diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25641 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3657,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00027344818156959256,
+      "grad_norm": 0.5185216069221497,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.952,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005468963631391851,
+      "grad_norm": 0.48264726996421814,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.9283,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008203445447087777,
+      "grad_norm": 0.3865518867969513,
+      "learning_rate": 1.5e-06,
+      "loss": 1.8876,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010937927262783702,
+      "grad_norm": 0.34167248010635376,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.9148,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013672409078479629,
+      "grad_norm": 0.6942479014396667,
+      "learning_rate": 2.5e-06,
+      "loss": 1.946,
+      "step": 5
+    },
+    {
+      "epoch": 0.0016406890894175555,
+      "grad_norm": 0.34968069195747375,
+      "learning_rate": 3e-06,
+      "loss": 1.9213,
+      "step": 6
+    },
+    {
+      "epoch": 0.0019141372709871479,
+      "grad_norm": 0.48089227080345154,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 1.8915,
+      "step": 7
+    },
+    {
+      "epoch": 0.0021875854525567405,
+      "grad_norm": 0.34136202931404114,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.9219,
+      "step": 8
+    },
+    {
+      "epoch": 0.002461033634126333,
+      "grad_norm": 0.529236912727356,
+      "learning_rate": 4.5e-06,
+      "loss": 1.9322,
+      "step": 9
+    },
+    {
+      "epoch": 0.0027344818156959257,
+      "grad_norm": 0.4663859009742737,
+      "learning_rate": 5e-06,
+      "loss": 1.8154,
+      "step": 10
+    },
+    {
+      "epoch": 0.0030079299972655183,
+      "grad_norm": 0.9236092567443848,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.9887,
+      "step": 11
+    },
+    {
+      "epoch": 0.003281378178835111,
+      "grad_norm": 0.43542373180389404,
+      "learning_rate": 6e-06,
+      "loss": 1.963,
+      "step": 12
+    },
+    {
+      "epoch": 0.003554826360404703,
+      "grad_norm": 0.33968186378479004,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.7446,
+      "step": 13
+    },
+    {
+      "epoch": 0.0038282745419742957,
+      "grad_norm": 0.2824525535106659,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.8923,
+      "step": 14
+    },
+    {
+      "epoch": 0.004101722723543888,
+      "grad_norm": 0.2588050663471222,
+      "learning_rate": 7.5e-06,
+      "loss": 1.7194,
+      "step": 15
+    },
+    {
+      "epoch": 0.004375170905113481,
+      "grad_norm": 0.3694664537906647,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.8807,
+      "step": 16
+    },
+    {
+      "epoch": 0.004648619086683074,
+      "grad_norm": 0.2177015244960785,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 1.8355,
+      "step": 17
+    },
+    {
+      "epoch": 0.004922067268252666,
+      "grad_norm": 0.395831435918808,
+      "learning_rate": 9e-06,
+      "loss": 1.9272,
+      "step": 18
+    },
+    {
+      "epoch": 0.005195515449822259,
+      "grad_norm": 0.23232288658618927,
+      "learning_rate": 9.5e-06,
+      "loss": 1.8483,
+      "step": 19
+    },
+    {
+      "epoch": 0.005468963631391851,
+      "grad_norm": 0.211470827460289,
+      "learning_rate": 1e-05,
+      "loss": 1.8662,
+      "step": 20
+    },
+    {
+      "epoch": 0.005742411812961444,
+      "grad_norm": 0.21474166214466095,
+      "learning_rate": 1.05e-05,
+      "loss": 1.8321,
+      "step": 21
+    },
+    {
+      "epoch": 0.006015859994531037,
+      "grad_norm": 0.22423861920833588,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.8677,
+      "step": 22
+    },
+    {
+      "epoch": 0.006289308176100629,
+      "grad_norm": 0.22832632064819336,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 1.8584,
+      "step": 23
+    },
+    {
+      "epoch": 0.006562756357670222,
+      "grad_norm": 0.2289043813943863,
+      "learning_rate": 1.2e-05,
+      "loss": 1.9455,
+      "step": 24
+    },
+    {
+      "epoch": 0.006836204539239814,
+      "grad_norm": 0.21795853972434998,
+      "learning_rate": 1.25e-05,
+      "loss": 1.8726,
+      "step": 25
+    },
+    {
+      "epoch": 0.007109652720809406,
+      "grad_norm": 0.2133929580450058,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.8685,
+      "step": 26
+    },
+    {
+      "epoch": 0.007383100902378999,
+      "grad_norm": 0.18260058760643005,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.8412,
+      "step": 27
+    },
+    {
+      "epoch": 0.0076565490839485915,
+      "grad_norm": 0.18040010333061218,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.8779,
+      "step": 28
+    },
+    {
+      "epoch": 0.007929997265518185,
+      "grad_norm": 0.17110276222229004,
+      "learning_rate": 1.45e-05,
+      "loss": 1.8708,
+      "step": 29
+    },
+    {
+      "epoch": 0.008203445447087777,
+      "grad_norm": 0.17357124388217926,
+      "learning_rate": 1.5e-05,
+      "loss": 1.8665,
+      "step": 30
+    },
+    {
+      "epoch": 0.00847689362865737,
+      "grad_norm": 0.1595795750617981,
+      "learning_rate": 1.55e-05,
+      "loss": 1.861,
+      "step": 31
+    },
+    {
+      "epoch": 0.008750341810226962,
+      "grad_norm": 0.18393680453300476,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.9642,
+      "step": 32
+    },
+    {
+      "epoch": 0.009023789991796555,
+      "grad_norm": 0.17827364802360535,
+      "learning_rate": 1.65e-05,
+      "loss": 1.8671,
+      "step": 33
+    },
+    {
+      "epoch": 0.009297238173366147,
+      "grad_norm": 0.1620740294456482,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.8421,
+      "step": 34
+    },
+    {
+      "epoch": 0.009570686354935739,
+      "grad_norm": 0.15932734310626984,
+      "learning_rate": 1.75e-05,
+      "loss": 1.6872,
+      "step": 35
+    },
+    {
+      "epoch": 0.009844134536505332,
+      "grad_norm": 0.1550036072731018,
+      "learning_rate": 1.8e-05,
+      "loss": 1.7036,
+      "step": 36
+    },
+    {
+      "epoch": 0.010117582718074924,
+      "grad_norm": 0.17445598542690277,
+      "learning_rate": 1.85e-05,
+      "loss": 1.9004,
+      "step": 37
+    },
+    {
+      "epoch": 0.010391030899644518,
+      "grad_norm": 0.15758390724658966,
+      "learning_rate": 1.9e-05,
+      "loss": 1.8423,
+      "step": 38
+    },
+    {
+      "epoch": 0.01066447908121411,
+      "grad_norm": 0.16238977015018463,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 1.7722,
+      "step": 39
+    },
+    {
+      "epoch": 0.010937927262783703,
+      "grad_norm": 0.14447326958179474,
+      "learning_rate": 2e-05,
+      "loss": 1.7876,
+      "step": 40
+    },
+    {
+      "epoch": 0.011211375444353295,
+      "grad_norm": 0.1725250482559204,
+      "learning_rate": 2.05e-05,
+      "loss": 1.7639,
+      "step": 41
+    },
+    {
+      "epoch": 0.011484823625922888,
+      "grad_norm": 0.14800074696540833,
+      "learning_rate": 2.1e-05,
+      "loss": 1.7457,
+      "step": 42
+    },
+    {
+      "epoch": 0.01175827180749248,
+      "grad_norm": 0.16171656548976898,
+      "learning_rate": 2.15e-05,
+      "loss": 1.8488,
+      "step": 43
+    },
+    {
+      "epoch": 0.012031719989062073,
+      "grad_norm": 0.17418281733989716,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.8017,
+      "step": 44
+    },
+    {
+      "epoch": 0.012305168170631665,
+      "grad_norm": 0.1688196063041687,
+      "learning_rate": 2.25e-05,
+      "loss": 1.9224,
+      "step": 45
+    },
+    {
+      "epoch": 0.012578616352201259,
+      "grad_norm": 0.1520584672689438,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.8044,
+      "step": 46
+    },
+    {
+      "epoch": 0.01285206453377085,
+      "grad_norm": 0.1390339732170105,
+      "learning_rate": 2.35e-05,
+      "loss": 1.709,
+      "step": 47
+    },
+    {
+      "epoch": 0.013125512715340444,
+      "grad_norm": 0.17691773176193237,
+      "learning_rate": 2.4e-05,
+      "loss": 1.8488,
+      "step": 48
+    },
+    {
+      "epoch": 0.013398960896910036,
+      "grad_norm": 0.1481151133775711,
+      "learning_rate": 2.45e-05,
+      "loss": 1.9118,
+      "step": 49
+    },
+    {
+      "epoch": 0.013672409078479627,
+      "grad_norm": 0.14990705251693726,
+      "learning_rate": 2.5e-05,
+      "loss": 1.7629,
+      "step": 50
+    },
+    {
+      "epoch": 0.01394585726004922,
+      "grad_norm": 0.15109795331954956,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 1.7675,
+      "step": 51
+    },
+    {
+      "epoch": 0.014219305441618812,
+      "grad_norm": 0.15651960670948029,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.7662,
+      "step": 52
+    },
+    {
+      "epoch": 0.014492753623188406,
+      "grad_norm": 0.15988552570343018,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 1.7708,
+      "step": 53
+    },
+    {
+      "epoch": 0.014766201804757998,
+      "grad_norm": 0.1471298187971115,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.7393,
+      "step": 54
+    },
+    {
+      "epoch": 0.015039649986327591,
+      "grad_norm": 0.1633843183517456,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 1.8271,
+      "step": 55
+    },
+    {
+      "epoch": 0.015313098167897183,
+      "grad_norm": 0.1614934802055359,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.7251,
+      "step": 56
+    },
+    {
+      "epoch": 0.015586546349466776,
+      "grad_norm": 0.15689364075660706,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.7851,
+      "step": 57
+    },
+    {
+      "epoch": 0.01585999453103637,
+      "grad_norm": 0.1521284133195877,
+      "learning_rate": 2.9e-05,
+      "loss": 1.8146,
+      "step": 58
+    },
+    {
+      "epoch": 0.01613344271260596,
+      "grad_norm": 0.15829654037952423,
+      "learning_rate": 2.95e-05,
+      "loss": 1.8003,
+      "step": 59
+    },
+    {
+      "epoch": 0.016406890894175553,
+      "grad_norm": 0.1448163092136383,
+      "learning_rate": 3e-05,
+      "loss": 1.7512,
+      "step": 60
+    },
+    {
+      "epoch": 0.016680339075745145,
+      "grad_norm": 0.1601008176803589,
+      "learning_rate": 3.05e-05,
+      "loss": 1.7764,
+      "step": 61
+    },
+    {
+      "epoch": 0.01695378725731474,
+      "grad_norm": 0.14987356960773468,
+      "learning_rate": 3.1e-05,
+      "loss": 1.7891,
+      "step": 62
+    },
+    {
+      "epoch": 0.017227235438884332,
+      "grad_norm": 0.16058160364627838,
+      "learning_rate": 3.15e-05,
+      "loss": 1.7912,
+      "step": 63
+    },
+    {
+      "epoch": 0.017500683620453924,
+      "grad_norm": 0.15416061878204346,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.8189,
+      "step": 64
+    },
+    {
+      "epoch": 0.017774131802023516,
+      "grad_norm": 0.15167462825775146,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.6827,
+      "step": 65
+    },
+    {
+      "epoch": 0.01804757998359311,
+      "grad_norm": 0.1508171409368515,
+      "learning_rate": 3.3e-05,
+      "loss": 1.7364,
+      "step": 66
+    },
+    {
+      "epoch": 0.018321028165162703,
+      "grad_norm": 0.15617215633392334,
+      "learning_rate": 3.35e-05,
+      "loss": 1.7085,
+      "step": 67
+    },
+    {
+      "epoch": 0.018594476346732294,
+      "grad_norm": 0.15548895299434662,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.8138,
+      "step": 68
+    },
+    {
+      "epoch": 0.018867924528301886,
+      "grad_norm": 0.17112572491168976,
+      "learning_rate": 3.45e-05,
+      "loss": 1.8459,
+      "step": 69
+    },
+    {
+      "epoch": 0.019141372709871478,
+      "grad_norm": 0.18400102853775024,
+      "learning_rate": 3.5e-05,
+      "loss": 1.9113,
+      "step": 70
+    },
+    {
+      "epoch": 0.019414820891441073,
+      "grad_norm": 0.16024211049079895,
+      "learning_rate": 3.55e-05,
+      "loss": 1.7861,
+      "step": 71
+    },
+    {
+      "epoch": 0.019688269073010665,
+      "grad_norm": 0.16651467978954315,
+      "learning_rate": 3.6e-05,
+      "loss": 1.626,
+      "step": 72
+    },
+    {
+      "epoch": 0.019961717254580257,
+      "grad_norm": 0.15933339297771454,
+      "learning_rate": 3.65e-05,
+      "loss": 1.7464,
+      "step": 73
+    },
+    {
+      "epoch": 0.02023516543614985,
+      "grad_norm": 0.15582682192325592,
+      "learning_rate": 3.7e-05,
+      "loss": 1.7713,
+      "step": 74
+    },
+    {
+      "epoch": 0.020508613617719443,
+      "grad_norm": 0.15848350524902344,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.7918,
+      "step": 75
+    },
+    {
+      "epoch": 0.020782061799289035,
+      "grad_norm": 0.1563899666070938,
+      "learning_rate": 3.8e-05,
+      "loss": 1.7257,
+      "step": 76
+    },
+    {
+      "epoch": 0.021055509980858627,
+      "grad_norm": 0.17177645862102509,
+      "learning_rate": 3.85e-05,
+      "loss": 1.784,
+      "step": 77
+    },
+    {
+      "epoch": 0.02132895816242822,
+      "grad_norm": 0.15895935893058777,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.7906,
+      "step": 78
+    },
+    {
+      "epoch": 0.021602406343997814,
+      "grad_norm": 0.15804462134838104,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 1.6952,
+      "step": 79
+    },
+    {
+      "epoch": 0.021875854525567406,
+      "grad_norm": 0.15477772057056427,
+      "learning_rate": 4e-05,
+      "loss": 1.6538,
+      "step": 80
+    },
+    {
+      "epoch": 0.022149302707136997,
+      "grad_norm": 0.1593542844057083,
+      "learning_rate": 4.05e-05,
+      "loss": 1.727,
+      "step": 81
+    },
+    {
+      "epoch": 0.02242275088870659,
+      "grad_norm": 0.164969801902771,
+      "learning_rate": 4.1e-05,
+      "loss": 1.6443,
+      "step": 82
+    },
+    {
+      "epoch": 0.02269619907027618,
+      "grad_norm": 0.16219045221805573,
+      "learning_rate": 4.15e-05,
+      "loss": 1.7775,
+      "step": 83
+    },
+    {
+      "epoch": 0.022969647251845776,
+      "grad_norm": 0.15507090091705322,
+      "learning_rate": 4.2e-05,
+      "loss": 1.6936,
+      "step": 84
+    },
+    {
+      "epoch": 0.023243095433415368,
+      "grad_norm": 0.17881833016872406,
+      "learning_rate": 4.25e-05,
+      "loss": 1.8355,
+      "step": 85
+    },
+    {
+      "epoch": 0.02351654361498496,
+      "grad_norm": 0.16070392727851868,
+      "learning_rate": 4.3e-05,
+      "loss": 1.6853,
+      "step": 86
+    },
+    {
+      "epoch": 0.02378999179655455,
+      "grad_norm": 0.16585201025009155,
+      "learning_rate": 4.35e-05,
+      "loss": 1.7516,
+      "step": 87
+    },
+    {
+      "epoch": 0.024063439978124147,
+      "grad_norm": 0.1578633040189743,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.7062,
+      "step": 88
+    },
+    {
+      "epoch": 0.02433688815969374,
+      "grad_norm": 0.17426982522010803,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 1.7312,
+      "step": 89
+    },
+    {
+      "epoch": 0.02461033634126333,
+      "grad_norm": 0.16550804674625397,
+      "learning_rate": 4.5e-05,
+      "loss": 1.7708,
+      "step": 90
+    },
+    {
+      "epoch": 0.024883784522832922,
+      "grad_norm": 0.17182576656341553,
+      "learning_rate": 4.55e-05,
+      "loss": 1.7964,
+      "step": 91
+    },
+    {
+      "epoch": 0.025157232704402517,
+      "grad_norm": 0.1642204076051712,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.7858,
+      "step": 92
+    },
+    {
+      "epoch": 0.02543068088597211,
+      "grad_norm": 0.17677852511405945,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 1.701,
+      "step": 93
+    },
+    {
+      "epoch": 0.0257041290675417,
+      "grad_norm": 0.17346423864364624,
+      "learning_rate": 4.7e-05,
+      "loss": 1.8136,
+      "step": 94
+    },
+    {
+      "epoch": 0.025977577249111292,
+      "grad_norm": 0.17641915380954742,
+      "learning_rate": 4.75e-05,
+      "loss": 1.7921,
+      "step": 95
+    },
+    {
+      "epoch": 0.026251025430680888,
+      "grad_norm": 0.19822247326374054,
+      "learning_rate": 4.8e-05,
+      "loss": 1.7887,
+      "step": 96
+    },
+    {
+      "epoch": 0.02652447361225048,
+      "grad_norm": 0.1817287653684616,
+      "learning_rate": 4.85e-05,
+      "loss": 1.7063,
+      "step": 97
+    },
+    {
+      "epoch": 0.02679792179382007,
+      "grad_norm": 0.1644916832447052,
+      "learning_rate": 4.9e-05,
+      "loss": 1.684,
+      "step": 98
+    },
+    {
+      "epoch": 0.027071369975389663,
+      "grad_norm": 0.19094440340995789,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 1.8106,
+      "step": 99
+    },
+    {
+      "epoch": 0.027344818156959255,
+      "grad_norm": 0.16572299599647522,
+      "learning_rate": 5e-05,
+      "loss": 1.6812,
+      "step": 100
+    },
+    {
+      "epoch": 0.02761826633852885,
+      "grad_norm": 0.16459035873413086,
+      "learning_rate": 5e-05,
+      "loss": 1.697,
+      "step": 101
+    },
+    {
+      "epoch": 0.02789171452009844,
+      "grad_norm": 0.18967324495315552,
+      "learning_rate": 5e-05,
+      "loss": 1.7078,
+      "step": 102
+    },
+    {
+      "epoch": 0.028165162701668033,
+      "grad_norm": 0.1775873899459839,
+      "learning_rate": 5e-05,
+      "loss": 1.7995,
+      "step": 103
+    },
+    {
+      "epoch": 0.028438610883237625,
+      "grad_norm": 0.18641340732574463,
+      "learning_rate": 5e-05,
+      "loss": 1.8392,
+      "step": 104
+    },
+    {
+      "epoch": 0.02871205906480722,
+      "grad_norm": 0.17424939572811127,
+      "learning_rate": 5e-05,
+      "loss": 1.7695,
+      "step": 105
+    },
+    {
+      "epoch": 0.028985507246376812,
+      "grad_norm": 0.19932492077350616,
+      "learning_rate": 5e-05,
+      "loss": 1.6987,
+      "step": 106
+    },
+    {
+      "epoch": 0.029258955427946404,
+      "grad_norm": 0.1666601300239563,
+      "learning_rate": 5e-05,
+      "loss": 1.6425,
+      "step": 107
+    },
+    {
+      "epoch": 0.029532403609515995,
+      "grad_norm": 0.19102485477924347,
+      "learning_rate": 5e-05,
+      "loss": 1.7754,
+      "step": 108
+    },
+    {
+      "epoch": 0.02980585179108559,
+      "grad_norm": 0.17251844704151154,
+      "learning_rate": 5e-05,
+      "loss": 1.7315,
+      "step": 109
+    },
+    {
+      "epoch": 0.030079299972655182,
+      "grad_norm": 0.1837855726480484,
+      "learning_rate": 5e-05,
+      "loss": 1.805,
+      "step": 110
+    },
+    {
+      "epoch": 0.030352748154224774,
+      "grad_norm": 0.19414560496807098,
+      "learning_rate": 5e-05,
+      "loss": 1.6129,
+      "step": 111
+    },
+    {
+      "epoch": 0.030626196335794366,
+      "grad_norm": 0.16896690428256989,
+      "learning_rate": 5e-05,
+      "loss": 1.8306,
+      "step": 112
+    },
+    {
+      "epoch": 0.03089964451736396,
+      "grad_norm": 0.18901929259300232,
+      "learning_rate": 5e-05,
+      "loss": 1.7591,
+      "step": 113
+    },
+    {
+      "epoch": 0.031173092698933553,
+      "grad_norm": 0.17992505431175232,
+      "learning_rate": 5e-05,
+      "loss": 1.7676,
+      "step": 114
+    },
+    {
+      "epoch": 0.031446540880503145,
+      "grad_norm": 0.1731676608324051,
+      "learning_rate": 5e-05,
+      "loss": 1.7749,
+      "step": 115
+    },
+    {
+      "epoch": 0.03171998906207274,
+      "grad_norm": 0.17806388437747955,
+      "learning_rate": 5e-05,
+      "loss": 1.7912,
+      "step": 116
+    },
+    {
+      "epoch": 0.03199343724364233,
+      "grad_norm": 0.176969513297081,
+      "learning_rate": 5e-05,
+      "loss": 1.7966,
+      "step": 117
+    },
+    {
+      "epoch": 0.03226688542521192,
+      "grad_norm": 0.18301358819007874,
+      "learning_rate": 5e-05,
+      "loss": 1.7454,
+      "step": 118
+    },
+    {
+      "epoch": 0.03254033360678151,
+      "grad_norm": 0.18621793389320374,
+      "learning_rate": 5e-05,
+      "loss": 1.8087,
+      "step": 119
+    },
+    {
+      "epoch": 0.03281378178835111,
+      "grad_norm": 0.17661884427070618,
+      "learning_rate": 5e-05,
+      "loss": 1.8529,
+      "step": 120
+    },
+    {
+      "epoch": 0.0330872299699207,
+      "grad_norm": 0.1590997874736786,
+      "learning_rate": 5e-05,
+      "loss": 1.6437,
+      "step": 121
+    },
+    {
+      "epoch": 0.03336067815149029,
+      "grad_norm": 0.15839175879955292,
+      "learning_rate": 5e-05,
+      "loss": 1.6921,
+      "step": 122
+    },
+    {
+      "epoch": 0.033634126333059886,
+      "grad_norm": 0.1695318967103958,
+      "learning_rate": 5e-05,
+      "loss": 1.7514,
+      "step": 123
+    },
+    {
+      "epoch": 0.03390757451462948,
+      "grad_norm": 0.17511457204818726,
+      "learning_rate": 5e-05,
+      "loss": 1.7665,
+      "step": 124
+    },
+    {
+      "epoch": 0.03418102269619907,
+      "grad_norm": 0.16365903615951538,
+      "learning_rate": 5e-05,
+      "loss": 1.7289,
+      "step": 125
+    },
+    {
+      "epoch": 0.034454470877768664,
+      "grad_norm": 0.16163618862628937,
+      "learning_rate": 5e-05,
+      "loss": 1.7243,
+      "step": 126
+    },
+    {
+      "epoch": 0.03472791905933825,
+      "grad_norm": 0.1713578999042511,
+      "learning_rate": 5e-05,
+      "loss": 1.8056,
+      "step": 127
+    },
+    {
+      "epoch": 0.03500136724090785,
+      "grad_norm": 0.16412892937660217,
+      "learning_rate": 5e-05,
+      "loss": 1.7385,
+      "step": 128
+    },
+    {
+      "epoch": 0.03527481542247744,
+      "grad_norm": 0.16879543662071228,
+      "learning_rate": 5e-05,
+      "loss": 1.7961,
+      "step": 129
+    },
+    {
+      "epoch": 0.03554826360404703,
+      "grad_norm": 0.1659931093454361,
+      "learning_rate": 5e-05,
+      "loss": 1.6833,
+      "step": 130
+    },
+    {
+      "epoch": 0.035821711785616626,
+      "grad_norm": 0.16397275030612946,
+      "learning_rate": 5e-05,
+      "loss": 1.7543,
+      "step": 131
+    },
+    {
+      "epoch": 0.03609515996718622,
+      "grad_norm": 0.17216797173023224,
+      "learning_rate": 5e-05,
+      "loss": 1.8237,
+      "step": 132
+    },
+    {
+      "epoch": 0.03636860814875581,
+      "grad_norm": 0.15959759056568146,
+      "learning_rate": 5e-05,
+      "loss": 1.6513,
+      "step": 133
+    },
+    {
+      "epoch": 0.036642056330325405,
+      "grad_norm": 0.17499953508377075,
+      "learning_rate": 5e-05,
+      "loss": 1.7132,
+      "step": 134
+    },
+    {
+      "epoch": 0.03691550451189499,
+      "grad_norm": 0.17289526760578156,
+      "learning_rate": 5e-05,
+      "loss": 1.6642,
+      "step": 135
+    },
+    {
+      "epoch": 0.03718895269346459,
+      "grad_norm": 0.1796165257692337,
+      "learning_rate": 5e-05,
+      "loss": 1.6765,
+      "step": 136
+    },
+    {
+      "epoch": 0.037462400875034184,
+      "grad_norm": 0.20457540452480316,
+      "learning_rate": 5e-05,
+      "loss": 1.7167,
+      "step": 137
+    },
+    {
+      "epoch": 0.03773584905660377,
+      "grad_norm": 0.18604455888271332,
+      "learning_rate": 5e-05,
+      "loss": 1.7398,
+      "step": 138
+    },
+    {
+      "epoch": 0.03800929723817337,
+      "grad_norm": 0.20968973636627197,
+      "learning_rate": 5e-05,
+      "loss": 1.7422,
+      "step": 139
+    },
+    {
+      "epoch": 0.038282745419742956,
+      "grad_norm": 0.21795013546943665,
+      "learning_rate": 5e-05,
+      "loss": 1.7568,
+      "step": 140
+    },
+    {
+      "epoch": 0.03855619360131255,
+      "grad_norm": 0.19364839792251587,
+      "learning_rate": 5e-05,
+      "loss": 1.7591,
+      "step": 141
+    },
+    {
+      "epoch": 0.038829641782882146,
+      "grad_norm": 0.2641007602214813,
+      "learning_rate": 5e-05,
+      "loss": 1.8116,
+      "step": 142
+    },
+    {
+      "epoch": 0.039103089964451734,
+      "grad_norm": 0.17127980291843414,
+      "learning_rate": 5e-05,
+      "loss": 1.7593,
+      "step": 143
+    },
+    {
+      "epoch": 0.03937653814602133,
+      "grad_norm": 0.2609815001487732,
+      "learning_rate": 5e-05,
+      "loss": 1.7366,
+      "step": 144
+    },
+    {
+      "epoch": 0.039649986327590925,
+      "grad_norm": 0.1736164689064026,
+      "learning_rate": 5e-05,
+      "loss": 1.6295,
+      "step": 145
+    },
+    {
+      "epoch": 0.03992343450916051,
+      "grad_norm": 0.17097944021224976,
+      "learning_rate": 5e-05,
+      "loss": 1.6762,
+      "step": 146
+    },
+    {
+      "epoch": 0.04019688269073011,
+      "grad_norm": 0.19413994252681732,
+      "learning_rate": 5e-05,
+      "loss": 1.7808,
+      "step": 147
+    },
+    {
+      "epoch": 0.0404703308722997,
+      "grad_norm": 0.19383320212364197,
+      "learning_rate": 5e-05,
+      "loss": 1.7765,
+      "step": 148
+    },
+    {
+      "epoch": 0.04074377905386929,
+      "grad_norm": 0.18459807336330414,
+      "learning_rate": 5e-05,
+      "loss": 1.7168,
+      "step": 149
+    },
+    {
+      "epoch": 0.04101722723543889,
+      "grad_norm": 0.18464331328868866,
+      "learning_rate": 5e-05,
+      "loss": 1.8481,
+      "step": 150
+    },
+    {
+      "epoch": 0.041290675417008475,
+      "grad_norm": 0.18524886667728424,
+      "learning_rate": 5e-05,
+      "loss": 1.7175,
+      "step": 151
+    },
+    {
+      "epoch": 0.04156412359857807,
+      "grad_norm": 0.1744994819164276,
+      "learning_rate": 5e-05,
+      "loss": 1.7207,
+      "step": 152
+    },
+    {
+      "epoch": 0.04183757178014766,
+      "grad_norm": 0.1815788894891739,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 153
+    },
+    {
+      "epoch": 0.042111019961717254,
+      "grad_norm": 0.21097545325756073,
+      "learning_rate": 5e-05,
+      "loss": 1.7963,
+      "step": 154
+    },
+    {
+      "epoch": 0.04238446814328685,
+      "grad_norm": 0.17117224633693695,
+      "learning_rate": 5e-05,
+      "loss": 1.7514,
+      "step": 155
+    },
+    {
+      "epoch": 0.04265791632485644,
+      "grad_norm": 0.17618225514888763,
+      "learning_rate": 5e-05,
+      "loss": 1.7029,
+      "step": 156
+    },
+    {
+      "epoch": 0.04293136450642603,
+      "grad_norm": 0.223564013838768,
+      "learning_rate": 5e-05,
+      "loss": 1.8719,
+      "step": 157
+    },
+    {
+      "epoch": 0.04320481268799563,
+      "grad_norm": 0.17528285086154938,
+      "learning_rate": 5e-05,
+      "loss": 1.8264,
+      "step": 158
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 0.20005618035793304,
+      "learning_rate": 5e-05,
+      "loss": 1.7559,
+      "step": 159
+    },
+    {
+      "epoch": 0.04375170905113481,
+      "grad_norm": 0.18977142870426178,
+      "learning_rate": 5e-05,
+      "loss": 1.7066,
+      "step": 160
+    },
+    {
+      "epoch": 0.0440251572327044,
+      "grad_norm": 0.18722118437290192,
+      "learning_rate": 5e-05,
+      "loss": 1.7769,
+      "step": 161
+    },
+    {
+      "epoch": 0.044298605414273995,
+      "grad_norm": 0.16823123395442963,
+      "learning_rate": 5e-05,
+      "loss": 1.7666,
+      "step": 162
+    },
+    {
+      "epoch": 0.04457205359584359,
+      "grad_norm": 0.18529324233531952,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 163
+    },
+    {
+      "epoch": 0.04484550177741318,
+      "grad_norm": 0.18550924956798553,
+      "learning_rate": 5e-05,
+      "loss": 1.6754,
+      "step": 164
+    },
+    {
+      "epoch": 0.045118949958982774,
+      "grad_norm": 0.15453596413135529,
+      "learning_rate": 5e-05,
+      "loss": 1.6601,
+      "step": 165
+    },
+    {
+      "epoch": 0.04539239814055236,
+      "grad_norm": 0.17858926951885223,
+      "learning_rate": 5e-05,
+      "loss": 1.6343,
+      "step": 166
+    },
+    {
+      "epoch": 0.04566584632212196,
+      "grad_norm": 0.1769731193780899,
+      "learning_rate": 5e-05,
+      "loss": 1.6508,
+      "step": 167
+    },
+    {
+      "epoch": 0.04593929450369155,
+      "grad_norm": 0.16900819540023804,
+      "learning_rate": 5e-05,
+      "loss": 1.7167,
+      "step": 168
+    },
+    {
+      "epoch": 0.04621274268526114,
+      "grad_norm": 0.18043388426303864,
+      "learning_rate": 5e-05,
+      "loss": 1.75,
+      "step": 169
+    },
+    {
+      "epoch": 0.046486190866830736,
+      "grad_norm": 0.2040599286556244,
+      "learning_rate": 5e-05,
+      "loss": 1.7314,
+      "step": 170
+    },
+    {
+      "epoch": 0.04675963904840033,
+      "grad_norm": 0.17198055982589722,
+      "learning_rate": 5e-05,
+      "loss": 1.6897,
+      "step": 171
+    },
+    {
+      "epoch": 0.04703308722996992,
+      "grad_norm": 0.16982243955135345,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 172
+    },
+    {
+      "epoch": 0.047306535411539515,
+      "grad_norm": 0.1677250862121582,
+      "learning_rate": 5e-05,
+      "loss": 1.7385,
+      "step": 173
+    },
+    {
+      "epoch": 0.0475799835931091,
+      "grad_norm": 0.16259929537773132,
+      "learning_rate": 5e-05,
+      "loss": 1.7417,
+      "step": 174
+    },
+    {
+      "epoch": 0.0478534317746787,
+      "grad_norm": 0.1767575442790985,
+      "learning_rate": 5e-05,
+      "loss": 1.7981,
+      "step": 175
+    },
+    {
+      "epoch": 0.04812687995624829,
+      "grad_norm": 0.17178016901016235,
+      "learning_rate": 5e-05,
+      "loss": 1.7238,
+      "step": 176
+    },
+    {
+      "epoch": 0.04840032813781788,
+      "grad_norm": 0.1756935715675354,
+      "learning_rate": 5e-05,
+      "loss": 1.6331,
+      "step": 177
+    },
+    {
+      "epoch": 0.04867377631938748,
+      "grad_norm": 0.15742701292037964,
+      "learning_rate": 5e-05,
+      "loss": 1.7367,
+      "step": 178
+    },
+    {
+      "epoch": 0.04894722450095707,
+      "grad_norm": 0.16502848267555237,
+      "learning_rate": 5e-05,
+      "loss": 1.6908,
+      "step": 179
+    },
+    {
+      "epoch": 0.04922067268252666,
+      "grad_norm": 0.1676185131072998,
+      "learning_rate": 5e-05,
+      "loss": 1.6611,
+      "step": 180
+    },
+    {
+      "epoch": 0.049494120864096255,
+      "grad_norm": 0.19482578337192535,
+      "learning_rate": 5e-05,
+      "loss": 1.7986,
+      "step": 181
+    },
+    {
+      "epoch": 0.049767569045665844,
+      "grad_norm": 0.15637359023094177,
+      "learning_rate": 5e-05,
+      "loss": 1.7225,
+      "step": 182
+    },
+    {
+      "epoch": 0.05004101722723544,
+      "grad_norm": 0.17269264161586761,
+      "learning_rate": 5e-05,
+      "loss": 1.692,
+      "step": 183
+    },
+    {
+      "epoch": 0.050314465408805034,
+      "grad_norm": 0.17836996912956238,
+      "learning_rate": 5e-05,
+      "loss": 1.7457,
+      "step": 184
+    },
+    {
+      "epoch": 0.05058791359037462,
+      "grad_norm": 0.1829950511455536,
+      "learning_rate": 5e-05,
+      "loss": 1.7991,
+      "step": 185
+    },
+    {
+      "epoch": 0.05086136177194422,
+      "grad_norm": 0.18094682693481445,
+      "learning_rate": 5e-05,
+      "loss": 1.7173,
+      "step": 186
+    },
+    {
+      "epoch": 0.051134809953513806,
+      "grad_norm": 0.14723382890224457,
+      "learning_rate": 5e-05,
+      "loss": 1.6355,
+      "step": 187
+    },
+    {
+      "epoch": 0.0514082581350834,
+      "grad_norm": 0.18650388717651367,
+      "learning_rate": 5e-05,
+      "loss": 1.7328,
+      "step": 188
+    },
+    {
+      "epoch": 0.051681706316652996,
+      "grad_norm": 0.1545383185148239,
+      "learning_rate": 5e-05,
+      "loss": 1.6386,
+      "step": 189
+    },
+    {
+      "epoch": 0.051955154498222585,
+      "grad_norm": 0.18034450709819794,
+      "learning_rate": 5e-05,
+      "loss": 1.7728,
+      "step": 190
+    },
+    {
+      "epoch": 0.05222860267979218,
+      "grad_norm": 0.17649757862091064,
+      "learning_rate": 5e-05,
+      "loss": 1.8434,
+      "step": 191
+    },
+    {
+      "epoch": 0.052502050861361775,
+      "grad_norm": 0.1520988494157791,
+      "learning_rate": 5e-05,
+      "loss": 1.6414,
+      "step": 192
+    },
+    {
+      "epoch": 0.05277549904293136,
+      "grad_norm": 0.1721171736717224,
+      "learning_rate": 5e-05,
+      "loss": 1.7516,
+      "step": 193
+    },
+    {
+      "epoch": 0.05304894722450096,
+      "grad_norm": 0.16801105439662933,
+      "learning_rate": 5e-05,
+      "loss": 1.741,
+      "step": 194
+    },
+    {
+      "epoch": 0.05332239540607055,
+      "grad_norm": 0.1697547286748886,
+      "learning_rate": 5e-05,
+      "loss": 1.7154,
+      "step": 195
+    },
+    {
+      "epoch": 0.05359584358764014,
+      "grad_norm": 0.163418710231781,
+      "learning_rate": 5e-05,
+      "loss": 1.7957,
+      "step": 196
+    },
+    {
+      "epoch": 0.05386929176920974,
+      "grad_norm": 0.15881727635860443,
+      "learning_rate": 5e-05,
+      "loss": 1.71,
+      "step": 197
+    },
+    {
+      "epoch": 0.054142739950779326,
+      "grad_norm": 0.18331541121006012,
+      "learning_rate": 5e-05,
+      "loss": 1.7412,
+      "step": 198
+    },
+    {
+      "epoch": 0.05441618813234892,
+      "grad_norm": 0.15311068296432495,
+      "learning_rate": 5e-05,
+      "loss": 1.7389,
+      "step": 199
+    },
+    {
+      "epoch": 0.05468963631391851,
+      "grad_norm": 0.17510437965393066,
+      "learning_rate": 5e-05,
+      "loss": 1.7061,
+      "step": 200
+    },
+    {
+      "epoch": 0.054963084495488104,
+      "grad_norm": 0.1618979126214981,
+      "learning_rate": 5e-05,
+      "loss": 1.7949,
+      "step": 201
+    },
+    {
+      "epoch": 0.0552365326770577,
+      "grad_norm": 0.16283638775348663,
+      "learning_rate": 5e-05,
+      "loss": 1.6262,
+      "step": 202
+    },
+    {
+      "epoch": 0.05550998085862729,
+      "grad_norm": 0.17936623096466064,
+      "learning_rate": 5e-05,
+      "loss": 1.7559,
+      "step": 203
+    },
+    {
+      "epoch": 0.05578342904019688,
+      "grad_norm": 0.15585114061832428,
+      "learning_rate": 5e-05,
+      "loss": 1.6809,
+      "step": 204
+    },
+    {
+      "epoch": 0.05605687722176648,
+      "grad_norm": 0.16536429524421692,
+      "learning_rate": 5e-05,
+      "loss": 1.8167,
+      "step": 205
+    },
+    {
+      "epoch": 0.056330325403336066,
+      "grad_norm": 0.16469433903694153,
+      "learning_rate": 5e-05,
+      "loss": 1.762,
+      "step": 206
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.15904970467090607,
+      "learning_rate": 5e-05,
+      "loss": 1.7354,
+      "step": 207
+    },
+    {
+      "epoch": 0.05687722176647525,
+      "grad_norm": 0.15505826473236084,
+      "learning_rate": 5e-05,
+      "loss": 1.6751,
+      "step": 208
+    },
+    {
+      "epoch": 0.057150669948044845,
+      "grad_norm": 0.1706695258617401,
+      "learning_rate": 5e-05,
+      "loss": 1.8423,
+      "step": 209
+    },
+    {
+      "epoch": 0.05742411812961444,
+      "grad_norm": 0.1645784229040146,
+      "learning_rate": 5e-05,
+      "loss": 1.7291,
+      "step": 210
+    },
+    {
+      "epoch": 0.05769756631118403,
+      "grad_norm": 0.15510506927967072,
+      "learning_rate": 5e-05,
+      "loss": 1.6877,
+      "step": 211
+    },
+    {
+      "epoch": 0.057971014492753624,
+      "grad_norm": 0.16437038779258728,
+      "learning_rate": 5e-05,
+      "loss": 1.6895,
+      "step": 212
+    },
+    {
+      "epoch": 0.05824446267432322,
+      "grad_norm": 0.15057580173015594,
+      "learning_rate": 5e-05,
+      "loss": 1.6953,
+      "step": 213
+    },
+    {
+      "epoch": 0.05851791085589281,
+      "grad_norm": 0.17037834227085114,
+      "learning_rate": 5e-05,
+      "loss": 1.6958,
+      "step": 214
+    },
+    {
+      "epoch": 0.0587913590374624,
+      "grad_norm": 0.1691209077835083,
+      "learning_rate": 5e-05,
+      "loss": 1.6792,
+      "step": 215
+    },
+    {
+      "epoch": 0.05906480721903199,
+      "grad_norm": 0.18123316764831543,
+      "learning_rate": 5e-05,
+      "loss": 1.7774,
+      "step": 216
+    },
+    {
+      "epoch": 0.059338255400601586,
+      "grad_norm": 0.17319968342781067,
+      "learning_rate": 5e-05,
+      "loss": 1.7516,
+      "step": 217
+    },
+    {
+      "epoch": 0.05961170358217118,
+      "grad_norm": 0.16802047193050385,
+      "learning_rate": 5e-05,
+      "loss": 1.6909,
+      "step": 218
+    },
+    {
+      "epoch": 0.05988515176374077,
+      "grad_norm": 0.16676077246665955,
+      "learning_rate": 5e-05,
+      "loss": 1.613,
+      "step": 219
+    },
+    {
+      "epoch": 0.060158599945310365,
+      "grad_norm": 0.15997642278671265,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 220
+    },
+    {
+      "epoch": 0.06043204812687995,
+      "grad_norm": 0.1547921746969223,
+      "learning_rate": 5e-05,
+      "loss": 1.6874,
+      "step": 221
+    },
+    {
+      "epoch": 0.06070549630844955,
+      "grad_norm": 0.17867213487625122,
+      "learning_rate": 5e-05,
+      "loss": 1.6957,
+      "step": 222
+    },
+    {
+      "epoch": 0.060978944490019144,
+      "grad_norm": 0.15124128758907318,
+      "learning_rate": 5e-05,
+      "loss": 1.6269,
+      "step": 223
+    },
+    {
+      "epoch": 0.06125239267158873,
+      "grad_norm": 0.15765389800071716,
+      "learning_rate": 5e-05,
+      "loss": 1.7699,
+      "step": 224
+    },
+    {
+      "epoch": 0.06152584085315833,
+      "grad_norm": 0.18673233687877655,
+      "learning_rate": 5e-05,
+      "loss": 1.8988,
+      "step": 225
+    },
+    {
+      "epoch": 0.06179928903472792,
+      "grad_norm": 0.148275688290596,
+      "learning_rate": 5e-05,
+      "loss": 1.6565,
+      "step": 226
+    },
+    {
+      "epoch": 0.06207273721629751,
+      "grad_norm": 0.17454782128334045,
+      "learning_rate": 5e-05,
+      "loss": 1.7273,
+      "step": 227
+    },
+    {
+      "epoch": 0.062346185397867106,
+      "grad_norm": 0.15484969317913055,
+      "learning_rate": 5e-05,
+      "loss": 1.6832,
+      "step": 228
+    },
+    {
+      "epoch": 0.0626196335794367,
+      "grad_norm": 0.16521553695201874,
+      "learning_rate": 5e-05,
+      "loss": 1.886,
+      "step": 229
+    },
+    {
+      "epoch": 0.06289308176100629,
+      "grad_norm": 0.1734921932220459,
+      "learning_rate": 5e-05,
+      "loss": 1.8453,
+      "step": 230
+    },
+    {
+      "epoch": 0.06316652994257588,
+      "grad_norm": 0.16037142276763916,
+      "learning_rate": 5e-05,
+      "loss": 1.6604,
+      "step": 231
+    },
+    {
+      "epoch": 0.06343997812414548,
+      "grad_norm": 0.194137305021286,
+      "learning_rate": 5e-05,
+      "loss": 1.7101,
+      "step": 232
+    },
+    {
+      "epoch": 0.06371342630571507,
+      "grad_norm": 0.16026362776756287,
+      "learning_rate": 5e-05,
+      "loss": 1.7296,
+      "step": 233
+    },
+    {
+      "epoch": 0.06398687448728466,
+      "grad_norm": 0.1773460954427719,
+      "learning_rate": 5e-05,
+      "loss": 1.6812,
+      "step": 234
+    },
+    {
+      "epoch": 0.06426032266885426,
+      "grad_norm": 0.19645363092422485,
+      "learning_rate": 5e-05,
+      "loss": 1.8694,
+      "step": 235
+    },
+    {
+      "epoch": 0.06453377085042385,
+      "grad_norm": 0.16966943442821503,
+      "learning_rate": 5e-05,
+      "loss": 1.6903,
+      "step": 236
+    },
+    {
+      "epoch": 0.06480721903199343,
+      "grad_norm": 0.21228881180286407,
+      "learning_rate": 5e-05,
+      "loss": 1.7405,
+      "step": 237
+    },
+    {
+      "epoch": 0.06508066721356302,
+      "grad_norm": 0.16073106229305267,
+      "learning_rate": 5e-05,
+      "loss": 1.7031,
+      "step": 238
+    },
+    {
+      "epoch": 0.06535411539513263,
+      "grad_norm": 0.1698150485754013,
+      "learning_rate": 5e-05,
+      "loss": 1.6663,
+      "step": 239
+    },
+    {
+      "epoch": 0.06562756357670221,
+      "grad_norm": 0.1578008383512497,
+      "learning_rate": 5e-05,
+      "loss": 1.6899,
+      "step": 240
+    },
+    {
+      "epoch": 0.0659010117582718,
+      "grad_norm": 0.1813380867242813,
+      "learning_rate": 5e-05,
+      "loss": 1.7034,
+      "step": 241
+    },
+    {
+      "epoch": 0.0661744599398414,
+      "grad_norm": 0.16481561958789825,
+      "learning_rate": 5e-05,
+      "loss": 1.7453,
+      "step": 242
+    },
+    {
+      "epoch": 0.06644790812141099,
+      "grad_norm": 0.15525515377521515,
+      "learning_rate": 5e-05,
+      "loss": 1.7296,
+      "step": 243
+    },
+    {
+      "epoch": 0.06672135630298058,
+      "grad_norm": 0.19922387599945068,
+      "learning_rate": 5e-05,
+      "loss": 1.7433,
+      "step": 244
+    },
+    {
+      "epoch": 0.06699480448455018,
+      "grad_norm": 0.18391215801239014,
+      "learning_rate": 5e-05,
+      "loss": 1.7203,
+      "step": 245
+    },
+    {
+      "epoch": 0.06726825266611977,
+      "grad_norm": 0.15986333787441254,
+      "learning_rate": 5e-05,
+      "loss": 1.6375,
+      "step": 246
+    },
+    {
+      "epoch": 0.06754170084768936,
+      "grad_norm": 0.15006937086582184,
+      "learning_rate": 5e-05,
+      "loss": 1.6448,
+      "step": 247
+    },
+    {
+      "epoch": 0.06781514902925896,
+      "grad_norm": 0.19520802795886993,
+      "learning_rate": 5e-05,
+      "loss": 1.6555,
+      "step": 248
+    },
+    {
+      "epoch": 0.06808859721082855,
+      "grad_norm": 0.14804011583328247,
+      "learning_rate": 5e-05,
+      "loss": 1.682,
+      "step": 249
+    },
+    {
+      "epoch": 0.06836204539239814,
+      "grad_norm": 0.18628214299678802,
+      "learning_rate": 5e-05,
+      "loss": 1.7169,
+      "step": 250
+    },
+    {
+      "epoch": 0.06863549357396773,
+      "grad_norm": 0.15606792271137238,
+      "learning_rate": 5e-05,
+      "loss": 1.7105,
+      "step": 251
+    },
+    {
+      "epoch": 0.06890894175553733,
+      "grad_norm": 0.17179042100906372,
+      "learning_rate": 5e-05,
+      "loss": 1.703,
+      "step": 252
+    },
+    {
+      "epoch": 0.06918238993710692,
+      "grad_norm": 0.1775059700012207,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 253
+    },
+    {
+      "epoch": 0.0694558381186765,
+      "grad_norm": 0.15647803246974945,
+      "learning_rate": 5e-05,
+      "loss": 1.6498,
+      "step": 254
+    },
+    {
+      "epoch": 0.06972928630024611,
+      "grad_norm": 0.1764082908630371,
+      "learning_rate": 5e-05,
+      "loss": 1.7309,
+      "step": 255
+    },
+    {
+      "epoch": 0.0700027344818157,
+      "grad_norm": 0.1760226935148239,
+      "learning_rate": 5e-05,
+      "loss": 1.7507,
+      "step": 256
+    },
+    {
+      "epoch": 0.07027618266338528,
+      "grad_norm": 0.1542169153690338,
+      "learning_rate": 5e-05,
+      "loss": 1.569,
+      "step": 257
+    },
+    {
+      "epoch": 0.07054963084495489,
+      "grad_norm": 0.21275527775287628,
+      "learning_rate": 5e-05,
+      "loss": 1.7928,
+      "step": 258
+    },
+    {
+      "epoch": 0.07082307902652447,
+      "grad_norm": 0.16845186054706573,
+      "learning_rate": 5e-05,
+      "loss": 1.7246,
+      "step": 259
+    },
+    {
+      "epoch": 0.07109652720809406,
+      "grad_norm": 0.2582179605960846,
+      "learning_rate": 5e-05,
+      "loss": 1.6447,
+      "step": 260
+    },
+    {
+      "epoch": 0.07136997538966366,
+      "grad_norm": 0.16100846230983734,
+      "learning_rate": 5e-05,
+      "loss": 1.7214,
+      "step": 261
+    },
+    {
+      "epoch": 0.07164342357123325,
+      "grad_norm": 0.19200801849365234,
+      "learning_rate": 5e-05,
+      "loss": 1.7948,
+      "step": 262
+    },
+    {
+      "epoch": 0.07191687175280284,
+      "grad_norm": 0.2010166496038437,
+      "learning_rate": 5e-05,
+      "loss": 1.6726,
+      "step": 263
+    },
+    {
+      "epoch": 0.07219031993437244,
+      "grad_norm": 0.15795177221298218,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 264
+    },
+    {
+      "epoch": 0.07246376811594203,
+      "grad_norm": 0.1857866495847702,
+      "learning_rate": 5e-05,
+      "loss": 1.6915,
+      "step": 265
+    },
+    {
+      "epoch": 0.07273721629751162,
+      "grad_norm": 0.16975003480911255,
+      "learning_rate": 5e-05,
+      "loss": 1.6847,
+      "step": 266
+    },
+    {
+      "epoch": 0.07301066447908121,
+      "grad_norm": 0.14858882129192352,
+      "learning_rate": 5e-05,
+      "loss": 1.7224,
+      "step": 267
+    },
+    {
+      "epoch": 0.07328411266065081,
+      "grad_norm": 0.20861102640628815,
+      "learning_rate": 5e-05,
+      "loss": 1.6836,
+      "step": 268
+    },
+    {
+      "epoch": 0.0735575608422204,
+      "grad_norm": 0.1622302383184433,
+      "learning_rate": 5e-05,
+      "loss": 1.6482,
+      "step": 269
+    },
+    {
+      "epoch": 0.07383100902378999,
+      "grad_norm": 0.1789664328098297,
+      "learning_rate": 5e-05,
+      "loss": 1.7508,
+      "step": 270
+    },
+    {
+      "epoch": 0.07410445720535959,
+      "grad_norm": 0.16430623829364777,
+      "learning_rate": 5e-05,
+      "loss": 1.735,
+      "step": 271
+    },
+    {
+      "epoch": 0.07437790538692918,
+      "grad_norm": 0.15610988438129425,
+      "learning_rate": 5e-05,
+      "loss": 1.6569,
+      "step": 272
+    },
+    {
+      "epoch": 0.07465135356849877,
+      "grad_norm": 0.16325096786022186,
+      "learning_rate": 5e-05,
+      "loss": 1.6165,
+      "step": 273
+    },
+    {
+      "epoch": 0.07492480175006837,
+      "grad_norm": 0.15770940482616425,
+      "learning_rate": 5e-05,
+      "loss": 1.727,
+      "step": 274
+    },
+    {
+      "epoch": 0.07519824993163796,
+      "grad_norm": 0.15711399912834167,
+      "learning_rate": 5e-05,
+      "loss": 1.6336,
+      "step": 275
+    },
+    {
+      "epoch": 0.07547169811320754,
+      "grad_norm": 0.16381201148033142,
+      "learning_rate": 5e-05,
+      "loss": 1.69,
+      "step": 276
+    },
+    {
+      "epoch": 0.07574514629477715,
+      "grad_norm": 0.15251821279525757,
+      "learning_rate": 5e-05,
+      "loss": 1.6241,
+      "step": 277
+    },
+    {
+      "epoch": 0.07601859447634673,
+      "grad_norm": 0.177230566740036,
+      "learning_rate": 5e-05,
+      "loss": 1.7341,
+      "step": 278
+    },
+    {
+      "epoch": 0.07629204265791632,
+      "grad_norm": 0.16340211033821106,
+      "learning_rate": 5e-05,
+      "loss": 1.726,
+      "step": 279
+    },
+    {
+      "epoch": 0.07656549083948591,
+      "grad_norm": 0.15226496756076813,
+      "learning_rate": 5e-05,
+      "loss": 1.7703,
+      "step": 280
+    },
+    {
+      "epoch": 0.07683893902105551,
+      "grad_norm": 0.1857025921344757,
+      "learning_rate": 5e-05,
+      "loss": 1.6637,
+      "step": 281
+    },
+    {
+      "epoch": 0.0771123872026251,
+      "grad_norm": 0.1753539890050888,
+      "learning_rate": 5e-05,
+      "loss": 1.8154,
+      "step": 282
+    },
+    {
+      "epoch": 0.07738583538419469,
+      "grad_norm": 0.16672004759311676,
+      "learning_rate": 5e-05,
+      "loss": 1.6617,
+      "step": 283
+    },
+    {
+      "epoch": 0.07765928356576429,
+      "grad_norm": 0.186684250831604,
+      "learning_rate": 5e-05,
+      "loss": 1.6501,
+      "step": 284
+    },
+    {
+      "epoch": 0.07793273174733388,
+      "grad_norm": 0.18949177861213684,
+      "learning_rate": 5e-05,
+      "loss": 1.7384,
+      "step": 285
+    },
+    {
+      "epoch": 0.07820617992890347,
+      "grad_norm": 0.1607401967048645,
+      "learning_rate": 5e-05,
+      "loss": 1.719,
+      "step": 286
+    },
+    {
+      "epoch": 0.07847962811047307,
+      "grad_norm": 0.15836010873317719,
+      "learning_rate": 5e-05,
+      "loss": 1.6533,
+      "step": 287
+    },
+    {
+      "epoch": 0.07875307629204266,
+      "grad_norm": 0.17572252452373505,
+      "learning_rate": 5e-05,
+      "loss": 1.7285,
+      "step": 288
+    },
+    {
+      "epoch": 0.07902652447361225,
+      "grad_norm": 0.159013569355011,
+      "learning_rate": 5e-05,
+      "loss": 1.704,
+      "step": 289
+    },
+    {
+      "epoch": 0.07929997265518185,
+      "grad_norm": 0.16590869426727295,
+      "learning_rate": 5e-05,
+      "loss": 1.7192,
+      "step": 290
+    },
+    {
+      "epoch": 0.07957342083675144,
+      "grad_norm": 0.1848197877407074,
+      "learning_rate": 5e-05,
+      "loss": 1.7157,
+      "step": 291
+    },
+    {
+      "epoch": 0.07984686901832103,
+      "grad_norm": 0.1681578904390335,
+      "learning_rate": 5e-05,
+      "loss": 1.729,
+      "step": 292
+    },
+    {
+      "epoch": 0.08012031719989061,
+      "grad_norm": 0.18026727437973022,
+      "learning_rate": 5e-05,
+      "loss": 1.6524,
+      "step": 293
+    },
+    {
+      "epoch": 0.08039376538146022,
+      "grad_norm": 0.18624775111675262,
+      "learning_rate": 5e-05,
+      "loss": 1.7211,
+      "step": 294
+    },
+    {
+      "epoch": 0.0806672135630298,
+      "grad_norm": 0.17745369672775269,
+      "learning_rate": 5e-05,
+      "loss": 1.7268,
+      "step": 295
+    },
+    {
+      "epoch": 0.0809406617445994,
+      "grad_norm": 0.196435809135437,
+      "learning_rate": 5e-05,
+      "loss": 1.6314,
+      "step": 296
+    },
+    {
+      "epoch": 0.081214109926169,
+      "grad_norm": 0.15979966521263123,
+      "learning_rate": 5e-05,
+      "loss": 1.73,
+      "step": 297
+    },
+    {
+      "epoch": 0.08148755810773858,
+      "grad_norm": 0.16943073272705078,
+      "learning_rate": 5e-05,
+      "loss": 1.7083,
+      "step": 298
+    },
+    {
+      "epoch": 0.08176100628930817,
+      "grad_norm": 0.18523457646369934,
+      "learning_rate": 5e-05,
+      "loss": 1.7286,
+      "step": 299
+    },
+    {
+      "epoch": 0.08203445447087777,
+      "grad_norm": 0.15192170441150665,
+      "learning_rate": 5e-05,
+      "loss": 1.7399,
+      "step": 300
+    },
+    {
+      "epoch": 0.08230790265244736,
+      "grad_norm": 0.18083369731903076,
+      "learning_rate": 5e-05,
+      "loss": 1.7006,
+      "step": 301
+    },
+    {
+      "epoch": 0.08258135083401695,
+      "grad_norm": 0.16836769878864288,
+      "learning_rate": 5e-05,
+      "loss": 1.8089,
+      "step": 302
+    },
+    {
+      "epoch": 0.08285479901558655,
+      "grad_norm": 0.15962150692939758,
+      "learning_rate": 5e-05,
+      "loss": 1.7109,
+      "step": 303
+    },
+    {
+      "epoch": 0.08312824719715614,
+      "grad_norm": 0.18868067860603333,
+      "learning_rate": 5e-05,
+      "loss": 1.7697,
+      "step": 304
+    },
+    {
+      "epoch": 0.08340169537872573,
+      "grad_norm": 0.16335056722164154,
+      "learning_rate": 5e-05,
+      "loss": 1.7866,
+      "step": 305
+    },
+    {
+      "epoch": 0.08367514356029532,
+      "grad_norm": 0.23511578142642975,
+      "learning_rate": 5e-05,
+      "loss": 1.8343,
+      "step": 306
+    },
+    {
+      "epoch": 0.08394859174186492,
+      "grad_norm": 0.15618833899497986,
+      "learning_rate": 5e-05,
+      "loss": 1.633,
+      "step": 307
+    },
+    {
+      "epoch": 0.08422203992343451,
+      "grad_norm": 0.16993245482444763,
+      "learning_rate": 5e-05,
+      "loss": 1.7104,
+      "step": 308
+    },
+    {
+      "epoch": 0.0844954881050041,
+      "grad_norm": 0.18835750222206116,
+      "learning_rate": 5e-05,
+      "loss": 1.7113,
+      "step": 309
+    },
+    {
+      "epoch": 0.0847689362865737,
+      "grad_norm": 0.16639582812786102,
+      "learning_rate": 5e-05,
+      "loss": 1.816,
+      "step": 310
+    },
+    {
+      "epoch": 0.08504238446814329,
+      "grad_norm": 0.15975739061832428,
+      "learning_rate": 5e-05,
+      "loss": 1.5883,
+      "step": 311
+    },
+    {
+      "epoch": 0.08531583264971287,
+      "grad_norm": 0.20657338201999664,
+      "learning_rate": 5e-05,
+      "loss": 1.6521,
+      "step": 312
+    },
+    {
+      "epoch": 0.08558928083128248,
+      "grad_norm": 0.15396147966384888,
+      "learning_rate": 5e-05,
+      "loss": 1.7602,
+      "step": 313
+    },
+    {
+      "epoch": 0.08586272901285207,
+      "grad_norm": 0.18306365609169006,
+      "learning_rate": 5e-05,
+      "loss": 1.6811,
+      "step": 314
+    },
+    {
+      "epoch": 0.08613617719442165,
+      "grad_norm": 0.1487811654806137,
+      "learning_rate": 5e-05,
+      "loss": 1.5397,
+      "step": 315
+    },
+    {
+      "epoch": 0.08640962537599126,
+      "grad_norm": 0.14201436936855316,
+      "learning_rate": 5e-05,
+      "loss": 1.5702,
+      "step": 316
+    },
+    {
+      "epoch": 0.08668307355756084,
+      "grad_norm": 0.17401176691055298,
+      "learning_rate": 5e-05,
+      "loss": 1.6514,
+      "step": 317
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.16217194497585297,
+      "learning_rate": 5e-05,
+      "loss": 1.7284,
+      "step": 318
+    },
+    {
+      "epoch": 0.08722996992070002,
+      "grad_norm": 0.1471739411354065,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 319
+    },
+    {
+      "epoch": 0.08750341810226962,
+      "grad_norm": 0.1779485046863556,
+      "learning_rate": 5e-05,
+      "loss": 1.7358,
+      "step": 320
+    },
+    {
+      "epoch": 0.08777686628383921,
+      "grad_norm": 0.15634022653102875,
+      "learning_rate": 5e-05,
+      "loss": 1.6477,
+      "step": 321
+    },
+    {
+      "epoch": 0.0880503144654088,
+      "grad_norm": 0.1602734923362732,
+      "learning_rate": 5e-05,
+      "loss": 1.7381,
+      "step": 322
+    },
+    {
+      "epoch": 0.0883237626469784,
+      "grad_norm": 0.1770986020565033,
+      "learning_rate": 5e-05,
+      "loss": 1.7719,
+      "step": 323
+    },
+    {
+      "epoch": 0.08859721082854799,
+      "grad_norm": 0.1577000766992569,
+      "learning_rate": 5e-05,
+      "loss": 1.7571,
+      "step": 324
+    },
+    {
+      "epoch": 0.08887065901011758,
+      "grad_norm": 0.158709317445755,
+      "learning_rate": 5e-05,
+      "loss": 1.7026,
+      "step": 325
+    },
+    {
+      "epoch": 0.08914410719168718,
+      "grad_norm": 0.1561996042728424,
+      "learning_rate": 5e-05,
+      "loss": 1.602,
+      "step": 326
+    },
+    {
+      "epoch": 0.08941755537325677,
+      "grad_norm": 0.14826953411102295,
+      "learning_rate": 5e-05,
+      "loss": 1.7566,
+      "step": 327
+    },
+    {
+      "epoch": 0.08969100355482636,
+      "grad_norm": 0.1718713790178299,
+      "learning_rate": 5e-05,
+      "loss": 1.7978,
+      "step": 328
+    },
+    {
+      "epoch": 0.08996445173639596,
+      "grad_norm": 0.1710497885942459,
+      "learning_rate": 5e-05,
+      "loss": 1.7204,
+      "step": 329
+    },
+    {
+      "epoch": 0.09023789991796555,
+      "grad_norm": 0.1561228483915329,
+      "learning_rate": 5e-05,
+      "loss": 1.7124,
+      "step": 330
+    },
+    {
+      "epoch": 0.09051134809953514,
+      "grad_norm": 0.18225868046283722,
+      "learning_rate": 5e-05,
+      "loss": 1.7286,
+      "step": 331
+    },
+    {
+      "epoch": 0.09078479628110472,
+      "grad_norm": 0.1495756208896637,
+      "learning_rate": 5e-05,
+      "loss": 1.6065,
+      "step": 332
+    },
+    {
+      "epoch": 0.09105824446267433,
+      "grad_norm": 0.15466246008872986,
+      "learning_rate": 5e-05,
+      "loss": 1.7078,
+      "step": 333
+    },
+    {
+      "epoch": 0.09133169264424391,
+      "grad_norm": 0.15771101415157318,
+      "learning_rate": 5e-05,
+      "loss": 1.691,
+      "step": 334
+    },
+    {
+      "epoch": 0.0916051408258135,
+      "grad_norm": 0.1566590517759323,
+      "learning_rate": 5e-05,
+      "loss": 1.8164,
+      "step": 335
+    },
+    {
+      "epoch": 0.0918785890073831,
+      "grad_norm": 0.1567210704088211,
+      "learning_rate": 5e-05,
+      "loss": 1.6887,
+      "step": 336
+    },
+    {
+      "epoch": 0.09215203718895269,
+      "grad_norm": 0.1516759693622589,
+      "learning_rate": 5e-05,
+      "loss": 1.5887,
+      "step": 337
+    },
+    {
+      "epoch": 0.09242548537052228,
+      "grad_norm": 0.15259622037410736,
+      "learning_rate": 5e-05,
+      "loss": 1.7337,
+      "step": 338
+    },
+    {
+      "epoch": 0.09269893355209188,
+      "grad_norm": 0.14548294246196747,
+      "learning_rate": 5e-05,
+      "loss": 1.6738,
+      "step": 339
+    },
+    {
+      "epoch": 0.09297238173366147,
+      "grad_norm": 0.14872747659683228,
+      "learning_rate": 5e-05,
+      "loss": 1.7034,
+      "step": 340
+    },
+    {
+      "epoch": 0.09324582991523106,
+      "grad_norm": 0.1520431935787201,
+      "learning_rate": 5e-05,
+      "loss": 1.6948,
+      "step": 341
+    },
+    {
+      "epoch": 0.09351927809680066,
+      "grad_norm": 0.15893866121768951,
+      "learning_rate": 5e-05,
+      "loss": 1.7664,
+      "step": 342
+    },
+    {
+      "epoch": 0.09379272627837025,
+      "grad_norm": 0.1537715345621109,
+      "learning_rate": 5e-05,
+      "loss": 1.6939,
+      "step": 343
+    },
+    {
+      "epoch": 0.09406617445993984,
+      "grad_norm": 0.15221412479877472,
+      "learning_rate": 5e-05,
+      "loss": 1.7216,
+      "step": 344
+    },
+    {
+      "epoch": 0.09433962264150944,
+      "grad_norm": 0.16553561389446259,
+      "learning_rate": 5e-05,
+      "loss": 1.7498,
+      "step": 345
+    },
+    {
+      "epoch": 0.09461307082307903,
+      "grad_norm": 0.148160919547081,
+      "learning_rate": 5e-05,
+      "loss": 1.6938,
+      "step": 346
+    },
+    {
+      "epoch": 0.09488651900464862,
+      "grad_norm": 0.14013923704624176,
+      "learning_rate": 5e-05,
+      "loss": 1.5675,
+      "step": 347
+    },
+    {
+      "epoch": 0.0951599671862182,
+      "grad_norm": 0.14623858034610748,
+      "learning_rate": 5e-05,
+      "loss": 1.6729,
+      "step": 348
+    },
+    {
+      "epoch": 0.09543341536778781,
+      "grad_norm": 0.16254989802837372,
+      "learning_rate": 5e-05,
+      "loss": 1.8071,
+      "step": 349
+    },
+    {
+      "epoch": 0.0957068635493574,
+      "grad_norm": 0.1588207483291626,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 350
+    },
+    {
+      "epoch": 0.09598031173092698,
+      "grad_norm": 0.15622512996196747,
+      "learning_rate": 5e-05,
+      "loss": 1.7482,
+      "step": 351
+    },
+    {
+      "epoch": 0.09625375991249659,
+      "grad_norm": 0.15577097237110138,
+      "learning_rate": 5e-05,
+      "loss": 1.7231,
+      "step": 352
+    },
+    {
+      "epoch": 0.09652720809406617,
+      "grad_norm": 0.1480865478515625,
+      "learning_rate": 5e-05,
+      "loss": 1.6877,
+      "step": 353
+    },
+    {
+      "epoch": 0.09680065627563576,
+      "grad_norm": 0.15626661479473114,
+      "learning_rate": 5e-05,
+      "loss": 1.7725,
+      "step": 354
+    },
+    {
+      "epoch": 0.09707410445720537,
+      "grad_norm": 0.17652183771133423,
+      "learning_rate": 5e-05,
+      "loss": 1.6995,
+      "step": 355
+    },
+    {
+      "epoch": 0.09734755263877495,
+      "grad_norm": 0.15815986692905426,
+      "learning_rate": 5e-05,
+      "loss": 1.7932,
+      "step": 356
+    },
+    {
+      "epoch": 0.09762100082034454,
+      "grad_norm": 0.1668958067893982,
+      "learning_rate": 5e-05,
+      "loss": 1.7085,
+      "step": 357
+    },
+    {
+      "epoch": 0.09789444900191414,
+      "grad_norm": 0.15382306277751923,
+      "learning_rate": 5e-05,
+      "loss": 1.776,
+      "step": 358
+    },
+    {
+      "epoch": 0.09816789718348373,
+      "grad_norm": 0.1443256139755249,
+      "learning_rate": 5e-05,
+      "loss": 1.6132,
+      "step": 359
+    },
+    {
+      "epoch": 0.09844134536505332,
+      "grad_norm": 0.14953619241714478,
+      "learning_rate": 5e-05,
+      "loss": 1.6197,
+      "step": 360
+    },
+    {
+      "epoch": 0.09871479354662291,
+      "grad_norm": 0.16327831149101257,
+      "learning_rate": 5e-05,
+      "loss": 1.6632,
+      "step": 361
+    },
+    {
+      "epoch": 0.09898824172819251,
+      "grad_norm": 0.16656070947647095,
+      "learning_rate": 5e-05,
+      "loss": 1.732,
+      "step": 362
+    },
+    {
+      "epoch": 0.0992616899097621,
+      "grad_norm": 0.14592863619327545,
+      "learning_rate": 5e-05,
+      "loss": 1.6502,
+      "step": 363
+    },
+    {
+      "epoch": 0.09953513809133169,
+      "grad_norm": 0.152685284614563,
+      "learning_rate": 5e-05,
+      "loss": 1.7162,
+      "step": 364
+    },
+    {
+      "epoch": 0.09980858627290129,
+      "grad_norm": 0.14510676264762878,
+      "learning_rate": 5e-05,
+      "loss": 1.5949,
+      "step": 365
+    },
+    {
+      "epoch": 0.10008203445447088,
+      "grad_norm": 0.16652299463748932,
+      "learning_rate": 5e-05,
+      "loss": 1.7656,
+      "step": 366
+    },
+    {
+      "epoch": 0.10035548263604047,
+      "grad_norm": 0.1451842486858368,
+      "learning_rate": 5e-05,
+      "loss": 1.6449,
+      "step": 367
+    },
+    {
+      "epoch": 0.10062893081761007,
+      "grad_norm": 0.18037185072898865,
+      "learning_rate": 5e-05,
+      "loss": 1.7886,
+      "step": 368
+    },
+    {
+      "epoch": 0.10090237899917966,
+      "grad_norm": 0.14976035058498383,
+      "learning_rate": 5e-05,
+      "loss": 1.5979,
+      "step": 369
+    },
+    {
+      "epoch": 0.10117582718074924,
+      "grad_norm": 0.15457609295845032,
+      "learning_rate": 5e-05,
+      "loss": 1.6899,
+      "step": 370
+    },
+    {
+      "epoch": 0.10144927536231885,
+      "grad_norm": 0.19134309887886047,
+      "learning_rate": 5e-05,
+      "loss": 1.7097,
+      "step": 371
+    },
+    {
+      "epoch": 0.10172272354388844,
+      "grad_norm": 0.14750947058200836,
+      "learning_rate": 5e-05,
+      "loss": 1.6993,
+      "step": 372
+    },
+    {
+      "epoch": 0.10199617172545802,
+      "grad_norm": 0.1667216420173645,
+      "learning_rate": 5e-05,
+      "loss": 1.7053,
+      "step": 373
+    },
+    {
+      "epoch": 0.10226961990702761,
+      "grad_norm": 0.1675473004579544,
+      "learning_rate": 5e-05,
+      "loss": 1.6688,
+      "step": 374
+    },
+    {
+      "epoch": 0.10254306808859721,
+      "grad_norm": 0.14588837325572968,
+      "learning_rate": 5e-05,
+      "loss": 1.6963,
+      "step": 375
+    },
+    {
+      "epoch": 0.1028165162701668,
+      "grad_norm": 0.17661041021347046,
+      "learning_rate": 5e-05,
+      "loss": 1.6563,
+      "step": 376
+    },
+    {
+      "epoch": 0.10308996445173639,
+      "grad_norm": 0.16083909571170807,
+      "learning_rate": 5e-05,
+      "loss": 1.6893,
+      "step": 377
+    },
+    {
+      "epoch": 0.10336341263330599,
+      "grad_norm": 0.14849358797073364,
+      "learning_rate": 5e-05,
+      "loss": 1.6621,
+      "step": 378
+    },
+    {
+      "epoch": 0.10363686081487558,
+      "grad_norm": 0.15476635098457336,
+      "learning_rate": 5e-05,
+      "loss": 1.7159,
+      "step": 379
+    },
+    {
+      "epoch": 0.10391030899644517,
+      "grad_norm": 0.16712796688079834,
+      "learning_rate": 5e-05,
+      "loss": 1.728,
+      "step": 380
+    },
+    {
+      "epoch": 0.10418375717801477,
+      "grad_norm": 0.15185219049453735,
+      "learning_rate": 5e-05,
+      "loss": 1.674,
+      "step": 381
+    },
+    {
+      "epoch": 0.10445720535958436,
+      "grad_norm": 0.16834664344787598,
+      "learning_rate": 5e-05,
+      "loss": 1.7249,
+      "step": 382
+    },
+    {
+      "epoch": 0.10473065354115395,
+      "grad_norm": 0.15080732107162476,
+      "learning_rate": 5e-05,
+      "loss": 1.7108,
+      "step": 383
+    },
+    {
+      "epoch": 0.10500410172272355,
+      "grad_norm": 0.17864330112934113,
+      "learning_rate": 5e-05,
+      "loss": 1.743,
+      "step": 384
+    },
+    {
+      "epoch": 0.10527754990429314,
+      "grad_norm": 0.1494104564189911,
+      "learning_rate": 5e-05,
+      "loss": 1.7065,
+      "step": 385
+    },
+    {
+      "epoch": 0.10555099808586273,
+      "grad_norm": 0.1590685397386551,
+      "learning_rate": 5e-05,
+      "loss": 1.6841,
+      "step": 386
+    },
+    {
+      "epoch": 0.10582444626743231,
+      "grad_norm": 0.15133905410766602,
+      "learning_rate": 5e-05,
+      "loss": 1.7493,
+      "step": 387
+    },
+    {
+      "epoch": 0.10609789444900192,
+      "grad_norm": 0.16563621163368225,
+      "learning_rate": 5e-05,
+      "loss": 1.7181,
+      "step": 388
+    },
+    {
+      "epoch": 0.1063713426305715,
+      "grad_norm": 0.14327523112297058,
+      "learning_rate": 5e-05,
+      "loss": 1.6653,
+      "step": 389
+    },
+    {
+      "epoch": 0.1066447908121411,
+      "grad_norm": 0.14864635467529297,
+      "learning_rate": 5e-05,
+      "loss": 1.5826,
+      "step": 390
+    },
+    {
+      "epoch": 0.1069182389937107,
+      "grad_norm": 0.15405511856079102,
+      "learning_rate": 5e-05,
+      "loss": 1.7333,
+      "step": 391
+    },
+    {
+      "epoch": 0.10719168717528028,
+      "grad_norm": 0.1498226821422577,
+      "learning_rate": 5e-05,
+      "loss": 1.7149,
+      "step": 392
+    },
+    {
+      "epoch": 0.10746513535684987,
+      "grad_norm": 0.14723950624465942,
+      "learning_rate": 5e-05,
+      "loss": 1.7088,
+      "step": 393
+    },
+    {
+      "epoch": 0.10773858353841947,
+      "grad_norm": 0.15685135126113892,
+      "learning_rate": 5e-05,
+      "loss": 1.717,
+      "step": 394
+    },
+    {
+      "epoch": 0.10801203171998906,
+      "grad_norm": 0.16080626845359802,
+      "learning_rate": 5e-05,
+      "loss": 1.7603,
+      "step": 395
+    },
+    {
+      "epoch": 0.10828547990155865,
+      "grad_norm": 0.17556969821453094,
+      "learning_rate": 5e-05,
+      "loss": 1.7453,
