epoch": 0.2655946770169116,
+      "grad_norm": 0.21311348676681519,
+      "learning_rate": 7.061541105564642e-05,
+      "loss": 1.6359,
+      "step": 479
+    },
+    {
+      "epoch": 0.2661491544219573,
+      "grad_norm": 0.21512793004512787,
+      "learning_rate": 7.056787003610667e-05,
+      "loss": 1.6067,
+      "step": 480
+    },
+    {
+      "epoch": 0.2667036318270031,
+      "grad_norm": 0.2027646005153656,
+      "learning_rate": 7.052022499209864e-05,
+      "loss": 1.6362,
+      "step": 481
+    },
+    {
+      "epoch": 0.2672581092320488,
+      "grad_norm": 0.23081980645656586,
+      "learning_rate": 7.047247608576157e-05,
+      "loss": 1.6021,
+      "step": 482
+    },
+    {
+      "epoch": 0.26781258663709456,
+      "grad_norm": 0.20609967410564423,
+      "learning_rate": 7.04246234795881e-05,
+      "loss": 1.6012,
+      "step": 483
+    },
+    {
+      "epoch": 0.2683670640421403,
+      "grad_norm": 0.20472081005573273,
+      "learning_rate": 7.03766673364238e-05,
+      "loss": 1.5602,
+      "step": 484
+    },
+    {
+      "epoch": 0.26892154144718605,
+      "grad_norm": 0.2030676007270813,
+      "learning_rate": 7.032860781946657e-05,
+      "loss": 1.6053,
+      "step": 485
+    },
+    {
+      "epoch": 0.26947601885223177,
+      "grad_norm": 0.22049987316131592,
+      "learning_rate": 7.028044509226612e-05,
+      "loss": 1.6074,
+      "step": 486
+    },
+    {
+      "epoch": 0.27003049625727754,
+      "grad_norm": 0.20548531413078308,
+      "learning_rate": 7.023217931872334e-05,
+      "loss": 1.5976,
+      "step": 487
+    },
+    {
+      "epoch": 0.27058497366232326,
+      "grad_norm": 0.22220125794410706,
+      "learning_rate": 7.018381066308984e-05,
+      "loss": 1.589,
+      "step": 488
+    },
+    {
+      "epoch": 0.27113945106736903,
+      "grad_norm": 0.209039568901062,
+      "learning_rate": 7.013533928996733e-05,
+      "loss": 1.6254,
+      "step": 489
+    },
+    {
+      "epoch": 0.27169392847241475,
+      "grad_norm": 0.21839891374111176,
+      "learning_rate": 7.00867653643071e-05,
+      "loss": 1.6726,
+      "step": 490
+    },
+    {
+      "epoch": 0.2722484058774605,
+      "grad_norm": 0.20095285773277283,
+      "learning_rate": 7.003808905140936e-05,
+      "loss": 1.5836,
+      "step": 491
+    },
+    {
+      "epoch": 0.27280288328250624,
+      "grad_norm": 0.19979004561901093,
+      "learning_rate": 6.998931051692283e-05,
+      "loss": 1.5875,
+      "step": 492
+    },
+    {
+      "epoch": 0.273357360687552,
+      "grad_norm": 0.20679202675819397,
+      "learning_rate": 6.994042992684406e-05,
+      "loss": 1.5333,
+      "step": 493
+    },
+    {
+      "epoch": 0.2739118380925977,
+      "grad_norm": 0.1942681074142456,
+      "learning_rate": 6.989144744751689e-05,
+      "loss": 1.5397,
+      "step": 494
+    },
+    {
+      "epoch": 0.2744663154976435,
+      "grad_norm": 0.210038959980011,
+      "learning_rate": 6.984236324563192e-05,
+      "loss": 1.6923,
+      "step": 495
+    },
+    {
+      "epoch": 0.2750207929026892,
+      "grad_norm": 0.213044673204422,
+      "learning_rate": 6.979317748822594e-05,
+      "loss": 1.6881,
+      "step": 496
+    },
+    {
+      "epoch": 0.275575270307735,
+      "grad_norm": 0.21259163320064545,
+      "learning_rate": 6.974389034268127e-05,
+      "loss": 1.6463,
+      "step": 497
+    },
+    {
+      "epoch": 0.2761297477127807,
+      "grad_norm": 0.2128421813249588,
+      "learning_rate": 6.969450197672534e-05,
+      "loss": 1.6319,
+      "step": 498
+    },
+    {
+      "epoch": 0.2766842251178264,
+      "grad_norm": 0.21015048027038574,
+      "learning_rate": 6.964501255842995e-05,
+      "loss": 1.5996,
+      "step": 499
+    },
+    {
+      "epoch": 0.2772387025228722,
+      "grad_norm": 0.20773227512836456,
+      "learning_rate": 6.959542225621087e-05,
+      "loss": 1.6476,
+      "step": 500
+    },
+    {
+      "epoch": 0.2777931799279179,
+      "grad_norm": 0.21952128410339355,
+      "learning_rate": 6.954573123882718e-05,
+      "loss": 1.5959,
+      "step": 501
+    },
+    {
+      "epoch": 0.2783476573329637,
+      "grad_norm": 0.20505724847316742,
+      "learning_rate": 6.949593967538062e-05,
+      "loss": 1.583,
+      "step": 502
+    },
+    {
+      "epoch": 0.2789021347380094,
+      "grad_norm": 0.21379536390304565,
+      "learning_rate": 6.94460477353152e-05,
+      "loss": 1.6331,
+      "step": 503
+    },
+    {
+      "epoch": 0.27945661214305517,
+      "grad_norm": 0.22115319967269897,
+      "learning_rate": 6.939605558841644e-05,
+      "loss": 1.5612,
+      "step": 504
+    },
+    {
+      "epoch": 0.2800110895481009,
+      "grad_norm": 0.2217654585838318,
+      "learning_rate": 6.934596340481088e-05,
+      "loss": 1.5818,
+      "step": 505
+    },
+    {
+      "epoch": 0.28056556695314666,
+      "grad_norm": 0.21580982208251953,
+      "learning_rate": 6.929577135496556e-05,
+      "loss": 1.6032,
+      "step": 506
+    },
+    {
+      "epoch": 0.2811200443581924,
+      "grad_norm": 0.2084866166114807,
+      "learning_rate": 6.924547960968726e-05,
+      "loss": 1.5739,
+      "step": 507
+    },
+    {
+      "epoch": 0.28167452176323815,
+      "grad_norm": 0.20981299877166748,
+      "learning_rate": 6.919508834012213e-05,
+      "loss": 1.6299,
+      "step": 508
+    },
+    {
+      "epoch": 0.28222899916828387,
+      "grad_norm": 0.20049329102039337,
+      "learning_rate": 6.914459771775496e-05,
+      "loss": 1.6606,
+      "step": 509
+    },
+    {
+      "epoch": 0.28278347657332964,
+      "grad_norm": 0.22596615552902222,
+      "learning_rate": 6.909400791440864e-05,
+      "loss": 1.5644,
+      "step": 510
+    },
+    {
+      "epoch": 0.28333795397837536,
+      "grad_norm": 0.20176082849502563,
+      "learning_rate": 6.904331910224361e-05,
+      "loss": 1.5579,
+      "step": 511
+    },
+    {
+      "epoch": 0.28389243138342113,
+      "grad_norm": 0.22138462960720062,
+      "learning_rate": 6.899253145375723e-05,
+      "loss": 1.6555,
+      "step": 512
+    },
+    {
+      "epoch": 0.28444690878846685,
+      "grad_norm": 0.2055101990699768,
+      "learning_rate": 6.894164514178317e-05,
+      "loss": 1.591,
+      "step": 513
+    },
+    {
+      "epoch": 0.2850013861935126,
+      "grad_norm": 0.2184118628501892,
+      "learning_rate": 6.88906603394909e-05,
+      "loss": 1.6579,
+      "step": 514
+    },
+    {
+      "epoch": 0.28555586359855833,
+      "grad_norm": 0.21609501540660858,
+      "learning_rate": 6.883957722038508e-05,
+      "loss": 1.6977,
+      "step": 515
+    },
+    {
+      "epoch": 0.2861103410036041,
+      "grad_norm": 0.20950822532176971,
+      "learning_rate": 6.878839595830489e-05,
+      "loss": 1.6842,
+      "step": 516
+    },
+    {
+      "epoch": 0.2866648184086498,
+      "grad_norm": 0.20638826489448547,
+      "learning_rate": 6.873711672742353e-05,
+      "loss": 1.5596,
+      "step": 517
+    },
+    {
+      "epoch": 0.2872192958136956,
+      "grad_norm": 0.20869210362434387,
+      "learning_rate": 6.868573970224757e-05,
+      "loss": 1.5563,
+      "step": 518
+    },
+    {
+      "epoch": 0.2877737732187413,
+      "grad_norm": 0.20834721624851227,
+      "learning_rate": 6.863426505761645e-05,
+      "loss": 1.6254,
+      "step": 519
+    },
+    {
+      "epoch": 0.2883282506237871,
+      "grad_norm": 0.20829017460346222,
+      "learning_rate": 6.85826929687017e-05,
+      "loss": 1.6216,
+      "step": 520
+    },
+    {
+      "epoch": 0.2888827280288328,
+      "grad_norm": 0.20338068902492523,
+      "learning_rate": 6.853102361100657e-05,
+      "loss": 1.5475,
+      "step": 521
+    },
+    {
+      "epoch": 0.2894372054338786,
+      "grad_norm": 0.21138431131839752,
+      "learning_rate": 6.847925716036526e-05,
+      "loss": 1.625,
+      "step": 522
+    },
+    {
+      "epoch": 0.2899916828389243,
+      "grad_norm": 0.21660204231739044,
+      "learning_rate": 6.842739379294241e-05,
+      "loss": 1.5498,
+      "step": 523
+    },
+    {
+      "epoch": 0.29054616024397006,
+      "grad_norm": 0.2074098438024521,
+      "learning_rate": 6.837543368523244e-05,
+      "loss": 1.6134,
+      "step": 524
+    },
+    {
+      "epoch": 0.2911006376490158,
+      "grad_norm": 0.20286381244659424,
+      "learning_rate": 6.832337701405904e-05,
+      "loss": 1.6212,
+      "step": 525
+    },
+    {
+      "epoch": 0.29165511505406155,
+      "grad_norm": 0.21887677907943726,
+      "learning_rate": 6.827122395657445e-05,
+      "loss": 1.6228,
+      "step": 526
+    },
+    {
+      "epoch": 0.29220959245910727,
+      "grad_norm": 0.20947344601154327,
+      "learning_rate": 6.821897469025895e-05,
+      "loss": 1.6106,
+      "step": 527
+    },
+    {
+      "epoch": 0.29276406986415304,
+      "grad_norm": 0.20572635531425476,
+      "learning_rate": 6.816662939292024e-05,
+      "loss": 1.5996,
+      "step": 528
+    },
+    {
+      "epoch": 0.29331854726919876,
+      "grad_norm": 0.20719175040721893,
+      "learning_rate": 6.81141882426928e-05,
+      "loss": 1.6113,
+      "step": 529
+    },
+    {
+      "epoch": 0.29387302467424453,
+      "grad_norm": 0.2026231288909912,
+      "learning_rate": 6.80616514180373e-05,
+      "loss": 1.5744,
+      "step": 530
+    },
+    {
+      "epoch": 0.29442750207929025,
+      "grad_norm": 0.19878005981445312,
+      "learning_rate": 6.800901909774e-05,
+      "loss": 1.5897,
+      "step": 531
+    },
+    {
+      "epoch": 0.294981979484336,
+      "grad_norm": 0.2110818773508072,
+      "learning_rate": 6.795629146091215e-05,
+      "loss": 1.5741,
+      "step": 532
+    },
+    {
+      "epoch": 0.29553645688938174,
+      "grad_norm": 0.20320414006710052,
+      "learning_rate": 6.790346868698936e-05,
+      "loss": 1.6345,
+      "step": 533
+    },
+    {
+      "epoch": 0.2960909342944275,
+      "grad_norm": 0.19918134808540344,
+      "learning_rate": 6.785055095573098e-05,
+      "loss": 1.5811,
+      "step": 534
+    },
+    {
+      "epoch": 0.2966454116994732,
+      "grad_norm": 0.20493772625923157,
+      "learning_rate": 6.779753844721955e-05,
+      "loss": 1.5529,
+      "step": 535
+    },
+    {
+      "epoch": 0.297199889104519,
+      "grad_norm": 0.20805048942565918,
+      "learning_rate": 6.774443134186008e-05,
+      "loss": 1.5974,
+      "step": 536
+    },
+    {
+      "epoch": 0.2977543665095647,
+      "grad_norm": 0.21310682594776154,
+      "learning_rate": 6.769122982037959e-05,
+      "loss": 1.6811,
+      "step": 537
+    },
+    {
+      "epoch": 0.2983088439146105,
+      "grad_norm": 0.21554915606975555,
+      "learning_rate": 6.763793406382629e-05,
+      "loss": 1.6707,
+      "step": 538
+    },
+    {
+      "epoch": 0.2988633213196562,
+      "grad_norm": 0.21235989034175873,
+      "learning_rate": 6.758454425356917e-05,
+      "loss": 1.6141,
+      "step": 539
+    },
+    {
+      "epoch": 0.299417798724702,
+      "grad_norm": 0.2248426377773285,
+      "learning_rate": 6.753106057129725e-05,
+      "loss": 1.6711,
+      "step": 540
+    },
+    {
+      "epoch": 0.2999722761297477,
+      "grad_norm": 0.20382218062877655,
+      "learning_rate": 6.747748319901902e-05,
+      "loss": 1.5494,
+      "step": 541
+    },
+    {
+      "epoch": 0.30052675353479347,
+      "grad_norm": 0.21593333780765533,
+      "learning_rate": 6.74238123190618e-05,
+      "loss": 1.6111,
+      "step": 542
+    },
+    {
+      "epoch": 0.3010812309398392,
+      "grad_norm": 0.20532597601413727,
+      "learning_rate": 6.737004811407109e-05,
+      "loss": 1.5824,
+      "step": 543
+    },
+    {
+      "epoch": 0.30163570834488496,
+      "grad_norm": 0.23563292622566223,
+      "learning_rate": 6.731619076701002e-05,
+      "loss": 1.5641,
+      "step": 544
+    },
+    {
+      "epoch": 0.3021901857499307,
+      "grad_norm": 0.2265416830778122,
+      "learning_rate": 6.726224046115871e-05,
+      "loss": 1.7597,
+      "step": 545
+    },
+    {
+      "epoch": 0.30274466315497645,
+      "grad_norm": 0.21692894399166107,
+      "learning_rate": 6.720819738011355e-05,
+      "loss": 1.5962,
+      "step": 546
+    },
+    {
+      "epoch": 0.30329914056002216,
+      "grad_norm": 0.2314797043800354,
+      "learning_rate": 6.715406170778671e-05,
+      "loss": 1.5945,
+      "step": 547
+    },
+    {
+      "epoch": 0.30385361796506793,
+      "grad_norm": 0.21994145214557648,
+      "learning_rate": 6.709983362840544e-05,
+      "loss": 1.6904,
+      "step": 548
+    },
+    {
+      "epoch": 0.30440809537011365,
+      "grad_norm": 0.21997472643852234,
+      "learning_rate": 6.704551332651144e-05,
+      "loss": 1.6322,
+      "step": 549
+    },
+    {
+      "epoch": 0.3049625727751594,
+      "grad_norm": 0.22159717977046967,
+      "learning_rate": 6.699110098696029e-05,
+      "loss": 1.6348,
+      "step": 550
+    },
+    {
+      "epoch": 0.30551705018020514,
+      "grad_norm": 0.20532500743865967,
+      "learning_rate": 6.693659679492072e-05,
+      "loss": 1.5946,
+      "step": 551
+    },
+    {
+      "epoch": 0.3060715275852509,
+      "grad_norm": 0.22436174750328064,
+      "learning_rate": 6.688200093587409e-05,
+      "loss": 1.5677,
+      "step": 552
+    },
+    {
+      "epoch": 0.30662600499029663,
+      "grad_norm": 0.20958715677261353,
+      "learning_rate": 6.682731359561369e-05,
+      "loss": 1.5468,
+      "step": 553
+    },
+    {
+      "epoch": 0.3071804823953424,
+      "grad_norm": 0.23304277658462524,
+      "learning_rate": 6.677253496024412e-05,
+      "loss": 1.6636,
+      "step": 554
+    },
+    {
+      "epoch": 0.3077349598003881,
+      "grad_norm": 0.2050829380750656,
+      "learning_rate": 6.671766521618069e-05,
+      "loss": 1.5629,
+      "step": 555
+    },
+    {
+      "epoch": 0.3082894372054339,
+      "grad_norm": 0.22310495376586914,
+      "learning_rate": 6.666270455014874e-05,
+      "loss": 1.6744,
+      "step": 556
+    },
+    {
+      "epoch": 0.3088439146104796,
+      "grad_norm": 0.21029001474380493,
+      "learning_rate": 6.6607653149183e-05,
+      "loss": 1.6397,
+      "step": 557
+    },
+    {
+      "epoch": 0.3093983920155254,
+      "grad_norm": 0.21131139993667603,
+      "learning_rate": 6.655251120062702e-05,
+      "loss": 1.62,
+      "step": 558
+    },
+    {
+      "epoch": 0.3099528694205711,
+      "grad_norm": 0.21726414561271667,
+      "learning_rate": 6.649727889213246e-05,
+      "loss": 1.601,
+      "step": 559
+    },
+    {
+      "epoch": 0.31050734682561687,
+      "grad_norm": 0.2165653556585312,
+      "learning_rate": 6.644195641165851e-05,
+      "loss": 1.577,
+      "step": 560
+    },
+    {
+      "epoch": 0.3110618242306626,
+      "grad_norm": 0.20710572600364685,
+      "learning_rate": 6.63865439474712e-05,
+      "loss": 1.5241,
+      "step": 561
+    },
+    {
+      "epoch": 0.31161630163570836,
+      "grad_norm": 0.20715083181858063,
+      "learning_rate": 6.633104168814276e-05,
+      "loss": 1.5959,
+      "step": 562
+    },
+    {
+      "epoch": 0.3121707790407541,
+      "grad_norm": 0.2072436511516571,
+      "learning_rate": 6.627544982255105e-05,
+      "loss": 1.669,
+      "step": 563
+    },
+    {
+      "epoch": 0.31272525644579985,
+      "grad_norm": 0.22219659388065338,
+      "learning_rate": 6.62197685398788e-05,
+      "loss": 1.6437,
+      "step": 564
+    },
+    {
+      "epoch": 0.31327973385084557,
+      "grad_norm": 0.209031879901886,
+      "learning_rate": 6.616399802961312e-05,
+      "loss": 1.6819,
+      "step": 565
+    },
+    {
+      "epoch": 0.31383421125589134,
+      "grad_norm": 0.21048156917095184,
+      "learning_rate": 6.610813848154467e-05,
+      "loss": 1.6213,
+      "step": 566
+    },
+    {
+      "epoch": 0.31438868866093705,
+      "grad_norm": 0.20893770456314087,
+      "learning_rate": 6.605219008576718e-05,
+      "loss": 1.6092,
+      "step": 567
+    },
+    {
+      "epoch": 0.3149431660659828,
+      "grad_norm": 0.20855359733104706,
+      "learning_rate": 6.599615303267672e-05,
+      "loss": 1.6285,
+      "step": 568
+    },
+    {
+      "epoch": 0.31549764347102854,
+      "grad_norm": 0.20627255737781525,
+      "learning_rate": 6.594002751297106e-05,
+      "loss": 1.5787,
+      "step": 569
+    },
+    {
+      "epoch": 0.3160521208760743,
+      "grad_norm": 0.21437032520771027,
+      "learning_rate": 6.588381371764903e-05,
+      "loss": 1.5824,
+      "step": 570
+    },
+    {
+      "epoch": 0.31660659828112003,
+      "grad_norm": 0.21238544583320618,
+      "learning_rate": 6.582751183800983e-05,
+      "loss": 1.6169,
+      "step": 571
+    },
+    {
+      "epoch": 0.3171610756861658,
+      "grad_norm": 0.20815801620483398,
+      "learning_rate": 6.57711220656525e-05,
+      "loss": 1.5866,
+      "step": 572
+    },
+    {
+      "epoch": 0.3177155530912115,
+      "grad_norm": 0.21221747994422913,
+      "learning_rate": 6.57146445924751e-05,
+      "loss": 1.6005,
+      "step": 573
+    },
+    {
+      "epoch": 0.3182700304962573,
+      "grad_norm": 0.22167359292507172,
+      "learning_rate": 6.565807961067421e-05,
+      "loss": 1.661,
+      "step": 574
+    },
+    {
+      "epoch": 0.318824507901303,
+      "grad_norm": 0.22634592652320862,
+      "learning_rate": 6.560142731274416e-05,
+      "loss": 1.6441,
+      "step": 575
+    },
+    {
+      "epoch": 0.3193789853063488,
+      "grad_norm": 0.20446883141994476,
+      "learning_rate": 6.554468789147644e-05,
+      "loss": 1.5857,
+      "step": 576
+    },
+    {
+      "epoch": 0.3199334627113945,
+      "grad_norm": 0.22486351430416107,
+      "learning_rate": 6.548786153995901e-05,
+      "loss": 1.6204,
+      "step": 577
+    },
+    {
+      "epoch": 0.3204879401164403,
+      "grad_norm": 0.2034136950969696,
+      "learning_rate": 6.543094845157569e-05,
+      "loss": 1.5765,
+      "step": 578
+    },
+    {
+      "epoch": 0.321042417521486,
+      "grad_norm": 0.21498997509479523,
+      "learning_rate": 6.537394882000545e-05,
+      "loss": 1.5481,
+      "step": 579
+    },
+    {
+      "epoch": 0.32159689492653176,
+      "grad_norm": 0.2029666006565094,
+      "learning_rate": 6.531686283922179e-05,
+      "loss": 1.578,
+      "step": 580
+    },
+    {
+      "epoch": 0.3221513723315775,
+      "grad_norm": 0.20419777929782867,
+      "learning_rate": 6.525969070349205e-05,
+      "loss": 1.5908,
+      "step": 581
+    },
+    {
+      "epoch": 0.32270584973662325,
+      "grad_norm": 0.2099248170852661,
+      "learning_rate": 6.520243260737674e-05,
+      "loss": 1.5928,
+      "step": 582
+    },
+    {
+      "epoch": 0.32326032714166897,
+      "grad_norm": 0.21225160360336304,
+      "learning_rate": 6.514508874572893e-05,
+      "loss": 1.5871,
+      "step": 583
+    },
+    {
+      "epoch": 0.32381480454671474,
+      "grad_norm": 0.20643840730190277,
+      "learning_rate": 6.508765931369356e-05,
+      "loss": 1.5607,
+      "step": 584
+    },
+    {
+      "epoch": 0.32436928195176046,
+      "grad_norm": 0.2086772918701172,
+      "learning_rate": 6.503014450670674e-05,
+      "loss": 1.5538,
+      "step": 585
+    },
+    {
+      "epoch": 0.32492375935680623,
+      "grad_norm": 0.2010130137205124,
+      "learning_rate": 6.497254452049513e-05,
+      "loss": 1.5856,
+      "step": 586
+    },
+    {
+      "epoch": 0.32547823676185195,
+      "grad_norm": 0.2193068265914917,
+      "learning_rate": 6.491485955107526e-05,
+      "loss": 1.7064,
+      "step": 587
+    },
+    {
+      "epoch": 0.3260327141668977,
+      "grad_norm": 0.2186783403158188,
+      "learning_rate": 6.485708979475291e-05,
+      "loss": 1.5548,
+      "step": 588
+    },
+    {
+      "epoch": 0.32658719157194344,
+      "grad_norm": 0.213240385055542,
+      "learning_rate": 6.47992354481223e-05,
+      "loss": 1.6239,
+      "step": 589
+    },
+    {
+      "epoch": 0.3271416689769892,
+      "grad_norm": 0.2521916627883911,
+      "learning_rate": 6.474129670806561e-05,
+      "loss": 1.6636,
+      "step": 590
+    },
+    {
+      "epoch": 0.3276961463820349,
+      "grad_norm": 0.21178847551345825,
+      "learning_rate": 6.468327377175214e-05,
+      "loss": 1.5943,
+      "step": 591
+    },
+    {
+      "epoch": 0.3282506237870807,
+      "grad_norm": 0.2146124243736267,
+      "learning_rate": 6.462516683663778e-05,
+      "loss": 1.615,
+      "step": 592
+    },
+    {
+      "epoch": 0.3288051011921264,
+      "grad_norm": 0.20442825555801392,
+      "learning_rate": 6.456697610046423e-05,
+      "loss": 1.5938,
+      "step": 593
+    },
+    {
+      "epoch": 0.3293595785971722,
+      "grad_norm": 0.2278599590063095,
+      "learning_rate": 6.450870176125838e-05,
+      "loss": 1.6331,
+      "step": 594
+    },
+    {
+      "epoch": 0.3299140560022179,
+      "grad_norm": 0.2228085696697235,
+      "learning_rate": 6.445034401733164e-05,
+      "loss": 1.6489,
+      "step": 595
+    },
+    {
+      "epoch": 0.3304685334072637,
+      "grad_norm": 0.24472852051258087,
+      "learning_rate": 6.439190306727926e-05,
+      "loss": 1.5416,
+      "step": 596
+    },
+    {
+      "epoch": 0.3310230108123094,
+      "grad_norm": 0.22040173411369324,
+      "learning_rate": 6.433337910997958e-05,
+      "loss": 1.5253,
+      "step": 597
+    },
+    {
+      "epoch": 0.33157748821735517,
+      "grad_norm": 0.27518245577812195,
+      "learning_rate": 6.427477234459353e-05,
+      "loss": 1.6411,
+      "step": 598
+    },
+    {
+      "epoch": 0.3321319656224009,
+      "grad_norm": 0.2157251238822937,
+      "learning_rate": 6.421608297056374e-05,
+      "loss": 1.6153,
+      "step": 599
+    },
+    {
+      "epoch": 0.33268644302744665,
+      "grad_norm": 0.2443874329328537,
+      "learning_rate": 6.415731118761401e-05,
+      "loss": 1.5696,
+      "step": 600
+    },
+    {
+      "epoch": 0.33324092043249237,
+      "grad_norm": 0.22603949904441833,
+      "learning_rate": 6.409845719574857e-05,
+      "loss": 1.6176,
+      "step": 601
+    },
+    {
+      "epoch": 0.33379539783753814,
+      "grad_norm": 0.20818479359149933,
+      "learning_rate": 6.403952119525143e-05,
+      "loss": 1.5308,
+      "step": 602
+    },
+    {
+      "epoch": 0.33434987524258386,
+      "grad_norm": 0.2425011247396469,
+      "learning_rate": 6.398050338668567e-05,
+      "loss": 1.5735,
+      "step": 603
+    },
+    {
+      "epoch": 0.33490435264762963,
+      "grad_norm": 0.21690739691257477,
+      "learning_rate": 6.392140397089275e-05,
+      "loss": 1.6721,
+      "step": 604
+    },
+    {
+      "epoch": 0.33545883005267535,
+      "grad_norm": 0.24877040088176727,
+      "learning_rate": 6.386222314899187e-05,
+      "loss": 1.6139,
+      "step": 605
+    },
+    {
+      "epoch": 0.3360133074577211,
+      "grad_norm": 0.2150987684726715,
+      "learning_rate": 6.380296112237926e-05,
+      "loss": 1.5697,
+      "step": 606
+    },
+    {