+      "step": 396
+    },
+    {
+      "epoch": 0.10855892808312825,
+      "grad_norm": 0.16690774261951447,
+      "learning_rate": 5e-05,
+      "loss": 1.7831,
+      "step": 397
+    },
+    {
+      "epoch": 0.10883237626469784,
+      "grad_norm": 0.16693341732025146,
+      "learning_rate": 5e-05,
+      "loss": 1.7015,
+      "step": 398
+    },
+    {
+      "epoch": 0.10910582444626743,
+      "grad_norm": 0.15416108071804047,
+      "learning_rate": 5e-05,
+      "loss": 1.6887,
+      "step": 399
+    },
+    {
+      "epoch": 0.10937927262783702,
+      "grad_norm": 0.15936125814914703,
+      "learning_rate": 5e-05,
+      "loss": 1.6774,
+      "step": 400
+    },
+    {
+      "epoch": 0.10965272080940662,
+      "grad_norm": 0.15093334019184113,
+      "learning_rate": 5e-05,
+      "loss": 1.6507,
+      "step": 401
+    },
+    {
+      "epoch": 0.10992616899097621,
+      "grad_norm": 0.1748283952474594,
+      "learning_rate": 5e-05,
+      "loss": 1.6893,
+      "step": 402
+    },
+    {
+      "epoch": 0.1101996171725458,
+      "grad_norm": 0.17557309567928314,
+      "learning_rate": 5e-05,
+      "loss": 1.7279,
+      "step": 403
+    },
+    {
+      "epoch": 0.1104730653541154,
+      "grad_norm": 0.15320509672164917,
+      "learning_rate": 5e-05,
+      "loss": 1.6621,
+      "step": 404
+    },
+    {
+      "epoch": 0.11074651353568499,
+      "grad_norm": 0.1499612033367157,
+      "learning_rate": 5e-05,
+      "loss": 1.7715,
+      "step": 405
+    },
+    {
+      "epoch": 0.11101996171725458,
+      "grad_norm": 0.16688010096549988,
+      "learning_rate": 5e-05,
+      "loss": 1.7823,
+      "step": 406
+    },
+    {
+      "epoch": 0.11129340989882418,
+      "grad_norm": 0.157332643866539,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 407
+    },
+    {
+      "epoch": 0.11156685808039377,
+      "grad_norm": 0.14416515827178955,
+      "learning_rate": 5e-05,
+      "loss": 1.6094,
+      "step": 408
+    },
+    {
+      "epoch": 0.11184030626196335,
+      "grad_norm": 0.1920742392539978,
+      "learning_rate": 5e-05,
+      "loss": 1.7436,
+      "step": 409
+    },
+    {
+      "epoch": 0.11211375444353296,
+      "grad_norm": 0.1514004021883011,
+      "learning_rate": 5e-05,
+      "loss": 1.7049,
+      "step": 410
+    },
+    {
+      "epoch": 0.11238720262510254,
+      "grad_norm": 0.16829192638397217,
+      "learning_rate": 5e-05,
+      "loss": 1.7128,
+      "step": 411
+    },
+    {
+      "epoch": 0.11266065080667213,
+      "grad_norm": 0.16184952855110168,
+      "learning_rate": 5e-05,
+      "loss": 1.697,
+      "step": 412
+    },
+    {
+      "epoch": 0.11293409898824174,
+      "grad_norm": 0.15670904517173767,
+      "learning_rate": 5e-05,
+      "loss": 1.6045,
+      "step": 413
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.17832054197788239,
+      "learning_rate": 5e-05,
+      "loss": 1.7311,
+      "step": 414
+    },
+    {
+      "epoch": 0.11348099535138091,
+      "grad_norm": 0.15141215920448303,
+      "learning_rate": 5e-05,
+      "loss": 1.7269,
+      "step": 415
+    },
+    {
+      "epoch": 0.1137544435329505,
+      "grad_norm": 0.16881325840950012,
+      "learning_rate": 5e-05,
+      "loss": 1.753,
+      "step": 416
+    },
+    {
+      "epoch": 0.1140278917145201,
+      "grad_norm": 0.14302976429462433,
+      "learning_rate": 5e-05,
+      "loss": 1.6762,
+      "step": 417
+    },
+    {
+      "epoch": 0.11430133989608969,
+      "grad_norm": 0.15213017165660858,
+      "learning_rate": 5e-05,
+      "loss": 1.6379,
+      "step": 418
+    },
+    {
+      "epoch": 0.11457478807765928,
+      "grad_norm": 0.14254313707351685,
+      "learning_rate": 5e-05,
+      "loss": 1.702,
+      "step": 419
+    },
+    {
+      "epoch": 0.11484823625922888,
+      "grad_norm": 0.14474402368068695,
+      "learning_rate": 5e-05,
+      "loss": 1.6177,
+      "step": 420
+    },
+    {
+      "epoch": 0.11512168444079847,
+      "grad_norm": 0.1474299430847168,
+      "learning_rate": 5e-05,
+      "loss": 1.6375,
+      "step": 421
+    },
+    {
+      "epoch": 0.11539513262236806,
+      "grad_norm": 0.15810638666152954,
+      "learning_rate": 5e-05,
+      "loss": 1.7308,
+      "step": 422
+    },
+    {
+      "epoch": 0.11566858080393766,
+      "grad_norm": 0.15957583487033844,
+      "learning_rate": 5e-05,
+      "loss": 1.662,
+      "step": 423
+    },
+    {
+      "epoch": 0.11594202898550725,
+      "grad_norm": 0.1710730344057083,
+      "learning_rate": 5e-05,
+      "loss": 1.8709,
+      "step": 424
+    },
+    {
+      "epoch": 0.11621547716707684,
+      "grad_norm": 0.15612466633319855,
+      "learning_rate": 5e-05,
+      "loss": 1.6212,
+      "step": 425
+    },
+    {
+      "epoch": 0.11648892534864644,
+      "grad_norm": 0.1628880500793457,
+      "learning_rate": 5e-05,
+      "loss": 1.8161,
+      "step": 426
+    },
+    {
+      "epoch": 0.11676237353021603,
+      "grad_norm": 0.15160787105560303,
+      "learning_rate": 5e-05,
+      "loss": 1.738,
+      "step": 427
+    },
+    {
+      "epoch": 0.11703582171178561,
+      "grad_norm": 0.15054672956466675,
+      "learning_rate": 5e-05,
+      "loss": 1.6689,
+      "step": 428
+    },
+    {
+      "epoch": 0.1173092698933552,
+      "grad_norm": 0.16296370327472687,
+      "learning_rate": 5e-05,
+      "loss": 1.652,
+      "step": 429
+    },
+    {
+      "epoch": 0.1175827180749248,
+      "grad_norm": 0.14984969794750214,
+      "learning_rate": 5e-05,
+      "loss": 1.678,
+      "step": 430
+    },
+    {
+      "epoch": 0.1178561662564944,
+      "grad_norm": 0.1585860401391983,
+      "learning_rate": 5e-05,
+      "loss": 1.5993,
+      "step": 431
+    },
+    {
+      "epoch": 0.11812961443806398,
+      "grad_norm": 0.16059096157550812,
+      "learning_rate": 5e-05,
+      "loss": 1.6874,
+      "step": 432
+    },
+    {
+      "epoch": 0.11840306261963358,
+      "grad_norm": 0.16525696218013763,
+      "learning_rate": 5e-05,
+      "loss": 1.6795,
+      "step": 433
+    },
+    {
+      "epoch": 0.11867651080120317,
+      "grad_norm": 0.18887005746364594,
+      "learning_rate": 5e-05,
+      "loss": 1.6089,
+      "step": 434
+    },
+    {
+      "epoch": 0.11894995898277276,
+      "grad_norm": 0.15065988898277283,
+      "learning_rate": 5e-05,
+      "loss": 1.6332,
+      "step": 435
+    },
+    {
+      "epoch": 0.11922340716434236,
+      "grad_norm": 0.19893988966941833,
+      "learning_rate": 5e-05,
+      "loss": 1.721,
+      "step": 436
+    },
+    {
+      "epoch": 0.11949685534591195,
+      "grad_norm": 0.14734850823879242,
+      "learning_rate": 5e-05,
+      "loss": 1.6332,
+      "step": 437
+    },
+    {
+      "epoch": 0.11977030352748154,
+      "grad_norm": 0.18723982572555542,
+      "learning_rate": 5e-05,
+      "loss": 1.777,
+      "step": 438
+    },
+    {
+      "epoch": 0.12004375170905114,
+      "grad_norm": 0.1655622124671936,
+      "learning_rate": 5e-05,
+      "loss": 1.835,
+      "step": 439
+    },
+    {
+      "epoch": 0.12031719989062073,
+      "grad_norm": 0.1495118886232376,
+      "learning_rate": 5e-05,
+      "loss": 1.5813,
+      "step": 440
+    },
+    {
+      "epoch": 0.12059064807219032,
+      "grad_norm": 0.1608804315328598,
+      "learning_rate": 5e-05,
+      "loss": 1.6728,
+      "step": 441
+    },
+    {
+      "epoch": 0.1208640962537599,
+      "grad_norm": 0.1557897925376892,
+      "learning_rate": 5e-05,
+      "loss": 1.7919,
+      "step": 442
+    },
+    {
+      "epoch": 0.12113754443532951,
+      "grad_norm": 0.14694997668266296,
+      "learning_rate": 5e-05,
+      "loss": 1.6916,
+      "step": 443
+    },
+    {
+      "epoch": 0.1214109926168991,
+      "grad_norm": 0.15379074215888977,
+      "learning_rate": 5e-05,
+      "loss": 1.7453,
+      "step": 444
+    },
+    {
+      "epoch": 0.12168444079846868,
+      "grad_norm": 0.1691819578409195,
+      "learning_rate": 5e-05,
+      "loss": 1.7035,
+      "step": 445
+    },
+    {
+      "epoch": 0.12195788898003829,
+      "grad_norm": 0.16844218969345093,
+      "learning_rate": 5e-05,
+      "loss": 1.6465,
+      "step": 446
+    },
+    {
+      "epoch": 0.12223133716160788,
+      "grad_norm": 0.1534705013036728,
+      "learning_rate": 5e-05,
+      "loss": 1.7147,
+      "step": 447
+    },
+    {
+      "epoch": 0.12250478534317746,
+      "grad_norm": 0.16286121308803558,
+      "learning_rate": 5e-05,
+      "loss": 1.795,
+      "step": 448
+    },
+    {
+      "epoch": 0.12277823352474707,
+      "grad_norm": 0.1552685797214508,
+      "learning_rate": 5e-05,
+      "loss": 1.7379,
+      "step": 449
+    },
+    {
+      "epoch": 0.12305168170631665,
+      "grad_norm": 0.15220941603183746,
+      "learning_rate": 5e-05,
+      "loss": 1.5912,
+      "step": 450
+    },
+    {
+      "epoch": 0.12332512988788624,
+      "grad_norm": 0.15760937333106995,
+      "learning_rate": 5e-05,
+      "loss": 1.723,
+      "step": 451
+    },
+    {
+      "epoch": 0.12359857806945584,
+      "grad_norm": 0.14814068377017975,
+      "learning_rate": 5e-05,
+      "loss": 1.6482,
+      "step": 452
+    },
+    {
+      "epoch": 0.12387202625102543,
+      "grad_norm": 0.15846781432628632,
+      "learning_rate": 5e-05,
+      "loss": 1.6243,
+      "step": 453
+    },
+    {
+      "epoch": 0.12414547443259502,
+      "grad_norm": 0.14988292753696442,
+      "learning_rate": 5e-05,
+      "loss": 1.6261,
+      "step": 454
+    },
+    {
+      "epoch": 0.12441892261416461,
+      "grad_norm": 0.16977062821388245,
+      "learning_rate": 5e-05,
+      "loss": 1.6638,
+      "step": 455
+    },
+    {
+      "epoch": 0.12469237079573421,
+      "grad_norm": 0.16247521340847015,
+      "learning_rate": 5e-05,
+      "loss": 1.6939,
+      "step": 456
+    },
+    {
+      "epoch": 0.1249658189773038,
+      "grad_norm": 0.15425889194011688,
+      "learning_rate": 5e-05,
+      "loss": 1.6645,
+      "step": 457
+    },
+    {
+      "epoch": 0.1252392671588734,
+      "grad_norm": 0.18221181631088257,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 458
+    },
+    {
+      "epoch": 0.12551271534044298,
+      "grad_norm": 0.14771276712417603,
+      "learning_rate": 5e-05,
+      "loss": 1.7018,
+      "step": 459
+    },
+    {
+      "epoch": 0.12578616352201258,
+      "grad_norm": 0.175098717212677,
+      "learning_rate": 5e-05,
+      "loss": 1.78,
+      "step": 460
+    },
+    {
+      "epoch": 0.12605961170358218,
+      "grad_norm": 0.16985467076301575,
+      "learning_rate": 5e-05,
+      "loss": 1.6305,
+      "step": 461
+    },
+    {
+      "epoch": 0.12633305988515175,
+      "grad_norm": 0.1552826166152954,
+      "learning_rate": 5e-05,
+      "loss": 1.6986,
+      "step": 462
+    },
+    {
+      "epoch": 0.12660650806672136,
+      "grad_norm": 0.17271657288074493,
+      "learning_rate": 5e-05,
+      "loss": 1.7182,
+      "step": 463
+    },
+    {
+      "epoch": 0.12687995624829096,
+      "grad_norm": 0.1511317491531372,
+      "learning_rate": 5e-05,
+      "loss": 1.6274,
+      "step": 464
+    },
+    {
+      "epoch": 0.12715340442986053,
+      "grad_norm": 0.1526670604944229,
+      "learning_rate": 5e-05,
+      "loss": 1.6481,
+      "step": 465
+    },
+    {
+      "epoch": 0.12742685261143014,
+      "grad_norm": 0.15212751924991608,
+      "learning_rate": 5e-05,
+      "loss": 1.7155,
+      "step": 466
+    },
+    {
+      "epoch": 0.12770030079299974,
+      "grad_norm": 0.16328099370002747,
+      "learning_rate": 5e-05,
+      "loss": 1.7052,
+      "step": 467
+    },
+    {
+      "epoch": 0.1279737489745693,
+      "grad_norm": 0.14235898852348328,
+      "learning_rate": 5e-05,
+      "loss": 1.6514,
+      "step": 468
+    },
+    {
+      "epoch": 0.12824719715613891,
+      "grad_norm": 0.17346210777759552,
+      "learning_rate": 5e-05,
+      "loss": 1.6711,
+      "step": 469
+    },
+    {
+      "epoch": 0.12852064533770852,
+      "grad_norm": 0.17061196267604828,
+      "learning_rate": 5e-05,
+      "loss": 1.6815,
+      "step": 470
+    },
+    {
+      "epoch": 0.1287940935192781,
+      "grad_norm": 0.16796523332595825,
+      "learning_rate": 5e-05,
+      "loss": 1.6304,
+      "step": 471
+    },
+    {
+      "epoch": 0.1290675417008477,
+      "grad_norm": 0.17930328845977783,
+      "learning_rate": 5e-05,
+      "loss": 1.7399,
+      "step": 472
+    },
+    {
+      "epoch": 0.1293409898824173,
+      "grad_norm": 0.16334758698940277,
+      "learning_rate": 5e-05,
+      "loss": 1.6031,
+      "step": 473
+    },
+    {
+      "epoch": 0.12961443806398687,
+      "grad_norm": 0.18290555477142334,
+      "learning_rate": 5e-05,
+      "loss": 1.7326,
+      "step": 474
+    },
+    {
+      "epoch": 0.12988788624555647,
+      "grad_norm": 0.15675663948059082,
+      "learning_rate": 5e-05,
+      "loss": 1.6692,
+      "step": 475
+    },
+    {
+      "epoch": 0.13016133442712605,
+      "grad_norm": 0.1945638507604599,
+      "learning_rate": 5e-05,
+      "loss": 1.6891,
+      "step": 476
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.19286639988422394,
+      "learning_rate": 5e-05,
+      "loss": 1.7076,
+      "step": 477
+    },
+    {
+      "epoch": 0.13070823079026525,
+      "grad_norm": 0.17249007523059845,
+      "learning_rate": 5e-05,
+      "loss": 1.706,
+      "step": 478
+    },
+    {
+      "epoch": 0.13098167897183483,
+      "grad_norm": 0.1968424767255783,
+      "learning_rate": 5e-05,
+      "loss": 1.6126,
+      "step": 479
+    },
+    {
+      "epoch": 0.13125512715340443,
+      "grad_norm": 0.19258543848991394,
+      "learning_rate": 5e-05,
+      "loss": 1.7572,
+      "step": 480
+    },
+    {
+      "epoch": 0.13152857533497403,
+      "grad_norm": 0.16817843914031982,
+      "learning_rate": 5e-05,
+      "loss": 1.6691,
+      "step": 481
+    },
+    {
+      "epoch": 0.1318020235165436,
+      "grad_norm": 0.21118032932281494,
+      "learning_rate": 5e-05,
+      "loss": 1.6716,
+      "step": 482
+    },
+    {
+      "epoch": 0.1320754716981132,
+      "grad_norm": 0.15985234081745148,
+      "learning_rate": 5e-05,
+      "loss": 1.7348,
+      "step": 483
+    },
+    {
+      "epoch": 0.1323489198796828,
+      "grad_norm": 0.20505684614181519,
+      "learning_rate": 5e-05,
+      "loss": 1.7032,
+      "step": 484
+    },
+    {
+      "epoch": 0.13262236806125238,
+      "grad_norm": 0.18702927231788635,
+      "learning_rate": 5e-05,
+      "loss": 1.6254,
+      "step": 485
+    },
+    {
+      "epoch": 0.13289581624282198,
+      "grad_norm": 0.17232342064380646,
+      "learning_rate": 5e-05,
+      "loss": 1.7039,
+      "step": 486
+    },
+    {
+      "epoch": 0.1331692644243916,
+      "grad_norm": 0.20174479484558105,
+      "learning_rate": 5e-05,
+      "loss": 1.7167,
+      "step": 487
+    },
+    {
+      "epoch": 0.13344271260596116,
+      "grad_norm": 0.15040408074855804,
+      "learning_rate": 5e-05,
+      "loss": 1.7868,
+      "step": 488
+    },
+    {
+      "epoch": 0.13371616078753076,
+      "grad_norm": 0.2027222216129303,
+      "learning_rate": 5e-05,
+      "loss": 1.7302,
+      "step": 489
+    },
+    {
+      "epoch": 0.13398960896910037,
+      "grad_norm": 0.1705685704946518,
+      "learning_rate": 5e-05,
+      "loss": 1.7658,
+      "step": 490
+    },
+    {
+      "epoch": 0.13426305715066994,
+      "grad_norm": 0.15521638095378876,
+      "learning_rate": 5e-05,
+      "loss": 1.7048,
+      "step": 491
+    },
+    {
+      "epoch": 0.13453650533223954,
+      "grad_norm": 0.16530796885490417,
+      "learning_rate": 5e-05,
+      "loss": 1.7096,
+      "step": 492
+    },
+    {
+      "epoch": 0.13480995351380914,
+      "grad_norm": 0.1526857316493988,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 493
+    },
+    {
+      "epoch": 0.13508340169537872,
+      "grad_norm": 0.15197399258613586,
+      "learning_rate": 5e-05,
+      "loss": 1.7544,
+      "step": 494
+    },
+    {
+      "epoch": 0.13535684987694832,
+      "grad_norm": 0.17518053948879242,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 495
+    },
+    {
+      "epoch": 0.13563029805851792,
+      "grad_norm": 0.18786345422267914,
+      "learning_rate": 5e-05,
+      "loss": 1.6955,
+      "step": 496
+    },
+    {
+      "epoch": 0.1359037462400875,
+      "grad_norm": 0.16572856903076172,
+      "learning_rate": 5e-05,
+      "loss": 1.8124,
+      "step": 497
+    },
+    {
+      "epoch": 0.1361771944216571,
+      "grad_norm": 0.20353186130523682,
+      "learning_rate": 5e-05,
+      "loss": 1.6766,
+      "step": 498
+    },
+    {
+      "epoch": 0.1364506426032267,
+      "grad_norm": 0.15591098368167877,
+      "learning_rate": 5e-05,
+      "loss": 1.6693,
+      "step": 499
+    },
+    {
+      "epoch": 0.13672409078479628,
+      "grad_norm": 0.17146877944469452,
+      "learning_rate": 5e-05,
+      "loss": 1.5863,
+      "step": 500
+    },
+    {
+      "epoch": 0.13699753896636588,
+      "grad_norm": 0.1932230293750763,
+      "learning_rate": 5e-05,
+      "loss": 1.6442,
+      "step": 501
+    },
+    {
+      "epoch": 0.13727098714793545,
+      "grad_norm": 0.14423272013664246,
+      "learning_rate": 5e-05,
+      "loss": 1.6168,
+      "step": 502
+    },
+    {
+      "epoch": 0.13754443532950505,
+      "grad_norm": 0.15283560752868652,
+      "learning_rate": 5e-05,
+      "loss": 1.661,
+      "step": 503
+    },
+    {
+      "epoch": 0.13781788351107466,
+      "grad_norm": 0.19705916941165924,
+      "learning_rate": 5e-05,
+      "loss": 1.7297,
+      "step": 504
+    },
+    {
+      "epoch": 0.13809133169264423,
+      "grad_norm": 0.16028250753879547,
+      "learning_rate": 5e-05,
+      "loss": 1.764,
+      "step": 505
+    },
+    {
+      "epoch": 0.13836477987421383,
+      "grad_norm": 0.18517790734767914,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 506
+    },
+    {
+      "epoch": 0.13863822805578344,
+      "grad_norm": 0.17298881709575653,
+      "learning_rate": 5e-05,
+      "loss": 1.6147,
+      "step": 507
+    },
+    {
+      "epoch": 0.138911676237353,
+      "grad_norm": 0.1618654429912567,
+      "learning_rate": 5e-05,
+      "loss": 1.7223,
+      "step": 508
+    },
+    {
+      "epoch": 0.1391851244189226,
+      "grad_norm": 0.17122521996498108,
+      "learning_rate": 5e-05,
+      "loss": 1.6623,
+      "step": 509
+    },
+    {
+      "epoch": 0.13945857260049221,
+      "grad_norm": 0.1496722251176834,
+      "learning_rate": 5e-05,
+      "loss": 1.6599,
+      "step": 510
+    },
+    {
+      "epoch": 0.1397320207820618,
+      "grad_norm": 0.14283980429172516,
+      "learning_rate": 5e-05,
+      "loss": 1.6108,
+      "step": 511
+    },
+    {
+      "epoch": 0.1400054689636314,
+      "grad_norm": 0.15560725331306458,
+      "learning_rate": 5e-05,
+      "loss": 1.6407,
+      "step": 512
+    },
+    {
+      "epoch": 0.140278917145201,
+      "grad_norm": 0.14533740282058716,
+      "learning_rate": 5e-05,
+      "loss": 1.6877,
+      "step": 513
+    },
+    {
+      "epoch": 0.14055236532677057,
+      "grad_norm": 0.15537096560001373,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 514
+    },
+    {
+      "epoch": 0.14082581350834017,
+      "grad_norm": 0.14118176698684692,
+      "learning_rate": 5e-05,
+      "loss": 1.6312,
+      "step": 515
+    },
+    {
+      "epoch": 0.14109926168990977,
+      "grad_norm": 0.14146625995635986,
+      "learning_rate": 5e-05,
+      "loss": 1.5738,
+      "step": 516
+    },
+    {
+      "epoch": 0.14137270987147935,
+      "grad_norm": 0.15033139288425446,
+      "learning_rate": 5e-05,
+      "loss": 1.7774,
+      "step": 517
+    },
+    {
+      "epoch": 0.14164615805304895,
+      "grad_norm": 0.141265869140625,
+      "learning_rate": 5e-05,
+      "loss": 1.6147,
+      "step": 518
+    },
+    {
+      "epoch": 0.14191960623461855,
+      "grad_norm": 0.14538274705410004,
+      "learning_rate": 5e-05,
+      "loss": 1.6542,
+      "step": 519
+    },
+    {
+      "epoch": 0.14219305441618812,
+      "grad_norm": 0.15939506888389587,
+      "learning_rate": 5e-05,
+      "loss": 1.6933,
+      "step": 520
+    },
+    {
+      "epoch": 0.14246650259775773,
+      "grad_norm": 0.1489834487438202,
+      "learning_rate": 5e-05,
+      "loss": 1.6392,
+      "step": 521
+    },
+    {
+      "epoch": 0.14273995077932733,
+      "grad_norm": 0.18363149464130402,
+      "learning_rate": 5e-05,
+      "loss": 1.7042,
+      "step": 522
+    },
+    {
+      "epoch": 0.1430133989608969,
+      "grad_norm": 0.16449148952960968,
+      "learning_rate": 5e-05,
+      "loss": 1.6792,
+      "step": 523
+    },
+    {
+      "epoch": 0.1432868471424665,
+      "grad_norm": 0.15585970878601074,
+      "learning_rate": 5e-05,
+      "loss": 1.6823,
+      "step": 524
+    },
+    {
+      "epoch": 0.1435602953240361,
+      "grad_norm": 0.1625368297100067,
+      "learning_rate": 5e-05,
+      "loss": 1.6709,
+      "step": 525
+    },
+    {
+      "epoch": 0.14383374350560568,
+      "grad_norm": 0.16228622198104858,
+      "learning_rate": 5e-05,
+      "loss": 1.7291,
+      "step": 526
+    },
+    {
+      "epoch": 0.14410719168717528,
+      "grad_norm": 0.16529570519924164,
+      "learning_rate": 5e-05,
+      "loss": 1.7617,
+      "step": 527
+    },
+    {
+      "epoch": 0.1443806398687449,
+      "grad_norm": 0.1438167542219162,
+      "learning_rate": 5e-05,
+      "loss": 1.6428,
+      "step": 528
+    },
+    {
+      "epoch": 0.14465408805031446,
+      "grad_norm": 0.16496515274047852,
+      "learning_rate": 5e-05,
+      "loss": 1.771,
+      "step": 529
+    },
+    {
+      "epoch": 0.14492753623188406,
+      "grad_norm": 0.16147810220718384,
+      "learning_rate": 5e-05,
+      "loss": 1.6306,
+      "step": 530
+    },
+    {
+      "epoch": 0.14520098441345364,
+      "grad_norm": 0.14431102573871613,
+      "learning_rate": 5e-05,
+      "loss": 1.5672,
+      "step": 531
+    },
+    {
+      "epoch": 0.14547443259502324,
+      "grad_norm": 0.1534319818019867,
+      "learning_rate": 5e-05,
+      "loss": 1.7233,
+      "step": 532
+    },
+    {
+      "epoch": 0.14574788077659284,
+      "grad_norm": 0.15924040973186493,
+      "learning_rate": 5e-05,
+      "loss": 1.6973,
+      "step": 533
+    },
+    {
+      "epoch": 0.14602132895816242,
+      "grad_norm": 0.14544963836669922,
+      "learning_rate": 5e-05,
+      "loss": 1.6773,
+      "step": 534
+    },
+    {
+      "epoch": 0.14629477713973202,
+      "grad_norm": 0.1628415733575821,
+      "learning_rate": 5e-05,
+      "loss": 1.6427,
+      "step": 535
+    },
+    {
+      "epoch": 0.14656822532130162,
+      "grad_norm": 0.1434401571750641,
+      "learning_rate": 5e-05,
+      "loss": 1.6043,
+      "step": 536
+    },
+    {
+      "epoch": 0.1468416735028712,
+      "grad_norm": 0.17323043942451477,
+      "learning_rate": 5e-05,
+      "loss": 1.7487,
+      "step": 537
+    },
+    {
+      "epoch": 0.1471151216844408,
+      "grad_norm": 0.15013372898101807,
+      "learning_rate": 5e-05,
+      "loss": 1.6936,
+      "step": 538
+    },
+    {
+      "epoch": 0.1473885698660104,
+      "grad_norm": 0.14302760362625122,
+      "learning_rate": 5e-05,
+      "loss": 1.6237,
+      "step": 539
+    },
+    {
+      "epoch": 0.14766201804757997,
+      "grad_norm": 0.16000758111476898,
+      "learning_rate": 5e-05,
+      "loss": 1.7007,
+      "step": 540
+    },
+    {
+      "epoch": 0.14793546622914958,
+      "grad_norm": 0.1614782214164734,
+      "learning_rate": 5e-05,
+      "loss": 1.794,
+      "step": 541
+    },
+    {
+      "epoch": 0.14820891441071918,
+      "grad_norm": 0.17829090356826782,
+      "learning_rate": 5e-05,
+      "loss": 1.7155,
+      "step": 542
+    },
+    {
+      "epoch": 0.14848236259228875,
+      "grad_norm": 0.16465310752391815,
+      "learning_rate": 5e-05,
+      "loss": 1.6136,
+      "step": 543
+    },
+    {
+      "epoch": 0.14875581077385835,
+      "grad_norm": 0.17050433158874512,
+      "learning_rate": 5e-05,
+      "loss": 1.6685,
+      "step": 544
+    },
+    {
+      "epoch": 0.14902925895542796,
+      "grad_norm": 0.1535949409008026,
+      "learning_rate": 5e-05,
+      "loss": 1.6712,
+      "step": 545
+    },
+    {
+      "epoch": 0.14930270713699753,
+      "grad_norm": 0.1774369478225708,
+      "learning_rate": 5e-05,
+      "loss": 1.6754,
+      "step": 546
+    },
+    {
+      "epoch": 0.14957615531856713,
+      "grad_norm": 0.14995476603507996,
+      "learning_rate": 5e-05,
+      "loss": 1.6214,
+      "step": 547
+    },
+    {
+      "epoch": 0.14984960350013674,
+      "grad_norm": 0.16445566713809967,
+      "learning_rate": 5e-05,
+      "loss": 1.6705,
+      "step": 548
+    },
+    {
+      "epoch": 0.1501230516817063,
+      "grad_norm": 0.1804472953081131,
+      "learning_rate": 5e-05,
+      "loss": 1.7822,
+      "step": 549
+    },
+    {
+      "epoch": 0.1503964998632759,
+      "grad_norm": 0.15526773035526276,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 550
+    },
+    {
+      "epoch": 0.15066994804484551,
+      "grad_norm": 0.1779652237892151,
+      "learning_rate": 5e-05,
+      "loss": 1.6667,
+      "step": 551
+    },
+    {
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.15680311620235443,
+      "learning_rate": 5e-05,
+      "loss": 1.663,
+      "step": 552
+    },
+    {
+      "epoch": 0.1512168444079847,
+      "grad_norm": 0.15443289279937744,
+      "learning_rate": 5e-05,
+      "loss": 1.65,
+      "step": 553
+    },
+    {
+      "epoch": 0.1514902925895543,
+      "grad_norm": 0.16194204986095428,
+      "learning_rate": 5e-05,
+      "loss": 1.7575,
+      "step": 554
+    },
+    {
+      "epoch": 0.15176374077112387,
+      "grad_norm": 0.1542530059814453,
+      "learning_rate": 5e-05,
+      "loss": 1.6408,
+      "step": 555
+    },
+    {
+      "epoch": 0.15203718895269347,
+      "grad_norm": 0.14716754853725433,
+      "learning_rate": 5e-05,
+      "loss": 1.6898,
+      "step": 556
+    },
+    {
+      "epoch": 0.15231063713426304,
+      "grad_norm": 0.15577493607997894,
+      "learning_rate": 5e-05,
+      "loss": 1.6614,
+      "step": 557
+    },
+    {
+      "epoch": 0.15258408531583265,
+      "grad_norm": 0.14232535660266876,
+      "learning_rate": 5e-05,
+      "loss": 1.6025,
+      "step": 558
+    },
+    {
+      "epoch": 0.15285753349740225,
+      "grad_norm": 0.15629936754703522,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 559
+    },
+    {
+      "epoch": 0.15313098167897182,
+      "grad_norm": 0.15671175718307495,
+      "learning_rate": 5e-05,
+      "loss": 1.6616,
+      "step": 560
+    },
+    {
+      "epoch": 0.15340442986054142,
+      "grad_norm": 0.15339986979961395,
+      "learning_rate": 5e-05,
+      "loss": 1.7227,
+      "step": 561
+    },
+    {
+      "epoch": 0.15367787804211103,
+      "grad_norm": 0.14904935657978058,
+      "learning_rate": 5e-05,
+      "loss": 1.7057,
+      "step": 562
+    },
+    {
+      "epoch": 0.1539513262236806,
+      "grad_norm": 0.16447263956069946,
+      "learning_rate": 5e-05,
+      "loss": 1.7246,
+      "step": 563
+    },
+    {
+      "epoch": 0.1542247744052502,
+      "grad_norm": 0.15060292184352875,
+      "learning_rate": 5e-05,
+      "loss": 1.731,
+      "step": 564
+    },
+    {
+      "epoch": 0.1544982225868198,
+      "grad_norm": 0.15564222633838654,
+      "learning_rate": 5e-05,
+      "loss": 1.662,
+      "step": 565
+    },
+    {
+      "epoch": 0.15477167076838938,
+      "grad_norm": 0.1540018767118454,
+      "learning_rate": 5e-05,
+      "loss": 1.6793,
+      "step": 566
+    },
+    {
+      "epoch": 0.15504511894995898,
+      "grad_norm": 0.14922669529914856,
+      "learning_rate": 5e-05,
+      "loss": 1.6619,
+      "step": 567
+    },
+    {
+      "epoch": 0.15531856713152858,
+      "grad_norm": 0.1617313027381897,
+      "learning_rate": 5e-05,
+      "loss": 1.6571,
+      "step": 568
+    },
+    {
+      "epoch": 0.15559201531309816,
+      "grad_norm": 0.15420418977737427,
+      "learning_rate": 5e-05,
+      "loss": 1.6897,
+      "step": 569
+    },
+    {
+      "epoch": 0.15586546349466776,
+      "grad_norm": 0.16023240983486176,
+      "learning_rate": 5e-05,
+      "loss": 1.7376,
+      "step": 570
+    },
+    {
+      "epoch": 0.15613891167623736,
+      "grad_norm": 0.17280931770801544,
+      "learning_rate": 5e-05,
+      "loss": 1.7339,
+      "step": 571
+    },
+    {
+      "epoch": 0.15641235985780694,
+      "grad_norm": 0.15177179872989655,
+      "learning_rate": 5e-05,
+      "loss": 1.7808,
+      "step": 572
+    },
+    {
+      "epoch": 0.15668580803937654,
+      "grad_norm": 0.14648132026195526,
+      "learning_rate": 5e-05,
+      "loss": 1.6798,
+      "step": 573
+    },
+    {
+      "epoch": 0.15695925622094614,
+      "grad_norm": 0.1551511436700821,
+      "learning_rate": 5e-05,
+      "loss": 1.8416,
+      "step": 574
+    },
+    {
+      "epoch": 0.15723270440251572,
+      "grad_norm": 0.14707240462303162,
+      "learning_rate": 5e-05,
+      "loss": 1.6569,
+      "step": 575
+    },
+    {
+      "epoch": 0.15750615258408532,
+      "grad_norm": 0.15362726151943207,
+      "learning_rate": 5e-05,
+      "loss": 1.6738,
+      "step": 576
+    },
+    {
+      "epoch": 0.15777960076565492,
+      "grad_norm": 0.1611640602350235,
+      "learning_rate": 5e-05,
+      "loss": 1.6615,
+      "step": 577
+    },
+    {
+      "epoch": 0.1580530489472245,
+      "grad_norm": 0.14445427060127258,
+      "learning_rate": 5e-05,
+      "loss": 1.6479,
+      "step": 578
+    },
+    {
+      "epoch": 0.1583264971287941,
+      "grad_norm": 0.1491660177707672,
+      "learning_rate": 5e-05,
+      "loss": 1.6616,
+      "step": 579
+    },
+    {
+      "epoch": 0.1585999453103637,
+      "grad_norm": 0.16628406941890717,
+      "learning_rate": 5e-05,
+      "loss": 1.7884,
+      "step": 580
+    },
+    {
+      "epoch": 0.15887339349193327,
+      "grad_norm": 0.1535821259021759,
+      "learning_rate": 5e-05,
+      "loss": 1.7276,
+      "step": 581
+    },
+    {
+      "epoch": 0.15914684167350288,
+      "grad_norm": 0.14763818681240082,
+      "learning_rate": 5e-05,
+      "loss": 1.526,
+      "step": 582
+    },
+    {
+      "epoch": 0.15942028985507245,
+      "grad_norm": 0.15529736876487732,
+      "learning_rate": 5e-05,
+      "loss": 1.7708,
+      "step": 583
+    },
+    {
+      "epoch": 0.15969373803664205,
+      "grad_norm": 0.15036877989768982,
+      "learning_rate": 5e-05,
+      "loss": 1.7016,
+      "step": 584
+    },
+    {
+      "epoch": 0.15996718621821165,
+      "grad_norm": 0.15483739972114563,
+      "learning_rate": 5e-05,
+      "loss": 1.5965,
+      "step": 585
+    },
+    {
+      "epoch": 0.16024063439978123,
+      "grad_norm": 0.15531662106513977,
+      "learning_rate": 5e-05,
+      "loss": 1.6769,
+      "step": 586
+    },
+    {
+      "epoch": 0.16051408258135083,
+      "grad_norm": 0.1560511589050293,
+      "learning_rate": 5e-05,
+      "loss": 1.7538,
+      "step": 587
+    },
+    {
+      "epoch": 0.16078753076292043,
+      "grad_norm": 0.1831214427947998,
+      "learning_rate": 5e-05,
+      "loss": 1.7097,
+      "step": 588
+    },
+    {
+      "epoch": 0.16106097894449,
+      "grad_norm": 0.15891355276107788,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 589
+    },
+    {
+      "epoch": 0.1613344271260596,
+      "grad_norm": 0.1625254601240158,
+      "learning_rate": 5e-05,
+      "loss": 1.644,
+      "step": 590
+    },
+    {
+      "epoch": 0.1616078753076292,
+      "grad_norm": 0.15376971662044525,
+      "learning_rate": 5e-05,
+      "loss": 1.701,
+      "step": 591
+    },
+    {
+      "epoch": 0.1618813234891988,
+      "grad_norm": 0.14270102977752686,
+      "learning_rate": 5e-05,
+      "loss": 1.5089,
+      "step": 592
+    },
+    {
+      "epoch": 0.1621547716707684,
+      "grad_norm": 0.14729395508766174,
+      "learning_rate": 5e-05,
+      "loss": 1.6989,
+      "step": 593
+    },
+    {
+      "epoch": 0.162428219852338,
+      "grad_norm": 0.1675315499305725,
+      "learning_rate": 5e-05,
+      "loss": 1.7195,
+      "step": 594
+    },
+    {
+      "epoch": 0.16270166803390756,
+      "grad_norm": 0.14959470927715302,
+      "learning_rate": 5e-05,
+      "loss": 1.678,
+      "step": 595
+    },
+    {
+      "epoch": 0.16297511621547717,
+      "grad_norm": 0.15854990482330322,
+      "learning_rate": 5e-05,
+      "loss": 1.6452,
+      "step": 596
+    },
+    {
+      "epoch": 0.16324856439704677,
+      "grad_norm": 0.151190385222435,
+      "learning_rate": 5e-05,
+      "loss": 1.6751,
+      "step": 597
+    },
+    {
+      "epoch": 0.16352201257861634,
+      "grad_norm": 0.1632450520992279,
+      "learning_rate": 5e-05,
+      "loss": 1.662,
+      "step": 598
+    },
+    {
+      "epoch": 0.16379546076018595,
+      "grad_norm": 0.1520984023809433,
+      "learning_rate": 5e-05,
+      "loss": 1.7377,
+      "step": 599
+    },
+    {
+      "epoch": 0.16406890894175555,
+      "grad_norm": 0.14863604307174683,
+      "learning_rate": 5e-05,
+      "loss": 1.5927,
+      "step": 600
+    },
+    {
+      "epoch": 0.16434235712332512,
+      "grad_norm": 0.15424251556396484,
+      "learning_rate": 5e-05,
+      "loss": 1.6265,
+      "step": 601
+    },
+    {
+      "epoch": 0.16461580530489472,
+      "grad_norm": 0.15759313106536865,
+      "learning_rate": 5e-05,
+      "loss": 1.6659,
+      "step": 602
+    },
+    {
+      "epoch": 0.16488925348646433,
+      "grad_norm": 0.1515471190214157,
+      "learning_rate": 5e-05,
+      "loss": 1.6585,
+      "step": 603
+    },
+    {
+      "epoch": 0.1651627016680339,
+      "grad_norm": 0.16653236746788025,
+      "learning_rate": 5e-05,
+      "loss": 1.6593,
+      "step": 604
+    },
+    {
+      "epoch": 0.1654361498496035,
+      "grad_norm": 0.15483258664608002,
+      "learning_rate": 5e-05,
+      "loss": 1.7138,
+      "step": 605
+    },
+    {
+      "epoch": 0.1657095980311731,
+      "grad_norm": 0.14448527991771698,
+      "learning_rate": 5e-05,
+      "loss": 1.6627,
+      "step": 606
+    },
+    {
+      "epoch": 0.16598304621274268,
+      "grad_norm": 0.15589672327041626,
+      "learning_rate": 5e-05,
+      "loss": 1.5996,
+      "step": 607
+    },
+    {
+      "epoch": 0.16625649439431228,
+      "grad_norm": 0.16023333370685577,
+      "learning_rate": 5e-05,
+      "loss": 1.6351,
+      "step": 608
+    },
+    {
+      "epoch": 0.16652994257588188,
+      "grad_norm": 0.16108067333698273,
+      "learning_rate": 5e-05,
+      "loss": 1.716,
+      "step": 609
+    },
+    {
+      "epoch": 0.16680339075745146,
+      "grad_norm": 0.14648942649364471,
+      "learning_rate": 5e-05,
+      "loss": 1.662,
+      "step": 610
+    },
+    {
+      "epoch": 0.16707683893902106,
+      "grad_norm": 0.14486436545848846,
+      "learning_rate": 5e-05,
+      "loss": 1.6219,
+      "step": 611
+    },
+    {
+      "epoch": 0.16735028712059064,
+      "grad_norm": 0.15563012659549713,
+      "learning_rate": 5e-05,
+      "loss": 1.6803,
+      "step": 612
+    },
+    {
+      "epoch": 0.16762373530216024,
+      "grad_norm": 0.1571030616760254,
+      "learning_rate": 5e-05,
+      "loss": 1.7195,
+      "step": 613
+    },
+    {
+      "epoch": 0.16789718348372984,
+      "grad_norm": 0.14509615302085876,
+      "learning_rate": 5e-05,
+      "loss": 1.6462,
+      "step": 614
+    },
+    {
+      "epoch": 0.1681706316652994,
+      "grad_norm": 0.15147417783737183,
+      "learning_rate": 5e-05,
+      "loss": 1.7463,
+      "step": 615
+    },
+    {
+      "epoch": 0.16844407984686902,
+      "grad_norm": 0.1549796462059021,
+      "learning_rate": 5e-05,
+      "loss": 1.6997,
+      "step": 616
+    },
+    {
+      "epoch": 0.16871752802843862,
+      "grad_norm": 0.16344091296195984,
+      "learning_rate": 5e-05,
+      "loss": 1.6065,
+      "step": 617
+    },
+    {
+      "epoch": 0.1689909762100082,
+      "grad_norm": 0.1531122624874115,
+      "learning_rate": 5e-05,
+      "loss": 1.5878,
+      "step": 618
+    },
+    {
+      "epoch": 0.1692644243915778,
+      "grad_norm": 0.17184039950370789,
+      "learning_rate": 5e-05,
+      "loss": 1.6921,
+      "step": 619
+    },
+    {
+      "epoch": 0.1695378725731474,
+      "grad_norm": 0.16684702038764954,
+      "learning_rate": 5e-05,
+      "loss": 1.7952,
+      "step": 620
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.1621457189321518,
+      "learning_rate": 5e-05,
+      "loss": 1.6935,
+      "step": 621
+    },
+    {
+      "epoch": 0.17008476893628657,
+      "grad_norm": 0.15157224237918854,
+      "learning_rate": 5e-05,
+      "loss": 1.747,
+      "step": 622
+    },
+    {
+      "epoch": 0.17035821711785618,
+      "grad_norm": 0.1498030573129654,
+      "learning_rate": 5e-05,
+      "loss": 1.6138,
+      "step": 623
+    },
+    {
+      "epoch": 0.17063166529942575,
+      "grad_norm": 0.15279249846935272,
+      "learning_rate": 5e-05,
+      "loss": 1.6191,
+      "step": 624
+    },
+    {
+      "epoch": 0.17090511348099535,
+      "grad_norm": 0.15653859078884125,
+      "learning_rate": 5e-05,
+      "loss": 1.6271,
+      "step": 625
+    },
+    {
+      "epoch": 0.17117856166256495,
+      "grad_norm": 0.17365135252475739,
+      "learning_rate": 5e-05,
+      "loss": 1.733,
+      "step": 626
+    },
+    {
+      "epoch": 0.17145200984413453,
+      "grad_norm": 0.14621610939502716,
+      "learning_rate": 5e-05,
+      "loss": 1.5571,
+      "step": 627
+    },
+    {
+      "epoch": 0.17172545802570413,
+      "grad_norm": 0.1693270057439804,
+      "learning_rate": 5e-05,
+      "loss": 1.7436,
+      "step": 628
+    },
+    {
+      "epoch": 0.17199890620727373,
+      "grad_norm": 0.15240801870822906,
+      "learning_rate": 5e-05,
+      "loss": 1.7051,
+      "step": 629
+    },
+    {
+      "epoch": 0.1722723543888433,
+      "grad_norm": 0.14783060550689697,
+      "learning_rate": 5e-05,
+      "loss": 1.6765,
+      "step": 630
+    },
+    {
+      "epoch": 0.1725458025704129,
+      "grad_norm": 0.15038736164569855,
+      "learning_rate": 5e-05,
+      "loss": 1.5878,
+      "step": 631
+    },
+    {
+      "epoch": 0.1728192507519825,
+      "grad_norm": 0.16009370982646942,
+      "learning_rate": 5e-05,
+      "loss": 1.7588,
+      "step": 632
+    },
+    {
+      "epoch": 0.17309269893355209,
+      "grad_norm": 0.15497462451457977,
+      "learning_rate": 5e-05,
+      "loss": 1.6953,
+      "step": 633
+    },
+    {
+      "epoch": 0.1733661471151217,
+      "grad_norm": 0.18641214072704315,
+      "learning_rate": 5e-05,
+      "loss": 1.6218,
+      "step": 634
+    },
+    {
+      "epoch": 0.1736395952966913,
+      "grad_norm": 0.1536916345357895,
+      "learning_rate": 5e-05,
+      "loss": 1.6898,
+      "step": 635
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.15958872437477112,
+      "learning_rate": 5e-05,
+      "loss": 1.6823,
+      "step": 636
+    },
+    {
+      "epoch": 0.17418649165983047,
+      "grad_norm": 0.16902394592761993,
+      "learning_rate": 5e-05,
+      "loss": 1.7029,
+      "step": 637
+    },
+    {
+      "epoch": 0.17445993984140004,
+      "grad_norm": 0.1553962379693985,
+      "learning_rate": 5e-05,
+      "loss": 1.7663,
+      "step": 638
+    },
+    {
+      "epoch": 0.17473338802296964,
+      "grad_norm": 0.1627410352230072,
+      "learning_rate": 5e-05,
+      "loss": 1.5562,
+      "step": 639
+    },
+    {
+      "epoch": 0.17500683620453925,
+      "grad_norm": 0.16163355112075806,
+      "learning_rate": 5e-05,
+      "loss": 1.6688,
+      "step": 640
+    },
+    {
+      "epoch": 0.17528028438610882,
+      "grad_norm": 0.14881187677383423,
+      "learning_rate": 5e-05,
+      "loss": 1.6582,
+      "step": 641
+    },
+    {
+      "epoch": 0.17555373256767842,
+      "grad_norm": 0.1964666247367859,
+      "learning_rate": 5e-05,
+      "loss": 1.6861,
+      "step": 642
+    },
+    {
+      "epoch": 0.17582718074924802,
+      "grad_norm": 0.15720294415950775,
+      "learning_rate": 5e-05,
+      "loss": 1.6249,
+      "step": 643
+    },
+    {
+      "epoch": 0.1761006289308176,
+      "grad_norm": 0.15254618227481842,
+      "learning_rate": 5e-05,
+      "loss": 1.6902,
+      "step": 644
+    },
+    {
+      "epoch": 0.1763740771123872,
+      "grad_norm": 0.19420553743839264,
+      "learning_rate": 5e-05,
+      "loss": 1.7313,
+      "step": 645
+    },
+    {
+      "epoch": 0.1766475252939568,
+      "grad_norm": 0.1590331643819809,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 646
+    },
+    {
+      "epoch": 0.17692097347552638,
+      "grad_norm": 0.18741661310195923,
+      "learning_rate": 5e-05,
+      "loss": 1.6959,
+      "step": 647
+    },
+    {
+      "epoch": 0.17719442165709598,
+      "grad_norm": 0.16829299926757812,
+      "learning_rate": 5e-05,
+      "loss": 1.7166,
+      "step": 648
+    },
+    {
+      "epoch": 0.17746786983866558,
+      "grad_norm": 0.16889558732509613,
+      "learning_rate": 5e-05,
+      "loss": 1.6289,
+      "step": 649
+    },
+    {
+      "epoch": 0.17774131802023516,
+      "grad_norm": 0.18458721041679382,
+      "learning_rate": 5e-05,
+      "loss": 1.6351,
+      "step": 650
+    },
+    {
+      "epoch": 0.17801476620180476,
+      "grad_norm": 0.1489873081445694,
+      "learning_rate": 5e-05,
+      "loss": 1.6426,
+      "step": 651
+    },
+    {
+      "epoch": 0.17828821438337436,
+      "grad_norm": 0.1619219332933426,
+      "learning_rate": 5e-05,
+      "loss": 1.6849,
+      "step": 652
+    },
+    {
+      "epoch": 0.17856166256494393,
+      "grad_norm": 0.15107835829257965,
+      "learning_rate": 5e-05,
+      "loss": 1.6455,
+      "step": 653
+    },
+    {
+      "epoch": 0.17883511074651354,
+      "grad_norm": 0.15547437965869904,
+      "learning_rate": 5e-05,
+      "loss": 1.7137,
+      "step": 654
+    },
+    {
+      "epoch": 0.17910855892808314,
+      "grad_norm": 0.18254241347312927,
+      "learning_rate": 5e-05,
+      "loss": 1.7565,
+      "step": 655
+    },
+    {
+      "epoch": 0.1793820071096527,
+      "grad_norm": 0.15430989861488342,
+      "learning_rate": 5e-05,
+      "loss": 1.7157,
+      "step": 656
+    },
+    {
+      "epoch": 0.17965545529122232,
+      "grad_norm": 0.14642038941383362,
+      "learning_rate": 5e-05,
+      "loss": 1.6962,
+      "step": 657
+    },
+    {
+      "epoch": 0.17992890347279192,
+      "grad_norm": 0.15086182951927185,
+      "learning_rate": 5e-05,
+      "loss": 1.6187,
+      "step": 658
+    },
+    {
+      "epoch": 0.1802023516543615,
+      "grad_norm": 0.1483525186777115,
+      "learning_rate": 5e-05,
+      "loss": 1.6572,
+      "step": 659
+    },
+    {
+      "epoch": 0.1804757998359311,
+      "grad_norm": 0.1709260642528534,
+      "learning_rate": 5e-05,
+      "loss": 1.6813,
+      "step": 660
+    },
+    {
+      "epoch": 0.1807492480175007,
+      "grad_norm": 0.17065215110778809,
+      "learning_rate": 5e-05,
+      "loss": 1.7901,
+      "step": 661
+    },
+    {
+      "epoch": 0.18102269619907027,
+      "grad_norm": 0.15385740995407104,
+      "learning_rate": 5e-05,
+      "loss": 1.6925,
+      "step": 662
+    },
+    {
+      "epoch": 0.18129614438063987,
+      "grad_norm": 0.16739514470100403,
+      "learning_rate": 5e-05,
+      "loss": 1.6025,
+      "step": 663
+    },
+    {
+      "epoch": 0.18156959256220945,
+      "grad_norm": 0.16431251168251038,
+      "learning_rate": 5e-05,
+      "loss": 1.6316,
+      "step": 664
+    },
+    {
+      "epoch": 0.18184304074377905,
+      "grad_norm": 0.15166325867176056,
+      "learning_rate": 5e-05,
+      "loss": 1.7378,
+      "step": 665
+    },
+    {
+      "epoch": 0.18211648892534865,
+      "grad_norm": 0.1769624650478363,
+      "learning_rate": 5e-05,
+      "loss": 1.6963,
+      "step": 666
+    },
+    {
+      "epoch": 0.18238993710691823,
+      "grad_norm": 0.17551544308662415,
+      "learning_rate": 5e-05,
+      "loss": 1.6722,
+      "step": 667
+    },
+    {
+      "epoch": 0.18266338528848783,
+      "grad_norm": 0.1519336998462677,
+      "learning_rate": 5e-05,
+      "loss": 1.7812,
+      "step": 668
+    },
+    {
+      "epoch": 0.18293683347005743,
+      "grad_norm": 0.1523965448141098,
+      "learning_rate": 5e-05,
+      "loss": 1.6374,
+      "step": 669
+    },
+    {
+      "epoch": 0.183210281651627,
+      "grad_norm": 0.16304127871990204,
+      "learning_rate": 5e-05,
+      "loss": 1.6436,
+      "step": 670
+    },
+    {
+      "epoch": 0.1834837298331966,
+      "grad_norm": 0.1389431655406952,
+      "learning_rate": 5e-05,
+      "loss": 1.6136,
+      "step": 671
+    },
+    {
+      "epoch": 0.1837571780147662,
+      "grad_norm": 0.15939609706401825,
+      "learning_rate": 5e-05,
+      "loss": 1.6407,
+      "step": 672
+    },
+    {
+      "epoch": 0.18403062619633578,
+      "grad_norm": 0.15122279524803162,
+      "learning_rate": 5e-05,
+      "loss": 1.605,
+      "step": 673
+    },
+    {
+      "epoch": 0.18430407437790539,
+      "grad_norm": 0.14799101650714874,
+      "learning_rate": 5e-05,
+      "loss": 1.6199,
+      "step": 674
+    },
+    {
+      "epoch": 0.184577522559475,
+      "grad_norm": 0.17853890359401703,
+      "learning_rate": 5e-05,
+      "loss": 1.7246,
+      "step": 675
+    },
+    {
+      "epoch": 0.18485097074104456,
+      "grad_norm": 0.14898012578487396,
+      "learning_rate": 5e-05,
+      "loss": 1.7212,
+      "step": 676
+    },
+    {
+      "epoch": 0.18512441892261416,
+      "grad_norm": 0.1488298773765564,
+      "learning_rate": 5e-05,
+      "loss": 1.6592,
+      "step": 677
+    },
+    {
+      "epoch": 0.18539786710418377,
+      "grad_norm": 0.14956548810005188,
+      "learning_rate": 5e-05,
+      "loss": 1.6025,
+      "step": 678
+    },
+    {
+      "epoch": 0.18567131528575334,
+      "grad_norm": 0.14764147996902466,
+      "learning_rate": 5e-05,
+      "loss": 1.6695,
+      "step": 679
+    },
+    {
+      "epoch": 0.18594476346732294,
+      "grad_norm": 0.16988615691661835,
+      "learning_rate": 5e-05,
+      "loss": 1.7755,
+      "step": 680
+    },
+    {
+      "epoch": 0.18621821164889255,
+      "grad_norm": 0.1601964235305786,
+      "learning_rate": 5e-05,
+      "loss": 1.7565,
+      "step": 681
+    },
+    {
+      "epoch": 0.18649165983046212,
+      "grad_norm": 0.17461629211902618,
+      "learning_rate": 5e-05,
+      "loss": 1.7552,
+      "step": 682
+    },
+    {
+      "epoch": 0.18676510801203172,
+      "grad_norm": 0.15619756281375885,
+      "learning_rate": 5e-05,
+      "loss": 1.7162,
+      "step": 683
+    },
+    {
+      "epoch": 0.18703855619360132,
+      "grad_norm": 0.15257790684700012,
+      "learning_rate": 5e-05,
+      "loss": 1.7154,
+      "step": 684
+    },
+    {
+      "epoch": 0.1873120043751709,
+      "grad_norm": 0.15032611787319183,
+      "learning_rate": 5e-05,
+      "loss": 1.5825,
+      "step": 685
+    },
+    {
+      "epoch": 0.1875854525567405,
+      "grad_norm": 0.1578359454870224,
+      "learning_rate": 5e-05,
+      "loss": 1.706,
+      "step": 686
+    },
+    {
+      "epoch": 0.1878589007383101,
+      "grad_norm": 0.16593731939792633,
+      "learning_rate": 5e-05,
+      "loss": 1.624,
+      "step": 687
+    },
+    {
+      "epoch": 0.18813234891987968,
+      "grad_norm": 0.17991593480110168,
+      "learning_rate": 5e-05,
+      "loss": 1.7293,
+      "step": 688
+    },
+    {
+      "epoch": 0.18840579710144928,
+      "grad_norm": 0.15078522264957428,
+      "learning_rate": 5e-05,
+      "loss": 1.6419,
+      "step": 689
+    },
+    {
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.20055970549583435,
+      "learning_rate": 5e-05,
+      "loss": 1.7743,
+      "step": 690
+    },
+    {
+      "epoch": 0.18895269346458846,
+      "grad_norm": 0.14143189787864685,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 691
+    },
+    {
+      "epoch": 0.18922614164615806,
+      "grad_norm": 0.17427237331867218,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 692
+    },
+    {
+      "epoch": 0.18949958982772763,
+      "grad_norm": 0.18519847095012665,
+      "learning_rate": 5e-05,
+      "loss": 1.7039,
+      "step": 693
+    },
+    {
+      "epoch": 0.18977303800929723,
+      "grad_norm": 0.15196627378463745,
+      "learning_rate": 5e-05,
+      "loss": 1.7847,
+      "step": 694
+    },
+    {
+      "epoch": 0.19004648619086684,
+      "grad_norm": 0.17246317863464355,
+      "learning_rate": 5e-05,
+      "loss": 1.7097,
+      "step": 695
+    },
+    {
+      "epoch": 0.1903199343724364,
+      "grad_norm": 0.16350534558296204,
+      "learning_rate": 5e-05,
+      "loss": 1.6783,
+      "step": 696
+    },
+    {
+      "epoch": 0.190593382554006,
+      "grad_norm": 0.15077216923236847,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 697
+    },
+    {
+      "epoch": 0.19086683073557562,
+      "grad_norm": 0.17161305248737335,
+      "learning_rate": 5e-05,
+      "loss": 1.757,
+      "step": 698
+    },
+    {
+      "epoch": 0.1911402789171452,
+      "grad_norm": 0.15666204690933228,
+      "learning_rate": 5e-05,
+      "loss": 1.6094,
+      "step": 699
+    },
+    {
+      "epoch": 0.1914137270987148,
+      "grad_norm": 0.15116041898727417,
+      "learning_rate": 5e-05,
+      "loss": 1.7604,
+      "step": 700
+    },
+    {
+      "epoch": 0.1916871752802844,
+      "grad_norm": 0.17067068815231323,
+      "learning_rate": 5e-05,
+      "loss": 1.6355,
+      "step": 701
+    },
+    {
+      "epoch": 0.19196062346185397,
+      "grad_norm": 0.1541707068681717,
+      "learning_rate": 5e-05,
+      "loss": 1.7279,
+      "step": 702
+    },
+    {
+      "epoch": 0.19223407164342357,
+      "grad_norm": 0.154357448220253,
+      "learning_rate": 5e-05,
+      "loss": 1.5981,
+      "step": 703
+    },
+    {
+      "epoch": 0.19250751982499317,
+      "grad_norm": 0.1600044220685959,
+      "learning_rate": 5e-05,
+      "loss": 1.7863,
+      "step": 704
+    },
+    {
+      "epoch": 0.19278096800656275,
+      "grad_norm": 0.15702758729457855,
+      "learning_rate": 5e-05,
+      "loss": 1.752,
+      "step": 705
+    },
+    {
+      "epoch": 0.19305441618813235,
+      "grad_norm": 0.15750904381275177,
+      "learning_rate": 5e-05,
+      "loss": 1.6763,
+      "step": 706
+    },
+    {
+      "epoch": 0.19332786436970195,
+      "grad_norm": 0.14942318201065063,
+      "learning_rate": 5e-05,
+      "loss": 1.5828,
+      "step": 707
+    },
+    {
+      "epoch": 0.19360131255127153,
+      "grad_norm": 0.14567220211029053,
+      "learning_rate": 5e-05,
+      "loss": 1.647,
+      "step": 708
+    },
+    {
+      "epoch": 0.19387476073284113,
+      "grad_norm": 0.15003998577594757,
+      "learning_rate": 5e-05,
+      "loss": 1.6545,
+      "step": 709
+    },
+    {
+      "epoch": 0.19414820891441073,
+      "grad_norm": 0.1552526205778122,
+      "learning_rate": 5e-05,
+      "loss": 1.7039,
+      "step": 710
+    },
+    {
+      "epoch": 0.1944216570959803,
+      "grad_norm": 0.16224539279937744,
+      "learning_rate": 5e-05,
+      "loss": 1.6792,
+      "step": 711
+    },
+    {
+      "epoch": 0.1946951052775499,
+      "grad_norm": 0.15637609362602234,
+      "learning_rate": 5e-05,
+      "loss": 1.7131,
+      "step": 712
+    },
+    {
+      "epoch": 0.1949685534591195,
+      "grad_norm": 0.14550897479057312,
+      "learning_rate": 5e-05,
+      "loss": 1.6662,
+      "step": 713
+    },
+    {
+      "epoch": 0.19524200164068908,
+      "grad_norm": 0.14311423897743225,
+      "learning_rate": 5e-05,
+      "loss": 1.5609,
+      "step": 714
+    },
+    {
+      "epoch": 0.19551544982225869,
+      "grad_norm": 0.1632792055606842,
+      "learning_rate": 5e-05,
+      "loss": 1.716,
+      "step": 715
+    },
+    {
+      "epoch": 0.1957888980038283,
+      "grad_norm": 0.14958421885967255,
+      "learning_rate": 5e-05,
+      "loss": 1.6868,
+      "step": 716
+    },
+    {
+      "epoch": 0.19606234618539786,
+      "grad_norm": 0.15640190243721008,
+      "learning_rate": 5e-05,
+      "loss": 1.6763,
+      "step": 717
+    },
+    {
+      "epoch": 0.19633579436696746,
+      "grad_norm": 0.1410387009382248,
+      "learning_rate": 5e-05,
+      "loss": 1.6593,
+      "step": 718
+    },
+    {
+      "epoch": 0.19660924254853704,
+      "grad_norm": 0.14812639355659485,
+      "learning_rate": 5e-05,
+      "loss": 1.6907,
+      "step": 719
+    },
+    {
+      "epoch": 0.19688269073010664,
+      "grad_norm": 0.15623793005943298,
+      "learning_rate": 5e-05,
+      "loss": 1.6879,
+      "step": 720
+    },
+    {
+      "epoch": 0.19715613891167624,
+      "grad_norm": 0.1545989066362381,
+      "learning_rate": 5e-05,
+      "loss": 1.7276,
+      "step": 721
+    },
+    {
+      "epoch": 0.19742958709324582,
+      "grad_norm": 0.15051992237567902,
+      "learning_rate": 5e-05,
+      "loss": 1.7184,
+      "step": 722
+    },
+    {
+      "epoch": 0.19770303527481542,
+      "grad_norm": 0.14712779223918915,
+      "learning_rate": 5e-05,
+      "loss": 1.6231,
+      "step": 723
+    },
+    {
+      "epoch": 0.19797648345638502,
+      "grad_norm": 0.15088264644145966,
+      "learning_rate": 5e-05,
+      "loss": 1.662,
+      "step": 724
+    },
+    {
+      "epoch": 0.1982499316379546,
+      "grad_norm": 0.14665238559246063,
+      "learning_rate": 5e-05,
+      "loss": 1.6626,
+      "step": 725
+    },
+    {
+      "epoch": 0.1985233798195242,
+      "grad_norm": 0.16032913327217102,
+      "learning_rate": 5e-05,
+      "loss": 1.7283,
+      "step": 726
+    },
+    {
+      "epoch": 0.1987968280010938,
+      "grad_norm": 0.15149696171283722,
+      "learning_rate": 5e-05,
+      "loss": 1.6387,
+      "step": 727
+    },
+    {
+      "epoch": 0.19907027618266337,
+      "grad_norm": 0.15610548853874207,
+      "learning_rate": 5e-05,
+      "loss": 1.7084,
+      "step": 728
+    },
+    {
+      "epoch": 0.19934372436423298,
+      "grad_norm": 0.16325855255126953,
+      "learning_rate": 5e-05,
+      "loss": 1.567,
+      "step": 729
+    },
+    {
+      "epoch": 0.19961717254580258,
+      "grad_norm": 0.1529161036014557,
+      "learning_rate": 5e-05,
+      "loss": 1.6343,
+      "step": 730
+    },
+    {
+      "epoch": 0.19989062072737215,
+      "grad_norm": 0.1644391119480133,
+      "learning_rate": 5e-05,
+      "loss": 1.6918,
+      "step": 731
+    },
+    {
+      "epoch": 0.20016406890894176,
+      "grad_norm": 0.14825589954853058,
+      "learning_rate": 5e-05,
+      "loss": 1.6954,
+      "step": 732
+    },
+    {
+      "epoch": 0.20043751709051136,
+      "grad_norm": 0.15434464812278748,
+      "learning_rate": 5e-05,
+      "loss": 1.751,
+      "step": 733
+    },
+    {
+      "epoch": 0.20071096527208093,
+      "grad_norm": 0.14505796134471893,
+      "learning_rate": 5e-05,
+      "loss": 1.6398,
+      "step": 734
+    },
+    {
+      "epoch": 0.20098441345365053,
+      "grad_norm": 0.14808180928230286,
+      "learning_rate": 5e-05,
+      "loss": 1.6475,
+      "step": 735
+    },
+    {
+      "epoch": 0.20125786163522014,
+      "grad_norm": 0.15832236409187317,
+      "learning_rate": 5e-05,
+      "loss": 1.6888,
+      "step": 736
+    },
+    {
+      "epoch": 0.2015313098167897,
+      "grad_norm": 0.15063433349132538,
+      "learning_rate": 5e-05,
+      "loss": 1.714,
+      "step": 737
+    },
+    {
+      "epoch": 0.2018047579983593,
+      "grad_norm": 0.14673678576946259,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 738
+    },
+    {
+      "epoch": 0.20207820617992892,
+      "grad_norm": 0.16134969890117645,
+      "learning_rate": 5e-05,
+      "loss": 1.618,
+      "step": 739
+    },
+    {
+      "epoch": 0.2023516543614985,
+      "grad_norm": 0.15674619376659393,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 740
+    },
+    {
+      "epoch": 0.2026251025430681,
+      "grad_norm": 0.16632983088493347,
+      "learning_rate": 5e-05,
+      "loss": 1.6608,
+      "step": 741
+    },
+    {
+      "epoch": 0.2028985507246377,
+      "grad_norm": 0.15681079030036926,
+      "learning_rate": 5e-05,
+      "loss": 1.6452,
+      "step": 742
+    },
+    {
+      "epoch": 0.20317199890620727,
+      "grad_norm": 0.17244625091552734,
+      "learning_rate": 5e-05,
+      "loss": 1.6092,
+      "step": 743
+    },
+    {
+      "epoch": 0.20344544708777687,
+      "grad_norm": 0.17212654650211334,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 744
+    },
+    {
+      "epoch": 0.20371889526934644,
+      "grad_norm": 0.14502903819084167,
+      "learning_rate": 5e-05,
+      "loss": 1.6578,
+      "step": 745
+    },
+    {
+      "epoch": 0.20399234345091605,
+      "grad_norm": 0.19574536383152008,
+      "learning_rate": 5e-05,
+      "loss": 1.7769,
+      "step": 746
+    },
+    {
+      "epoch": 0.20426579163248565,
+      "grad_norm": 0.16160033643245697,
+      "learning_rate": 5e-05,
+      "loss": 1.6885,
+      "step": 747
+    },
+    {
+      "epoch": 0.20453923981405522,
+      "grad_norm": 0.15171994268894196,
+      "learning_rate": 5e-05,
+      "loss": 1.6761,
+      "step": 748
+    },
+    {
+      "epoch": 0.20481268799562483,
+      "grad_norm": 0.16578879952430725,
+      "learning_rate": 5e-05,
+      "loss": 1.6024,
+      "step": 749
+    },
+    {
+      "epoch": 0.20508613617719443,
+      "grad_norm": 0.1527402251958847,
+      "learning_rate": 5e-05,
+      "loss": 1.6214,
+      "step": 750
+    },
+    {
+      "epoch": 0.205359584358764,
+      "grad_norm": 0.15434326231479645,
+      "learning_rate": 5e-05,
+      "loss": 1.591,
+      "step": 751
+    },
+    {
+      "epoch": 0.2056330325403336,
+      "grad_norm": 0.15459883213043213,
+      "learning_rate": 5e-05,
+      "loss": 1.604,
+      "step": 752
+    },
+    {
+      "epoch": 0.2059064807219032,
+      "grad_norm": 0.17635492980480194,
+      "learning_rate": 5e-05,
+      "loss": 1.7674,
+      "step": 753
+    },
+    {
+      "epoch": 0.20617992890347278,
+      "grad_norm": 0.1530025154352188,
+      "learning_rate": 5e-05,
+      "loss": 1.5982,
+      "step": 754
+    },
+    {
+      "epoch": 0.20645337708504238,
+      "grad_norm": 0.16364556550979614,
+      "learning_rate": 5e-05,
+      "loss": 1.5648,
+      "step": 755
+    },
+    {
+      "epoch": 0.20672682526661199,
+      "grad_norm": 0.1753033846616745,
+      "learning_rate": 5e-05,
+      "loss": 1.6603,
+      "step": 756
+    },
+    {
+      "epoch": 0.20700027344818156,
+      "grad_norm": 0.17838945984840393,
+      "learning_rate": 5e-05,
+      "loss": 1.6572,
+      "step": 757
+    },
+    {
+      "epoch": 0.20727372162975116,
+      "grad_norm": 0.20289616286754608,
+      "learning_rate": 5e-05,
+      "loss": 1.713,
+      "step": 758
+    },
+    {
+      "epoch": 0.20754716981132076,
+      "grad_norm": 0.1460665762424469,
+      "learning_rate": 5e-05,
+      "loss": 1.6701,
+      "step": 759
+    },
+    {
+      "epoch": 0.20782061799289034,
+      "grad_norm": 0.1894819289445877,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 760
+    },
+    {
+      "epoch": 0.20809406617445994,
+      "grad_norm": 0.17014984786510468,
+      "learning_rate": 5e-05,
+      "loss": 1.6229,
+      "step": 761
+    },
+    {
+      "epoch": 0.20836751435602954,
+      "grad_norm": 0.14133282005786896,
+      "learning_rate": 5e-05,
+      "loss": 1.4855,
+      "step": 762
+    },
+    {
+      "epoch": 0.20864096253759912,
+      "grad_norm": 0.17830294370651245,
+      "learning_rate": 5e-05,
+      "loss": 1.8294,
+      "step": 763
+    },
+    {
+      "epoch": 0.20891441071916872,
+      "grad_norm": 0.17372062802314758,
+      "learning_rate": 5e-05,
+      "loss": 1.6214,
+      "step": 764
+    },
+    {
+      "epoch": 0.20918785890073832,
+      "grad_norm": 0.14572595059871674,
+      "learning_rate": 5e-05,
+      "loss": 1.6391,
+      "step": 765
+    },
+    {
+      "epoch": 0.2094613070823079,
+      "grad_norm": 0.19754233956336975,
+      "learning_rate": 5e-05,
+      "loss": 1.7374,
+      "step": 766
+    },
+    {
+      "epoch": 0.2097347552638775,
+      "grad_norm": 0.17328870296478271,
+      "learning_rate": 5e-05,
+      "loss": 1.6893,
+      "step": 767
+    },
+    {
+      "epoch": 0.2100082034454471,
+      "grad_norm": 0.15320096909999847,
+      "learning_rate": 5e-05,
+      "loss": 1.6703,
+      "step": 768
+    },
+    {
+      "epoch": 0.21028165162701667,
+      "grad_norm": 0.19963279366493225,
+      "learning_rate": 5e-05,
+      "loss": 1.8128,
+      "step": 769
+    },
+    {
+      "epoch": 0.21055509980858628,
+      "grad_norm": 0.15706828236579895,
+      "learning_rate": 5e-05,
+      "loss": 1.6998,
+      "step": 770
+    },
+    {
+      "epoch": 0.21082854799015588,
+      "grad_norm": 0.18131129443645477,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 771
+    },
+    {
+      "epoch": 0.21110199617172545,
+      "grad_norm": 0.19108878076076508,
+      "learning_rate": 5e-05,
+      "loss": 1.7074,
+      "step": 772
+    },
+    {
+      "epoch": 0.21137544435329506,
+      "grad_norm": 0.14925052225589752,
+      "learning_rate": 5e-05,
+      "loss": 1.6908,
+      "step": 773
+    },
+    {
+      "epoch": 0.21164889253486463,
+      "grad_norm": 0.1783328801393509,
+      "learning_rate": 5e-05,
+      "loss": 1.7353,
+      "step": 774
+    },
+    {
+      "epoch": 0.21192234071643423,
+      "grad_norm": 0.17631973326206207,
+      "learning_rate": 5e-05,
+      "loss": 1.683,
+      "step": 775
+    },
+    {
+      "epoch": 0.21219578889800383,
+      "grad_norm": 0.15554718673229218,
+      "learning_rate": 5e-05,
+      "loss": 1.6239,
+      "step": 776
+    },
+    {
+      "epoch": 0.2124692370795734,
+      "grad_norm": 0.1897638440132141,
+      "learning_rate": 5e-05,
+      "loss": 1.713,
+      "step": 777
+    },
+    {
+      "epoch": 0.212742685261143,
+      "grad_norm": 0.14565777778625488,
+      "learning_rate": 5e-05,
+      "loss": 1.5607,
+      "step": 778
+    },
+    {
+      "epoch": 0.2130161334427126,
+      "grad_norm": 0.14770552515983582,
+      "learning_rate": 5e-05,
+      "loss": 1.6054,
+      "step": 779
+    },
+    {
+      "epoch": 0.2132895816242822,
+      "grad_norm": 0.1784149557352066,
+      "learning_rate": 5e-05,
+      "loss": 1.7054,
+      "step": 780
+    },
+    {
+      "epoch": 0.2135630298058518,
+      "grad_norm": 0.15223895013332367,
+      "learning_rate": 5e-05,
+      "loss": 1.756,
+      "step": 781
+    },
+    {
+      "epoch": 0.2138364779874214,
+      "grad_norm": 0.15329818427562714,
+      "learning_rate": 5e-05,
+      "loss": 1.7035,
+      "step": 782
+    },
+    {
+      "epoch": 0.21410992616899097,
+      "grad_norm": 0.1558707058429718,
+      "learning_rate": 5e-05,
+      "loss": 1.6835,
+      "step": 783
+    },
+    {
+      "epoch": 0.21438337435056057,
+      "grad_norm": 0.15981443226337433,
+      "learning_rate": 5e-05,
+      "loss": 1.7418,
+      "step": 784
+    },
+    {
+      "epoch": 0.21465682253213017,
+      "grad_norm": 0.1494167298078537,
+      "learning_rate": 5e-05,
+      "loss": 1.5906,
+      "step": 785
+    },
+    {
+      "epoch": 0.21493027071369974,
+      "grad_norm": 0.15564508736133575,
+      "learning_rate": 5e-05,
+      "loss": 1.6873,
+      "step": 786
+    },
+    {
+      "epoch": 0.21520371889526935,
+      "grad_norm": 0.1639227420091629,
+      "learning_rate": 5e-05,
+      "loss": 1.7083,
+      "step": 787
+    },
+    {
+      "epoch": 0.21547716707683895,
+      "grad_norm": 0.14128105342388153,
+      "learning_rate": 5e-05,
+      "loss": 1.608,
+      "step": 788
+    },
+    {
+      "epoch": 0.21575061525840852,
+      "grad_norm": 0.16327965259552002,
+      "learning_rate": 5e-05,
+      "loss": 1.6022,
+      "step": 789
+    },
+    {
+      "epoch": 0.21602406343997813,
+      "grad_norm": 0.1559935361146927,
+      "learning_rate": 5e-05,
+      "loss": 1.6483,
+      "step": 790
+    },
+    {
+      "epoch": 0.21629751162154773,
+      "grad_norm": 0.1511768251657486,
+      "learning_rate": 5e-05,
+      "loss": 1.5955,
+      "step": 791
+    },
+    {
+      "epoch": 0.2165709598031173,
+      "grad_norm": 0.16056060791015625,
+      "learning_rate": 5e-05,
+      "loss": 1.6471,
+      "step": 792
+    },
+    {
+      "epoch": 0.2168444079846869,
+      "grad_norm": 0.16120120882987976,
+      "learning_rate": 5e-05,
+      "loss": 1.8007,
+      "step": 793
+    },
+    {
+      "epoch": 0.2171178561662565,
+      "grad_norm": 0.1515311449766159,
+      "learning_rate": 5e-05,
+      "loss": 1.6438,
+      "step": 794
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.15457142889499664,
+      "learning_rate": 5e-05,
+      "loss": 1.7115,
+      "step": 795
+    },
+    {
+      "epoch": 0.21766475252939568,
+      "grad_norm": 0.15303310751914978,
+      "learning_rate": 5e-05,
+      "loss": 1.6677,
+      "step": 796
+    },
+    {
+      "epoch": 0.21793820071096529,