+      "epoch": 0.33656778486276684,
+      "grad_norm": 0.22079655528068542,
+      "learning_rate": 6.374361809272749e-05,
+      "loss": 1.6363,
+      "step": 607
+    },
+    {
+      "epoch": 0.3371222622678126,
+      "grad_norm": 0.20685946941375732,
+      "learning_rate": 6.368419426198475e-05,
+      "loss": 1.6146,
+      "step": 608
+    },
+    {
+      "epoch": 0.33767673967285833,
+      "grad_norm": 0.21263012290000916,
+      "learning_rate": 6.362468983237427e-05,
+      "loss": 1.5774,
+      "step": 609
+    },
+    {
+      "epoch": 0.3382312170779041,
+      "grad_norm": 0.21780216693878174,
+      "learning_rate": 6.356510500639353e-05,
+      "loss": 1.6147,
+      "step": 610
+    },
+    {
+      "epoch": 0.3387856944829498,
+      "grad_norm": 0.21480686962604523,
+      "learning_rate": 6.350543998681358e-05,
+      "loss": 1.6554,
+      "step": 611
+    },
+    {
+      "epoch": 0.3393401718879956,
+      "grad_norm": 0.2028435915708542,
+      "learning_rate": 6.344569497667843e-05,
+      "loss": 1.546,
+      "step": 612
+    },
+    {
+      "epoch": 0.3398946492930413,
+      "grad_norm": 0.22665052115917206,
+      "learning_rate": 6.338587017930425e-05,
+      "loss": 1.571,
+      "step": 613
+    },
+    {
+      "epoch": 0.3404491266980871,
+      "grad_norm": 0.20434421300888062,
+      "learning_rate": 6.332596579827876e-05,
+      "loss": 1.586,
+      "step": 614
+    },
+    {
+      "epoch": 0.3410036041031328,
+      "grad_norm": 0.20660161972045898,
+      "learning_rate": 6.326598203746049e-05,
+      "loss": 1.5352,
+      "step": 615
+    },
+    {
+      "epoch": 0.34155808150817857,
+      "grad_norm": 0.19511684775352478,
+      "learning_rate": 6.320591910097813e-05,
+      "loss": 1.5336,
+      "step": 616
+    },
+    {
+      "epoch": 0.3421125589132243,
+      "grad_norm": 0.21570967137813568,
+      "learning_rate": 6.314577719322978e-05,
+      "loss": 1.625,
+      "step": 617
+    },
+    {
+      "epoch": 0.34266703631827006,
+      "grad_norm": 0.20801737904548645,
+      "learning_rate": 6.308555651888233e-05,
+      "loss": 1.5529,
+      "step": 618
+    },
+    {
+      "epoch": 0.3432215137233158,
+      "grad_norm": 0.21180784702301025,
+      "learning_rate": 6.302525728287064e-05,
+      "loss": 1.6666,
+      "step": 619
+    },
+    {
+      "epoch": 0.34377599112836155,
+      "grad_norm": 0.20075224339962006,
+      "learning_rate": 6.296487969039701e-05,
+      "loss": 1.6188,
+      "step": 620
+    },
+    {
+      "epoch": 0.34433046853340726,
+      "grad_norm": 0.22708876430988312,
+      "learning_rate": 6.290442394693033e-05,
+      "loss": 1.6372,
+      "step": 621
+    },
+    {
+      "epoch": 0.344884945938453,
+      "grad_norm": 0.20980872213840485,
+      "learning_rate": 6.284389025820547e-05,
+      "loss": 1.584,
+      "step": 622
+    },
+    {
+      "epoch": 0.34543942334349875,
+      "grad_norm": 0.20637759566307068,
+      "learning_rate": 6.278327883022255e-05,
+      "loss": 1.5419,
+      "step": 623
+    },
+    {
+      "epoch": 0.34599390074854447,
+      "grad_norm": 0.2151980698108673,
+      "learning_rate": 6.272258986924624e-05,
+      "loss": 1.6066,
+      "step": 624
+    },
+    {
+      "epoch": 0.34654837815359024,
+      "grad_norm": 0.21410825848579407,
+      "learning_rate": 6.266182358180504e-05,
+      "loss": 1.657,
+      "step": 625
+    },
+    {
+      "epoch": 0.34710285555863596,
+      "grad_norm": 0.20414431393146515,
+      "learning_rate": 6.260098017469063e-05,
+      "loss": 1.6131,
+      "step": 626
+    },
+    {
+      "epoch": 0.34765733296368173,
+      "grad_norm": 0.19871225953102112,
+      "learning_rate": 6.254005985495711e-05,
+      "loss": 1.5843,
+      "step": 627
+    },
+    {
+      "epoch": 0.34821181036872745,
+      "grad_norm": 0.20605245232582092,
+      "learning_rate": 6.247906282992034e-05,
+      "loss": 1.6049,
+      "step": 628
+    },
+    {
+      "epoch": 0.3487662877737732,
+      "grad_norm": 0.20710204541683197,
+      "learning_rate": 6.241798930715719e-05,
+      "loss": 1.6005,
+      "step": 629
+    },
+    {
+      "epoch": 0.34932076517881894,
+      "grad_norm": 0.212701216340065,
+      "learning_rate": 6.235683949450486e-05,
+      "loss": 1.5958,
+      "step": 630
+    },
+    {
+      "epoch": 0.3498752425838647,
+      "grad_norm": 0.2133304476737976,
+      "learning_rate": 6.229561360006019e-05,
+      "loss": 1.6924,
+      "step": 631
+    },
+    {
+      "epoch": 0.3504297199889104,
+      "grad_norm": 0.208997905254364,
+      "learning_rate": 6.223431183217892e-05,
+      "loss": 1.6412,
+      "step": 632
+    },
+    {
+      "epoch": 0.3509841973939562,
+      "grad_norm": 0.2044183760881424,
+      "learning_rate": 6.217293439947498e-05,
+      "loss": 1.6015,
+      "step": 633
+    },
+    {
+      "epoch": 0.3515386747990019,
+      "grad_norm": 0.2030702829360962,
+      "learning_rate": 6.211148151081978e-05,
+      "loss": 1.5971,
+      "step": 634
+    },
+    {
+      "epoch": 0.3520931522040477,
+      "grad_norm": 0.20268899202346802,
+      "learning_rate": 6.204995337534159e-05,
+      "loss": 1.5271,
+      "step": 635
+    },
+    {
+      "epoch": 0.3526476296090934,
+      "grad_norm": 0.2157873958349228,
+      "learning_rate": 6.198835020242467e-05,
+      "loss": 1.6008,
+      "step": 636
+    },
+    {
+      "epoch": 0.3532021070141392,
+      "grad_norm": 0.20912663638591766,
+      "learning_rate": 6.192667220170863e-05,
+      "loss": 1.5696,
+      "step": 637
+    },
+    {
+      "epoch": 0.3537565844191849,
+      "grad_norm": 0.21048259735107422,
+      "learning_rate": 6.18649195830878e-05,
+      "loss": 1.5968,
+      "step": 638
+    },
+    {
+      "epoch": 0.35431106182423067,
+      "grad_norm": 0.21538151800632477,
+      "learning_rate": 6.180309255671035e-05,
+      "loss": 1.5876,
+      "step": 639
+    },
+    {
+      "epoch": 0.3548655392292764,
+      "grad_norm": 0.22569550573825836,
+      "learning_rate": 6.174119133297775e-05,
+      "loss": 1.5357,
+      "step": 640
+    },
+    {
+      "epoch": 0.35542001663432216,
+      "grad_norm": 0.20763731002807617,
+      "learning_rate": 6.167921612254391e-05,
+      "loss": 1.6395,
+      "step": 641
+    },
+    {
+      "epoch": 0.3559744940393679,
+      "grad_norm": 0.23909436166286469,
+      "learning_rate": 6.161716713631453e-05,
+      "loss": 1.6401,
+      "step": 642
+    },
+    {
+      "epoch": 0.35652897144441364,
+      "grad_norm": 0.20901797711849213,
+      "learning_rate": 6.155504458544641e-05,
+      "loss": 1.6965,
+      "step": 643
+    },
+    {
+      "epoch": 0.35708344884945936,
+      "grad_norm": 0.21390089392662048,
+      "learning_rate": 6.149284868134663e-05,
+      "loss": 1.5991,
+      "step": 644
+    },
+    {
+      "epoch": 0.35763792625450513,
+      "grad_norm": 0.2071395069360733,
+      "learning_rate": 6.143057963567198e-05,
+      "loss": 1.6255,
+      "step": 645
+    },
+    {
+      "epoch": 0.35819240365955085,
+      "grad_norm": 0.23318380117416382,
+      "learning_rate": 6.136823766032808e-05,
+      "loss": 1.6149,
+      "step": 646
+    },
+    {
+      "epoch": 0.3587468810645966,
+      "grad_norm": 0.21669737994670868,
+      "learning_rate": 6.130582296746876e-05,
+      "loss": 1.6068,
+      "step": 647
+    },
+    {
+      "epoch": 0.35930135846964234,
+      "grad_norm": 0.20202122628688812,
+      "learning_rate": 6.124333576949533e-05,
+      "loss": 1.5232,
+      "step": 648
+    },
+    {
+      "epoch": 0.3598558358746881,
+      "grad_norm": 0.20784695446491241,
+      "learning_rate": 6.118077627905584e-05,
+      "loss": 1.5963,
+      "step": 649
+    },
+    {
+      "epoch": 0.36041031327973383,
+      "grad_norm": 0.20467756688594818,
+      "learning_rate": 6.111814470904431e-05,
+      "loss": 1.5672,
+      "step": 650
+    },
+    {
+      "epoch": 0.3609647906847796,
+      "grad_norm": 0.20664702355861664,
+      "learning_rate": 6.105544127260012e-05,
+      "loss": 1.5897,
+      "step": 651
+    },
+    {
+      "epoch": 0.3615192680898253,
+      "grad_norm": 0.21211595833301544,
+      "learning_rate": 6.0992666183107134e-05,
+      "loss": 1.5792,
+      "step": 652
+    },
+    {
+      "epoch": 0.3620737454948711,
+      "grad_norm": 0.22114971280097961,
+      "learning_rate": 6.092981965419313e-05,
+      "loss": 1.5907,
+      "step": 653
+    },
+    {
+      "epoch": 0.3626282228999168,
+      "grad_norm": 0.22703395783901215,
+      "learning_rate": 6.086690189972898e-05,
+      "loss": 1.5498,
+      "step": 654
+    },
+    {
+      "epoch": 0.3631827003049626,
+      "grad_norm": 0.22778987884521484,
+      "learning_rate": 6.080391313382793e-05,
+      "loss": 1.6132,
+      "step": 655
+    },
+    {
+      "epoch": 0.3637371777100083,
+      "grad_norm": 0.21941795945167542,
+      "learning_rate": 6.074085357084487e-05,
+      "loss": 1.6606,
+      "step": 656
+    },
+    {
+      "epoch": 0.36429165511505407,
+      "grad_norm": 0.20801624655723572,
+      "learning_rate": 6.0677723425375636e-05,
+      "loss": 1.558,
+      "step": 657
+    },
+    {
+      "epoch": 0.3648461325200998,
+      "grad_norm": 0.22957980632781982,
+      "learning_rate": 6.061452291225627e-05,
+      "loss": 1.4949,
+      "step": 658
+    },
+    {
+      "epoch": 0.36540060992514556,
+      "grad_norm": 0.20350852608680725,
+      "learning_rate": 6.055125224656225e-05,
+      "loss": 1.5406,
+      "step": 659
+    },
+    {
+      "epoch": 0.3659550873301913,
+      "grad_norm": 0.21100889146327972,
+      "learning_rate": 6.048791164360781e-05,
+      "loss": 1.5888,
+      "step": 660
+    },
+    {
+      "epoch": 0.36650956473523705,
+      "grad_norm": 0.21262866258621216,
+      "learning_rate": 6.0424501318945194e-05,
+      "loss": 1.61,
+      "step": 661
+    },
+    {
+      "epoch": 0.36706404214028276,
+      "grad_norm": 0.22323627769947052,
+      "learning_rate": 6.036102148836387e-05,
+      "loss": 1.6143,
+      "step": 662
+    },
+    {
+      "epoch": 0.36761851954532854,
+      "grad_norm": 0.22186584770679474,
+      "learning_rate": 6.02974723678899e-05,
+      "loss": 1.5533,
+      "step": 663
+    },
+    {
+      "epoch": 0.36817299695037425,
+      "grad_norm": 0.2202465534210205,
+      "learning_rate": 6.0233854173785086e-05,
+      "loss": 1.6063,
+      "step": 664
+    },
+    {
+      "epoch": 0.36872747435542,
+      "grad_norm": 0.2212614119052887,
+      "learning_rate": 6.017016712254635e-05,
+      "loss": 1.5558,
+      "step": 665
+    },
+    {
+      "epoch": 0.36928195176046574,
+      "grad_norm": 0.21273072063922882,
+      "learning_rate": 6.0106411430904865e-05,
+      "loss": 1.5717,
+      "step": 666
+    },
+    {
+      "epoch": 0.3698364291655115,
+      "grad_norm": 0.24066902697086334,
+      "learning_rate": 6.004258731582546e-05,
+      "loss": 1.6385,
+      "step": 667
+    },
+    {
+      "epoch": 0.37039090657055723,
+      "grad_norm": 0.2065984606742859,
+      "learning_rate": 5.997869499450581e-05,
+      "loss": 1.6573,
+      "step": 668
+    },
+    {
+      "epoch": 0.370945383975603,
+      "grad_norm": 0.29595062136650085,
+      "learning_rate": 5.991473468437562e-05,
+      "loss": 1.6945,
+      "step": 669
+    },
+    {
+      "epoch": 0.3714998613806487,
+      "grad_norm": 0.20868165791034698,
+      "learning_rate": 5.985070660309609e-05,
+      "loss": 1.5274,
+      "step": 670
+    },
+    {
+      "epoch": 0.3720543387856945,
+      "grad_norm": 0.2694888412952423,
+      "learning_rate": 5.978661096855893e-05,
+      "loss": 1.5829,
+      "step": 671
+    },
+    {
+      "epoch": 0.3726088161907402,
+      "grad_norm": 0.21415413916110992,
+      "learning_rate": 5.972244799888583e-05,
+      "loss": 1.6068,
+      "step": 672
+    },
+    {
+      "epoch": 0.373163293595786,
+      "grad_norm": 0.2538868486881256,
+      "learning_rate": 5.9658217912427554e-05,
+      "loss": 1.586,
+      "step": 673
+    },
+    {
+      "epoch": 0.3737177710008317,
+      "grad_norm": 0.22458183765411377,
+      "learning_rate": 5.959392092776333e-05,
+      "loss": 1.5573,
+      "step": 674
+    },
+    {
+      "epoch": 0.3742722484058775,
+      "grad_norm": 0.23578275740146637,
+      "learning_rate": 5.952955726370001e-05,
+      "loss": 1.5725,
+      "step": 675
+    },
+    {
+      "epoch": 0.3748267258109232,
+      "grad_norm": 0.24922388792037964,
+      "learning_rate": 5.946512713927135e-05,
+      "loss": 1.6212,
+      "step": 676
+    },
+    {
+      "epoch": 0.37538120321596896,
+      "grad_norm": 0.21490588784217834,
+      "learning_rate": 5.940063077373732e-05,
+      "loss": 1.61,
+      "step": 677
+    },
+    {
+      "epoch": 0.3759356806210147,
+      "grad_norm": 0.26785025000572205,
+      "learning_rate": 5.933606838658328e-05,
+      "loss": 1.5897,
+      "step": 678
+    },
+    {
+      "epoch": 0.37649015802606045,
+      "grad_norm": 0.2139958143234253,
+      "learning_rate": 5.927144019751925e-05,
+      "loss": 1.5807,
+      "step": 679
+    },
+    {
+      "epoch": 0.37704463543110617,
+      "grad_norm": 0.25820040702819824,
+      "learning_rate": 5.9206746426479215e-05,
+      "loss": 1.5907,
+      "step": 680
+    },
+    {
+      "epoch": 0.37759911283615194,
+      "grad_norm": 0.21500596404075623,
+      "learning_rate": 5.9141987293620334e-05,
+      "loss": 1.5283,
+      "step": 681
+    },
+    {
+      "epoch": 0.37815359024119766,
+      "grad_norm": 0.21151086688041687,
+      "learning_rate": 5.907716301932217e-05,
+      "loss": 1.644,
+      "step": 682
+    },
+    {
+      "epoch": 0.37870806764624343,
+      "grad_norm": 0.24681684374809265,
+      "learning_rate": 5.901227382418599e-05,
+      "loss": 1.55,
+      "step": 683
+    },
+    {
+      "epoch": 0.37926254505128915,
+      "grad_norm": 0.20953518152236938,
+      "learning_rate": 5.894731992903399e-05,
+      "loss": 1.5991,
+      "step": 684
+    },
+    {
+      "epoch": 0.3798170224563349,
+      "grad_norm": 0.2509237825870514,
+      "learning_rate": 5.888230155490853e-05,
+      "loss": 1.6488,
+      "step": 685
+    },
+    {
+      "epoch": 0.38037149986138064,
+      "grad_norm": 0.22760917246341705,
+      "learning_rate": 5.8817218923071406e-05,
+      "loss": 1.5948,
+      "step": 686
+    },
+    {
+      "epoch": 0.3809259772664264,
+      "grad_norm": 0.20528970658779144,
+      "learning_rate": 5.875207225500308e-05,
+      "loss": 1.53,
+      "step": 687
+    },
+    {
+      "epoch": 0.3814804546714721,
+      "grad_norm": 0.3011578619480133,
+      "learning_rate": 5.8686861772401974e-05,
+      "loss": 1.6489,
+      "step": 688
+    },
+    {
+      "epoch": 0.3820349320765179,
+      "grad_norm": 0.20146943628787994,
+      "learning_rate": 5.8621587697183595e-05,
+      "loss": 1.5495,
+      "step": 689
+    },
+    {
+      "epoch": 0.3825894094815636,
+      "grad_norm": 0.24754847586154938,
+      "learning_rate": 5.8556250251479945e-05,
+      "loss": 1.6115,
+      "step": 690
+    },
+    {
+      "epoch": 0.3831438868866094,
+      "grad_norm": 0.21535150706768036,
+      "learning_rate": 5.84908496576386e-05,
+      "loss": 1.6324,
+      "step": 691
+    },
+    {
+      "epoch": 0.3836983642916551,
+      "grad_norm": 0.21527227759361267,
+      "learning_rate": 5.8425386138222116e-05,
+      "loss": 1.5944,
+      "step": 692
+    },
+    {
+      "epoch": 0.3842528416967009,
+      "grad_norm": 0.2230699062347412,
+      "learning_rate": 5.8359859916007116e-05,
+      "loss": 1.5704,
+      "step": 693
+    },
+    {
+      "epoch": 0.3848073191017466,
+      "grad_norm": 0.20711487531661987,
+      "learning_rate": 5.8294271213983646e-05,
+      "loss": 1.614,
+      "step": 694
+    },
+    {
+      "epoch": 0.38536179650679236,
+      "grad_norm": 0.23301945626735687,
+      "learning_rate": 5.822862025535436e-05,
+      "loss": 1.6114,
+      "step": 695
+    },
+    {
+      "epoch": 0.3859162739118381,
+      "grad_norm": 0.22132247686386108,
+      "learning_rate": 5.816290726353378e-05,
+      "loss": 1.6219,
+      "step": 696
+    },
+    {
+      "epoch": 0.38647075131688385,
+      "grad_norm": 0.22144806385040283,
+      "learning_rate": 5.809713246214756e-05,
+      "loss": 1.6163,
+      "step": 697
+    },
+    {
+      "epoch": 0.38702522872192957,
+      "grad_norm": 0.21476450562477112,
+      "learning_rate": 5.8031296075031625e-05,
+      "loss": 1.5815,
+      "step": 698
+    },
+    {
+      "epoch": 0.38757970612697534,
+      "grad_norm": 0.2073630541563034,
+      "learning_rate": 5.7965398326231535e-05,
+      "loss": 1.5726,
+      "step": 699
+    },
+    {
+      "epoch": 0.38813418353202106,
+      "grad_norm": 0.23282212018966675,
+      "learning_rate": 5.7899439440001656e-05,
+      "loss": 1.5905,
+      "step": 700
+    },
+    {
+      "epoch": 0.38868866093706683,
+      "grad_norm": 0.2030622363090515,
+      "learning_rate": 5.7833419640804426e-05,
+      "loss": 1.606,
+      "step": 701
+    },
+    {
+      "epoch": 0.38924313834211255,
+      "grad_norm": 0.22797855734825134,
+      "learning_rate": 5.7767339153309526e-05,
+      "loss": 1.6297,
+      "step": 702
+    },
+    {
+      "epoch": 0.3897976157471583,
+      "grad_norm": 0.21430207788944244,
+      "learning_rate": 5.770119820239321e-05,
+      "loss": 1.647,
+      "step": 703
+    },
+    {
+      "epoch": 0.39035209315220404,
+      "grad_norm": 0.21015015244483948,
+      "learning_rate": 5.7634997013137465e-05,
+      "loss": 1.5667,
+      "step": 704
+    },
+    {
+      "epoch": 0.3909065705572498,
+      "grad_norm": 0.2301322966814041,
+      "learning_rate": 5.7568735810829294e-05,
+      "loss": 1.6313,
+      "step": 705
+    },
+    {
+      "epoch": 0.3914610479622955,
+      "grad_norm": 0.21517878770828247,
+      "learning_rate": 5.750241482095993e-05,
+      "loss": 1.671,
+      "step": 706
+    },
+    {
+      "epoch": 0.3920155253673413,
+      "grad_norm": 0.2164376974105835,
+      "learning_rate": 5.743603426922401e-05,
+      "loss": 1.6186,
+      "step": 707
+    },
+    {
+      "epoch": 0.392570002772387,
+      "grad_norm": 0.2114233523607254,
+      "learning_rate": 5.736959438151895e-05,
+      "loss": 1.7116,
+      "step": 708
+    },
+    {
+      "epoch": 0.3931244801774328,
+      "grad_norm": 0.22350315749645233,
+      "learning_rate": 5.730309538394404e-05,
+      "loss": 1.6869,
+      "step": 709
+    },
+    {
+      "epoch": 0.3936789575824785,
+      "grad_norm": 0.2107817828655243,
+      "learning_rate": 5.723653750279974e-05,
+      "loss": 1.6738,
+      "step": 710
+    },
+    {
+      "epoch": 0.3942334349875243,
+      "grad_norm": 0.20846299827098846,
+      "learning_rate": 5.716992096458686e-05,
+      "loss": 1.5339,
+      "step": 711
+    },
+    {
+      "epoch": 0.39478791239257,
+      "grad_norm": 0.2204621285200119,
+      "learning_rate": 5.710324599600589e-05,
+      "loss": 1.5552,
+      "step": 712
+    },
+    {
+      "epoch": 0.39534238979761577,
+      "grad_norm": 0.23123124241828918,
+      "learning_rate": 5.7036512823956085e-05,
+      "loss": 1.7236,
+      "step": 713
+    },
+    {
+      "epoch": 0.3958968672026615,
+      "grad_norm": 0.24838416278362274,
+      "learning_rate": 5.696972167553485e-05,
+      "loss": 1.6249,
+      "step": 714
+    },
+    {
+      "epoch": 0.39645134460770726,
+      "grad_norm": 0.2113085836172104,
+      "learning_rate": 5.6902872778036825e-05,
+      "loss": 1.5164,
+      "step": 715
+    },
+    {
+      "epoch": 0.397005822012753,
+      "grad_norm": 0.2129892259836197,
+      "learning_rate": 5.6835966358953186e-05,
+      "loss": 1.6242,
+      "step": 716
+    },
+    {
+      "epoch": 0.39756029941779875,
+      "grad_norm": 0.21519723534584045,
+      "learning_rate": 5.67690026459709e-05,
+      "loss": 1.5984,
+      "step": 717
+    },
+    {
+      "epoch": 0.39811477682284446,
+      "grad_norm": 0.20214593410491943,
+      "learning_rate": 5.670198186697185e-05,
+      "loss": 1.6229,
+      "step": 718
+    },
+    {
+      "epoch": 0.39866925422789024,
+      "grad_norm": 0.21002614498138428,
+      "learning_rate": 5.6634904250032166e-05,
+      "loss": 1.6142,
+      "step": 719
+    },
+    {
+      "epoch": 0.39922373163293595,
+      "grad_norm": 0.20920030772686005,
+      "learning_rate": 5.656777002342136e-05,
+      "loss": 1.5819,
+      "step": 720
+    },
+    {
+      "epoch": 0.3997782090379817,
+      "grad_norm": 0.21228238940238953,
+      "learning_rate": 5.650057941560164e-05,
+      "loss": 1.6284,
+      "step": 721
+    },
+    {
+      "epoch": 0.40033268644302744,
+      "grad_norm": 0.331759512424469,
+      "learning_rate": 5.643333265522702e-05,
+      "loss": 1.5552,
+      "step": 722
+    },
+    {
+      "epoch": 0.4008871638480732,
+      "grad_norm": 0.20200534164905548,
+      "learning_rate": 5.636602997114268e-05,
+      "loss": 1.545,
+      "step": 723
+    },
+    {
+      "epoch": 0.40144164125311893,
+      "grad_norm": 0.2071356177330017,
+      "learning_rate": 5.629867159238404e-05,
+      "loss": 1.5618,
+      "step": 724
+    },
+    {
+      "epoch": 0.4019961186581647,
+      "grad_norm": 0.2137397825717926,
+      "learning_rate": 5.62312577481761e-05,
+      "loss": 1.5858,
+      "step": 725
+    },
+    {
+      "epoch": 0.4025505960632104,
+      "grad_norm": 0.2188277244567871,
+      "learning_rate": 5.616378866793259e-05,
+      "loss": 1.6123,
+      "step": 726
+    },
+    {
+      "epoch": 0.4031050734682562,
+      "grad_norm": 0.21393775939941406,
+      "learning_rate": 5.609626458125521e-05,
+      "loss": 1.6253,
+      "step": 727
+    },
+    {
+      "epoch": 0.4036595508733019,
+      "grad_norm": 0.2320735603570938,
+      "learning_rate": 5.6028685717932895e-05,
+      "loss": 1.5904,
+      "step": 728
+    },
+    {
+      "epoch": 0.4042140282783477,
+      "grad_norm": 0.22155635058879852,
+      "learning_rate": 5.596105230794091e-05,
+      "loss": 1.6295,
+      "step": 729
+    },
+    {
+      "epoch": 0.4047685056833934,
+      "grad_norm": 0.21663862466812134,
+      "learning_rate": 5.589336458144023e-05,
+      "loss": 1.5455,
+      "step": 730
+    },
+    {
+      "epoch": 0.40532298308843917,
+      "grad_norm": 0.2116219699382782,
+      "learning_rate": 5.582562276877659e-05,
+      "loss": 1.6089,
+      "step": 731
+    },
+    {
+      "epoch": 0.4058774604934849,
+      "grad_norm": 0.21755720674991608,
+      "learning_rate": 5.575782710047985e-05,
+      "loss": 1.5895,
+      "step": 732
+    },
+    {
+      "epoch": 0.40643193789853066,
+      "grad_norm": 0.21230702102184296,
+      "learning_rate": 5.5689977807263105e-05,
+      "loss": 1.5425,
+      "step": 733
+    },
+    {
+      "epoch": 0.4069864153035764,
+      "grad_norm": 0.21914036571979523,
+      "learning_rate": 5.5622075120021976e-05,
+      "loss": 1.659,
+      "step": 734
+    },
+    {
+      "epoch": 0.40754089270862215,
+      "grad_norm": 0.2092135101556778,
+      "learning_rate": 5.5554119269833746e-05,
+      "loss": 1.5671,
+      "step": 735
+    },
+    {
+      "epoch": 0.40809537011366787,
+      "grad_norm": 0.20937521755695343,
+      "learning_rate": 5.548611048795663e-05,
+      "loss": 1.552,
+      "step": 736
+    },
+    {
+      "epoch": 0.40864984751871364,
+      "grad_norm": 0.20987878739833832,
+      "learning_rate": 5.5418049005828994e-05,
+      "loss": 1.565,
+      "step": 737
+    },
+    {
+      "epoch": 0.40920432492375935,
+      "grad_norm": 0.22819262742996216,
+      "learning_rate": 5.534993505506851e-05,
+      "loss": 1.5928,
+      "step": 738
+    },
+    {
+      "epoch": 0.4097588023288051,
+      "grad_norm": 0.2156732827425003,
+      "learning_rate": 5.5281768867471455e-05,
+      "loss": 1.6158,
+      "step": 739
+    },
+    {
+      "epoch": 0.41031327973385084,
+      "grad_norm": 0.21924850344657898,
+      "learning_rate": 5.521355067501181e-05,
+      "loss": 1.5497,
+      "step": 740
+    },
+    {
+      "epoch": 0.4108677571388966,
+      "grad_norm": 0.20754189789295197,
+      "learning_rate": 5.5145280709840566e-05,
+      "loss": 1.5645,
+      "step": 741
+    },
+    {
+      "epoch": 0.41142223454394233,
+      "grad_norm": 0.22220148146152496,
+      "learning_rate": 5.5076959204284915e-05,
+      "loss": 1.5868,
+      "step": 742
+    },
+    {
+      "epoch": 0.4119767119489881,
+      "grad_norm": 0.20873475074768066,
+      "learning_rate": 5.5008586390847404e-05,
+      "loss": 1.616,
+      "step": 743
+    },
+    {
+      "epoch": 0.4125311893540338,
+      "grad_norm": 0.2225092649459839,
+      "learning_rate": 5.494016250220521e-05,
+      "loss": 1.6132,
+      "step": 744
+    },
+    {
+      "epoch": 0.4130856667590796,
+      "grad_norm": 0.20154620707035065,
+      "learning_rate": 5.487168777120932e-05,
+      "loss": 1.5352,
+      "step": 745
+    },
+    {
+      "epoch": 0.4136401441641253,
+      "grad_norm": 0.20975901186466217,
+      "learning_rate": 5.480316243088375e-05,
+      "loss": 1.5762,
+      "step": 746
+    },
+    {
+      "epoch": 0.41419462156917103,
+      "grad_norm": 0.2048894613981247,
+      "learning_rate": 5.4734586714424706e-05,
+      "loss": 1.5744,
+      "step": 747
+    },
+    {
+      "epoch": 0.4147490989742168,
+      "grad_norm": 0.2058396339416504,
+      "learning_rate": 5.466596085519988e-05,
+      "loss": 1.6035,
+      "step": 748
+    },
+    {
+      "epoch": 0.4153035763792625,
+      "grad_norm": 0.21916894614696503,
+      "learning_rate": 5.459728508674756e-05,
+      "loss": 1.6108,
+      "step": 749
+    },
+    {
+      "epoch": 0.4158580537843083,
+      "grad_norm": 0.2189890742301941,
+      "learning_rate": 5.4528559642775885e-05,
+      "loss": 1.6031,
+      "step": 750
+    },
+    {
+      "epoch": 0.416412531189354,
+      "grad_norm": 0.2107512503862381,
+      "learning_rate": 5.445978475716207e-05,
+      "loss": 1.5418,
+      "step": 751
+    },
+    {
+      "epoch": 0.4169670085943998,
+      "grad_norm": 0.20556119084358215,
+      "learning_rate": 5.4390960663951565e-05,
+      "loss": 1.597,
+      "step": 752
+    },
+    {
+      "epoch": 0.4175214859994455,
+      "grad_norm": 0.20787164568901062,
+      "learning_rate": 5.4322087597357264e-05,
+      "loss": 1.6246,
+      "step": 753
+    },
+    {
+      "epoch": 0.41807596340449127,
+      "grad_norm": 0.22257111966609955,
+      "learning_rate": 5.4253165791758743e-05,
+      "loss": 1.5532,
+      "step": 754
+    },
+    {
+      "epoch": 0.418630440809537,
+      "grad_norm": 0.20410916209220886,
+      "learning_rate": 5.4184195481701425e-05,
+      "loss": 1.5668,
+      "step": 755
+    },
+    {
+      "epoch": 0.41918491821458276,
+      "grad_norm": 0.234087273478508,
+      "learning_rate": 5.411517690189581e-05,
+      "loss": 1.5941,
+      "step": 756
+    },
+    {
+      "epoch": 0.4197393956196285,
+      "grad_norm": 0.21471847593784332,
+      "learning_rate": 5.404611028721665e-05,
+      "loss": 1.5839,
+      "step": 757
+    },
+    {
+      "epoch": 0.42029387302467425,
+      "grad_norm": 0.2194783240556717,
+      "learning_rate": 5.3976995872702174e-05,
+      "loss": 1.664,
+      "step": 758
+    },
+    {
+      "epoch": 0.42084835042971996,
+      "grad_norm": 0.21059168875217438,
+      "learning_rate": 5.390783389355326e-05,
+      "loss": 1.5688,
+      "step": 759
+    },
+    {
+      "epoch": 0.42140282783476574,
+      "grad_norm": 0.22869731485843658,
+      "learning_rate": 5.3838624585132666e-05,
+      "loss": 1.5823,
+      "step": 760
+    },
+    {
+      "epoch": 0.42195730523981145,
+      "grad_norm": 0.21715888381004333,
+      "learning_rate": 5.3769368182964226e-05,
+      "loss": 1.6105,
+      "step": 761
+    },
+    {
+      "epoch": 0.4225117826448572,
+      "grad_norm": 0.2125747799873352,
+      "learning_rate": 5.3700064922732e-05,
+      "loss": 1.5806,
+      "step": 762
+    },
+    {
+      "epoch": 0.42306626004990294,
+      "grad_norm": 0.20356424152851105,
+      "learning_rate": 5.363071504027956e-05,
+      "loss": 1.5695,
+      "step": 763
+    },
+    {
+      "epoch": 0.4236207374549487,
+      "grad_norm": 0.21644245088100433,
+      "learning_rate": 5.3561318771609076e-05,
+      "loss": 1.6602,
+      "step": 764
+    },
+    {
+      "epoch": 0.42417521485999443,
+      "grad_norm": 0.21127082407474518,
+      "learning_rate": 5.349187635288063e-05,
+      "loss": 1.554,
+      "step": 765
+    },
+    {
+      "epoch": 0.4247296922650402,
+      "grad_norm": 0.22828276455402374,
+      "learning_rate": 5.3422388020411325e-05,
+      "loss": 1.6498,
+      "step": 766
+    },
+    {
+      "epoch": 0.4252841696700859,
+      "grad_norm": 0.22633306682109833,
+      "learning_rate": 5.3352854010674515e-05,
+      "loss": 1.6079,
+      "step": 767
+    },
+    {
+      "epoch": 0.4258386470751317,
+      "grad_norm": 0.2162792831659317,
+      "learning_rate": 5.3283274560299e-05,
+      "loss": 1.6315,
+      "step": 768
+    },
+    {
+      "epoch": 0.4263931244801774,
+      "grad_norm": 0.23089319467544556,
+      "learning_rate": 5.321364990606821e-05,
+      "loss": 1.638,
+      "step": 769
+    },
+    {
+      "epoch": 0.4269476018852232,
+      "grad_norm": 0.2013845294713974,
+      "learning_rate": 5.3143980284919424e-05,
+      "loss": 1.5903,
+      "step": 770
+    },
+    {
+      "epoch": 0.4275020792902689,
+      "grad_norm": 0.21818867325782776,
+      "learning_rate": 5.307426593394294e-05,
+      "loss": 1.5645,
+      "step": 771
+    },
+    {
+      "epoch": 0.42805655669531467,
+      "grad_norm": 0.2172323316335678,
+      "learning_rate": 5.300450709038126e-05,
+      "loss": 1.5598,
+      "step": 772
+    },
+    {
+      "epoch": 0.4286110341003604,
+      "grad_norm": 0.21557918190956116,
+      "learning_rate": 5.29347039916283e-05,
+      "loss": 1.5937,
+      "step": 773
+    },
+    {
+      "epoch": 0.42916551150540616,
+      "grad_norm": 0.6955245733261108,
+      "learning_rate": 5.286485687522861e-05,
+      "loss": 1.6414,
+      "step": 774
+    },
+    {
+      "epoch": 0.4297199889104519,
+      "grad_norm": 0.21430703997612,
+      "learning_rate": 5.27949659788765e-05,
+      "loss": 1.5782,
+      "step": 775
+    },
+    {
+      "epoch": 0.43027446631549765,
+      "grad_norm": 0.2163209319114685,
+      "learning_rate": 5.272503154041527e-05,
+      "loss": 1.5992,
+      "step": 776
+    },
+    {
+      "epoch": 0.43082894372054337,
+      "grad_norm": 0.21228471398353577,
+      "learning_rate": 5.265505379783642e-05,
+      "loss": 1.5764,
+      "step": 777
+    },
+    {
+      "epoch": 0.43138342112558914,
+      "grad_norm": 0.21286910772323608,
+      "learning_rate": 5.258503298927879e-05,
+      "loss": 1.5892,
+      "step": 778
+    },
+    {
+      "epoch": 0.43193789853063486,
+      "grad_norm": 0.24243953824043274,
+      "learning_rate": 5.25149693530278e-05,
+      "loss": 1.6385,
+      "step": 779
+    },
+    {
+      "epoch": 0.43249237593568063,
+      "grad_norm": 0.20691627264022827,
+      "learning_rate": 5.244486312751459e-05,
+      "loss": 1.5992,
+      "step": 780
+    },
+    {
+      "epoch": 0.43304685334072635,
+      "grad_norm": 0.21702216565608978,
+      "learning_rate": 5.237471455131526e-05,
+      "loss": 1.5866,
+      "step": 781
+    },
+    {
+      "epoch": 0.4336013307457721,
+      "grad_norm": 0.22178849577903748,
+      "learning_rate": 5.230452386315e-05,
+      "loss": 1.6211,
+      "step": 782
+    },
+    {
+      "epoch": 0.43415580815081783,
+      "grad_norm": 0.2103152573108673,
+      "learning_rate": 5.223429130188235e-05,
+      "loss": 1.5822,
+      "step": 783
+    },
+    {
+      "epoch": 0.4347102855558636,
+      "grad_norm": 0.2260008603334427,
+      "learning_rate": 5.216401710651831e-05,
+      "loss": 1.5565,
+      "step": 784
+    },
+    {
+      "epoch": 0.4352647629609093,
+      "grad_norm": 0.22004656493663788,
+      "learning_rate": 5.20937015162056e-05,
+      "loss": 1.6813,
+      "step": 785
+    },
+    {
+      "epoch": 0.4358192403659551,
+      "grad_norm": 0.21554167568683624,
+      "learning_rate": 5.202334477023277e-05,
+      "loss": 1.6137,
+      "step": 786
+    },
+    {
+      "epoch": 0.4363737177710008,
+      "grad_norm": 0.2462761402130127,
+      "learning_rate": 5.195294710802845e-05,
+      "loss": 1.6396,
+      "step": 787
+    },
+    {
+      "epoch": 0.4369281951760466,
+      "grad_norm": 0.20096971094608307,
+      "learning_rate": 5.1882508769160504e-05,
+      "loss": 1.5586,
+      "step": 788
+    },
+    {
+      "epoch": 0.4374826725810923,
+      "grad_norm": 0.21837452054023743,
+      "learning_rate": 5.1812029993335226e-05,
+      "loss": 1.5882,
+      "step": 789
+    },
+    {
+      "epoch": 0.4380371499861381,
+      "grad_norm": 0.20460233092308044,
+      "learning_rate": 5.174151102039653e-05,
+      "loss": 1.5031,
+      "step": 790
+    },
+    {
+      "epoch": 0.4385916273911838,
+      "grad_norm": 0.2106686383485794,
+      "learning_rate": 5.167095209032509e-05,
+      "loss": 1.5377,
+      "step": 791
+    },
+    {
+      "epoch": 0.43914610479622956,
+      "grad_norm": 0.24171209335327148,
+      "learning_rate": 5.160035344323758e-05,
+      "loss": 1.6242,
+      "step": 792
+    },
+    {
+      "epoch": 0.4397005822012753,
+      "grad_norm": 0.21198348701000214,
+      "learning_rate": 5.152971531938583e-05,
+      "loss": 1.5975,
+      "step": 793
+    },
+    {
+      "epoch": 0.44025505960632105,
+      "grad_norm": 0.2270112782716751,
+      "learning_rate": 5.145903795915603e-05,
+      "loss": 1.5772,
+      "step": 794
+    },
+    {
+      "epoch": 0.44080953701136677,
+      "grad_norm": 0.20378631353378296,
+      "learning_rate": 5.138832160306785e-05,
+      "loss": 1.5778,
+      "step": 795
+    },
+    {
+      "epoch": 0.44136401441641254,
+      "grad_norm": 0.23949572443962097,
+      "learning_rate": 5.1317566491773714e-05,
+      "loss": 1.6321,
+      "step": 796
+    },
+    {
+      "epoch": 0.44191849182145826,
+      "grad_norm": 0.20986564457416534,
+      "learning_rate": 5.1246772866057885e-05,
+      "loss": 1.5273,
+      "step": 797
+    },
+    {
+      "epoch": 0.44247296922650403,
+      "grad_norm": 0.20840664207935333,
+      "learning_rate": 5.117594096683574e-05,
+      "loss": 1.5354,
+      "step": 798
+    },
+    {
+      "epoch": 0.44302744663154975,
+      "grad_norm": 0.24688231945037842,
+      "learning_rate": 5.1105071035152884e-05,
+      "loss": 1.637,
+      "step": 799
+    },
+    {
+      "epoch": 0.4435819240365955,
+      "grad_norm": 0.20614393055438995,
+      "learning_rate": 5.1034163312184325e-05,
+      "loss": 1.5573,
+      "step": 800
+    },
+    {
+      "epoch": 0.44413640144164124,
+      "grad_norm": 0.22784316539764404,
+      "learning_rate": 5.096321803923372e-05,
+      "loss": 1.6117,
+      "step": 801
+    },
+    {
+      "epoch": 0.444690878846687,
+      "grad_norm": 0.22619880735874176,
+      "learning_rate": 5.089223545773248e-05,
+      "loss": 1.5151,
+      "step": 802
+    },
+    {
+      "epoch": 0.4452453562517327,
+      "grad_norm": 0.2142421305179596,
+      "learning_rate": 5.082121580923899e-05,
+      "loss": 1.6123,
+      "step": 803
+    },
+    {
+      "epoch": 0.4457998336567785,
+      "grad_norm": 0.21790969371795654,
+      "learning_rate": 5.0750159335437775e-05,
+      "loss": 1.5859,
+      "step": 804
+    },
+    {
+      "epoch": 0.4463543110618242,
+      "grad_norm": 0.21449251472949982,
+      "learning_rate": 5.067906627813868e-05,
+      "loss": 1.6656,
+      "step": 805
+    },
+    {
+      "epoch": 0.44690878846687,
+      "grad_norm": 0.221047043800354,
+      "learning_rate": 5.0607936879276055e-05,
+      "loss": 1.5889,
+      "step": 806
+    },
+    {
+      "epoch": 0.4474632658719157,
+      "grad_norm": 0.22916845977306366,
+      "learning_rate": 5.0536771380907885e-05,
+      "loss": 1.5901,
+      "step": 807
+    },
+    {
+      "epoch": 0.4480177432769615,
+      "grad_norm": 0.20636655390262604,
+      "learning_rate": 5.046557002521507e-05,
+      "loss": 1.5883,
+      "step": 808
+    },
+    {
+      "epoch": 0.4485722206820072,
+      "grad_norm": 0.23739656805992126,
+      "learning_rate": 5.039433305450047e-05,
+      "loss": 1.6089,
+      "step": 809
+    },
+    {
+      "epoch": 0.44912669808705297,
+      "grad_norm": 0.2077774852514267,
+      "learning_rate": 5.03230607111882e-05,
+      "loss": 1.5868,
+      "step": 810
+    },
+    {
+      "epoch": 0.4496811754920987,
+      "grad_norm": 0.21799373626708984,
+      "learning_rate": 5.025175323782267e-05,
+      "loss": 1.565,
+      "step": 811
+    },
+    {
+      "epoch": 0.45023565289714446,
+      "grad_norm": 0.21249794960021973,
+      "learning_rate": 5.0180410877067955e-05,
+      "loss": 1.5063,
+      "step": 812
+    },
+    {
+      "epoch": 0.4507901303021902,
+      "grad_norm": 0.21143661439418793,
+      "learning_rate": 5.0109033871706754e-05,
+      "loss": 1.62,
+      "step": 813
+    },
+    {
+      "epoch": 0.45134460770723595,
+      "grad_norm": 0.21807655692100525,
+      "learning_rate": 5.0037622464639724e-05,
+      "loss": 1.585,
+      "step": 814
+    },
+    {
+      "epoch": 0.45189908511228166,
+      "grad_norm": 0.2183266580104828,
+      "learning_rate": 4.996617689888455e-05,
+      "loss": 1.6266,
+      "step": 815
+    },
+    {
+      "epoch": 0.45245356251732743,
+      "grad_norm": 0.20334866642951965,
+      "learning_rate": 4.989469741757519e-05,
+      "loss": 1.6247,
+      "step": 816
+    },
+    {
+      "epoch": 0.45300803992237315,
+      "grad_norm": 0.21051491796970367,
+      "learning_rate": 4.9823184263961e-05,
+      "loss": 1.5917,
+      "step": 817
+    },
+    {
+      "epoch": 0.4535625173274189,
+      "grad_norm": 0.2286101132631302,
+      "learning_rate": 4.975163768140596e-05,
+      "loss": 1.6318,
+      "step": 818
+    },
+    {
+      "epoch": 0.45411699473246464,
+      "grad_norm": 0.20518524944782257,
+      "learning_rate": 4.9680057913387775e-05,
+      "loss": 1.5723,
+      "step": 819
+    },
+    {
+      "epoch": 0.4546714721375104,
+      "grad_norm": 0.2158391773700714,
+      "learning_rate": 4.960844520349709e-05,
+      "loss": 1.6655,
+      "step": 820
+    },
+    {
+      "epoch": 0.45522594954255613,
+      "grad_norm": 0.20380595326423645,
+      "learning_rate": 4.953679979543666e-05,
+      "loss": 1.5681,
+      "step": 821
+    },
+    {
+      "epoch": 0.4557804269476019,
+      "grad_norm": 0.2011675089597702,
+      "learning_rate": 4.946512193302051e-05,
+      "loss": 1.5654,
+      "step": 822
+    },
+    {
+      "epoch": 0.4563349043526476,
+      "grad_norm": 0.21012163162231445,
+      "learning_rate": 4.939341186017313e-05,
+      "loss": 1.6037,
+      "step": 823
+    },
+    {
+      "epoch": 0.4568893817576934,
+      "grad_norm": 0.2087511122226715,
+      "learning_rate": 4.932166982092858e-05,
+      "loss": 1.5656,
+      "step": 824
+    },
+    {
+      "epoch": 0.4574438591627391,
+      "grad_norm": 0.21227777004241943,
+      "learning_rate": 4.924989605942973e-05,
+      "loss": 1.6027,
+      "step": 825
+    },
+    {
+      "epoch": 0.4579983365677849,
+      "grad_norm": 0.20791299641132355,
+      "learning_rate": 4.9178090819927414e-05,
+      "loss": 1.5582,
+      "step": 826
+    },
+    {
+      "epoch": 0.4585528139728306,
+      "grad_norm": 0.2057400643825531,
+      "learning_rate": 4.910625434677956e-05,
+      "loss": 1.6189,
+      "step": 827
+    },
+    {
+      "epoch": 0.45910729137787637,
+      "grad_norm": 0.20824283361434937,
+      "learning_rate": 4.903438688445043e-05,
+      "loss": 1.5573,
+      "step": 828
+    },
+    {
+      "epoch": 0.4596617687829221,
+      "grad_norm": 0.20382678508758545,
+      "learning_rate": 4.896248867750969e-05,
+      "loss": 1.5942,
+      "step": 829
+    },
+    {
+      "epoch": 0.46021624618796786,
+      "grad_norm": 0.21372130513191223,
+      "learning_rate": 4.8890559970631667e-05,
+      "loss": 1.5624,
+      "step": 830
+    },
+    {
+      "epoch": 0.4607707235930136,
+      "grad_norm": 0.20644085109233856,
+      "learning_rate": 4.881860100859446e-05,
+      "loss": 1.6326,
+      "step": 831
+    },
+    {
+      "epoch": 0.46132520099805935,
+      "grad_norm": 0.20801600813865662,
+      "learning_rate": 4.874661203627917e-05,
+      "loss": 1.5424,
+      "step": 832
+    },
+    {
+      "epoch": 0.46187967840310507,
+      "grad_norm": 0.21283288300037384,
+      "learning_rate": 4.867459329866897e-05,
+      "loss": 1.6228,
+      "step": 833
+    },
+    {
+      "epoch": 0.46243415580815084,
+      "grad_norm": 0.21098175644874573,
+      "learning_rate": 4.860254504084835e-05,
+      "loss": 1.5839,
+      "step": 834
+    },
+    {
+      "epoch": 0.46298863321319655,
+      "grad_norm": 0.20932459831237793,
+      "learning_rate": 4.853046750800228e-05,
+      "loss": 1.5911,
+      "step": 835
+    },
+    {
+      "epoch": 0.4635431106182423,
+      "grad_norm": 0.2054491639137268,
+      "learning_rate": 4.8458360945415317e-05,
+      "loss": 1.6222,
+      "step": 836
+    },
+    {
+      "epoch": 0.46409758802328804,
+      "grad_norm": 0.19949305057525635,
+      "learning_rate": 4.838622559847084e-05,
+      "loss": 1.5386,
+      "step": 837
+    },
+    {
+      "epoch": 0.4646520654283338,
+      "grad_norm": 0.21647600829601288,
+      "learning_rate": 4.831406171265015e-05,
+      "loss": 1.5216,
+      "step": 838
+    },
+    {
+      "epoch": 0.46520654283337953,
+      "grad_norm": 0.20176930725574493,
+      "learning_rate": 4.824186953353171e-05,
+      "loss": 1.5054,
+      "step": 839
+    },
+    {
+      "epoch": 0.4657610202384253,
+      "grad_norm": 0.20189246535301208,
+      "learning_rate": 4.816964930679024e-05,
+      "loss": 1.5422,
+      "step": 840
+    },
+    {
+      "epoch": 0.466315497643471,
+      "grad_norm": 0.2124892771244049,
+      "learning_rate": 4.8097401278195904e-05,
+      "loss": 1.6112,
+      "step": 841
+    },
+    {
+      "epoch": 0.4668699750485168,
+      "grad_norm": 0.20635344088077545,
+      "learning_rate": 4.8025125693613485e-05,
+      "loss": 1.55,
+      "step": 842
+    },
+    {
+      "epoch": 0.4674244524535625,
+      "grad_norm": 0.20471568405628204,
+      "learning_rate": 4.7952822799001564e-05,
+      "loss": 1.5934,
+      "step": 843
+    },
+    {
+      "epoch": 0.4679789298586083,
+      "grad_norm": 0.20640766620635986,
+      "learning_rate": 4.78804928404116e-05,
+      "loss": 1.5859,
+      "step": 844
+    },
+    {
+      "epoch": 0.468533407263654,
+      "grad_norm": 0.2087356001138687,
+      "learning_rate": 4.780813606398722e-05,
+      "loss": 1.6468,
+      "step": 845
+    },
+    {
+      "epoch": 0.4690878846686998,
+      "grad_norm": 0.20477420091629028,
+      "learning_rate": 4.7735752715963265e-05,
+      "loss": 1.5926,
+      "step": 846
+    },
+    {
+      "epoch": 0.4696423620737455,
+      "grad_norm": 0.20607273280620575,
+      "learning_rate": 4.766334304266503e-05,
+      "loss": 1.5688,
+      "step": 847
+    },
+    {
+      "epoch": 0.47019683947879126,
+      "grad_norm": 0.20167583227157593,
+      "learning_rate": 4.7590907290507396e-05,
+      "loss": 1.5408,
+      "step": 848
+    },
+    {
+      "epoch": 0.470751316883837,
+      "grad_norm": 0.22671446204185486,
+      "learning_rate": 4.751844570599395e-05,
+      "loss": 1.5693,
+      "step": 849
+    },
+    {
+      "epoch": 0.47130579428888275,
+      "grad_norm": 0.2094809114933014,
+      "learning_rate": 4.7445958535716265e-05,
+      "loss": 1.586,
+      "step": 850
+    },
+    {
+      "epoch": 0.47186027169392847,
+      "grad_norm": 0.21534985303878784,
+      "learning_rate": 4.73734460263529e-05,
+      "loss": 1.6699,
+      "step": 851
+    },
+    {
+      "epoch": 0.47241474909897424,
+      "grad_norm": 0.21628102660179138,
+      "learning_rate": 4.730090842466871e-05,
+      "loss": 1.6307,
+      "step": 852
+    },
+    {
+      "epoch": 0.47296922650401996,
+      "grad_norm": 0.21070408821105957,
+      "learning_rate": 4.7228345977513886e-05,
+      "loss": 1.5859,
+      "step": 853
+    },
+    {
+      "epoch": 0.47352370390906573,
+      "grad_norm": 0.21383161842823029,
+      "learning_rate": 4.715575893182324e-05,
+      "loss": 1.672,
+      "step": 854
+    },
+    {
+      "epoch": 0.47407818131411145,
+      "grad_norm": 0.2086363434791565,
+      "learning_rate": 4.7083147534615224e-05,
+      "loss": 1.5431,
+      "step": 855
+    },
+    {
+      "epoch": 0.4746326587191572,
+      "grad_norm": 0.1970445066690445,
+      "learning_rate": 4.70105120329912e-05,
+      "loss": 1.4742,
+      "step": 856
+    },
+    {
+      "epoch": 0.47518713612420294,
+      "grad_norm": 0.20317339897155762,
+      "learning_rate": 4.6937852674134555e-05,
+      "loss": 1.5643,
+      "step": 857
+    },
+    {
+      "epoch": 0.4757416135292487,
+      "grad_norm": 0.20548830926418304,
+      "learning_rate": 4.6865169705309815e-05,
+      "loss": 1.5369,
+      "step": 858
+    },
+    {
+      "epoch": 0.4762960909342944,
+      "grad_norm": 0.20916157960891724,
+      "learning_rate": 4.679246337386195e-05,
+      "loss": 1.544,
+      "step": 859
+    },
+    {
+      "epoch": 0.4768505683393402,
+      "grad_norm": 0.2089296132326126,
+      "learning_rate": 4.671973392721535e-05,
+      "loss": 1.5906,
+      "step": 860
+    },
+    {
+      "epoch": 0.4774050457443859,
+      "grad_norm": 0.21110884845256805,
+      "learning_rate": 4.6646981612873105e-05,
+      "loss": 1.6617,
+      "step": 861
+    },
+    {
+      "epoch": 0.4779595231494317,
+      "grad_norm": 0.20762455463409424,
+      "learning_rate": 4.65742066784161e-05,