+      "grad_norm": 0.1515769511461258,
+      "learning_rate": 5e-05,
+      "loss": 1.7371,
+      "step": 797
+    },
+    {
+      "epoch": 0.21821164889253486,
+      "grad_norm": 0.1483011692762375,
+      "learning_rate": 5e-05,
+      "loss": 1.6359,
+      "step": 798
+    },
+    {
+      "epoch": 0.21848509707410446,
+      "grad_norm": 0.1543753296136856,
+      "learning_rate": 5e-05,
+      "loss": 1.7473,
+      "step": 799
+    },
+    {
+      "epoch": 0.21875854525567404,
+      "grad_norm": 0.14442037045955658,
+      "learning_rate": 5e-05,
+      "loss": 1.6086,
+      "step": 800
+    },
+    {
+      "epoch": 0.21903199343724364,
+      "grad_norm": 0.14965881407260895,
+      "learning_rate": 5e-05,
+      "loss": 1.6522,
+      "step": 801
+    },
+    {
+      "epoch": 0.21930544161881324,
+      "grad_norm": 0.16104258596897125,
+      "learning_rate": 5e-05,
+      "loss": 1.5625,
+      "step": 802
+    },
+    {
+      "epoch": 0.21957888980038281,
+      "grad_norm": 0.16634979844093323,
+      "learning_rate": 5e-05,
+      "loss": 1.6574,
+      "step": 803
+    },
+    {
+      "epoch": 0.21985233798195242,
+      "grad_norm": 0.1422521471977234,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 804
+    },
+    {
+      "epoch": 0.22012578616352202,
+      "grad_norm": 0.17888352274894714,
+      "learning_rate": 5e-05,
+      "loss": 1.7694,
+      "step": 805
+    },
+    {
+      "epoch": 0.2203992343450916,
+      "grad_norm": 0.16180840134620667,
+      "learning_rate": 5e-05,
+      "loss": 1.6889,
+      "step": 806
+    },
+    {
+      "epoch": 0.2206726825266612,
+      "grad_norm": 0.1583503931760788,
+      "learning_rate": 5e-05,
+      "loss": 1.6678,
+      "step": 807
+    },
+    {
+      "epoch": 0.2209461307082308,
+      "grad_norm": 0.19179575145244598,
+      "learning_rate": 5e-05,
+      "loss": 1.7058,
+      "step": 808
+    },
+    {
+      "epoch": 0.22121957888980037,
+      "grad_norm": 0.14800883829593658,
+      "learning_rate": 5e-05,
+      "loss": 1.6305,
+      "step": 809
+    },
+    {
+      "epoch": 0.22149302707136997,
+      "grad_norm": 0.17706219851970673,
+      "learning_rate": 5e-05,
+      "loss": 1.691,
+      "step": 810
+    },
+    {
+      "epoch": 0.22176647525293958,
+      "grad_norm": 0.1569800078868866,
+      "learning_rate": 5e-05,
+      "loss": 1.5772,
+      "step": 811
+    },
+    {
+      "epoch": 0.22203992343450915,
+      "grad_norm": 0.1444752961397171,
+      "learning_rate": 5e-05,
+      "loss": 1.6368,
+      "step": 812
+    },
+    {
+      "epoch": 0.22231337161607875,
+      "grad_norm": 0.15736012160778046,
+      "learning_rate": 5e-05,
+      "loss": 1.5703,
+      "step": 813
+    },
+    {
+      "epoch": 0.22258681979764836,
+      "grad_norm": 0.18501217663288116,
+      "learning_rate": 5e-05,
+      "loss": 1.6198,
+      "step": 814
+    },
+    {
+      "epoch": 0.22286026797921793,
+      "grad_norm": 0.14157734811306,
+      "learning_rate": 5e-05,
+      "loss": 1.672,
+      "step": 815
+    },
+    {
+      "epoch": 0.22313371616078753,
+      "grad_norm": 0.16897475719451904,
+      "learning_rate": 5e-05,
+      "loss": 1.6156,
+      "step": 816
+    },
+    {
+      "epoch": 0.22340716434235713,
+      "grad_norm": 0.15974189341068268,
+      "learning_rate": 5e-05,
+      "loss": 1.5237,
+      "step": 817
+    },
+    {
+      "epoch": 0.2236806125239267,
+      "grad_norm": 0.14449931681156158,
+      "learning_rate": 5e-05,
+      "loss": 1.6376,
+      "step": 818
+    },
+    {
+      "epoch": 0.2239540607054963,
+      "grad_norm": 0.1783899962902069,
+      "learning_rate": 5e-05,
+      "loss": 1.707,
+      "step": 819
+    },
+    {
+      "epoch": 0.2242275088870659,
+      "grad_norm": 0.15125927329063416,
+      "learning_rate": 5e-05,
+      "loss": 1.6233,
+      "step": 820
+    },
+    {
+      "epoch": 0.2245009570686355,
+      "grad_norm": 0.15262438356876373,
+      "learning_rate": 5e-05,
+      "loss": 1.6687,
+      "step": 821
+    },
+    {
+      "epoch": 0.2247744052502051,
+      "grad_norm": 0.16108544170856476,
+      "learning_rate": 5e-05,
+      "loss": 1.647,
+      "step": 822
+    },
+    {
+      "epoch": 0.2250478534317747,
+      "grad_norm": 0.1504691243171692,
+      "learning_rate": 5e-05,
+      "loss": 1.6531,
+      "step": 823
+    },
+    {
+      "epoch": 0.22532130161334427,
+      "grad_norm": 0.1500689536333084,
+      "learning_rate": 5e-05,
+      "loss": 1.5914,
+      "step": 824
+    },
+    {
+      "epoch": 0.22559474979491387,
+      "grad_norm": 0.15878424048423767,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 825
+    },
+    {
+      "epoch": 0.22586819797648347,
+      "grad_norm": 0.15859322249889374,
+      "learning_rate": 5e-05,
+      "loss": 1.6787,
+      "step": 826
+    },
+    {
+      "epoch": 0.22614164615805304,
+      "grad_norm": 0.15319658815860748,
+      "learning_rate": 5e-05,
+      "loss": 1.6832,
+      "step": 827
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.15381459891796112,
+      "learning_rate": 5e-05,
+      "loss": 1.6422,
+      "step": 828
+    },
+    {
+      "epoch": 0.22668854252119222,
+      "grad_norm": 0.16640834510326385,
+      "learning_rate": 5e-05,
+      "loss": 1.7306,
+      "step": 829
+    },
+    {
+      "epoch": 0.22696199070276182,
+      "grad_norm": 0.1528118997812271,
+      "learning_rate": 5e-05,
+      "loss": 1.6858,
+      "step": 830
+    },
+    {
+      "epoch": 0.22723543888433143,
+      "grad_norm": 0.15058903396129608,
+      "learning_rate": 5e-05,
+      "loss": 1.6336,
+      "step": 831
+    },
+    {
+      "epoch": 0.227508887065901,
+      "grad_norm": 0.1552654355764389,
+      "learning_rate": 5e-05,
+      "loss": 1.677,
+      "step": 832
+    },
+    {
+      "epoch": 0.2277823352474706,
+      "grad_norm": 0.16159558296203613,
+      "learning_rate": 5e-05,
+      "loss": 1.682,
+      "step": 833
+    },
+    {
+      "epoch": 0.2280557834290402,
+      "grad_norm": 0.14523838460445404,
+      "learning_rate": 5e-05,
+      "loss": 1.5887,
+      "step": 834
+    },
+    {
+      "epoch": 0.22832923161060978,
+      "grad_norm": 0.16024388372898102,
+      "learning_rate": 5e-05,
+      "loss": 1.7047,
+      "step": 835
+    },
+    {
+      "epoch": 0.22860267979217938,
+      "grad_norm": 0.15451057255268097,
+      "learning_rate": 5e-05,
+      "loss": 1.6663,
+      "step": 836
+    },
+    {
+      "epoch": 0.22887612797374898,
+      "grad_norm": 0.14806225895881653,
+      "learning_rate": 5e-05,
+      "loss": 1.6782,
+      "step": 837
+    },
+    {
+      "epoch": 0.22914957615531856,
+      "grad_norm": 0.15798652172088623,
+      "learning_rate": 5e-05,
+      "loss": 1.7092,
+      "step": 838
+    },
+    {
+      "epoch": 0.22942302433688816,
+      "grad_norm": 0.15452131628990173,
+      "learning_rate": 5e-05,
+      "loss": 1.7137,
+      "step": 839
+    },
+    {
+      "epoch": 0.22969647251845776,
+      "grad_norm": 0.15365323424339294,
+      "learning_rate": 5e-05,
+      "loss": 1.7464,
+      "step": 840
+    },
+    {
+      "epoch": 0.22996992070002734,
+      "grad_norm": 0.14897583425045013,
+      "learning_rate": 5e-05,
+      "loss": 1.6712,
+      "step": 841
+    },
+    {
+      "epoch": 0.23024336888159694,
+      "grad_norm": 0.1519531011581421,
+      "learning_rate": 5e-05,
+      "loss": 1.7296,
+      "step": 842
+    },
+    {
+      "epoch": 0.23051681706316654,
+      "grad_norm": 0.14781694114208221,
+      "learning_rate": 5e-05,
+      "loss": 1.6932,
+      "step": 843
+    },
+    {
+      "epoch": 0.23079026524473611,
+      "grad_norm": 0.1538834124803543,
+      "learning_rate": 5e-05,
+      "loss": 1.6682,
+      "step": 844
+    },
+    {
+      "epoch": 0.23106371342630572,
+      "grad_norm": 0.14690068364143372,
+      "learning_rate": 5e-05,
+      "loss": 1.6194,
+      "step": 845
+    },
+    {
+      "epoch": 0.23133716160787532,
+      "grad_norm": 0.14810669422149658,
+      "learning_rate": 5e-05,
+      "loss": 1.6565,
+      "step": 846
+    },
+    {
+      "epoch": 0.2316106097894449,
+      "grad_norm": 0.1599397510290146,
+      "learning_rate": 5e-05,
+      "loss": 1.7253,
+      "step": 847
+    },
+    {
+      "epoch": 0.2318840579710145,
+      "grad_norm": 0.14897406101226807,
+      "learning_rate": 5e-05,
+      "loss": 1.6383,
+      "step": 848
+    },
+    {
+      "epoch": 0.2321575061525841,
+      "grad_norm": 0.14833606779575348,
+      "learning_rate": 5e-05,
+      "loss": 1.5975,
+      "step": 849
+    },
+    {
+      "epoch": 0.23243095433415367,
+      "grad_norm": 0.15134021639823914,
+      "learning_rate": 5e-05,
+      "loss": 1.6181,
+      "step": 850
+    },
+    {
+      "epoch": 0.23270440251572327,
+      "grad_norm": 0.1754770278930664,
+      "learning_rate": 5e-05,
+      "loss": 1.6905,
+      "step": 851
+    },
+    {
+      "epoch": 0.23297785069729288,
+      "grad_norm": 0.15956006944179535,
+      "learning_rate": 5e-05,
+      "loss": 1.7139,
+      "step": 852
+    },
+    {
+      "epoch": 0.23325129887886245,
+      "grad_norm": 0.1927836388349533,
+      "learning_rate": 5e-05,
+      "loss": 1.7274,
+      "step": 853
+    },
+    {
+      "epoch": 0.23352474706043205,
+      "grad_norm": 0.17408336699008942,
+      "learning_rate": 5e-05,
+      "loss": 1.6502,
+      "step": 854
+    },
+    {
+      "epoch": 0.23379819524200163,
+      "grad_norm": 0.15393292903900146,
+      "learning_rate": 5e-05,
+      "loss": 1.6537,
+      "step": 855
+    },
+    {
+      "epoch": 0.23407164342357123,
+      "grad_norm": 0.1970151662826538,
+      "learning_rate": 5e-05,
+      "loss": 1.6247,
+      "step": 856
+    },
+    {
+      "epoch": 0.23434509160514083,
+      "grad_norm": 0.14329130947589874,
+      "learning_rate": 5e-05,
+      "loss": 1.5819,
+      "step": 857
+    },
+    {
+      "epoch": 0.2346185397867104,
+      "grad_norm": 0.14831414818763733,
+      "learning_rate": 5e-05,
+      "loss": 1.6632,
+      "step": 858
+    },
+    {
+      "epoch": 0.23489198796828,
+      "grad_norm": 0.16687729954719543,
+      "learning_rate": 5e-05,
+      "loss": 1.6668,
+      "step": 859
+    },
+    {
+      "epoch": 0.2351654361498496,
+      "grad_norm": 0.14544977247714996,
+      "learning_rate": 5e-05,
+      "loss": 1.6716,
+      "step": 860
+    },
+    {
+      "epoch": 0.23543888433141918,
+      "grad_norm": 0.15175144374370575,
+      "learning_rate": 5e-05,
+      "loss": 1.7159,
+      "step": 861
+    },
+    {
+      "epoch": 0.2357123325129888,
+      "grad_norm": 0.17007999122142792,
+      "learning_rate": 5e-05,
+      "loss": 1.8236,
+      "step": 862
+    },
+    {
+      "epoch": 0.2359857806945584,
+      "grad_norm": 0.1416562795639038,
+      "learning_rate": 5e-05,
+      "loss": 1.6291,
+      "step": 863
+    },
+    {
+      "epoch": 0.23625922887612796,
+      "grad_norm": 0.16543252766132355,
+      "learning_rate": 5e-05,
+      "loss": 1.695,
+      "step": 864
+    },
+    {
+      "epoch": 0.23653267705769757,
+      "grad_norm": 0.16213998198509216,
+      "learning_rate": 5e-05,
+      "loss": 1.7187,
+      "step": 865
+    },
+    {
+      "epoch": 0.23680612523926717,
+      "grad_norm": 0.15842589735984802,
+      "learning_rate": 5e-05,
+      "loss": 1.7074,
+      "step": 866
+    },
+    {
+      "epoch": 0.23707957342083674,
+      "grad_norm": 0.16753311455249786,
+      "learning_rate": 5e-05,
+      "loss": 1.6487,
+      "step": 867
+    },
+    {
+      "epoch": 0.23735302160240634,
+      "grad_norm": 0.15180423855781555,
+      "learning_rate": 5e-05,
+      "loss": 1.6396,
+      "step": 868
+    },
+    {
+      "epoch": 0.23762646978397595,
+      "grad_norm": 0.1585211455821991,
+      "learning_rate": 5e-05,
+      "loss": 1.6959,
+      "step": 869
+    },
+    {
+      "epoch": 0.23789991796554552,
+      "grad_norm": 0.1608756184577942,
+      "learning_rate": 5e-05,
+      "loss": 1.7013,
+      "step": 870
+    },
+    {
+      "epoch": 0.23817336614711512,
+      "grad_norm": 0.14348694682121277,
+      "learning_rate": 5e-05,
+      "loss": 1.5302,
+      "step": 871
+    },
+    {
+      "epoch": 0.23844681432868473,
+      "grad_norm": 0.14560513198375702,
+      "learning_rate": 5e-05,
+      "loss": 1.6678,
+      "step": 872
+    },
+    {
+      "epoch": 0.2387202625102543,
+      "grad_norm": 0.1609930694103241,
+      "learning_rate": 5e-05,
+      "loss": 1.6599,
+      "step": 873
+    },
+    {
+      "epoch": 0.2389937106918239,
+      "grad_norm": 0.15456150472164154,
+      "learning_rate": 5e-05,
+      "loss": 1.6087,
+      "step": 874
+    },
+    {
+      "epoch": 0.2392671588733935,
+      "grad_norm": 0.14654546976089478,
+      "learning_rate": 5e-05,
+      "loss": 1.6199,
+      "step": 875
+    },
+    {
+      "epoch": 0.23954060705496308,
+      "grad_norm": 0.14911627769470215,
+      "learning_rate": 5e-05,
+      "loss": 1.7099,
+      "step": 876
+    },
+    {
+      "epoch": 0.23981405523653268,
+      "grad_norm": 0.17598745226860046,
+      "learning_rate": 5e-05,
+      "loss": 1.6854,
+      "step": 877
+    },
+    {
+      "epoch": 0.24008750341810228,
+      "grad_norm": 0.1485157608985901,
+      "learning_rate": 5e-05,
+      "loss": 1.652,
+      "step": 878
+    },
+    {
+      "epoch": 0.24036095159967186,
+      "grad_norm": 0.19263607263565063,
+      "learning_rate": 5e-05,
+      "loss": 1.6623,
+      "step": 879
+    },
+    {
+      "epoch": 0.24063439978124146,
+      "grad_norm": 0.15963439643383026,
+      "learning_rate": 5e-05,
+      "loss": 1.7693,
+      "step": 880
+    },
+    {
+      "epoch": 0.24090784796281103,
+      "grad_norm": 0.1660141795873642,
+      "learning_rate": 5e-05,
+      "loss": 1.5823,
+      "step": 881
+    },
+    {
+      "epoch": 0.24118129614438064,
+      "grad_norm": 0.16761627793312073,
+      "learning_rate": 5e-05,
+      "loss": 1.6647,
+      "step": 882
+    },
+    {
+      "epoch": 0.24145474432595024,
+      "grad_norm": 0.1527351289987564,
+      "learning_rate": 5e-05,
+      "loss": 1.5621,
+      "step": 883
+    },
+    {
+      "epoch": 0.2417281925075198,
+      "grad_norm": 0.1732681542634964,
+      "learning_rate": 5e-05,
+      "loss": 1.8039,
+      "step": 884
+    },
+    {
+      "epoch": 0.24200164068908941,
+      "grad_norm": 0.15724296867847443,
+      "learning_rate": 5e-05,
+      "loss": 1.5098,
+      "step": 885
+    },
+    {
+      "epoch": 0.24227508887065902,
+      "grad_norm": 0.1623707413673401,
+      "learning_rate": 5e-05,
+      "loss": 1.7287,
+      "step": 886
+    },
+    {
+      "epoch": 0.2425485370522286,
+      "grad_norm": 0.15608251094818115,
+      "learning_rate": 5e-05,
+      "loss": 1.6904,
+      "step": 887
+    },
+    {
+      "epoch": 0.2428219852337982,
+      "grad_norm": 0.16214315593242645,
+      "learning_rate": 5e-05,
+      "loss": 1.6739,
+      "step": 888
+    },
+    {
+      "epoch": 0.2430954334153678,
+      "grad_norm": 0.15404769778251648,
+      "learning_rate": 5e-05,
+      "loss": 1.6115,
+      "step": 889
+    },
+    {
+      "epoch": 0.24336888159693737,
+      "grad_norm": 0.17034713923931122,
+      "learning_rate": 5e-05,
+      "loss": 1.6412,
+      "step": 890
+    },
+    {
+      "epoch": 0.24364232977850697,
+      "grad_norm": 0.14412027597427368,
+      "learning_rate": 5e-05,
+      "loss": 1.5481,
+      "step": 891
+    },
+    {
+      "epoch": 0.24391577796007657,
+      "grad_norm": 0.1591709554195404,
+      "learning_rate": 5e-05,
+      "loss": 1.7115,
+      "step": 892
+    },
+    {
+      "epoch": 0.24418922614164615,
+      "grad_norm": 0.18128591775894165,
+      "learning_rate": 5e-05,
+      "loss": 1.6538,
+      "step": 893
+    },
+    {
+      "epoch": 0.24446267432321575,
+      "grad_norm": 0.1539849489927292,
+      "learning_rate": 5e-05,
+      "loss": 1.6899,
+      "step": 894
+    },
+    {
+      "epoch": 0.24473612250478535,
+      "grad_norm": 0.14450113475322723,
+      "learning_rate": 5e-05,
+      "loss": 1.5931,
+      "step": 895
+    },
+    {
+      "epoch": 0.24500957068635493,
+      "grad_norm": 0.152107372879982,
+      "learning_rate": 5e-05,
+      "loss": 1.6001,
+      "step": 896
+    },
+    {
+      "epoch": 0.24528301886792453,
+      "grad_norm": 0.14510129392147064,
+      "learning_rate": 5e-05,
+      "loss": 1.6048,
+      "step": 897
+    },
+    {
+      "epoch": 0.24555646704949413,
+      "grad_norm": 0.1490333080291748,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 898
+    },
+    {
+      "epoch": 0.2458299152310637,
+      "grad_norm": 0.14951343834400177,
+      "learning_rate": 5e-05,
+      "loss": 1.7005,
+      "step": 899
+    },
+    {
+      "epoch": 0.2461033634126333,
+      "grad_norm": 0.1492195427417755,
+      "learning_rate": 5e-05,
+      "loss": 1.6831,
+      "step": 900
+    },
+    {
+      "epoch": 0.2463768115942029,
+      "grad_norm": 0.15410034358501434,
+      "learning_rate": 5e-05,
+      "loss": 1.6933,
+      "step": 901
+    },
+    {
+      "epoch": 0.24665025977577248,
+      "grad_norm": 0.14823299646377563,
+      "learning_rate": 5e-05,
+      "loss": 1.637,
+      "step": 902
+    },
+    {
+      "epoch": 0.2469237079573421,
+      "grad_norm": 0.14969344437122345,
+      "learning_rate": 5e-05,
+      "loss": 1.6546,
+      "step": 903
+    },
+    {
+      "epoch": 0.2471971561389117,
+      "grad_norm": 0.14941661059856415,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 904
+    },
+    {
+      "epoch": 0.24747060432048126,
+      "grad_norm": 0.1853816956281662,
+      "learning_rate": 5e-05,
+      "loss": 1.6285,
+      "step": 905
+    },
+    {
+      "epoch": 0.24774405250205087,
+      "grad_norm": 0.15117646753787994,
+      "learning_rate": 5e-05,
+      "loss": 1.6274,
+      "step": 906
+    },
+    {
+      "epoch": 0.24801750068362047,
+      "grad_norm": 0.1522308588027954,
+      "learning_rate": 5e-05,
+      "loss": 1.735,
+      "step": 907
+    },
+    {
+      "epoch": 0.24829094886519004,
+      "grad_norm": 0.13882996141910553,
+      "learning_rate": 5e-05,
+      "loss": 1.5867,
+      "step": 908
+    },
+    {
+      "epoch": 0.24856439704675964,
+      "grad_norm": 0.1577073335647583,
+      "learning_rate": 5e-05,
+      "loss": 1.6647,
+      "step": 909
+    },
+    {
+      "epoch": 0.24883784522832922,
+      "grad_norm": 0.15113559365272522,
+      "learning_rate": 5e-05,
+      "loss": 1.6303,
+      "step": 910
+    },
+    {
+      "epoch": 0.24911129340989882,
+      "grad_norm": 0.15019430220127106,
+      "learning_rate": 5e-05,
+      "loss": 1.6976,
+      "step": 911
+    },
+    {
+      "epoch": 0.24938474159146842,
+      "grad_norm": 0.14112697541713715,
+      "learning_rate": 5e-05,
+      "loss": 1.54,
+      "step": 912
+    },
+    {
+      "epoch": 0.249658189773038,
+      "grad_norm": 0.14087800681591034,
+      "learning_rate": 5e-05,
+      "loss": 1.5316,
+      "step": 913
+    },
+    {
+      "epoch": 0.2499316379546076,
+      "grad_norm": 0.14896658062934875,
+      "learning_rate": 5e-05,
+      "loss": 1.561,
+      "step": 914
+    },
+    {
+      "epoch": 0.2502050861361772,
+      "grad_norm": 0.15148289501667023,
+      "learning_rate": 5e-05,
+      "loss": 1.6698,
+      "step": 915
+    },
+    {
+      "epoch": 0.2504785343177468,
+      "grad_norm": 0.1539052277803421,
+      "learning_rate": 5e-05,
+      "loss": 1.6017,
+      "step": 916
+    },
+    {
+      "epoch": 0.2507519824993164,
+      "grad_norm": 0.16018076241016388,
+      "learning_rate": 5e-05,
+      "loss": 1.6956,
+      "step": 917
+    },
+    {
+      "epoch": 0.25102543068088595,
+      "grad_norm": 0.17696644365787506,
+      "learning_rate": 5e-05,
+      "loss": 1.7196,
+      "step": 918
+    },
+    {
+      "epoch": 0.25129887886245555,
+      "grad_norm": 0.1509835124015808,
+      "learning_rate": 5e-05,
+      "loss": 1.6243,
+      "step": 919
+    },
+    {
+      "epoch": 0.25157232704402516,
+      "grad_norm": 0.15202875435352325,
+      "learning_rate": 5e-05,
+      "loss": 1.6868,
+      "step": 920
+    },
+    {
+      "epoch": 0.25184577522559476,
+      "grad_norm": 0.16320888698101044,
+      "learning_rate": 5e-05,
+      "loss": 1.697,
+      "step": 921
+    },
+    {
+      "epoch": 0.25211922340716436,
+      "grad_norm": 0.15281890332698822,
+      "learning_rate": 5e-05,
+      "loss": 1.5824,
+      "step": 922
+    },
+    {
+      "epoch": 0.2523926715887339,
+      "grad_norm": 0.1686651110649109,
+      "learning_rate": 5e-05,
+      "loss": 1.6362,
+      "step": 923
+    },
+    {
+      "epoch": 0.2526661197703035,
+      "grad_norm": 0.16698114573955536,
+      "learning_rate": 5e-05,
+      "loss": 1.6557,
+      "step": 924
+    },
+    {
+      "epoch": 0.2529395679518731,
+      "grad_norm": 0.16273106634616852,
+      "learning_rate": 5e-05,
+      "loss": 1.7441,
+      "step": 925
+    },
+    {
+      "epoch": 0.2532130161334427,
+      "grad_norm": 0.16961312294006348,
+      "learning_rate": 5e-05,
+      "loss": 1.7053,
+      "step": 926
+    },
+    {
+      "epoch": 0.2534864643150123,
+      "grad_norm": 0.1489211916923523,
+      "learning_rate": 5e-05,
+      "loss": 1.6257,
+      "step": 927
+    },
+    {
+      "epoch": 0.2537599124965819,
+      "grad_norm": 0.15058746933937073,
+      "learning_rate": 5e-05,
+      "loss": 1.61,
+      "step": 928
+    },
+    {
+      "epoch": 0.25403336067815147,
+      "grad_norm": 0.16194166243076324,
+      "learning_rate": 5e-05,
+      "loss": 1.6667,
+      "step": 929
+    },
+    {
+      "epoch": 0.25430680885972107,
+      "grad_norm": 0.1765557825565338,
+      "learning_rate": 5e-05,
+      "loss": 1.6572,
+      "step": 930
+    },
+    {
+      "epoch": 0.25458025704129067,
+      "grad_norm": 0.14746791124343872,
+      "learning_rate": 5e-05,
+      "loss": 1.6596,
+      "step": 931
+    },
+    {
+      "epoch": 0.25485370522286027,
+      "grad_norm": 0.14352820813655853,
+      "learning_rate": 5e-05,
+      "loss": 1.5291,
+      "step": 932
+    },
+    {
+      "epoch": 0.2551271534044299,
+      "grad_norm": 0.15766541659832,
+      "learning_rate": 5e-05,
+      "loss": 1.6496,
+      "step": 933
+    },
+    {
+      "epoch": 0.2554006015859995,
+      "grad_norm": 0.1461249589920044,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 934
+    },
+    {
+      "epoch": 0.255674049767569,
+      "grad_norm": 0.15647587180137634,
+      "learning_rate": 5e-05,
+      "loss": 1.7245,
+      "step": 935
+    },
+    {
+      "epoch": 0.2559474979491386,
+      "grad_norm": 0.16043874621391296,
+      "learning_rate": 5e-05,
+      "loss": 1.7584,
+      "step": 936
+    },
+    {
+      "epoch": 0.2562209461307082,
+      "grad_norm": 0.16398166120052338,
+      "learning_rate": 5e-05,
+      "loss": 1.7888,
+      "step": 937
+    },
+    {
+      "epoch": 0.25649439431227783,
+      "grad_norm": 0.1568370908498764,
+      "learning_rate": 5e-05,
+      "loss": 1.6886,
+      "step": 938
+    },
+    {
+      "epoch": 0.25676784249384743,
+      "grad_norm": 0.14432042837142944,
+      "learning_rate": 5e-05,
+      "loss": 1.5691,
+      "step": 939
+    },
+    {
+      "epoch": 0.25704129067541703,
+      "grad_norm": 0.15498638153076172,
+      "learning_rate": 5e-05,
+      "loss": 1.6414,
+      "step": 940
+    },
+    {
+      "epoch": 0.2573147388569866,
+      "grad_norm": 0.1573934257030487,
+      "learning_rate": 5e-05,
+      "loss": 1.7171,
+      "step": 941
+    },
+    {
+      "epoch": 0.2575881870385562,
+      "grad_norm": 0.16507598757743835,
+      "learning_rate": 5e-05,
+      "loss": 1.7697,
+      "step": 942
+    },
+    {
+      "epoch": 0.2578616352201258,
+      "grad_norm": 0.16209067404270172,
+      "learning_rate": 5e-05,
+      "loss": 1.7732,
+      "step": 943
+    },
+    {
+      "epoch": 0.2581350834016954,
+      "grad_norm": 0.16477470099925995,
+      "learning_rate": 5e-05,
+      "loss": 1.6439,
+      "step": 944
+    },
+    {
+      "epoch": 0.258408531583265,
+      "grad_norm": 0.15217360854148865,
+      "learning_rate": 5e-05,
+      "loss": 1.6341,
+      "step": 945
+    },
+    {
+      "epoch": 0.2586819797648346,
+      "grad_norm": 0.16021955013275146,
+      "learning_rate": 5e-05,
+      "loss": 1.6829,
+      "step": 946
+    },
+    {
+      "epoch": 0.25895542794640414,
+      "grad_norm": 0.16192196309566498,
+      "learning_rate": 5e-05,
+      "loss": 1.6225,
+      "step": 947
+    },
+    {
+      "epoch": 0.25922887612797374,
+      "grad_norm": 0.15227045118808746,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 948
+    },
+    {
+      "epoch": 0.25950232430954334,
+      "grad_norm": 0.17690598964691162,
+      "learning_rate": 5e-05,
+      "loss": 1.7092,
+      "step": 949
+    },
+    {
+      "epoch": 0.25977577249111294,
+      "grad_norm": 0.1463916003704071,
+      "learning_rate": 5e-05,
+      "loss": 1.5948,
+      "step": 950
+    },
+    {
+      "epoch": 0.26004922067268255,
+      "grad_norm": 0.16608351469039917,
+      "learning_rate": 5e-05,
+      "loss": 1.6339,
+      "step": 951
+    },
+    {
+      "epoch": 0.2603226688542521,
+      "grad_norm": 0.16047005355358124,
+      "learning_rate": 5e-05,
+      "loss": 1.6928,
+      "step": 952
+    },
+    {
+      "epoch": 0.2605961170358217,
+      "grad_norm": 0.1434023380279541,
+      "learning_rate": 5e-05,
+      "loss": 1.5613,
+      "step": 953
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.1558619737625122,
+      "learning_rate": 5e-05,
+      "loss": 1.6868,
+      "step": 954
+    },
+    {
+      "epoch": 0.2611430133989609,
+      "grad_norm": 0.14743700623512268,
+      "learning_rate": 5e-05,
+      "loss": 1.612,
+      "step": 955
+    },
+    {
+      "epoch": 0.2614164615805305,
+      "grad_norm": 0.15097011625766754,
+      "learning_rate": 5e-05,
+      "loss": 1.5625,
+      "step": 956
+    },
+    {
+      "epoch": 0.2616899097621001,
+      "grad_norm": 0.14291320741176605,
+      "learning_rate": 5e-05,
+      "loss": 1.6787,
+      "step": 957
+    },
+    {
+      "epoch": 0.26196335794366965,
+      "grad_norm": 0.15289993584156036,
+      "learning_rate": 5e-05,
+      "loss": 1.754,
+      "step": 958
+    },
+    {
+      "epoch": 0.26223680612523925,
+      "grad_norm": 0.15618576109409332,
+      "learning_rate": 5e-05,
+      "loss": 1.7212,
+      "step": 959
+    },
+    {
+      "epoch": 0.26251025430680885,
+      "grad_norm": 0.14384949207305908,
+      "learning_rate": 5e-05,
+      "loss": 1.6108,
+      "step": 960
+    },
+    {
+      "epoch": 0.26278370248837846,
+      "grad_norm": 0.15583495795726776,
+      "learning_rate": 5e-05,
+      "loss": 1.6998,
+      "step": 961
+    },
+    {
+      "epoch": 0.26305715066994806,
+      "grad_norm": 0.1544669270515442,
+      "learning_rate": 5e-05,
+      "loss": 1.646,
+      "step": 962
+    },
+    {
+      "epoch": 0.26333059885151766,
+      "grad_norm": 0.15387408435344696,
+      "learning_rate": 5e-05,
+      "loss": 1.7415,
+      "step": 963
+    },
+    {
+      "epoch": 0.2636040470330872,
+      "grad_norm": 0.15197156369686127,
+      "learning_rate": 5e-05,
+      "loss": 1.6324,
+      "step": 964
+    },
+    {
+      "epoch": 0.2638774952146568,
+      "grad_norm": 0.15596544742584229,
+      "learning_rate": 5e-05,
+      "loss": 1.6116,
+      "step": 965
+    },
+    {
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.15174002945423126,
+      "learning_rate": 5e-05,
+      "loss": 1.6552,
+      "step": 966
+    },
+    {
+      "epoch": 0.264424391577796,
+      "grad_norm": 0.16541074216365814,
+      "learning_rate": 5e-05,
+      "loss": 1.7092,
+      "step": 967
+    },
+    {
+      "epoch": 0.2646978397593656,
+      "grad_norm": 0.15054158866405487,
+      "learning_rate": 5e-05,
+      "loss": 1.6584,
+      "step": 968
+    },
+    {
+      "epoch": 0.2649712879409352,
+      "grad_norm": 0.16845272481441498,
+      "learning_rate": 5e-05,
+      "loss": 1.7485,
+      "step": 969
+    },
+    {
+      "epoch": 0.26524473612250477,
+      "grad_norm": 0.15951582789421082,
+      "learning_rate": 5e-05,
+      "loss": 1.6003,
+      "step": 970
+    },
+    {
+      "epoch": 0.26551818430407437,
+      "grad_norm": 0.15601181983947754,
+      "learning_rate": 5e-05,
+      "loss": 1.6222,
+      "step": 971
+    },
+    {
+      "epoch": 0.26579163248564397,
+      "grad_norm": 0.1529085487127304,
+      "learning_rate": 5e-05,
+      "loss": 1.687,
+      "step": 972
+    },
+    {
+      "epoch": 0.26606508066721357,
+      "grad_norm": 0.15299002826213837,
+      "learning_rate": 5e-05,
+      "loss": 1.716,
+      "step": 973
+    },
+    {
+      "epoch": 0.2663385288487832,
+      "grad_norm": 0.17247562110424042,
+      "learning_rate": 5e-05,
+      "loss": 1.6112,
+      "step": 974
+    },
+    {
+      "epoch": 0.2666119770303528,
+      "grad_norm": 0.16275718808174133,
+      "learning_rate": 5e-05,
+      "loss": 1.6481,
+      "step": 975
+    },
+    {
+      "epoch": 0.2668854252119223,
+      "grad_norm": 0.18116293847560883,
+      "learning_rate": 5e-05,
+      "loss": 1.5794,
+      "step": 976
+    },
+    {
+      "epoch": 0.2671588733934919,
+      "grad_norm": 0.14527344703674316,
+      "learning_rate": 5e-05,
+      "loss": 1.6521,
+      "step": 977
+    },
+    {
+      "epoch": 0.2674323215750615,
+      "grad_norm": 0.15483027696609497,
+      "learning_rate": 5e-05,
+      "loss": 1.637,
+      "step": 978
+    },
+    {
+      "epoch": 0.26770576975663113,
+      "grad_norm": 0.15565639734268188,
+      "learning_rate": 5e-05,
+      "loss": 1.6142,
+      "step": 979
+    },
+    {
+      "epoch": 0.26797921793820073,
+      "grad_norm": 0.15443767607212067,
+      "learning_rate": 5e-05,
+      "loss": 1.5859,
+      "step": 980
+    },
+    {
+      "epoch": 0.2682526661197703,
+      "grad_norm": 0.15313716232776642,
+      "learning_rate": 5e-05,
+      "loss": 1.6508,
+      "step": 981
+    },
+    {
+      "epoch": 0.2685261143013399,
+      "grad_norm": 0.15430563688278198,
+      "learning_rate": 5e-05,
+      "loss": 1.6352,
+      "step": 982
+    },
+    {
+      "epoch": 0.2687995624829095,
+      "grad_norm": 0.15906836092472076,
+      "learning_rate": 5e-05,
+      "loss": 1.7201,
+      "step": 983
+    },
+    {
+      "epoch": 0.2690730106644791,
+      "grad_norm": 0.151002898812294,
+      "learning_rate": 5e-05,
+      "loss": 1.7418,
+      "step": 984
+    },
+    {
+      "epoch": 0.2693464588460487,
+      "grad_norm": 0.1649433970451355,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 985
+    },
+    {
+      "epoch": 0.2696199070276183,
+      "grad_norm": 0.15000388026237488,
+      "learning_rate": 5e-05,
+      "loss": 1.5924,
+      "step": 986
+    },
+    {
+      "epoch": 0.26989335520918784,
+      "grad_norm": 0.1750802844762802,
+      "learning_rate": 5e-05,
+      "loss": 1.621,
+      "step": 987
+    },
+    {
+      "epoch": 0.27016680339075744,
+      "grad_norm": 0.14742594957351685,
+      "learning_rate": 5e-05,
+      "loss": 1.6103,
+      "step": 988
+    },
+    {
+      "epoch": 0.27044025157232704,
+      "grad_norm": 0.1554790437221527,
+      "learning_rate": 5e-05,
+      "loss": 1.6996,
+      "step": 989
+    },
+    {
+      "epoch": 0.27071369975389664,
+      "grad_norm": 0.16735535860061646,
+      "learning_rate": 5e-05,
+      "loss": 1.6705,
+      "step": 990
+    },
+    {
+      "epoch": 0.27098714793546624,
+      "grad_norm": 0.1675184816122055,
+      "learning_rate": 5e-05,
+      "loss": 1.7114,
+      "step": 991
+    },
+    {
+      "epoch": 0.27126059611703585,
+      "grad_norm": 0.1542404443025589,
+      "learning_rate": 5e-05,
+      "loss": 1.5953,
+      "step": 992
+    },
+    {
+      "epoch": 0.2715340442986054,
+      "grad_norm": 0.1605929136276245,
+      "learning_rate": 5e-05,
+      "loss": 1.5541,
+      "step": 993
+    },
+    {
+      "epoch": 0.271807492480175,
+      "grad_norm": 0.17429675161838531,
+      "learning_rate": 5e-05,
+      "loss": 1.6393,
+      "step": 994
+    },
+    {
+      "epoch": 0.2720809406617446,
+      "grad_norm": 0.16409185528755188,
+      "learning_rate": 5e-05,
+      "loss": 1.7244,
+      "step": 995
+    },
+    {
+      "epoch": 0.2723543888433142,
+      "grad_norm": 0.16131426393985748,
+      "learning_rate": 5e-05,
+      "loss": 1.6339,
+      "step": 996
+    },
+    {
+      "epoch": 0.2726278370248838,
+      "grad_norm": 0.14714978635311127,
+      "learning_rate": 5e-05,
+      "loss": 1.5154,
+      "step": 997
+    },
+    {
+      "epoch": 0.2729012852064534,
+      "grad_norm": 0.14420582354068756,
+      "learning_rate": 5e-05,
+      "loss": 1.654,
+      "step": 998
+    },
+    {
+      "epoch": 0.27317473338802295,
+      "grad_norm": 0.15642261505126953,
+      "learning_rate": 5e-05,
+      "loss": 1.5478,
+      "step": 999
+    },
+    {
+      "epoch": 0.27344818156959255,
+      "grad_norm": 0.14731954038143158,
+      "learning_rate": 5e-05,
+      "loss": 1.5659,
+      "step": 1000
+    },
+    {
+      "epoch": 0.27372162975116215,
+      "grad_norm": 0.16006474196910858,
+      "learning_rate": 5e-05,
+      "loss": 1.6048,
+      "step": 1001
+    },
+    {
+      "epoch": 0.27399507793273176,
+      "grad_norm": 0.15757764875888824,
+      "learning_rate": 5e-05,
+      "loss": 1.6572,
+      "step": 1002
+    },
+    {
+      "epoch": 0.27426852611430136,
+      "grad_norm": 0.15396282076835632,
+      "learning_rate": 5e-05,
+      "loss": 1.7424,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2745419742958709,
+      "grad_norm": 0.15785497426986694,
+      "learning_rate": 5e-05,
+      "loss": 1.7142,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2748154224774405,
+      "grad_norm": 0.14254657924175262,
+      "learning_rate": 5e-05,
+      "loss": 1.5812,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2750888706590101,
+      "grad_norm": 0.1501447856426239,
+      "learning_rate": 5e-05,
+      "loss": 1.7112,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2753623188405797,
+      "grad_norm": 0.16908079385757446,
+      "learning_rate": 5e-05,
+      "loss": 1.7688,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2756357670221493,
+      "grad_norm": 0.1612824648618698,
+      "learning_rate": 5e-05,
+      "loss": 1.6473,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2759092152037189,
+      "grad_norm": 0.1502557396888733,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 1009
+    },
+    {
+      "epoch": 0.27618266338528846,
+      "grad_norm": 0.1549758017063141,
+      "learning_rate": 5e-05,
+      "loss": 1.6473,
+      "step": 1010
+    },
+    {
+      "epoch": 0.27645611156685806,
+      "grad_norm": 0.1512656807899475,
+      "learning_rate": 5e-05,
+      "loss": 1.7336,
+      "step": 1011
+    },
+    {
+      "epoch": 0.27672955974842767,
+      "grad_norm": 0.1543295830488205,
+      "learning_rate": 5e-05,
+      "loss": 1.6293,
+      "step": 1012
+    },
+    {
+      "epoch": 0.27700300792999727,
+      "grad_norm": 0.15347859263420105,
+      "learning_rate": 5e-05,
+      "loss": 1.589,
+      "step": 1013
+    },
+    {
+      "epoch": 0.27727645611156687,
+      "grad_norm": 0.15677101910114288,
+      "learning_rate": 5e-05,
+      "loss": 1.7283,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2775499042931365,
+      "grad_norm": 0.16513799130916595,
+      "learning_rate": 5e-05,
+      "loss": 1.6725,
+      "step": 1015
+    },
+    {
+      "epoch": 0.277823352474706,
+      "grad_norm": 0.15116244554519653,
+      "learning_rate": 5e-05,
+      "loss": 1.639,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2780968006562756,
+      "grad_norm": 0.1464737504720688,
+      "learning_rate": 5e-05,
+      "loss": 1.6598,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2783702488378452,
+      "grad_norm": 0.1716412454843521,
+      "learning_rate": 5e-05,
+      "loss": 1.6941,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2786436970194148,
+      "grad_norm": 0.15639688074588776,
+      "learning_rate": 5e-05,
+      "loss": 1.6334,
+      "step": 1019
+    },
+    {
+      "epoch": 0.27891714520098443,
+      "grad_norm": 0.1594480276107788,
+      "learning_rate": 5e-05,
+      "loss": 1.6916,
+      "step": 1020
+    },
+    {
+      "epoch": 0.27919059338255403,
+      "grad_norm": 0.14737538993358612,
+      "learning_rate": 5e-05,
+      "loss": 1.5393,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2794640415641236,
+      "grad_norm": 0.15937146544456482,
+      "learning_rate": 5e-05,
+      "loss": 1.7071,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2797374897456932,
+      "grad_norm": 0.15143409371376038,
+      "learning_rate": 5e-05,
+      "loss": 1.6276,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2800109379272628,
+      "grad_norm": 0.16173769533634186,
+      "learning_rate": 5e-05,
+      "loss": 1.6131,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2802843861088324,
+      "grad_norm": 0.14734511077404022,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 1025
+    },
+    {
+      "epoch": 0.280557834290402,
+      "grad_norm": 0.1571275144815445,
+      "learning_rate": 5e-05,
+      "loss": 1.7169,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2808312824719716,
+      "grad_norm": 0.15668824315071106,
+      "learning_rate": 5e-05,
+      "loss": 1.8378,
+      "step": 1027
+    },
+    {
+      "epoch": 0.28110473065354113,
+      "grad_norm": 0.15768049657344818,
+      "learning_rate": 5e-05,
+      "loss": 1.6911,
+      "step": 1028
+    },
+    {
+      "epoch": 0.28137817883511074,
+      "grad_norm": 0.15700852870941162,
+      "learning_rate": 5e-05,
+      "loss": 1.7142,
+      "step": 1029
+    },
+    {
+      "epoch": 0.28165162701668034,
+      "grad_norm": 0.14887461066246033,
+      "learning_rate": 5e-05,
+      "loss": 1.6241,
+      "step": 1030
+    },
+    {
+      "epoch": 0.28192507519824994,
+      "grad_norm": 0.15155330300331116,
+      "learning_rate": 5e-05,
+      "loss": 1.7063,
+      "step": 1031
+    },
+    {
+      "epoch": 0.28219852337981954,
+      "grad_norm": 0.1655176430940628,
+      "learning_rate": 5e-05,
+      "loss": 1.6063,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2824719715613891,
+      "grad_norm": 0.1478133499622345,
+      "learning_rate": 5e-05,
+      "loss": 1.6977,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2827454197429587,
+      "grad_norm": 0.14731581509113312,
+      "learning_rate": 5e-05,
+      "loss": 1.6238,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.16165371239185333,
+      "learning_rate": 5e-05,
+      "loss": 1.5795,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2832923161060979,
+      "grad_norm": 0.1721602976322174,
+      "learning_rate": 5e-05,
+      "loss": 1.7447,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2835657642876675,
+      "grad_norm": 0.14430297911167145,
+      "learning_rate": 5e-05,
+      "loss": 1.5664,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2838392124692371,
+      "grad_norm": 0.15474802255630493,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28411266065080665,
+      "grad_norm": 0.16973662376403809,
+      "learning_rate": 5e-05,
+      "loss": 1.601,
+      "step": 1039
+    },
+    {
+      "epoch": 0.28438610883237625,
+      "grad_norm": 0.170043483376503,
+      "learning_rate": 5e-05,
+      "loss": 1.6629,
+      "step": 1040
+    },
+    {
+      "epoch": 0.28465955701394585,
+      "grad_norm": 0.1592838615179062,
+      "learning_rate": 5e-05,
+      "loss": 1.6638,
+      "step": 1041
+    },
+    {
+      "epoch": 0.28493300519551545,
+      "grad_norm": 0.1679082214832306,
+      "learning_rate": 5e-05,
+      "loss": 1.6667,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28520645337708506,
+      "grad_norm": 0.17934280633926392,
+      "learning_rate": 5e-05,
+      "loss": 1.6224,
+      "step": 1043
+    },
+    {
+      "epoch": 0.28547990155865466,
+      "grad_norm": 0.14445669949054718,
+      "learning_rate": 5e-05,
+      "loss": 1.6861,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2857533497402242,
+      "grad_norm": 0.15397311747074127,
+      "learning_rate": 5e-05,
+      "loss": 1.7506,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2860267979217938,
+      "grad_norm": 0.15878726541996002,
+      "learning_rate": 5e-05,
+      "loss": 1.7283,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2863002461033634,
+      "grad_norm": 0.15634529292583466,
+      "learning_rate": 5e-05,
+      "loss": 1.6385,
+      "step": 1047
+    },
+    {
+      "epoch": 0.286573694284933,
+      "grad_norm": 0.15060849487781525,
+      "learning_rate": 5e-05,
+      "loss": 1.6744,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2868471424665026,
+      "grad_norm": 0.16258785128593445,
+      "learning_rate": 5e-05,
+      "loss": 1.6701,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2871205906480722,
+      "grad_norm": 0.14815811812877655,
+      "learning_rate": 5e-05,
+      "loss": 1.5947,
+      "step": 1050
+    },
+    {
+      "epoch": 0.28739403882964176,
+      "grad_norm": 0.17626261711120605,
+      "learning_rate": 5e-05,
+      "loss": 1.6407,
+      "step": 1051
+    },
+    {
+      "epoch": 0.28766748701121136,
+      "grad_norm": 0.1558643877506256,
+      "learning_rate": 5e-05,
+      "loss": 1.6002,
+      "step": 1052
+    },
+    {
+      "epoch": 0.28794093519278097,
+      "grad_norm": 0.14758311212062836,
+      "learning_rate": 5e-05,
+      "loss": 1.5985,
+      "step": 1053
+    },
+    {
+      "epoch": 0.28821438337435057,
+      "grad_norm": 0.16778376698493958,
+      "learning_rate": 5e-05,
+      "loss": 1.6799,
+      "step": 1054
+    },
+    {
+      "epoch": 0.28848783155592017,
+      "grad_norm": 0.16925834119319916,
+      "learning_rate": 5e-05,
+      "loss": 1.6951,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2887612797374898,
+      "grad_norm": 0.16058339178562164,
+      "learning_rate": 5e-05,
+      "loss": 1.6221,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2890347279190593,
+      "grad_norm": 0.16010229289531708,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2893081761006289,
+      "grad_norm": 0.15044192969799042,
+      "learning_rate": 5e-05,
+      "loss": 1.6247,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2895816242821985,
+      "grad_norm": 0.14547835290431976,
+      "learning_rate": 5e-05,
+      "loss": 1.6111,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.15208472311496735,
+      "learning_rate": 5e-05,
+      "loss": 1.7198,
+      "step": 1060
+    },
+    {
+      "epoch": 0.29012852064533773,
+      "grad_norm": 0.15193967521190643,
+      "learning_rate": 5e-05,
+      "loss": 1.6775,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2904019688269073,
+      "grad_norm": 0.15141284465789795,
+      "learning_rate": 5e-05,
+      "loss": 1.6727,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2906754170084769,
+      "grad_norm": 0.16151681542396545,
+      "learning_rate": 5e-05,
+      "loss": 1.7054,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2909488651900465,
+      "grad_norm": 0.15598168969154358,
+      "learning_rate": 5e-05,
+      "loss": 1.6536,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2912223133716161,
+      "grad_norm": 0.147295281291008,
+      "learning_rate": 5e-05,
+      "loss": 1.658,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2914957615531857,
+      "grad_norm": 0.1515830159187317,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2917692097347553,
+      "grad_norm": 0.1516982913017273,
+      "learning_rate": 5e-05,
+      "loss": 1.6866,
+      "step": 1067
+    },
+    {
+      "epoch": 0.29204265791632483,
+      "grad_norm": 0.16418761014938354,
+      "learning_rate": 5e-05,
+      "loss": 1.7676,
+      "step": 1068
+    },
+    {
+      "epoch": 0.29231610609789443,
+      "grad_norm": 0.1566287726163864,
+      "learning_rate": 5e-05,
+      "loss": 1.6747,
+      "step": 1069
+    },
+    {
+      "epoch": 0.29258955427946404,
+      "grad_norm": 0.16158966720104218,
+      "learning_rate": 5e-05,
+      "loss": 1.6606,
+      "step": 1070
+    },
+    {
+      "epoch": 0.29286300246103364,
+      "grad_norm": 0.1478201299905777,
+      "learning_rate": 5e-05,
+      "loss": 1.708,
+      "step": 1071
+    },
+    {
+      "epoch": 0.29313645064260324,
+      "grad_norm": 0.15915492177009583,
+      "learning_rate": 5e-05,
+      "loss": 1.6656,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29340989882417284,
+      "grad_norm": 0.16170357167720795,
+      "learning_rate": 5e-05,
+      "loss": 1.6458,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2936833470057424,
+      "grad_norm": 0.1563751995563507,
+      "learning_rate": 5e-05,
+      "loss": 1.6192,
+      "step": 1074
+    },
+    {
+      "epoch": 0.293956795187312,
+      "grad_norm": 0.1516135334968567,
+      "learning_rate": 5e-05,
+      "loss": 1.6441,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2942302433688816,
+      "grad_norm": 0.14756453037261963,
+      "learning_rate": 5e-05,
+      "loss": 1.5897,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2945036915504512,
+      "grad_norm": 0.1514975130558014,
+      "learning_rate": 5e-05,
+      "loss": 1.6956,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2947771397320208,
+      "grad_norm": 0.1433197408914566,
+      "learning_rate": 5e-05,
+      "loss": 1.5484,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2950505879135904,
+      "grad_norm": 0.14783865213394165,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 1079
+    },
+    {
+      "epoch": 0.29532403609515995,
+      "grad_norm": 0.1563805192708969,
+      "learning_rate": 5e-05,
+      "loss": 1.6578,
+      "step": 1080
+    },
+    {
+      "epoch": 0.29559748427672955,
+      "grad_norm": 0.15550269186496735,
+      "learning_rate": 5e-05,
+      "loss": 1.6739,
+      "step": 1081
+    },
+    {
+      "epoch": 0.29587093245829915,
+      "grad_norm": 0.1647748500108719,
+      "learning_rate": 5e-05,
+      "loss": 1.8268,
+      "step": 1082
+    },
+    {
+      "epoch": 0.29614438063986875,
+      "grad_norm": 0.14956046640872955,
+      "learning_rate": 5e-05,
+      "loss": 1.6989,
+      "step": 1083
+    },
+    {
+      "epoch": 0.29641782882143836,
+      "grad_norm": 0.1558298021554947,
+      "learning_rate": 5e-05,
+      "loss": 1.6784,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2966912770030079,
+      "grad_norm": 0.16264335811138153,
+      "learning_rate": 5e-05,
+      "loss": 1.6733,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2969647251845775,
+      "grad_norm": 0.15490394830703735,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2972381733661471,
+      "grad_norm": 0.15151673555374146,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2975116215477167,
+      "grad_norm": 0.15661917626857758,
+      "learning_rate": 5e-05,
+      "loss": 1.6329,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2977850697292863,
+      "grad_norm": 0.1620190590620041,
+      "learning_rate": 5e-05,
+      "loss": 1.6779,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2980585179108559,
+      "grad_norm": 0.15718665719032288,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 1090
+    },
+    {
+      "epoch": 0.29833196609242546,
+      "grad_norm": 0.15119051933288574,
+      "learning_rate": 5e-05,
+      "loss": 1.6514,
+      "step": 1091
+    },
+    {
+      "epoch": 0.29860541427399506,
+      "grad_norm": 0.172483429312706,
+      "learning_rate": 5e-05,
+      "loss": 1.6253,
+      "step": 1092
+    },
+    {
+      "epoch": 0.29887886245556466,
+      "grad_norm": 0.17236994206905365,
+      "learning_rate": 5e-05,
+      "loss": 1.6927,
+      "step": 1093
+    },
+    {
+      "epoch": 0.29915231063713427,
+      "grad_norm": 0.1658979058265686,
+      "learning_rate": 5e-05,
+      "loss": 1.6684,
+      "step": 1094
+    },
+    {
+      "epoch": 0.29942575881870387,
+      "grad_norm": 0.16414695978164673,
+      "learning_rate": 5e-05,
+      "loss": 1.5933,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29969920700027347,
+      "grad_norm": 0.156681627035141,
+      "learning_rate": 5e-05,
+      "loss": 1.6016,
+      "step": 1096
+    },
+    {
+      "epoch": 0.299972655181843,
+      "grad_norm": 0.14425235986709595,
+      "learning_rate": 5e-05,
+      "loss": 1.5941,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3002461033634126,
+      "grad_norm": 0.16902539134025574,
+      "learning_rate": 5e-05,
+      "loss": 1.5992,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3005195515449822,
+      "grad_norm": 0.17275625467300415,
+      "learning_rate": 5e-05,
+      "loss": 1.7861,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3007929997265518,
+      "grad_norm": 0.15677183866500854,
+      "learning_rate": 5e-05,
+      "loss": 1.6925,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3010664479081214,
+      "grad_norm": 0.1713411509990692,
+      "learning_rate": 5e-05,
+      "loss": 1.6661,
+      "step": 1101
+    },
+    {
+      "epoch": 0.30133989608969103,
+      "grad_norm": 0.15756134688854218,
+      "learning_rate": 5e-05,
+      "loss": 1.6203,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3016133442712606,
+      "grad_norm": 0.15823891758918762,
+      "learning_rate": 5e-05,
+      "loss": 1.6594,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.15777519345283508,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3021602406343998,
+      "grad_norm": 0.15784233808517456,
+      "learning_rate": 5e-05,
+      "loss": 1.6849,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3024336888159694,
+      "grad_norm": 0.1561124622821808,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 1106
+    },
+    {
+      "epoch": 0.302707136997539,
+      "grad_norm": 0.1538151651620865,
+      "learning_rate": 5e-05,
+      "loss": 1.6838,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3029805851791086,
+      "grad_norm": 0.14938168227672577,
+      "learning_rate": 5e-05,
+      "loss": 1.6129,
+      "step": 1108
+    },
+    {
+      "epoch": 0.30325403336067813,
+      "grad_norm": 0.1569468080997467,
+      "learning_rate": 5e-05,
+      "loss": 1.6651,
+      "step": 1109
+    },
+    {
+      "epoch": 0.30352748154224773,
+      "grad_norm": 0.15352605283260345,
+      "learning_rate": 5e-05,
+      "loss": 1.7078,
+      "step": 1110
+    },
+    {
+      "epoch": 0.30380092972381734,
+      "grad_norm": 0.15539826452732086,
+      "learning_rate": 5e-05,
+      "loss": 1.643,
+      "step": 1111
+    },
+    {
+      "epoch": 0.30407437790538694,
+      "grad_norm": 0.14606958627700806,
+      "learning_rate": 5e-05,
+      "loss": 1.6365,
+      "step": 1112
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.15335801243782043,
+      "learning_rate": 5e-05,
+      "loss": 1.6818,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3046212742685261,
+      "grad_norm": 0.15371185541152954,
+      "learning_rate": 5e-05,
+      "loss": 1.6413,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3048947224500957,
+      "grad_norm": 0.15357623994350433,
+      "learning_rate": 5e-05,
+      "loss": 1.6887,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3051681706316653,
+      "grad_norm": 0.14678330719470978,
+      "learning_rate": 5e-05,
+      "loss": 1.7126,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3054416188132349,
+      "grad_norm": 0.1587059050798416,
+      "learning_rate": 5e-05,
+      "loss": 1.6373,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3057150669948045,
+      "grad_norm": 0.1466260552406311,
+      "learning_rate": 5e-05,
+      "loss": 1.6006,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3059885151763741,
+      "grad_norm": 0.14602194726467133,
+      "learning_rate": 5e-05,
+      "loss": 1.6627,
+      "step": 1119
+    },
+    {
+      "epoch": 0.30626196335794365,
+      "grad_norm": 0.1472383588552475,
+      "learning_rate": 5e-05,
+      "loss": 1.5288,
+      "step": 1120
+    },
+    {
+      "epoch": 0.30653541153951325,
+      "grad_norm": 0.1505843549966812,
+      "learning_rate": 5e-05,
+      "loss": 1.6624,
+      "step": 1121
+    },
+    {
+      "epoch": 0.30680885972108285,
+      "grad_norm": 0.15707598626613617,
+      "learning_rate": 5e-05,
+      "loss": 1.6482,
+      "step": 1122
+    },
+    {
+      "epoch": 0.30708230790265245,
+      "grad_norm": 0.15016111731529236,
+      "learning_rate": 5e-05,
+      "loss": 1.5778,
+      "step": 1123
+    },
+    {
+      "epoch": 0.30735575608422205,
+      "grad_norm": 0.1513780951499939,
+      "learning_rate": 5e-05,
+      "loss": 1.5779,
+      "step": 1124
+    },
+    {
+      "epoch": 0.30762920426579166,
+      "grad_norm": 0.15368494391441345,
+      "learning_rate": 5e-05,
+      "loss": 1.6531,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3079026524473612,
+      "grad_norm": 0.1544412523508072,
+      "learning_rate": 5e-05,
+      "loss": 1.7159,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3081761006289308,
+      "grad_norm": 0.14770282804965973,
+      "learning_rate": 5e-05,
+      "loss": 1.6338,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3084495488105004,
+      "grad_norm": 0.14697682857513428,
+      "learning_rate": 5e-05,
+      "loss": 1.6817,
+      "step": 1128
+    },
+    {
+      "epoch": 0.30872299699207,
+      "grad_norm": 0.15029069781303406,
+      "learning_rate": 5e-05,
+      "loss": 1.7186,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3089964451736396,
+      "grad_norm": 0.15909984707832336,
+      "learning_rate": 5e-05,
+      "loss": 1.7119,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3092698933552092,
+      "grad_norm": 0.14838500320911407,
+      "learning_rate": 5e-05,
+      "loss": 1.615,
+      "step": 1131
+    },
+    {
+      "epoch": 0.30954334153677876,
+      "grad_norm": 0.15336006879806519,
+      "learning_rate": 5e-05,
+      "loss": 1.6544,
+      "step": 1132
+    },
+    {
+      "epoch": 0.30981678971834836,
+      "grad_norm": 0.15481799840927124,
+      "learning_rate": 5e-05,
+      "loss": 1.6279,
+      "step": 1133
+    },
+    {
+      "epoch": 0.31009023789991796,
+      "grad_norm": 0.1593121588230133,
+      "learning_rate": 5e-05,
+      "loss": 1.6697,
+      "step": 1134
+    },
+    {
+      "epoch": 0.31036368608148757,
+      "grad_norm": 0.16930562257766724,
+      "learning_rate": 5e-05,
+      "loss": 1.6706,
+      "step": 1135
+    },
+    {
+      "epoch": 0.31063713426305717,
+      "grad_norm": 0.15615466237068176,
+      "learning_rate": 5e-05,
+      "loss": 1.6018,
+      "step": 1136
+    },
+    {
+      "epoch": 0.31091058244462677,
+      "grad_norm": 0.15975496172904968,
+      "learning_rate": 5e-05,
+      "loss": 1.696,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3111840306261963,
+      "grad_norm": 0.14851494133472443,
+      "learning_rate": 5e-05,
+      "loss": 1.5931,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3114574788077659,
+      "grad_norm": 0.1657806932926178,
+      "learning_rate": 5e-05,
+      "loss": 1.7412,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3117309269893355,
+      "grad_norm": 0.15335924923419952,
+      "learning_rate": 5e-05,
+      "loss": 1.6903,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3120043751709051,
+      "grad_norm": 0.15214680135250092,
+      "learning_rate": 5e-05,
+      "loss": 1.7338,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3122778233524747,
+      "grad_norm": 0.16068771481513977,
+      "learning_rate": 5e-05,
+      "loss": 1.712,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3125512715340443,
+      "grad_norm": 0.15814535319805145,
+      "learning_rate": 5e-05,
+      "loss": 1.7013,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3128247197156139,
+      "grad_norm": 0.1487572193145752,
+      "learning_rate": 5e-05,
+      "loss": 1.6186,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3130981678971835,
+      "grad_norm": 0.15032756328582764,
+      "learning_rate": 5e-05,
+      "loss": 1.6842,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3133716160787531,
+      "grad_norm": 0.15075697004795074,
+      "learning_rate": 5e-05,
+      "loss": 1.536,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3136450642603227,
+      "grad_norm": 0.15556646883487701,
+      "learning_rate": 5e-05,
+      "loss": 1.6043,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3139185124418923,
+      "grad_norm": 0.16194835305213928,
+      "learning_rate": 5e-05,
+      "loss": 1.7423,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31419196062346183,
+      "grad_norm": 0.1608458310365677,
+      "learning_rate": 5e-05,
+      "loss": 1.6599,
+      "step": 1149
+    },
+    {
+      "epoch": 0.31446540880503143,
+      "grad_norm": 0.16816149652004242,
+      "learning_rate": 5e-05,
+      "loss": 1.7054,
+      "step": 1150
+    },
+    {
+      "epoch": 0.31473885698660103,
+      "grad_norm": 0.14813366532325745,
+      "learning_rate": 5e-05,
+      "loss": 1.7635,
+      "step": 1151
+    },
+    {
+      "epoch": 0.31501230516817064,
+      "grad_norm": 0.16501514613628387,
+      "learning_rate": 5e-05,
+      "loss": 1.6414,
+      "step": 1152
+    },
+    {
+      "epoch": 0.31528575334974024,
+      "grad_norm": 0.15683704614639282,
+      "learning_rate": 5e-05,
+      "loss": 1.6307,
+      "step": 1153
+    },
+    {
+      "epoch": 0.31555920153130984,
+      "grad_norm": 0.1599043607711792,
+      "learning_rate": 5e-05,
+      "loss": 1.678,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3158326497128794,
+      "grad_norm": 0.15843378007411957,
+      "learning_rate": 5e-05,
+      "loss": 1.6729,
+      "step": 1155
+    },
+    {
+      "epoch": 0.316106097894449,
+      "grad_norm": 0.14226911962032318,
+      "learning_rate": 5e-05,
+      "loss": 1.584,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3163795460760186,
+      "grad_norm": 0.17751207947731018,
+      "learning_rate": 5e-05,
+      "loss": 1.6461,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3166529942575882,
+      "grad_norm": 0.1599876433610916,
+      "learning_rate": 5e-05,
+      "loss": 1.698,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3169264424391578,
+      "grad_norm": 0.14740917086601257,
+      "learning_rate": 5e-05,
+      "loss": 1.6424,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3171998906207274,
+      "grad_norm": 0.15979237854480743,
+      "learning_rate": 5e-05,
+      "loss": 1.6335,
+      "step": 1160
+    },
+    {
+      "epoch": 0.31747333880229694,
+      "grad_norm": 0.16154277324676514,
+      "learning_rate": 5e-05,
+      "loss": 1.7442,
+      "step": 1161
+    },
+    {
+      "epoch": 0.31774678698386655,
+      "grad_norm": 0.15989361703395844,
+      "learning_rate": 5e-05,
+      "loss": 1.639,
+      "step": 1162
+    },
+    {
+      "epoch": 0.31802023516543615,
+      "grad_norm": 0.15814067423343658,
+      "learning_rate": 5e-05,
+      "loss": 1.7051,
+      "step": 1163
+    },
+    {
+      "epoch": 0.31829368334700575,
+      "grad_norm": 0.15459835529327393,
+      "learning_rate": 5e-05,
+      "loss": 1.5727,
+      "step": 1164
+    },
+    {
+      "epoch": 0.31856713152857535,
+      "grad_norm": 0.15153050422668457,
+      "learning_rate": 5e-05,
+      "loss": 1.6371,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3188405797101449,
+      "grad_norm": 0.1500144898891449,
+      "learning_rate": 5e-05,
+      "loss": 1.6598,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3191140278917145,
+      "grad_norm": 0.15817582607269287,
+      "learning_rate": 5e-05,
+      "loss": 1.6488,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3193874760732841,
+      "grad_norm": 0.15822099149227142,
+      "learning_rate": 5e-05,
+      "loss": 1.5906,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3196609242548537,
+      "grad_norm": 0.14896897971630096,
+      "learning_rate": 5e-05,
+      "loss": 1.6791,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3199343724364233,
+      "grad_norm": 0.1585737019777298,
+      "learning_rate": 5e-05,
+      "loss": 1.7674,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3202078206179929,
+      "grad_norm": 0.16357015073299408,
+      "learning_rate": 5e-05,
+      "loss": 1.6621,
+      "step": 1171
+    },
+    {
+      "epoch": 0.32048126879956246,
+      "grad_norm": 0.14537674188613892,
+      "learning_rate": 5e-05,
+      "loss": 1.5276,
+      "step": 1172
+    },
+    {
+      "epoch": 0.32075471698113206,
+      "grad_norm": 0.15648779273033142,
+      "learning_rate": 5e-05,
+      "loss": 1.6323,
+      "step": 1173
+    },
+    {
+      "epoch": 0.32102816516270166,
+      "grad_norm": 0.14875225722789764,
+      "learning_rate": 5e-05,
+      "loss": 1.6257,
+      "step": 1174
+    },
+    {
+      "epoch": 0.32130161334427126,
+      "grad_norm": 0.14988994598388672,
+      "learning_rate": 5e-05,
+      "loss": 1.642,
+      "step": 1175
+    },
+    {
+      "epoch": 0.32157506152584087,
+      "grad_norm": 0.14954978227615356,
+      "learning_rate": 5e-05,
+      "loss": 1.6477,
+      "step": 1176
+    },
+    {
+      "epoch": 0.32184850970741047,
+      "grad_norm": 0.15913893282413483,
+      "learning_rate": 5e-05,
+      "loss": 1.7523,
+      "step": 1177
+    },
+    {
+      "epoch": 0.32212195788898,
+      "grad_norm": 0.17450571060180664,
+      "learning_rate": 5e-05,
+      "loss": 1.7723,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3223954060705496,
+      "grad_norm": 0.14600080251693726,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3226688542521192,
+      "grad_norm": 0.14759889245033264,
+      "learning_rate": 5e-05,
+      "loss": 1.5792,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3229423024336888,
+      "grad_norm": 0.17649635672569275,
+      "learning_rate": 5e-05,
+      "loss": 1.7302,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3232157506152584,
+      "grad_norm": 0.15906043350696564,
+      "learning_rate": 5e-05,
+      "loss": 1.6046,
+      "step": 1182
+    },
+    {
+      "epoch": 0.323489198796828,
+      "grad_norm": 0.14562730491161346,
+      "learning_rate": 5e-05,
+      "loss": 1.566,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3237626469783976,
+      "grad_norm": 0.15766160190105438,
+      "learning_rate": 5e-05,
+      "loss": 1.6397,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3240360951599672,
+      "grad_norm": 0.16190050542354584,
+      "learning_rate": 5e-05,
+      "loss": 1.5876,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3243095433415368,
+      "grad_norm": 0.15443246066570282,
+      "learning_rate": 5e-05,
+      "loss": 1.6817,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3245829915231064,
+      "grad_norm": 0.1686965525150299,
+      "learning_rate": 5e-05,
+      "loss": 1.5725,
+      "step": 1187
+    },
+    {
+      "epoch": 0.324856439704676,
+      "grad_norm": 0.16837440431118011,
+      "learning_rate": 5e-05,
+      "loss": 1.681,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3251298878862456,
+      "grad_norm": 0.15975917875766754,
+      "learning_rate": 5e-05,
+      "loss": 1.6988,
+      "step": 1189
+    },
+    {
+      "epoch": 0.32540333606781513,
+      "grad_norm": 0.16166743636131287,
+      "learning_rate": 5e-05,
+      "loss": 1.6254,
+      "step": 1190
+    },
+    {
+      "epoch": 0.32567678424938473,
+      "grad_norm": 0.15348750352859497,
+      "learning_rate": 5e-05,
+      "loss": 1.6757,
+      "step": 1191
+    },
+    {
+      "epoch": 0.32595023243095433,
+      "grad_norm": 0.16833312809467316,
+      "learning_rate": 5e-05,
+      "loss": 1.6262,
+      "step": 1192
+    },
+    {
+      "epoch": 0.32622368061252394,
+      "grad_norm": 0.1633581817150116,
+      "learning_rate": 5e-05,
+      "loss": 1.7312,
+      "step": 1193
+    },
+    {
+      "epoch": 0.32649712879409354,
+      "grad_norm": 0.14953148365020752,
+      "learning_rate": 5e-05,
+      "loss": 1.5993,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3267705769756631,
+      "grad_norm": 0.1685009002685547,
+      "learning_rate": 5e-05,
+      "loss": 1.7026,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3270440251572327,
+      "grad_norm": 0.1520211100578308,
+      "learning_rate": 5e-05,
+      "loss": 1.6178,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3273174733388023,
+      "grad_norm": 0.15548743307590485,
+      "learning_rate": 5e-05,
+      "loss": 1.6335,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3275909215203719,
+      "grad_norm": 0.1540190726518631,
+      "learning_rate": 5e-05,
+      "loss": 1.7661,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3278643697019415,
+      "grad_norm": 0.15100663900375366,
+      "learning_rate": 5e-05,
+      "loss": 1.5361,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3281378178835111,