+      "loss": 1.5733,
+      "step": 862
+    },
+    {
+      "epoch": 0.4785140005544774,
+      "grad_norm": 0.20924974977970123,
+      "learning_rate": 4.650140937150222e-05,
+      "loss": 1.6104,
+      "step": 863
+    },
+    {
+      "epoch": 0.4790684779595232,
+      "grad_norm": 0.21603722870349884,
+      "learning_rate": 4.642858993986549e-05,
+      "loss": 1.5838,
+      "step": 864
+    },
+    {
+      "epoch": 0.4796229553645689,
+      "grad_norm": 0.20661696791648865,
+      "learning_rate": 4.635574863131522e-05,
+      "loss": 1.6063,
+      "step": 865
+    },
+    {
+      "epoch": 0.48017743276961466,
+      "grad_norm": 0.2147568166255951,
+      "learning_rate": 4.6282885693735145e-05,
+      "loss": 1.5424,
+      "step": 866
+    },
+    {
+      "epoch": 0.4807319101746604,
+      "grad_norm": 0.20484501123428345,
+      "learning_rate": 4.621000137508263e-05,
+      "loss": 1.5754,
+      "step": 867
+    },
+    {
+      "epoch": 0.48128638757970615,
+      "grad_norm": 0.22287361323833466,
+      "learning_rate": 4.61370959233878e-05,
+      "loss": 1.6868,
+      "step": 868
+    },
+    {
+      "epoch": 0.48184086498475187,
+      "grad_norm": 0.21111956238746643,
+      "learning_rate": 4.6064169586752706e-05,
+      "loss": 1.4827,
+      "step": 869
+    },
+    {
+      "epoch": 0.48239534238979764,
+      "grad_norm": 0.20634272694587708,
+      "learning_rate": 4.599122261335044e-05,
+      "loss": 1.5634,
+      "step": 870
+    },
+    {
+      "epoch": 0.48294981979484336,
+      "grad_norm": 0.2151550054550171,
+      "learning_rate": 4.591825525142433e-05,
+      "loss": 1.5334,
+      "step": 871
+    },
+    {
+      "epoch": 0.4835042971998891,
+      "grad_norm": 0.21538786590099335,
+      "learning_rate": 4.584526774928713e-05,
+      "loss": 1.563,
+      "step": 872
+    },
+    {
+      "epoch": 0.48405877460493485,
+      "grad_norm": 0.2079581320285797,
+      "learning_rate": 4.5772260355320075e-05,
+      "loss": 1.5912,
+      "step": 873
+    },
+    {
+      "epoch": 0.48461325200998057,
+      "grad_norm": 0.21243925392627716,
+      "learning_rate": 4.5699233317972145e-05,
+      "loss": 1.5245,
+      "step": 874
+    },
+    {
+      "epoch": 0.48516772941502634,
+      "grad_norm": 0.20607562363147736,
+      "learning_rate": 4.562618688575911e-05,
+      "loss": 1.5385,
+      "step": 875
+    },
+    {
+      "epoch": 0.48572220682007206,
+      "grad_norm": 0.2081446647644043,
+      "learning_rate": 4.555312130726279e-05,
+      "loss": 1.589,
+      "step": 876
+    },
+    {
+      "epoch": 0.48627668422511783,
+      "grad_norm": 0.20849229395389557,
+      "learning_rate": 4.5480036831130144e-05,
+      "loss": 1.6432,
+      "step": 877
+    },
+    {
+      "epoch": 0.48683116163016354,
+      "grad_norm": 0.220262348651886,
+      "learning_rate": 4.540693370607244e-05,
+      "loss": 1.5364,
+      "step": 878
+    },
+    {
+      "epoch": 0.4873856390352093,
+      "grad_norm": 0.2163216471672058,
+      "learning_rate": 4.533381218086443e-05,
+      "loss": 1.6256,
+      "step": 879
+    },
+    {
+      "epoch": 0.48794011644025503,
+      "grad_norm": 0.21877123415470123,
+      "learning_rate": 4.5260672504343436e-05,
+      "loss": 1.6153,
+      "step": 880
+    },
+    {
+      "epoch": 0.4884945938453008,
+      "grad_norm": 0.20430409908294678,
+      "learning_rate": 4.518751492540859e-05,
+      "loss": 1.5887,
+      "step": 881
+    },
+    {
+      "epoch": 0.4890490712503465,
+      "grad_norm": 0.21504053473472595,
+      "learning_rate": 4.5114339693019924e-05,
+      "loss": 1.6251,
+      "step": 882
+    },
+    {
+      "epoch": 0.4896035486553923,
+      "grad_norm": 0.21704961359500885,
+      "learning_rate": 4.504114705619758e-05,
+      "loss": 1.6214,
+      "step": 883
+    },
+    {
+      "epoch": 0.490158026060438,
+      "grad_norm": 0.2103128433227539,
+      "learning_rate": 4.4967937264020896e-05,
+      "loss": 1.5078,
+      "step": 884
+    },
+    {
+      "epoch": 0.4907125034654838,
+      "grad_norm": 0.21497170627117157,
+      "learning_rate": 4.4894710565627585e-05,
+      "loss": 1.6292,
+      "step": 885
+    },
+    {
+      "epoch": 0.4912669808705295,
+      "grad_norm": 0.20846691727638245,
+      "learning_rate": 4.4821467210212924e-05,
+      "loss": 1.5809,
+      "step": 886
+    },
+    {
+      "epoch": 0.4918214582755753,
+      "grad_norm": 0.20751187205314636,
+      "learning_rate": 4.474820744702887e-05,
+      "loss": 1.5428,
+      "step": 887
+    },
+    {
+      "epoch": 0.492375935680621,
+      "grad_norm": 0.22106949985027313,
+      "learning_rate": 4.4674931525383176e-05,
+      "loss": 1.57,
+      "step": 888
+    },
+    {
+      "epoch": 0.49293041308566676,
+      "grad_norm": 0.21874569356441498,
+      "learning_rate": 4.460163969463864e-05,
+      "loss": 1.565,
+      "step": 889
+    },
+    {
+      "epoch": 0.4934848904907125,
+      "grad_norm": 0.21572619676589966,
+      "learning_rate": 4.452833220421216e-05,
+      "loss": 1.5302,
+      "step": 890
+    },
+    {
+      "epoch": 0.49403936789575825,
+      "grad_norm": 0.2403019368648529,
+      "learning_rate": 4.445500930357393e-05,
+      "loss": 1.5533,
+      "step": 891
+    },
+    {
+      "epoch": 0.49459384530080397,
+      "grad_norm": 0.2276349812746048,
+      "learning_rate": 4.438167124224663e-05,
+      "loss": 1.5833,
+      "step": 892
+    },
+    {
+      "epoch": 0.49514832270584974,
+      "grad_norm": 0.20668740570545197,
+      "learning_rate": 4.430831826980445e-05,
+      "loss": 1.5467,
+      "step": 893
+    },
+    {
+      "epoch": 0.49570280011089546,
+      "grad_norm": 0.21800528466701508,
+      "learning_rate": 4.4234950635872406e-05,
+      "loss": 1.5423,
+      "step": 894
+    },
+    {
+      "epoch": 0.49625727751594123,
+      "grad_norm": 0.21564136445522308,
+      "learning_rate": 4.416156859012534e-05,
+      "loss": 1.5414,
+      "step": 895
+    },
+    {
+      "epoch": 0.49681175492098695,
+      "grad_norm": 0.21869921684265137,
+      "learning_rate": 4.4088172382287205e-05,
+      "loss": 1.5703,
+      "step": 896
+    },
+    {
+      "epoch": 0.4973662323260327,
+      "grad_norm": 0.2215128093957901,
+      "learning_rate": 4.401476226213009e-05,
+      "loss": 1.5246,
+      "step": 897
+    },
+    {
+      "epoch": 0.49792070973107844,
+      "grad_norm": 0.24564118683338165,
+      "learning_rate": 4.394133847947346e-05,
+      "loss": 1.6237,
+      "step": 898
+    },
+    {
+      "epoch": 0.4984751871361242,
+      "grad_norm": 0.20941196382045746,
+      "learning_rate": 4.386790128418328e-05,
+      "loss": 1.5816,
+      "step": 899
+    },
+    {
+      "epoch": 0.4990296645411699,
+      "grad_norm": 0.21743817627429962,
+      "learning_rate": 4.3794450926171106e-05,
+      "loss": 1.5763,
+      "step": 900
+    },
+    {
+      "epoch": 0.4995841419462157,
+      "grad_norm": 0.2064601182937622,
+      "learning_rate": 4.3720987655393384e-05,
+      "loss": 1.549,
+      "step": 901
+    },
+    {
+      "epoch": 0.5001386193512615,
+      "grad_norm": 0.2088851034641266,
+      "learning_rate": 4.36475117218504e-05,
+      "loss": 1.5859,
+      "step": 902
+    },
+    {
+      "epoch": 0.5006930967563071,
+      "grad_norm": 0.21022847294807434,
+      "learning_rate": 4.357402337558561e-05,
+      "loss": 1.5749,
+      "step": 903
+    },
+    {
+      "epoch": 0.5012475741613529,
+      "grad_norm": 0.21769824624061584,
+      "learning_rate": 4.350052286668466e-05,
+      "loss": 1.6526,
+      "step": 904
+    },
+    {
+      "epoch": 0.5018020515663987,
+      "grad_norm": 0.2308962196111679,
+      "learning_rate": 4.342701044527461e-05,
+      "loss": 1.6642,
+      "step": 905
+    },
+    {
+      "epoch": 0.5023565289714444,
+      "grad_norm": 0.21604010462760925,
+      "learning_rate": 4.335348636152306e-05,
+      "loss": 1.6659,
+      "step": 906
+    },
+    {
+      "epoch": 0.5029110063764901,
+      "grad_norm": 0.20606866478919983,
+      "learning_rate": 4.3279950865637296e-05,
+      "loss": 1.5198,
+      "step": 907
+    },
+    {
+      "epoch": 0.5034654837815359,
+      "grad_norm": 0.20940081775188446,
+      "learning_rate": 4.320640420786344e-05,
+      "loss": 1.572,
+      "step": 908
+    },
+    {
+      "epoch": 0.5040199611865817,
+      "grad_norm": 0.22230613231658936,
+      "learning_rate": 4.313284663848558e-05,
+      "loss": 1.5601,
+      "step": 909
+    },
+    {
+      "epoch": 0.5045744385916274,
+      "grad_norm": 0.21619489789009094,
+      "learning_rate": 4.305927840782497e-05,
+      "loss": 1.6177,
+      "step": 910
+    },
+    {
+      "epoch": 0.5051289159966731,
+      "grad_norm": 0.2147083431482315,
+      "learning_rate": 4.298569976623912e-05,
+      "loss": 1.6277,
+      "step": 911
+    },
+    {
+      "epoch": 0.5056833934017189,
+      "grad_norm": 0.21617257595062256,
+      "learning_rate": 4.291211096412099e-05,
+      "loss": 1.6043,
+      "step": 912
+    },
+    {
+      "epoch": 0.5062378708067646,
+      "grad_norm": 0.20666974782943726,
+      "learning_rate": 4.283851225189807e-05,
+      "loss": 1.5511,
+      "step": 913
+    },
+    {
+      "epoch": 0.5067923482118104,
+      "grad_norm": 0.22238971292972565,
+      "learning_rate": 4.276490388003164e-05,
+      "loss": 1.6442,
+      "step": 914
+    },
+    {
+      "epoch": 0.5073468256168561,
+      "grad_norm": 0.21716102957725525,
+      "learning_rate": 4.269128609901581e-05,
+      "loss": 1.5887,
+      "step": 915
+    },
+    {
+      "epoch": 0.5079013030219018,
+      "grad_norm": 0.20268242061138153,
+      "learning_rate": 4.261765915937674e-05,
+      "loss": 1.5302,
+      "step": 916
+    },
+    {
+      "epoch": 0.5084557804269476,
+      "grad_norm": 0.21744856238365173,
+      "learning_rate": 4.254402331167171e-05,
+      "loss": 1.5459,
+      "step": 917
+    },
+    {
+      "epoch": 0.5090102578319934,
+      "grad_norm": 0.20796500146389008,
+      "learning_rate": 4.247037880648836e-05,
+      "loss": 1.5942,
+      "step": 918
+    },
+    {
+      "epoch": 0.509564735237039,
+      "grad_norm": 0.22836215794086456,
+      "learning_rate": 4.239672589444376e-05,
+      "loss": 1.6449,
+      "step": 919
+    },
+    {
+      "epoch": 0.5101192126420848,
+      "grad_norm": 0.20672906935214996,
+      "learning_rate": 4.232306482618362e-05,
+      "loss": 1.5657,
+      "step": 920
+    },
+    {
+      "epoch": 0.5106736900471306,
+      "grad_norm": 0.21046310663223267,
+      "learning_rate": 4.22493958523814e-05,
+      "loss": 1.5618,
+      "step": 921
+    },
+    {
+      "epoch": 0.5112281674521764,
+      "grad_norm": 0.2109476923942566,
+      "learning_rate": 4.2175719223737426e-05,
+      "loss": 1.5723,
+      "step": 922
+    },
+    {
+      "epoch": 0.511782644857222,
+      "grad_norm": 0.2226995974779129,
+      "learning_rate": 4.210203519097813e-05,
+      "loss": 1.6005,
+      "step": 923
+    },
+    {
+      "epoch": 0.5123371222622678,
+      "grad_norm": 0.21203988790512085,
+      "learning_rate": 4.202834400485508e-05,
+      "loss": 1.5898,
+      "step": 924
+    },
+    {
+      "epoch": 0.5128915996673136,
+      "grad_norm": 0.21835650503635406,
+      "learning_rate": 4.1954645916144245e-05,
+      "loss": 1.5458,
+      "step": 925
+    },
+    {
+      "epoch": 0.5134460770723593,
+      "grad_norm": 0.22072114050388336,
+      "learning_rate": 4.188094117564505e-05,
+      "loss": 1.5888,
+      "step": 926
+    },
+    {
+      "epoch": 0.514000554477405,
+      "grad_norm": 0.20924319326877594,
+      "learning_rate": 4.1807230034179566e-05,
+      "loss": 1.5856,
+      "step": 927
+    },
+    {
+      "epoch": 0.5145550318824508,
+      "grad_norm": 0.21937541663646698,
+      "learning_rate": 4.1733512742591646e-05,
+      "loss": 1.5418,
+      "step": 928
+    },
+    {
+      "epoch": 0.5151095092874965,
+      "grad_norm": 0.2233019769191742,
+      "learning_rate": 4.165978955174606e-05,
+      "loss": 1.6723,
+      "step": 929
+    },
+    {
+      "epoch": 0.5156639866925423,
+      "grad_norm": 0.22729487717151642,
+      "learning_rate": 4.15860607125277e-05,
+      "loss": 1.5608,
+      "step": 930
+    },
+    {
+      "epoch": 0.516218464097588,
+      "grad_norm": 0.20513324439525604,
+      "learning_rate": 4.151232647584061e-05,
+      "loss": 1.5652,
+      "step": 931
+    },
+    {
+      "epoch": 0.5167729415026338,
+      "grad_norm": 0.2118518203496933,
+      "learning_rate": 4.143858709260726e-05,
+      "loss": 1.6201,
+      "step": 932
+    },
+    {
+      "epoch": 0.5173274189076795,
+      "grad_norm": 0.20294101536273956,
+      "learning_rate": 4.136484281376758e-05,
+      "loss": 1.5273,
+      "step": 933
+    },
+    {
+      "epoch": 0.5178818963127253,
+      "grad_norm": 0.21510812640190125,
+      "learning_rate": 4.1291093890278244e-05,
+      "loss": 1.5741,
+      "step": 934
+    },
+    {
+      "epoch": 0.518436373717771,
+      "grad_norm": 0.20787851512432098,
+      "learning_rate": 4.1217340573111625e-05,
+      "loss": 1.6002,
+      "step": 935
+    },
+    {
+      "epoch": 0.5189908511228167,
+      "grad_norm": 0.20799653232097626,
+      "learning_rate": 4.114358311325513e-05,
+      "loss": 1.541,
+      "step": 936
+    },
+    {
+      "epoch": 0.5195453285278625,
+      "grad_norm": 0.20286118984222412,
+      "learning_rate": 4.106982176171025e-05,
+      "loss": 1.4923,
+      "step": 937
+    },
+    {
+      "epoch": 0.5200998059329083,
+      "grad_norm": 0.2020486742258072,
+      "learning_rate": 4.099605676949169e-05,
+      "loss": 1.4772,
+      "step": 938
+    },
+    {
+      "epoch": 0.5206542833379539,
+      "grad_norm": 0.21534226834774017,
+      "learning_rate": 4.0922288387626536e-05,
+      "loss": 1.5663,
+      "step": 939
+    },
+    {
+      "epoch": 0.5212087607429997,
+      "grad_norm": 0.24164481461048126,
+      "learning_rate": 4.0848516867153474e-05,
+      "loss": 1.6155,
+      "step": 940
+    },
+    {
+      "epoch": 0.5217632381480455,
+      "grad_norm": 0.20027059316635132,
+      "learning_rate": 4.077474245912182e-05,
+      "loss": 1.5143,
+      "step": 941
+    },
+    {
+      "epoch": 0.5223177155530913,
+      "grad_norm": 0.2217823565006256,
+      "learning_rate": 4.070096541459071e-05,
+      "loss": 1.5583,
+      "step": 942
+    },
+    {
+      "epoch": 0.5228721929581369,
+      "grad_norm": 0.21679724752902985,
+      "learning_rate": 4.0627185984628295e-05,
+      "loss": 1.5588,
+      "step": 943
+    },
+    {
+      "epoch": 0.5234266703631827,
+      "grad_norm": 0.20818306505680084,
+      "learning_rate": 4.055340442031079e-05,
+      "loss": 1.5417,
+      "step": 944
+    },
+    {
+      "epoch": 0.5239811477682285,
+      "grad_norm": 0.21085157990455627,
+      "learning_rate": 4.0479620972721726e-05,
+      "loss": 1.5735,
+      "step": 945
+    },
+    {
+      "epoch": 0.5245356251732742,
+      "grad_norm": 0.21629397571086884,
+      "learning_rate": 4.040583589295101e-05,
+      "loss": 1.6296,
+      "step": 946
+    },
+    {
+      "epoch": 0.5250901025783199,
+      "grad_norm": 0.2124132364988327,
+      "learning_rate": 4.03320494320941e-05,
+      "loss": 1.5781,
+      "step": 947
+    },
+    {
+      "epoch": 0.5256445799833657,
+      "grad_norm": 0.20531564950942993,
+      "learning_rate": 4.0258261841251175e-05,
+      "loss": 1.5726,
+      "step": 948
+    },
+    {
+      "epoch": 0.5261990573884114,
+      "grad_norm": 0.20956727862358093,
+      "learning_rate": 4.018447337152626e-05,
+      "loss": 1.5653,
+      "step": 949
+    },
+    {
+      "epoch": 0.5267535347934572,
+      "grad_norm": 0.20362792909145355,
+      "learning_rate": 4.0110684274026365e-05,
+      "loss": 1.576,
+      "step": 950
+    },
+    {
+      "epoch": 0.5273080121985029,
+      "grad_norm": 0.20399406552314758,
+      "learning_rate": 4.00368947998606e-05,
+      "loss": 1.5732,
+      "step": 951
+    },
+    {
+      "epoch": 0.5278624896035486,
+      "grad_norm": 0.21037329733371735,
+      "learning_rate": 3.9963105200139415e-05,
+      "loss": 1.5909,
+      "step": 952
+    },
+    {
+      "epoch": 0.5284169670085944,
+      "grad_norm": 0.2091919481754303,
+      "learning_rate": 3.9889315725973655e-05,
+      "loss": 1.5478,
+      "step": 953
+    },
+    {
+      "epoch": 0.5289714444136402,
+      "grad_norm": 0.2109626978635788,
+      "learning_rate": 3.981552662847375e-05,
+      "loss": 1.5515,
+      "step": 954
+    },
+    {
+      "epoch": 0.5295259218186859,
+      "grad_norm": 0.21235403418540955,
+      "learning_rate": 3.9741738158748824e-05,
+      "loss": 1.5435,
+      "step": 955
+    },
+    {
+      "epoch": 0.5300803992237316,
+      "grad_norm": 0.2205905020236969,
+      "learning_rate": 3.966795056790591e-05,
+      "loss": 1.6312,
+      "step": 956
+    },
+    {
+      "epoch": 0.5306348766287774,
+      "grad_norm": 0.2057129591703415,
+      "learning_rate": 3.959416410704901e-05,
+      "loss": 1.5718,
+      "step": 957
+    },
+    {
+      "epoch": 0.5311893540338232,
+      "grad_norm": 0.2083214968442917,
+      "learning_rate": 3.952037902727829e-05,
+      "loss": 1.5801,
+      "step": 958
+    },
+    {
+      "epoch": 0.5317438314388688,
+      "grad_norm": 0.21389059722423553,
+      "learning_rate": 3.944659557968922e-05,
+      "loss": 1.5781,
+      "step": 959
+    },
+    {
+      "epoch": 0.5322983088439146,
+      "grad_norm": 0.20008264482021332,
+      "learning_rate": 3.937281401537171e-05,
+      "loss": 1.5221,
+      "step": 960
+    },
+    {
+      "epoch": 0.5328527862489604,
+      "grad_norm": 0.22723457217216492,
+      "learning_rate": 3.92990345854093e-05,
+      "loss": 1.6409,
+      "step": 961
+    },
+    {
+      "epoch": 0.5334072636540061,
+      "grad_norm": 0.21602332592010498,
+      "learning_rate": 3.922525754087819e-05,
+      "loss": 1.624,
+      "step": 962
+    },
+    {
+      "epoch": 0.5339617410590518,
+      "grad_norm": 0.2093724012374878,
+      "learning_rate": 3.915148313284653e-05,
+      "loss": 1.5583,
+      "step": 963
+    },
+    {
+      "epoch": 0.5345162184640976,
+      "grad_norm": 0.20897488296031952,
+      "learning_rate": 3.907771161237347e-05,
+      "loss": 1.6248,
+      "step": 964
+    },
+    {
+      "epoch": 0.5350706958691434,
+      "grad_norm": 0.21805384755134583,
+      "learning_rate": 3.900394323050833e-05,
+      "loss": 1.6227,
+      "step": 965
+    },
+    {
+      "epoch": 0.5356251732741891,
+      "grad_norm": 0.22409921884536743,
+      "learning_rate": 3.893017823828977e-05,
+      "loss": 1.5582,
+      "step": 966
+    },
+    {
+      "epoch": 0.5361796506792348,
+      "grad_norm": 0.22005674242973328,
+      "learning_rate": 3.8856416886744874e-05,
+      "loss": 1.6032,
+      "step": 967
+    },
+    {
+      "epoch": 0.5367341280842806,
+      "grad_norm": 0.20452825725078583,
+      "learning_rate": 3.878265942688838e-05,
+      "loss": 1.5501,
+      "step": 968
+    },
+    {
+      "epoch": 0.5372886054893263,
+      "grad_norm": 0.21274332702159882,
+      "learning_rate": 3.8708906109721776e-05,
+      "loss": 1.5386,
+      "step": 969
+    },
+    {
+      "epoch": 0.5378430828943721,
+      "grad_norm": 0.21413595974445343,
+      "learning_rate": 3.863515718623242e-05,
+      "loss": 1.6382,
+      "step": 970
+    },
+    {
+      "epoch": 0.5383975602994178,
+      "grad_norm": 0.22043178975582123,
+      "learning_rate": 3.856141290739276e-05,
+      "loss": 1.5306,
+      "step": 971
+    },
+    {
+      "epoch": 0.5389520377044635,
+      "grad_norm": 0.2085334211587906,
+      "learning_rate": 3.8487673524159404e-05,
+      "loss": 1.5452,
+      "step": 972
+    },
+    {
+      "epoch": 0.5395065151095093,
+      "grad_norm": 0.2164732664823532,
+      "learning_rate": 3.8413939287472305e-05,
+      "loss": 1.6273,
+      "step": 973
+    },
+    {
+      "epoch": 0.5400609925145551,
+      "grad_norm": 0.23173940181732178,
+      "learning_rate": 3.8340210448253945e-05,
+      "loss": 1.629,
+      "step": 974
+    },
+    {
+      "epoch": 0.5406154699196007,
+      "grad_norm": 0.20250822603702545,
+      "learning_rate": 3.826648725740836e-05,
+      "loss": 1.5313,
+      "step": 975
+    },
+    {
+      "epoch": 0.5411699473246465,
+      "grad_norm": 0.20862893760204315,
+      "learning_rate": 3.819276996582045e-05,
+      "loss": 1.574,
+      "step": 976
+    },
+    {
+      "epoch": 0.5417244247296923,
+      "grad_norm": 0.2386544793844223,
+      "learning_rate": 3.8119058824354966e-05,
+      "loss": 1.6064,
+      "step": 977
+    },
+    {
+      "epoch": 0.5422789021347381,
+      "grad_norm": 0.2093050479888916,
+      "learning_rate": 3.804535408385577e-05,
+      "loss": 1.5596,
+      "step": 978
+    },
+    {
+      "epoch": 0.5428333795397837,
+      "grad_norm": 0.22440102696418762,
+      "learning_rate": 3.7971655995144937e-05,
+      "loss": 1.5967,
+      "step": 979
+    },
+    {
+      "epoch": 0.5433878569448295,
+      "grad_norm": 0.21898169815540314,
+      "learning_rate": 3.789796480902188e-05,
+      "loss": 1.5639,
+      "step": 980
+    },
+    {
+      "epoch": 0.5439423343498753,
+      "grad_norm": 0.21143299341201782,
+      "learning_rate": 3.782428077626259e-05,
+      "loss": 1.5549,
+      "step": 981
+    },
+    {
+      "epoch": 0.544496811754921,
+      "grad_norm": 0.21122704446315765,
+      "learning_rate": 3.775060414761861e-05,
+      "loss": 1.5778,
+      "step": 982
+    },
+    {
+      "epoch": 0.5450512891599667,
+      "grad_norm": 0.22585324943065643,
+      "learning_rate": 3.767693517381638e-05,
+      "loss": 1.6119,
+      "step": 983
+    },
+    {
+      "epoch": 0.5456057665650125,
+      "grad_norm": 0.1983765810728073,
+      "learning_rate": 3.760327410555625e-05,
+      "loss": 1.4554,
+      "step": 984
+    },
+    {
+      "epoch": 0.5461602439700582,
+      "grad_norm": 0.210382342338562,
+      "learning_rate": 3.752962119351166e-05,
+      "loss": 1.562,
+      "step": 985
+    },
+    {
+      "epoch": 0.546714721375104,
+      "grad_norm": 0.2129284292459488,
+      "learning_rate": 3.745597668832831e-05,
+      "loss": 1.6004,
+      "step": 986
+    },
+    {
+      "epoch": 0.5472691987801497,
+      "grad_norm": 0.20643486082553864,
+      "learning_rate": 3.738234084062327e-05,
+      "loss": 1.5137,
+      "step": 987
+    },
+    {
+      "epoch": 0.5478236761851955,
+      "grad_norm": 0.20947743952274323,
+      "learning_rate": 3.730871390098419e-05,
+      "loss": 1.5833,
+      "step": 988
+    },
+    {
+      "epoch": 0.5483781535902412,
+      "grad_norm": 0.2061213254928589,
+      "learning_rate": 3.723509611996837e-05,
+      "loss": 1.5231,
+      "step": 989
+    },
+    {
+      "epoch": 0.548932630995287,
+      "grad_norm": 0.21855583786964417,
+      "learning_rate": 3.716148774810194e-05,