+      "grad_norm": 0.15432646870613098,
+      "learning_rate": 5e-05,
+      "loss": 1.7029,
+      "step": 1200
+    },
+    {
+      "epoch": 0.32841126606508064,
+      "grad_norm": 0.15209780633449554,
+      "learning_rate": 5e-05,
+      "loss": 1.6857,
+      "step": 1201
+    },
+    {
+      "epoch": 0.32868471424665024,
+      "grad_norm": 0.1498226374387741,
+      "learning_rate": 5e-05,
+      "loss": 1.7318,
+      "step": 1202
+    },
+    {
+      "epoch": 0.32895816242821985,
+      "grad_norm": 0.149881511926651,
+      "learning_rate": 5e-05,
+      "loss": 1.6591,
+      "step": 1203
+    },
+    {
+      "epoch": 0.32923161060978945,
+      "grad_norm": 0.15110871195793152,
+      "learning_rate": 5e-05,
+      "loss": 1.6238,
+      "step": 1204
+    },
+    {
+      "epoch": 0.32950505879135905,
+      "grad_norm": 0.1504354178905487,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 1205
+    },
+    {
+      "epoch": 0.32977850697292865,
+      "grad_norm": 0.16923676431179047,
+      "learning_rate": 5e-05,
+      "loss": 1.7914,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3300519551544982,
+      "grad_norm": 0.17524860799312592,
+      "learning_rate": 5e-05,
+      "loss": 1.6875,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3303254033360678,
+      "grad_norm": 0.1538068652153015,
+      "learning_rate": 5e-05,
+      "loss": 1.6375,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3305988515176374,
+      "grad_norm": 0.15115374326705933,
+      "learning_rate": 5e-05,
+      "loss": 1.6759,
+      "step": 1209
+    },
+    {
+      "epoch": 0.330872299699207,
+      "grad_norm": 0.1774495393037796,
+      "learning_rate": 5e-05,
+      "loss": 1.6865,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3311457478807766,
+      "grad_norm": 0.16683407127857208,
+      "learning_rate": 5e-05,
+      "loss": 1.6031,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3314191960623462,
+      "grad_norm": 0.16064640879631042,
+      "learning_rate": 5e-05,
+      "loss": 1.7608,
+      "step": 1212
+    },
+    {
+      "epoch": 0.33169264424391576,
+      "grad_norm": 0.16114073991775513,
+      "learning_rate": 5e-05,
+      "loss": 1.6658,
+      "step": 1213
+    },
+    {
+      "epoch": 0.33196609242548536,
+      "grad_norm": 0.161068856716156,
+      "learning_rate": 5e-05,
+      "loss": 1.6989,
+      "step": 1214
+    },
+    {
+      "epoch": 0.33223954060705496,
+      "grad_norm": 0.155606210231781,
+      "learning_rate": 5e-05,
+      "loss": 1.6235,
+      "step": 1215
+    },
+    {
+      "epoch": 0.33251298878862456,
+      "grad_norm": 0.18368175625801086,
+      "learning_rate": 5e-05,
+      "loss": 1.6516,
+      "step": 1216
+    },
+    {
+      "epoch": 0.33278643697019417,
+      "grad_norm": 0.1500716358423233,
+      "learning_rate": 5e-05,
+      "loss": 1.6361,
+      "step": 1217
+    },
+    {
+      "epoch": 0.33305988515176377,
+      "grad_norm": 0.15658661723136902,
+      "learning_rate": 5e-05,
+      "loss": 1.5645,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.16541090607643127,
+      "learning_rate": 5e-05,
+      "loss": 1.7556,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3336067815149029,
+      "grad_norm": 0.15288498997688293,
+      "learning_rate": 5e-05,
+      "loss": 1.6349,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3338802296964725,
+      "grad_norm": 0.16073960065841675,
+      "learning_rate": 5e-05,
+      "loss": 1.6157,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3341536778780421,
+      "grad_norm": 0.1543978601694107,
+      "learning_rate": 5e-05,
+      "loss": 1.6628,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3344271260596117,
+      "grad_norm": 0.14811809360980988,
+      "learning_rate": 5e-05,
+      "loss": 1.5668,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33470057424118127,
+      "grad_norm": 0.15395627915859222,
+      "learning_rate": 5e-05,
+      "loss": 1.6593,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3349740224227509,
+      "grad_norm": 0.16432489454746246,
+      "learning_rate": 5e-05,
+      "loss": 1.6334,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3352474706043205,
+      "grad_norm": 0.15168853104114532,
+      "learning_rate": 5e-05,
+      "loss": 1.6247,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3355209187858901,
+      "grad_norm": 0.15608245134353638,
+      "learning_rate": 5e-05,
+      "loss": 1.6651,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3357943669674597,
+      "grad_norm": 0.16598603129386902,
+      "learning_rate": 5e-05,
+      "loss": 1.7173,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3360678151490293,
+      "grad_norm": 0.14476749300956726,
+      "learning_rate": 5e-05,
+      "loss": 1.6363,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3363412633305988,
+      "grad_norm": 0.16102361679077148,
+      "learning_rate": 5e-05,
+      "loss": 1.6763,
+      "step": 1230
+    },
+    {
+      "epoch": 0.33661471151216843,
+      "grad_norm": 0.16813768446445465,
+      "learning_rate": 5e-05,
+      "loss": 1.6754,
+      "step": 1231
+    },
+    {
+      "epoch": 0.33688815969373803,
+      "grad_norm": 0.15631164610385895,
+      "learning_rate": 5e-05,
+      "loss": 1.6529,
+      "step": 1232
+    },
+    {
+      "epoch": 0.33716160787530763,
+      "grad_norm": 0.1545805037021637,
+      "learning_rate": 5e-05,
+      "loss": 1.7156,
+      "step": 1233
+    },
+    {
+      "epoch": 0.33743505605687724,
+      "grad_norm": 0.17586275935173035,
+      "learning_rate": 5e-05,
+      "loss": 1.6615,
+      "step": 1234
+    },
+    {
+      "epoch": 0.33770850423844684,
+      "grad_norm": 0.15202665328979492,
+      "learning_rate": 5e-05,
+      "loss": 1.6291,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3379819524200164,
+      "grad_norm": 0.16369853913784027,
+      "learning_rate": 5e-05,
+      "loss": 1.6827,
+      "step": 1236
+    },
+    {
+      "epoch": 0.338255400601586,
+      "grad_norm": 0.15997561812400818,
+      "learning_rate": 5e-05,
+      "loss": 1.7339,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3385288487831556,
+      "grad_norm": 0.14640896022319794,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3388022969647252,
+      "grad_norm": 0.1671455204486847,
+      "learning_rate": 5e-05,
+      "loss": 1.6769,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3390757451462948,
+      "grad_norm": 0.16016924381256104,
+      "learning_rate": 5e-05,
+      "loss": 1.7308,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3393491933278644,
+      "grad_norm": 0.16020850837230682,
+      "learning_rate": 5e-05,
+      "loss": 1.7288,
+      "step": 1241
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.14866626262664795,
+      "learning_rate": 5e-05,
+      "loss": 1.5764,
+      "step": 1242
+    },
+    {
+      "epoch": 0.33989608969100354,
+      "grad_norm": 0.17916180193424225,
+      "learning_rate": 5e-05,
+      "loss": 1.7789,
+      "step": 1243
+    },
+    {
+      "epoch": 0.34016953787257315,
+      "grad_norm": 0.14788402616977692,
+      "learning_rate": 5e-05,
+      "loss": 1.63,
+      "step": 1244
+    },
+    {
+      "epoch": 0.34044298605414275,
+      "grad_norm": 0.15443041920661926,
+      "learning_rate": 5e-05,
+      "loss": 1.6499,
+      "step": 1245
+    },
+    {
+      "epoch": 0.34071643423571235,
+      "grad_norm": 0.15852518379688263,
+      "learning_rate": 5e-05,
+      "loss": 1.6446,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3409898824172819,
+      "grad_norm": 0.15751297771930695,
+      "learning_rate": 5e-05,
+      "loss": 1.6294,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3412633305988515,
+      "grad_norm": 0.15762586891651154,
+      "learning_rate": 5e-05,
+      "loss": 1.6864,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3415367787804211,
+      "grad_norm": 0.15331421792507172,
+      "learning_rate": 5e-05,
+      "loss": 1.6898,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3418102269619907,
+      "grad_norm": 0.165785014629364,
+      "learning_rate": 5e-05,
+      "loss": 1.703,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3420836751435603,
+      "grad_norm": 0.1514601707458496,
+      "learning_rate": 5e-05,
+      "loss": 1.6489,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3423571233251299,
+      "grad_norm": 0.187610924243927,
+      "learning_rate": 5e-05,
+      "loss": 1.6109,
+      "step": 1252
+    },
+    {
+      "epoch": 0.34263057150669946,
+      "grad_norm": 0.1479722261428833,
+      "learning_rate": 5e-05,
+      "loss": 1.6528,
+      "step": 1253
+    },
+    {
+      "epoch": 0.34290401968826906,
+      "grad_norm": 0.1570201814174652,
+      "learning_rate": 5e-05,
+      "loss": 1.6706,
+      "step": 1254
+    },
+    {
+      "epoch": 0.34317746786983866,
+      "grad_norm": 0.1705133467912674,
+      "learning_rate": 5e-05,
+      "loss": 1.7262,
+      "step": 1255
+    },
+    {
+      "epoch": 0.34345091605140826,
+      "grad_norm": 0.15776453912258148,
+      "learning_rate": 5e-05,
+      "loss": 1.6374,
+      "step": 1256
+    },
+    {
+      "epoch": 0.34372436423297786,
+      "grad_norm": 0.16335074603557587,
+      "learning_rate": 5e-05,
+      "loss": 1.6261,
+      "step": 1257
+    },
+    {
+      "epoch": 0.34399781241454747,
+      "grad_norm": 0.17720459401607513,
+      "learning_rate": 5e-05,
+      "loss": 1.5963,
+      "step": 1258
+    },
+    {
+      "epoch": 0.344271260596117,
+      "grad_norm": 0.14796927571296692,
+      "learning_rate": 5e-05,
+      "loss": 1.5923,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3445447087776866,
+      "grad_norm": 0.16118891537189484,
+      "learning_rate": 5e-05,
+      "loss": 1.5986,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3448181569592562,
+      "grad_norm": 0.16453957557678223,
+      "learning_rate": 5e-05,
+      "loss": 1.7498,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3450916051408258,
+      "grad_norm": 0.15277455747127533,
+      "learning_rate": 5e-05,
+      "loss": 1.6196,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3453650533223954,
+      "grad_norm": 0.1496528536081314,
+      "learning_rate": 5e-05,
+      "loss": 1.635,
+      "step": 1263
+    },
+    {
+      "epoch": 0.345638501503965,
+      "grad_norm": 0.1616893708705902,
+      "learning_rate": 5e-05,
+      "loss": 1.591,
+      "step": 1264
+    },
+    {
+      "epoch": 0.34591194968553457,
+      "grad_norm": 0.1540791094303131,
+      "learning_rate": 5e-05,
+      "loss": 1.6713,
+      "step": 1265
+    },
+    {
+      "epoch": 0.34618539786710417,
+      "grad_norm": 0.15063992142677307,
+      "learning_rate": 5e-05,
+      "loss": 1.6739,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3464588460486738,
+      "grad_norm": 0.15687578916549683,
+      "learning_rate": 5e-05,
+      "loss": 1.7105,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3467322942302434,
+      "grad_norm": 0.14987704157829285,
+      "learning_rate": 5e-05,
+      "loss": 1.7308,
+      "step": 1268
+    },
+    {
+      "epoch": 0.347005742411813,
+      "grad_norm": 0.1480540633201599,
+      "learning_rate": 5e-05,
+      "loss": 1.7043,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3472791905933826,
+      "grad_norm": 0.1529364138841629,
+      "learning_rate": 5e-05,
+      "loss": 1.5753,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3475526387749521,
+      "grad_norm": 0.14511734247207642,
+      "learning_rate": 5e-05,
+      "loss": 1.5622,
+      "step": 1271
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.14741328358650208,
+      "learning_rate": 5e-05,
+      "loss": 1.6804,
+      "step": 1272
+    },
+    {
+      "epoch": 0.34809953513809133,
+      "grad_norm": 0.14820526540279388,
+      "learning_rate": 5e-05,
+      "loss": 1.5904,
+      "step": 1273
+    },
+    {
+      "epoch": 0.34837298331966093,
+      "grad_norm": 0.1521809697151184,
+      "learning_rate": 5e-05,
+      "loss": 1.7002,
+      "step": 1274
+    },
+    {
+      "epoch": 0.34864643150123054,
+      "grad_norm": 0.1577511727809906,
+      "learning_rate": 5e-05,
+      "loss": 1.6897,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3489198796828001,
+      "grad_norm": 0.14861604571342468,
+      "learning_rate": 5e-05,
+      "loss": 1.6671,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3491933278643697,
+      "grad_norm": 0.15119168162345886,
+      "learning_rate": 5e-05,
+      "loss": 1.6387,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3494667760459393,
+      "grad_norm": 0.1554817259311676,
+      "learning_rate": 5e-05,
+      "loss": 1.5855,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3497402242275089,
+      "grad_norm": 0.151669442653656,
+      "learning_rate": 5e-05,
+      "loss": 1.6634,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3500136724090785,
+      "grad_norm": 0.1499747335910797,
+      "learning_rate": 5e-05,
+      "loss": 1.627,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3502871205906481,
+      "grad_norm": 0.15303084254264832,
+      "learning_rate": 5e-05,
+      "loss": 1.6594,
+      "step": 1281
+    },
+    {
+      "epoch": 0.35056056877221764,
+      "grad_norm": 0.1563277691602707,
+      "learning_rate": 5e-05,
+      "loss": 1.8,
+      "step": 1282
+    },
+    {
+      "epoch": 0.35083401695378724,
+      "grad_norm": 0.16208653151988983,
+      "learning_rate": 5e-05,
+      "loss": 1.7052,
+      "step": 1283
+    },
+    {
+      "epoch": 0.35110746513535684,
+      "grad_norm": 0.14594794809818268,
+      "learning_rate": 5e-05,
+      "loss": 1.5883,
+      "step": 1284
+    },
+    {
+      "epoch": 0.35138091331692645,
+      "grad_norm": 0.1515040099620819,
+      "learning_rate": 5e-05,
+      "loss": 1.681,
+      "step": 1285
+    },
+    {
+      "epoch": 0.35165436149849605,
+      "grad_norm": 0.15303464233875275,
+      "learning_rate": 5e-05,
+      "loss": 1.6776,
+      "step": 1286
+    },
+    {
+      "epoch": 0.35192780968006565,
+      "grad_norm": 0.15074992179870605,
+      "learning_rate": 5e-05,
+      "loss": 1.6131,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3522012578616352,
+      "grad_norm": 0.15136325359344482,
+      "learning_rate": 5e-05,
+      "loss": 1.6393,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3524747060432048,
+      "grad_norm": 0.161278635263443,
+      "learning_rate": 5e-05,
+      "loss": 1.5921,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3527481542247744,
+      "grad_norm": 0.16516782343387604,
+      "learning_rate": 5e-05,
+      "loss": 1.6514,
+      "step": 1290
+    },
+    {
+      "epoch": 0.353021602406344,
+      "grad_norm": 0.1504773199558258,
+      "learning_rate": 5e-05,
+      "loss": 1.6053,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3532950505879136,
+      "grad_norm": 0.15545381605625153,
+      "learning_rate": 5e-05,
+      "loss": 1.6231,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3535684987694832,
+      "grad_norm": 0.14545585215091705,
+      "learning_rate": 5e-05,
+      "loss": 1.6238,
+      "step": 1293
+    },
+    {
+      "epoch": 0.35384194695105275,
+      "grad_norm": 0.15290163457393646,
+      "learning_rate": 5e-05,
+      "loss": 1.6625,
+      "step": 1294
+    },
+    {
+      "epoch": 0.35411539513262236,
+      "grad_norm": 0.15580037236213684,
+      "learning_rate": 5e-05,
+      "loss": 1.6871,
+      "step": 1295
+    },
+    {
+      "epoch": 0.35438884331419196,
+      "grad_norm": 0.15643970668315887,
+      "learning_rate": 5e-05,
+      "loss": 1.6451,
+      "step": 1296
+    },
+    {
+      "epoch": 0.35466229149576156,
+      "grad_norm": 0.14689311385154724,
+      "learning_rate": 5e-05,
+      "loss": 1.612,
+      "step": 1297
+    },
+    {
+      "epoch": 0.35493573967733116,
+      "grad_norm": 0.152411088347435,
+      "learning_rate": 5e-05,
+      "loss": 1.5722,
+      "step": 1298
+    },
+    {
+      "epoch": 0.35520918785890077,
+      "grad_norm": 0.16959436237812042,
+      "learning_rate": 5e-05,
+      "loss": 1.7114,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3554826360404703,
+      "grad_norm": 0.15447081625461578,
+      "learning_rate": 5e-05,
+      "loss": 1.7206,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3557560842220399,
+      "grad_norm": 0.17055825889110565,
+      "learning_rate": 5e-05,
+      "loss": 1.5612,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3560295324036095,
+      "grad_norm": 0.157841756939888,
+      "learning_rate": 5e-05,
+      "loss": 1.6815,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3563029805851791,
+      "grad_norm": 0.17073825001716614,
+      "learning_rate": 5e-05,
+      "loss": 1.6889,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3565764287667487,
+      "grad_norm": 0.17082183063030243,
+      "learning_rate": 5e-05,
+      "loss": 1.6979,
+      "step": 1304
+    },
+    {
+      "epoch": 0.35684987694831827,
+      "grad_norm": 0.15137112140655518,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 1305
+    },
+    {
+      "epoch": 0.35712332512988787,
+      "grad_norm": 0.15363116562366486,
+      "learning_rate": 5e-05,
+      "loss": 1.6242,
+      "step": 1306
+    },
+    {
+      "epoch": 0.35739677331145747,
+      "grad_norm": 0.15610884130001068,
+      "learning_rate": 5e-05,
+      "loss": 1.5751,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3576702214930271,
+      "grad_norm": 0.15649494528770447,
+      "learning_rate": 5e-05,
+      "loss": 1.6683,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3579436696745967,
+      "grad_norm": 0.1602485328912735,
+      "learning_rate": 5e-05,
+      "loss": 1.7109,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3582171178561663,
+      "grad_norm": 0.16601736843585968,
+      "learning_rate": 5e-05,
+      "loss": 1.6664,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3584905660377358,
+      "grad_norm": 0.15012311935424805,
+      "learning_rate": 5e-05,
+      "loss": 1.5708,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3587640142193054,
+      "grad_norm": 0.16618122160434723,
+      "learning_rate": 5e-05,
+      "loss": 1.7141,
+      "step": 1312
+    },
+    {
+      "epoch": 0.35903746240087503,
+      "grad_norm": 0.1614745706319809,
+      "learning_rate": 5e-05,
+      "loss": 1.6907,
+      "step": 1313
+    },
+    {
+      "epoch": 0.35931091058244463,
+      "grad_norm": 0.1518169492483139,
+      "learning_rate": 5e-05,
+      "loss": 1.6233,
+      "step": 1314
+    },
+    {
+      "epoch": 0.35958435876401423,
+      "grad_norm": 0.16219857335090637,
+      "learning_rate": 5e-05,
+      "loss": 1.6707,
+      "step": 1315
+    },
+    {
+      "epoch": 0.35985780694558384,
+      "grad_norm": 0.15628166496753693,
+      "learning_rate": 5e-05,
+      "loss": 1.7647,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3601312551271534,
+      "grad_norm": 0.14191019535064697,
+      "learning_rate": 5e-05,
+      "loss": 1.5061,
+      "step": 1317
+    },
+    {
+      "epoch": 0.360404703308723,
+      "grad_norm": 0.148356094956398,
+      "learning_rate": 5e-05,
+      "loss": 1.6642,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3606781514902926,
+      "grad_norm": 0.14994607865810394,
+      "learning_rate": 5e-05,
+      "loss": 1.655,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3609515996718622,
+      "grad_norm": 0.15752872824668884,
+      "learning_rate": 5e-05,
+      "loss": 1.5816,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3612250478534318,
+      "grad_norm": 0.15151529014110565,
+      "learning_rate": 5e-05,
+      "loss": 1.6372,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3614984960350014,
+      "grad_norm": 0.15662340819835663,
+      "learning_rate": 5e-05,
+      "loss": 1.6894,
+      "step": 1322
+    },
+    {
+      "epoch": 0.36177194421657094,
+      "grad_norm": 0.16049139201641083,
+      "learning_rate": 5e-05,
+      "loss": 1.616,
+      "step": 1323
+    },
+    {
+      "epoch": 0.36204539239814054,
+      "grad_norm": 0.15439613163471222,
+      "learning_rate": 5e-05,
+      "loss": 1.6672,
+      "step": 1324
+    },
+    {
+      "epoch": 0.36231884057971014,
+      "grad_norm": 0.15884248912334442,
+      "learning_rate": 5e-05,
+      "loss": 1.6366,
+      "step": 1325
+    },
+    {
+      "epoch": 0.36259228876127975,
+      "grad_norm": 0.15194907784461975,
+      "learning_rate": 5e-05,
+      "loss": 1.7004,
+      "step": 1326
+    },
+    {
+      "epoch": 0.36286573694284935,
+      "grad_norm": 0.1542571783065796,
+      "learning_rate": 5e-05,
+      "loss": 1.6714,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3631391851244189,
+      "grad_norm": 0.15640923380851746,
+      "learning_rate": 5e-05,
+      "loss": 1.7018,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3634126333059885,
+      "grad_norm": 0.15510603785514832,
+      "learning_rate": 5e-05,
+      "loss": 1.6173,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3636860814875581,
+      "grad_norm": 0.14958932995796204,
+      "learning_rate": 5e-05,
+      "loss": 1.6507,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3639595296691277,
+      "grad_norm": 0.15672361850738525,
+      "learning_rate": 5e-05,
+      "loss": 1.6679,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3642329778506973,
+      "grad_norm": 0.16806292533874512,
+      "learning_rate": 5e-05,
+      "loss": 1.6097,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3645064260322669,
+      "grad_norm": 0.14968685805797577,
+      "learning_rate": 5e-05,
+      "loss": 1.6181,
+      "step": 1333
+    },
+    {
+      "epoch": 0.36477987421383645,
+      "grad_norm": 0.14979314804077148,
+      "learning_rate": 5e-05,
+      "loss": 1.705,
+      "step": 1334
+    },
+    {
+      "epoch": 0.36505332239540605,
+      "grad_norm": 0.14740002155303955,
+      "learning_rate": 5e-05,
+      "loss": 1.6532,
+      "step": 1335
+    },
+    {
+      "epoch": 0.36532677057697566,
+      "grad_norm": 0.15307697653770447,
+      "learning_rate": 5e-05,
+      "loss": 1.5616,
+      "step": 1336
+    },
+    {
+      "epoch": 0.36560021875854526,
+      "grad_norm": 0.165946364402771,
+      "learning_rate": 5e-05,
+      "loss": 1.7643,
+      "step": 1337
+    },
+    {
+      "epoch": 0.36587366694011486,
+      "grad_norm": 0.1671787053346634,
+      "learning_rate": 5e-05,
+      "loss": 1.7023,
+      "step": 1338
+    },
+    {
+      "epoch": 0.36614711512168446,
+      "grad_norm": 0.1597784012556076,
+      "learning_rate": 5e-05,
+      "loss": 1.6975,
+      "step": 1339
+    },
+    {
+      "epoch": 0.366420563303254,
+      "grad_norm": 0.1569102704524994,
+      "learning_rate": 5e-05,
+      "loss": 1.7043,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3666940114848236,
+      "grad_norm": 0.1625957041978836,
+      "learning_rate": 5e-05,
+      "loss": 1.6712,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3669674596663932,
+      "grad_norm": 0.16208802163600922,
+      "learning_rate": 5e-05,
+      "loss": 1.6891,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3672409078479628,
+      "grad_norm": 0.1636815369129181,
+      "learning_rate": 5e-05,
+      "loss": 1.7703,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3675143560295324,
+      "grad_norm": 0.1954984962940216,
+      "learning_rate": 5e-05,
+      "loss": 1.6462,
+      "step": 1344
+    },
+    {
+      "epoch": 0.367787804211102,
+      "grad_norm": 0.16720376908779144,
+      "learning_rate": 5e-05,
+      "loss": 1.6952,
+      "step": 1345
+    },
+    {
+      "epoch": 0.36806125239267157,
+      "grad_norm": 0.1478559672832489,
+      "learning_rate": 5e-05,
+      "loss": 1.6585,
+      "step": 1346
+    },
+    {
+      "epoch": 0.36833470057424117,
+      "grad_norm": 0.18975849449634552,
+      "learning_rate": 5e-05,
+      "loss": 1.6765,
+      "step": 1347
+    },
+    {
+      "epoch": 0.36860814875581077,
+      "grad_norm": 0.1447249799966812,
+      "learning_rate": 5e-05,
+      "loss": 1.633,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3688815969373804,
+      "grad_norm": 0.19498814642429352,
+      "learning_rate": 5e-05,
+      "loss": 1.7036,
+      "step": 1349
+    },
+    {
+      "epoch": 0.36915504511895,
+      "grad_norm": 0.16138319671154022,
+      "learning_rate": 5e-05,
+      "loss": 1.6062,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3694284933005196,
+      "grad_norm": 0.15296588838100433,
+      "learning_rate": 5e-05,
+      "loss": 1.6813,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3697019414820891,
+      "grad_norm": 0.1728629469871521,
+      "learning_rate": 5e-05,
+      "loss": 1.723,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3699753896636587,
+      "grad_norm": 0.17153020203113556,
+      "learning_rate": 5e-05,
+      "loss": 1.6255,
+      "step": 1353
+    },
+    {
+      "epoch": 0.37024883784522833,
+      "grad_norm": 0.1482992023229599,
+      "learning_rate": 5e-05,
+      "loss": 1.5935,
+      "step": 1354
+    },
+    {
+      "epoch": 0.37052228602679793,
+      "grad_norm": 0.19278480112552643,
+      "learning_rate": 5e-05,
+      "loss": 1.7194,
+      "step": 1355
+    },
+    {
+      "epoch": 0.37079573420836753,
+      "grad_norm": 0.15969671308994293,
+      "learning_rate": 5e-05,
+      "loss": 1.7102,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3710691823899371,
+      "grad_norm": 0.1841089278459549,
+      "learning_rate": 5e-05,
+      "loss": 1.6842,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3713426305715067,
+      "grad_norm": 0.1696590632200241,
+      "learning_rate": 5e-05,
+      "loss": 1.6301,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3716160787530763,
+      "grad_norm": 0.15877504646778107,
+      "learning_rate": 5e-05,
+      "loss": 1.6211,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3718895269346459,
+      "grad_norm": 0.17480207979679108,
+      "learning_rate": 5e-05,
+      "loss": 1.5707,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3721629751162155,
+      "grad_norm": 0.1526334136724472,
+      "learning_rate": 5e-05,
+      "loss": 1.6062,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3724364232977851,
+      "grad_norm": 0.1520829051733017,
+      "learning_rate": 5e-05,
+      "loss": 1.713,
+      "step": 1362
+    },
+    {
+      "epoch": 0.37270987147935464,
+      "grad_norm": 0.17460103332996368,
+      "learning_rate": 5e-05,
+      "loss": 1.5544,
+      "step": 1363
+    },
+    {
+      "epoch": 0.37298331966092424,
+      "grad_norm": 0.16600289940834045,
+      "learning_rate": 5e-05,
+      "loss": 1.6942,
+      "step": 1364
+    },
+    {
+      "epoch": 0.37325676784249384,
+      "grad_norm": 0.15874885022640228,
+      "learning_rate": 5e-05,
+      "loss": 1.5408,
+      "step": 1365
+    },
+    {
+      "epoch": 0.37353021602406344,
+      "grad_norm": 0.15118831396102905,
+      "learning_rate": 5e-05,
+      "loss": 1.6963,
+      "step": 1366
+    },
+    {
+      "epoch": 0.37380366420563305,
+      "grad_norm": 0.16530974209308624,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 1367
+    },
+    {
+      "epoch": 0.37407711238720265,
+      "grad_norm": 0.16382399201393127,
+      "learning_rate": 5e-05,
+      "loss": 1.5668,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3743505605687722,
+      "grad_norm": 0.16075266897678375,
+      "learning_rate": 5e-05,
+      "loss": 1.5714,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3746240087503418,
+      "grad_norm": 0.16619382798671722,
+      "learning_rate": 5e-05,
+      "loss": 1.6636,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3748974569319114,
+      "grad_norm": 0.16388960182666779,
+      "learning_rate": 5e-05,
+      "loss": 1.5565,
+      "step": 1371
+    },
+    {
+      "epoch": 0.375170905113481,
+      "grad_norm": 0.17126598954200745,
+      "learning_rate": 5e-05,
+      "loss": 1.6027,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3754443532950506,
+      "grad_norm": 0.15221962332725525,
+      "learning_rate": 5e-05,
+      "loss": 1.7189,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3757178014766202,
+      "grad_norm": 0.1918559968471527,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 1374
+    },
+    {
+      "epoch": 0.37599124965818975,
+      "grad_norm": 0.1808595359325409,
+      "learning_rate": 5e-05,
+      "loss": 1.7059,
+      "step": 1375
+    },
+    {
+      "epoch": 0.37626469783975935,
+      "grad_norm": 0.16812893748283386,
+      "learning_rate": 5e-05,
+      "loss": 1.8234,
+      "step": 1376
+    },
+    {
+      "epoch": 0.37653814602132896,
+      "grad_norm": 0.17743167281150818,
+      "learning_rate": 5e-05,
+      "loss": 1.6533,
+      "step": 1377
+    },
+    {
+      "epoch": 0.37681159420289856,
+      "grad_norm": 0.16644933819770813,
+      "learning_rate": 5e-05,
+      "loss": 1.6681,
+      "step": 1378
+    },
+    {
+      "epoch": 0.37708504238446816,
+      "grad_norm": 0.1525644063949585,
+      "learning_rate": 5e-05,
+      "loss": 1.6302,
+      "step": 1379
+    },
+    {
+      "epoch": 0.37735849056603776,
+      "grad_norm": 0.1539047807455063,
+      "learning_rate": 5e-05,
+      "loss": 1.6559,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3776319387476073,
+      "grad_norm": 0.17377246916294098,
+      "learning_rate": 5e-05,
+      "loss": 1.6563,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3779053869291769,
+      "grad_norm": 0.1529269814491272,
+      "learning_rate": 5e-05,
+      "loss": 1.6278,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3781788351107465,
+      "grad_norm": 0.17307884991168976,
+      "learning_rate": 5e-05,
+      "loss": 1.8326,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3784522832923161,
+      "grad_norm": 0.15691839158535004,
+      "learning_rate": 5e-05,
+      "loss": 1.652,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3787257314738857,
+      "grad_norm": 0.16453763842582703,
+      "learning_rate": 5e-05,
+      "loss": 1.6961,
+      "step": 1385
+    },
+    {
+      "epoch": 0.37899917965545526,
+      "grad_norm": 0.17871487140655518,
+      "learning_rate": 5e-05,
+      "loss": 1.6779,
+      "step": 1386
+    },
+    {
+      "epoch": 0.37927262783702487,
+      "grad_norm": 0.15350158512592316,
+      "learning_rate": 5e-05,
+      "loss": 1.6196,
+      "step": 1387
+    },
+    {
+      "epoch": 0.37954607601859447,
+      "grad_norm": 0.1600034087896347,
+      "learning_rate": 5e-05,
+      "loss": 1.6356,
+      "step": 1388
+    },
+    {
+      "epoch": 0.37981952420016407,
+      "grad_norm": 0.17216049134731293,
+      "learning_rate": 5e-05,
+      "loss": 1.6116,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3800929723817337,
+      "grad_norm": 0.15600605309009552,
+      "learning_rate": 5e-05,
+      "loss": 1.6993,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3803664205633033,
+      "grad_norm": 0.14978420734405518,
+      "learning_rate": 5e-05,
+      "loss": 1.7346,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3806398687448728,
+      "grad_norm": 0.176108255982399,
+      "learning_rate": 5e-05,
+      "loss": 1.6644,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3809133169264424,
+      "grad_norm": 0.1513052135705948,
+      "learning_rate": 5e-05,
+      "loss": 1.648,
+      "step": 1393
+    },
+    {
+      "epoch": 0.381186765108012,
+      "grad_norm": 0.16158834099769592,
+      "learning_rate": 5e-05,
+      "loss": 1.6868,
+      "step": 1394
+    },
+    {
+      "epoch": 0.38146021328958163,
+      "grad_norm": 0.1669863909482956,
+      "learning_rate": 5e-05,
+      "loss": 1.6286,
+      "step": 1395
+    },
+    {
+      "epoch": 0.38173366147115123,
+      "grad_norm": 0.16049465537071228,
+      "learning_rate": 5e-05,
+      "loss": 1.7456,
+      "step": 1396
+    },
+    {
+      "epoch": 0.38200710965272083,
+      "grad_norm": 0.17042703926563263,
+      "learning_rate": 5e-05,
+      "loss": 1.663,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3822805578342904,
+      "grad_norm": 0.1597413718700409,
+      "learning_rate": 5e-05,
+      "loss": 1.6331,
+      "step": 1398
+    },
+    {
+      "epoch": 0.38255400601586,
+      "grad_norm": 0.15566933155059814,
+      "learning_rate": 5e-05,
+      "loss": 1.6423,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3828274541974296,
+      "grad_norm": 0.15969716012477875,
+      "learning_rate": 5e-05,
+      "loss": 1.7039,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3831009023789992,
+      "grad_norm": 0.15060196816921234,
+      "learning_rate": 5e-05,
+      "loss": 1.6484,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3833743505605688,
+      "grad_norm": 0.15443742275238037,
+      "learning_rate": 5e-05,
+      "loss": 1.5378,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3836477987421384,
+      "grad_norm": 0.15732194483280182,
+      "learning_rate": 5e-05,
+      "loss": 1.6646,
+      "step": 1403
+    },
+    {
+      "epoch": 0.38392124692370794,
+      "grad_norm": 0.15805989503860474,
+      "learning_rate": 5e-05,
+      "loss": 1.5314,
+      "step": 1404
+    },
+    {
+      "epoch": 0.38419469510527754,
+      "grad_norm": 0.1553335189819336,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 1405
+    },
+    {
+      "epoch": 0.38446814328684714,
+      "grad_norm": 0.16199639439582825,
+      "learning_rate": 5e-05,
+      "loss": 1.6392,
+      "step": 1406
+    },
+    {
+      "epoch": 0.38474159146841674,
+      "grad_norm": 0.15157069265842438,
+      "learning_rate": 5e-05,
+      "loss": 1.5499,
+      "step": 1407
+    },
+    {
+      "epoch": 0.38501503964998635,
+      "grad_norm": 0.15281042456626892,
+      "learning_rate": 5e-05,
+      "loss": 1.6623,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3852884878315559,
+      "grad_norm": 0.16575610637664795,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3855619360131255,
+      "grad_norm": 0.17022867500782013,
+      "learning_rate": 5e-05,
+      "loss": 1.7497,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3858353841946951,
+      "grad_norm": 0.1417384147644043,
+      "learning_rate": 5e-05,
+      "loss": 1.615,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3861088323762647,
+      "grad_norm": 0.18117718398571014,
+      "learning_rate": 5e-05,
+      "loss": 1.7536,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3863822805578343,
+      "grad_norm": 0.15728381276130676,
+      "learning_rate": 5e-05,
+      "loss": 1.577,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3866557287394039,
+      "grad_norm": 0.163984015583992,
+      "learning_rate": 5e-05,
+      "loss": 1.6305,
+      "step": 1414
+    },
+    {
+      "epoch": 0.38692917692097345,
+      "grad_norm": 0.16205452382564545,
+      "learning_rate": 5e-05,
+      "loss": 1.5719,
+      "step": 1415
+    },
+    {
+      "epoch": 0.38720262510254305,
+      "grad_norm": 0.16791151463985443,
+      "learning_rate": 5e-05,
+      "loss": 1.6911,
+      "step": 1416
+    },
+    {
+      "epoch": 0.38747607328411265,
+      "grad_norm": 0.165225088596344,
+      "learning_rate": 5e-05,
+      "loss": 1.5915,
+      "step": 1417
+    },
+    {
+      "epoch": 0.38774952146568226,
+      "grad_norm": 0.1673455834388733,
+      "learning_rate": 5e-05,
+      "loss": 1.6484,
+      "step": 1418
+    },
+    {
+      "epoch": 0.38802296964725186,
+      "grad_norm": 0.15316179394721985,
+      "learning_rate": 5e-05,
+      "loss": 1.6133,
+      "step": 1419
+    },
+    {
+      "epoch": 0.38829641782882146,
+      "grad_norm": 0.17565707862377167,
+      "learning_rate": 5e-05,
+      "loss": 1.7247,
+      "step": 1420
+    },
+    {
+      "epoch": 0.388569866010391,
+      "grad_norm": 0.15834634006023407,
+      "learning_rate": 5e-05,
+      "loss": 1.6689,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3888433141919606,
+      "grad_norm": 0.16178841888904572,
+      "learning_rate": 5e-05,
+      "loss": 1.6276,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3891167623735302,
+      "grad_norm": 0.15787442028522491,
+      "learning_rate": 5e-05,
+      "loss": 1.6655,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3893902105550998,
+      "grad_norm": 0.17375800013542175,
+      "learning_rate": 5e-05,
+      "loss": 1.6643,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3896636587366694,
+      "grad_norm": 0.1527184098958969,
+      "learning_rate": 5e-05,
+      "loss": 1.5199,
+      "step": 1425
+    },
+    {
+      "epoch": 0.389937106918239,
+      "grad_norm": 0.17630016803741455,
+      "learning_rate": 5e-05,
+      "loss": 1.9053,
+      "step": 1426
+    },
+    {
+      "epoch": 0.39021055509980856,
+      "grad_norm": 0.16821622848510742,
+      "learning_rate": 5e-05,
+      "loss": 1.6229,
+      "step": 1427
+    },
+    {
+      "epoch": 0.39048400328137817,
+      "grad_norm": 0.1571454554796219,
+      "learning_rate": 5e-05,
+      "loss": 1.5945,
+      "step": 1428
+    },
+    {
+      "epoch": 0.39075745146294777,
+      "grad_norm": 0.16041232645511627,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 1429
+    },
+    {
+      "epoch": 0.39103089964451737,
+      "grad_norm": 0.15614961087703705,
+      "learning_rate": 5e-05,
+      "loss": 1.6234,
+      "step": 1430
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.1785866618156433,
+      "learning_rate": 5e-05,
+      "loss": 1.5821,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3915777960076566,
+      "grad_norm": 0.1583407074213028,
+      "learning_rate": 5e-05,
+      "loss": 1.6031,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3918512441892261,
+      "grad_norm": 0.15513792634010315,
+      "learning_rate": 5e-05,
+      "loss": 1.6701,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3921246923707957,
+      "grad_norm": 0.15645377337932587,
+      "learning_rate": 5e-05,
+      "loss": 1.6072,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3923981405523653,
+      "grad_norm": 0.17222340404987335,
+      "learning_rate": 5e-05,
+      "loss": 1.7521,
+      "step": 1435
+    },
+    {
+      "epoch": 0.39267158873393493,
+      "grad_norm": 0.16205964982509613,
+      "learning_rate": 5e-05,
+      "loss": 1.5481,
+      "step": 1436
+    },
+    {
+      "epoch": 0.39294503691550453,
+      "grad_norm": 0.16628527641296387,
+      "learning_rate": 5e-05,
+      "loss": 1.6072,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3932184850970741,
+      "grad_norm": 0.15461629629135132,
+      "learning_rate": 5e-05,
+      "loss": 1.6637,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3934919332786437,
+      "grad_norm": 0.17771334946155548,
+      "learning_rate": 5e-05,
+      "loss": 1.7054,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3937653814602133,
+      "grad_norm": 0.16035428643226624,
+      "learning_rate": 5e-05,
+      "loss": 1.6933,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3940388296417829,
+      "grad_norm": 0.15316614508628845,
+      "learning_rate": 5e-05,
+      "loss": 1.4949,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3943122778233525,
+      "grad_norm": 0.15236356854438782,
+      "learning_rate": 5e-05,
+      "loss": 1.6445,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3945857260049221,
+      "grad_norm": 0.15374061465263367,
+      "learning_rate": 5e-05,
+      "loss": 1.7432,
+      "step": 1443
+    },
+    {
+      "epoch": 0.39485917418649163,
+      "grad_norm": 0.15344202518463135,
+      "learning_rate": 5e-05,
+      "loss": 1.584,
+      "step": 1444
+    },
+    {
+      "epoch": 0.39513262236806124,
+      "grad_norm": 0.15145696699619293,
+      "learning_rate": 5e-05,
+      "loss": 1.6804,
+      "step": 1445
+    },
+    {
+      "epoch": 0.39540607054963084,
+      "grad_norm": 0.15984676778316498,
+      "learning_rate": 5e-05,
+      "loss": 1.7636,
+      "step": 1446
+    },
+    {
+      "epoch": 0.39567951873120044,
+      "grad_norm": 0.14700670540332794,
+      "learning_rate": 5e-05,
+      "loss": 1.5888,
+      "step": 1447
+    },
+    {
+      "epoch": 0.39595296691277004,
+      "grad_norm": 0.1534331887960434,
+      "learning_rate": 5e-05,
+      "loss": 1.5919,
+      "step": 1448
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.15421104431152344,
+      "learning_rate": 5e-05,
+      "loss": 1.5659,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3964998632759092,
+      "grad_norm": 0.17723333835601807,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3967733114574788,
+      "grad_norm": 0.1645742654800415,
+      "learning_rate": 5e-05,
+      "loss": 1.6573,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3970467596390484,
+      "grad_norm": 0.1458151489496231,
+      "learning_rate": 5e-05,
+      "loss": 1.4709,
+      "step": 1452
+    },
+    {
+      "epoch": 0.397320207820618,
+      "grad_norm": 0.15043888986110687,
+      "learning_rate": 5e-05,
+      "loss": 1.6267,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3975936560021876,
+      "grad_norm": 0.16159576177597046,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3978671041837572,
+      "grad_norm": 0.15626250207424164,
+      "learning_rate": 5e-05,
+      "loss": 1.6397,
+      "step": 1455
+    },
+    {
+      "epoch": 0.39814055236532675,
+      "grad_norm": 0.15309499204158783,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 1456
+    },
+    {
+      "epoch": 0.39841400054689635,
+      "grad_norm": 0.1638927310705185,
+      "learning_rate": 5e-05,
+      "loss": 1.6134,
+      "step": 1457
+    },
+    {
+      "epoch": 0.39868744872846595,
+      "grad_norm": 0.156220942735672,
+      "learning_rate": 5e-05,
+      "loss": 1.6544,
+      "step": 1458
+    },
+    {
+      "epoch": 0.39896089691003556,
+      "grad_norm": 0.15519675612449646,
+      "learning_rate": 5e-05,
+      "loss": 1.6289,
+      "step": 1459
+    },
+    {
+      "epoch": 0.39923434509160516,
+      "grad_norm": 0.16165190935134888,
+      "learning_rate": 5e-05,
+      "loss": 1.6655,
+      "step": 1460
+    },
+    {
+      "epoch": 0.39950779327317476,
+      "grad_norm": 0.14794325828552246,
+      "learning_rate": 5e-05,
+      "loss": 1.6321,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3997812414547443,
+      "grad_norm": 0.1512671858072281,
+      "learning_rate": 5e-05,
+      "loss": 1.6119,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4000546896363139,
+      "grad_norm": 0.14749833941459656,
+      "learning_rate": 5e-05,
+      "loss": 1.6302,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4003281378178835,
+      "grad_norm": 0.1552843600511551,
+      "learning_rate": 5e-05,
+      "loss": 1.6113,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4006015859994531,
+      "grad_norm": 0.15012173354625702,
+      "learning_rate": 5e-05,
+      "loss": 1.5396,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4008750341810227,
+      "grad_norm": 0.15311282873153687,
+      "learning_rate": 5e-05,
+      "loss": 1.6266,
+      "step": 1466
+    },
+    {
+      "epoch": 0.40114848236259226,
+      "grad_norm": 0.15125569701194763,
+      "learning_rate": 5e-05,
+      "loss": 1.5979,
+      "step": 1467
+    },
+    {
+      "epoch": 0.40142193054416186,
+      "grad_norm": 0.1545860469341278,
+      "learning_rate": 5e-05,
+      "loss": 1.6872,
+      "step": 1468
+    },
+    {
+      "epoch": 0.40169537872573147,
+      "grad_norm": 0.1852022111415863,
+      "learning_rate": 5e-05,
+      "loss": 1.6847,
+      "step": 1469
+    },
+    {
+      "epoch": 0.40196882690730107,
+      "grad_norm": 0.1469736099243164,
+      "learning_rate": 5e-05,
+      "loss": 1.58,
+      "step": 1470
+    },
+    {
+      "epoch": 0.40224227508887067,
+      "grad_norm": 0.17848168313503265,
+      "learning_rate": 5e-05,
+      "loss": 1.6899,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4025157232704403,
+      "grad_norm": 0.16272708773612976,
+      "learning_rate": 5e-05,
+      "loss": 1.6655,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4027891714520098,
+      "grad_norm": 0.1590905487537384,
+      "learning_rate": 5e-05,
+      "loss": 1.6239,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4030626196335794,
+      "grad_norm": 0.16318002343177795,
+      "learning_rate": 5e-05,
+      "loss": 1.6324,
+      "step": 1474
+    },
+    {
+      "epoch": 0.403336067815149,
+      "grad_norm": 0.1604347825050354,
+      "learning_rate": 5e-05,
+      "loss": 1.6793,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4036095159967186,
+      "grad_norm": 0.1481696516275406,
+      "learning_rate": 5e-05,
+      "loss": 1.6391,
+      "step": 1476
+    },
+    {
+      "epoch": 0.40388296417828823,
+      "grad_norm": 0.1675114631652832,
+      "learning_rate": 5e-05,
+      "loss": 1.5925,
+      "step": 1477
+    },
+    {
+      "epoch": 0.40415641235985783,
+      "grad_norm": 0.15092286467552185,
+      "learning_rate": 5e-05,
+      "loss": 1.6294,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4044298605414274,
+      "grad_norm": 0.1638384312391281,
+      "learning_rate": 5e-05,
+      "loss": 1.6552,
+      "step": 1479
+    },
+    {
+      "epoch": 0.404703308722997,
+      "grad_norm": 0.1602986752986908,
+      "learning_rate": 5e-05,
+      "loss": 1.6264,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4049767569045666,
+      "grad_norm": 0.1523655354976654,
+      "learning_rate": 5e-05,
+      "loss": 1.6013,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4052502050861362,
+      "grad_norm": 0.1629789173603058,
+      "learning_rate": 5e-05,
+      "loss": 1.6301,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4055236532677058,
+      "grad_norm": 0.14943981170654297,
+      "learning_rate": 5e-05,
+      "loss": 1.5619,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4057971014492754,
+      "grad_norm": 0.15840235352516174,
+      "learning_rate": 5e-05,
+      "loss": 1.6715,
+      "step": 1484
+    },
+    {
+      "epoch": 0.40607054963084493,
+      "grad_norm": 0.15214814245700836,
+      "learning_rate": 5e-05,
+      "loss": 1.6118,
+      "step": 1485
+    },
+    {
+      "epoch": 0.40634399781241454,
+      "grad_norm": 0.15232683718204498,
+      "learning_rate": 5e-05,
+      "loss": 1.604,
+      "step": 1486
+    },
+    {
+      "epoch": 0.40661744599398414,
+      "grad_norm": 0.15846048295497894,
+      "learning_rate": 5e-05,
+      "loss": 1.6588,
+      "step": 1487
+    },
+    {
+      "epoch": 0.40689089417555374,
+      "grad_norm": 0.15377342700958252,
+      "learning_rate": 5e-05,
+      "loss": 1.5945,
+      "step": 1488
+    },
+    {
+      "epoch": 0.40716434235712334,
+      "grad_norm": 0.14967899024486542,
+      "learning_rate": 5e-05,
+      "loss": 1.5724,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4074377905386929,
+      "grad_norm": 0.15246756374835968,
+      "learning_rate": 5e-05,
+      "loss": 1.651,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4077112387202625,
+      "grad_norm": 0.1686270534992218,
+      "learning_rate": 5e-05,
+      "loss": 1.519,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4079846869018321,
+      "grad_norm": 0.1610393226146698,
+      "learning_rate": 5e-05,
+      "loss": 1.6535,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4082581350834017,
+      "grad_norm": 0.17595553398132324,
+      "learning_rate": 5e-05,
+      "loss": 1.6219,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4085315832649713,
+      "grad_norm": 0.1568283587694168,
+      "learning_rate": 5e-05,
+      "loss": 1.6203,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4088050314465409,
+      "grad_norm": 0.15611572563648224,
+      "learning_rate": 5e-05,
+      "loss": 1.7053,
+      "step": 1495
+    },
+    {
+      "epoch": 0.40907847962811045,
+      "grad_norm": 0.16013391315937042,
+      "learning_rate": 5e-05,
+      "loss": 1.6718,
+      "step": 1496
+    },
+    {
+      "epoch": 0.40935192780968005,
+      "grad_norm": 0.14747720956802368,
+      "learning_rate": 5e-05,
+      "loss": 1.5962,
+      "step": 1497
+    },
+    {
+      "epoch": 0.40962537599124965,
+      "grad_norm": 0.1694934070110321,
+      "learning_rate": 5e-05,
+      "loss": 1.6709,
+      "step": 1498
+    },
+    {
+      "epoch": 0.40989882417281925,
+      "grad_norm": 0.17449548840522766,
+      "learning_rate": 5e-05,
+      "loss": 1.6565,
+      "step": 1499
+    },
+    {
+      "epoch": 0.41017227235438886,
+      "grad_norm": 0.15884292125701904,
+      "learning_rate": 5e-05,
+      "loss": 1.6468,
+      "step": 1500
+    },
+    {
+      "epoch": 0.41044572053595846,
+      "grad_norm": 0.1549280881881714,
+      "learning_rate": 5e-05,
+      "loss": 1.5994,
+      "step": 1501
+    },
+    {
+      "epoch": 0.410719168717528,
+      "grad_norm": 0.15527546405792236,
+      "learning_rate": 5e-05,
+      "loss": 1.555,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4109926168990976,
+      "grad_norm": 0.1679718941450119,
+      "learning_rate": 5e-05,
+      "loss": 1.6539,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4112660650806672,
+      "grad_norm": 0.14265212416648865,
+      "learning_rate": 5e-05,
+      "loss": 1.434,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4115395132622368,
+      "grad_norm": 0.1542886644601822,
+      "learning_rate": 5e-05,
+      "loss": 1.5932,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4118129614438064,
+      "grad_norm": 0.1690497249364853,
+      "learning_rate": 5e-05,
+      "loss": 1.6344,
+      "step": 1506
+    },
+    {
+      "epoch": 0.412086409625376,
+      "grad_norm": 0.1507546305656433,
+      "learning_rate": 5e-05,
+      "loss": 1.5773,
+      "step": 1507
+    },
+    {
+      "epoch": 0.41235985780694556,
+      "grad_norm": 0.15997721254825592,
+      "learning_rate": 5e-05,
+      "loss": 1.694,
+      "step": 1508
+    },
+    {
+      "epoch": 0.41263330598851516,
+      "grad_norm": 0.15315738320350647,
+      "learning_rate": 5e-05,
+      "loss": 1.6413,
+      "step": 1509
+    },
+    {
+      "epoch": 0.41290675417008477,
+      "grad_norm": 0.14834025502204895,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 1510
+    },
+    {
+      "epoch": 0.41318020235165437,
+      "grad_norm": 0.15537337958812714,
+      "learning_rate": 5e-05,
+      "loss": 1.6266,
+      "step": 1511
+    },
+    {
+      "epoch": 0.41345365053322397,
+      "grad_norm": 0.15986233949661255,
+      "learning_rate": 5e-05,
+      "loss": 1.5657,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4137270987147936,
+      "grad_norm": 0.15568973124027252,
+      "learning_rate": 5e-05,
+      "loss": 1.6403,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4140005468963631,
+      "grad_norm": 0.16333115100860596,
+      "learning_rate": 5e-05,
+      "loss": 1.6504,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4142739950779327,
+      "grad_norm": 0.1546471118927002,
+      "learning_rate": 5e-05,
+      "loss": 1.6104,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4145474432595023,
+      "grad_norm": 0.16248218715190887,
+      "learning_rate": 5e-05,
+      "loss": 1.6648,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4148208914410719,
+      "grad_norm": 0.16074463725090027,
+      "learning_rate": 5e-05,
+      "loss": 1.6331,
+      "step": 1517
+    },
+    {
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.15557120740413666,
+      "learning_rate": 5e-05,
+      "loss": 1.6255,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4153677878042111,
+      "grad_norm": 0.15570150315761566,
+      "learning_rate": 5e-05,
+      "loss": 1.6743,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4156412359857807,
+      "grad_norm": 0.17130300402641296,
+      "learning_rate": 5e-05,
+      "loss": 1.6354,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4159146841673503,
+      "grad_norm": 0.1627659946680069,
+      "learning_rate": 5e-05,
+      "loss": 1.6975,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4161881323489199,
+      "grad_norm": 0.1583612710237503,
+      "learning_rate": 5e-05,
+      "loss": 1.6427,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4164615805304895,
+      "grad_norm": 0.15877516567707062,
+      "learning_rate": 5e-05,
+      "loss": 1.6016,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4167350287120591,
+      "grad_norm": 0.16064798831939697,
+      "learning_rate": 5e-05,
+      "loss": 1.605,
+      "step": 1524
+    },
+    {
+      "epoch": 0.41700847689362863,
+      "grad_norm": 0.16150295734405518,
+      "learning_rate": 5e-05,
+      "loss": 1.7196,
+      "step": 1525
+    },
+    {
+      "epoch": 0.41728192507519823,
+      "grad_norm": 0.1557587832212448,
+      "learning_rate": 5e-05,
+      "loss": 1.6024,
+      "step": 1526
+    },
+    {
+      "epoch": 0.41755537325676784,
+      "grad_norm": 0.14889568090438843,
+      "learning_rate": 5e-05,
+      "loss": 1.6842,
+      "step": 1527
+    },
+    {
+      "epoch": 0.41782882143833744,
+      "grad_norm": 0.15397529304027557,
+      "learning_rate": 5e-05,
+      "loss": 1.6452,
+      "step": 1528
+    },
+    {
+      "epoch": 0.41810226961990704,
+      "grad_norm": 0.14965930581092834,
+      "learning_rate": 5e-05,
+      "loss": 1.599,
+      "step": 1529
+    },
+    {
+      "epoch": 0.41837571780147664,
+      "grad_norm": 0.14950762689113617,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4186491659830462,
+      "grad_norm": 0.1583251804113388,
+      "learning_rate": 5e-05,
+      "loss": 1.661,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4189226141646158,
+      "grad_norm": 0.1470499336719513,
+      "learning_rate": 5e-05,
+      "loss": 1.4718,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4191960623461854,
+      "grad_norm": 0.15115216374397278,
+      "learning_rate": 5e-05,
+      "loss": 1.6577,
+      "step": 1533
+    },
+    {
+      "epoch": 0.419469510527755,
+      "grad_norm": 0.1576387733221054,
+      "learning_rate": 5e-05,
+      "loss": 1.6743,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4197429587093246,
+      "grad_norm": 0.15287528932094574,
+      "learning_rate": 5e-05,
+      "loss": 1.619,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4200164068908942,
+      "grad_norm": 0.15793505311012268,
+      "learning_rate": 5e-05,
+      "loss": 1.604,
+      "step": 1536
+    },
+    {
+      "epoch": 0.42028985507246375,
+      "grad_norm": 0.14916017651557922,
+      "learning_rate": 5e-05,
+      "loss": 1.6273,
+      "step": 1537
+    },
+    {
+      "epoch": 0.42056330325403335,
+      "grad_norm": 0.17765149474143982,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 1538
+    },
+    {
+      "epoch": 0.42083675143560295,
+      "grad_norm": 0.15169952809810638,
+      "learning_rate": 5e-05,
+      "loss": 1.5879,
+      "step": 1539
+    },
+    {
+      "epoch": 0.42111019961717255,
+      "grad_norm": 0.15626677870750427,
+      "learning_rate": 5e-05,
+      "loss": 1.5916,
+      "step": 1540
+    },
+    {
+      "epoch": 0.42138364779874216,
+      "grad_norm": 0.17149586975574493,
+      "learning_rate": 5e-05,
+      "loss": 1.5988,
+      "step": 1541
+    },
+    {
+      "epoch": 0.42165709598031176,
+      "grad_norm": 0.15350784361362457,
+      "learning_rate": 5e-05,
+      "loss": 1.7237,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4219305441618813,
+      "grad_norm": 0.1620538979768753,
+      "learning_rate": 5e-05,
+      "loss": 1.6646,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4222039923434509,
+      "grad_norm": 0.19055283069610596,
+      "learning_rate": 5e-05,
+      "loss": 1.6122,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4224774405250205,
+      "grad_norm": 0.16404415667057037,
+      "learning_rate": 5e-05,
+      "loss": 1.718,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4227508887065901,
+      "grad_norm": 0.18905016779899597,
+      "learning_rate": 5e-05,
+      "loss": 1.6112,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4230243368881597,
+      "grad_norm": 0.15663960576057434,
+      "learning_rate": 5e-05,
+      "loss": 1.5915,
+      "step": 1547
+    },
+    {
+      "epoch": 0.42329778506972926,
+      "grad_norm": 0.15725626051425934,
+      "learning_rate": 5e-05,
+      "loss": 1.6258,
+      "step": 1548
+    },
+    {
+      "epoch": 0.42357123325129886,
+      "grad_norm": 0.14791657030582428,
+      "learning_rate": 5e-05,
+      "loss": 1.6198,
+      "step": 1549
+    },
+    {
+      "epoch": 0.42384468143286846,
+      "grad_norm": 0.1585807204246521,
+      "learning_rate": 5e-05,
+      "loss": 1.6669,
+      "step": 1550
+    },
+    {
+      "epoch": 0.42411812961443807,
+      "grad_norm": 0.1611957997083664,
+      "learning_rate": 5e-05,
+      "loss": 1.6661,
+      "step": 1551
+    },
+    {
+      "epoch": 0.42439157779600767,
+      "grad_norm": 0.15032406151294708,
+      "learning_rate": 5e-05,
+      "loss": 1.6543,
+      "step": 1552
+    },
+    {
+      "epoch": 0.42466502597757727,
+      "grad_norm": 0.15738679468631744,
+      "learning_rate": 5e-05,
+      "loss": 1.6384,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4249384741591468,
+      "grad_norm": 0.1558157354593277,
+      "learning_rate": 5e-05,
+      "loss": 1.6816,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4252119223407164,
+      "grad_norm": 0.15282602608203888,
+      "learning_rate": 5e-05,
+      "loss": 1.6191,
+      "step": 1555
+    },
+    {
+      "epoch": 0.425485370522286,
+      "grad_norm": 0.16039276123046875,
+      "learning_rate": 5e-05,
+      "loss": 1.6248,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4257588187038556,
+      "grad_norm": 0.1603170931339264,
+      "learning_rate": 5e-05,
+      "loss": 1.6796,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4260322668854252,
+      "grad_norm": 0.16182708740234375,
+      "learning_rate": 5e-05,
+      "loss": 1.62,
+      "step": 1558
+    },
+    {
+      "epoch": 0.42630571506699483,
+      "grad_norm": 0.15992410480976105,
+      "learning_rate": 5e-05,
+      "loss": 1.5315,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4265791632485644,
+      "grad_norm": 0.15438152849674225,
+      "learning_rate": 5e-05,
+      "loss": 1.6707,
+      "step": 1560
+    },
+    {
+      "epoch": 0.426852611430134,
+      "grad_norm": 0.1619657427072525,
+      "learning_rate": 5e-05,
+      "loss": 1.6402,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4271260596117036,
+      "grad_norm": 0.16311468183994293,
+      "learning_rate": 5e-05,
+      "loss": 1.6292,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4273995077932732,
+      "grad_norm": 0.15871946513652802,
+      "learning_rate": 5e-05,
+      "loss": 1.7462,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4276729559748428,
+      "grad_norm": 0.17374621331691742,
+      "learning_rate": 5e-05,
+      "loss": 1.6766,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4279464041564124,
+      "grad_norm": 0.1653493344783783,
+      "learning_rate": 5e-05,
+      "loss": 1.6076,
+      "step": 1565
+    },
+    {
+      "epoch": 0.42821985233798193,
+      "grad_norm": 0.16838444769382477,
+      "learning_rate": 5e-05,
+      "loss": 1.754,
+      "step": 1566
+    },
+    {
+      "epoch": 0.42849330051955153,
+      "grad_norm": 0.14829325675964355,
+      "learning_rate": 5e-05,
+      "loss": 1.5052,
+      "step": 1567
+    },
+    {
+      "epoch": 0.42876674870112114,
+      "grad_norm": 0.1606733649969101,
+      "learning_rate": 5e-05,
+      "loss": 1.7269,
+      "step": 1568
+    },
+    {
+      "epoch": 0.42904019688269074,
+      "grad_norm": 0.16772723197937012,
+      "learning_rate": 5e-05,
+      "loss": 1.7469,
+      "step": 1569
+    },
+    {
+      "epoch": 0.42931364506426034,
+      "grad_norm": 0.1608705222606659,
+      "learning_rate": 5e-05,
+      "loss": 1.6371,
+      "step": 1570
+    },
+    {
+      "epoch": 0.42958709324582994,
+      "grad_norm": 0.16294115781784058,
+      "learning_rate": 5e-05,
+      "loss": 1.5675,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4298605414273995,
+      "grad_norm": 0.16304922103881836,
+      "learning_rate": 5e-05,
+      "loss": 1.7321,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4301339896089691,
+      "grad_norm": 0.18761757016181946,
+      "learning_rate": 5e-05,
+      "loss": 1.7382,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4304074377905387,
+      "grad_norm": 0.18934416770935059,
+      "learning_rate": 5e-05,
+      "loss": 1.5394,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4306808859721083,
+      "grad_norm": 0.16312015056610107,
+      "learning_rate": 5e-05,
+      "loss": 1.6435,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4309543341536779,
+      "grad_norm": 0.20754271745681763,
+      "learning_rate": 5e-05,
+      "loss": 1.625,
+      "step": 1576
+    },
+    {
+      "epoch": 0.43122778233524744,
+      "grad_norm": 0.19495531916618347,
+      "learning_rate": 5e-05,
+      "loss": 1.6369,
+      "step": 1577
+    },
+    {
+      "epoch": 0.43150123051681705,
+      "grad_norm": 0.1650926023721695,
+      "learning_rate": 5e-05,
+      "loss": 1.7474,
+      "step": 1578
+    },
+    {
+      "epoch": 0.43177467869838665,
+      "grad_norm": 0.20554234087467194,
+      "learning_rate": 5e-05,
+      "loss": 1.5715,
+      "step": 1579
+    },
+    {
+      "epoch": 0.43204812687995625,
+      "grad_norm": 0.18859036266803741,
+      "learning_rate": 5e-05,
+      "loss": 1.6938,
+      "step": 1580
+    },
+    {
+      "epoch": 0.43232157506152585,
+      "grad_norm": 0.16272279620170593,
+      "learning_rate": 5e-05,
+      "loss": 1.583,
+      "step": 1581
+    },
+    {
+      "epoch": 0.43259502324309546,
+      "grad_norm": 0.20159992575645447,
+      "learning_rate": 5e-05,
+      "loss": 1.594,
+      "step": 1582
+    },
+    {
+      "epoch": 0.432868471424665,
+      "grad_norm": 0.16306354105472565,
+      "learning_rate": 5e-05,
+      "loss": 1.6359,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4331419196062346,
+      "grad_norm": 0.16837598383426666,
+      "learning_rate": 5e-05,
+      "loss": 1.6118,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4334153677878042,
+      "grad_norm": 0.19100947678089142,
+      "learning_rate": 5e-05,
+      "loss": 1.6267,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4336888159693738,
+      "grad_norm": 0.15260100364685059,
+      "learning_rate": 5e-05,
+      "loss": 1.5452,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4339622641509434,
+      "grad_norm": 0.15616540610790253,
+      "learning_rate": 5e-05,
+      "loss": 1.7126,
+      "step": 1587
+    },
+    {
+      "epoch": 0.434235712332513,
+      "grad_norm": 0.16332991421222687,
+      "learning_rate": 5e-05,
+      "loss": 1.6253,
+      "step": 1588
+    },
+    {
+      "epoch": 0.43450916051408256,
+      "grad_norm": 0.16623741388320923,
+      "learning_rate": 5e-05,
+      "loss": 1.6323,
+      "step": 1589
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.15148966014385223,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 1590
+    },
+    {
+      "epoch": 0.43505605687722176,
+      "grad_norm": 0.1654033362865448,
+      "learning_rate": 5e-05,
+      "loss": 1.5779,
+      "step": 1591
+    },
+    {
+      "epoch": 0.43532950505879137,
+      "grad_norm": 0.17548300325870514,
+      "learning_rate": 5e-05,
+      "loss": 1.6037,
+      "step": 1592
+    },
+    {
+      "epoch": 0.43560295324036097,
+      "grad_norm": 0.1487036496400833,
+      "learning_rate": 5e-05,
+      "loss": 1.5709,
+      "step": 1593
+    },
+    {
+      "epoch": 0.43587640142193057,
+      "grad_norm": 0.17089228332042694,
+      "learning_rate": 5e-05,
+      "loss": 1.6508,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4361498496035001,
+      "grad_norm": 0.17811010777950287,
+      "learning_rate": 5e-05,
+      "loss": 1.5717,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4364232977850697,
+      "grad_norm": 0.16123969852924347,
+      "learning_rate": 5e-05,
+      "loss": 1.657,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4366967459666393,
+      "grad_norm": 0.15267348289489746,
+      "learning_rate": 5e-05,
+      "loss": 1.637,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4369701941482089,
+      "grad_norm": 0.19433604180812836,
+      "learning_rate": 5e-05,
+      "loss": 1.5795,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4372436423297785,
+      "grad_norm": 0.15938717126846313,
+      "learning_rate": 5e-05,
+      "loss": 1.719,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4375170905113481,
+      "grad_norm": 0.16735820472240448,
+      "learning_rate": 5e-05,
+      "loss": 1.7389,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4377905386929177,
+      "grad_norm": 0.19536766409873962,
+      "learning_rate": 5e-05,
+      "loss": 1.7317,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4380639868744873,
+      "grad_norm": 0.18374434113502502,
+      "learning_rate": 5e-05,
+      "loss": 1.541,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4383374350560569,
+      "grad_norm": 0.15462787449359894,
+      "learning_rate": 5e-05,
+      "loss": 1.6388,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4386108832376265,
+      "grad_norm": 0.18189293146133423,
+      "learning_rate": 5e-05,
+      "loss": 1.6179,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4388843314191961,
+      "grad_norm": 0.17478562891483307,
+      "learning_rate": 5e-05,
+      "loss": 1.6332,
+      "step": 1605
+    },
+    {
+      "epoch": 0.43915777960076563,
+      "grad_norm": 0.16604724526405334,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 1606
+    },
+    {
+      "epoch": 0.43943122778233523,
+      "grad_norm": 0.14763274788856506,
+      "learning_rate": 5e-05,
+      "loss": 1.5179,
+      "step": 1607
+    },
+    {
+      "epoch": 0.43970467596390483,
+      "grad_norm": 0.18372495472431183,
+      "learning_rate": 5e-05,
+      "loss": 1.6489,
+      "step": 1608
+    },
+    {
+      "epoch": 0.43997812414547444,
+      "grad_norm": 0.19224227964878082,
+      "learning_rate": 5e-05,
+      "loss": 1.6223,
+      "step": 1609
+    },
+    {
+      "epoch": 0.44025157232704404,
+      "grad_norm": 0.1518845111131668,
+      "learning_rate": 5e-05,
+      "loss": 1.6314,
+      "step": 1610
+    },
+    {
+      "epoch": 0.44052502050861364,
+      "grad_norm": 0.18382224440574646,
+      "learning_rate": 5e-05,
+      "loss": 1.6356,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4407984686901832,
+      "grad_norm": 0.18745778501033783,
+      "learning_rate": 5e-05,
+      "loss": 1.6189,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4410719168717528,
+      "grad_norm": 0.1606750637292862,
+      "learning_rate": 5e-05,
+      "loss": 1.7485,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4413453650533224,
+      "grad_norm": 0.176387757062912,
+      "learning_rate": 5e-05,
+      "loss": 1.6141,
+      "step": 1614
+    },
+    {
+      "epoch": 0.441618813234892,
+      "grad_norm": 0.17925933003425598,
+      "learning_rate": 5e-05,
+      "loss": 1.6457,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4418922614164616,
+      "grad_norm": 0.15073414146900177,
+      "learning_rate": 5e-05,
+      "loss": 1.5818,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4421657095980312,
+      "grad_norm": 0.18990382552146912,
+      "learning_rate": 5e-05,
+      "loss": 1.6838,
+      "step": 1617
+    },
+    {
+      "epoch": 0.44243915777960074,
+      "grad_norm": 0.1909227818250656,
+      "learning_rate": 5e-05,
+      "loss": 1.7176,
+      "step": 1618
+    },
+    {
+      "epoch": 0.44271260596117035,
+      "grad_norm": 0.14983290433883667,
+      "learning_rate": 5e-05,
+      "loss": 1.5743,
+      "step": 1619
+    },
+    {
+      "epoch": 0.44298605414273995,
+      "grad_norm": 0.18672508001327515,
+      "learning_rate": 5e-05,
+      "loss": 1.7437,
+      "step": 1620
+    },
+    {
+      "epoch": 0.44325950232430955,
+      "grad_norm": 0.1633402407169342,
+      "learning_rate": 5e-05,
+      "loss": 1.6302,
+      "step": 1621
+    },
+    {
+      "epoch": 0.44353295050587915,
+      "grad_norm": 0.15954262018203735,
+      "learning_rate": 5e-05,
+      "loss": 1.6534,
+      "step": 1622
+    },
+    {
+      "epoch": 0.44380639868744876,
+      "grad_norm": 0.17406605184078217,
+      "learning_rate": 5e-05,
+      "loss": 1.6408,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4440798468690183,
+      "grad_norm": 0.15827330946922302,
+      "learning_rate": 5e-05,
+      "loss": 1.7256,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4443532950505879,
+      "grad_norm": 0.1576947122812271,
+      "learning_rate": 5e-05,
+      "loss": 1.6882,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4446267432321575,
+      "grad_norm": 0.161362424492836,
+      "learning_rate": 5e-05,
+      "loss": 1.5669,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4449001914137271,
+      "grad_norm": 0.15085799992084503,
+      "learning_rate": 5e-05,
+      "loss": 1.6761,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4451736395952967,
+      "grad_norm": 0.14607954025268555,
+      "learning_rate": 5e-05,
+      "loss": 1.5689,
+      "step": 1628
+    },
+    {
+      "epoch": 0.44544708777686626,
+      "grad_norm": 0.149903804063797,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 1629
+    },
+    {
+      "epoch": 0.44572053595843586,
+      "grad_norm": 0.15304753184318542,
+      "learning_rate": 5e-05,
+      "loss": 1.6986,
+      "step": 1630
+    },
+    {
+      "epoch": 0.44599398414000546,
+      "grad_norm": 0.15582279860973358,
+      "learning_rate": 5e-05,
+      "loss": 1.6101,
+      "step": 1631
+    },
+    {
+      "epoch": 0.44626743232157506,
+      "grad_norm": 0.15496088564395905,
+      "learning_rate": 5e-05,
+      "loss": 1.6459,
+      "step": 1632
+    },
+    {
+      "epoch": 0.44654088050314467,
+      "grad_norm": 0.1606382578611374,
+      "learning_rate": 5e-05,
+      "loss": 1.6516,
+      "step": 1633
+    },
+    {
+      "epoch": 0.44681432868471427,
+      "grad_norm": 0.16134901344776154,
+      "learning_rate": 5e-05,
+      "loss": 1.7784,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4470877768662838,
+      "grad_norm": 0.16512495279312134,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4473612250478534,
+      "grad_norm": 0.15310251712799072,
+      "learning_rate": 5e-05,
+      "loss": 1.6425,
+      "step": 1636
+    },
+    {
+      "epoch": 0.447634673229423,
+      "grad_norm": 0.1573248952627182,
+      "learning_rate": 5e-05,
+      "loss": 1.6168,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4479081214109926,
+      "grad_norm": 0.15116660296916962,
+      "learning_rate": 5e-05,
+      "loss": 1.5542,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4481815695925622,
+      "grad_norm": 0.1584998220205307,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4484550177741318,
+      "grad_norm": 0.15654048323631287,
+      "learning_rate": 5e-05,
+      "loss": 1.6022,
+      "step": 1640
+    },
+    {
+      "epoch": 0.44872846595570137,
+      "grad_norm": 0.1555994600057602,
+      "learning_rate": 5e-05,
+      "loss": 1.5843,
+      "step": 1641
+    },
+    {
+      "epoch": 0.449001914137271,
+      "grad_norm": 0.1540311872959137,
+      "learning_rate": 5e-05,
+      "loss": 1.6218,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4492753623188406,
+      "grad_norm": 0.16408203542232513,
+      "learning_rate": 5e-05,
+      "loss": 1.6489,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4495488105004102,
+      "grad_norm": 0.15430676937103271,
+      "learning_rate": 5e-05,
+      "loss": 1.6169,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4498222586819798,
+      "grad_norm": 0.15662290155887604,
+      "learning_rate": 5e-05,
+      "loss": 1.7067,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4500957068635494,
+      "grad_norm": 0.16638533771038055,
+      "learning_rate": 5e-05,
+      "loss": 1.6786,
+      "step": 1646
+    },
+    {
+      "epoch": 0.45036915504511893,
+      "grad_norm": 0.15840692818164825,
+      "learning_rate": 5e-05,
+      "loss": 1.5811,
+      "step": 1647
+    },
+    {
+      "epoch": 0.45064260322668853,
+      "grad_norm": 0.15642379224300385,
+      "learning_rate": 5e-05,
+      "loss": 1.6151,
+      "step": 1648
+    },
+    {
+      "epoch": 0.45091605140825813,
+      "grad_norm": 0.15364129841327667,
+      "learning_rate": 5e-05,
+      "loss": 1.5745,
+      "step": 1649
+    },
+    {
+      "epoch": 0.45118949958982774,
+      "grad_norm": 0.16230598092079163,
+      "learning_rate": 5e-05,
+      "loss": 1.5954,
+      "step": 1650
+    },
+    {
+      "epoch": 0.45146294777139734,
+      "grad_norm": 0.16471756994724274,
+      "learning_rate": 5e-05,
+      "loss": 1.6509,
+      "step": 1651
+    },
+    {
+      "epoch": 0.45173639595296694,
+      "grad_norm": 0.16866976022720337,
+      "learning_rate": 5e-05,
+      "loss": 1.7927,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4520098441345365,
+      "grad_norm": 0.16828210651874542,
+      "learning_rate": 5e-05,
+      "loss": 1.7027,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4522832923161061,
+      "grad_norm": 0.15160506963729858,
+      "learning_rate": 5e-05,
+      "loss": 1.6926,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4525567404976757,
+      "grad_norm": 0.1498889923095703,
+      "learning_rate": 5e-05,
+      "loss": 1.6022,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.16272372007369995,
+      "learning_rate": 5e-05,
+      "loss": 1.6532,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4531036368608149,
+      "grad_norm": 0.16485629975795746,
+      "learning_rate": 5e-05,
+      "loss": 1.6615,
+      "step": 1657
+    },
+    {
+      "epoch": 0.45337708504238444,
+      "grad_norm": 0.15402807295322418,
+      "learning_rate": 5e-05,
+      "loss": 1.6162,
+      "step": 1658
+    },
+    {
+      "epoch": 0.45365053322395404,
+      "grad_norm": 0.14957159757614136,
+      "learning_rate": 5e-05,
+      "loss": 1.646,
+      "step": 1659
+    },
+    {
+      "epoch": 0.45392398140552365,
+      "grad_norm": 0.15733934938907623,
+      "learning_rate": 5e-05,
+      "loss": 1.5475,
+      "step": 1660
+    },
+    {
+      "epoch": 0.45419742958709325,
+      "grad_norm": 0.16374224424362183,
+      "learning_rate": 5e-05,
+      "loss": 1.6937,
+      "step": 1661
+    },
+    {
+      "epoch": 0.45447087776866285,
+      "grad_norm": 0.1516236960887909,
+      "learning_rate": 5e-05,
+      "loss": 1.5963,
+      "step": 1662
+    },
+    {
+      "epoch": 0.45474432595023245,
+      "grad_norm": 0.1528540402650833,
+      "learning_rate": 5e-05,
+      "loss": 1.5636,
+      "step": 1663
+    },
+    {
+      "epoch": 0.455017774131802,
+      "grad_norm": 0.1831309050321579,
+      "learning_rate": 5e-05,
+      "loss": 1.7391,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4552912223133716,
+      "grad_norm": 0.1571461260318756,
+      "learning_rate": 5e-05,
+      "loss": 1.56,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4555646704949412,
+      "grad_norm": 0.17489275336265564,
+      "learning_rate": 5e-05,
+      "loss": 1.7431,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4558381186765108,
+      "grad_norm": 0.16362959146499634,
+      "learning_rate": 5e-05,
+      "loss": 1.6041,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4561115668580804,
+      "grad_norm": 0.1590225100517273,
+      "learning_rate": 5e-05,
+      "loss": 1.7021,
+      "step": 1668
+    },
+    {
+      "epoch": 0.45638501503965,
+      "grad_norm": 0.1953715831041336,
+      "learning_rate": 5e-05,
+      "loss": 1.7408,
+      "step": 1669
+    },
+    {
+      "epoch": 0.45665846322121956,
+      "grad_norm": 0.1608944684267044,
+      "learning_rate": 5e-05,
+      "loss": 1.6262,
+      "step": 1670
+    },
+    {
+      "epoch": 0.45693191140278916,
+      "grad_norm": 0.16777822375297546,
+      "learning_rate": 5e-05,
+      "loss": 1.6342,
+      "step": 1671
+    },
+    {
+      "epoch": 0.45720535958435876,
+      "grad_norm": 0.16327768564224243,
+      "learning_rate": 5e-05,
+      "loss": 1.6965,
+      "step": 1672
+    },
+    {
+      "epoch": 0.45747880776592836,
+      "grad_norm": 0.15659572184085846,
+      "learning_rate": 5e-05,
+      "loss": 1.6095,
+      "step": 1673
+    },
+    {
+      "epoch": 0.45775225594749797,
+      "grad_norm": 0.17218998074531555,
+      "learning_rate": 5e-05,
+      "loss": 1.6831,
+      "step": 1674
+    },
+    {
+      "epoch": 0.45802570412906757,
+      "grad_norm": 0.1536104679107666,
+      "learning_rate": 5e-05,
+      "loss": 1.5643,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4582991523106371,
+      "grad_norm": 0.15404142439365387,
+      "learning_rate": 5e-05,
+      "loss": 1.5892,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4585726004922067,
+      "grad_norm": 0.15226425230503082,
+      "learning_rate": 5e-05,
+      "loss": 1.582,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4588460486737763,
+      "grad_norm": 0.1722135990858078,
+      "learning_rate": 5e-05,
+      "loss": 1.6726,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4591194968553459,
+      "grad_norm": 0.14889506995677948,
+      "learning_rate": 5e-05,
+      "loss": 1.5599,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4593929450369155,
+      "grad_norm": 0.15106317400932312,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 1680
+    },
+    {
+      "epoch": 0.45966639321848507,
+      "grad_norm": 0.1514967978000641,
+      "learning_rate": 5e-05,
+      "loss": 1.5257,
+      "step": 1681
+    },
+    {
+      "epoch": 0.45993984140005467,
+      "grad_norm": 0.1474781632423401,
+      "learning_rate": 5e-05,
+      "loss": 1.5199,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4602132895816243,
+      "grad_norm": 0.15649180114269257,
+      "learning_rate": 5e-05,
+      "loss": 1.7461,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4604867377631939,
+      "grad_norm": 0.15975254774093628,
+      "learning_rate": 5e-05,
+      "loss": 1.6344,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4607601859447635,
+      "grad_norm": 0.15253420174121857,
+      "learning_rate": 5e-05,
+      "loss": 1.5446,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4610336341263331,
+      "grad_norm": 0.16303133964538574,
+      "learning_rate": 5e-05,
+      "loss": 1.6806,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4613070823079026,
+      "grad_norm": 0.15411800146102905,
+      "learning_rate": 5e-05,
+      "loss": 1.6656,
+      "step": 1687
+    },
+    {
+      "epoch": 0.46158053048947223,
+      "grad_norm": 0.15066716074943542,
+      "learning_rate": 5e-05,
+      "loss": 1.6403,
+      "step": 1688
+    },
+    {
+      "epoch": 0.46185397867104183,
+      "grad_norm": 0.15379618108272552,
+      "learning_rate": 5e-05,
+      "loss": 1.5969,
+      "step": 1689
+    },
+    {
+      "epoch": 0.46212742685261143,
+      "grad_norm": 0.15352702140808105,
+      "learning_rate": 5e-05,
+      "loss": 1.6403,
+      "step": 1690
+    },
+    {
+      "epoch": 0.46240087503418104,
+      "grad_norm": 0.158903107047081,
+      "learning_rate": 5e-05,
+      "loss": 1.6892,
+      "step": 1691
+    },
+    {
+      "epoch": 0.46267432321575064,
+      "grad_norm": 0.15684601664543152,
+      "learning_rate": 5e-05,
+      "loss": 1.6744,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4629477713973202,
+      "grad_norm": 0.1492249220609665,
+      "learning_rate": 5e-05,
+      "loss": 1.613,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4632212195788898,
+      "grad_norm": 0.1778024435043335,
+      "learning_rate": 5e-05,
+      "loss": 1.7122,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4634946677604594,
+      "grad_norm": 0.16280174255371094,
+      "learning_rate": 5e-05,
+      "loss": 1.6553,
+      "step": 1695
+    },
+    {
+      "epoch": 0.463768115942029,
+      "grad_norm": 0.1631561666727066,
+      "learning_rate": 5e-05,
+      "loss": 1.646,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4640415641235986,
+      "grad_norm": 0.1574755609035492,
+      "learning_rate": 5e-05,
+      "loss": 1.7003,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4643150123051682,
+      "grad_norm": 0.15220339596271515,
+      "learning_rate": 5e-05,
+      "loss": 1.5751,
+      "step": 1698
+    },
+    {
+      "epoch": 0.46458846048673774,
+      "grad_norm": 0.14914649724960327,
+      "learning_rate": 5e-05,
+      "loss": 1.498,
+      "step": 1699
+    },
+    {
+      "epoch": 0.46486190866830734,
+      "grad_norm": 0.19843582808971405,
+      "learning_rate": 5e-05,
+      "loss": 1.7326,
+      "step": 1700
+    },
+    {
+      "epoch": 0.46513535684987695,
+      "grad_norm": 0.1573631912469864,
+      "learning_rate": 5e-05,
+      "loss": 1.6891,
+      "step": 1701
+    },
+    {
+      "epoch": 0.46540880503144655,
+      "grad_norm": 0.16770999133586884,
+      "learning_rate": 5e-05,
+      "loss": 1.6852,
+      "step": 1702
+    },
+    {
+      "epoch": 0.46568225321301615,
+      "grad_norm": 0.17956487834453583,
+      "learning_rate": 5e-05,
+      "loss": 1.6751,
+      "step": 1703
+    },
+    {
+      "epoch": 0.46595570139458575,
+      "grad_norm": 0.1517421156167984,
+      "learning_rate": 5e-05,
+      "loss": 1.6928,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4662291495761553,
+      "grad_norm": 0.164058655500412,
+      "learning_rate": 5e-05,
+      "loss": 1.6313,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4665025977577249,
+      "grad_norm": 0.16080081462860107,
+      "learning_rate": 5e-05,
+      "loss": 1.5901,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4667760459392945,
+      "grad_norm": 0.149660125374794,
+      "learning_rate": 5e-05,
+      "loss": 1.6411,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4670494941208641,
+      "grad_norm": 0.1580061912536621,
+      "learning_rate": 5e-05,
+      "loss": 1.6815,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4673229423024337,
+      "grad_norm": 0.16102102398872375,
+      "learning_rate": 5e-05,
+      "loss": 1.6063,
+      "step": 1709
+    },
+    {
+      "epoch": 0.46759639048400325,
+      "grad_norm": 0.16004008054733276,
+      "learning_rate": 5e-05,
+      "loss": 1.5849,
+      "step": 1710
+    },
+    {
+      "epoch": 0.46786983866557286,
+      "grad_norm": 0.14842753112316132,
+      "learning_rate": 5e-05,
+      "loss": 1.6713,
+      "step": 1711
+    },
+    {
+      "epoch": 0.46814328684714246,
+      "grad_norm": 0.17115214467048645,
+      "learning_rate": 5e-05,
+      "loss": 1.7549,
+      "step": 1712
+    },
+    {
+      "epoch": 0.46841673502871206,
+      "grad_norm": 0.16055969893932343,
+      "learning_rate": 5e-05,
+      "loss": 1.7193,
+      "step": 1713
+    },
+    {
+      "epoch": 0.46869018321028166,
+      "grad_norm": 0.14992430806159973,
+      "learning_rate": 5e-05,
+      "loss": 1.6529,
+      "step": 1714
+    },
+    {
+      "epoch": 0.46896363139185127,
+      "grad_norm": 0.1594884842634201,
+      "learning_rate": 5e-05,
+      "loss": 1.5389,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4692370795734208,
+      "grad_norm": 0.15089979767799377,
+      "learning_rate": 5e-05,
+      "loss": 1.5729,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4695105277549904,
+      "grad_norm": 0.15678352117538452,
+      "learning_rate": 5e-05,
+      "loss": 1.6118,
+      "step": 1717
+    },
+    {
+      "epoch": 0.46978397593656,
+      "grad_norm": 0.15891622006893158,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4700574241181296,
+      "grad_norm": 0.15619421005249023,
+      "learning_rate": 5e-05,
+      "loss": 1.657,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4703308722996992,
+      "grad_norm": 0.15846212208271027,
+      "learning_rate": 5e-05,
+      "loss": 1.649,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4706043204812688,
+      "grad_norm": 0.1617954969406128,
+      "learning_rate": 5e-05,
+      "loss": 1.665,
+      "step": 1721
+    },
+    {
+      "epoch": 0.47087776866283837,
+      "grad_norm": 0.15690676867961884,
+      "learning_rate": 5e-05,
+      "loss": 1.5683,
+      "step": 1722
+    },
+    {
+      "epoch": 0.47115121684440797,
+      "grad_norm": 0.15898874402046204,
+      "learning_rate": 5e-05,
+      "loss": 1.5895,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4714246650259776,
+      "grad_norm": 0.1485942155122757,
+      "learning_rate": 5e-05,
+      "loss": 1.5482,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4716981132075472,
+      "grad_norm": 0.15413321554660797,
+      "learning_rate": 5e-05,
+      "loss": 1.6355,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4719715613891168,
+      "grad_norm": 0.16040156781673431,
+      "learning_rate": 5e-05,
+      "loss": 1.6984,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4722450095706864,
+      "grad_norm": 0.1513047069311142,
+      "learning_rate": 5e-05,
+      "loss": 1.5586,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4725184577522559,
+      "grad_norm": 0.1601109653711319,
+      "learning_rate": 5e-05,
+      "loss": 1.7306,
+      "step": 1728
+    },
+    {
+      "epoch": 0.47279190593382553,
+      "grad_norm": 0.15264980494976044,
+      "learning_rate": 5e-05,
+      "loss": 1.6797,
+      "step": 1729
+    },
+    {
+      "epoch": 0.47306535411539513,
+      "grad_norm": 0.15382279455661774,
+      "learning_rate": 5e-05,
+      "loss": 1.7204,
+      "step": 1730
+    },
+    {
+      "epoch": 0.47333880229696473,
+      "grad_norm": 0.15572312474250793,
+      "learning_rate": 5e-05,
+      "loss": 1.5855,
+      "step": 1731
+    },
+    {
+      "epoch": 0.47361225047853434,
+      "grad_norm": 0.15744370222091675,
+      "learning_rate": 5e-05,
+      "loss": 1.6487,
+      "step": 1732
+    },
+    {
+      "epoch": 0.47388569866010394,
+      "grad_norm": 0.1619214117527008,
+      "learning_rate": 5e-05,
+      "loss": 1.6556,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4741591468416735,
+      "grad_norm": 0.1546183079481125,
+      "learning_rate": 5e-05,
+      "loss": 1.585,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4744325950232431,
+      "grad_norm": 0.1744302213191986,
+      "learning_rate": 5e-05,
+      "loss": 1.7555,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4747060432048127,
+      "grad_norm": 0.16364151239395142,
+      "learning_rate": 5e-05,
+      "loss": 1.6622,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4749794913863823,
+      "grad_norm": 0.14857520163059235,
+      "learning_rate": 5e-05,
+      "loss": 1.5803,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4752529395679519,
+      "grad_norm": 0.1706264317035675,
+      "learning_rate": 5e-05,
+      "loss": 1.6314,
+      "step": 1738
+    },
+    {
+      "epoch": 0.47552638774952144,
+      "grad_norm": 0.15458574891090393,
+      "learning_rate": 5e-05,
+      "loss": 1.5577,
+      "step": 1739
+    },
+    {
+      "epoch": 0.47579983593109104,
+      "grad_norm": 0.1472868025302887,
+      "learning_rate": 5e-05,
+      "loss": 1.5581,
+      "step": 1740
+    },
+    {
+      "epoch": 0.47607328411266064,
+      "grad_norm": 0.16088488698005676,
+      "learning_rate": 5e-05,
+      "loss": 1.6711,
+      "step": 1741
+    },
+    {
+      "epoch": 0.47634673229423025,
+      "grad_norm": 0.16059233248233795,
+      "learning_rate": 5e-05,
+      "loss": 1.6239,
+      "step": 1742
+    },
+    {
+      "epoch": 0.47662018047579985,
+      "grad_norm": 0.15403006970882416,
+      "learning_rate": 5e-05,
+      "loss": 1.5916,
+      "step": 1743
+    },
+    {
+      "epoch": 0.47689362865736945,
+      "grad_norm": 0.15987733006477356,
+      "learning_rate": 5e-05,
+      "loss": 1.5877,
+      "step": 1744
+    },
+    {
+      "epoch": 0.477167076838939,
+      "grad_norm": 0.1711868941783905,
+      "learning_rate": 5e-05,
+      "loss": 1.6065,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4774405250205086,
+      "grad_norm": 0.16569697856903076,
+      "learning_rate": 5e-05,
+      "loss": 1.6181,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4777139732020782,
+      "grad_norm": 0.15485012531280518,
+      "learning_rate": 5e-05,
+      "loss": 1.6009,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4779874213836478,
+      "grad_norm": 0.16187280416488647,
+      "learning_rate": 5e-05,
+      "loss": 1.5976,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.1522330790758133,
+      "learning_rate": 5e-05,
+      "loss": 1.7015,
+      "step": 1749
+    },
+    {
+      "epoch": 0.478534317746787,
+      "grad_norm": 0.1496533900499344,
+      "learning_rate": 5e-05,
+      "loss": 1.6114,
+      "step": 1750
+    },
+    {
+      "epoch": 0.47880776592835655,
+      "grad_norm": 0.16666877269744873,
+      "learning_rate": 5e-05,
+      "loss": 1.6728,
+      "step": 1751
+    },
+    {
+      "epoch": 0.47908121410992616,
+      "grad_norm": 0.1623934507369995,
+      "learning_rate": 5e-05,
+      "loss": 1.5624,
+      "step": 1752
+    },
+    {
+      "epoch": 0.47935466229149576,
+      "grad_norm": 0.1671043336391449,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 1753
+    },
+    {
+      "epoch": 0.47962811047306536,
+      "grad_norm": 0.16084609925746918,
+      "learning_rate": 5e-05,
+      "loss": 1.5419,
+      "step": 1754
+    },
+    {
+      "epoch": 0.47990155865463496,
+      "grad_norm": 0.18174776434898376,
+      "learning_rate": 5e-05,
+      "loss": 1.6554,
+      "step": 1755
+    },
+    {
+      "epoch": 0.48017500683620457,
+      "grad_norm": 0.15653514862060547,
+      "learning_rate": 5e-05,
+      "loss": 1.5938,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4804484550177741,
+      "grad_norm": 0.16829371452331543,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4807219031993437,
+      "grad_norm": 0.16628386080265045,
+      "learning_rate": 5e-05,
+      "loss": 1.6325,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4809953513809133,
+      "grad_norm": 0.15540596842765808,
+      "learning_rate": 5e-05,
+      "loss": 1.5648,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4812687995624829,
+      "grad_norm": 0.18352627754211426,
+      "learning_rate": 5e-05,
+      "loss": 1.7056,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4815422477440525,
+      "grad_norm": 0.16480763256549835,
+      "learning_rate": 5e-05,
+      "loss": 1.6272,
+      "step": 1761
+    },
+    {
+      "epoch": 0.48181569592562207,
+      "grad_norm": 0.16389521956443787,
+      "learning_rate": 5e-05,
+      "loss": 1.7246,
+      "step": 1762
+    },
+    {
+      "epoch": 0.48208914410719167,
+      "grad_norm": 0.14809884130954742,
+      "learning_rate": 5e-05,
+      "loss": 1.5785,
+      "step": 1763
+    },
+    {
+      "epoch": 0.48236259228876127,
+      "grad_norm": 0.1639075130224228,
+      "learning_rate": 5e-05,
+      "loss": 1.5561,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4826360404703309,
+      "grad_norm": 0.17026962339878082,
+      "learning_rate": 5e-05,
+      "loss": 1.599,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4829094886519005,
+      "grad_norm": 0.16182227432727814,
+      "learning_rate": 5e-05,
+      "loss": 1.5801,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4831829368334701,
+      "grad_norm": 0.17417606711387634,
+      "learning_rate": 5e-05,
+      "loss": 1.5894,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4834563850150396,
+      "grad_norm": 0.1606941670179367,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4837298331966092,
+      "grad_norm": 0.152814581990242,
+      "learning_rate": 5e-05,
+      "loss": 1.6011,
+      "step": 1769
+    },
+    {
+      "epoch": 0.48400328137817883,
+      "grad_norm": 0.16164222359657288,
+      "learning_rate": 5e-05,
+      "loss": 1.5673,
+      "step": 1770
+    },
+    {
+      "epoch": 0.48427672955974843,
+      "grad_norm": 0.16344451904296875,
+      "learning_rate": 5e-05,
+      "loss": 1.6209,
+      "step": 1771
+    },
+    {
+      "epoch": 0.48455017774131803,
+      "grad_norm": 0.1603250950574875,
+      "learning_rate": 5e-05,
+      "loss": 1.6105,
+      "step": 1772
+    },
+    {
+      "epoch": 0.48482362592288764,
+      "grad_norm": 0.16991026699543,
+      "learning_rate": 5e-05,
+      "loss": 1.6723,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4850970741044572,
+      "grad_norm": 0.1601463109254837,
+      "learning_rate": 5e-05,
+      "loss": 1.5577,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4853705222860268,
+      "grad_norm": 0.16674864292144775,
+      "learning_rate": 5e-05,
+      "loss": 1.6598,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4856439704675964,
+      "grad_norm": 0.16150332987308502,
+      "learning_rate": 5e-05,
+      "loss": 1.6948,
+      "step": 1776
+    },
+    {
+      "epoch": 0.485917418649166,
+      "grad_norm": 0.16552074253559113,
+      "learning_rate": 5e-05,
+      "loss": 1.6746,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4861908668307356,
+      "grad_norm": 0.1552310287952423,
+      "learning_rate": 5e-05,
+      "loss": 1.5961,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4864643150123052,
+      "grad_norm": 0.16018088161945343,
+      "learning_rate": 5e-05,
+      "loss": 1.6024,
+      "step": 1779
+    },
+    {
+      "epoch": 0.48673776319387474,
+      "grad_norm": 0.16856145858764648,
+      "learning_rate": 5e-05,
+      "loss": 1.6187,
+      "step": 1780
+    },
+    {
+      "epoch": 0.48701121137544434,
+      "grad_norm": 0.16507881879806519,
+      "learning_rate": 5e-05,
+      "loss": 1.6644,
+      "step": 1781
+    },
+    {
+      "epoch": 0.48728465955701394,
+      "grad_norm": 0.1524297147989273,
+      "learning_rate": 5e-05,
+      "loss": 1.6013,
+      "step": 1782
+    },
+    {
+      "epoch": 0.48755810773858355,
+      "grad_norm": 0.16545777022838593,
+      "learning_rate": 5e-05,
+      "loss": 1.6813,
+      "step": 1783
+    },
+    {
+      "epoch": 0.48783155592015315,
+      "grad_norm": 0.17331644892692566,
+      "learning_rate": 5e-05,
+      "loss": 1.6797,
+      "step": 1784
+    },
+    {
+      "epoch": 0.48810500410172275,
+      "grad_norm": 0.16021324694156647,
+      "learning_rate": 5e-05,
+      "loss": 1.5828,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4883784522832923,
+      "grad_norm": 0.15434862673282623,
+      "learning_rate": 5e-05,
+      "loss": 1.5449,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4886519004648619,
+      "grad_norm": 0.15682034194469452,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4889253486464315,
+      "grad_norm": 0.17640839517116547,
+      "learning_rate": 5e-05,
+      "loss": 1.7292,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4891987968280011,
+      "grad_norm": 0.16206398606300354,
+      "learning_rate": 5e-05,
+      "loss": 1.5844,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4894722450095707,
+      "grad_norm": 0.15871930122375488,
+      "learning_rate": 5e-05,
+      "loss": 1.685,
+      "step": 1790
+    },
+    {
+      "epoch": 0.48974569319114025,
+      "grad_norm": 0.16885024309158325,
+      "learning_rate": 5e-05,
+      "loss": 1.7065,
+      "step": 1791
+    },
+    {
+      "epoch": 0.49001914137270985,
+      "grad_norm": 0.16324368119239807,
+      "learning_rate": 5e-05,
+      "loss": 1.6381,
+      "step": 1792
+    },
+    {
+      "epoch": 0.49029258955427946,
+      "grad_norm": 0.15525028109550476,
+      "learning_rate": 5e-05,
+      "loss": 1.6529,
+      "step": 1793
+    },
+    {
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.15970031917095184,
+      "learning_rate": 5e-05,
+      "loss": 1.751,
+      "step": 1794
+    },
+    {
+      "epoch": 0.49083948591741866,
+      "grad_norm": 0.16955524682998657,
+      "learning_rate": 5e-05,
+      "loss": 1.7307,
+      "step": 1795
+    },
+    {
+      "epoch": 0.49111293409898826,
+      "grad_norm": 0.153351828455925,
+      "learning_rate": 5e-05,
+      "loss": 1.5379,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4913863822805578,
+      "grad_norm": 0.16103506088256836,
+      "learning_rate": 5e-05,
+      "loss": 1.6945,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4916598304621274,
+      "grad_norm": 0.17233219742774963,
+      "learning_rate": 5e-05,
+      "loss": 1.6656,
+      "step": 1798
+    },
+    {
+      "epoch": 0.491933278643697,
+      "grad_norm": 0.15539239346981049,
+      "learning_rate": 5e-05,
+      "loss": 1.6,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4922067268252666,
+      "grad_norm": 0.1534857302904129,
+      "learning_rate": 5e-05,
+      "loss": 1.5498,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4924801750068362,
+      "grad_norm": 0.18539227545261383,
+      "learning_rate": 5e-05,
+      "loss": 1.5413,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4927536231884058,
+      "grad_norm": 0.15599320828914642,
+      "learning_rate": 5e-05,
+      "loss": 1.7108,
+      "step": 1802
+    },
+    {
+      "epoch": 0.49302707136997537,
+      "grad_norm": 0.15700851380825043,
+      "learning_rate": 5e-05,
+      "loss": 1.6198,
+      "step": 1803
+    },
+    {
+      "epoch": 0.49330051955154497,
+      "grad_norm": 0.15811684727668762,
+      "learning_rate": 5e-05,
+      "loss": 1.5085,
+      "step": 1804
+    },
+    {
+      "epoch": 0.49357396773311457,
+      "grad_norm": 0.15955877304077148,
+      "learning_rate": 5e-05,
+      "loss": 1.5673,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4938474159146842,
+      "grad_norm": 0.15809409320354462,
+      "learning_rate": 5e-05,
+      "loss": 1.6075,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4941208640962538,
+      "grad_norm": 0.17937301099300385,
+      "learning_rate": 5e-05,
+      "loss": 1.5832,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4943943122778234,
+      "grad_norm": 0.16430814564228058,
+      "learning_rate": 5e-05,
+      "loss": 1.6945,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4946677604593929,
+      "grad_norm": 0.1564474105834961,
+      "learning_rate": 5e-05,
+      "loss": 1.6259,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4949412086409625,
+      "grad_norm": 0.17254872620105743,
+      "learning_rate": 5e-05,
+      "loss": 1.7017,
+      "step": 1810
+    },
+    {
+      "epoch": 0.49521465682253213,
+      "grad_norm": 0.16017508506774902,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 1811
+    },
+    {
+      "epoch": 0.49548810500410173,
+      "grad_norm": 0.15552417933940887,
+      "learning_rate": 5e-05,
+      "loss": 1.5943,
+      "step": 1812
+    },
+    {
+      "epoch": 0.49576155318567133,
+      "grad_norm": 0.14992982149124146,
+      "learning_rate": 5e-05,
+      "loss": 1.5989,
+      "step": 1813
+    },
+    {
+      "epoch": 0.49603500136724094,
+      "grad_norm": 0.14720216393470764,
+      "learning_rate": 5e-05,
+      "loss": 1.4911,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4963084495488105,
+      "grad_norm": 0.15855912864208221,
+      "learning_rate": 5e-05,
+      "loss": 1.5896,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4965818977303801,
+      "grad_norm": 0.1551193743944168,
+      "learning_rate": 5e-05,
+      "loss": 1.58,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4968553459119497,
+      "grad_norm": 0.15198324620723724,
+      "learning_rate": 5e-05,
+      "loss": 1.6181,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4971287940935193,
+      "grad_norm": 0.1556388884782791,
+      "learning_rate": 5e-05,
+      "loss": 1.6119,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4974022422750889,
+      "grad_norm": 0.1501542180776596,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 1819
+    },
+    {
+      "epoch": 0.49767569045665844,
+      "grad_norm": 0.1648813784122467,
+      "learning_rate": 5e-05,
+      "loss": 1.6296,
+      "step": 1820
+    },
+    {
+      "epoch": 0.49794913863822804,
+      "grad_norm": 0.15707071125507355,
+      "learning_rate": 5e-05,
+      "loss": 1.6604,
+      "step": 1821
+    },
+    {
+      "epoch": 0.49822258681979764,
+      "grad_norm": 0.1539454311132431,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 1822
+    },
+    {
+      "epoch": 0.49849603500136724,
+      "grad_norm": 0.15157034993171692,
+      "learning_rate": 5e-05,
+      "loss": 1.6315,
+      "step": 1823
+    },
+    {
+      "epoch": 0.49876948318293685,
+      "grad_norm": 0.15487664937973022,
+      "learning_rate": 5e-05,
+      "loss": 1.676,
+      "step": 1824
+    },
+    {
+      "epoch": 0.49904293136450645,
+      "grad_norm": 0.15155303478240967,
+      "learning_rate": 5e-05,
+      "loss": 1.5685,
+      "step": 1825
+    },
+    {
+      "epoch": 0.499316379546076,
+      "grad_norm": 0.14079029858112335,
+      "learning_rate": 5e-05,
+      "loss": 1.5199,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4995898277276456,
+      "grad_norm": 0.15700970590114594,
+      "learning_rate": 5e-05,
+      "loss": 1.6312,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4998632759092152,
+      "grad_norm": 0.1630944162607193,
+      "learning_rate": 5e-05,
+      "loss": 1.7161,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5001367240907848,
+      "grad_norm": 0.15556710958480835,
+      "learning_rate": 5e-05,
+      "loss": 1.6417,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5004101722723544,
+      "grad_norm": 0.15315458178520203,
+      "learning_rate": 5e-05,
+      "loss": 1.6343,
+      "step": 1830
+    },
+    {
+      "epoch": 0.500683620453924,
+      "grad_norm": 0.15903054177761078,
+      "learning_rate": 5e-05,
+      "loss": 1.6433,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5009570686354936,
+      "grad_norm": 0.16064853966236115,
+      "learning_rate": 5e-05,
+      "loss": 1.6739,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5012305168170632,
+      "grad_norm": 0.14736616611480713,
+      "learning_rate": 5e-05,
+      "loss": 1.547,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5015039649986328,
+      "grad_norm": 0.17235925793647766,
+      "learning_rate": 5e-05,
+      "loss": 1.5768,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5017774131802023,
+      "grad_norm": 0.15816731750965118,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5020508613617719,
+      "grad_norm": 0.15823869407176971,
+      "learning_rate": 5e-05,
+      "loss": 1.6582,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5023243095433415,
+      "grad_norm": 0.14915357530117035,
+      "learning_rate": 5e-05,
+      "loss": 1.5551,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5025977577249111,
+      "grad_norm": 0.15896442532539368,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5028712059064807,
+      "grad_norm": 0.15622593462467194,
+      "learning_rate": 5e-05,
+      "loss": 1.5816,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5031446540880503,
+      "grad_norm": 0.16117733716964722,
+      "learning_rate": 5e-05,
+      "loss": 1.6266,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5034181022696199,
+      "grad_norm": 0.15878431499004364,
+      "learning_rate": 5e-05,
+      "loss": 1.6835,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5036915504511895,
+      "grad_norm": 0.15273039042949677,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5039649986327591,
+      "grad_norm": 0.1612463891506195,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5042384468143287,
+      "grad_norm": 0.17296503484249115,
+      "learning_rate": 5e-05,
+      "loss": 1.664,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5045118949958983,
+      "grad_norm": 0.1609872728586197,
+      "learning_rate": 5e-05,
+      "loss": 1.727,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5047853431774678,
+      "grad_norm": 0.18367880582809448,
+      "learning_rate": 5e-05,
+      "loss": 1.6324,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5050587913590374,
+      "grad_norm": 0.1622532606124878,
+      "learning_rate": 5e-05,
+      "loss": 1.6035,
+      "step": 1847
+    },
+    {
+      "epoch": 0.505332239540607,
+      "grad_norm": 0.16167710721492767,
+      "learning_rate": 5e-05,
+      "loss": 1.7208,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5056056877221766,
+      "grad_norm": 0.17007885873317719,
+      "learning_rate": 5e-05,
+      "loss": 1.4988,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5058791359037462,
+      "grad_norm": 0.1878514140844345,
+      "learning_rate": 5e-05,
+      "loss": 1.698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5061525840853158,
+      "grad_norm": 0.1594289392232895,
+      "learning_rate": 5e-05,
+      "loss": 1.6375,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5064260322668854,
+      "grad_norm": 0.17619235813617706,
+      "learning_rate": 5e-05,
+      "loss": 1.5621,
+      "step": 1852
+    },
+    {
+      "epoch": 0.506699480448455,
+      "grad_norm": 0.15811192989349365,
+      "learning_rate": 5e-05,
+      "loss": 1.4968,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5069729286300246,
+      "grad_norm": 0.15754997730255127,
+      "learning_rate": 5e-05,
+      "loss": 1.6578,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5072463768115942,
+      "grad_norm": 0.17825719714164734,
+      "learning_rate": 5e-05,
+      "loss": 1.6794,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5075198249931638,
+      "grad_norm": 0.1577797830104828,
+      "learning_rate": 5e-05,
+      "loss": 1.5553,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5077932731747334,
+      "grad_norm": 0.164150670170784,
+      "learning_rate": 5e-05,
+      "loss": 1.6357,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5080667213563029,
+      "grad_norm": 0.16957907378673553,
+      "learning_rate": 5e-05,
+      "loss": 1.6714,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5083401695378725,
+      "grad_norm": 0.15430040657520294,
+      "learning_rate": 5e-05,
+      "loss": 1.5349,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5086136177194421,
+      "grad_norm": 0.1596827656030655,
+      "learning_rate": 5e-05,
+      "loss": 1.6752,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5088870659010117,
+      "grad_norm": 0.15648046135902405,
+      "learning_rate": 5e-05,
+      "loss": 1.5519,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5091605140825813,
+      "grad_norm": 0.154343381524086,
+      "learning_rate": 5e-05,
+      "loss": 1.5832,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5094339622641509,
+      "grad_norm": 0.15734641253948212,
+      "learning_rate": 5e-05,
+      "loss": 1.6465,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5097074104457205,
+      "grad_norm": 0.16264241933822632,
+      "learning_rate": 5e-05,
+      "loss": 1.7062,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5099808586272901,
+      "grad_norm": 0.1580963432788849,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5102543068088597,
+      "grad_norm": 0.15470340847969055,
+      "learning_rate": 5e-05,
+      "loss": 1.591,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5105277549904293,
+      "grad_norm": 0.18182386457920074,
+      "learning_rate": 5e-05,
+      "loss": 1.7503,
+      "step": 1867
+    },
+    {
+      "epoch": 0.510801203171999,
+      "grad_norm": 0.16213788092136383,
+      "learning_rate": 5e-05,
+      "loss": 1.5868,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5110746513535686,
+      "grad_norm": 0.1534777283668518,
+      "learning_rate": 5e-05,
+      "loss": 1.5489,
+      "step": 1869
+    },
+    {
+      "epoch": 0.511348099535138,
+      "grad_norm": 0.1524716168642044,
+      "learning_rate": 5e-05,
+      "loss": 1.5555,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5116215477167076,
+      "grad_norm": 0.15701556205749512,
+      "learning_rate": 5e-05,
+      "loss": 1.5622,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5118949958982772,
+      "grad_norm": 0.1485898792743683,
+      "learning_rate": 5e-05,
+      "loss": 1.5619,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5121684440798469,
+      "grad_norm": 0.16039031744003296,
+      "learning_rate": 5e-05,
+      "loss": 1.6172,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5124418922614165,
+      "grad_norm": 0.167929545044899,
+      "learning_rate": 5e-05,
+      "loss": 1.6437,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5127153404429861,
+      "grad_norm": 0.16350795328617096,
+      "learning_rate": 5e-05,
+      "loss": 1.6416,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5129887886245557,
+      "grad_norm": 0.15972229838371277,
+      "learning_rate": 5e-05,
+      "loss": 1.7117,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5132622368061253,
+      "grad_norm": 0.15788406133651733,
+      "learning_rate": 5e-05,
+      "loss": 1.5657,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5135356849876949,
+      "grad_norm": 0.1540132462978363,
+      "learning_rate": 5e-05,
+      "loss": 1.6087,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5138091331692645,
+      "grad_norm": 0.16247431933879852,
+      "learning_rate": 5e-05,
+      "loss": 1.6432,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5140825813508341,
+      "grad_norm": 0.15716880559921265,
+      "learning_rate": 5e-05,
+      "loss": 1.5565,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5143560295324036,
+      "grad_norm": 0.15379799902439117,
+      "learning_rate": 5e-05,
+      "loss": 1.6237,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5146294777139732,
+      "grad_norm": 0.15428978204727173,
+      "learning_rate": 5e-05,
+      "loss": 1.5128,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5149029258955428,
+      "grad_norm": 0.15110410749912262,
+      "learning_rate": 5e-05,
+      "loss": 1.5246,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5151763740771124,
+      "grad_norm": 0.15420377254486084,
+      "learning_rate": 5e-05,
+      "loss": 1.6952,
+      "step": 1884
+    },
+    {
+      "epoch": 0.515449822258682,
+      "grad_norm": 0.15766222774982452,
+      "learning_rate": 5e-05,
+      "loss": 1.6611,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5157232704402516,
+      "grad_norm": 0.1532919555902481,
+      "learning_rate": 5e-05,
+      "loss": 1.5322,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5159967186218212,
+      "grad_norm": 0.15919700264930725,
+      "learning_rate": 5e-05,
+      "loss": 1.6387,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5162701668033908,
+      "grad_norm": 0.15913300216197968,
+      "learning_rate": 5e-05,
+      "loss": 1.6679,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5165436149849604,
+      "grad_norm": 0.16154128313064575,
+      "learning_rate": 5e-05,
+      "loss": 1.72,
+      "step": 1889
+    },
+    {
+      "epoch": 0.51681706316653,
+      "grad_norm": 0.16438822448253632,
+      "learning_rate": 5e-05,
+      "loss": 1.7099,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5170905113480996,
+      "grad_norm": 0.1546657532453537,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5173639595296692,
+      "grad_norm": 0.15779191255569458,
+      "learning_rate": 5e-05,
+      "loss": 1.6176,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5176374077112387,
+      "grad_norm": 0.15966758131980896,
+      "learning_rate": 5e-05,
+      "loss": 1.665,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5179108558928083,
+      "grad_norm": 0.17288130521774292,
+      "learning_rate": 5e-05,
+      "loss": 1.6342,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5181843040743779,
+      "grad_norm": 0.16509395837783813,
+      "learning_rate": 5e-05,
+      "loss": 1.649,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5184577522559475,
+      "grad_norm": 0.14995107054710388,
+      "learning_rate": 5e-05,
+      "loss": 1.4966,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5187312004375171,
+      "grad_norm": 0.15514351427555084,
+      "learning_rate": 5e-05,
+      "loss": 1.5691,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5190046486190867,
+      "grad_norm": 0.15947668254375458,
+      "learning_rate": 5e-05,
+      "loss": 1.534,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5192780968006563,
+      "grad_norm": 0.16156336665153503,
+      "learning_rate": 5e-05,
+      "loss": 1.7277,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5195515449822259,
+      "grad_norm": 0.17509359121322632,
+      "learning_rate": 5e-05,
+      "loss": 1.8131,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5198249931637955,
+      "grad_norm": 0.1574561595916748,
+      "learning_rate": 5e-05,
+      "loss": 1.618,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5200984413453651,
+      "grad_norm": 0.1618000864982605,
+      "learning_rate": 5e-05,
+      "loss": 1.5925,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5203718895269347,
+      "grad_norm": 0.15169738233089447,
+      "learning_rate": 5e-05,
+      "loss": 1.6704,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5206453377085042,
+      "grad_norm": 0.15747463703155518,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5209187858900738,
+      "grad_norm": 0.15733812749385834,
+      "learning_rate": 5e-05,
+      "loss": 1.6501,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5211922340716434,
+      "grad_norm": 0.15596060454845428,
+      "learning_rate": 5e-05,
+      "loss": 1.6661,
+      "step": 1906
+    },
+    {
+      "epoch": 0.521465682253213,
+      "grad_norm": 0.15629130601882935,
+      "learning_rate": 5e-05,
+      "loss": 1.652,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.15206284821033478,
+      "learning_rate": 5e-05,
+      "loss": 1.5696,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5220125786163522,
+      "grad_norm": 0.16571788489818573,
+      "learning_rate": 5e-05,
+      "loss": 1.6589,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5222860267979218,
+      "grad_norm": 0.1548844426870346,
+      "learning_rate": 5e-05,
+      "loss": 1.6127,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5225594749794914,
+      "grad_norm": 0.16375820338726044,
+      "learning_rate": 5e-05,
+      "loss": 1.5907,
+      "step": 1911
+    },
+    {
+      "epoch": 0.522832923161061,
+      "grad_norm": 0.15710307657718658,
+      "learning_rate": 5e-05,
+      "loss": 1.5947,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5231063713426306,
+      "grad_norm": 0.15390416979789734,
+      "learning_rate": 5e-05,
+      "loss": 1.6072,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5233798195242002,
+      "grad_norm": 0.15938615798950195,
+      "learning_rate": 5e-05,
+      "loss": 1.6348,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5236532677057698,
+      "grad_norm": 0.14976823329925537,
+      "learning_rate": 5e-05,
+      "loss": 1.5721,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5239267158873393,
+      "grad_norm": 0.15962772071361542,
+      "learning_rate": 5e-05,
+      "loss": 1.6077,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5242001640689089,
+      "grad_norm": 0.1549055576324463,
+      "learning_rate": 5e-05,
+      "loss": 1.5865,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5244736122504785,
+      "grad_norm": 0.1605559140443802,
+      "learning_rate": 5e-05,
+      "loss": 1.5902,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5247470604320481,
+      "grad_norm": 0.15829849243164062,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5250205086136177,
+      "grad_norm": 0.1654774397611618,
+      "learning_rate": 5e-05,
+      "loss": 1.6368,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5252939567951873,
+      "grad_norm": 0.16315412521362305,
+      "learning_rate": 5e-05,
+      "loss": 1.6415,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5255674049767569,
+      "grad_norm": 0.14765046536922455,
+      "learning_rate": 5e-05,
+      "loss": 1.6004,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5258408531583265,
+      "grad_norm": 0.15793031454086304,
+      "learning_rate": 5e-05,
+      "loss": 1.6902,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5261143013398961,
+      "grad_norm": 0.165854349732399,
+      "learning_rate": 5e-05,
+      "loss": 1.6598,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5263877495214657,
+      "grad_norm": 0.15993940830230713,
+      "learning_rate": 5e-05,
+      "loss": 1.6446,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5266611977030353,
+      "grad_norm": 0.1579129993915558,
+      "learning_rate": 5e-05,
+      "loss": 1.6392,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5269346458846048,
+      "grad_norm": 0.1714421659708023,
+      "learning_rate": 5e-05,
+      "loss": 1.6083,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5272080940661744,
+      "grad_norm": 0.15757611393928528,
+      "learning_rate": 5e-05,
+      "loss": 1.6344,
+      "step": 1928
+    },
+    {
+      "epoch": 0.527481542247744,
+      "grad_norm": 0.15477389097213745,
+      "learning_rate": 5e-05,
+      "loss": 1.6428,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5277549904293136,
+      "grad_norm": 0.14917831122875214,
+      "learning_rate": 5e-05,
+      "loss": 1.4957,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5280284386108832,
+      "grad_norm": 0.17004439234733582,
+      "learning_rate": 5e-05,
+      "loss": 1.7057,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.1737544983625412,
+      "learning_rate": 5e-05,
+      "loss": 1.6653,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5285753349740224,
+      "grad_norm": 0.15638841688632965,
+      "learning_rate": 5e-05,
+      "loss": 1.581,
+      "step": 1933
+    },
+    {
+      "epoch": 0.528848783155592,
+      "grad_norm": 0.19510416686534882,
+      "learning_rate": 5e-05,
+      "loss": 1.7202,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5291222313371616,
+      "grad_norm": 0.16090962290763855,
+      "learning_rate": 5e-05,
+      "loss": 1.6374,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5293956795187312,
+      "grad_norm": 0.1696002036333084,
+      "learning_rate": 5e-05,
+      "loss": 1.7889,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5296691277003008,
+      "grad_norm": 0.16609039902687073,
+      "learning_rate": 5e-05,
+      "loss": 1.6189,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5299425758818704,
+      "grad_norm": 0.1632707715034485,
+      "learning_rate": 5e-05,
+      "loss": 1.6715,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5302160240634399,
+      "grad_norm": 0.1675024926662445,
+      "learning_rate": 5e-05,
+      "loss": 1.6406,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5304894722450095,
+      "grad_norm": 0.1700305938720703,
+      "learning_rate": 5e-05,
+      "loss": 1.5702,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5307629204265791,
+      "grad_norm": 0.16249525547027588,
+      "learning_rate": 5e-05,
+      "loss": 1.6847,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5310363686081487,
+      "grad_norm": 0.16194301843643188,
+      "learning_rate": 5e-05,
+      "loss": 1.5392,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5313098167897183,
+      "grad_norm": 0.16195613145828247,
+      "learning_rate": 5e-05,
+      "loss": 1.6185,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5315832649712879,
+      "grad_norm": 0.15754897892475128,
+      "learning_rate": 5e-05,
+      "loss": 1.6188,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5318567131528575,
+      "grad_norm": 0.16981816291809082,
+      "learning_rate": 5e-05,
+      "loss": 1.6742,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5321301613344271,
+      "grad_norm": 0.16634345054626465,
+      "learning_rate": 5e-05,
+      "loss": 1.5054,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5324036095159967,
+      "grad_norm": 0.16170655190944672,
+      "learning_rate": 5e-05,
+      "loss": 1.6678,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5326770576975663,
+      "grad_norm": 0.1492849886417389,
+      "learning_rate": 5e-05,
+      "loss": 1.6136,
+      "step": 1948
+    },
+    {
+      "epoch": 0.532950505879136,
+      "grad_norm": 0.18234175443649292,
+      "learning_rate": 5e-05,
+      "loss": 1.7196,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5332239540607056,
+      "grad_norm": 0.158365860581398,
+      "learning_rate": 5e-05,
+      "loss": 1.4944,
+      "step": 1950
+    },
+    {
+      "epoch": 0.533497402242275,
+      "grad_norm": 0.1587635576725006,
+      "learning_rate": 5e-05,
+      "loss": 1.6334,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5337708504238446,
+      "grad_norm": 0.18955689668655396,
+      "learning_rate": 5e-05,
+      "loss": 1.6253,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5340442986054142,
+      "grad_norm": 0.1624482125043869,
+      "learning_rate": 5e-05,
+      "loss": 1.6511,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5343177467869838,
+      "grad_norm": 0.1539737582206726,
+      "learning_rate": 5e-05,
+      "loss": 1.4702,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5345911949685535,
+      "grad_norm": 0.17508608102798462,
+      "learning_rate": 5e-05,
+      "loss": 1.6406,
+      "step": 1955
+    },
+    {
+      "epoch": 0.534864643150123,
+      "grad_norm": 0.15725727379322052,
+      "learning_rate": 5e-05,
+      "loss": 1.5327,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5351380913316927,
+      "grad_norm": 0.16343533992767334,
+      "learning_rate": 5e-05,
+      "loss": 1.6075,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5354115395132623,
+      "grad_norm": 0.16122648119926453,
+      "learning_rate": 5e-05,
+      "loss": 1.577,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5356849876948319,
+      "grad_norm": 0.16027076542377472,
+      "learning_rate": 5e-05,
+      "loss": 1.6378,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5359584358764015,
+      "grad_norm": 0.15738053619861603,
+      "learning_rate": 5e-05,
+      "loss": 1.702,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5362318840579711,
+      "grad_norm": 0.15926112234592438,
+      "learning_rate": 5e-05,
+      "loss": 1.6522,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5365053322395406,
+      "grad_norm": 0.15066231787204742,
+      "learning_rate": 5e-05,
+      "loss": 1.4936,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5367787804211102,
+      "grad_norm": 0.14897476136684418,
+      "learning_rate": 5e-05,
+      "loss": 1.5098,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5370522286026798,
+      "grad_norm": 0.1546986699104309,
+      "learning_rate": 5e-05,
+      "loss": 1.5904,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5373256767842494,
+      "grad_norm": 0.15693390369415283,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 1965
+    },
+    {
+      "epoch": 0.537599124965819,
+      "grad_norm": 0.16184082627296448,
+      "learning_rate": 5e-05,
+      "loss": 1.7093,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5378725731473886,
+      "grad_norm": 0.157283753156662,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5381460213289582,
+      "grad_norm": 0.16195395588874817,
+      "learning_rate": 5e-05,
+      "loss": 1.6401,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5384194695105278,
+      "grad_norm": 0.16151578724384308,
+      "learning_rate": 5e-05,
+      "loss": 1.6325,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5386929176920974,
+      "grad_norm": 0.1499180942773819,
+      "learning_rate": 5e-05,
+      "loss": 1.5444,
+      "step": 1970
+    },
+    {
+      "epoch": 0.538966365873667,
+      "grad_norm": 0.16066747903823853,
+      "learning_rate": 5e-05,
+      "loss": 1.5921,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5392398140552366,
+      "grad_norm": 0.15843364596366882,
+      "learning_rate": 5e-05,
+      "loss": 1.5952,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5395132622368062,
+      "grad_norm": 0.15702027082443237,
+      "learning_rate": 5e-05,
+      "loss": 1.5061,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5397867104183757,
+      "grad_norm": 0.15099968016147614,
+      "learning_rate": 5e-05,
+      "loss": 1.6227,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5400601585999453,
+      "grad_norm": 0.15741010010242462,
+      "learning_rate": 5e-05,
+      "loss": 1.647,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5403336067815149,
+      "grad_norm": 0.1735711693763733,
+      "learning_rate": 5e-05,
+      "loss": 1.5785,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5406070549630845,
+      "grad_norm": 0.1666189730167389,
+      "learning_rate": 5e-05,
+      "loss": 1.6775,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5408805031446541,
+      "grad_norm": 0.17314961552619934,
+      "learning_rate": 5e-05,
+      "loss": 1.6201,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5411539513262237,
+      "grad_norm": 0.17330363392829895,
+      "learning_rate": 5e-05,
+      "loss": 1.6672,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5414273995077933,
+      "grad_norm": 0.16131727397441864,
+      "learning_rate": 5e-05,
+      "loss": 1.6451,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5417008476893629,
+      "grad_norm": 0.16308879852294922,
+      "learning_rate": 5e-05,
+      "loss": 1.5366,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5419742958709325,
+      "grad_norm": 0.1608009785413742,
+      "learning_rate": 5e-05,
+      "loss": 1.636,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5422477440525021,
+      "grad_norm": 0.1665000170469284,
+      "learning_rate": 5e-05,
+      "loss": 1.6159,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5425211922340717,
+      "grad_norm": 0.15207409858703613,
+      "learning_rate": 5e-05,
+      "loss": 1.5366,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5427946404156412,
+      "grad_norm": 0.17597833275794983,
+      "learning_rate": 5e-05,
+      "loss": 1.6525,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5430680885972108,
+      "grad_norm": 0.15393896400928497,
+      "learning_rate": 5e-05,
+      "loss": 1.5363,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5433415367787804,
+      "grad_norm": 0.17519448697566986,
+      "learning_rate": 5e-05,
+      "loss": 1.6806,
+      "step": 1987
+    },
+    {
+      "epoch": 0.54361498496035,
+      "grad_norm": 0.1612955778837204,
+      "learning_rate": 5e-05,
+      "loss": 1.531,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5438884331419196,
+      "grad_norm": 0.15321476757526398,
+      "learning_rate": 5e-05,
+      "loss": 1.4887,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5441618813234892,
+      "grad_norm": 0.1556374877691269,
+      "learning_rate": 5e-05,
+      "loss": 1.625,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5444353295050588,
+      "grad_norm": 0.1725417673587799,
+      "learning_rate": 5e-05,
+      "loss": 1.6286,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5447087776866284,
+      "grad_norm": 0.15200626850128174,
+      "learning_rate": 5e-05,
+      "loss": 1.5596,
+      "step": 1992
+    },
+    {
+      "epoch": 0.544982225868198,
+      "grad_norm": 0.1692056953907013,
+      "learning_rate": 5e-05,
+      "loss": 1.7184,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5452556740497676,
+      "grad_norm": 0.15530715882778168,
+      "learning_rate": 5e-05,
+      "loss": 1.493,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5455291222313372,
+      "grad_norm": 0.1559220254421234,
+      "learning_rate": 5e-05,
+      "loss": 1.5517,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5458025704129068,
+      "grad_norm": 0.15481142699718475,
+      "learning_rate": 5e-05,
+      "loss": 1.5246,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5460760185944763,
+      "grad_norm": 0.1693953573703766,
+      "learning_rate": 5e-05,
+      "loss": 1.6766,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5463494667760459,
+      "grad_norm": 0.14999975264072418,
+      "learning_rate": 5e-05,
+      "loss": 1.5715,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5466229149576155,
+      "grad_norm": 0.17395493388175964,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5468963631391851,
+      "grad_norm": 0.16458867490291595,
+      "learning_rate": 5e-05,
+      "loss": 1.5627,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5471698113207547,
+      "grad_norm": 0.15648111701011658,
+      "learning_rate": 5e-05,
+      "loss": 1.6565,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5474432595023243,
+      "grad_norm": 0.16169799864292145,
+      "learning_rate": 5e-05,
+      "loss": 1.6626,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5477167076838939,
+      "grad_norm": 0.17330998182296753,
+      "learning_rate": 5e-05,
+      "loss": 1.6268,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5479901558654635,
+      "grad_norm": 0.1588200032711029,
+      "learning_rate": 5e-05,
+      "loss": 1.5532,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5482636040470331,
+      "grad_norm": 0.16601422429084778,
+      "learning_rate": 5e-05,
+      "loss": 1.5976,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5485370522286027,
+      "grad_norm": 0.17725588381290436,
+      "learning_rate": 5e-05,
+      "loss": 1.6347,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5488105004101723,