+      "loss": 1.5806,
+      "step": 990
+    },
+    {
+      "epoch": 0.5494871084003327,
+      "grad_norm": 0.21223531663417816,
+      "learning_rate": 3.708788903587904e-05,
+      "loss": 1.5188,
+      "step": 991
+    },
+    {
+      "epoch": 0.5500415858053784,
+      "grad_norm": 0.20913442969322205,
+      "learning_rate": 3.701430023376089e-05,
+      "loss": 1.6229,
+      "step": 992
+    },
+    {
+      "epoch": 0.5505960632104242,
+      "grad_norm": 0.2168632298707962,
+      "learning_rate": 3.6940721592175026e-05,
+      "loss": 1.5483,
+      "step": 993
+    },
+    {
+      "epoch": 0.55115054061547,
+      "grad_norm": 0.21936000883579254,
+      "learning_rate": 3.686715336151443e-05,
+      "loss": 1.5724,
+      "step": 994
+    },
+    {
+      "epoch": 0.5517050180205156,
+      "grad_norm": 0.21921208500862122,
+      "learning_rate": 3.6793595792136565e-05,
+      "loss": 1.6074,
+      "step": 995
+    },
+    {
+      "epoch": 0.5522594954255614,
+      "grad_norm": 0.21475403010845184,
+      "learning_rate": 3.672004913436271e-05,
+      "loss": 1.6166,
+      "step": 996
+    },
+    {
+      "epoch": 0.5528139728306072,
+      "grad_norm": 0.2194800078868866,
+      "learning_rate": 3.664651363847695e-05,
+      "loss": 1.5691,
+      "step": 997
+    },
+    {
+      "epoch": 0.5533684502356528,
+      "grad_norm": 0.21276038885116577,
+      "learning_rate": 3.65729895547254e-05,
+      "loss": 1.5916,
+      "step": 998
+    },
+    {
+      "epoch": 0.5539229276406986,
+      "grad_norm": 0.2259768396615982,
+      "learning_rate": 3.649947713331536e-05,
+      "loss": 1.5617,
+      "step": 999
+    },
+    {
+      "epoch": 0.5544774050457444,
+      "grad_norm": 0.21104209125041962,
+      "learning_rate": 3.64259766244144e-05,
+      "loss": 1.5882,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5550318824507902,
+      "grad_norm": 0.21075722575187683,
+      "learning_rate": 3.63524882781496e-05,
+      "loss": 1.6089,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5555863598558358,
+      "grad_norm": 0.23340259492397308,
+      "learning_rate": 3.627901234460663e-05,
+      "loss": 1.6235,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5561408372608816,
+      "grad_norm": 0.22706355154514313,
+      "learning_rate": 3.6205549073828894e-05,
+      "loss": 1.5388,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5566953146659274,
+      "grad_norm": 0.22355878353118896,
+      "learning_rate": 3.613209871581674e-05,
+      "loss": 1.5329,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5572497920709731,
+      "grad_norm": 0.22762571275234222,
+      "learning_rate": 3.6058661520526555e-05,
+      "loss": 1.5367,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5578042694760188,
+      "grad_norm": 0.21231630444526672,
+      "learning_rate": 3.598523773786991e-05,
+      "loss": 1.5366,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5583587468810646,
+      "grad_norm": 0.22039973735809326,
+      "learning_rate": 3.591182761771281e-05,
+      "loss": 1.6085,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5589132242861103,
+      "grad_norm": 0.22262850403785706,
+      "learning_rate": 3.5838431409874666e-05,
+      "loss": 1.5297,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5594677016911561,
+      "grad_norm": 0.20765410363674164,
+      "learning_rate": 3.576504936412762e-05,
+      "loss": 1.5488,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5600221790962018,
+      "grad_norm": 0.21379908919334412,
+      "learning_rate": 3.5691681730195564e-05,
+      "loss": 1.581,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5605766565012475,
+      "grad_norm": 0.2365419566631317,
+      "learning_rate": 3.561832875775338e-05,
+      "loss": 1.6571,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5611311339062933,
+      "grad_norm": 0.2150985300540924,
+      "learning_rate": 3.5544990696426077e-05,
+      "loss": 1.546,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5616856113113391,
+      "grad_norm": 0.20792889595031738,
+      "learning_rate": 3.5471667795787845e-05,
+      "loss": 1.5404,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5622400887163848,
+      "grad_norm": 0.23230226337909698,
+      "learning_rate": 3.5398360305361375e-05,
+      "loss": 1.5393,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5627945661214305,
+      "grad_norm": 0.20976661145687103,
+      "learning_rate": 3.532506847461684e-05,
+      "loss": 1.5753,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5633490435264763,
+      "grad_norm": 0.21504442393779755,
+      "learning_rate": 3.5251792552971146e-05,
+      "loss": 1.5944,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5639035209315221,
+      "grad_norm": 0.21869289875030518,
+      "learning_rate": 3.517853278978708e-05,
+      "loss": 1.5166,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5644579983365677,
+      "grad_norm": 0.22892804443836212,
+      "learning_rate": 3.510528943437243e-05,
+      "loss": 1.6643,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5650124757416135,
+      "grad_norm": 0.21030141413211823,
+      "learning_rate": 3.503206273597913e-05,
+      "loss": 1.5408,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5655669531466593,
+      "grad_norm": 0.23448987305164337,
+      "learning_rate": 3.495885294380243e-05,
+      "loss": 1.5352,
+      "step": 1020
+    },
+    {
+      "epoch": 0.566121430551705,
+      "grad_norm": 0.20386260747909546,
+      "learning_rate": 3.488566030698008e-05,
+      "loss": 1.5978,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5666759079567507,
+      "grad_norm": 0.22522738575935364,
+      "learning_rate": 3.481248507459143e-05,
+      "loss": 1.6675,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5672303853617965,
+      "grad_norm": 0.2048107385635376,
+      "learning_rate": 3.4739327495656584e-05,
+      "loss": 1.4263,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5677848627668423,
+      "grad_norm": 0.21060484647750854,
+      "learning_rate": 3.466618781913559e-05,
+      "loss": 1.5268,
+      "step": 1024
+    },
+    {
+      "epoch": 0.568339340171888,
+      "grad_norm": 0.22075191140174866,
+      "learning_rate": 3.459306629392757e-05,
+      "loss": 1.6015,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5688938175769337,
+      "grad_norm": 0.20704005658626556,
+      "learning_rate": 3.4519963168869855e-05,
+      "loss": 1.4969,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5694482949819795,
+      "grad_norm": 0.21155935525894165,
+      "learning_rate": 3.444687869273722e-05,
+      "loss": 1.5355,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5700027723870252,
+      "grad_norm": 0.21162542700767517,
+      "learning_rate": 3.437381311424091e-05,
+      "loss": 1.5899,
+      "step": 1028
+    },
+    {
+      "epoch": 0.570557249792071,
+      "grad_norm": 0.2058079093694687,
+      "learning_rate": 3.4300766682027875e-05,
+      "loss": 1.4967,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5711117271971167,
+      "grad_norm": 0.21342620253562927,
+      "learning_rate": 3.422773964467994e-05,
+      "loss": 1.5941,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5716662046021624,
+      "grad_norm": 0.21482619643211365,
+      "learning_rate": 3.4154732250712876e-05,
+      "loss": 1.5887,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5722206820072082,
+      "grad_norm": 0.20948712527751923,
+      "learning_rate": 3.4081744748575684e-05,
+      "loss": 1.5695,
+      "step": 1032
+    },
+    {
+      "epoch": 0.572775159412254,
+      "grad_norm": 0.2241201102733612,
+      "learning_rate": 3.400877738664958e-05,
+      "loss": 1.5542,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5733296368172996,
+      "grad_norm": 0.21154385805130005,
+      "learning_rate": 3.39358304132473e-05,
+      "loss": 1.5349,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5738841142223454,
+      "grad_norm": 0.22654007375240326,
+      "learning_rate": 3.386290407661221e-05,
+      "loss": 1.6839,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5744385916273912,
+      "grad_norm": 0.2175297737121582,
+      "learning_rate": 3.3789998624917376e-05,
+      "loss": 1.5791,
+      "step": 1036
+    },
+    {
+      "epoch": 0.574993069032437,
+      "grad_norm": 0.2157989740371704,
+      "learning_rate": 3.3717114306264875e-05,
+      "loss": 1.5716,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5755475464374826,
+      "grad_norm": 0.2078414410352707,
+      "learning_rate": 3.364425136868479e-05,
+      "loss": 1.586,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5761020238425284,
+      "grad_norm": 0.2155274897813797,
+      "learning_rate": 3.357141006013451e-05,
+      "loss": 1.5882,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5766565012475742,
+      "grad_norm": 0.21952180564403534,
+      "learning_rate": 3.349859062849779e-05,
+      "loss": 1.6528,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5772109786526199,
+      "grad_norm": 0.2106204628944397,
+      "learning_rate": 3.3425793321583914e-05,
+      "loss": 1.5356,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5777654560576656,
+      "grad_norm": 0.2236543446779251,
+      "learning_rate": 3.335301838712692e-05,
+      "loss": 1.5639,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5783199334627114,
+      "grad_norm": 0.21134650707244873,
+      "learning_rate": 3.328026607278466e-05,
+      "loss": 1.5635,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5788744108677571,
+      "grad_norm": 0.21437637507915497,
+      "learning_rate": 3.3207536626138046e-05,
+      "loss": 1.557,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5794288882728029,
+      "grad_norm": 0.21564576029777527,
+      "learning_rate": 3.313483029469019e-05,
+      "loss": 1.548,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5799833656778486,
+      "grad_norm": 0.21482598781585693,
+      "learning_rate": 3.306214732586546e-05,
+      "loss": 1.5844,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5805378430828944,
+      "grad_norm": 0.20883773267269135,
+      "learning_rate": 3.2989487967008806e-05,
+      "loss": 1.5653,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5810923204879401,
+      "grad_norm": 0.20583604276180267,
+      "learning_rate": 3.291685246538478e-05,
+      "loss": 1.5298,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5816467978929859,
+      "grad_norm": 0.20545633137226105,
+      "learning_rate": 3.2844241068176766e-05,
+      "loss": 1.5625,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5822012752980316,
+      "grad_norm": 0.2127029448747635,
+      "learning_rate": 3.277165402248612e-05,
+      "loss": 1.6048,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5827557527030773,
+      "grad_norm": 0.20467516779899597,
+      "learning_rate": 3.26990915753313e-05,
+      "loss": 1.5292,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5833102301081231,
+      "grad_norm": 0.20428596436977386,
+      "learning_rate": 3.2626553973647115e-05,
+      "loss": 1.5542,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5838647075131689,
+      "grad_norm": 0.21240785717964172,
+      "learning_rate": 3.255404146428375e-05,
+      "loss": 1.5601,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5844191849182145,
+      "grad_norm": 0.20584693551063538,
+      "learning_rate": 3.248155429400605e-05,
+      "loss": 1.5476,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5849736623232603,
+      "grad_norm": 0.21363112330436707,
+      "learning_rate": 3.2409092709492624e-05,
+      "loss": 1.6318,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5855281397283061,
+      "grad_norm": 0.22125588357448578,
+      "learning_rate": 3.233665695733498e-05,
+      "loss": 1.5496,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5860826171333519,
+      "grad_norm": 0.22506000101566315,
+      "learning_rate": 3.2264247284036755e-05,
+      "loss": 1.547,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5866370945383975,
+      "grad_norm": 0.21581967175006866,
+      "learning_rate": 3.2191863936012794e-05,
+      "loss": 1.5397,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5871915719434433,
+      "grad_norm": 0.21308638155460358,
+      "learning_rate": 3.211950715958841e-05,
+      "loss": 1.5873,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5877460493484891,
+      "grad_norm": 0.20769579708576202,
+      "learning_rate": 3.2047177200998456e-05,
+      "loss": 1.5548,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5883005267535348,
+      "grad_norm": 0.20457741618156433,
+      "learning_rate": 3.197487430638652e-05,
+      "loss": 1.55,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5888550041585805,
+      "grad_norm": 0.21090541779994965,
+      "learning_rate": 3.1902598721804096e-05,
+      "loss": 1.6617,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5894094815636263,
+      "grad_norm": 0.21099694073200226,
+      "learning_rate": 3.1830350693209774e-05,
+      "loss": 1.5077,
+      "step": 1063
+    },
+    {
+      "epoch": 0.589963958968672,
+      "grad_norm": 0.21446220576763153,
+      "learning_rate": 3.1758130466468293e-05,
+      "loss": 1.6078,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5905184363737178,
+      "grad_norm": 0.21016094088554382,
+      "learning_rate": 3.168593828734986e-05,
+      "loss": 1.5688,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5910729137787635,
+      "grad_norm": 0.21141645312309265,
+      "learning_rate": 3.161377440152918e-05,
+      "loss": 1.5527,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5916273911838092,
+      "grad_norm": 0.2074812948703766,
+      "learning_rate": 3.154163905458469e-05,
+      "loss": 1.5399,
+      "step": 1067
+    },
+    {
+      "epoch": 0.592181868588855,
+      "grad_norm": 0.2177177220582962,
+      "learning_rate": 3.146953249199774e-05,
+      "loss": 1.599,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5927363459939008,
+      "grad_norm": 0.20859509706497192,
+      "learning_rate": 3.139745495915166e-05,
+      "loss": 1.5818,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5932908233989465,
+      "grad_norm": 0.20710986852645874,
+      "learning_rate": 3.1325406701331056e-05,
+      "loss": 1.559,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5938453008039922,
+      "grad_norm": 0.21879532933235168,
+      "learning_rate": 3.1253387963720835e-05,
+      "loss": 1.5392,
+      "step": 1071
+    },
+    {
+      "epoch": 0.594399778209038,
+      "grad_norm": 0.2117648422718048,
+      "learning_rate": 3.118139899140553e-05,
+      "loss": 1.6009,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5949542556140838,
+      "grad_norm": 0.21254301071166992,
+      "learning_rate": 3.110944002936835e-05,
+      "loss": 1.5663,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5955087330191294,
+      "grad_norm": 0.2058122456073761,
+      "learning_rate": 3.1037511322490324e-05,
+      "loss": 1.5066,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5960632104241752,
+      "grad_norm": 0.2119075357913971,
+      "learning_rate": 3.096561311554959e-05,
+      "loss": 1.5925,
+      "step": 1075
+    },
+    {
+      "epoch": 0.596617687829221,
+      "grad_norm": 0.2224556803703308,
+      "learning_rate": 3.089374565322045e-05,
+      "loss": 1.6041,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5971721652342667,
+      "grad_norm": 0.21044711768627167,
+      "learning_rate": 3.082190918007259e-05,
+      "loss": 1.5778,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5977266426393124,
+      "grad_norm": 0.21917963027954102,
+      "learning_rate": 3.0750103940570284e-05,
+      "loss": 1.668,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5982811200443582,
+      "grad_norm": 0.21519285440444946,
+      "learning_rate": 3.067833017907144e-05,
+      "loss": 1.567,
+      "step": 1079
+    },
+    {
+      "epoch": 0.598835597449404,
+      "grad_norm": 0.21112053096294403,
+      "learning_rate": 3.0606588139826884e-05,
+      "loss": 1.5426,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5993900748544497,
+      "grad_norm": 0.21877877414226532,
+      "learning_rate": 3.0534878066979494e-05,
+      "loss": 1.5749,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5999445522594954,
+      "grad_norm": 0.2124583125114441,
+      "learning_rate": 3.046320020456334e-05,
+      "loss": 1.4941,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6004990296645412,
+      "grad_norm": 0.21378760039806366,
+      "learning_rate": 3.0391554796502925e-05,
+      "loss": 1.5903,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6010535070695869,
+      "grad_norm": 0.21284620463848114,
+      "learning_rate": 3.031994208661223e-05,
+      "loss": 1.5439,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6016079844746327,
+      "grad_norm": 0.2211058884859085,
+      "learning_rate": 3.0248362318594055e-05,
+      "loss": 1.5932,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6021624618796784,
+      "grad_norm": 0.20590293407440186,
+      "learning_rate": 3.0176815736039007e-05,
+      "loss": 1.5591,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6027169392847241,
+      "grad_norm": 0.21580100059509277,
+      "learning_rate": 3.010530258242483e-05,
+      "loss": 1.5505,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6032714166897699,
+      "grad_norm": 0.2178206592798233,
+      "learning_rate": 3.0033823101115473e-05,
+      "loss": 1.4999,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6038258940948157,
+      "grad_norm": 0.22579710185527802,
+      "learning_rate": 2.9962377535360286e-05,
+      "loss": 1.622,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6043803714998613,
+      "grad_norm": 0.22102071344852448,
+      "learning_rate": 2.9890966128293243e-05,
+      "loss": 1.5054,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6049348489049071,
+      "grad_norm": 0.2073606252670288,
+      "learning_rate": 2.9819589122932055e-05,
+      "loss": 1.4963,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6054893263099529,
+      "grad_norm": 0.20773765444755554,
+      "learning_rate": 2.9748246762177326e-05,
+      "loss": 1.5046,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6060438037149987,
+      "grad_norm": 0.2178242802619934,
+      "learning_rate": 2.9676939288811825e-05,
+      "loss": 1.623,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6065982811200443,
+      "grad_norm": 0.21794721484184265,
+      "learning_rate": 2.960566694549954e-05,
+      "loss": 1.5074,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6071527585250901,
+      "grad_norm": 0.22172760963439941,
+      "learning_rate": 2.953442997478494e-05,
+      "loss": 1.6091,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6077072359301359,
+      "grad_norm": 0.20793460309505463,
+      "learning_rate": 2.9463228619092132e-05,
+      "loss": 1.5604,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6082617133351816,
+      "grad_norm": 0.2183390110731125,
+      "learning_rate": 2.9392063120723962e-05,
+      "loss": 1.4959,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6088161907402273,
+      "grad_norm": 0.2197534292936325,
+      "learning_rate": 2.9320933721861335e-05,
+      "loss": 1.536,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6093706681452731,
+      "grad_norm": 0.2130371481180191,
+      "learning_rate": 2.9249840664562242e-05,
+      "loss": 1.6402,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6099251455503188,
+      "grad_norm": 0.21440336108207703,
+      "learning_rate": 2.917878419076102e-05,
+      "loss": 1.5631,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6104796229553646,
+      "grad_norm": 0.22678157687187195,
+      "learning_rate": 2.9107764542267536e-05,
+      "loss": 1.5603,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6110341003604103,
+      "grad_norm": 0.21499857306480408,
+      "learning_rate": 2.903678196076628e-05,
+      "loss": 1.642,
+      "step": 1102
+    },
+    {
+      "epoch": 0.611588577765456,
+      "grad_norm": 0.20951659977436066,
+      "learning_rate": 2.896583668781569e-05,
+      "loss": 1.5311,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6121430551705018,
+      "grad_norm": 0.2207638919353485,
+      "learning_rate": 2.8894928964847133e-05,
+      "loss": 1.5827,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6126975325755476,
+      "grad_norm": 0.2097187638282776,
+      "learning_rate": 2.8824059033164262e-05,
+      "loss": 1.5433,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6132520099805933,
+      "grad_norm": 0.20598512887954712,
+      "learning_rate": 2.875322713394213e-05,
+      "loss": 1.5966,
+      "step": 1106
+    },
+    {
+      "epoch": 0.613806487385639,
+      "grad_norm": 0.20717155933380127,
+      "learning_rate": 2.8682433508226303e-05,
+      "loss": 1.4861,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6143609647906848,
+      "grad_norm": 0.20514699816703796,
+      "learning_rate": 2.8611678396932164e-05,
+      "loss": 1.5664,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6149154421957306,
+      "grad_norm": 0.21211731433868408,
+      "learning_rate": 2.8540962040843982e-05,
+      "loss": 1.5026,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6154699196007762,
+      "grad_norm": 0.21061904728412628,
+      "learning_rate": 2.8470284680614172e-05,
+      "loss": 1.5795,
+      "step": 1110
+    },
+    {
+      "epoch": 0.616024397005822,
+      "grad_norm": 0.2107418179512024,
+      "learning_rate": 2.8399646556762435e-05,
+      "loss": 1.6298,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6165788744108678,
+      "grad_norm": 0.20848070085048676,
+      "learning_rate": 2.832904790967492e-05,
+      "loss": 1.5619,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6171333518159136,
+      "grad_norm": 0.21443457901477814,
+      "learning_rate": 2.8258488979603488e-05,
+      "loss": 1.5624,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6176878292209592,
+      "grad_norm": 0.21567028760910034,
+      "learning_rate": 2.8187970006664784e-05,
+      "loss": 1.5691,
+      "step": 1114
+    },
+    {
+      "epoch": 0.618242306626005,
+      "grad_norm": 0.20712579786777496,
+      "learning_rate": 2.8117491230839502e-05,
+      "loss": 1.5106,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6187967840310508,
+      "grad_norm": 0.2100321501493454,
+      "learning_rate": 2.8047052891971565e-05,
+      "loss": 1.5948,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6193512614360965,
+      "grad_norm": 0.21011658012866974,
+      "learning_rate": 2.7976655229767248e-05,
+      "loss": 1.5616,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6199057388411422,
+      "grad_norm": 0.22047735750675201,