+      "grad_norm": 0.15480291843414307,
+      "learning_rate": 5e-05,
+      "loss": 1.585,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5490839485917418,
+      "grad_norm": 0.1640474647283554,
+      "learning_rate": 5e-05,
+      "loss": 1.4857,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5493573967733114,
+      "grad_norm": 0.17515790462493896,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 2009
+    },
+    {
+      "epoch": 0.549630844954881,
+      "grad_norm": 0.15881620347499847,
+      "learning_rate": 5e-05,
+      "loss": 1.5243,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5499042931364506,
+      "grad_norm": 0.1577688604593277,
+      "learning_rate": 5e-05,
+      "loss": 1.6015,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5501777413180202,
+      "grad_norm": 0.16591089963912964,
+      "learning_rate": 5e-05,
+      "loss": 1.6122,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5504511894995898,
+      "grad_norm": 0.16563619673252106,
+      "learning_rate": 5e-05,
+      "loss": 1.6371,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5507246376811594,
+      "grad_norm": 0.17261956632137299,
+      "learning_rate": 5e-05,
+      "loss": 1.6834,
+      "step": 2014
+    },
+    {
+      "epoch": 0.550998085862729,
+      "grad_norm": 0.15990975499153137,
+      "learning_rate": 5e-05,
+      "loss": 1.6495,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5512715340442986,
+      "grad_norm": 0.15920180082321167,
+      "learning_rate": 5e-05,
+      "loss": 1.6299,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5515449822258682,
+      "grad_norm": 0.15567980706691742,
+      "learning_rate": 5e-05,
+      "loss": 1.5456,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5518184304074378,
+      "grad_norm": 0.155466690659523,
+      "learning_rate": 5e-05,
+      "loss": 1.6061,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5520918785890074,
+      "grad_norm": 0.16620704531669617,
+      "learning_rate": 5e-05,
+      "loss": 1.6645,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5523653267705769,
+      "grad_norm": 0.16219502687454224,
+      "learning_rate": 5e-05,
+      "loss": 1.7455,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5526387749521465,
+      "grad_norm": 0.1692618727684021,
+      "learning_rate": 5e-05,
+      "loss": 1.6036,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5529122231337161,
+      "grad_norm": 0.1535937339067459,
+      "learning_rate": 5e-05,
+      "loss": 1.614,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5531856713152857,
+      "grad_norm": 0.16089798510074615,
+      "learning_rate": 5e-05,
+      "loss": 1.585,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5534591194968553,
+      "grad_norm": 0.16011877357959747,
+      "learning_rate": 5e-05,
+      "loss": 1.6358,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5537325676784249,
+      "grad_norm": 0.1536594033241272,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5540060158599945,
+      "grad_norm": 0.15482646226882935,
+      "learning_rate": 5e-05,
+      "loss": 1.5291,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5542794640415641,
+      "grad_norm": 0.16264241933822632,
+      "learning_rate": 5e-05,
+      "loss": 1.5984,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5545529122231337,
+      "grad_norm": 0.16012269258499146,
+      "learning_rate": 5e-05,
+      "loss": 1.6847,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5548263604047033,
+      "grad_norm": 0.16615892946720123,
+      "learning_rate": 5e-05,
+      "loss": 1.5873,
+      "step": 2029
+    },
+    {
+      "epoch": 0.555099808586273,
+      "grad_norm": 0.15969812870025635,
+      "learning_rate": 5e-05,
+      "loss": 1.5295,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5553732567678425,
+      "grad_norm": 0.17502804100513458,
+      "learning_rate": 5e-05,
+      "loss": 1.6885,
+      "step": 2031
+    },
+    {
+      "epoch": 0.555646704949412,
+      "grad_norm": 0.15848223865032196,
+      "learning_rate": 5e-05,
+      "loss": 1.5479,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5559201531309816,
+      "grad_norm": 0.16243097186088562,
+      "learning_rate": 5e-05,
+      "loss": 1.6315,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5561936013125512,
+      "grad_norm": 0.15360400080680847,
+      "learning_rate": 5e-05,
+      "loss": 1.5927,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5564670494941208,
+      "grad_norm": 0.17348462343215942,
+      "learning_rate": 5e-05,
+      "loss": 1.6338,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5567404976756904,
+      "grad_norm": 0.14669735729694366,
+      "learning_rate": 5e-05,
+      "loss": 1.5976,
+      "step": 2036
+    },
+    {
+      "epoch": 0.55701394585726,
+      "grad_norm": 0.17374375462532043,
+      "learning_rate": 5e-05,
+      "loss": 1.6696,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5572873940388297,
+      "grad_norm": 0.1698565036058426,
+      "learning_rate": 5e-05,
+      "loss": 1.7155,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5575608422203993,
+      "grad_norm": 0.16311299800872803,
+      "learning_rate": 5e-05,
+      "loss": 1.5581,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5578342904019689,
+      "grad_norm": 0.1789819747209549,
+      "learning_rate": 5e-05,
+      "loss": 1.6041,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5581077385835385,
+      "grad_norm": 0.15918207168579102,
+      "learning_rate": 5e-05,
+      "loss": 1.6056,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5583811867651081,
+      "grad_norm": 0.1748282015323639,
+      "learning_rate": 5e-05,
+      "loss": 1.6903,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5586546349466776,
+      "grad_norm": 0.15822182595729828,
+      "learning_rate": 5e-05,
+      "loss": 1.6031,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5589280831282472,
+      "grad_norm": 0.15568383038043976,
+      "learning_rate": 5e-05,
+      "loss": 1.6616,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5592015313098168,
+      "grad_norm": 0.1490369737148285,
+      "learning_rate": 5e-05,
+      "loss": 1.554,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5594749794913864,
+      "grad_norm": 0.1597105860710144,
+      "learning_rate": 5e-05,
+      "loss": 1.5907,
+      "step": 2046
+    },
+    {
+      "epoch": 0.559748427672956,
+      "grad_norm": 0.15021972358226776,
+      "learning_rate": 5e-05,
+      "loss": 1.5364,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5600218758545256,
+      "grad_norm": 0.16100069880485535,
+      "learning_rate": 5e-05,
+      "loss": 1.6302,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5602953240360952,
+      "grad_norm": 0.1672513782978058,
+      "learning_rate": 5e-05,
+      "loss": 1.6362,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5605687722176648,
+      "grad_norm": 0.14868006110191345,
+      "learning_rate": 5e-05,
+      "loss": 1.5026,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5608422203992344,
+      "grad_norm": 0.15839874744415283,
+      "learning_rate": 5e-05,
+      "loss": 1.676,
+      "step": 2051
+    },
+    {
+      "epoch": 0.561115668580804,
+      "grad_norm": 0.15299920737743378,
+      "learning_rate": 5e-05,
+      "loss": 1.6009,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5613891167623736,
+      "grad_norm": 0.15907810628414154,
+      "learning_rate": 5e-05,
+      "loss": 1.6038,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5616625649439432,
+      "grad_norm": 0.15567612648010254,
+      "learning_rate": 5e-05,
+      "loss": 1.6237,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5619360131255127,
+      "grad_norm": 0.1545831263065338,
+      "learning_rate": 5e-05,
+      "loss": 1.5655,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5622094613070823,
+      "grad_norm": 0.15580891072750092,
+      "learning_rate": 5e-05,
+      "loss": 1.5292,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5624829094886519,
+      "grad_norm": 0.1534157246351242,
+      "learning_rate": 5e-05,
+      "loss": 1.5798,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5627563576702215,
+      "grad_norm": 0.17224334180355072,
+      "learning_rate": 5e-05,
+      "loss": 1.705,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5630298058517911,
+      "grad_norm": 0.15283888578414917,
+      "learning_rate": 5e-05,
+      "loss": 1.6189,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5633032540333607,
+      "grad_norm": 0.1531190425157547,
+      "learning_rate": 5e-05,
+      "loss": 1.4706,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5635767022149303,
+      "grad_norm": 0.15777452290058136,
+      "learning_rate": 5e-05,
+      "loss": 1.5925,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5638501503964999,
+      "grad_norm": 0.16896897554397583,
+      "learning_rate": 5e-05,
+      "loss": 1.6667,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5641235985780695,
+      "grad_norm": 0.15329276025295258,
+      "learning_rate": 5e-05,
+      "loss": 1.5491,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5643970467596391,
+      "grad_norm": 0.16101738810539246,
+      "learning_rate": 5e-05,
+      "loss": 1.6393,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5646704949412087,
+      "grad_norm": 0.1673561930656433,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5649439431227782,
+      "grad_norm": 0.1541615128517151,
+      "learning_rate": 5e-05,
+      "loss": 1.559,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 0.16007159650325775,
+      "learning_rate": 5e-05,
+      "loss": 1.584,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5654908394859174,
+      "grad_norm": 0.1529233604669571,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 2068
+    },
+    {
+      "epoch": 0.565764287667487,
+      "grad_norm": 0.1614188849925995,
+      "learning_rate": 5e-05,
+      "loss": 1.6406,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.15437230467796326,
+      "learning_rate": 5e-05,
+      "loss": 1.6067,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5663111840306262,
+      "grad_norm": 0.15617471933364868,
+      "learning_rate": 5e-05,
+      "loss": 1.6215,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5665846322121958,
+      "grad_norm": 0.15771955251693726,
+      "learning_rate": 5e-05,
+      "loss": 1.5109,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5668580803937654,
+      "grad_norm": 0.1565471887588501,
+      "learning_rate": 5e-05,
+      "loss": 1.5407,
+      "step": 2073
+    },
+    {
+      "epoch": 0.567131528575335,
+      "grad_norm": 0.16085505485534668,
+      "learning_rate": 5e-05,
+      "loss": 1.6118,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5674049767569046,
+      "grad_norm": 0.15981802344322205,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5676784249384742,
+      "grad_norm": 0.16176685690879822,
+      "learning_rate": 5e-05,
+      "loss": 1.6723,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5679518731200438,
+      "grad_norm": 0.17309653759002686,
+      "learning_rate": 5e-05,
+      "loss": 1.7196,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5682253213016133,
+      "grad_norm": 0.15585723519325256,
+      "learning_rate": 5e-05,
+      "loss": 1.5884,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5684987694831829,
+      "grad_norm": 0.1600128412246704,
+      "learning_rate": 5e-05,
+      "loss": 1.6116,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5687722176647525,
+      "grad_norm": 0.16564618051052094,
+      "learning_rate": 5e-05,
+      "loss": 1.5695,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5690456658463221,
+      "grad_norm": 0.15419046580791473,
+      "learning_rate": 5e-05,
+      "loss": 1.5458,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5693191140278917,
+      "grad_norm": 0.17247512936592102,
+      "learning_rate": 5e-05,
+      "loss": 1.6021,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5695925622094613,
+      "grad_norm": 0.16553881764411926,
+      "learning_rate": 5e-05,
+      "loss": 1.5771,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5698660103910309,
+      "grad_norm": 0.16219867765903473,
+      "learning_rate": 5e-05,
+      "loss": 1.6749,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5701394585726005,
+      "grad_norm": 0.17330633103847504,
+      "learning_rate": 5e-05,
+      "loss": 1.6143,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5704129067541701,
+      "grad_norm": 0.16274374723434448,
+      "learning_rate": 5e-05,
+      "loss": 1.6205,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5706863549357397,
+      "grad_norm": 0.15556305646896362,
+      "learning_rate": 5e-05,
+      "loss": 1.5875,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5709598031173093,
+      "grad_norm": 0.15635040402412415,
+      "learning_rate": 5e-05,
+      "loss": 1.6349,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5712332512988788,
+      "grad_norm": 0.160813108086586,
+      "learning_rate": 5e-05,
+      "loss": 1.604,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5715066994804484,
+      "grad_norm": 0.15628398954868317,
+      "learning_rate": 5e-05,
+      "loss": 1.5023,
+      "step": 2090
+    },
+    {
+      "epoch": 0.571780147662018,
+      "grad_norm": 0.18704648315906525,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5720535958435876,
+      "grad_norm": 0.1672285944223404,
+      "learning_rate": 5e-05,
+      "loss": 1.643,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5723270440251572,
+      "grad_norm": 0.15511822700500488,
+      "learning_rate": 5e-05,
+      "loss": 1.6143,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5726004922067268,
+      "grad_norm": 0.16968627274036407,
+      "learning_rate": 5e-05,
+      "loss": 1.649,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5728739403882964,
+      "grad_norm": 0.16604706645011902,
+      "learning_rate": 5e-05,
+      "loss": 1.6111,
+      "step": 2095
+    },
+    {
+      "epoch": 0.573147388569866,
+      "grad_norm": 0.16677343845367432,
+      "learning_rate": 5e-05,
+      "loss": 1.6142,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5734208367514356,
+      "grad_norm": 0.16563932597637177,
+      "learning_rate": 5e-05,
+      "loss": 1.6043,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5736942849330052,
+      "grad_norm": 0.15738388895988464,
+      "learning_rate": 5e-05,
+      "loss": 1.5803,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5739677331145748,
+      "grad_norm": 0.15241163969039917,
+      "learning_rate": 5e-05,
+      "loss": 1.5756,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5742411812961444,
+      "grad_norm": 0.1751902997493744,
+      "learning_rate": 5e-05,
+      "loss": 1.6738,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5745146294777139,
+      "grad_norm": 0.15986758470535278,
+      "learning_rate": 5e-05,
+      "loss": 1.5693,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5747880776592835,
+      "grad_norm": 0.19663068652153015,
+      "learning_rate": 5e-05,
+      "loss": 1.59,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5750615258408531,
+      "grad_norm": 0.15884363651275635,
+      "learning_rate": 5e-05,
+      "loss": 1.6378,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5753349740224227,
+      "grad_norm": 0.1701025813817978,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5756084222039923,
+      "grad_norm": 0.1757257878780365,
+      "learning_rate": 5e-05,
+      "loss": 1.7625,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5758818703855619,
+      "grad_norm": 0.15682871639728546,
+      "learning_rate": 5e-05,
+      "loss": 1.6039,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5761553185671315,
+      "grad_norm": 0.15779271721839905,
+      "learning_rate": 5e-05,
+      "loss": 1.6915,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5764287667487011,
+      "grad_norm": 0.15956608951091766,
+      "learning_rate": 5e-05,
+      "loss": 1.622,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5767022149302707,
+      "grad_norm": 0.15776754915714264,
+      "learning_rate": 5e-05,
+      "loss": 1.6435,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5769756631118403,
+      "grad_norm": 0.17491094768047333,
+      "learning_rate": 5e-05,
+      "loss": 1.5058,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5772491112934099,
+      "grad_norm": 0.15431523323059082,
+      "learning_rate": 5e-05,
+      "loss": 1.6608,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5775225594749795,
+      "grad_norm": 0.15970687568187714,
+      "learning_rate": 5e-05,
+      "loss": 1.6203,
+      "step": 2112
+    },
+    {
+      "epoch": 0.577796007656549,
+      "grad_norm": 0.16885408759117126,
+      "learning_rate": 5e-05,
+      "loss": 1.6354,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5780694558381186,
+      "grad_norm": 0.1878814995288849,
+      "learning_rate": 5e-05,
+      "loss": 1.6947,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5783429040196882,
+      "grad_norm": 0.17708611488342285,
+      "learning_rate": 5e-05,
+      "loss": 1.5832,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5786163522012578,
+      "grad_norm": 0.16223332285881042,
+      "learning_rate": 5e-05,
+      "loss": 1.7159,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5788898003828274,
+      "grad_norm": 0.1625894010066986,
+      "learning_rate": 5e-05,
+      "loss": 1.6246,
+      "step": 2117
+    },
+    {
+      "epoch": 0.579163248564397,
+      "grad_norm": 0.1753711998462677,
+      "learning_rate": 5e-05,
+      "loss": 1.629,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5794366967459667,
+      "grad_norm": 0.15611834824085236,
+      "learning_rate": 5e-05,
+      "loss": 1.6562,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5797101449275363,
+      "grad_norm": 0.1676332801580429,
+      "learning_rate": 5e-05,
+      "loss": 1.6435,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5799835931091059,
+      "grad_norm": 0.16473741829395294,
+      "learning_rate": 5e-05,
+      "loss": 1.5665,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5802570412906755,
+      "grad_norm": 0.1505088359117508,
+      "learning_rate": 5e-05,
+      "loss": 1.5784,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5805304894722451,
+      "grad_norm": 0.1565895527601242,
+      "learning_rate": 5e-05,
+      "loss": 1.6327,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5808039376538146,
+      "grad_norm": 0.16196846961975098,
+      "learning_rate": 5e-05,
+      "loss": 1.503,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5810773858353842,
+      "grad_norm": 0.1584513634443283,
+      "learning_rate": 5e-05,
+      "loss": 1.6884,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5813508340169538,
+      "grad_norm": 0.1517631560564041,
+      "learning_rate": 5e-05,
+      "loss": 1.4673,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5816242821985234,
+      "grad_norm": 0.15401685237884521,
+      "learning_rate": 5e-05,
+      "loss": 1.5718,
+      "step": 2127
+    },
+    {
+      "epoch": 0.581897730380093,
+      "grad_norm": 0.1545703262090683,
+      "learning_rate": 5e-05,
+      "loss": 1.6156,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5821711785616626,
+      "grad_norm": 0.1678171306848526,
+      "learning_rate": 5e-05,
+      "loss": 1.7368,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5824446267432322,
+      "grad_norm": 0.15641961991786957,
+      "learning_rate": 5e-05,
+      "loss": 1.7208,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5827180749248018,
+      "grad_norm": 0.16505669057369232,
+      "learning_rate": 5e-05,
+      "loss": 1.6648,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5829915231063714,
+      "grad_norm": 0.15975305438041687,
+      "learning_rate": 5e-05,
+      "loss": 1.577,
+      "step": 2132
+    },
+    {
+      "epoch": 0.583264971287941,
+      "grad_norm": 0.15941214561462402,
+      "learning_rate": 5e-05,
+      "loss": 1.6433,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5835384194695106,
+      "grad_norm": 0.15701207518577576,
+      "learning_rate": 5e-05,
+      "loss": 1.5609,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5838118676510802,
+      "grad_norm": 0.1584734469652176,
+      "learning_rate": 5e-05,
+      "loss": 1.5644,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5840853158326497,
+      "grad_norm": 0.149795264005661,
+      "learning_rate": 5e-05,
+      "loss": 1.5459,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5843587640142193,
+      "grad_norm": 0.16000156104564667,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5846322121957889,
+      "grad_norm": 0.14752228558063507,
+      "learning_rate": 5e-05,
+      "loss": 1.4835,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5849056603773585,
+      "grad_norm": 0.1613810956478119,
+      "learning_rate": 5e-05,
+      "loss": 1.5714,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5851791085589281,
+      "grad_norm": 0.15406599640846252,
+      "learning_rate": 5e-05,
+      "loss": 1.5929,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5854525567404977,
+      "grad_norm": 0.15958181023597717,
+      "learning_rate": 5e-05,
+      "loss": 1.6232,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5857260049220673,
+      "grad_norm": 0.16171377897262573,
+      "learning_rate": 5e-05,
+      "loss": 1.7343,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5859994531036369,
+      "grad_norm": 0.15244810283184052,
+      "learning_rate": 5e-05,
+      "loss": 1.6367,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5862729012852065,
+      "grad_norm": 0.16295252740383148,
+      "learning_rate": 5e-05,
+      "loss": 1.6822,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5865463494667761,
+      "grad_norm": 0.16270601749420166,
+      "learning_rate": 5e-05,
+      "loss": 1.5552,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5868197976483457,
+      "grad_norm": 0.1615038365125656,
+      "learning_rate": 5e-05,
+      "loss": 1.6273,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5870932458299152,
+      "grad_norm": 0.17501547932624817,
+      "learning_rate": 5e-05,
+      "loss": 1.6598,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5873666940114848,
+      "grad_norm": 0.15854844450950623,
+      "learning_rate": 5e-05,
+      "loss": 1.6657,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5876401421930544,
+      "grad_norm": 0.15156009793281555,
+      "learning_rate": 5e-05,
+      "loss": 1.6571,
+      "step": 2149
+    },
+    {
+      "epoch": 0.587913590374624,
+      "grad_norm": 0.15848124027252197,
+      "learning_rate": 5e-05,
+      "loss": 1.6176,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5881870385561936,
+      "grad_norm": 0.15376807749271393,
+      "learning_rate": 5e-05,
+      "loss": 1.659,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5884604867377632,
+      "grad_norm": 0.15940286219120026,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5887339349193328,
+      "grad_norm": 0.1563788503408432,
+      "learning_rate": 5e-05,
+      "loss": 1.7226,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5890073831009024,
+      "grad_norm": 0.14890335500240326,
+      "learning_rate": 5e-05,
+      "loss": 1.5368,
+      "step": 2154
+    },
+    {
+      "epoch": 0.589280831282472,
+      "grad_norm": 0.16119109094142914,
+      "learning_rate": 5e-05,
+      "loss": 1.64,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5895542794640416,
+      "grad_norm": 0.15738506615161896,
+      "learning_rate": 5e-05,
+      "loss": 1.6306,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5898277276456112,
+      "grad_norm": 0.15417338907718658,
+      "learning_rate": 5e-05,
+      "loss": 1.5886,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5901011758271808,
+      "grad_norm": 0.15935613214969635,
+      "learning_rate": 5e-05,
+      "loss": 1.7004,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5903746240087503,
+      "grad_norm": 0.15843814611434937,
+      "learning_rate": 5e-05,
+      "loss": 1.684,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5906480721903199,
+      "grad_norm": 0.15500840544700623,
+      "learning_rate": 5e-05,
+      "loss": 1.6215,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5909215203718895,
+      "grad_norm": 0.15437741577625275,
+      "learning_rate": 5e-05,
+      "loss": 1.5473,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5911949685534591,
+      "grad_norm": 0.17060165107250214,
+      "learning_rate": 5e-05,
+      "loss": 1.6937,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5914684167350287,
+      "grad_norm": 0.15199008584022522,
+      "learning_rate": 5e-05,
+      "loss": 1.5557,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5917418649165983,
+      "grad_norm": 0.16401898860931396,
+      "learning_rate": 5e-05,
+      "loss": 1.6955,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5920153130981679,
+      "grad_norm": 0.16618569195270538,
+      "learning_rate": 5e-05,
+      "loss": 1.6962,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5922887612797375,
+      "grad_norm": 0.1566200703382492,
+      "learning_rate": 5e-05,
+      "loss": 1.6495,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5925622094613071,
+      "grad_norm": 0.15811263024806976,
+      "learning_rate": 5e-05,
+      "loss": 1.6775,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5928356576428767,
+      "grad_norm": 0.16176708042621613,
+      "learning_rate": 5e-05,
+      "loss": 1.5412,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5931091058244463,
+      "grad_norm": 0.15909145772457123,
+      "learning_rate": 5e-05,
+      "loss": 1.6401,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5933825540060158,
+      "grad_norm": 0.15169541537761688,
+      "learning_rate": 5e-05,
+      "loss": 1.6311,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5936560021875854,
+      "grad_norm": 0.1585574597120285,
+      "learning_rate": 5e-05,
+      "loss": 1.613,
+      "step": 2171
+    },
+    {
+      "epoch": 0.593929450369155,
+      "grad_norm": 0.1560421884059906,
+      "learning_rate": 5e-05,
+      "loss": 1.581,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5942028985507246,
+      "grad_norm": 0.1590547412633896,
+      "learning_rate": 5e-05,
+      "loss": 1.581,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5944763467322942,
+      "grad_norm": 0.1633371263742447,
+      "learning_rate": 5e-05,
+      "loss": 1.5969,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5947497949138638,
+      "grad_norm": 0.16914770007133484,
+      "learning_rate": 5e-05,
+      "loss": 1.6087,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5950232430954334,
+      "grad_norm": 0.15823769569396973,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 2176
+    },
+    {
+      "epoch": 0.595296691277003,
+      "grad_norm": 0.1536397933959961,
+      "learning_rate": 5e-05,
+      "loss": 1.5132,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5955701394585726,
+      "grad_norm": 0.16653746366500854,
+      "learning_rate": 5e-05,
+      "loss": 1.5937,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5958435876401422,
+      "grad_norm": 0.17655491828918457,
+      "learning_rate": 5e-05,
+      "loss": 1.7522,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5961170358217118,
+      "grad_norm": 0.16123391687870026,
+      "learning_rate": 5e-05,
+      "loss": 1.6493,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5963904840032814,
+      "grad_norm": 0.16177600622177124,
+      "learning_rate": 5e-05,
+      "loss": 1.6801,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5966639321848509,
+      "grad_norm": 0.16101764142513275,
+      "learning_rate": 5e-05,
+      "loss": 1.6148,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5969373803664205,
+      "grad_norm": 0.15031690895557404,
+      "learning_rate": 5e-05,
+      "loss": 1.5556,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5972108285479901,
+      "grad_norm": 0.15750014781951904,
+      "learning_rate": 5e-05,
+      "loss": 1.6032,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5974842767295597,
+      "grad_norm": 0.16449135541915894,
+      "learning_rate": 5e-05,
+      "loss": 1.6762,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5977577249111293,
+      "grad_norm": 0.1698635369539261,
+      "learning_rate": 5e-05,
+      "loss": 1.6279,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5980311730926989,
+      "grad_norm": 0.16983488202095032,
+      "learning_rate": 5e-05,
+      "loss": 1.5975,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5983046212742685,
+      "grad_norm": 0.16298316419124603,
+      "learning_rate": 5e-05,
+      "loss": 1.603,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5985780694558381,
+      "grad_norm": 0.16714924573898315,
+      "learning_rate": 5e-05,
+      "loss": 1.6296,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5988515176374077,
+      "grad_norm": 0.16718335449695587,
+      "learning_rate": 5e-05,
+      "loss": 1.7187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5991249658189773,
+      "grad_norm": 0.16621999442577362,
+      "learning_rate": 5e-05,
+      "loss": 1.631,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5993984140005469,
+      "grad_norm": 0.15571008622646332,
+      "learning_rate": 5e-05,
+      "loss": 1.6298,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5996718621821165,
+      "grad_norm": 0.16420917212963104,
+      "learning_rate": 5e-05,
+      "loss": 1.721,
+      "step": 2193
+    },
+    {
+      "epoch": 0.599945310363686,
+      "grad_norm": 0.174006387591362,
+      "learning_rate": 5e-05,
+      "loss": 1.6524,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6002187585452556,
+      "grad_norm": 0.1774541437625885,
+      "learning_rate": 5e-05,
+      "loss": 1.7686,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6004922067268252,
+      "grad_norm": 0.1665443480014801,
+      "learning_rate": 5e-05,
+      "loss": 1.6123,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6007656549083948,
+      "grad_norm": 0.1589983105659485,
+      "learning_rate": 5e-05,
+      "loss": 1.6771,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6010391030899644,
+      "grad_norm": 0.171598419547081,
+      "learning_rate": 5e-05,
+      "loss": 1.6487,
+      "step": 2198
+    },
+    {
+      "epoch": 0.601312551271534,
+      "grad_norm": 0.16494937241077423,
+      "learning_rate": 5e-05,
+      "loss": 1.5661,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6015859994531036,
+      "grad_norm": 0.16418486833572388,
+      "learning_rate": 5e-05,
+      "loss": 1.5924,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6018594476346733,
+      "grad_norm": 0.16469943523406982,
+      "learning_rate": 5e-05,
+      "loss": 1.6722,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6021328958162429,
+      "grad_norm": 0.1632019281387329,
+      "learning_rate": 5e-05,
+      "loss": 1.5975,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6024063439978125,
+      "grad_norm": 0.1550121158361435,
+      "learning_rate": 5e-05,
+      "loss": 1.6071,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6026797921793821,
+      "grad_norm": 0.16901268064975739,
+      "learning_rate": 5e-05,
+      "loss": 1.7557,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6029532403609515,
+      "grad_norm": 0.16295814514160156,
+      "learning_rate": 5e-05,
+      "loss": 1.6243,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6032266885425212,
+      "grad_norm": 0.15237173438072205,
+      "learning_rate": 5e-05,
+      "loss": 1.5015,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6035001367240908,
+      "grad_norm": 0.17193639278411865,
+      "learning_rate": 5e-05,
+      "loss": 1.6746,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.154008686542511,
+      "learning_rate": 5e-05,
+      "loss": 1.5166,
+      "step": 2208
+    },
+    {
+      "epoch": 0.60404703308723,
+      "grad_norm": 0.1634528785943985,
+      "learning_rate": 5e-05,
+      "loss": 1.6208,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6043204812687996,
+      "grad_norm": 0.16617950797080994,
+      "learning_rate": 5e-05,
+      "loss": 1.6649,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6045939294503692,
+      "grad_norm": 0.1584254801273346,
+      "learning_rate": 5e-05,
+      "loss": 1.6875,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6048673776319388,
+      "grad_norm": 0.15937229990959167,
+      "learning_rate": 5e-05,
+      "loss": 1.5211,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6051408258135084,
+      "grad_norm": 0.15841583907604218,
+      "learning_rate": 5e-05,
+      "loss": 1.5409,
+      "step": 2213
+    },
+    {
+      "epoch": 0.605414273995078,
+      "grad_norm": 0.15823234617710114,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6056877221766476,
+      "grad_norm": 0.15863433480262756,
+      "learning_rate": 5e-05,
+      "loss": 1.59,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6059611703582172,
+      "grad_norm": 0.16127796471118927,
+      "learning_rate": 5e-05,
+      "loss": 1.6605,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6062346185397867,
+      "grad_norm": 0.15523792803287506,
+      "learning_rate": 5e-05,
+      "loss": 1.551,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6065080667213563,
+      "grad_norm": 0.16734172403812408,
+      "learning_rate": 5e-05,
+      "loss": 1.5959,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6067815149029259,
+      "grad_norm": 0.16805590689182281,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6070549630844955,
+      "grad_norm": 0.1635158360004425,
+      "learning_rate": 5e-05,
+      "loss": 1.7377,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6073284112660651,
+      "grad_norm": 0.17108480632305145,
+      "learning_rate": 5e-05,
+      "loss": 1.6587,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6076018594476347,
+      "grad_norm": 0.1763361096382141,
+      "learning_rate": 5e-05,
+      "loss": 1.735,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6078753076292043,
+      "grad_norm": 0.153673455119133,
+      "learning_rate": 5e-05,
+      "loss": 1.5807,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6081487558107739,
+      "grad_norm": 0.17884378135204315,
+      "learning_rate": 5e-05,
+      "loss": 1.7078,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6084222039923435,
+      "grad_norm": 0.2724404036998749,
+      "learning_rate": 5e-05,
+      "loss": 1.72,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.16155698895454407,
+      "learning_rate": 5e-05,
+      "loss": 1.6043,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6089691003554827,
+      "grad_norm": 0.17532581090927124,
+      "learning_rate": 5e-05,
+      "loss": 1.6597,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6092425485370522,
+      "grad_norm": 0.1655532866716385,
+      "learning_rate": 5e-05,
+      "loss": 1.6069,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6095159967186218,
+      "grad_norm": 0.158722385764122,
+      "learning_rate": 5e-05,
+      "loss": 1.5885,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6097894449001914,
+      "grad_norm": 0.17299902439117432,
+      "learning_rate": 5e-05,
+      "loss": 1.5863,
+      "step": 2230
+    },
+    {
+      "epoch": 0.610062893081761,
+      "grad_norm": 0.17708896100521088,
+      "learning_rate": 5e-05,
+      "loss": 1.656,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6103363412633306,
+      "grad_norm": 0.16440723836421967,
+      "learning_rate": 5e-05,
+      "loss": 1.6651,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6106097894449002,
+      "grad_norm": 0.16679327189922333,
+      "learning_rate": 5e-05,
+      "loss": 1.6108,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6108832376264698,
+      "grad_norm": 0.18106336891651154,
+      "learning_rate": 5e-05,
+      "loss": 1.6973,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6111566858080394,
+      "grad_norm": 0.1723693609237671,
+      "learning_rate": 5e-05,
+      "loss": 1.6928,
+      "step": 2235
+    },
+    {
+      "epoch": 0.611430133989609,
+      "grad_norm": 0.1619867980480194,
+      "learning_rate": 5e-05,
+      "loss": 1.5845,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6117035821711786,
+      "grad_norm": 0.18252336978912354,
+      "learning_rate": 5e-05,
+      "loss": 1.6004,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6119770303527482,
+      "grad_norm": 0.166873037815094,
+      "learning_rate": 5e-05,
+      "loss": 1.6284,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6122504785343178,
+      "grad_norm": 0.16315174102783203,
+      "learning_rate": 5e-05,
+      "loss": 1.607,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6125239267158873,
+      "grad_norm": 0.17975984513759613,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6127973748974569,
+      "grad_norm": 0.1571521908044815,
+      "learning_rate": 5e-05,
+      "loss": 1.6401,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6130708230790265,
+      "grad_norm": 0.16314658522605896,
+      "learning_rate": 5e-05,
+      "loss": 1.5639,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6133442712605961,
+      "grad_norm": 0.1598280519247055,
+      "learning_rate": 5e-05,
+      "loss": 1.5334,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6136177194421657,
+      "grad_norm": 0.15419380366802216,
+      "learning_rate": 5e-05,
+      "loss": 1.5093,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6138911676237353,
+      "grad_norm": 0.15684397518634796,
+      "learning_rate": 5e-05,
+      "loss": 1.6236,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6141646158053049,
+      "grad_norm": 0.1650727540254593,
+      "learning_rate": 5e-05,
+      "loss": 1.7009,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6144380639868745,
+      "grad_norm": 0.16277483105659485,
+      "learning_rate": 5e-05,
+      "loss": 1.6878,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6147115121684441,
+      "grad_norm": 0.1605841964483261,
+      "learning_rate": 5e-05,
+      "loss": 1.6445,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6149849603500137,
+      "grad_norm": 0.16627003252506256,
+      "learning_rate": 5e-05,
+      "loss": 1.6217,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6152584085315833,
+      "grad_norm": 0.15815889835357666,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6155318567131528,
+      "grad_norm": 0.15406928956508636,
+      "learning_rate": 5e-05,
+      "loss": 1.6202,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6158053048947224,
+      "grad_norm": 0.1534588783979416,
+      "learning_rate": 5e-05,
+      "loss": 1.5494,
+      "step": 2252
+    },
+    {
+      "epoch": 0.616078753076292,
+      "grad_norm": 0.15792147815227509,
+      "learning_rate": 5e-05,
+      "loss": 1.5803,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6163522012578616,
+      "grad_norm": 0.15371176600456238,
+      "learning_rate": 5e-05,
+      "loss": 1.5671,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6166256494394312,
+      "grad_norm": 0.15262968838214874,
+      "learning_rate": 5e-05,
+      "loss": 1.5196,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6168990976210008,
+      "grad_norm": 0.15933872759342194,
+      "learning_rate": 5e-05,
+      "loss": 1.5735,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6171725458025704,
+      "grad_norm": 0.15623560547828674,
+      "learning_rate": 5e-05,
+      "loss": 1.529,
+      "step": 2257
+    },
+    {
+      "epoch": 0.61744599398414,
+      "grad_norm": 0.15916681289672852,
+      "learning_rate": 5e-05,
+      "loss": 1.5828,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6177194421657096,
+      "grad_norm": 0.15046894550323486,
+      "learning_rate": 5e-05,
+      "loss": 1.6456,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6179928903472792,
+      "grad_norm": 0.15784558653831482,
+      "learning_rate": 5e-05,
+      "loss": 1.5167,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6182663385288488,
+      "grad_norm": 0.16395771503448486,
+      "learning_rate": 5e-05,
+      "loss": 1.5806,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6185397867104184,
+      "grad_norm": 0.15254901349544525,
+      "learning_rate": 5e-05,
+      "loss": 1.5617,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6188132348919879,
+      "grad_norm": 0.15680168569087982,
+      "learning_rate": 5e-05,
+      "loss": 1.5795,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6190866830735575,
+      "grad_norm": 0.1724022775888443,
+      "learning_rate": 5e-05,
+      "loss": 1.6245,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6193601312551271,
+      "grad_norm": 0.1564149111509323,
+      "learning_rate": 5e-05,
+      "loss": 1.5742,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6196335794366967,
+      "grad_norm": 0.18149738013744354,
+      "learning_rate": 5e-05,
+      "loss": 1.5351,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6199070276182663,
+      "grad_norm": 0.15708599984645844,
+      "learning_rate": 5e-05,
+      "loss": 1.6182,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6201804757998359,
+      "grad_norm": 0.16995084285736084,
+      "learning_rate": 5e-05,
+      "loss": 1.6893,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6204539239814055,
+      "grad_norm": 0.16211822628974915,
+      "learning_rate": 5e-05,
+      "loss": 1.702,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6207273721629751,
+      "grad_norm": 0.16163219511508942,
+      "learning_rate": 5e-05,
+      "loss": 1.6688,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6210008203445447,
+      "grad_norm": 0.16371332108974457,
+      "learning_rate": 5e-05,
+      "loss": 1.5554,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6212742685261143,
+      "grad_norm": 0.1574903279542923,
+      "learning_rate": 5e-05,
+      "loss": 1.6372,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6215477167076839,
+      "grad_norm": 0.15815797448158264,
+      "learning_rate": 5e-05,
+      "loss": 1.5674,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6218211648892535,
+      "grad_norm": 0.1578124612569809,
+      "learning_rate": 5e-05,
+      "loss": 1.5226,
+      "step": 2274
+    },
+    {
+      "epoch": 0.622094613070823,
+      "grad_norm": 0.1595599204301834,
+      "learning_rate": 5e-05,
+      "loss": 1.5657,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6223680612523926,
+      "grad_norm": 0.15654776990413666,
+      "learning_rate": 5e-05,
+      "loss": 1.6009,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6226415094339622,
+      "grad_norm": 0.1564917415380478,
+      "learning_rate": 5e-05,
+      "loss": 1.6166,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6229149576155318,
+      "grad_norm": 0.1595001518726349,
+      "learning_rate": 5e-05,
+      "loss": 1.5389,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6231884057971014,
+      "grad_norm": 0.16073504090309143,
+      "learning_rate": 5e-05,
+      "loss": 1.5948,
+      "step": 2279
+    },
+    {
+      "epoch": 0.623461853978671,
+      "grad_norm": 0.15986725687980652,
+      "learning_rate": 5e-05,
+      "loss": 1.6057,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6237353021602406,
+      "grad_norm": 0.1577160656452179,
+      "learning_rate": 5e-05,
+      "loss": 1.6144,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6240087503418102,
+      "grad_norm": 0.1593600958585739,
+      "learning_rate": 5e-05,
+      "loss": 1.6538,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6242821985233798,
+      "grad_norm": 0.15497250854969025,
+      "learning_rate": 5e-05,
+      "loss": 1.5714,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6245556467049495,
+      "grad_norm": 0.1659761667251587,
+      "learning_rate": 5e-05,
+      "loss": 1.6146,
+      "step": 2284
+    },
+    {
+      "epoch": 0.624829094886519,
+      "grad_norm": 0.16305339336395264,
+      "learning_rate": 5e-05,
+      "loss": 1.6935,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6251025430680885,
+      "grad_norm": 0.16828620433807373,
+      "learning_rate": 5e-05,
+      "loss": 1.6341,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6253759912496581,
+      "grad_norm": 0.16056501865386963,
+      "learning_rate": 5e-05,
+      "loss": 1.6103,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6256494394312277,
+      "grad_norm": 0.1658848077058792,
+      "learning_rate": 5e-05,
+      "loss": 1.6548,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6259228876127974,
+      "grad_norm": 0.1599251627922058,
+      "learning_rate": 5e-05,
+      "loss": 1.5715,
+      "step": 2289
+    },
+    {
+      "epoch": 0.626196335794367,
+      "grad_norm": 0.15655700862407684,
+      "learning_rate": 5e-05,
+      "loss": 1.5145,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6264697839759366,
+      "grad_norm": 0.16632331907749176,
+      "learning_rate": 5e-05,
+      "loss": 1.6599,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6267432321575062,
+      "grad_norm": 0.16444595158100128,
+      "learning_rate": 5e-05,
+      "loss": 1.6591,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6270166803390758,
+      "grad_norm": 0.1640361100435257,
+      "learning_rate": 5e-05,
+      "loss": 1.5751,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6272901285206454,
+      "grad_norm": 0.16003577411174774,
+      "learning_rate": 5e-05,
+      "loss": 1.6096,
+      "step": 2294
+    },
+    {
+      "epoch": 0.627563576702215,
+      "grad_norm": 0.16625234484672546,
+      "learning_rate": 5e-05,
+      "loss": 1.6685,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6278370248837846,
+      "grad_norm": 0.15295128524303436,
+      "learning_rate": 5e-05,
+      "loss": 1.569,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6281104730653542,
+      "grad_norm": 0.16834108531475067,
+      "learning_rate": 5e-05,
+      "loss": 1.6813,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6283839212469237,
+      "grad_norm": 0.16331374645233154,
+      "learning_rate": 5e-05,
+      "loss": 1.5879,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6286573694284933,
+      "grad_norm": 0.17699094116687775,
+      "learning_rate": 5e-05,
+      "loss": 1.673,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6289308176100629,
+      "grad_norm": 0.1626228392124176,
+      "learning_rate": 5e-05,
+      "loss": 1.6703,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6292042657916325,
+      "grad_norm": 0.16163185238838196,
+      "learning_rate": 5e-05,
+      "loss": 1.6249,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6294777139732021,
+      "grad_norm": 0.17444822192192078,
+      "learning_rate": 5e-05,
+      "loss": 1.5542,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6297511621547717,
+      "grad_norm": 0.1661493182182312,
+      "learning_rate": 5e-05,
+      "loss": 1.6295,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6300246103363413,
+      "grad_norm": 0.1628965586423874,
+      "learning_rate": 5e-05,
+      "loss": 1.5952,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6302980585179109,
+      "grad_norm": 0.1656888723373413,
+      "learning_rate": 5e-05,
+      "loss": 1.594,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6305715066994805,
+      "grad_norm": 0.15425047278404236,
+      "learning_rate": 5e-05,
+      "loss": 1.4985,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6308449548810501,
+      "grad_norm": 0.15567493438720703,
+      "learning_rate": 5e-05,
+      "loss": 1.5571,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6311184030626197,
+      "grad_norm": 0.17899873852729797,
+      "learning_rate": 5e-05,
+      "loss": 1.6662,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6313918512441892,
+      "grad_norm": 0.15900376439094543,
+      "learning_rate": 5e-05,
+      "loss": 1.6517,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6316652994257588,
+      "grad_norm": 0.15737038850784302,
+      "learning_rate": 5e-05,
+      "loss": 1.5635,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6319387476073284,
+      "grad_norm": 0.16273202002048492,
+      "learning_rate": 5e-05,
+      "loss": 1.5987,
+      "step": 2311
+    },
+    {
+      "epoch": 0.632212195788898,
+      "grad_norm": 0.15607643127441406,
+      "learning_rate": 5e-05,
+      "loss": 1.5978,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6324856439704676,
+      "grad_norm": 0.16705310344696045,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6327590921520372,
+      "grad_norm": 0.15748460590839386,
+      "learning_rate": 5e-05,
+      "loss": 1.5379,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6330325403336068,
+      "grad_norm": 0.16817143559455872,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6333059885151764,
+      "grad_norm": 0.15591007471084595,
+      "learning_rate": 5e-05,
+      "loss": 1.539,
+      "step": 2316
+    },
+    {
+      "epoch": 0.633579436696746,
+      "grad_norm": 0.1618664264678955,
+      "learning_rate": 5e-05,
+      "loss": 1.5866,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6338528848783156,
+      "grad_norm": 0.1688256859779358,
+      "learning_rate": 5e-05,
+      "loss": 1.7367,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6341263330598852,
+      "grad_norm": 0.17073442041873932,
+      "learning_rate": 5e-05,
+      "loss": 1.6241,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6343997812414548,
+      "grad_norm": 0.16293637454509735,
+      "learning_rate": 5e-05,
+      "loss": 1.5913,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6346732294230243,
+      "grad_norm": 0.15301881730556488,
+      "learning_rate": 5e-05,
+      "loss": 1.5685,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6349466776045939,
+      "grad_norm": 0.1594955027103424,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6352201257861635,
+      "grad_norm": 0.15445536375045776,
+      "learning_rate": 5e-05,
+      "loss": 1.5536,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6354935739677331,
+      "grad_norm": 0.166269913315773,
+      "learning_rate": 5e-05,
+      "loss": 1.6064,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6357670221493027,
+      "grad_norm": 0.15730029344558716,
+      "learning_rate": 5e-05,
+      "loss": 1.5387,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6360404703308723,
+      "grad_norm": 0.15941552817821503,
+      "learning_rate": 5e-05,
+      "loss": 1.6213,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6363139185124419,
+      "grad_norm": 0.16579844057559967,
+      "learning_rate": 5e-05,
+      "loss": 1.5993,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6365873666940115,
+      "grad_norm": 0.17182214558124542,
+      "learning_rate": 5e-05,
+      "loss": 1.6941,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6368608148755811,
+      "grad_norm": 0.1528700590133667,
+      "learning_rate": 5e-05,
+      "loss": 1.5408,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6371342630571507,
+      "grad_norm": 0.15841877460479736,
+      "learning_rate": 5e-05,
+      "loss": 1.5145,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6374077112387203,
+      "grad_norm": 0.17149239778518677,
+      "learning_rate": 5e-05,
+      "loss": 1.7896,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6376811594202898,
+      "grad_norm": 0.17149433493614197,
+      "learning_rate": 5e-05,
+      "loss": 1.6635,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6379546076018594,
+      "grad_norm": 0.1553071290254593,
+      "learning_rate": 5e-05,
+      "loss": 1.6115,
+      "step": 2333
+    },
+    {
+      "epoch": 0.638228055783429,
+      "grad_norm": 0.17740413546562195,
+      "learning_rate": 5e-05,
+      "loss": 1.6419,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6385015039649986,
+      "grad_norm": 0.16571162641048431,
+      "learning_rate": 5e-05,
+      "loss": 1.6287,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6387749521465682,
+      "grad_norm": 0.16401976346969604,
+      "learning_rate": 5e-05,
+      "loss": 1.5881,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6390484003281378,
+      "grad_norm": 0.16186074912548065,
+      "learning_rate": 5e-05,
+      "loss": 1.644,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6393218485097074,
+      "grad_norm": 0.16843372583389282,
+      "learning_rate": 5e-05,
+      "loss": 1.708,
+      "step": 2338
+    },
+    {
+      "epoch": 0.639595296691277,
+      "grad_norm": 0.1627928465604782,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6398687448728466,
+      "grad_norm": 0.15548692643642426,
+      "learning_rate": 5e-05,
+      "loss": 1.6938,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6401421930544162,
+      "grad_norm": 0.15752455592155457,
+      "learning_rate": 5e-05,
+      "loss": 1.6264,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6404156412359858,
+      "grad_norm": 0.1618221253156662,
+      "learning_rate": 5e-05,
+      "loss": 1.6214,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6406890894175554,
+      "grad_norm": 0.15189188718795776,
+      "learning_rate": 5e-05,
+      "loss": 1.6024,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6409625375991249,
+      "grad_norm": 0.16147050261497498,
+      "learning_rate": 5e-05,
+      "loss": 1.6617,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6412359857806945,
+      "grad_norm": 0.16319455206394196,
+      "learning_rate": 5e-05,
+      "loss": 1.6428,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.15911765396595,
+      "learning_rate": 5e-05,
+      "loss": 1.5411,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6417828821438337,
+      "grad_norm": 0.15844199061393738,
+      "learning_rate": 5e-05,
+      "loss": 1.5859,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6420563303254033,
+      "grad_norm": 0.16009950637817383,
+      "learning_rate": 5e-05,
+      "loss": 1.6266,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6423297785069729,
+      "grad_norm": 0.15512198209762573,
+      "learning_rate": 5e-05,
+      "loss": 1.6013,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6426032266885425,
+      "grad_norm": 0.1645142138004303,
+      "learning_rate": 5e-05,
+      "loss": 1.545,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6428766748701121,
+      "grad_norm": 0.15614928305149078,
+      "learning_rate": 5e-05,
+      "loss": 1.6131,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6431501230516817,
+      "grad_norm": 0.16520023345947266,
+      "learning_rate": 5e-05,
+      "loss": 1.6631,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6434235712332513,
+      "grad_norm": 0.164496511220932,
+      "learning_rate": 5e-05,
+      "loss": 1.6092,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6436970194148209,
+      "grad_norm": 0.16124795377254486,
+      "learning_rate": 5e-05,
+      "loss": 1.5647,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6439704675963905,
+      "grad_norm": 0.1593211442232132,
+      "learning_rate": 5e-05,
+      "loss": 1.5575,
+      "step": 2355
+    },
+    {
+      "epoch": 0.64424391577796,
+      "grad_norm": 0.15209631621837616,
+      "learning_rate": 5e-05,
+      "loss": 1.5939,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6445173639595296,
+      "grad_norm": 0.1607188880443573,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6447908121410992,
+      "grad_norm": 0.1642126739025116,
+      "learning_rate": 5e-05,
+      "loss": 1.6879,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6450642603226688,
+      "grad_norm": 0.15646396577358246,
+      "learning_rate": 5e-05,
+      "loss": 1.5971,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6453377085042384,
+      "grad_norm": 0.16889351606369019,
+      "learning_rate": 5e-05,
+      "loss": 1.6108,
+      "step": 2360
+    },
+    {
+      "epoch": 0.645611156685808,
+      "grad_norm": 0.15696920454502106,
+      "learning_rate": 5e-05,
+      "loss": 1.6325,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6458846048673776,
+      "grad_norm": 0.17429569363594055,
+      "learning_rate": 5e-05,
+      "loss": 1.6995,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6461580530489472,
+      "grad_norm": 0.16672879457473755,
+      "learning_rate": 5e-05,
+      "loss": 1.6846,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6464315012305168,
+      "grad_norm": 0.15300826728343964,
+      "learning_rate": 5e-05,
+      "loss": 1.5222,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6467049494120864,
+      "grad_norm": 0.16159702837467194,
+      "learning_rate": 5e-05,
+      "loss": 1.6988,
+      "step": 2365
+    },
+    {
+      "epoch": 0.646978397593656,
+      "grad_norm": 0.1599888950586319,
+      "learning_rate": 5e-05,
+      "loss": 1.6864,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6472518457752255,
+      "grad_norm": 0.15492287278175354,
+      "learning_rate": 5e-05,
+      "loss": 1.5455,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6475252939567951,
+      "grad_norm": 0.16030439734458923,
+      "learning_rate": 5e-05,
+      "loss": 1.6159,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6477987421383647,
+      "grad_norm": 0.16615253686904907,
+      "learning_rate": 5e-05,
+      "loss": 1.6585,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6480721903199343,
+      "grad_norm": 0.15947528183460236,
+      "learning_rate": 5e-05,
+      "loss": 1.5924,
+      "step": 2370
+    },
+    {
+      "epoch": 0.648345638501504,
+      "grad_norm": 0.1585034430027008,
+      "learning_rate": 5e-05,
+      "loss": 1.5964,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6486190866830736,
+      "grad_norm": 0.15772749483585358,
+      "learning_rate": 5e-05,
+      "loss": 1.5869,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6488925348646432,
+      "grad_norm": 0.16060665249824524,
+      "learning_rate": 5e-05,
+      "loss": 1.6111,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6491659830462128,
+      "grad_norm": 0.15704810619354248,
+      "learning_rate": 5e-05,
+      "loss": 1.6257,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6494394312277824,
+      "grad_norm": 0.16538581252098083,
+      "learning_rate": 5e-05,
+      "loss": 1.5543,
+      "step": 2375
+    },
+    {
+      "epoch": 0.649712879409352,
+      "grad_norm": 0.15830865502357483,
+      "learning_rate": 5e-05,
+      "loss": 1.6593,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6499863275909216,
+      "grad_norm": 0.17541712522506714,
+      "learning_rate": 5e-05,
+      "loss": 1.8108,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6502597757724912,
+      "grad_norm": 0.15826934576034546,
+      "learning_rate": 5e-05,
+      "loss": 1.6312,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6505332239540607,
+      "grad_norm": 0.1586860567331314,
+      "learning_rate": 5e-05,
+      "loss": 1.6282,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6508066721356303,
+      "grad_norm": 0.15930341184139252,
+      "learning_rate": 5e-05,
+      "loss": 1.6039,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6510801203171999,
+      "grad_norm": 0.16288048028945923,
+      "learning_rate": 5e-05,
+      "loss": 1.64,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6513535684987695,
+      "grad_norm": 0.15390625596046448,
+      "learning_rate": 5e-05,
+      "loss": 1.5283,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6516270166803391,
+      "grad_norm": 0.15672685205936432,
+      "learning_rate": 5e-05,
+      "loss": 1.5796,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6519004648619087,
+      "grad_norm": 0.16665740311145782,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.16864562034606934,
+      "learning_rate": 5e-05,
+      "loss": 1.5693,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6524473612250479,
+      "grad_norm": 0.17148347198963165,
+      "learning_rate": 5e-05,
+      "loss": 1.6832,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6527208094066175,
+      "grad_norm": 0.15762430429458618,
+      "learning_rate": 5e-05,
+      "loss": 1.5791,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6529942575881871,
+      "grad_norm": 0.16125012934207916,
+      "learning_rate": 5e-05,
+      "loss": 1.6339,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6532677057697567,
+      "grad_norm": 0.17223118245601654,
+      "learning_rate": 5e-05,
+      "loss": 1.6022,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6535411539513262,
+      "grad_norm": 0.16604146361351013,
+      "learning_rate": 5e-05,
+      "loss": 1.6289,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6538146021328958,
+      "grad_norm": 0.16918644309043884,
+      "learning_rate": 5e-05,
+      "loss": 1.6586,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6540880503144654,
+      "grad_norm": 0.16766510903835297,
+      "learning_rate": 5e-05,
+      "loss": 1.6329,
+      "step": 2392
+    },
+    {
+      "epoch": 0.654361498496035,
+      "grad_norm": 0.1656326800584793,
+      "learning_rate": 5e-05,
+      "loss": 1.6551,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6546349466776046,
+      "grad_norm": 0.16642306745052338,
+      "learning_rate": 5e-05,
+      "loss": 1.589,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6549083948591742,
+      "grad_norm": 0.15996260941028595,
+      "learning_rate": 5e-05,
+      "loss": 1.5731,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6551818430407438,
+      "grad_norm": 0.16654618084430695,
+      "learning_rate": 5e-05,
+      "loss": 1.7088,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6554552912223134,
+      "grad_norm": 0.16626910865306854,
+      "learning_rate": 5e-05,
+      "loss": 1.5951,
+      "step": 2397
+    },
+    {
+      "epoch": 0.655728739403883,
+      "grad_norm": 0.16408471763134003,
+      "learning_rate": 5e-05,
+      "loss": 1.5914,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6560021875854526,
+      "grad_norm": 0.15765917301177979,
+      "learning_rate": 5e-05,
+      "loss": 1.5525,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6562756357670222,
+      "grad_norm": 0.15567491948604584,
+      "learning_rate": 5e-05,
+      "loss": 1.5724,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6565490839485918,
+      "grad_norm": 0.1593087911605835,
+      "learning_rate": 5e-05,
+      "loss": 1.6831,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6568225321301613,
+      "grad_norm": 0.17593322694301605,
+      "learning_rate": 5e-05,
+      "loss": 1.6298,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6570959803117309,
+      "grad_norm": 0.16059966385364532,
+      "learning_rate": 5e-05,
+      "loss": 1.6232,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6573694284933005,
+      "grad_norm": 0.15784414112567902,
+      "learning_rate": 5e-05,
+      "loss": 1.5765,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6576428766748701,
+      "grad_norm": 0.16477881371974945,
+      "learning_rate": 5e-05,
+      "loss": 1.6309,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6579163248564397,
+      "grad_norm": 0.15968461334705353,
+      "learning_rate": 5e-05,
+      "loss": 1.5916,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6581897730380093,
+      "grad_norm": 0.15735867619514465,
+      "learning_rate": 5e-05,
+      "loss": 1.5926,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6584632212195789,
+      "grad_norm": 0.15733473002910614,
+      "learning_rate": 5e-05,
+      "loss": 1.6399,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6587366694011485,
+      "grad_norm": 0.16281574964523315,
+      "learning_rate": 5e-05,
+      "loss": 1.7522,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6590101175827181,
+      "grad_norm": 0.17003872990608215,
+      "learning_rate": 5e-05,
+      "loss": 1.6451,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6592835657642877,
+      "grad_norm": 0.15980958938598633,
+      "learning_rate": 5e-05,
+      "loss": 1.6791,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6595570139458573,
+      "grad_norm": 0.15881507098674774,
+      "learning_rate": 5e-05,
+      "loss": 1.6218,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6598304621274268,
+      "grad_norm": 0.1568790227174759,
+      "learning_rate": 5e-05,
+      "loss": 1.6652,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6601039103089964,
+      "grad_norm": 0.16632792353630066,
+      "learning_rate": 5e-05,
+      "loss": 1.6607,
+      "step": 2414
+    },
+    {
+      "epoch": 0.660377358490566,
+      "grad_norm": 0.15554708242416382,
+      "learning_rate": 5e-05,
+      "loss": 1.6423,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6606508066721356,
+      "grad_norm": 0.16607338190078735,
+      "learning_rate": 5e-05,
+      "loss": 1.5636,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6609242548537052,
+      "grad_norm": 0.1500883400440216,
+      "learning_rate": 5e-05,
+      "loss": 1.5665,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6611977030352748,
+      "grad_norm": 0.1519675999879837,
+      "learning_rate": 5e-05,
+      "loss": 1.5243,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6614711512168444,
+      "grad_norm": 0.1638784110546112,
+      "learning_rate": 5e-05,
+      "loss": 1.6462,
+      "step": 2419
+    },
+    {
+      "epoch": 0.661744599398414,
+      "grad_norm": 0.15792463719844818,
+      "learning_rate": 5e-05,
+      "loss": 1.5241,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6620180475799836,
+      "grad_norm": 0.15407374501228333,
+      "learning_rate": 5e-05,
+      "loss": 1.5822,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6622914957615532,
+      "grad_norm": 0.1588737964630127,
+      "learning_rate": 5e-05,
+      "loss": 1.7011,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6625649439431228,
+      "grad_norm": 0.15952999889850616,
+      "learning_rate": 5e-05,
+      "loss": 1.5825,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6628383921246924,
+      "grad_norm": 0.1603320688009262,
+      "learning_rate": 5e-05,
+      "loss": 1.5174,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6631118403062619,
+      "grad_norm": 0.1592378169298172,
+      "learning_rate": 5e-05,
+      "loss": 1.5717,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6633852884878315,
+      "grad_norm": 0.15770871937274933,
+      "learning_rate": 5e-05,
+      "loss": 1.603,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6636587366694011,
+      "grad_norm": 0.14396196603775024,
+      "learning_rate": 5e-05,
+      "loss": 1.4309,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6639321848509707,
+      "grad_norm": 0.15671992301940918,
+      "learning_rate": 5e-05,
+      "loss": 1.6278,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6642056330325403,
+      "grad_norm": 0.157623752951622,
+      "learning_rate": 5e-05,
+      "loss": 1.5608,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6644790812141099,
+      "grad_norm": 0.1747284233570099,
+      "learning_rate": 5e-05,
+      "loss": 1.6987,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6647525293956795,
+      "grad_norm": 0.16402825713157654,
+      "learning_rate": 5e-05,
+      "loss": 1.6901,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6650259775772491,
+      "grad_norm": 0.15663176774978638,
+      "learning_rate": 5e-05,
+      "loss": 1.5185,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6652994257588187,
+      "grad_norm": 0.14828836917877197,
+      "learning_rate": 5e-05,
+      "loss": 1.525,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6655728739403883,
+      "grad_norm": 0.15145589411258698,
+      "learning_rate": 5e-05,
+      "loss": 1.5436,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6658463221219579,
+      "grad_norm": 0.166320338845253,
+      "learning_rate": 5e-05,
+      "loss": 1.6605,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6661197703035275,
+      "grad_norm": 0.15976975858211517,
+      "learning_rate": 5e-05,
+      "loss": 1.6091,
+      "step": 2436
+    },
+    {
+      "epoch": 0.666393218485097,
+      "grad_norm": 0.1547703742980957,
+      "learning_rate": 5e-05,
+      "loss": 1.594,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.15781809389591217,
+      "learning_rate": 5e-05,
+      "loss": 1.6204,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6669401148482362,
+      "grad_norm": 0.16567228734493256,
+      "learning_rate": 5e-05,
+      "loss": 1.5822,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6672135630298058,
+      "grad_norm": 0.16364511847496033,
+      "learning_rate": 5e-05,
+      "loss": 1.606,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6674870112113754,
+      "grad_norm": 0.17084093391895294,
+      "learning_rate": 5e-05,
+      "loss": 1.6213,
+      "step": 2441
+    },
+    {
+      "epoch": 0.667760459392945,
+      "grad_norm": 0.16154657304286957,
+      "learning_rate": 5e-05,
+      "loss": 1.5344,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6680339075745146,
+      "grad_norm": 0.1677396446466446,
+      "learning_rate": 5e-05,
+      "loss": 1.5043,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6683073557560842,
+      "grad_norm": 0.16632121801376343,
+      "learning_rate": 5e-05,
+      "loss": 1.6781,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6685808039376538,
+      "grad_norm": 0.16119202971458435,
+      "learning_rate": 5e-05,
+      "loss": 1.5862,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6688542521192234,
+      "grad_norm": 0.16189579665660858,
+      "learning_rate": 5e-05,
+      "loss": 1.5532,
+      "step": 2446
+    },
+    {
+      "epoch": 0.669127700300793,
+      "grad_norm": 0.16972647607326508,
+      "learning_rate": 5e-05,
+      "loss": 1.634,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6694011484823625,
+      "grad_norm": 0.1550510823726654,
+      "learning_rate": 5e-05,
+      "loss": 1.6588,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6696745966639321,
+      "grad_norm": 0.16665950417518616,
+      "learning_rate": 5e-05,
+      "loss": 1.6175,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6699480448455017,
+      "grad_norm": 0.17640990018844604,
+      "learning_rate": 5e-05,
+      "loss": 1.5892,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6702214930270713,
+      "grad_norm": 0.1729980856180191,
+      "learning_rate": 5e-05,
+      "loss": 1.6211,
+      "step": 2451
+    },
+    {
+      "epoch": 0.670494941208641,
+      "grad_norm": 0.1840631514787674,
+      "learning_rate": 5e-05,
+      "loss": 1.6796,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6707683893902106,
+      "grad_norm": 0.16333408653736115,
+      "learning_rate": 5e-05,
+      "loss": 1.6035,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6710418375717802,
+      "grad_norm": 0.15522709488868713,
+      "learning_rate": 5e-05,
+      "loss": 1.5799,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6713152857533498,
+      "grad_norm": 0.1594085544347763,
+      "learning_rate": 5e-05,
+      "loss": 1.6033,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6715887339349194,
+      "grad_norm": 0.174288809299469,
+      "learning_rate": 5e-05,
+      "loss": 1.5725,
+      "step": 2456
+    },
+    {
+      "epoch": 0.671862182116489,
+      "grad_norm": 0.1599048525094986,
+      "learning_rate": 5e-05,
+      "loss": 1.5534,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6721356302980586,
+      "grad_norm": 0.17705102264881134,
+      "learning_rate": 5e-05,
+      "loss": 1.6193,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6724090784796282,
+      "grad_norm": 0.17394457757472992,
+      "learning_rate": 5e-05,
+      "loss": 1.5991,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6726825266611977,
+      "grad_norm": 0.17613448202610016,
+      "learning_rate": 5e-05,
+      "loss": 1.6804,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6729559748427673,
+      "grad_norm": 0.1572728008031845,
+      "learning_rate": 5e-05,
+      "loss": 1.6653,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6732294230243369,
+      "grad_norm": 0.17541466653347015,
+      "learning_rate": 5e-05,
+      "loss": 1.5699,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6735028712059065,
+      "grad_norm": 0.1598716527223587,
+      "learning_rate": 5e-05,
+      "loss": 1.5847,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6737763193874761,
+      "grad_norm": 0.16459186375141144,
+      "learning_rate": 5e-05,
+      "loss": 1.5796,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6740497675690457,
+      "grad_norm": 0.16075541079044342,
+      "learning_rate": 5e-05,
+      "loss": 1.6633,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6743232157506153,
+      "grad_norm": 0.16022174060344696,
+      "learning_rate": 5e-05,
+      "loss": 1.574,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6745966639321849,
+      "grad_norm": 0.16279369592666626,
+      "learning_rate": 5e-05,
+      "loss": 1.6139,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6748701121137545,
+      "grad_norm": 0.15538600087165833,
+      "learning_rate": 5e-05,
+      "loss": 1.6165,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6751435602953241,
+      "grad_norm": 0.15767286717891693,
+      "learning_rate": 5e-05,
+      "loss": 1.6236,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6754170084768937,
+      "grad_norm": 0.15616439282894135,
+      "learning_rate": 5e-05,
+      "loss": 1.6096,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6756904566584632,
+      "grad_norm": 0.1593167930841446,
+      "learning_rate": 5e-05,
+      "loss": 1.5664,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6759639048400328,
+      "grad_norm": 0.15188159048557281,
+      "learning_rate": 5e-05,
+      "loss": 1.6384,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6762373530216024,
+      "grad_norm": 0.1587432473897934,
+      "learning_rate": 5e-05,
+      "loss": 1.6228,
+      "step": 2473
+    },
+    {
+      "epoch": 0.676510801203172,
+      "grad_norm": 0.15247757732868195,
+      "learning_rate": 5e-05,
+      "loss": 1.4628,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6767842493847416,
+      "grad_norm": 0.1572297066450119,
+      "learning_rate": 5e-05,
+      "loss": 1.5996,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6770576975663112,
+      "grad_norm": 0.1596899777650833,
+      "learning_rate": 5e-05,
+      "loss": 1.6908,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6773311457478808,
+      "grad_norm": 0.15775376558303833,
+      "learning_rate": 5e-05,
+      "loss": 1.5248,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6776045939294504,
+      "grad_norm": 0.16116631031036377,
+      "learning_rate": 5e-05,
+      "loss": 1.5714,
+      "step": 2478
+    },
+    {
+      "epoch": 0.67787804211102,
+      "grad_norm": 0.15727153420448303,
+      "learning_rate": 5e-05,
+      "loss": 1.5822,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6781514902925896,
+      "grad_norm": 0.15772849321365356,
+      "learning_rate": 5e-05,
+      "loss": 1.64,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6784249384741592,
+      "grad_norm": 0.15587928891181946,
+      "learning_rate": 5e-05,
+      "loss": 1.5437,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6786983866557288,
+      "grad_norm": 0.16047807037830353,
+      "learning_rate": 5e-05,
+      "loss": 1.5917,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6789718348372983,
+      "grad_norm": 0.1718113124370575,
+      "learning_rate": 5e-05,
+      "loss": 1.6332,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.1675024926662445,
+      "learning_rate": 5e-05,
+      "loss": 1.6186,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6795187312004375,
+      "grad_norm": 0.16398410499095917,
+      "learning_rate": 5e-05,
+      "loss": 1.6521,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6797921793820071,
+      "grad_norm": 0.18062223494052887,
+      "learning_rate": 5e-05,
+      "loss": 1.5848,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6800656275635767,
+      "grad_norm": 0.15654748678207397,
+      "learning_rate": 5e-05,
+      "loss": 1.5452,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6803390757451463,
+      "grad_norm": 0.1599476933479309,
+      "learning_rate": 5e-05,
+      "loss": 1.6255,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6806125239267159,
+      "grad_norm": 0.16449692845344543,
+      "learning_rate": 5e-05,
+      "loss": 1.6717,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6808859721082855,
+      "grad_norm": 0.16174077987670898,
+      "learning_rate": 5e-05,
+      "loss": 1.5248,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6811594202898551,
+      "grad_norm": 0.1614287793636322,
+      "learning_rate": 5e-05,
+      "loss": 1.5408,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6814328684714247,
+      "grad_norm": 0.1738748550415039,
+      "learning_rate": 5e-05,
+      "loss": 1.6189,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6817063166529943,
+      "grad_norm": 0.1673515886068344,
+      "learning_rate": 5e-05,
+      "loss": 1.68,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6819797648345638,
+      "grad_norm": 0.16801506280899048,
+      "learning_rate": 5e-05,
+      "loss": 1.6737,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6822532130161334,
+      "grad_norm": 0.1836167722940445,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 2495
+    },
+    {
+      "epoch": 0.682526661197703,
+      "grad_norm": 0.1564912497997284,
+      "learning_rate": 5e-05,
+      "loss": 1.5649,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6828001093792726,
+      "grad_norm": 0.161629781126976,
+      "learning_rate": 5e-05,
+      "loss": 1.6278,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6830735575608422,
+      "grad_norm": 0.18047556281089783,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6833470057424118,
+      "grad_norm": 0.1663326621055603,
+      "learning_rate": 5e-05,
+      "loss": 1.6222,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6836204539239814,
+      "grad_norm": 0.16856786608695984,
+      "learning_rate": 5e-05,
+      "loss": 1.6573,
+      "step": 2500
+    },
+    {
+      "epoch": 0.683893902105551,
+      "grad_norm": 0.17234691977500916,
+      "learning_rate": 5e-05,
+      "loss": 1.6514,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6841673502871206,
+      "grad_norm": 0.15840747952461243,
+      "learning_rate": 5e-05,
+      "loss": 1.5428,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6844407984686902,
+      "grad_norm": 0.17479360103607178,
+      "learning_rate": 5e-05,
+      "loss": 1.6271,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6847142466502598,
+      "grad_norm": 0.16867142915725708,
+      "learning_rate": 5e-05,
+      "loss": 1.5434,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6849876948318294,
+      "grad_norm": 0.16198626160621643,
+      "learning_rate": 5e-05,
+      "loss": 1.6075,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6852611430133989,
+      "grad_norm": 0.16754168272018433,
+      "learning_rate": 5e-05,
+      "loss": 1.7111,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6855345911949685,
+      "grad_norm": 0.1667136549949646,
+      "learning_rate": 5e-05,
+      "loss": 1.582,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6858080393765381,
+      "grad_norm": 0.15942475199699402,
+      "learning_rate": 5e-05,
+      "loss": 1.6081,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6860814875581077,
+      "grad_norm": 0.1539759784936905,
+      "learning_rate": 5e-05,
+      "loss": 1.5774,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6863549357396773,
+      "grad_norm": 0.1718335896730423,
+      "learning_rate": 5e-05,
+      "loss": 1.7125,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6866283839212469,
+      "grad_norm": 0.16027317941188812,
+      "learning_rate": 5e-05,
+      "loss": 1.634,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6869018321028165,
+      "grad_norm": 0.1722375601530075,
+      "learning_rate": 5e-05,
+      "loss": 1.6105,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6871752802843861,
+      "grad_norm": 0.15657594799995422,
+      "learning_rate": 5e-05,
+      "loss": 1.5136,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6874487284659557,
+      "grad_norm": 0.16483189165592194,
+      "learning_rate": 5e-05,
+      "loss": 1.629,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6877221766475253,
+      "grad_norm": 0.1574966162443161,
+      "learning_rate": 5e-05,
+      "loss": 1.5831,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6879956248290949,
+      "grad_norm": 0.16608235239982605,
+      "learning_rate": 5e-05,
+      "loss": 1.6769,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6882690730106645,
+      "grad_norm": 0.1636161059141159,
+      "learning_rate": 5e-05,
+      "loss": 1.6551,
+      "step": 2517
+    },
+    {
+      "epoch": 0.688542521192234,
+      "grad_norm": 0.15895555913448334,
+      "learning_rate": 5e-05,
+      "loss": 1.5672,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6888159693738036,
+      "grad_norm": 0.17441825568675995,
+      "learning_rate": 5e-05,
+      "loss": 1.6559,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6890894175553732,
+      "grad_norm": 0.15818622708320618,
+      "learning_rate": 5e-05,
+      "loss": 1.6146,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6893628657369428,
+      "grad_norm": 0.16466504335403442,
+      "learning_rate": 5e-05,
+      "loss": 1.6309,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6896363139185124,
+      "grad_norm": 0.1637655645608902,
+      "learning_rate": 5e-05,
+      "loss": 1.6369,
+      "step": 2522
+    },
+    {
+      "epoch": 0.689909762100082,
+      "grad_norm": 0.17039266228675842,
+      "learning_rate": 5e-05,
+      "loss": 1.6315,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6901832102816516,
+      "grad_norm": 0.1624341756105423,
+      "learning_rate": 5e-05,
+      "loss": 1.5878,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6904566584632212,
+      "grad_norm": 0.17213301360607147,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6907301066447908,
+      "grad_norm": 0.17695990204811096,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6910035548263604,
+      "grad_norm": 0.1731920838356018,
+      "learning_rate": 5e-05,
+      "loss": 1.7317,
+      "step": 2527
+    },
+    {
+      "epoch": 0.69127700300793,
+      "grad_norm": 0.16777953505516052,
+      "learning_rate": 5e-05,
+      "loss": 1.608,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6915504511894995,
+      "grad_norm": 0.16122910380363464,
+      "learning_rate": 5e-05,
+      "loss": 1.5529,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6918238993710691,
+      "grad_norm": 0.16527071595191956,
+      "learning_rate": 5e-05,
+      "loss": 1.621,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6920973475526387,
+      "grad_norm": 0.1596560925245285,
+      "learning_rate": 5e-05,
+      "loss": 1.6079,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6923707957342083,
+      "grad_norm": 0.15807481110095978,
+      "learning_rate": 5e-05,
+      "loss": 1.5349,
+      "step": 2532
+    },
+    {
+      "epoch": 0.692644243915778,
+      "grad_norm": 0.16306331753730774,
+      "learning_rate": 5e-05,
+      "loss": 1.6461,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6929176920973475,
+      "grad_norm": 0.16564686596393585,
+      "learning_rate": 5e-05,
+      "loss": 1.5821,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6931911402789172,
+      "grad_norm": 0.1704019457101822,
+      "learning_rate": 5e-05,
+      "loss": 1.5851,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6934645884604868,
+      "grad_norm": 0.15806765854358673,
+      "learning_rate": 5e-05,
+      "loss": 1.6363,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6937380366420564,
+      "grad_norm": 0.15724487602710724,
+      "learning_rate": 5e-05,
+      "loss": 1.608,
+      "step": 2537
+    },
+    {
+      "epoch": 0.694011484823626,
+      "grad_norm": 0.16402623057365417,
+      "learning_rate": 5e-05,
+      "loss": 1.5813,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6942849330051956,
+      "grad_norm": 0.15913555026054382,
+      "learning_rate": 5e-05,
+      "loss": 1.6284,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6945583811867652,
+      "grad_norm": 0.16074508428573608,
+      "learning_rate": 5e-05,
+      "loss": 1.528,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6948318293683347,
+      "grad_norm": 0.15752652287483215,
+      "learning_rate": 5e-05,
+      "loss": 1.5023,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6951052775499043,
+      "grad_norm": 0.15707096457481384,
+      "learning_rate": 5e-05,
+      "loss": 1.6251,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6953787257314739,
+      "grad_norm": 0.15630987286567688,
+      "learning_rate": 5e-05,
+      "loss": 1.5003,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.15872447192668915,
+      "learning_rate": 5e-05,
+      "loss": 1.5698,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6959256220946131,
+      "grad_norm": 0.16033059358596802,
+      "learning_rate": 5e-05,
+      "loss": 1.5535,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6961990702761827,
+      "grad_norm": 0.1571992188692093,
+      "learning_rate": 5e-05,
+      "loss": 1.573,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6964725184577523,
+      "grad_norm": 0.15957675874233246,
+      "learning_rate": 5e-05,
+      "loss": 1.6449,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6967459666393219,
+      "grad_norm": 0.17184406518936157,
+      "learning_rate": 5e-05,
+      "loss": 1.7537,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6970194148208915,
+      "grad_norm": 0.1593320220708847,
+      "learning_rate": 5e-05,
+      "loss": 1.6327,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6972928630024611,
+      "grad_norm": 0.16837316751480103,
+      "learning_rate": 5e-05,
+      "loss": 1.6125,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6975663111840307,
+      "grad_norm": 0.1569320261478424,
+      "learning_rate": 5e-05,
+      "loss": 1.6095,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6978397593656002,
+      "grad_norm": 0.1526784896850586,
+      "learning_rate": 5e-05,
+      "loss": 1.5959,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6981132075471698,
+      "grad_norm": 0.16312266886234283,
+      "learning_rate": 5e-05,
+      "loss": 1.6402,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6983866557287394,
+      "grad_norm": 0.16702058911323547,
+      "learning_rate": 5e-05,
+      "loss": 1.5529,
+      "step": 2554
+    },
+    {
+      "epoch": 0.698660103910309,
+      "grad_norm": 0.16597692668437958,
+      "learning_rate": 5e-05,
+      "loss": 1.6619,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6989335520918786,
+      "grad_norm": 0.16194956004619598,
+      "learning_rate": 5e-05,
+      "loss": 1.7461,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6992070002734482,
+      "grad_norm": 0.15624722838401794,
+      "learning_rate": 5e-05,
+      "loss": 1.6271,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6994804484550178,
+      "grad_norm": 0.15689989924430847,
+      "learning_rate": 5e-05,
+      "loss": 1.5491,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6997538966365874,
+      "grad_norm": 0.1616222858428955,
+      "learning_rate": 5e-05,
+      "loss": 1.5906,
+      "step": 2559
+    },
+    {
+      "epoch": 0.700027344818157,
+      "grad_norm": 0.15930163860321045,
+      "learning_rate": 5e-05,
+      "loss": 1.5583,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7003007929997266,
+      "grad_norm": 0.16471946239471436,
+      "learning_rate": 5e-05,
+      "loss": 1.6327,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7005742411812962,
+      "grad_norm": 0.16374364495277405,
+      "learning_rate": 5e-05,
+      "loss": 1.6228,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7008476893628658,
+      "grad_norm": 0.1631327122449875,
+      "learning_rate": 5e-05,
+      "loss": 1.6547,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7011211375444353,
+      "grad_norm": 0.15994025766849518,
+      "learning_rate": 5e-05,
+      "loss": 1.5833,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7013945857260049,
+      "grad_norm": 0.16121256351470947,
+      "learning_rate": 5e-05,
+      "loss": 1.6305,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7016680339075745,
+      "grad_norm": 0.16354970633983612,
+      "learning_rate": 5e-05,
+      "loss": 1.5552,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7019414820891441,
+      "grad_norm": 0.161824032664299,
+      "learning_rate": 5e-05,
+      "loss": 1.6749,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7022149302707137,
+      "grad_norm": 0.17040131986141205,
+      "learning_rate": 5e-05,
+      "loss": 1.6291,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7024883784522833,
+      "grad_norm": 0.1799769103527069,
+      "learning_rate": 5e-05,
+      "loss": 1.657,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7027618266338529,
+      "grad_norm": 0.15718813240528107,
+      "learning_rate": 5e-05,
+      "loss": 1.5858,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7030352748154225,
+      "grad_norm": 0.16162623465061188,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7033087229969921,
+      "grad_norm": 0.16257646679878235,
+      "learning_rate": 5e-05,
+      "loss": 1.624,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7035821711785617,
+      "grad_norm": 0.16675184667110443,
+      "learning_rate": 5e-05,
+      "loss": 1.6738,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7038556193601313,
+      "grad_norm": 0.17329177260398865,
+      "learning_rate": 5e-05,
+      "loss": 1.6952,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7041290675417008,
+      "grad_norm": 0.15674902498722076,
+      "learning_rate": 5e-05,
+      "loss": 1.5858,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7044025157232704,
+      "grad_norm": 0.15690606832504272,
+      "learning_rate": 5e-05,
+      "loss": 1.5772,
+      "step": 2576
+    },
+    {
+      "epoch": 0.70467596390484,
+      "grad_norm": 0.16210415959358215,
+      "learning_rate": 5e-05,
+      "loss": 1.6705,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7049494120864096,
+      "grad_norm": 0.16525010764598846,
+      "learning_rate": 5e-05,
+      "loss": 1.6188,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7052228602679792,
+      "grad_norm": 0.1738116294145584,
+      "learning_rate": 5e-05,
+      "loss": 1.5788,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7054963084495488,
+      "grad_norm": 0.1632319688796997,
+      "learning_rate": 5e-05,
+      "loss": 1.6327,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7057697566311184,
+      "grad_norm": 0.16341067850589752,
+      "learning_rate": 5e-05,
+      "loss": 1.6567,
+      "step": 2581
+    },
+    {
+      "epoch": 0.706043204812688,
+      "grad_norm": 0.17945139110088348,
+      "learning_rate": 5e-05,
+      "loss": 1.6919,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7063166529942576,
+      "grad_norm": 0.15467429161071777,
+      "learning_rate": 5e-05,
+      "loss": 1.618,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7065901011758272,
+      "grad_norm": 0.16300177574157715,
+      "learning_rate": 5e-05,
+      "loss": 1.6526,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7068635493573968,
+      "grad_norm": 0.15929557383060455,
+      "learning_rate": 5e-05,
+      "loss": 1.5512,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7071369975389664,
+      "grad_norm": 0.16156993806362152,
+      "learning_rate": 5e-05,
+      "loss": 1.6352,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7074104457205359,
+      "grad_norm": 0.15560369193553925,
+      "learning_rate": 5e-05,
+      "loss": 1.5727,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7076838939021055,
+      "grad_norm": 0.16337203979492188,
+      "learning_rate": 5e-05,
+      "loss": 1.561,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7079573420836751,
+      "grad_norm": 0.1646268218755722,
+      "learning_rate": 5e-05,
+      "loss": 1.5965,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7082307902652447,
+      "grad_norm": 0.16066445410251617,
+      "learning_rate": 5e-05,
+      "loss": 1.6132,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7085042384468143,
+      "grad_norm": 0.14791713654994965,
+      "learning_rate": 5e-05,
+      "loss": 1.5,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7087776866283839,
+      "grad_norm": 0.15416625142097473,
+      "learning_rate": 5e-05,
+      "loss": 1.5718,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7090511348099535,
+      "grad_norm": 0.1604636311531067,
+      "learning_rate": 5e-05,
+      "loss": 1.6064,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7093245829915231,
+      "grad_norm": 0.15781953930854797,
+      "learning_rate": 5e-05,
+      "loss": 1.5769,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7095980311730927,
+      "grad_norm": 0.15604273974895477,
+      "learning_rate": 5e-05,
+      "loss": 1.588,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7098714793546623,
+      "grad_norm": 0.15518909692764282,
+      "learning_rate": 5e-05,
+      "loss": 1.59,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7101449275362319,
+      "grad_norm": 0.16104191541671753,
+      "learning_rate": 5e-05,
+      "loss": 1.6747,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7104183757178015,
+      "grad_norm": 0.16382278501987457,
+      "learning_rate": 5e-05,
+      "loss": 1.6737,
+      "step": 2598
+    },
+    {
+      "epoch": 0.710691823899371,
+      "grad_norm": 0.16717152297496796,
+      "learning_rate": 5e-05,
+      "loss": 1.6239,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7109652720809406,
+      "grad_norm": 0.16594311594963074,
+      "learning_rate": 5e-05,
+      "loss": 1.7318,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7112387202625102,
+      "grad_norm": 0.16552476584911346,
+      "learning_rate": 5e-05,
+      "loss": 1.7376,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7115121684440798,
+      "grad_norm": 0.17463426291942596,
+      "learning_rate": 5e-05,
+      "loss": 1.6286,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7117856166256494,
+      "grad_norm": 0.159657821059227,
+      "learning_rate": 5e-05,
+      "loss": 1.5664,
+      "step": 2603
+    },
+    {
+      "epoch": 0.712059064807219,
+      "grad_norm": 0.16044557094573975,
+      "learning_rate": 5e-05,
+      "loss": 1.5265,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7123325129887886,
+      "grad_norm": 0.15536926686763763,
+      "learning_rate": 5e-05,
+      "loss": 1.5381,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7126059611703582,
+      "grad_norm": 0.1617160588502884,
+      "learning_rate": 5e-05,
+      "loss": 1.6025,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7128794093519278,
+      "grad_norm": 0.16367921233177185,
+      "learning_rate": 5e-05,
+      "loss": 1.7194,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7131528575334974,
+      "grad_norm": 0.16112937033176422,
+      "learning_rate": 5e-05,
+      "loss": 1.4914,
+      "step": 2608
+    },
+    {
+      "epoch": 0.713426305715067,
+      "grad_norm": 0.16444730758666992,
+      "learning_rate": 5e-05,
+      "loss": 1.625,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7136997538966365,
+      "grad_norm": 0.16794142127037048,
+      "learning_rate": 5e-05,
+      "loss": 1.6348,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7139732020782061,
+      "grad_norm": 0.1628817617893219,
+      "learning_rate": 5e-05,
+      "loss": 1.6083,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7142466502597757,
+      "grad_norm": 0.18700134754180908,
+      "learning_rate": 5e-05,
+      "loss": 1.6996,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7145200984413453,
+      "grad_norm": 0.15966500341892242,
+      "learning_rate": 5e-05,
+      "loss": 1.6113,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7147935466229149,
+      "grad_norm": 0.18392211198806763,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7150669948044845,
+      "grad_norm": 0.17130230367183685,
+      "learning_rate": 5e-05,
+      "loss": 1.6511,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7153404429860541,
+      "grad_norm": 0.16397760808467865,
+      "learning_rate": 5e-05,
+      "loss": 1.6283,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7156138911676238,
+      "grad_norm": 0.16686753928661346,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7158873393491934,
+      "grad_norm": 0.1645701676607132,
+      "learning_rate": 5e-05,
+      "loss": 1.6268,
+      "step": 2618
+    },
+    {
+      "epoch": 0.716160787530763,
+      "grad_norm": 0.1619407832622528,
+      "learning_rate": 5e-05,
+      "loss": 1.6309,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7164342357123326,
+      "grad_norm": 0.17755883932113647,
+      "learning_rate": 5e-05,
+      "loss": 1.6246,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7167076838939022,
+      "grad_norm": 0.17687168717384338,
+      "learning_rate": 5e-05,
+      "loss": 1.6039,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.16282087564468384,
+      "learning_rate": 5e-05,
+      "loss": 1.5709,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7172545802570413,
+      "grad_norm": 0.1705690324306488,
+      "learning_rate": 5e-05,
+      "loss": 1.4957,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7175280284386109,
+      "grad_norm": 0.17099738121032715,
+      "learning_rate": 5e-05,
+      "loss": 1.5331,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7178014766201805,
+      "grad_norm": 0.1703016757965088,
+      "learning_rate": 5e-05,
+      "loss": 1.6962,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7180749248017501,
+      "grad_norm": 0.1631249189376831,
+      "learning_rate": 5e-05,
+      "loss": 1.5764,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7183483729833197,
+      "grad_norm": 0.1670769453048706,
+      "learning_rate": 5e-05,
+      "loss": 1.5328,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7186218211648893,
+      "grad_norm": 0.15617656707763672,
+      "learning_rate": 5e-05,
+      "loss": 1.5934,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7188952693464589,
+      "grad_norm": 0.15787526965141296,
+      "learning_rate": 5e-05,
+      "loss": 1.5847,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7191687175280285,
+      "grad_norm": 0.160158172249794,
+      "learning_rate": 5e-05,
+      "loss": 1.5974,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7194421657095981,
+      "grad_norm": 0.16433510184288025,
+      "learning_rate": 5e-05,
+      "loss": 1.573,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7197156138911677,
+      "grad_norm": 0.15415172278881073,
+      "learning_rate": 5e-05,
+      "loss": 1.5221,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7199890620727372,
+      "grad_norm": 0.16252250969409943,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7202625102543068,
+      "grad_norm": 0.1922326236963272,
+      "learning_rate": 5e-05,
+      "loss": 1.5648,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7205359584358764,
+      "grad_norm": 0.16628074645996094,
+      "learning_rate": 5e-05,
+      "loss": 1.5623,
+      "step": 2635
+    },
+    {
+      "epoch": 0.720809406617446,
+      "grad_norm": 0.19099196791648865,
+      "learning_rate": 5e-05,
+      "loss": 1.6423,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7210828547990156,
+      "grad_norm": 0.20283576846122742,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7213563029805852,
+      "grad_norm": 0.15855461359024048,
+      "learning_rate": 5e-05,
+      "loss": 1.6438,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7216297511621548,
+      "grad_norm": 0.21607458591461182,
+      "learning_rate": 5e-05,
+      "loss": 1.6484,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7219031993437244,
+      "grad_norm": 0.1732168048620224,
+      "learning_rate": 5e-05,
+      "loss": 1.5945,
+      "step": 2640
+    },
+    {
+      "epoch": 0.722176647525294,
+      "grad_norm": 0.18066106736660004,
+      "learning_rate": 5e-05,
+      "loss": 1.6576,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7224500957068636,
+      "grad_norm": 0.1668356955051422,
+      "learning_rate": 5e-05,
+      "loss": 1.6666,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7227235438884332,
+      "grad_norm": 0.17484630644321442,
+      "learning_rate": 5e-05,
+      "loss": 1.5694,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7229969920700028,
+      "grad_norm": 0.16421037912368774,
+      "learning_rate": 5e-05,
+      "loss": 1.5838,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7232704402515723,
+      "grad_norm": 0.16860564053058624,
+      "learning_rate": 5e-05,
+      "loss": 1.6144,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7235438884331419,
+      "grad_norm": 0.15611010789871216,
+      "learning_rate": 5e-05,
+      "loss": 1.5933,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7238173366147115,
+      "grad_norm": 0.1699523776769638,
+      "learning_rate": 5e-05,
+      "loss": 1.6102,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7240907847962811,
+      "grad_norm": 0.15928548574447632,
+      "learning_rate": 5e-05,
+      "loss": 1.5433,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7243642329778507,
+      "grad_norm": 0.15467146039009094,
+      "learning_rate": 5e-05,
+      "loss": 1.5628,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7246376811594203,
+      "grad_norm": 0.1566615253686905,
+      "learning_rate": 5e-05,
+      "loss": 1.6284,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7249111293409899,
+      "grad_norm": 0.1829017996788025,
+      "learning_rate": 5e-05,
+      "loss": 1.5719,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7251845775225595,
+      "grad_norm": 0.20479872822761536,
+      "learning_rate": 5e-05,
+      "loss": 1.483,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7254580257041291,
+      "grad_norm": 0.16582070291042328,
+      "learning_rate": 5e-05,
+      "loss": 1.6091,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7257314738856987,
+      "grad_norm": 0.16045349836349487,
+      "learning_rate": 5e-05,
+      "loss": 1.603,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7260049220672683,
+      "grad_norm": 0.21731440722942352,
+      "learning_rate": 5e-05,
+      "loss": 1.5638,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7262783702488378,
+      "grad_norm": 0.16208265721797943,
+      "learning_rate": 5e-05,
+      "loss": 1.5898,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7265518184304074,
+      "grad_norm": 0.1588432937860489,
+      "learning_rate": 5e-05,
+      "loss": 1.5425,
+      "step": 2657
+    },
+    {
+      "epoch": 0.726825266611977,
+      "grad_norm": 0.1615055352449417,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7270987147935466,
+      "grad_norm": 0.14946486055850983,
+      "learning_rate": 5e-05,
+      "loss": 1.5624,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7273721629751162,
+      "grad_norm": 0.15998360514640808,
+      "learning_rate": 5e-05,
+      "loss": 1.7129,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7276456111566858,
+      "grad_norm": 0.15510539710521698,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7279190593382554,
+      "grad_norm": 0.15238802134990692,
+      "learning_rate": 5e-05,
+      "loss": 1.5503,
+      "step": 2662
+    },
+    {
+      "epoch": 0.728192507519825,
+      "grad_norm": 0.16379080712795258,
+      "learning_rate": 5e-05,
+      "loss": 1.5657,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7284659557013946,
+      "grad_norm": 0.1604292243719101,
+      "learning_rate": 5e-05,
+      "loss": 1.5491,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7287394038829642,
+      "grad_norm": 0.16709019243717194,
+      "learning_rate": 5e-05,
+      "loss": 1.5678,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7290128520645338,
+      "grad_norm": 0.15374033153057098,
+      "learning_rate": 5e-05,
+      "loss": 1.5231,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7292863002461034,
+      "grad_norm": 0.165635883808136,
+      "learning_rate": 5e-05,
+      "loss": 1.6576,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7295597484276729,
+      "grad_norm": 0.15312857925891876,
+      "learning_rate": 5e-05,
+      "loss": 1.5964,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7298331966092425,
+      "grad_norm": 0.16065338253974915,
+      "learning_rate": 5e-05,
+      "loss": 1.6054,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7301066447908121,
+      "grad_norm": 0.16220992803573608,
+      "learning_rate": 5e-05,
+      "loss": 1.6551,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7303800929723817,
+      "grad_norm": 0.16524186730384827,
+      "learning_rate": 5e-05,
+      "loss": 1.5719,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7306535411539513,
+      "grad_norm": 0.1688978374004364,
+      "learning_rate": 5e-05,
+      "loss": 1.5573,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7309269893355209,
+      "grad_norm": 0.16838806867599487,
+      "learning_rate": 5e-05,
+      "loss": 1.6978,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7312004375170905,
+      "grad_norm": 0.15754219889640808,
+      "learning_rate": 5e-05,
+      "loss": 1.5535,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7314738856986601,
+      "grad_norm": 0.17631256580352783,
+      "learning_rate": 5e-05,
+      "loss": 1.6562,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7317473338802297,
+      "grad_norm": 0.17020723223686218,
+      "learning_rate": 5e-05,
+      "loss": 1.6649,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7320207820617993,
+      "grad_norm": 0.17932482063770294,
+      "learning_rate": 5e-05,
+      "loss": 1.6102,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7322942302433689,
+      "grad_norm": 0.1644601821899414,
+      "learning_rate": 5e-05,
+      "loss": 1.5042,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7325676784249385,
+      "grad_norm": 0.18472503125667572,
+      "learning_rate": 5e-05,
+      "loss": 1.5805,
+      "step": 2679
+    },
+    {
+      "epoch": 0.732841126606508,
+      "grad_norm": 0.16369123756885529,
+      "learning_rate": 5e-05,
+      "loss": 1.6079,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7331145747880776,
+      "grad_norm": 0.16636909544467926,
+      "learning_rate": 5e-05,
+      "loss": 1.5607,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7333880229696472,
+      "grad_norm": 0.16647064685821533,
+      "learning_rate": 5e-05,
+      "loss": 1.5819,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7336614711512168,
+      "grad_norm": 0.17708119750022888,
+      "learning_rate": 5e-05,
+      "loss": 1.6366,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7339349193327864,
+      "grad_norm": 0.15985938906669617,
+      "learning_rate": 5e-05,
+      "loss": 1.6187,
+      "step": 2684
+    },
+    {
+      "epoch": 0.734208367514356,
+      "grad_norm": 0.1802692860364914,
+      "learning_rate": 5e-05,
+      "loss": 1.5459,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7344818156959256,
+      "grad_norm": 0.156411275267601,
+      "learning_rate": 5e-05,
+      "loss": 1.5549,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7347552638774952,
+      "grad_norm": 0.1642201840877533,
+      "learning_rate": 5e-05,
+      "loss": 1.6463,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7350287120590648,
+      "grad_norm": 0.16783371567726135,
+      "learning_rate": 5e-05,
+      "loss": 1.5811,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7353021602406344,
+      "grad_norm": 0.15901906788349152,
+      "learning_rate": 5e-05,
+      "loss": 1.5787,
+      "step": 2689
+    },
+    {
+      "epoch": 0.735575608422204,
+      "grad_norm": 0.1638929694890976,
+      "learning_rate": 5e-05,
+      "loss": 1.6526,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7358490566037735,
+      "grad_norm": 0.15809820592403412,
+      "learning_rate": 5e-05,
+      "loss": 1.6082,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7361225047853431,
+      "grad_norm": 0.1644376665353775,
+      "learning_rate": 5e-05,
+      "loss": 1.5305,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7363959529669127,
+      "grad_norm": 0.16332222521305084,
+      "learning_rate": 5e-05,
+      "loss": 1.6534,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7366694011484823,
+      "grad_norm": 0.18329092860221863,
+      "learning_rate": 5e-05,
+      "loss": 1.603,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7369428493300519,
+      "grad_norm": 0.1716667115688324,
+      "learning_rate": 5e-05,
+      "loss": 1.7023,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7372162975116215,
+      "grad_norm": 0.16460417211055756,
+      "learning_rate": 5e-05,
+      "loss": 1.6165,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7374897456931911,
+      "grad_norm": 0.16401362419128418,
+      "learning_rate": 5e-05,
+      "loss": 1.5556,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7377631938747607,
+      "grad_norm": 0.15876410901546478,
+      "learning_rate": 5e-05,
+      "loss": 1.5701,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7380366420563303,
+      "grad_norm": 0.18006089329719543,
+      "learning_rate": 5e-05,
+      "loss": 1.6061,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7383100902379,
+      "grad_norm": 0.16289302706718445,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7385835384194696,
+      "grad_norm": 0.15855976939201355,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7388569866010392,
+      "grad_norm": 0.16138465702533722,
+      "learning_rate": 5e-05,
+      "loss": 1.6222,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 0.1679658591747284,
+      "learning_rate": 5e-05,
+      "loss": 1.6472,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7394038829641782,
+      "grad_norm": 0.16092632710933685,
+      "learning_rate": 5e-05,
+      "loss": 1.6278,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7396773311457479,
+      "grad_norm": 0.17156141996383667,
+      "learning_rate": 5e-05,
+      "loss": 1.6397,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7399507793273175,
+      "grad_norm": 0.17136305570602417,
+      "learning_rate": 5e-05,
+      "loss": 1.6552,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7402242275088871,
+      "grad_norm": 0.16168268024921417,
+      "learning_rate": 5e-05,
+      "loss": 1.5059,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7404976756904567,
+      "grad_norm": 0.16185155510902405,
+      "learning_rate": 5e-05,
+      "loss": 1.5465,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7407711238720263,
+      "grad_norm": 0.15480349957942963,
+      "learning_rate": 5e-05,
+      "loss": 1.5585,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7410445720535959,
+      "grad_norm": 0.16501635313034058,
+      "learning_rate": 5e-05,
+      "loss": 1.6614,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7413180202351655,
+      "grad_norm": 0.15960432589054108,
+      "learning_rate": 5e-05,
+      "loss": 1.5774,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7415914684167351,
+      "grad_norm": 0.16600382328033447,
+      "learning_rate": 5e-05,
+      "loss": 1.6463,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7418649165983047,
+      "grad_norm": 0.16479994356632233,
+      "learning_rate": 5e-05,
+      "loss": 1.6348,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7421383647798742,
+      "grad_norm": 0.15985356271266937,
+      "learning_rate": 5e-05,
+      "loss": 1.5811,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7424118129614438,
+      "grad_norm": 0.16183726489543915,
+      "learning_rate": 5e-05,
+      "loss": 1.6204,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7426852611430134,
+      "grad_norm": 0.1608124077320099,
+      "learning_rate": 5e-05,
+      "loss": 1.5919,
+      "step": 2716
+    },
+    {
+      "epoch": 0.742958709324583,
+      "grad_norm": 0.16460396349430084,
+      "learning_rate": 5e-05,
+      "loss": 1.5784,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7432321575061526,
+      "grad_norm": 0.1570330709218979,
+      "learning_rate": 5e-05,
+      "loss": 1.577,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7435056056877222,
+      "grad_norm": 0.17111632227897644,
+      "learning_rate": 5e-05,
+      "loss": 1.5384,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7437790538692918,
+      "grad_norm": 0.1649581640958786,
+      "learning_rate": 5e-05,
+      "loss": 1.5283,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7440525020508614,
+      "grad_norm": 0.16623181104660034,
+      "learning_rate": 5e-05,
+      "loss": 1.6578,
+      "step": 2721
+    },
+    {
+      "epoch": 0.744325950232431,
+      "grad_norm": 0.16478942334651947,
+      "learning_rate": 5e-05,
+      "loss": 1.6588,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7445993984140006,
+      "grad_norm": 0.16344524919986725,
+      "learning_rate": 5e-05,
+      "loss": 1.6124,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7448728465955702,
+      "grad_norm": 0.1609012335538864,
+      "learning_rate": 5e-05,
+      "loss": 1.6005,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7451462947771398,
+      "grad_norm": 0.17075218260288239,
+      "learning_rate": 5e-05,
+      "loss": 1.7217,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7454197429587093,
+      "grad_norm": 0.16447240114212036,
+      "learning_rate": 5e-05,
+      "loss": 1.6142,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7456931911402789,
+      "grad_norm": 0.15842659771442413,
+      "learning_rate": 5e-05,
+      "loss": 1.566,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7459666393218485,
+      "grad_norm": 0.1670447736978531,
+      "learning_rate": 5e-05,
+      "loss": 1.6299,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7462400875034181,
+      "grad_norm": 0.1572618931531906,
+      "learning_rate": 5e-05,
+      "loss": 1.5555,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7465135356849877,
+      "grad_norm": 0.16439321637153625,
+      "learning_rate": 5e-05,
+      "loss": 1.6149,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7467869838665573,
+      "grad_norm": 0.15821623802185059,
+      "learning_rate": 5e-05,
+      "loss": 1.6452,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7470604320481269,
+      "grad_norm": 0.16016224026679993,
+      "learning_rate": 5e-05,
+      "loss": 1.5371,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7473338802296965,
+      "grad_norm": 0.1666966676712036,
+      "learning_rate": 5e-05,
+      "loss": 1.5912,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7476073284112661,
+      "grad_norm": 0.1630515605211258,
+      "learning_rate": 5e-05,
+      "loss": 1.5903,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7478807765928357,
+      "grad_norm": 0.1738271415233612,
+      "learning_rate": 5e-05,
+      "loss": 1.6498,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7481542247744053,
+      "grad_norm": 0.16823971271514893,
+      "learning_rate": 5e-05,
+      "loss": 1.6878,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7484276729559748,
+      "grad_norm": 0.16174787282943726,
+      "learning_rate": 5e-05,
+      "loss": 1.5538,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7487011211375444,
+      "grad_norm": 0.15992052853107452,
+      "learning_rate": 5e-05,
+      "loss": 1.5307,
+      "step": 2738
+    },
+    {
+      "epoch": 0.748974569319114,
+      "grad_norm": 0.17191752791404724,
+      "learning_rate": 5e-05,
+      "loss": 1.6057,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7492480175006836,
+      "grad_norm": 0.16447550058364868,
+      "learning_rate": 5e-05,
+      "loss": 1.6311,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7495214656822532,
+      "grad_norm": 0.16319876909255981,
+      "learning_rate": 5e-05,
+      "loss": 1.5762,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7497949138638228,
+      "grad_norm": 0.16644878685474396,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7500683620453924,
+      "grad_norm": 0.16339637339115143,
+      "learning_rate": 5e-05,
+      "loss": 1.6049,
+      "step": 2743
+    },
+    {
+      "epoch": 0.750341810226962,
+      "grad_norm": 0.1628854125738144,
+      "learning_rate": 5e-05,
+      "loss": 1.5755,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7506152584085316,
+      "grad_norm": 0.15712201595306396,
+      "learning_rate": 5e-05,
+      "loss": 1.5888,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7508887065901012,
+      "grad_norm": 0.15863806009292603,
+      "learning_rate": 5e-05,
+      "loss": 1.5662,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7511621547716708,
+      "grad_norm": 0.16975349187850952,
+      "learning_rate": 5e-05,
+      "loss": 1.6032,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7514356029532404,
+      "grad_norm": 0.15435737371444702,
+      "learning_rate": 5e-05,
+      "loss": 1.5288,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7517090511348099,
+      "grad_norm": 0.16747300326824188,
+      "learning_rate": 5e-05,
+      "loss": 1.59,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7519824993163795,
+      "grad_norm": 0.17562800645828247,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7522559474979491,
+      "grad_norm": 0.1631203293800354,
+      "learning_rate": 5e-05,
+      "loss": 1.6197,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7525293956795187,
+      "grad_norm": 0.17813941836357117,
+      "learning_rate": 5e-05,
+      "loss": 1.607,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7528028438610883,
+      "grad_norm": 0.15916167199611664,
+      "learning_rate": 5e-05,
+      "loss": 1.6042,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7530762920426579,
+      "grad_norm": 0.18315176665782928,
+      "learning_rate": 5e-05,
+      "loss": 1.6809,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7533497402242275,
+      "grad_norm": 0.16371318697929382,
+      "learning_rate": 5e-05,
+      "loss": 1.58,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7536231884057971,
+      "grad_norm": 0.1591997742652893,
+      "learning_rate": 5e-05,
+      "loss": 1.5595,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7538966365873667,
+      "grad_norm": 0.1630491465330124,
+      "learning_rate": 5e-05,
+      "loss": 1.6129,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7541700847689363,
+      "grad_norm": 0.1691833734512329,
+      "learning_rate": 5e-05,
+      "loss": 1.6539,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7544435329505059,
+      "grad_norm": 0.16124382615089417,
+      "learning_rate": 5e-05,
+      "loss": 1.6068,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.16120778024196625,
+      "learning_rate": 5e-05,
+      "loss": 1.6477,
+      "step": 2760
+    },
+    {
+      "epoch": 0.754990429313645,
+      "grad_norm": 0.16382825374603271,
+      "learning_rate": 5e-05,
+      "loss": 1.6095,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7552638774952146,
+      "grad_norm": 0.16977065801620483,
+      "learning_rate": 5e-05,
+      "loss": 1.6844,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7555373256767842,
+      "grad_norm": 0.16113348305225372,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7558107738583538,
+      "grad_norm": 0.17500002682209015,
+      "learning_rate": 5e-05,
+      "loss": 1.5729,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7560842220399234,
+      "grad_norm": 0.16505929827690125,
+      "learning_rate": 5e-05,
+      "loss": 1.5522,
+      "step": 2765
+    },
+    {
+      "epoch": 0.756357670221493,
+      "grad_norm": 0.16675904393196106,
+      "learning_rate": 5e-05,
+      "loss": 1.6734,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7566311184030626,
+      "grad_norm": 0.20538048446178436,
+      "learning_rate": 5e-05,
+      "loss": 1.6777,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7569045665846322,
+      "grad_norm": 0.17019402980804443,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7571780147662018,
+      "grad_norm": 0.18559719622135162,
+      "learning_rate": 5e-05,
+      "loss": 1.6489,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7574514629477714,
+      "grad_norm": 0.16804775595664978,
+      "learning_rate": 5e-05,
+      "loss": 1.5647,
+      "step": 2770
+    },
+    {
+      "epoch": 0.757724911129341,
+      "grad_norm": 0.1561867594718933,
+      "learning_rate": 5e-05,
+      "loss": 1.5402,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7579983593109105,
+      "grad_norm": 0.1730283796787262,
+      "learning_rate": 5e-05,
+      "loss": 1.6805,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7582718074924801,
+      "grad_norm": 0.17423667013645172,
+      "learning_rate": 5e-05,
+      "loss": 1.6453,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7585452556740497,
+      "grad_norm": 0.1615607887506485,
+      "learning_rate": 5e-05,
+      "loss": 1.5584,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7588187038556193,
+      "grad_norm": 0.16401253640651703,
+      "learning_rate": 5e-05,
+      "loss": 1.5841,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7590921520371889,
+      "grad_norm": 0.1691298633813858,
+      "learning_rate": 5e-05,
+      "loss": 1.6887,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7593656002187585,
+      "grad_norm": 0.15560850501060486,
+      "learning_rate": 5e-05,
+      "loss": 1.4923,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7596390484003281,
+      "grad_norm": 0.1585303395986557,
+      "learning_rate": 5e-05,
+      "loss": 1.5857,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7599124965818977,
+      "grad_norm": 0.16869275271892548,
+      "learning_rate": 5e-05,
+      "loss": 1.5873,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7601859447634673,
+      "grad_norm": 0.17015528678894043,
+      "learning_rate": 5e-05,
+      "loss": 1.7419,
+      "step": 2780
+    },
+    {
+      "epoch": 0.760459392945037,
+      "grad_norm": 0.1577947735786438,
+      "learning_rate": 5e-05,
+      "loss": 1.5961,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7607328411266066,
+      "grad_norm": 0.16766513884067535,
+      "learning_rate": 5e-05,
+      "loss": 1.6623,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7610062893081762,
+      "grad_norm": 0.15825945138931274,
+      "learning_rate": 5e-05,
+      "loss": 1.6,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7612797374897456,
+      "grad_norm": 0.16338619589805603,
+      "learning_rate": 5e-05,
+      "loss": 1.677,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7615531856713152,
+      "grad_norm": 0.16439124941825867,
+      "learning_rate": 5e-05,
+      "loss": 1.6444,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7618266338528848,
+      "grad_norm": 0.1587378978729248,
+      "learning_rate": 5e-05,
+      "loss": 1.6477,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7621000820344545,
+      "grad_norm": 0.16001760959625244,
+      "learning_rate": 5e-05,
+      "loss": 1.638,
+      "step": 2787
+    },
+    {
+      "epoch": 0.762373530216024,
+      "grad_norm": 0.16488565504550934,
+      "learning_rate": 5e-05,
+      "loss": 1.5758,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7626469783975937,
+      "grad_norm": 0.17125697433948517,
+      "learning_rate": 5e-05,
+      "loss": 1.6488,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7629204265791633,
+      "grad_norm": 0.16552145779132843,
+      "learning_rate": 5e-05,
+      "loss": 1.6288,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7631938747607329,
+      "grad_norm": 0.16142013669013977,
+      "learning_rate": 5e-05,
+      "loss": 1.5717,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7634673229423025,
+      "grad_norm": 0.17424419522285461,
+      "learning_rate": 5e-05,
+      "loss": 1.6741,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7637407711238721,
+      "grad_norm": 0.1656377911567688,
+      "learning_rate": 5e-05,
+      "loss": 1.5975,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7640142193054417,
+      "grad_norm": 0.1610795110464096,
+      "learning_rate": 5e-05,
+      "loss": 1.6058,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7642876674870112,
+      "grad_norm": 0.160441592335701,
+      "learning_rate": 5e-05,
+      "loss": 1.5343,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7645611156685808,
+      "grad_norm": 0.16838759183883667,
+      "learning_rate": 5e-05,
+      "loss": 1.5573,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7648345638501504,
+      "grad_norm": 0.16315698623657227,
+      "learning_rate": 5e-05,
+      "loss": 1.5484,
+      "step": 2797
+    },
+    {
+      "epoch": 0.76510801203172,
+      "grad_norm": 0.17179346084594727,
+      "learning_rate": 5e-05,
+      "loss": 1.7635,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7653814602132896,
+      "grad_norm": 0.15899762511253357,
+      "learning_rate": 5e-05,
+      "loss": 1.645,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7656549083948592,
+      "grad_norm": 0.1585722267627716,
+      "learning_rate": 5e-05,
+      "loss": 1.5657,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7659283565764288,
+      "grad_norm": 0.16446919739246368,
+      "learning_rate": 5e-05,
+      "loss": 1.6099,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7662018047579984,
+      "grad_norm": 0.16195422410964966,
+      "learning_rate": 5e-05,
+      "loss": 1.5312,
+      "step": 2802
+    },
+    {
+      "epoch": 0.766475252939568,
+      "grad_norm": 0.1638331413269043,
+      "learning_rate": 5e-05,
+      "loss": 1.573,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7667487011211376,
+      "grad_norm": 0.16117534041404724,
+      "learning_rate": 5e-05,
+      "loss": 1.5535,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7670221493027072,
+      "grad_norm": 0.1906486600637436,
+      "learning_rate": 5e-05,
+      "loss": 1.6745,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7672955974842768,
+      "grad_norm": 0.1637091189622879,
+      "learning_rate": 5e-05,
+      "loss": 1.6076,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7675690456658463,
+      "grad_norm": 0.16975869238376617,
+      "learning_rate": 5e-05,
+      "loss": 1.5807,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7678424938474159,
+      "grad_norm": 0.16839958727359772,
+      "learning_rate": 5e-05,
+      "loss": 1.5136,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7681159420289855,
+      "grad_norm": 0.16886620223522186,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7683893902105551,
+      "grad_norm": 0.16913080215454102,
+      "learning_rate": 5e-05,
+      "loss": 1.547,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7686628383921247,
+      "grad_norm": 0.17105981707572937,
+      "learning_rate": 5e-05,
+      "loss": 1.6501,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7689362865736943,
+      "grad_norm": 0.16617277264595032,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7692097347552639,
+      "grad_norm": 0.16640552878379822,
+      "learning_rate": 5e-05,
+      "loss": 1.5152,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7694831829368335,
+      "grad_norm": 0.1717168390750885,
+      "learning_rate": 5e-05,
+      "loss": 1.5904,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7697566311184031,
+      "grad_norm": 0.15507300198078156,
+      "learning_rate": 5e-05,
+      "loss": 1.4552,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7700300792999727,
+      "grad_norm": 0.16422231495380402,
+      "learning_rate": 5e-05,
+      "loss": 1.6003,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7703035274815423,
+      "grad_norm": 0.16993476450443268,
+      "learning_rate": 5e-05,
+      "loss": 1.5668,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7705769756631118,
+      "grad_norm": 0.16527430713176727,
+      "learning_rate": 5e-05,
+      "loss": 1.5809,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7708504238446814,
+      "grad_norm": 0.16234123706817627,
+      "learning_rate": 5e-05,
+      "loss": 1.5393,
+      "step": 2819
+    },
+    {
+      "epoch": 0.771123872026251,
+      "grad_norm": 0.17365649342536926,
+      "learning_rate": 5e-05,
+      "loss": 1.5758,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7713973202078206,
+      "grad_norm": 0.17129479348659515,
+      "learning_rate": 5e-05,