+      "learning_rate": 2.790629848379442e-05,
+      "loss": 1.6123,
+      "step": 1118
+    },
+    {
+      "epoch": 0.620460216246188,
+      "grad_norm": 0.2138945609331131,
+      "learning_rate": 2.78359828934817e-05,
+      "loss": 1.5528,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6210146936512337,
+      "grad_norm": 0.216939777135849,
+      "learning_rate": 2.7765708698117663e-05,
+      "loss": 1.5472,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6215691710562794,
+      "grad_norm": 0.22868293523788452,
+      "learning_rate": 2.7695476136850018e-05,
+      "loss": 1.5588,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6221236484613252,
+      "grad_norm": 0.2176186591386795,
+      "learning_rate": 2.7625285448684757e-05,
+      "loss": 1.5469,
+      "step": 1122
+    },
+    {
+      "epoch": 0.622678125866371,
+      "grad_norm": 0.21710044145584106,
+      "learning_rate": 2.7555136872485415e-05,
+      "loss": 1.5845,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6232326032714167,
+      "grad_norm": 0.21951667964458466,
+      "learning_rate": 2.748503064697221e-05,
+      "loss": 1.5436,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6237870806764624,
+      "grad_norm": 0.2149830311536789,
+      "learning_rate": 2.7414967010721212e-05,
+      "loss": 1.5896,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6243415580815082,
+      "grad_norm": 0.21103326976299286,
+      "learning_rate": 2.7344946202163592e-05,
+      "loss": 1.6166,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6248960354865539,
+      "grad_norm": 0.22287732362747192,
+      "learning_rate": 2.727496845958474e-05,
+      "loss": 1.5062,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6254505128915997,
+      "grad_norm": 0.2150767594575882,
+      "learning_rate": 2.7205034021123505e-05,
+      "loss": 1.5854,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6260049902966454,
+      "grad_norm": 0.21351411938667297,
+      "learning_rate": 2.71351431247714e-05,
+      "loss": 1.6161,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6265594677016911,
+      "grad_norm": 0.2130012810230255,
+      "learning_rate": 2.7065296008371703e-05,
+      "loss": 1.5533,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6271139451067369,
+      "grad_norm": 0.22186146676540375,
+      "learning_rate": 2.699549290961876e-05,
+      "loss": 1.5779,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6276684225117827,
+      "grad_norm": 0.2098151296377182,
+      "learning_rate": 2.6925734066057076e-05,
+      "loss": 1.5365,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6282228999168283,
+      "grad_norm": 0.22018581628799438,
+      "learning_rate": 2.6856019715080576e-05,
+      "loss": 1.5884,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6287773773218741,
+      "grad_norm": 0.2144719660282135,
+      "learning_rate": 2.6786350093931805e-05,
+      "loss": 1.5707,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6293318547269199,
+      "grad_norm": 0.2225160151720047,
+      "learning_rate": 2.6716725439701013e-05,
+      "loss": 1.6019,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6298863321319657,
+      "grad_norm": 0.21820110082626343,
+      "learning_rate": 2.66471459893255e-05,
+      "loss": 1.6112,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6304408095370113,
+      "grad_norm": 0.2195034772157669,
+      "learning_rate": 2.6577611979588685e-05,
+      "loss": 1.5993,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6309952869420571,
+      "grad_norm": 0.214851975440979,
+      "learning_rate": 2.6508123647119376e-05,
+      "loss": 1.544,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6315497643471029,
+      "grad_norm": 0.21557554602622986,
+      "learning_rate": 2.643868122839093e-05,
+      "loss": 1.569,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6321042417521486,
+      "grad_norm": 0.21042174100875854,
+      "learning_rate": 2.6369284959720447e-05,
+      "loss": 1.5386,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6326587191571943,
+      "grad_norm": 0.20624153316020966,
+      "learning_rate": 2.629993507726801e-05,
+      "loss": 1.5237,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6332131965622401,
+      "grad_norm": 0.208335280418396,
+      "learning_rate": 2.6230631817035784e-05,
+      "loss": 1.5465,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6337676739672858,
+      "grad_norm": 0.21543239057064056,
+      "learning_rate": 2.6161375414867337e-05,
+      "loss": 1.5521,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6343221513723316,
+      "grad_norm": 0.21806393563747406,
+      "learning_rate": 2.6092166106446753e-05,
+      "loss": 1.504,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6348766287773773,
+      "grad_norm": 0.20788079500198364,
+      "learning_rate": 2.602300412729784e-05,
+      "loss": 1.5175,
+      "step": 1145
+    },
+    {
+      "epoch": 0.635431106182423,
+      "grad_norm": 0.21668802201747894,
+      "learning_rate": 2.5953889712783364e-05,
+      "loss": 1.5238,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6359855835874688,
+      "grad_norm": 0.219620943069458,
+      "learning_rate": 2.5884823098104202e-05,
+      "loss": 1.5186,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6365400609925146,
+      "grad_norm": 0.22174449265003204,
+      "learning_rate": 2.5815804518298575e-05,
+      "loss": 1.6057,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6370945383975603,
+      "grad_norm": 0.2142484486103058,
+      "learning_rate": 2.5746834208241266e-05,
+      "loss": 1.5701,
+      "step": 1149
+    },
+    {
+      "epoch": 0.637649015802606,
+      "grad_norm": 0.21401485800743103,
+      "learning_rate": 2.5677912402642742e-05,
+      "loss": 1.5556,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6382034932076518,
+      "grad_norm": 0.21097297966480255,
+      "learning_rate": 2.560903933604844e-05,
+      "loss": 1.561,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6387579706126976,
+      "grad_norm": 0.21315309405326843,
+      "learning_rate": 2.554021524283794e-05,
+      "loss": 1.5939,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6393124480177432,
+      "grad_norm": 0.21164047718048096,
+      "learning_rate": 2.5471440357224112e-05,
+      "loss": 1.5832,
+      "step": 1153
+    },
+    {
+      "epoch": 0.639866925422789,
+      "grad_norm": 0.1997397392988205,
+      "learning_rate": 2.5402714913252463e-05,
+      "loss": 1.4702,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6404214028278348,
+      "grad_norm": 0.2216353565454483,
+      "learning_rate": 2.5334039144800128e-05,
+      "loss": 1.6679,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6409758802328805,
+      "grad_norm": 0.20689931511878967,
+      "learning_rate": 2.5265413285575297e-05,
+      "loss": 1.5088,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6415303576379262,
+      "grad_norm": 0.21255846321582794,
+      "learning_rate": 2.5196837569116267e-05,
+      "loss": 1.5848,
+      "step": 1157
+    },
+    {
+      "epoch": 0.642084835042972,
+      "grad_norm": 0.19903656840324402,
+      "learning_rate": 2.5128312228790684e-05,
+      "loss": 1.5309,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6426393124480178,
+      "grad_norm": 0.2166842818260193,
+      "learning_rate": 2.505983749779481e-05,
+      "loss": 1.6536,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6431937898530635,
+      "grad_norm": 0.21221613883972168,
+      "learning_rate": 2.4991413609152606e-05,
+      "loss": 1.5502,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6437482672581092,
+      "grad_norm": 0.2121618390083313,
+      "learning_rate": 2.4923040795715095e-05,
+      "loss": 1.566,
+      "step": 1161
+    },
+    {
+      "epoch": 0.644302744663155,
+      "grad_norm": 0.21857164800167084,
+      "learning_rate": 2.485471929015944e-05,
+      "loss": 1.6212,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6448572220682007,
+      "grad_norm": 0.21056576073169708,
+      "learning_rate": 2.4786449324988203e-05,
+      "loss": 1.5362,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6454116994732465,
+      "grad_norm": 0.21527813374996185,
+      "learning_rate": 2.4718231132528562e-05,
+      "loss": 1.5372,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6459661768782922,
+      "grad_norm": 0.20520104467868805,
+      "learning_rate": 2.4650064944931495e-05,
+      "loss": 1.5219,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6465206542833379,
+      "grad_norm": 0.20994901657104492,
+      "learning_rate": 2.4581950994171013e-05,
+      "loss": 1.5247,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6470751316883837,
+      "grad_norm": 0.21351756155490875,
+      "learning_rate": 2.4513889512043383e-05,
+      "loss": 1.572,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6476296090934295,
+      "grad_norm": 0.21030496060848236,
+      "learning_rate": 2.444588073016627e-05,
+      "loss": 1.5433,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6481840864984751,
+      "grad_norm": 0.2144717425107956,
+      "learning_rate": 2.4377924879978045e-05,
+      "loss": 1.563,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6487385639035209,
+      "grad_norm": 0.20969878137111664,
+      "learning_rate": 2.4310022192736905e-05,
+      "loss": 1.5527,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6492930413085667,
+      "grad_norm": 0.20833313465118408,
+      "learning_rate": 2.4242172899520164e-05,
+      "loss": 1.5475,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6498475187136125,
+      "grad_norm": 0.21155394613742828,
+      "learning_rate": 2.417437723122343e-05,
+      "loss": 1.565,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6504019961186581,
+      "grad_norm": 0.2123716175556183,
+      "learning_rate": 2.4106635418559786e-05,
+      "loss": 1.5333,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6509564735237039,
+      "grad_norm": 0.21838416159152985,
+      "learning_rate": 2.40389476920591e-05,
+      "loss": 1.6055,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6515109509287497,
+      "grad_norm": 0.21394363045692444,
+      "learning_rate": 2.397131428206711e-05,
+      "loss": 1.5401,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6520654283337954,
+      "grad_norm": 0.21762172877788544,
+      "learning_rate": 2.3903735418744783e-05,
+      "loss": 1.6162,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6526199057388411,
+      "grad_norm": 0.21325650811195374,
+      "learning_rate": 2.3836211332067426e-05,
+      "loss": 1.5997,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6531743831438869,
+      "grad_norm": 0.2168748527765274,
+      "learning_rate": 2.376874225182391e-05,
+      "loss": 1.5998,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6537288605489326,
+      "grad_norm": 0.22087283432483673,
+      "learning_rate": 2.3701328407615977e-05,
+      "loss": 1.6222,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6542833379539784,
+      "grad_norm": 0.21224388480186462,
+      "learning_rate": 2.3633970028857327e-05,
+      "loss": 1.4607,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6548378153590241,
+      "grad_norm": 0.21814611554145813,
+      "learning_rate": 2.356666734477298e-05,
+      "loss": 1.575,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6553922927640699,
+      "grad_norm": 0.21967004239559174,
+      "learning_rate": 2.3499420584398377e-05,
+      "loss": 1.5645,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6559467701691156,
+      "grad_norm": 0.2046661376953125,
+      "learning_rate": 2.343222997657865e-05,
+      "loss": 1.5153,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6565012475741614,
+      "grad_norm": 0.21323862671852112,
+      "learning_rate": 2.336509574996784e-05,
+      "loss": 1.5637,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6570557249792071,
+      "grad_norm": 0.21644696593284607,
+      "learning_rate": 2.3298018133028166e-05,
+      "loss": 1.5501,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6576102023842528,
+      "grad_norm": 0.22033213078975677,
+      "learning_rate": 2.3230997354029116e-05,
+      "loss": 1.5611,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6581646797892986,
+      "grad_norm": 0.2171270102262497,
+      "learning_rate": 2.3164033641046824e-05,
+      "loss": 1.578,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6587191571943444,
+      "grad_norm": 0.2252931147813797,
+      "learning_rate": 2.309712722196319e-05,
+      "loss": 1.62,
+      "step": 1188
+    },
+    {
+      "epoch": 0.65927363459939,
+      "grad_norm": 0.20670726895332336,
+      "learning_rate": 2.303027832446516e-05,
+      "loss": 1.4831,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6598281120044358,
+      "grad_norm": 0.2236548513174057,
+      "learning_rate": 2.296348717604392e-05,
+      "loss": 1.6453,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6603825894094816,
+      "grad_norm": 0.22350645065307617,
+      "learning_rate": 2.2896754003994124e-05,
+      "loss": 1.5666,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6609370668145274,
+      "grad_norm": 0.21167486906051636,
+      "learning_rate": 2.2830079035413153e-05,
+      "loss": 1.5366,
+      "step": 1192
+    },
+    {
+      "epoch": 0.661491544219573,
+      "grad_norm": 0.2091861069202423,
+      "learning_rate": 2.2763462497200282e-05,
+      "loss": 1.5729,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6620460216246188,
+      "grad_norm": 0.20955854654312134,
+      "learning_rate": 2.2696904616055953e-05,
+      "loss": 1.5978,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6626004990296646,
+      "grad_norm": 0.20902609825134277,
+      "learning_rate": 2.2630405618481052e-05,
+      "loss": 1.5898,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6631549764347103,
+      "grad_norm": 0.2072347104549408,
+      "learning_rate": 2.2563965730775995e-05,
+      "loss": 1.517,
+      "step": 1196
+    },
+    {
+      "epoch": 0.663709453839756,
+      "grad_norm": 0.20107337832450867,
+      "learning_rate": 2.2497585179040087e-05,
+      "loss": 1.5093,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6642639312448018,
+      "grad_norm": 0.2148445099592209,
+      "learning_rate": 2.243126418917071e-05,
+      "loss": 1.5597,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6648184086498475,
+      "grad_norm": 0.20870907604694366,
+      "learning_rate": 2.236500298686253e-05,
+      "loss": 1.5352,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6653728860548933,
+      "grad_norm": 0.21237953007221222,
+      "learning_rate": 2.229880179760681e-05,
+      "loss": 1.5499,
+      "step": 1200
+    },
+    {
+      "epoch": 0.665927363459939,
+      "grad_norm": 0.21513739228248596,
+      "learning_rate": 2.2232660846690477e-05,
+      "loss": 1.5859,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6664818408649847,
+      "grad_norm": 0.22024357318878174,
+      "learning_rate": 2.2166580359195594e-05,
+      "loss": 1.6493,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6670363182700305,
+      "grad_norm": 0.21224506199359894,
+      "learning_rate": 2.210056055999835e-05,
+      "loss": 1.588,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6675907956750763,
+      "grad_norm": 0.21365848183631897,
+      "learning_rate": 2.2034601673768475e-05,
+      "loss": 1.5165,
+      "step": 1204
+    },
+    {
+      "epoch": 0.668145273080122,
+      "grad_norm": 0.21693235635757446,
+      "learning_rate": 2.1968703924968388e-05,
+      "loss": 1.6435,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6686997504851677,
+      "grad_norm": 0.2145548313856125,
+      "learning_rate": 2.1902867537852453e-05,
+      "loss": 1.5788,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6692542278902135,
+      "grad_norm": 0.21968676149845123,
+      "learning_rate": 2.1837092736466233e-05,
+      "loss": 1.6045,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6698087052952593,
+      "grad_norm": 0.2109319120645523,
+      "learning_rate": 2.1771379744645643e-05,
+      "loss": 1.5182,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6703631827003049,
+      "grad_norm": 0.21400365233421326,
+      "learning_rate": 2.170572878601636e-05,
+      "loss": 1.5532,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6709176601053507,
+      "grad_norm": 0.20740005373954773,
+      "learning_rate": 2.1640140083992908e-05,
+      "loss": 1.5601,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6714721375103965,
+      "grad_norm": 0.21574941277503967,
+      "learning_rate": 2.1574613861777904e-05,
+      "loss": 1.5462,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6720266149154422,
+      "grad_norm": 0.2179066389799118,
+      "learning_rate": 2.1509150342361407e-05,
+      "loss": 1.6072,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6725810923204879,
+      "grad_norm": 0.21342717111110687,
+      "learning_rate": 2.144374974852007e-05,
+      "loss": 1.6028,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6731355697255337,
+      "grad_norm": 0.21398906409740448,
+      "learning_rate": 2.1378412302816408e-05,
+      "loss": 1.595,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6736900471305795,
+      "grad_norm": 0.20966409146785736,
+      "learning_rate": 2.1313138227598053e-05,
+      "loss": 1.5391,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6742445245356252,
+      "grad_norm": 0.20586159825325012,
+      "learning_rate": 2.1247927744996913e-05,
+      "loss": 1.5392,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6747990019406709,
+      "grad_norm": 0.21044990420341492,
+      "learning_rate": 2.11827810769286e-05,
+      "loss": 1.6125,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6753534793457167,
+      "grad_norm": 0.22008857131004333,
+      "learning_rate": 2.111769844509149e-05,
+      "loss": 1.5878,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6759079567507624,
+      "grad_norm": 0.21583212912082672,
+      "learning_rate": 2.105268007096603e-05,
+      "loss": 1.5553,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6764624341558082,
+      "grad_norm": 0.20590542256832123,
+      "learning_rate": 2.0987726175814025e-05,
+      "loss": 1.5257,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6770169115608539,
+      "grad_norm": 0.21506734192371368,
+      "learning_rate": 2.0922836980677844e-05,
+      "loss": 1.6258,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6775713889658996,
+      "grad_norm": 0.21825283765792847,
+      "learning_rate": 2.085801270637968e-05,
+      "loss": 1.5429,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6781258663709454,
+      "grad_norm": 0.2100907862186432,
+      "learning_rate": 2.0793253573520785e-05,
+      "loss": 1.5905,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6786803437759912,
+      "grad_norm": 0.21609370410442352,
+      "learning_rate": 2.0728559802480754e-05,
+      "loss": 1.6102,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6792348211810368,
+      "grad_norm": 0.21235951781272888,
+      "learning_rate": 2.0663931613416746e-05,
+      "loss": 1.4997,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6797892985860826,
+      "grad_norm": 0.20869648456573486,
+      "learning_rate": 2.0599369226262693e-05,
+      "loss": 1.5351,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6803437759911284,
+      "grad_norm": 0.21568715572357178,
+      "learning_rate": 2.053487286072865e-05,
+      "loss": 1.5294,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6808982533961742,
+      "grad_norm": 0.21656538546085358,
+      "learning_rate": 2.0470442736300013e-05,
+      "loss": 1.5462,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6814527308012198,
+      "grad_norm": 0.21754758059978485,
+      "learning_rate": 2.0406079072236684e-05,
+      "loss": 1.5732,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6820072082062656,
+      "grad_norm": 0.21004228293895721,
+      "learning_rate": 2.0341782087572453e-05,
+      "loss": 1.5723,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6825616856113114,
+      "grad_norm": 0.22100019454956055,
+      "learning_rate": 2.0277552001114183e-05,
+      "loss": 1.5375,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6831161630163571,
+      "grad_norm": 0.21162226796150208,
+      "learning_rate": 2.0213389031441072e-05,
+      "loss": 1.5535,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6836706404214028,
+      "grad_norm": 0.21510981023311615,
+      "learning_rate": 2.0149293396903936e-05,
+      "loss": 1.5419,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6842251178264486,
+      "grad_norm": 0.22378797829151154,
+      "learning_rate": 2.0085265315624375e-05,
+      "loss": 1.5729,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6847795952314943,
+      "grad_norm": 0.2192632257938385,
+      "learning_rate": 2.002130500549422e-05,
+      "loss": 1.4287,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6853340726365401,
+      "grad_norm": 0.22077576816082,
+      "learning_rate": 1.995741268417455e-05,
+      "loss": 1.5635,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6858885500415858,
+      "grad_norm": 0.22820354998111725,
+      "learning_rate": 1.9893588569095148e-05,
+      "loss": 1.5514,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6864430274466315,
+      "grad_norm": 0.22799494862556458,
+      "learning_rate": 1.9829832877453673e-05,
+      "loss": 1.6064,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6869975048516773,
+      "grad_norm": 0.21330177783966064,
+      "learning_rate": 1.976614582621492e-05,
+      "loss": 1.5972,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6875519822567231,
+      "grad_norm": 0.22794507443904877,
+      "learning_rate": 1.9702527632110128e-05,
+      "loss": 1.5362,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6881064596617688,
+      "grad_norm": 0.2161223590373993,
+      "learning_rate": 1.9638978511636133e-05,
+      "loss": 1.5349,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6886609370668145,
+      "grad_norm": 0.21102319657802582,
+      "learning_rate": 1.9575498681054816e-05,
+      "loss": 1.5777,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6892154144718603,
+      "grad_norm": 0.2153400331735611,
+      "learning_rate": 1.9512088356392206e-05,
+      "loss": 1.5413,
+      "step": 1243
+    },
+    {
+      "epoch": 0.689769891876906,
+      "grad_norm": 0.23227231204509735,
+      "learning_rate": 1.9448747753437766e-05,
+      "loss": 1.5721,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6903243692819517,
+      "grad_norm": 0.21719689667224884,
+      "learning_rate": 1.938547708774373e-05,
+      "loss": 1.5343,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6908788466869975,
+      "grad_norm": 0.20928537845611572,
+      "learning_rate": 1.9322276574624374e-05,
+      "loss": 1.5005,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6914333240920433,
+      "grad_norm": 0.23064006865024567,
+      "learning_rate": 1.925914642915514e-05,
+      "loss": 1.5943,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6919878014970889,
+      "grad_norm": 0.22050850093364716,
+      "learning_rate": 1.919608686617208e-05,
+      "loss": 1.5496,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6925422789021347,
+      "grad_norm": 0.22092626988887787,
+      "learning_rate": 1.9133098100271018e-05,
+      "loss": 1.6028,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6930967563071805,
+      "grad_norm": 0.23645174503326416,
+      "learning_rate": 1.9070180345806867e-05,
+      "loss": 1.5723,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6936512337122263,
+      "grad_norm": 0.22588878870010376,
+      "learning_rate": 1.9007333816892886e-05,
+      "loss": 1.59,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6942057111172719,
+      "grad_norm": 0.2189507782459259,
+      "learning_rate": 1.8944558727399894e-05,
+      "loss": 1.5932,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6947601885223177,
+      "grad_norm": 0.21700263023376465,
+      "learning_rate": 1.8881855290955702e-05,
+      "loss": 1.5475,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6953146659273635,
+      "grad_norm": 0.21820667386054993,
+      "learning_rate": 1.8819223720944176e-05,
+      "loss": 1.534,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6958691433324092,
+      "grad_norm": 0.22174763679504395,
+      "learning_rate": 1.875666423050467e-05,
+      "loss": 1.6515,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6964236207374549,
+      "grad_norm": 0.22205999493598938,
+      "learning_rate": 1.8694177032531247e-05,
+      "loss": 1.6275,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6969780981425007,
+      "grad_norm": 0.21358875930309296,
+      "learning_rate": 1.863176233967193e-05,
+      "loss": 1.6198,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6975325755475464,
+      "grad_norm": 0.21394740045070648,
+      "learning_rate": 1.856942036432804e-05,
+      "loss": 1.5314,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6980870529525922,
+      "grad_norm": 0.21883390843868256,
+      "learning_rate": 1.850715131865336e-05,
+      "loss": 1.6133,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6986415303576379,
+      "grad_norm": 0.21168088912963867,
+      "learning_rate": 1.8444955414553595e-05,
+      "loss": 1.6031,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6991960077626836,
+      "grad_norm": 0.2189856618642807,
+      "learning_rate": 1.8382832863685477e-05,
+      "loss": 1.5858,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6997504851677294,
+      "grad_norm": 0.20865415036678314,
+      "learning_rate": 1.8320783877456107e-05,
+      "loss": 1.5458,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7003049625727752,
+      "grad_norm": 0.21896782517433167,
+      "learning_rate": 1.825880866702226e-05,
+      "loss": 1.5763,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7008594399778209,
+      "grad_norm": 0.2206234335899353,
+      "learning_rate": 1.8196907443289656e-05,
+      "loss": 1.5721,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7014139173828666,
+      "grad_norm": 0.217427596449852,
+      "learning_rate": 1.813508041691222e-05,
+      "loss": 1.6216,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7019683947879124,
+      "grad_norm": 0.2126351296901703,
+      "learning_rate": 1.8073327798291387e-05,
+      "loss": 1.5936,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7025228721929582,
+      "grad_norm": 0.2115585058927536,
+      "learning_rate": 1.8011649797575343e-05,
+      "loss": 1.5893,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7030773495980038,
+      "grad_norm": 0.21458736062049866,
+      "learning_rate": 1.7950046624658418e-05,
+      "loss": 1.4703,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7036318270030496,
+      "grad_norm": 0.21214710175991058,
+      "learning_rate": 1.788851848918022e-05,
+      "loss": 1.6151,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7041863044080954,
+      "grad_norm": 0.21619988977909088,
+      "learning_rate": 1.7827065600525043e-05,
+      "loss": 1.4581,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7047407818131411,
+      "grad_norm": 0.21274301409721375,
+      "learning_rate": 1.77656881678211e-05,
+      "loss": 1.5052,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7052952592181868,
+      "grad_norm": 0.2131747454404831,
+      "learning_rate": 1.7704386399939818e-05,
+      "loss": 1.5616,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7058497366232326,
+      "grad_norm": 0.22459973394870758,
+      "learning_rate": 1.7643160505495146e-05,
+      "loss": 1.6162,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7064042140282784,
+      "grad_norm": 0.21642784774303436,
+      "learning_rate": 1.7582010692842823e-05,
+      "loss": 1.5565,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7069586914333241,
+      "grad_norm": 0.22068682312965393,
+      "learning_rate": 1.7520937170079667e-05,
+      "loss": 1.5865,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7075131688383698,
+      "grad_norm": 0.2210376113653183,
+      "learning_rate": 1.7459940145042904e-05,
+      "loss": 1.5605,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7080676462434156,
+      "grad_norm": 0.21524764597415924,
+      "learning_rate": 1.7399019825309387e-05,
+      "loss": 1.568,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7086221236484613,
+      "grad_norm": 0.20720575749874115,
+      "learning_rate": 1.733817641819496e-05,
+      "loss": 1.5723,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7091766010535071,
+      "grad_norm": 0.22238507866859436,
+      "learning_rate": 1.7277410130753775e-05,
+      "loss": 1.5773,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7097310784585528,
+      "grad_norm": 0.20956286787986755,
+      "learning_rate": 1.7216721169777452e-05,
+      "loss": 1.5252,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7102855558635985,
+      "grad_norm": 0.21438319981098175,
+      "learning_rate": 1.7156109741794533e-05,
+      "loss": 1.4835,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7108400332686443,
+      "grad_norm": 0.21574389934539795,
+      "learning_rate": 1.709557605306967e-05,
+      "loss": 1.588,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7113945106736901,
+      "grad_norm": 0.2138158082962036,
+      "learning_rate": 1.7035120309602994e-05,
+      "loss": 1.5567,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7119489880787357,
+      "grad_norm": 0.21723328530788422,
+      "learning_rate": 1.6974742717129373e-05,
+      "loss": 1.5732,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7125034654837815,
+      "grad_norm": 0.2203451544046402,
+      "learning_rate": 1.6914443481117678e-05,
+      "loss": 1.6133,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7130579428888273,
+      "grad_norm": 0.20896215736865997,
+      "learning_rate": 1.6854222806770228e-05,
+      "loss": 1.5527,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7136124202938731,
+      "grad_norm": 0.20967014133930206,
+      "learning_rate": 1.679408089902188e-05,
+      "loss": 1.616,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7141668976989187,
+      "grad_norm": 0.21301911771297455,
+      "learning_rate": 1.673401796253952e-05,
+      "loss": 1.5695,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7147213751039645,
+      "grad_norm": 0.22176018357276917,
+      "learning_rate": 1.667403420172125e-05,
+      "loss": 1.5707,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7152758525090103,
+      "grad_norm": 0.20806194841861725,
+      "learning_rate": 1.6614129820695755e-05,
+      "loss": 1.547,
+      "step": 1290
+    },
+    {
+      "epoch": 0.715830329914056,
+      "grad_norm": 0.21885931491851807,
+      "learning_rate": 1.6554305023321587e-05,
+      "loss": 1.5618,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7163848073191017,
+      "grad_norm": 0.21963781118392944,
+      "learning_rate": 1.6494560013186413e-05,
+      "loss": 1.5697,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7169392847241475,
+      "grad_norm": 0.20642498135566711,
+      "learning_rate": 1.6434894993606474e-05,
+      "loss": 1.5343,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7174937621291932,
+      "grad_norm": 0.21944020688533783,
+      "learning_rate": 1.6375310167625736e-05,
+      "loss": 1.589,
+      "step": 1294
+    },
+    {
+      "epoch": 0.718048239534239,
+      "grad_norm": 0.20910033583641052,
+      "learning_rate": 1.631580573801526e-05,
+      "loss": 1.5539,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7186027169392847,
+      "grad_norm": 0.21879972517490387,
+      "learning_rate": 1.625638190727253e-05,
+      "loss": 1.557,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7191571943443305,
+      "grad_norm": 0.21549879014492035,
+      "learning_rate": 1.6197038877620745e-05,
+      "loss": 1.5955,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7197116717493762,
+      "grad_norm": 0.21491822600364685,
+      "learning_rate": 1.6137776851008135e-05,
+      "loss": 1.5934,
+      "step": 1298
+    },
+    {
+      "epoch": 0.720266149154422,
+      "grad_norm": 0.21684475243091583,
+      "learning_rate": 1.607859602910726e-05,
+      "loss": 1.5928,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7208206265594677,
+      "grad_norm": 0.2135036587715149,
+      "learning_rate": 1.601949661331434e-05,
+      "loss": 1.5901,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7213751039645134,
+      "grad_norm": 0.21174384653568268,
+      "learning_rate": 1.596047880474859e-05,
+      "loss": 1.5478,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7219295813695592,
+      "grad_norm": 0.21973606944084167,
+      "learning_rate": 1.5901542804251446e-05,
+      "loss": 1.5873,
+      "step": 1302
+    },
+    {
+      "epoch": 0.722484058774605,
+      "grad_norm": 0.21199335157871246,
+      "learning_rate": 1.5842688812385997e-05,
+      "loss": 1.5681,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7230385361796506,
+      "grad_norm": 0.20926666259765625,
+      "learning_rate": 1.578391702943628e-05,
+      "loss": 1.5635,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7235930135846964,
+      "grad_norm": 0.20591633021831512,
+      "learning_rate": 1.5725227655406485e-05,
+      "loss": 1.5811,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7241474909897422,
+      "grad_norm": 0.21061578392982483,
+      "learning_rate": 1.5666620890020423e-05,
+      "loss": 1.5401,
+      "step": 1306
+    },
+    {
+      "epoch": 0.724701968394788,
+      "grad_norm": 0.21330875158309937,
+      "learning_rate": 1.5608096932720758e-05,
+      "loss": 1.5631,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7252564457998336,
+      "grad_norm": 0.21160577237606049,
+      "learning_rate": 1.5549655982668365e-05,
+      "loss": 1.5541,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7258109232048794,
+      "grad_norm": 0.21944500505924225,
+      "learning_rate": 1.549129823874164e-05,
+      "loss": 1.5407,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7263654006099252,
+      "grad_norm": 0.21115024387836456,
+      "learning_rate": 1.543302389953578e-05,
+      "loss": 1.5101,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7269198780149709,
+      "grad_norm": 0.21656592190265656,
+      "learning_rate": 1.5374833163362228e-05,
+      "loss": 1.5505,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7274743554200166,
+      "grad_norm": 0.20774325728416443,
+      "learning_rate": 1.5316726228247874e-05,
+      "loss": 1.5188,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7280288328250624,
+      "grad_norm": 0.22355897724628448,
+      "learning_rate": 1.525870329193441e-05,
+      "loss": 1.5638,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7285833102301081,
+      "grad_norm": 0.20786559581756592,
+      "learning_rate": 1.5200764551877707e-05,
+      "loss": 1.5139,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7291377876351539,
+      "grad_norm": 0.21378779411315918,
+      "learning_rate": 1.51429102052471e-05,
+      "loss": 1.5272,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7296922650401996,
+      "grad_norm": 0.2113804817199707,
+      "learning_rate": 1.5085140448924738e-05,
+      "loss": 1.5503,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7302467424452453,
+      "grad_norm": 0.2093944549560547,
+      "learning_rate": 1.5027455479504896e-05,
+      "loss": 1.5027,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7308012198502911,
+      "grad_norm": 0.2111719399690628,
+      "learning_rate": 1.4969855493293275e-05,
+      "loss": 1.4586,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7313556972553369,
+      "grad_norm": 0.21029268205165863,
+      "learning_rate": 1.491234068630646e-05,
+      "loss": 1.6253,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7319101746603826,
+      "grad_norm": 0.2066301554441452,
+      "learning_rate": 1.4854911254271086e-05,
+      "loss": 1.4414,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7324646520654283,
+      "grad_norm": 0.2162986844778061,
+      "learning_rate": 1.4797567392623275e-05,
+      "loss": 1.5728,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7330191294704741,
+      "grad_norm": 0.21040524542331696,
+      "learning_rate": 1.4740309296507969e-05,
+      "loss": 1.5839,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7335736068755199,
+      "grad_norm": 0.22399760782718658,
+      "learning_rate": 1.4683137160778218e-05,
+      "loss": 1.6573,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7341280842805655,
+      "grad_norm": 0.20881177484989166,
+      "learning_rate": 1.4626051179994569e-05,
+      "loss": 1.5418,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7346825616856113,
+      "grad_norm": 0.21787725389003754,
+      "learning_rate": 1.4569051548424323e-05,
+      "loss": 1.6005,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7352370390906571,
+      "grad_norm": 0.2111303210258484,
+      "learning_rate": 1.4512138460041004e-05,
+      "loss": 1.5171,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7357915164957028,
+      "grad_norm": 0.20768657326698303,
+      "learning_rate": 1.4455312108523587e-05,
+      "loss": 1.5723,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7363459939007485,
+      "grad_norm": 0.20773515105247498,
+      "learning_rate": 1.4398572687255858e-05,
+      "loss": 1.5349,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7369004713057943,
+      "grad_norm": 0.20800316333770752,
+      "learning_rate": 1.43419203893258e-05,
+      "loss": 1.5316,
+      "step": 1329
+    },
+    {
+      "epoch": 0.73745494871084,
+      "grad_norm": 0.21494507789611816,
+      "learning_rate": 1.42853554075249e-05,
+      "loss": 1.5631,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7380094261158858,
+      "grad_norm": 0.2120676040649414,
+      "learning_rate": 1.4228877934347506e-05,
+      "loss": 1.5709,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7385639035209315,
+      "grad_norm": 0.20988710224628448,
+      "learning_rate": 1.4172488161990168e-05,
+      "loss": 1.5334,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7391183809259773,
+      "grad_norm": 0.21031689643859863,
+      "learning_rate": 1.4116186282350981e-05,
+      "loss": 1.5124,
+      "step": 1333
+    },
+    {
+      "epoch": 0.739672858331023,
+      "grad_norm": 0.21639370918273926,
+      "learning_rate": 1.4059972487028936e-05,
+      "loss": 1.567,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7402273357360688,
+      "grad_norm": 0.21508349478244781,
+      "learning_rate": 1.4003846967323283e-05,
+      "loss": 1.5132,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7407818131411145,
+      "grad_norm": 0.21374095976352692,
+      "learning_rate": 1.3947809914232808e-05,
+      "loss": 1.555,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7413362905461602,
+      "grad_norm": 0.22325333952903748,
+      "learning_rate": 1.3891861518455342e-05,
+      "loss": 1.5703,
+      "step": 1337
+    },
+    {
+      "epoch": 0.741890767951206,
+      "grad_norm": 0.22245821356773376,
+      "learning_rate": 1.3836001970386898e-05,
+      "loss": 1.5424,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7424452453562518,
+      "grad_norm": 0.2141106277704239,
+      "learning_rate": 1.3780231460121204e-05,
+      "loss": 1.601,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7429997227612974,
+      "grad_norm": 0.21222813427448273,
+      "learning_rate": 1.3724550177448968e-05,
+      "loss": 1.5908,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7435542001663432,
+      "grad_norm": 0.21440914273262024,
+      "learning_rate": 1.3668958311857247e-05,
+      "loss": 1.6119,
+      "step": 1341
+    },
+    {
+      "epoch": 0.744108677571389,
+      "grad_norm": 0.21355807781219482,
+      "learning_rate": 1.3613456052528822e-05,
+      "loss": 1.5494,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7446631549764348,
+      "grad_norm": 0.22582007944583893,
+      "learning_rate": 1.3558043588341488e-05,
+      "loss": 1.6013,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7452176323814804,
+      "grad_norm": 0.22058305144309998,
+      "learning_rate": 1.3502721107867536e-05,
+      "loss": 1.6472,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7457721097865262,
+      "grad_norm": 0.21767617762088776,
+      "learning_rate": 1.3447488799372997e-05,
+      "loss": 1.5594,
+      "step": 1345
+    },
+    {
+      "epoch": 0.746326587191572,
+      "grad_norm": 0.21455064415931702,
+      "learning_rate": 1.339234685081702e-05,
+      "loss": 1.4785,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7468810645966177,
+      "grad_norm": 0.2150326520204544,
+      "learning_rate": 1.3337295449851282e-05,
+      "loss": 1.5622,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7474355420016634,
+      "grad_norm": 0.21228265762329102,
+      "learning_rate": 1.328233478381932e-05,
+      "loss": 1.5579,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7479900194067092,
+      "grad_norm": 0.21863175928592682,
+      "learning_rate": 1.3227465039755889e-05,
+      "loss": 1.5563,
+      "step": 1349
+    },
+    {
+      "epoch": 0.748544496811755,
+      "grad_norm": 0.22516675293445587,
+      "learning_rate": 1.3172686404386323e-05,
+      "loss": 1.5843,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7490989742168007,
+      "grad_norm": 0.21071338653564453,
+      "learning_rate": 1.3117999064125923e-05,
+      "loss": 1.4893,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7496534516218464,
+      "grad_norm": 0.21058571338653564,
+      "learning_rate": 1.3063403205079302e-05,
+      "loss": 1.4929,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7502079290268922,
+      "grad_norm": 0.2110683172941208,
+      "learning_rate": 1.300889901303973e-05,
+      "loss": 1.5716,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7507624064319379,
+      "grad_norm": 0.22310392558574677,
+      "learning_rate": 1.2954486673488554e-05,
+      "loss": 1.5591,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7513168838369837,
+      "grad_norm": 0.21150308847427368,
+      "learning_rate": 1.290016637159457e-05,
+      "loss": 1.5552,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7518713612420294,
+      "grad_norm": 0.2172410786151886,
+      "learning_rate": 1.2845938292213296e-05,
+      "loss": 1.6018,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7524258386470751,
+      "grad_norm": 0.21438592672348022,
+      "learning_rate": 1.2791802619886457e-05,
+      "loss": 1.5254,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7529803160521209,
+      "grad_norm": 0.2097012847661972,
+      "learning_rate": 1.2737759538841297e-05,
+      "loss": 1.4957,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7535347934571667,
+      "grad_norm": 0.22424669563770294,
+      "learning_rate": 1.2683809232989975e-05,
+      "loss": 1.5994,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7540892708622123,
+      "grad_norm": 0.23129421472549438,
+      "learning_rate": 1.2629951885928931e-05,
+      "loss": 1.6463,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7546437482672581,
+      "grad_norm": 0.2219121903181076,
+      "learning_rate": 1.2576187680938214e-05,
+      "loss": 1.641,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7551982256723039,
+      "grad_norm": 0.21605202555656433,
+      "learning_rate": 1.2522516800980995e-05,
+      "loss": 1.6419,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7557527030773497,
+      "grad_norm": 0.2168269157409668,
+      "learning_rate": 1.2468939428702762e-05,
+      "loss": 1.5664,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7563071804823953,
+      "grad_norm": 0.21760636568069458,
+      "learning_rate": 1.2415455746430846e-05,
+      "loss": 1.5758,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7568616578874411,
+      "grad_norm": 0.2120659053325653,
+      "learning_rate": 1.2362065936173728e-05,
+      "loss": 1.584,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7574161352924869,
+      "grad_norm": 0.20472942292690277,
+      "learning_rate": 1.230877017962043e-05,
+      "loss": 1.5304,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7579706126975326,
+      "grad_norm": 0.21528665721416473,
+      "learning_rate": 1.2255568658139918e-05,
+      "loss": 1.6002,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7585250901025783,
+      "grad_norm": 0.21428853273391724,
+      "learning_rate": 1.2202461552780473e-05,
+      "loss": 1.6109,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7590795675076241,
+      "grad_norm": 0.21304574608802795,
+      "learning_rate": 1.214944904426902e-05,
+      "loss": 1.5883,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7596340449126698,
+      "grad_norm": 0.22913344204425812,
+      "learning_rate": 1.209653131301066e-05,
+      "loss": 1.6736,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7601885223177155,
+      "grad_norm": 0.2191346287727356,
+      "learning_rate": 1.2043708539087865e-05,
+      "loss": 1.5518,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7607429997227613,
+      "grad_norm": 0.21301528811454773,
+      "learning_rate": 1.1990980902260008e-05,
+      "loss": 1.5366,
+      "step": 1372
+    },
+    {
+      "epoch": 0.761297477127807,
+      "grad_norm": 0.21601028740406036,
+      "learning_rate": 1.1938348581962713e-05,
+      "loss": 1.5593,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7618519545328528,
+      "grad_norm": 0.21673454344272614,
+      "learning_rate": 1.1885811757307209e-05,
+      "loss": 1.589,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7624064319378985,
+      "grad_norm": 0.20797033607959747,
+      "learning_rate": 1.1833370607079778e-05,
+      "loss": 1.5379,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7629609093429442,
+      "grad_norm": 0.22476620972156525,
+      "learning_rate": 1.1781025309741056e-05,
+      "loss": 1.5536,
+      "step": 1376
+    },
+    {
+      "epoch": 0.76351538674799,
+      "grad_norm": 0.2215532660484314,
+      "learning_rate": 1.1728776043425563e-05,
+      "loss": 1.5604,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7640698641530358,
+      "grad_norm": 0.214621901512146,
+      "learning_rate": 1.1676622985940983e-05,
+      "loss": 1.5223,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7646243415580815,
+      "grad_norm": 0.2120085209608078,
+      "learning_rate": 1.1624566314767573e-05,
+      "loss": 1.5189,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7651788189631272,
+      "grad_norm": 0.21431514620780945,
+      "learning_rate": 1.1572606207057607e-05,
+      "loss": 1.5565,
+      "step": 1380
+    },
+    {
+      "epoch": 0.765733296368173,
+      "grad_norm": 0.21584293246269226,
+      "learning_rate": 1.152074283963475e-05,
+      "loss": 1.5717,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7662877737732188,
+      "grad_norm": 0.22375768423080444,
+      "learning_rate": 1.1468976388993438e-05,
+      "loss": 1.6198,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7668422511782644,
+      "grad_norm": 0.22090241312980652,
+      "learning_rate": 1.1417307031298304e-05,
+      "loss": 1.5913,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7673967285833102,
+      "grad_norm": 0.20872661471366882,
+      "learning_rate": 1.1365734942383565e-05,
+      "loss": 1.5244,
+      "step": 1384
+    },
+    {
+      "epoch": 0.767951205988356,
+      "grad_norm": 0.218302920460701,
+      "learning_rate": 1.131426029775244e-05,
+      "loss": 1.5787,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7685056833934018,
+      "grad_norm": 0.21635451912879944,
+      "learning_rate": 1.1262883272576492e-05,
+      "loss": 1.5958,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7690601607984474,
+      "grad_norm": 0.21018511056900024,
+      "learning_rate": 1.1211604041695114e-05,
+      "loss": 1.5487,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7696146382034932,
+      "grad_norm": 0.23798473179340363,
+      "learning_rate": 1.1160422779614928e-05,
+      "loss": 1.6402,
+      "step": 1388
+    },
+    {
+      "epoch": 0.770169115608539,
+      "grad_norm": 0.21271711587905884,
+      "learning_rate": 1.1109339660509098e-05,
+      "loss": 1.5796,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7707235930135847,
+      "grad_norm": 0.20341956615447998,
+      "learning_rate": 1.1058354858216842e-05,
+      "loss": 1.5225,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7712780704186304,
+      "grad_norm": 0.23549644649028778,
+      "learning_rate": 1.1007468546242786e-05,
+      "loss": 1.5797,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7718325478236762,
+      "grad_norm": 0.22186161577701569,
+      "learning_rate": 1.0956680897756394e-05,
+      "loss": 1.5828,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7723870252287219,
+      "grad_norm": 0.22768884897232056,
+      "learning_rate": 1.0905992085591373e-05,
+      "loss": 1.5781,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7729415026337677,
+      "grad_norm": 0.20948225259780884,
+      "learning_rate": 1.0855402282245047e-05,
+      "loss": 1.5476,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7734959800388134,
+      "grad_norm": 0.21078969538211823,
+      "learning_rate": 1.0804911659877874e-05,
+      "loss": 1.4978,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7740504574438591,
+      "grad_norm": 0.22138863801956177,
+      "learning_rate": 1.0754520390312755e-05,
+      "loss": 1.5183,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7746049348489049,
+      "grad_norm": 0.2178879678249359,
+      "learning_rate": 1.0704228645034464e-05,
+      "loss": 1.5743,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7751594122539507,
+      "grad_norm": 0.20594577491283417,
+      "learning_rate": 1.0654036595189124e-05,
+      "loss": 1.4646,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7757138896589963,
+      "grad_norm": 0.21161887049674988,
+      "learning_rate": 1.0603944411583576e-05,
+      "loss": 1.5743,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7762683670640421,
+      "grad_norm": 0.21460282802581787,
+      "learning_rate": 1.0553952264684804e-05,
+      "loss": 1.628,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7768228444690879,
+      "grad_norm": 0.2193065732717514,
+      "learning_rate": 1.0504060324619374e-05,
+      "loss": 1.5382,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7773773218741337,
+      "grad_norm": 0.21738368272781372,
+      "learning_rate": 1.0454268761172824e-05,
+      "loss": 1.5957,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7779317992791793,
+      "grad_norm": 0.21127985417842865,
+      "learning_rate": 1.040457774378913e-05,
+      "loss": 1.5551,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7784862766842251,
+      "grad_norm": 0.20940497517585754,
+      "learning_rate": 1.0354987441570064e-05,
+      "loss": 1.5498,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7790407540892709,
+      "grad_norm": 0.21508832275867462,
+      "learning_rate": 1.0305498023274674e-05,
+      "loss": 1.6172,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7795952314943166,
+      "grad_norm": 0.21197104454040527,
+      "learning_rate": 1.025610965731874e-05,
+      "loss": 1.5321,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7801497088993623,
+      "grad_norm": 0.21619556844234467,
+      "learning_rate": 1.0206822511774073e-05,
+      "loss": 1.5089,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7807041863044081,
+      "grad_norm": 0.2153014987707138,
+      "learning_rate": 1.015763675436808e-05,
+      "loss": 1.5755,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7812586637094538,
+      "grad_norm": 0.20888642966747284,
+      "learning_rate": 1.0108552552483122e-05,
+      "loss": 1.5995,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7818131411144996,
+      "grad_norm": 0.2115897834300995,
+      "learning_rate": 1.0059570073155953e-05,
+      "loss": 1.5222,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7823676185195453,
+      "grad_norm": 0.20816390216350555,
+      "learning_rate": 1.0010689483077187e-05,
+      "loss": 1.517,
+      "step": 1411
+    },
+    {
+      "epoch": 0.782922095924591,
+      "grad_norm": 0.21498888731002808,
+      "learning_rate": 9.961910948590643e-06,
+      "loss": 1.5944,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7834765733296368,
+      "grad_norm": 0.2085815966129303,
+      "learning_rate": 9.91323463569292e-06,
+      "loss": 1.4606,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7840310507346826,
+      "grad_norm": 0.21854795515537262,
+      "learning_rate": 9.864660710032669e-06,
+      "loss": 1.6048,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7845855281397283,
+      "grad_norm": 0.21158936619758606,
+      "learning_rate": 9.816189336910166e-06,
+      "loss": 1.5845,
+      "step": 1415
+    },
+    {
+      "epoch": 0.785140005544774,
+      "grad_norm": 0.21523204445838928,
+      "learning_rate": 9.767820681276671e-06,
+      "loss": 1.5289,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7856944829498198,
+      "grad_norm": 0.2180699110031128,
+      "learning_rate": 9.719554907733895e-06,
+      "loss": 1.5658,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7862489603548656,
+      "grad_norm": 0.21391287446022034,
+      "learning_rate": 9.671392180533443e-06,
+      "loss": 1.579,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7868034377599112,
+      "grad_norm": 0.20858129858970642,
+      "learning_rate": 9.62333266357622e-06,
+      "loss": 1.5127,
+      "step": 1419
+    },
+    {
+      "epoch": 0.787357915164957,
+      "grad_norm": 0.2106478065252304,
+      "learning_rate": 9.575376520411907e-06,
+      "loss": 1.5466,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7879123925700028,
+      "grad_norm": 0.20766615867614746,
+      "learning_rate": 9.527523914238452e-06,
+      "loss": 1.5704,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7884668699750486,
+      "grad_norm": 0.21175894141197205,
+      "learning_rate": 9.479775007901378e-06,
+      "loss": 1.5494,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7890213473800942,
+      "grad_norm": 0.2133675217628479,
+      "learning_rate": 9.43212996389335e-06,
+      "loss": 1.5659,
+      "step": 1423
+    },
+    {
+      "epoch": 0.78957582478514,
+      "grad_norm": 0.2181132733821869,
+      "learning_rate": 9.384588944353594e-06,
+      "loss": 1.6126,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7901303021901858,
+      "grad_norm": 0.22013451159000397,
+      "learning_rate": 9.337152111067312e-06,
+      "loss": 1.5657,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7906847795952315,
+      "grad_norm": 0.22332167625427246,
+      "learning_rate": 9.28981962546517e-06,
+      "loss": 1.5943,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7912392570002772,
+      "grad_norm": 0.20821382105350494,
+      "learning_rate": 9.242591648622698e-06,
+      "loss": 1.5187,
+      "step": 1427
+    },
+    {
+      "epoch": 0.791793734405323,
+      "grad_norm": 0.2182924449443817,
+      "learning_rate": 9.19546834125983e-06,
+      "loss": 1.585,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7923482118103687,
+      "grad_norm": 0.20459789037704468,
+      "learning_rate": 9.148449863740279e-06,
+      "loss": 1.52,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7929026892154145,
+      "grad_norm": 0.218237042427063,
+      "learning_rate": 9.101536376070981e-06,
+      "loss": 1.5027,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7934571666204602,
+      "grad_norm": 0.2164810746908188,
+      "learning_rate": 9.05472803790163e-06,
+      "loss": 1.5413,
+      "step": 1431
+    },
+    {
+      "epoch": 0.794011644025506,
+      "grad_norm": 0.21726112067699432,
+      "learning_rate": 9.008025008524068e-06,
+      "loss": 1.5963,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7945661214305517,
+      "grad_norm": 0.22692659497261047,
+      "learning_rate": 8.96142744687178e-06,
+      "loss": 1.5632,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7951205988355975,
+      "grad_norm": 0.21242159605026245,
+      "learning_rate": 8.914935511519313e-06,
+      "loss": 1.5366,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7956750762406432,
+      "grad_norm": 0.214253231883049,
+      "learning_rate": 8.868549360681791e-06,
+      "loss": 1.5982,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7962295536456889,
+      "grad_norm": 0.21711012721061707,
+      "learning_rate": 8.822269152214327e-06,
+      "loss": 1.5109,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7967840310507347,
+      "grad_norm": 0.22008074820041656,
+      "learning_rate": 8.776095043611494e-06,
+      "loss": 1.5869,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7973385084557805,
+      "grad_norm": 0.21308183670043945,
+      "learning_rate": 8.730027192006809e-06,
+      "loss": 1.5816,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7978929858608261,
+      "grad_norm": 0.21756571531295776,
+      "learning_rate": 8.684065754172235e-06,
+      "loss": 1.5787,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7984474632658719,
+      "grad_norm": 0.21786224842071533,
+      "learning_rate": 8.638210886517524e-06,
+      "loss": 1.5704,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7990019406709177,
+      "grad_norm": 0.21699659526348114,
+      "learning_rate": 8.592462745089815e-06,
+      "loss": 1.5663,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7995564180759634,
+      "grad_norm": 0.21736855804920197,
+      "learning_rate": 8.546821485573033e-06,
+      "loss": 1.5873,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8001108954810091,
+      "grad_norm": 0.20840026438236237,
+      "learning_rate": 8.501287263287383e-06,
+      "loss": 1.543,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8006653728860549,
+      "grad_norm": 0.21741265058517456,
+      "learning_rate": 8.45586023318882e-06,
+      "loss": 1.5909,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8012198502911007,
+      "grad_norm": 0.20903275907039642,
+      "learning_rate": 8.410540549868478e-06,
+      "loss": 1.5251,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8017743276961464,
+      "grad_norm": 0.21058255434036255,
+      "learning_rate": 8.365328367552257e-06,
+      "loss": 1.5511,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8023288051011921,
+      "grad_norm": 0.21151955425739288,
+      "learning_rate": 8.320223840100152e-06,
+      "loss": 1.6012,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8028832825062379,
+      "grad_norm": 0.2185509204864502,
+      "learning_rate": 8.27522712100584e-06,
+      "loss": 1.5945,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8034377599112836,
+      "grad_norm": 0.22160445153713226,
+      "learning_rate": 8.230338363396107e-06,
+      "loss": 1.5107,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8039922373163294,
+      "grad_norm": 0.2089175581932068,
+      "learning_rate": 8.185557720030347e-06,
+      "loss": 1.5586,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8045467147213751,
+      "grad_norm": 0.20976027846336365,
+      "learning_rate": 8.140885343300034e-06,
+      "loss": 1.5384,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8051011921264208,
+      "grad_norm": 0.21759721636772156,
+      "learning_rate": 8.09632138522817e-06,
+      "loss": 1.5683,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8056556695314666,
+      "grad_norm": 0.20990976691246033,
+      "learning_rate": 8.051865997468856e-06,
+      "loss": 1.5553,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8062101469365124,
+      "grad_norm": 0.21652382612228394,
+      "learning_rate": 8.007519331306701e-06,
+      "loss": 1.5073,
+      "step": 1454
+    },
+    {
+      "epoch": 0.806764624341558,
+      "grad_norm": 0.21122099459171295,
+      "learning_rate": 7.9632815376563e-06,
+      "loss": 1.5442,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8073191017466038,
+      "grad_norm": 0.21625083684921265,
+      "learning_rate": 7.919152767061765e-06,
+      "loss": 1.5177,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8078735791516496,
+      "grad_norm": 0.21200041472911835,
+      "learning_rate": 7.875133169696231e-06,
+      "loss": 1.62,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8084280565566954,
+      "grad_norm": 0.21699056029319763,
+      "learning_rate": 7.83122289536125e-06,
+      "loss": 1.6298,
+      "step": 1458
+    },
+    {
+      "epoch": 0.808982533961741,
+      "grad_norm": 0.21260713040828705,
+      "learning_rate": 7.78742209348637e-06,
+      "loss": 1.5463,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8095370113667868,
+      "grad_norm": 0.21496307849884033,
+      "learning_rate": 7.743730913128606e-06,
+      "loss": 1.5319,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8100914887718326,
+      "grad_norm": 0.2123148888349533,
+      "learning_rate": 7.700149502971901e-06,
+      "loss": 1.5688,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8106459661768783,
+      "grad_norm": 0.2114924192428589,
+      "learning_rate": 7.656678011326674e-06,
+      "loss": 1.4625,
+      "step": 1462
+    },
+    {
+      "epoch": 0.811200443581924,
+      "grad_norm": 0.21606843173503876,
+      "learning_rate": 7.613316586129231e-06,
+      "loss": 1.6005,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8117549209869698,
+      "grad_norm": 0.21045103669166565,
+      "learning_rate": 7.570065374941386e-06,
+      "loss": 1.5143,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8123093983920155,
+      "grad_norm": 0.21124376356601715,
+      "learning_rate": 7.526924524949826e-06,
+      "loss": 1.5344,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8128638757970613,
+      "grad_norm": 0.2107442021369934,
+      "learning_rate": 7.483894182965699e-06,
+      "loss": 1.5411,
+      "step": 1466
+    },
+    {
+      "epoch": 0.813418353202107,
+      "grad_norm": 0.21770091354846954,
+      "learning_rate": 7.440974495424087e-06,
+      "loss": 1.6174,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8139728306071528,
+      "grad_norm": 0.21479737758636475,
+      "learning_rate": 7.398165608383499e-06,
+      "loss": 1.5734,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8145273080121985,
+      "grad_norm": 0.22530944645404816,
+      "learning_rate": 7.355467667525404e-06,
+      "loss": 1.5711,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8150817854172443,
+      "grad_norm": 0.2182978242635727,
+      "learning_rate": 7.312880818153676e-06,
+      "loss": 1.5933,
+      "step": 1470
+    },
+    {
+      "epoch": 0.81563626282229,
+      "grad_norm": 0.20854482054710388,
+      "learning_rate": 7.270405205194158e-06,
+      "loss": 1.5411,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8161907402273357,
+      "grad_norm": 0.21064384281635284,
+      "learning_rate": 7.228040973194175e-06,
+      "loss": 1.5622,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8167452176323815,
+      "grad_norm": 0.2184503674507141,
+      "learning_rate": 7.185788266321969e-06,
+      "loss": 1.5769,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8172996950374273,
+      "grad_norm": 0.2059141993522644,
+      "learning_rate": 7.143647228366277e-06,
+      "loss": 1.5357,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8178541724424729,
+      "grad_norm": 0.21199941635131836,
+      "learning_rate": 7.10161800273582e-06,
+      "loss": 1.5796,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8184086498475187,
+      "grad_norm": 0.2183256447315216,
+      "learning_rate": 7.059700732458807e-06,
+      "loss": 1.5835,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8189631272525645,
+      "grad_norm": 0.21991661190986633,
+      "learning_rate": 7.0178955601824816e-06,
+      "loss": 1.6238,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8195176046576103,
+      "grad_norm": 0.2150619775056839,
+      "learning_rate": 6.976202628172548e-06,
+      "loss": 1.6039,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8200720820626559,
+      "grad_norm": 0.21119360625743866,
+      "learning_rate": 6.934622078312836e-06,
+      "loss": 1.5284,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8206265594677017,
+      "grad_norm": 0.21574904024600983,
+      "learning_rate": 6.8931540521046536e-06,
+      "loss": 1.5784,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8211810368727475,
+      "grad_norm": 0.21076872944831848,
+      "learning_rate": 6.851798690666429e-06,
+      "loss": 1.5671,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8217355142777932,
+      "grad_norm": 0.21518339216709137,
+      "learning_rate": 6.810556134733173e-06,
+      "loss": 1.5842,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8222899916828389,
+      "grad_norm": 0.21064461767673492,
+      "learning_rate": 6.769426524656007e-06,
+      "loss": 1.5794,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8228444690878847,
+      "grad_norm": 0.20996730029582977,
+      "learning_rate": 6.728410000401693e-06,
+      "loss": 1.5662,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8233989464929304,
+      "grad_norm": 0.21565823256969452,
+      "learning_rate": 6.687506701552156e-06,
+      "loss": 1.5227,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8239534238979762,
+      "grad_norm": 0.20833303034305573,
+      "learning_rate": 6.6467167673040134e-06,
+      "loss": 1.5601,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8245079013030219,
+      "grad_norm": 0.2161036729812622,
+      "learning_rate": 6.606040336468091e-06,
+      "loss": 1.5762,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8250623787080676,
+      "grad_norm": 0.20862828195095062,
+      "learning_rate": 6.565477547468933e-06,
+      "loss": 1.5346,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8256168561131134,
+      "grad_norm": 0.20955637097358704,
+      "learning_rate": 6.52502853834438e-06,
+      "loss": 1.5954,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8261713335181592,
+      "grad_norm": 0.21076586842536926,
+      "learning_rate": 6.484693446745089e-06,
+      "loss": 1.5716,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8267258109232049,
+      "grad_norm": 0.21574625372886658,
+      "learning_rate": 6.4444724099339996e-06,
+      "loss": 1.5426,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8272802883282506,
+      "grad_norm": 0.207157164812088,
+      "learning_rate": 6.404365564785946e-06,
+      "loss": 1.5192,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8278347657332964,
+      "grad_norm": 0.21191108226776123,
+      "learning_rate": 6.364373047787156e-06,
+      "loss": 1.6491,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8283892431383421,
+      "grad_norm": 0.21367955207824707,
+      "learning_rate": 6.324494995034789e-06,
+      "loss": 1.5497,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8289437205433878,
+      "grad_norm": 0.20905102789402008,
+      "learning_rate": 6.28473154223649e-06,
+      "loss": 1.5512,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8294981979484336,
+      "grad_norm": 0.21540366113185883,
+      "learning_rate": 6.245082824709863e-06,
+      "loss": 1.5571,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8300526753534794,
+      "grad_norm": 0.21667438745498657,
+      "learning_rate": 6.205548977382143e-06,
+      "loss": 1.6235,
+      "step": 1497
+    },
+    {
+      "epoch": 0.830607152758525,
+      "grad_norm": 0.215713232755661,
+      "learning_rate": 6.166130134789572e-06,
+      "loss": 1.5784,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8311616301635708,
+      "grad_norm": 0.20865316689014435,
+      "learning_rate": 6.126826431077071e-06,
+      "loss": 1.5567,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8317161075686166,
+      "grad_norm": 0.2152477353811264,
+      "learning_rate": 6.087637999997729e-06,
+      "loss": 1.6069,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1803,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.575018759487488e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}