+      "loss": 1.6084,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7716707683893902,
+      "grad_norm": 0.16345179080963135,
+      "learning_rate": 5e-05,
+      "loss": 1.5886,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7719442165709598,
+      "grad_norm": 0.17358049750328064,
+      "learning_rate": 5e-05,
+      "loss": 1.6457,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7722176647525294,
+      "grad_norm": 0.17579670250415802,
+      "learning_rate": 5e-05,
+      "loss": 1.7123,
+      "step": 2824
+    },
+    {
+      "epoch": 0.772491112934099,
+      "grad_norm": 0.1691892147064209,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7727645611156686,
+      "grad_norm": 0.17657522857189178,
+      "learning_rate": 5e-05,
+      "loss": 1.6072,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7730380092972382,
+      "grad_norm": 0.17486031353473663,
+      "learning_rate": 5e-05,
+      "loss": 1.5497,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7733114574788078,
+      "grad_norm": 0.16051729023456573,
+      "learning_rate": 5e-05,
+      "loss": 1.4969,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7735849056603774,
+      "grad_norm": 0.16621479392051697,
+      "learning_rate": 5e-05,
+      "loss": 1.6151,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7738583538419469,
+      "grad_norm": 0.16369469463825226,
+      "learning_rate": 5e-05,
+      "loss": 1.6707,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7741318020235165,
+      "grad_norm": 0.1664346307516098,
+      "learning_rate": 5e-05,
+      "loss": 1.552,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7744052502050861,
+      "grad_norm": 0.1688821017742157,
+      "learning_rate": 5e-05,
+      "loss": 1.566,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7746786983866557,
+      "grad_norm": 0.15843422710895538,
+      "learning_rate": 5e-05,
+      "loss": 1.6504,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7749521465682253,
+      "grad_norm": 0.16451962292194366,
+      "learning_rate": 5e-05,
+      "loss": 1.5611,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7752255947497949,
+      "grad_norm": 0.17255376279354095,
+      "learning_rate": 5e-05,
+      "loss": 1.5388,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7754990429313645,
+      "grad_norm": 0.1702892780303955,
+      "learning_rate": 5e-05,
+      "loss": 1.6152,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7757724911129341,
+      "grad_norm": 0.16479864716529846,
+      "learning_rate": 5e-05,
+      "loss": 1.677,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7760459392945037,
+      "grad_norm": 0.16758742928504944,
+      "learning_rate": 5e-05,
+      "loss": 1.5406,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7763193874760733,
+      "grad_norm": 0.1617206186056137,
+      "learning_rate": 5e-05,
+      "loss": 1.635,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7765928356576429,
+      "grad_norm": 0.1629093438386917,
+      "learning_rate": 5e-05,
+      "loss": 1.5516,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7768662838392125,
+      "grad_norm": 0.16901403665542603,
+      "learning_rate": 5e-05,
+      "loss": 1.6223,
+      "step": 2841
+    },
+    {
+      "epoch": 0.777139732020782,
+      "grad_norm": 0.16104839742183685,
+      "learning_rate": 5e-05,
+      "loss": 1.6432,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7774131802023516,
+      "grad_norm": 0.15974918007850647,
+      "learning_rate": 5e-05,
+      "loss": 1.5034,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7776866283839212,
+      "grad_norm": 0.1669033169746399,
+      "learning_rate": 5e-05,
+      "loss": 1.6008,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7779600765654908,
+      "grad_norm": 0.17953120172023773,
+      "learning_rate": 5e-05,
+      "loss": 1.6714,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7782335247470604,
+      "grad_norm": 0.16718675196170807,
+      "learning_rate": 5e-05,
+      "loss": 1.5205,
+      "step": 2846
+    },
+    {
+      "epoch": 0.77850697292863,
+      "grad_norm": 0.16724152863025665,
+      "learning_rate": 5e-05,
+      "loss": 1.5884,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7787804211101996,
+      "grad_norm": 0.16082365810871124,
+      "learning_rate": 5e-05,
+      "loss": 1.6615,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7790538692917692,
+      "grad_norm": 0.15853464603424072,
+      "learning_rate": 5e-05,
+      "loss": 1.5306,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7793273174733388,
+      "grad_norm": 0.1667121946811676,
+      "learning_rate": 5e-05,
+      "loss": 1.5941,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7796007656549084,
+      "grad_norm": 0.1697937250137329,
+      "learning_rate": 5e-05,
+      "loss": 1.6127,
+      "step": 2851
+    },
+    {
+      "epoch": 0.779874213836478,
+      "grad_norm": 0.158981055021286,
+      "learning_rate": 5e-05,
+      "loss": 1.5672,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7801476620180475,
+      "grad_norm": 0.1826319545507431,
+      "learning_rate": 5e-05,
+      "loss": 1.6604,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7804211101996171,
+      "grad_norm": 0.17479664087295532,
+      "learning_rate": 5e-05,
+      "loss": 1.6348,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7806945583811867,
+      "grad_norm": 0.17905382812023163,
+      "learning_rate": 5e-05,
+      "loss": 1.5434,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7809680065627563,
+      "grad_norm": 0.1738382875919342,
+      "learning_rate": 5e-05,
+      "loss": 1.6045,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7812414547443259,
+      "grad_norm": 0.16724707186222076,
+      "learning_rate": 5e-05,
+      "loss": 1.5591,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7815149029258955,
+      "grad_norm": 0.17346110939979553,
+      "learning_rate": 5e-05,
+      "loss": 1.5721,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7817883511074651,
+      "grad_norm": 0.16252191364765167,
+      "learning_rate": 5e-05,
+      "loss": 1.5935,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7820617992890347,
+      "grad_norm": 0.1672922968864441,
+      "learning_rate": 5e-05,
+      "loss": 1.6467,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7823352474706043,
+      "grad_norm": 0.1666378527879715,
+      "learning_rate": 5e-05,
+      "loss": 1.6392,
+      "step": 2861
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.1650533229112625,
+      "learning_rate": 5e-05,
+      "loss": 1.5737,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7828821438337435,
+      "grad_norm": 0.16479378938674927,
+      "learning_rate": 5e-05,
+      "loss": 1.6066,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7831555920153132,
+      "grad_norm": 0.16486263275146484,
+      "learning_rate": 5e-05,
+      "loss": 1.5617,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7834290401968826,
+      "grad_norm": 0.1614285111427307,
+      "learning_rate": 5e-05,
+      "loss": 1.6226,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7837024883784522,
+      "grad_norm": 0.176860049366951,
+      "learning_rate": 5e-05,
+      "loss": 1.5129,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7839759365600218,
+      "grad_norm": 0.16381007432937622,
+      "learning_rate": 5e-05,
+      "loss": 1.6524,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7842493847415914,
+      "grad_norm": 0.15981391072273254,
+      "learning_rate": 5e-05,
+      "loss": 1.5611,
+      "step": 2868
+    },
+    {
+      "epoch": 0.784522832923161,
+      "grad_norm": 0.17717792093753815,
+      "learning_rate": 5e-05,
+      "loss": 1.691,
+      "step": 2869
+    },
+    {
+      "epoch": 0.7847962811047307,
+      "grad_norm": 0.15418937802314758,
+      "learning_rate": 5e-05,
+      "loss": 1.5502,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7850697292863003,
+      "grad_norm": 0.17375600337982178,
+      "learning_rate": 5e-05,
+      "loss": 1.6708,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7853431774678699,
+      "grad_norm": 0.16681233048439026,
+      "learning_rate": 5e-05,
+      "loss": 1.6323,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7856166256494395,
+      "grad_norm": 0.15948568284511566,
+      "learning_rate": 5e-05,
+      "loss": 1.5582,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7858900738310091,
+      "grad_norm": 0.16835032403469086,
+      "learning_rate": 5e-05,
+      "loss": 1.5979,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7861635220125787,
+      "grad_norm": 0.16147814691066742,
+      "learning_rate": 5e-05,
+      "loss": 1.5618,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7864369701941482,
+      "grad_norm": 0.15999336540699005,
+      "learning_rate": 5e-05,
+      "loss": 1.5351,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7867104183757178,
+      "grad_norm": 0.17828235030174255,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7869838665572874,
+      "grad_norm": 0.16279451549053192,
+      "learning_rate": 5e-05,
+      "loss": 1.5923,
+      "step": 2878
+    },
+    {
+      "epoch": 0.787257314738857,
+      "grad_norm": 0.16363869607448578,
+      "learning_rate": 5e-05,
+      "loss": 1.6645,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7875307629204266,
+      "grad_norm": 0.17845980823040009,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7878042111019962,
+      "grad_norm": 0.17278538644313812,
+      "learning_rate": 5e-05,
+      "loss": 1.6211,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7880776592835658,
+      "grad_norm": 0.15572743117809296,
+      "learning_rate": 5e-05,
+      "loss": 1.5377,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7883511074651354,
+      "grad_norm": 0.18251000344753265,
+      "learning_rate": 5e-05,
+      "loss": 1.769,
+      "step": 2883
+    },
+    {
+      "epoch": 0.788624555646705,
+      "grad_norm": 0.17334499955177307,
+      "learning_rate": 5e-05,
+      "loss": 1.6428,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7888980038282746,
+      "grad_norm": 0.17535856366157532,
+      "learning_rate": 5e-05,
+      "loss": 1.7187,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7891714520098442,
+      "grad_norm": 0.16317224502563477,
+      "learning_rate": 5e-05,
+      "loss": 1.4863,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7894449001914138,
+      "grad_norm": 0.18952353298664093,
+      "learning_rate": 5e-05,
+      "loss": 1.7284,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7897183483729833,
+      "grad_norm": 0.17403848469257355,
+      "learning_rate": 5e-05,
+      "loss": 1.5511,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7899917965545529,
+      "grad_norm": 0.17016200721263885,
+      "learning_rate": 5e-05,
+      "loss": 1.5673,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7902652447361225,
+      "grad_norm": 0.16698943078517914,
+      "learning_rate": 5e-05,
+      "loss": 1.5506,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7905386929176921,
+      "grad_norm": 0.1672915667295456,
+      "learning_rate": 5e-05,
+      "loss": 1.5712,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7908121410992617,
+      "grad_norm": 0.17247925698757172,
+      "learning_rate": 5e-05,
+      "loss": 1.4931,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7910855892808313,
+      "grad_norm": 0.15129195153713226,
+      "learning_rate": 5e-05,
+      "loss": 1.5219,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7913590374624009,
+      "grad_norm": 0.16134491562843323,
+      "learning_rate": 5e-05,
+      "loss": 1.5749,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7916324856439705,
+      "grad_norm": 0.17356084287166595,
+      "learning_rate": 5e-05,
+      "loss": 1.607,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7919059338255401,
+      "grad_norm": 0.16742588579654694,
+      "learning_rate": 5e-05,
+      "loss": 1.515,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7921793820071097,
+      "grad_norm": 0.16483165323734283,
+      "learning_rate": 5e-05,
+      "loss": 1.5906,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.1711149662733078,
+      "learning_rate": 5e-05,
+      "loss": 1.6759,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7927262783702488,
+      "grad_norm": 0.1694236695766449,
+      "learning_rate": 5e-05,
+      "loss": 1.5804,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7929997265518184,
+      "grad_norm": 0.15667274594306946,
+      "learning_rate": 5e-05,
+      "loss": 1.5717,
+      "step": 2900
+    },
+    {
+      "epoch": 0.793273174733388,
+      "grad_norm": 0.18144182860851288,
+      "learning_rate": 5e-05,
+      "loss": 1.5945,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7935466229149576,
+      "grad_norm": 0.16796544194221497,
+      "learning_rate": 5e-05,
+      "loss": 1.6503,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7938200710965272,
+      "grad_norm": 0.16234607994556427,
+      "learning_rate": 5e-05,
+      "loss": 1.6427,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7940935192780968,
+      "grad_norm": 0.16550585627555847,
+      "learning_rate": 5e-05,
+      "loss": 1.6079,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7943669674596664,
+      "grad_norm": 0.1626591831445694,
+      "learning_rate": 5e-05,
+      "loss": 1.5215,
+      "step": 2905
+    },
+    {
+      "epoch": 0.794640415641236,
+      "grad_norm": 0.1648007035255432,
+      "learning_rate": 5e-05,
+      "loss": 1.4877,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7949138638228056,
+      "grad_norm": 0.1771146059036255,
+      "learning_rate": 5e-05,
+      "loss": 1.5802,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7951873120043752,
+      "grad_norm": 0.16548393666744232,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7954607601859448,
+      "grad_norm": 0.18224813044071198,
+      "learning_rate": 5e-05,
+      "loss": 1.5849,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7957342083675144,
+      "grad_norm": 0.19087089598178864,
+      "learning_rate": 5e-05,
+      "loss": 1.5959,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7960076565490839,
+      "grad_norm": 0.16123077273368835,
+      "learning_rate": 5e-05,
+      "loss": 1.5631,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7962811047306535,
+      "grad_norm": 0.18250906467437744,
+      "learning_rate": 5e-05,
+      "loss": 1.6527,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7965545529122231,
+      "grad_norm": 0.18301981687545776,
+      "learning_rate": 5e-05,
+      "loss": 1.6438,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7968280010937927,
+      "grad_norm": 0.15846803784370422,
+      "learning_rate": 5e-05,
+      "loss": 1.5266,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7971014492753623,
+      "grad_norm": 0.16750794649124146,
+      "learning_rate": 5e-05,
+      "loss": 1.6276,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7973748974569319,
+      "grad_norm": 0.16995219886302948,
+      "learning_rate": 5e-05,
+      "loss": 1.7035,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7976483456385015,
+      "grad_norm": 0.16649973392486572,
+      "learning_rate": 5e-05,
+      "loss": 1.5616,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7979217938200711,
+      "grad_norm": 0.14843766391277313,
+      "learning_rate": 5e-05,
+      "loss": 1.428,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7981952420016407,
+      "grad_norm": 0.17352357506752014,
+      "learning_rate": 5e-05,
+      "loss": 1.6872,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7984686901832103,
+      "grad_norm": 0.16598494350910187,
+      "learning_rate": 5e-05,
+      "loss": 1.6047,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7987421383647799,
+      "grad_norm": 0.1559354066848755,
+      "learning_rate": 5e-05,
+      "loss": 1.5798,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7990155865463495,
+      "grad_norm": 0.1671651154756546,
+      "learning_rate": 5e-05,
+      "loss": 1.599,
+      "step": 2922
+    },
+    {
+      "epoch": 0.799289034727919,
+      "grad_norm": 0.15780937671661377,
+      "learning_rate": 5e-05,
+      "loss": 1.5554,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7995624829094886,
+      "grad_norm": 0.16018469631671906,
+      "learning_rate": 5e-05,
+      "loss": 1.5614,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7998359310910582,
+      "grad_norm": 0.16393598914146423,
+      "learning_rate": 5e-05,
+      "loss": 1.6024,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8001093792726278,
+      "grad_norm": 0.15699848532676697,
+      "learning_rate": 5e-05,
+      "loss": 1.5343,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8003828274541974,
+      "grad_norm": 0.1754962056875229,
+      "learning_rate": 5e-05,
+      "loss": 1.6674,
+      "step": 2927
+    },
+    {
+      "epoch": 0.800656275635767,
+      "grad_norm": 0.16628246009349823,
+      "learning_rate": 5e-05,
+      "loss": 1.6243,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8009297238173366,
+      "grad_norm": 0.1607159823179245,
+      "learning_rate": 5e-05,
+      "loss": 1.6163,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8012031719989062,
+      "grad_norm": 0.17041060328483582,
+      "learning_rate": 5e-05,
+      "loss": 1.6411,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8014766201804758,
+      "grad_norm": 0.1616741120815277,
+      "learning_rate": 5e-05,
+      "loss": 1.4652,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8017500683620454,
+      "grad_norm": 0.16323620080947876,
+      "learning_rate": 5e-05,
+      "loss": 1.6336,
+      "step": 2932
+    },
+    {
+      "epoch": 0.802023516543615,
+      "grad_norm": 0.16059629619121552,
+      "learning_rate": 5e-05,
+      "loss": 1.6735,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8022969647251845,
+      "grad_norm": 0.17597347497940063,
+      "learning_rate": 5e-05,
+      "loss": 1.6076,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8025704129067541,
+      "grad_norm": 0.16613614559173584,
+      "learning_rate": 5e-05,
+      "loss": 1.607,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8028438610883237,
+      "grad_norm": 0.17486806213855743,
+      "learning_rate": 5e-05,
+      "loss": 1.6023,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8031173092698933,
+      "grad_norm": 0.1755034178495407,
+      "learning_rate": 5e-05,
+      "loss": 1.601,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8033907574514629,
+      "grad_norm": 0.16530238091945648,
+      "learning_rate": 5e-05,
+      "loss": 1.5338,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8036642056330325,
+      "grad_norm": 0.18131126463413239,
+      "learning_rate": 5e-05,
+      "loss": 1.7062,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8039376538146021,
+      "grad_norm": 0.1728467047214508,
+      "learning_rate": 5e-05,
+      "loss": 1.6545,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8042111019961717,
+      "grad_norm": 0.15401597321033478,
+      "learning_rate": 5e-05,
+      "loss": 1.4845,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8044845501777413,
+      "grad_norm": 0.18329520523548126,
+      "learning_rate": 5e-05,
+      "loss": 1.6014,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8047579983593109,
+      "grad_norm": 0.16439451277256012,
+      "learning_rate": 5e-05,
+      "loss": 1.5848,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8050314465408805,
+      "grad_norm": 0.16697899997234344,
+      "learning_rate": 5e-05,
+      "loss": 1.6285,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8053048947224501,
+      "grad_norm": 0.1735549122095108,
+      "learning_rate": 5e-05,
+      "loss": 1.7079,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8055783429040196,
+      "grad_norm": 0.16093853116035461,
+      "learning_rate": 5e-05,
+      "loss": 1.5457,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8058517910855892,
+      "grad_norm": 0.1812943071126938,
+      "learning_rate": 5e-05,
+      "loss": 1.7846,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8061252392671588,
+      "grad_norm": 0.1680847555398941,
+      "learning_rate": 5e-05,
+      "loss": 1.6172,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8063986874487284,
+      "grad_norm": 0.1760333627462387,
+      "learning_rate": 5e-05,
+      "loss": 1.6645,
+      "step": 2949
+    },
+    {
+      "epoch": 0.806672135630298,
+      "grad_norm": 0.16148562729358673,
+      "learning_rate": 5e-05,
+      "loss": 1.5527,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8069455838118677,
+      "grad_norm": 0.16784906387329102,
+      "learning_rate": 5e-05,
+      "loss": 1.6316,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8072190319934373,
+      "grad_norm": 0.16758820414543152,
+      "learning_rate": 5e-05,
+      "loss": 1.566,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8074924801750069,
+      "grad_norm": 0.16588006913661957,
+      "learning_rate": 5e-05,
+      "loss": 1.5411,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8077659283565765,
+      "grad_norm": 0.15759024024009705,
+      "learning_rate": 5e-05,
+      "loss": 1.497,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8080393765381461,
+      "grad_norm": 0.1606815755367279,
+      "learning_rate": 5e-05,
+      "loss": 1.5874,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8083128247197157,
+      "grad_norm": 0.171406090259552,
+      "learning_rate": 5e-05,
+      "loss": 1.6286,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8085862729012852,
+      "grad_norm": 0.1659841537475586,
+      "learning_rate": 5e-05,
+      "loss": 1.642,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8088597210828548,
+      "grad_norm": 0.1642453819513321,
+      "learning_rate": 5e-05,
+      "loss": 1.6138,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8091331692644244,
+      "grad_norm": 0.1737990528345108,
+      "learning_rate": 5e-05,
+      "loss": 1.6755,
+      "step": 2959
+    },
+    {
+      "epoch": 0.809406617445994,
+      "grad_norm": 0.16041895747184753,
+      "learning_rate": 5e-05,
+      "loss": 1.5752,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8096800656275636,
+      "grad_norm": 0.16759610176086426,
+      "learning_rate": 5e-05,
+      "loss": 1.5574,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8099535138091332,
+      "grad_norm": 0.17192721366882324,
+      "learning_rate": 5e-05,
+      "loss": 1.6143,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8102269619907028,
+      "grad_norm": 0.17108719050884247,
+      "learning_rate": 5e-05,
+      "loss": 1.668,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8105004101722724,
+      "grad_norm": 0.1675434559583664,
+      "learning_rate": 5e-05,
+      "loss": 1.596,
+      "step": 2964
+    },
+    {
+      "epoch": 0.810773858353842,
+      "grad_norm": 0.16867220401763916,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8110473065354116,
+      "grad_norm": 0.16799405217170715,
+      "learning_rate": 5e-05,
+      "loss": 1.6316,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8113207547169812,
+      "grad_norm": 0.1661425083875656,
+      "learning_rate": 5e-05,
+      "loss": 1.6354,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8115942028985508,
+      "grad_norm": 0.17176170647144318,
+      "learning_rate": 5e-05,
+      "loss": 1.5873,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8118676510801203,
+      "grad_norm": 0.16512513160705566,
+      "learning_rate": 5e-05,
+      "loss": 1.5885,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8121410992616899,
+      "grad_norm": 0.1690187305212021,
+      "learning_rate": 5e-05,
+      "loss": 1.6306,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8124145474432595,
+      "grad_norm": 0.17420509457588196,
+      "learning_rate": 5e-05,
+      "loss": 1.49,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8126879956248291,
+      "grad_norm": 0.16634975373744965,
+      "learning_rate": 5e-05,
+      "loss": 1.5559,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8129614438063987,
+      "grad_norm": 0.1683768481016159,
+      "learning_rate": 5e-05,
+      "loss": 1.5532,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8132348919879683,
+      "grad_norm": 0.16414394974708557,
+      "learning_rate": 5e-05,
+      "loss": 1.5992,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8135083401695379,
+      "grad_norm": 0.16777318716049194,
+      "learning_rate": 5e-05,
+      "loss": 1.6227,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8137817883511075,
+      "grad_norm": 0.16975970566272736,
+      "learning_rate": 5e-05,
+      "loss": 1.589,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8140552365326771,
+      "grad_norm": 0.16181518137454987,
+      "learning_rate": 5e-05,
+      "loss": 1.6104,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8143286847142467,
+      "grad_norm": 0.16820965707302094,
+      "learning_rate": 5e-05,
+      "loss": 1.597,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8146021328958163,
+      "grad_norm": 0.17384564876556396,
+      "learning_rate": 5e-05,
+      "loss": 1.61,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8148755810773858,
+      "grad_norm": 0.17363837361335754,
+      "learning_rate": 5e-05,
+      "loss": 1.6574,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8151490292589554,
+      "grad_norm": 0.17065000534057617,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 2981
+    },
+    {
+      "epoch": 0.815422477440525,
+      "grad_norm": 0.1675603985786438,
+      "learning_rate": 5e-05,
+      "loss": 1.6028,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8156959256220946,
+      "grad_norm": 0.16641339659690857,
+      "learning_rate": 5e-05,
+      "loss": 1.6176,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8159693738036642,
+      "grad_norm": 0.16285206377506256,
+      "learning_rate": 5e-05,
+      "loss": 1.6211,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8162428219852338,
+      "grad_norm": 0.1820947378873825,
+      "learning_rate": 5e-05,
+      "loss": 1.6473,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8165162701668034,
+      "grad_norm": 0.16042275726795197,
+      "learning_rate": 5e-05,
+      "loss": 1.5818,
+      "step": 2986
+    },
+    {
+      "epoch": 0.816789718348373,
+      "grad_norm": 0.1615065634250641,
+      "learning_rate": 5e-05,
+      "loss": 1.5165,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8170631665299426,
+      "grad_norm": 0.17090201377868652,
+      "learning_rate": 5e-05,
+      "loss": 1.653,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8173366147115122,
+      "grad_norm": 0.1570872813463211,
+      "learning_rate": 5e-05,
+      "loss": 1.5234,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8176100628930818,
+      "grad_norm": 0.16220347583293915,
+      "learning_rate": 5e-05,
+      "loss": 1.6026,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8178835110746514,
+      "grad_norm": 0.17709346115589142,
+      "learning_rate": 5e-05,
+      "loss": 1.5405,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8181569592562209,
+      "grad_norm": 0.15809805691242218,
+      "learning_rate": 5e-05,
+      "loss": 1.5744,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8184304074377905,
+      "grad_norm": 0.17646804451942444,
+      "learning_rate": 5e-05,
+      "loss": 1.6328,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8187038556193601,
+      "grad_norm": 0.19047993421554565,
+      "learning_rate": 5e-05,
+      "loss": 1.6451,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8189773038009297,
+      "grad_norm": 0.16458064317703247,
+      "learning_rate": 5e-05,
+      "loss": 1.6942,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8192507519824993,
+      "grad_norm": 0.17090070247650146,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8195242001640689,
+      "grad_norm": 0.1800215095281601,
+      "learning_rate": 5e-05,
+      "loss": 1.5693,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8197976483456385,
+      "grad_norm": 0.16883789002895355,
+      "learning_rate": 5e-05,
+      "loss": 1.5624,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8200710965272081,
+      "grad_norm": 0.16943061351776123,
+      "learning_rate": 5e-05,
+      "loss": 1.6307,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8203445447087777,
+      "grad_norm": 0.1675194501876831,
+      "learning_rate": 5e-05,
+      "loss": 1.5902,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8206179928903473,
+      "grad_norm": 0.16143114864826202,
+      "learning_rate": 5e-05,
+      "loss": 1.5696,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8208914410719169,
+      "grad_norm": 0.16645751893520355,
+      "learning_rate": 5e-05,
+      "loss": 1.6472,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8211648892534865,
+      "grad_norm": 0.16916631162166595,
+      "learning_rate": 5e-05,
+      "loss": 1.5685,
+      "step": 3003
+    },
+    {
+      "epoch": 0.821438337435056,
+      "grad_norm": 0.1605367660522461,
+      "learning_rate": 5e-05,
+      "loss": 1.6069,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8217117856166256,
+      "grad_norm": 0.1694014072418213,
+      "learning_rate": 5e-05,
+      "loss": 1.6411,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8219852337981952,
+      "grad_norm": 0.15868619084358215,
+      "learning_rate": 5e-05,
+      "loss": 1.5634,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8222586819797648,
+      "grad_norm": 0.16402362287044525,
+      "learning_rate": 5e-05,
+      "loss": 1.6254,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8225321301613344,
+      "grad_norm": 0.16820335388183594,
+      "learning_rate": 5e-05,
+      "loss": 1.6344,
+      "step": 3008
+    },
+    {
+      "epoch": 0.822805578342904,
+      "grad_norm": 0.1659119576215744,
+      "learning_rate": 5e-05,
+      "loss": 1.6044,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8230790265244736,
+      "grad_norm": 0.177097886800766,
+      "learning_rate": 5e-05,
+      "loss": 1.7114,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8233524747060432,
+      "grad_norm": 0.16565869748592377,
+      "learning_rate": 5e-05,
+      "loss": 1.5614,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8236259228876128,
+      "grad_norm": 0.16346898674964905,
+      "learning_rate": 5e-05,
+      "loss": 1.6737,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8238993710691824,
+      "grad_norm": 0.16942520439624786,
+      "learning_rate": 5e-05,
+      "loss": 1.6536,
+      "step": 3013
+    },
+    {
+      "epoch": 0.824172819250752,
+      "grad_norm": 0.16573195159435272,
+      "learning_rate": 5e-05,
+      "loss": 1.5418,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8244462674323215,
+      "grad_norm": 0.16428519785404205,
+      "learning_rate": 5e-05,
+      "loss": 1.5565,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8247197156138911,
+      "grad_norm": 0.1609538197517395,
+      "learning_rate": 5e-05,
+      "loss": 1.5613,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8249931637954607,
+      "grad_norm": 0.16682593524456024,
+      "learning_rate": 5e-05,
+      "loss": 1.6829,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8252666119770303,
+      "grad_norm": 0.16691312193870544,
+      "learning_rate": 5e-05,
+      "loss": 1.5133,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8255400601585999,
+      "grad_norm": 0.16984447836875916,
+      "learning_rate": 5e-05,
+      "loss": 1.6283,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8258135083401695,
+      "grad_norm": 0.1617315411567688,
+      "learning_rate": 5e-05,
+      "loss": 1.5689,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 0.17331373691558838,
+      "learning_rate": 5e-05,
+      "loss": 1.635,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8263604047033087,
+      "grad_norm": 0.1681394726037979,
+      "learning_rate": 5e-05,
+      "loss": 1.6635,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8266338528848783,
+      "grad_norm": 0.15742860734462738,
+      "learning_rate": 5e-05,
+      "loss": 1.4818,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8269073010664479,
+      "grad_norm": 0.18009516596794128,
+      "learning_rate": 5e-05,
+      "loss": 1.6404,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8271807492480175,
+      "grad_norm": 0.16786369681358337,
+      "learning_rate": 5e-05,
+      "loss": 1.5971,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8274541974295871,
+      "grad_norm": 0.16309179365634918,
+      "learning_rate": 5e-05,
+      "loss": 1.5124,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8277276456111566,
+      "grad_norm": 0.15728282928466797,
+      "learning_rate": 5e-05,
+      "loss": 1.5749,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8280010937927262,
+      "grad_norm": 0.16731971502304077,
+      "learning_rate": 5e-05,
+      "loss": 1.5051,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8282745419742958,
+      "grad_norm": 0.17260520160198212,
+      "learning_rate": 5e-05,
+      "loss": 1.5307,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8285479901558654,
+      "grad_norm": 0.16881436109542847,
+      "learning_rate": 5e-05,
+      "loss": 1.6312,
+      "step": 3030
+    },
+    {
+      "epoch": 0.828821438337435,
+      "grad_norm": 0.16781583428382874,
+      "learning_rate": 5e-05,
+      "loss": 1.5454,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8290948865190046,
+      "grad_norm": 0.16846032440662384,
+      "learning_rate": 5e-05,
+      "loss": 1.6269,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8293683347005743,
+      "grad_norm": 0.1699826866388321,
+      "learning_rate": 5e-05,
+      "loss": 1.6368,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8296417828821439,
+      "grad_norm": 0.16753311455249786,
+      "learning_rate": 5e-05,
+      "loss": 1.6949,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8299152310637135,
+      "grad_norm": 0.17179977893829346,
+      "learning_rate": 5e-05,
+      "loss": 1.6023,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.1626930981874466,
+      "learning_rate": 5e-05,
+      "loss": 1.6162,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8304621274268527,
+      "grad_norm": 0.1600092500448227,
+      "learning_rate": 5e-05,
+      "loss": 1.5765,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8307355756084221,
+      "grad_norm": 0.1597457230091095,
+      "learning_rate": 5e-05,
+      "loss": 1.6114,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8310090237899918,
+      "grad_norm": 0.16018109023571014,
+      "learning_rate": 5e-05,
+      "loss": 1.6279,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8312824719715614,
+      "grad_norm": 0.15814651548862457,
+      "learning_rate": 5e-05,
+      "loss": 1.5031,
+      "step": 3040
+    },
+    {
+      "epoch": 0.831555920153131,
+      "grad_norm": 0.1627386063337326,
+      "learning_rate": 5e-05,
+      "loss": 1.5399,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8318293683347006,
+      "grad_norm": 0.17200924456119537,
+      "learning_rate": 5e-05,
+      "loss": 1.6497,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8321028165162702,
+      "grad_norm": 0.16421593725681305,
+      "learning_rate": 5e-05,
+      "loss": 1.653,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8323762646978398,
+      "grad_norm": 0.16235610842704773,
+      "learning_rate": 5e-05,
+      "loss": 1.498,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8326497128794094,
+      "grad_norm": 0.16004261374473572,
+      "learning_rate": 5e-05,
+      "loss": 1.5331,
+      "step": 3045
+    },
+    {
+      "epoch": 0.832923161060979,
+      "grad_norm": 0.17818929255008698,
+      "learning_rate": 5e-05,
+      "loss": 1.6328,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8331966092425486,
+      "grad_norm": 0.17339615523815155,
+      "learning_rate": 5e-05,
+      "loss": 1.6716,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8334700574241182,
+      "grad_norm": 0.16837354004383087,
+      "learning_rate": 5e-05,
+      "loss": 1.6282,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8337435056056878,
+      "grad_norm": 0.16777902841567993,
+      "learning_rate": 5e-05,
+      "loss": 1.596,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8340169537872573,
+      "grad_norm": 0.16719725728034973,
+      "learning_rate": 5e-05,
+      "loss": 1.6055,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8342904019688269,
+      "grad_norm": 0.16437400877475739,
+      "learning_rate": 5e-05,
+      "loss": 1.6341,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8345638501503965,
+      "grad_norm": 0.1672685593366623,
+      "learning_rate": 5e-05,
+      "loss": 1.6135,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8348372983319661,
+      "grad_norm": 0.16196739673614502,
+      "learning_rate": 5e-05,
+      "loss": 1.5584,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8351107465135357,
+      "grad_norm": 0.15699756145477295,
+      "learning_rate": 5e-05,
+      "loss": 1.5569,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8353841946951053,
+      "grad_norm": 0.1717248409986496,
+      "learning_rate": 5e-05,
+      "loss": 1.6093,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8356576428766749,
+      "grad_norm": 0.17324498295783997,
+      "learning_rate": 5e-05,
+      "loss": 1.6929,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8359310910582445,
+      "grad_norm": 0.17252692580223083,
+      "learning_rate": 5e-05,
+      "loss": 1.6724,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8362045392398141,
+      "grad_norm": 0.17730343341827393,
+      "learning_rate": 5e-05,
+      "loss": 1.6094,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8364779874213837,
+      "grad_norm": 0.15670277178287506,
+      "learning_rate": 5e-05,
+      "loss": 1.5384,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8367514356029533,
+      "grad_norm": 0.171427920460701,
+      "learning_rate": 5e-05,
+      "loss": 1.5808,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8370248837845229,
+      "grad_norm": 0.1834273487329483,
+      "learning_rate": 5e-05,
+      "loss": 1.6069,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8372983319660924,
+      "grad_norm": 0.1576196849346161,
+      "learning_rate": 5e-05,
+      "loss": 1.6032,
+      "step": 3062
+    },
+    {
+      "epoch": 0.837571780147662,
+      "grad_norm": 0.17979919910430908,
+      "learning_rate": 5e-05,
+      "loss": 1.6884,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8378452283292316,
+      "grad_norm": 0.16846168041229248,
+      "learning_rate": 5e-05,
+      "loss": 1.4665,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8381186765108012,
+      "grad_norm": 0.17571622133255005,
+      "learning_rate": 5e-05,
+      "loss": 1.5971,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8383921246923708,
+      "grad_norm": 0.1625003069639206,
+      "learning_rate": 5e-05,
+      "loss": 1.6018,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8386655728739404,
+      "grad_norm": 0.16521279513835907,
+      "learning_rate": 5e-05,
+      "loss": 1.5108,
+      "step": 3067
+    },
+    {
+      "epoch": 0.83893902105551,
+      "grad_norm": 0.19198334217071533,
+      "learning_rate": 5e-05,
+      "loss": 1.6102,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8392124692370796,
+      "grad_norm": 0.16152355074882507,
+      "learning_rate": 5e-05,
+      "loss": 1.5801,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8394859174186492,
+      "grad_norm": 0.159002423286438,
+      "learning_rate": 5e-05,
+      "loss": 1.5931,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8397593656002188,
+      "grad_norm": 0.16774891316890717,
+      "learning_rate": 5e-05,
+      "loss": 1.613,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8400328137817884,
+      "grad_norm": 0.17755338549613953,
+      "learning_rate": 5e-05,
+      "loss": 1.6408,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8403062619633579,
+      "grad_norm": 0.15868759155273438,
+      "learning_rate": 5e-05,
+      "loss": 1.6276,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8405797101449275,
+      "grad_norm": 0.1694575846195221,
+      "learning_rate": 5e-05,
+      "loss": 1.6421,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8408531583264971,
+      "grad_norm": 0.1652621626853943,
+      "learning_rate": 5e-05,
+      "loss": 1.5798,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8411266065080667,
+      "grad_norm": 0.1710357666015625,
+      "learning_rate": 5e-05,
+      "loss": 1.5947,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8414000546896363,
+      "grad_norm": 0.16072213649749756,
+      "learning_rate": 5e-05,
+      "loss": 1.5291,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8416735028712059,
+      "grad_norm": 0.18232187628746033,
+      "learning_rate": 5e-05,
+      "loss": 1.6492,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8419469510527755,
+      "grad_norm": 0.1688283234834671,
+      "learning_rate": 5e-05,
+      "loss": 1.6395,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8422203992343451,
+      "grad_norm": 0.170608788728714,
+      "learning_rate": 5e-05,
+      "loss": 1.6103,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8424938474159147,
+      "grad_norm": 0.16807672381401062,
+      "learning_rate": 5e-05,
+      "loss": 1.535,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8427672955974843,
+      "grad_norm": 0.17390793561935425,
+      "learning_rate": 5e-05,
+      "loss": 1.6763,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8430407437790539,
+      "grad_norm": 0.1748964488506317,
+      "learning_rate": 5e-05,
+      "loss": 1.6523,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8433141919606235,
+      "grad_norm": 0.17339888215065002,
+      "learning_rate": 5e-05,
+      "loss": 1.4646,
+      "step": 3084
+    },
+    {
+      "epoch": 0.843587640142193,
+      "grad_norm": 0.16728384792804718,
+      "learning_rate": 5e-05,
+      "loss": 1.5108,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8438610883237626,
+      "grad_norm": 0.1811836063861847,
+      "learning_rate": 5e-05,
+      "loss": 1.6013,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8441345365053322,
+      "grad_norm": 0.15502838790416718,
+      "learning_rate": 5e-05,
+      "loss": 1.5567,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8444079846869018,
+      "grad_norm": 0.17469522356987,
+      "learning_rate": 5e-05,
+      "loss": 1.6128,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8446814328684714,
+      "grad_norm": 0.1718490719795227,
+      "learning_rate": 5e-05,
+      "loss": 1.6778,
+      "step": 3089
+    },
+    {
+      "epoch": 0.844954881050041,
+      "grad_norm": 0.1672654151916504,
+      "learning_rate": 5e-05,
+      "loss": 1.5281,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8452283292316106,
+      "grad_norm": 0.18899762630462646,
+      "learning_rate": 5e-05,
+      "loss": 1.5707,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8455017774131802,
+      "grad_norm": 0.16991832852363586,
+      "learning_rate": 5e-05,
+      "loss": 1.6655,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8457752255947498,
+      "grad_norm": 0.16526849567890167,
+      "learning_rate": 5e-05,
+      "loss": 1.5944,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8460486737763194,
+      "grad_norm": 0.1665259599685669,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 3094
+    },
+    {
+      "epoch": 0.846322121957889,
+      "grad_norm": 0.16578637063503265,
+      "learning_rate": 5e-05,
+      "loss": 1.5696,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8465955701394585,
+      "grad_norm": 0.15842179954051971,
+      "learning_rate": 5e-05,
+      "loss": 1.5215,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8468690183210281,
+      "grad_norm": 0.1621725708246231,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8471424665025977,
+      "grad_norm": 0.15811792016029358,
+      "learning_rate": 5e-05,
+      "loss": 1.5293,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8474159146841673,
+      "grad_norm": 0.17002242803573608,
+      "learning_rate": 5e-05,
+      "loss": 1.5781,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8476893628657369,
+      "grad_norm": 0.16223548352718353,
+      "learning_rate": 5e-05,
+      "loss": 1.5776,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8479628110473065,
+      "grad_norm": 0.17608049511909485,
+      "learning_rate": 5e-05,
+      "loss": 1.5716,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8482362592288761,
+      "grad_norm": 0.16243280470371246,
+      "learning_rate": 5e-05,
+      "loss": 1.582,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8485097074104457,
+      "grad_norm": 0.1691652089357376,
+      "learning_rate": 5e-05,
+      "loss": 1.5969,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8487831555920153,
+      "grad_norm": 0.17655591666698456,
+      "learning_rate": 5e-05,
+      "loss": 1.5828,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8490566037735849,
+      "grad_norm": 0.16086623072624207,
+      "learning_rate": 5e-05,
+      "loss": 1.5631,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8493300519551545,
+      "grad_norm": 0.1816483736038208,
+      "learning_rate": 5e-05,
+      "loss": 1.723,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8496035001367241,
+      "grad_norm": 0.17240236699581146,
+      "learning_rate": 5e-05,
+      "loss": 1.669,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8498769483182936,
+      "grad_norm": 0.16476856172084808,
+      "learning_rate": 5e-05,
+      "loss": 1.5872,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8501503964998632,
+      "grad_norm": 0.15981562435626984,
+      "learning_rate": 5e-05,
+      "loss": 1.5647,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8504238446814328,
+      "grad_norm": 0.1652696579694748,
+      "learning_rate": 5e-05,
+      "loss": 1.5882,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8506972928630024,
+      "grad_norm": 0.16290616989135742,
+      "learning_rate": 5e-05,
+      "loss": 1.6325,
+      "step": 3111
+    },
+    {
+      "epoch": 0.850970741044572,
+      "grad_norm": 0.16608236730098724,
+      "learning_rate": 5e-05,
+      "loss": 1.6638,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8512441892261416,
+      "grad_norm": 0.16186489164829254,
+      "learning_rate": 5e-05,
+      "loss": 1.5525,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8515176374077112,
+      "grad_norm": 0.1654919683933258,
+      "learning_rate": 5e-05,
+      "loss": 1.5305,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8517910855892808,
+      "grad_norm": 0.16424906253814697,
+      "learning_rate": 5e-05,
+      "loss": 1.5848,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8520645337708505,
+      "grad_norm": 0.16823723912239075,
+      "learning_rate": 5e-05,
+      "loss": 1.5737,
+      "step": 3116
+    },
+    {
+      "epoch": 0.85233798195242,
+      "grad_norm": 0.16373160481452942,
+      "learning_rate": 5e-05,
+      "loss": 1.6274,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8526114301339897,
+      "grad_norm": 0.18197789788246155,
+      "learning_rate": 5e-05,
+      "loss": 1.6227,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8528848783155591,
+      "grad_norm": 0.15607775747776031,
+      "learning_rate": 5e-05,
+      "loss": 1.5371,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8531583264971287,
+      "grad_norm": 0.16798537969589233,
+      "learning_rate": 5e-05,
+      "loss": 1.5632,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8534317746786984,
+      "grad_norm": 0.164012148976326,
+      "learning_rate": 5e-05,
+      "loss": 1.5629,
+      "step": 3121
+    },
+    {
+      "epoch": 0.853705222860268,
+      "grad_norm": 0.176783949136734,
+      "learning_rate": 5e-05,
+      "loss": 1.5977,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8539786710418376,
+      "grad_norm": 0.15592290461063385,
+      "learning_rate": 5e-05,
+      "loss": 1.5529,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8542521192234072,
+      "grad_norm": 0.18195119500160217,
+      "learning_rate": 5e-05,
+      "loss": 1.6908,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8545255674049768,
+      "grad_norm": 0.174040749669075,
+      "learning_rate": 5e-05,
+      "loss": 1.5646,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8547990155865464,
+      "grad_norm": 0.16298840939998627,
+      "learning_rate": 5e-05,
+      "loss": 1.5824,
+      "step": 3126
+    },
+    {
+      "epoch": 0.855072463768116,
+      "grad_norm": 0.16243524849414825,
+      "learning_rate": 5e-05,
+      "loss": 1.603,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8553459119496856,
+      "grad_norm": 0.16204892098903656,
+      "learning_rate": 5e-05,
+      "loss": 1.595,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8556193601312552,
+      "grad_norm": 0.16708557307720184,
+      "learning_rate": 5e-05,
+      "loss": 1.6143,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8558928083128248,
+      "grad_norm": 0.1747390627861023,
+      "learning_rate": 5e-05,
+      "loss": 1.5863,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8561662564943943,
+      "grad_norm": 0.17126700282096863,
+      "learning_rate": 5e-05,
+      "loss": 1.6091,
+      "step": 3131
+    },
+    {
+      "epoch": 0.8564397046759639,
+      "grad_norm": 0.1731352061033249,
+      "learning_rate": 5e-05,
+      "loss": 1.6073,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8567131528575335,
+      "grad_norm": 0.17451922595500946,
+      "learning_rate": 5e-05,
+      "loss": 1.7312,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8569866010391031,
+      "grad_norm": 0.156746968626976,
+      "learning_rate": 5e-05,
+      "loss": 1.4434,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8572600492206727,
+      "grad_norm": 0.1696559190750122,
+      "learning_rate": 5e-05,
+      "loss": 1.6185,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8575334974022423,
+      "grad_norm": 0.1689455807209015,
+      "learning_rate": 5e-05,
+      "loss": 1.5066,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8578069455838119,
+      "grad_norm": 0.16820566356182098,
+      "learning_rate": 5e-05,
+      "loss": 1.5294,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8580803937653815,
+      "grad_norm": 0.18438181281089783,
+      "learning_rate": 5e-05,
+      "loss": 1.6513,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8583538419469511,
+      "grad_norm": 0.1611601859331131,
+      "learning_rate": 5e-05,
+      "loss": 1.4604,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8586272901285207,
+      "grad_norm": 0.16456107795238495,
+      "learning_rate": 5e-05,
+      "loss": 1.5664,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8589007383100903,
+      "grad_norm": 0.16825945675373077,
+      "learning_rate": 5e-05,
+      "loss": 1.5777,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8591741864916599,
+      "grad_norm": 0.16525815427303314,
+      "learning_rate": 5e-05,
+      "loss": 1.5464,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8594476346732294,
+      "grad_norm": 0.17219798266887665,
+      "learning_rate": 5e-05,
+      "loss": 1.6255,
+      "step": 3143
+    },
+    {
+      "epoch": 0.859721082854799,
+      "grad_norm": 0.1633169800043106,
+      "learning_rate": 5e-05,
+      "loss": 1.5723,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8599945310363686,
+      "grad_norm": 0.15741117298603058,
+      "learning_rate": 5e-05,
+      "loss": 1.5098,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8602679792179382,
+      "grad_norm": 0.17558981478214264,
+      "learning_rate": 5e-05,
+      "loss": 1.6152,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8605414273995078,
+      "grad_norm": 0.16872353851795197,
+      "learning_rate": 5e-05,
+      "loss": 1.6884,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8608148755810774,
+      "grad_norm": 0.18673337996006012,
+      "learning_rate": 5e-05,
+      "loss": 1.6064,
+      "step": 3148
+    },
+    {
+      "epoch": 0.861088323762647,
+      "grad_norm": 0.16644282639026642,
+      "learning_rate": 5e-05,
+      "loss": 1.518,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8613617719442166,
+      "grad_norm": 0.16721606254577637,
+      "learning_rate": 5e-05,
+      "loss": 1.5891,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8616352201257862,
+      "grad_norm": 0.17002660036087036,
+      "learning_rate": 5e-05,
+      "loss": 1.5312,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8619086683073558,
+      "grad_norm": 0.16748939454555511,
+      "learning_rate": 5e-05,
+      "loss": 1.6485,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8621821164889254,
+      "grad_norm": 0.15232722461223602,
+      "learning_rate": 5e-05,
+      "loss": 1.4527,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8624555646704949,
+      "grad_norm": 0.16974018514156342,
+      "learning_rate": 5e-05,
+      "loss": 1.5227,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8627290128520645,
+      "grad_norm": 0.16853350400924683,
+      "learning_rate": 5e-05,
+      "loss": 1.5233,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8630024610336341,
+      "grad_norm": 0.16759264469146729,
+      "learning_rate": 5e-05,
+      "loss": 1.6178,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8632759092152037,
+      "grad_norm": 0.18532946705818176,
+      "learning_rate": 5e-05,
+      "loss": 1.4664,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8635493573967733,
+      "grad_norm": 0.16451618075370789,
+      "learning_rate": 5e-05,
+      "loss": 1.5393,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8638228055783429,
+      "grad_norm": 0.17090554535388947,
+      "learning_rate": 5e-05,
+      "loss": 1.6055,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8640962537599125,
+      "grad_norm": 0.16245882213115692,
+      "learning_rate": 5e-05,
+      "loss": 1.5864,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8643697019414821,
+      "grad_norm": 0.16497839987277985,
+      "learning_rate": 5e-05,
+      "loss": 1.5725,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8646431501230517,
+      "grad_norm": 0.1678629368543625,
+      "learning_rate": 5e-05,
+      "loss": 1.6436,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8649165983046213,
+      "grad_norm": 0.17666424810886383,
+      "learning_rate": 5e-05,
+      "loss": 1.5639,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8651900464861909,
+      "grad_norm": 0.16444604098796844,
+      "learning_rate": 5e-05,
+      "loss": 1.6184,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8654634946677605,
+      "grad_norm": 0.1679319143295288,
+      "learning_rate": 5e-05,
+      "loss": 1.5206,
+      "step": 3165
+    },
+    {
+      "epoch": 0.86573694284933,
+      "grad_norm": 0.16246174275875092,
+      "learning_rate": 5e-05,
+      "loss": 1.5794,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8660103910308996,
+      "grad_norm": 0.17036394774913788,
+      "learning_rate": 5e-05,
+      "loss": 1.6325,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8662838392124692,
+      "grad_norm": 0.1656508892774582,
+      "learning_rate": 5e-05,
+      "loss": 1.5285,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8665572873940388,
+      "grad_norm": 0.17411690950393677,
+      "learning_rate": 5e-05,
+      "loss": 1.5536,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8668307355756084,
+      "grad_norm": 0.1696336418390274,
+      "learning_rate": 5e-05,
+      "loss": 1.5778,
+      "step": 3170
+    },
+    {
+      "epoch": 0.867104183757178,
+      "grad_norm": 0.17345334589481354,
+      "learning_rate": 5e-05,
+      "loss": 1.5805,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8673776319387476,
+      "grad_norm": 0.1722104400396347,
+      "learning_rate": 5e-05,
+      "loss": 1.6734,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8676510801203172,
+      "grad_norm": 0.17595550417900085,
+      "learning_rate": 5e-05,
+      "loss": 1.6609,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.1680113971233368,
+      "learning_rate": 5e-05,
+      "loss": 1.6642,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8681979764834564,
+      "grad_norm": 0.16808103024959564,
+      "learning_rate": 5e-05,
+      "loss": 1.5834,
+      "step": 3175
+    },
+    {
+      "epoch": 0.868471424665026,
+      "grad_norm": 0.16665758192539215,
+      "learning_rate": 5e-05,
+      "loss": 1.6229,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8687448728465955,
+      "grad_norm": 0.16951392590999603,
+      "learning_rate": 5e-05,
+      "loss": 1.5884,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8690183210281651,
+      "grad_norm": 0.16685131192207336,
+      "learning_rate": 5e-05,
+      "loss": 1.632,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8692917692097347,
+      "grad_norm": 0.1711168885231018,
+      "learning_rate": 5e-05,
+      "loss": 1.6524,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.16593927145004272,
+      "learning_rate": 5e-05,
+      "loss": 1.6038,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8698386655728739,
+      "grad_norm": 0.16577638685703278,
+      "learning_rate": 5e-05,
+      "loss": 1.6469,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8701121137544435,
+      "grad_norm": 0.17342355847358704,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8703855619360131,
+      "grad_norm": 0.15923261642456055,
+      "learning_rate": 5e-05,
+      "loss": 1.5506,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8706590101175827,
+      "grad_norm": 0.18326669931411743,
+      "learning_rate": 5e-05,
+      "loss": 1.5866,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8709324582991523,
+      "grad_norm": 0.17581258714199066,
+      "learning_rate": 5e-05,
+      "loss": 1.5521,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8712059064807219,
+      "grad_norm": 0.1683703362941742,
+      "learning_rate": 5e-05,
+      "loss": 1.6314,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8714793546622915,
+      "grad_norm": 0.17400458455085754,
+      "learning_rate": 5e-05,
+      "loss": 1.586,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8717528028438611,
+      "grad_norm": 0.1864955574274063,
+      "learning_rate": 5e-05,
+      "loss": 1.6241,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8720262510254306,
+      "grad_norm": 0.16646967828273773,
+      "learning_rate": 5e-05,
+      "loss": 1.568,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8722996992070002,
+      "grad_norm": 0.17674346268177032,
+      "learning_rate": 5e-05,
+      "loss": 1.6457,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8725731473885698,
+      "grad_norm": 0.18272368609905243,
+      "learning_rate": 5e-05,
+      "loss": 1.6196,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8728465955701394,
+      "grad_norm": 0.16906492412090302,
+      "learning_rate": 5e-05,
+      "loss": 1.5682,
+      "step": 3192
+    },
+    {
+      "epoch": 0.873120043751709,
+      "grad_norm": 0.16809342801570892,
+      "learning_rate": 5e-05,
+      "loss": 1.5799,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8733934919332786,
+      "grad_norm": 0.1753772646188736,
+      "learning_rate": 5e-05,
+      "loss": 1.5376,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8736669401148482,
+      "grad_norm": 0.16110126674175262,
+      "learning_rate": 5e-05,
+      "loss": 1.6114,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8739403882964178,
+      "grad_norm": 0.16913153231143951,
+      "learning_rate": 5e-05,
+      "loss": 1.5882,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8742138364779874,
+      "grad_norm": 0.17413096129894257,
+      "learning_rate": 5e-05,
+      "loss": 1.577,
+      "step": 3197
+    },
+    {
+      "epoch": 0.874487284659557,
+      "grad_norm": 0.16093304753303528,
+      "learning_rate": 5e-05,
+      "loss": 1.5553,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8747607328411267,
+      "grad_norm": 0.1645004004240036,
+      "learning_rate": 5e-05,
+      "loss": 1.5537,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8750341810226961,
+      "grad_norm": 0.1645384430885315,
+      "learning_rate": 5e-05,
+      "loss": 1.629,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8753076292042657,
+      "grad_norm": 0.17203739285469055,
+      "learning_rate": 5e-05,
+      "loss": 1.6529,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8755810773858353,
+      "grad_norm": 0.1711164116859436,
+      "learning_rate": 5e-05,
+      "loss": 1.63,
+      "step": 3202
+    },
+    {
+      "epoch": 0.875854525567405,
+      "grad_norm": 0.17426781356334686,
+      "learning_rate": 5e-05,
+      "loss": 1.7013,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8761279737489746,
+      "grad_norm": 0.1662566065788269,
+      "learning_rate": 5e-05,
+      "loss": 1.5933,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8764014219305442,
+      "grad_norm": 0.17674902081489563,
+      "learning_rate": 5e-05,
+      "loss": 1.5578,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8766748701121138,
+      "grad_norm": 0.15999096632003784,
+      "learning_rate": 5e-05,
+      "loss": 1.6195,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8769483182936834,
+      "grad_norm": 0.16697584092617035,
+      "learning_rate": 5e-05,
+      "loss": 1.589,
+      "step": 3207
+    },
+    {
+      "epoch": 0.877221766475253,
+      "grad_norm": 0.16489756107330322,
+      "learning_rate": 5e-05,
+      "loss": 1.6149,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8774952146568226,
+      "grad_norm": 0.16519208252429962,
+      "learning_rate": 5e-05,
+      "loss": 1.5,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8777686628383922,
+      "grad_norm": 0.17452332377433777,
+      "learning_rate": 5e-05,
+      "loss": 1.6418,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8780421110199618,
+      "grad_norm": 0.16637583076953888,
+      "learning_rate": 5e-05,
+      "loss": 1.5775,
+      "step": 3211
+    },
+    {
+      "epoch": 0.8783155592015313,
+      "grad_norm": 0.15892595052719116,
+      "learning_rate": 5e-05,
+      "loss": 1.498,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8785890073831009,
+      "grad_norm": 0.17607276141643524,
+      "learning_rate": 5e-05,
+      "loss": 1.6636,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8788624555646705,
+      "grad_norm": 0.1674719601869583,
+      "learning_rate": 5e-05,
+      "loss": 1.6106,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8791359037462401,
+      "grad_norm": 0.17102982103824615,
+      "learning_rate": 5e-05,
+      "loss": 1.6174,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8794093519278097,
+      "grad_norm": 0.16860932111740112,
+      "learning_rate": 5e-05,
+      "loss": 1.5382,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8796828001093793,
+      "grad_norm": 0.15637625753879547,
+      "learning_rate": 5e-05,
+      "loss": 1.5326,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8799562482909489,
+      "grad_norm": 0.16792798042297363,
+      "learning_rate": 5e-05,
+      "loss": 1.6486,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8802296964725185,
+      "grad_norm": 0.17826606333255768,
+      "learning_rate": 5e-05,
+      "loss": 1.5327,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8805031446540881,
+      "grad_norm": 0.1626116782426834,
+      "learning_rate": 5e-05,
+      "loss": 1.5084,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8807765928356577,
+      "grad_norm": 0.16171365976333618,
+      "learning_rate": 5e-05,
+      "loss": 1.559,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8810500410172273,
+      "grad_norm": 0.16499905288219452,
+      "learning_rate": 5e-05,
+      "loss": 1.4079,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8813234891987969,
+      "grad_norm": 0.16512024402618408,
+      "learning_rate": 5e-05,
+      "loss": 1.5858,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8815969373803664,
+      "grad_norm": 0.1706002801656723,
+      "learning_rate": 5e-05,
+      "loss": 1.6038,
+      "step": 3224
+    },
+    {
+      "epoch": 0.881870385561936,
+      "grad_norm": 0.17629894614219666,
+      "learning_rate": 5e-05,
+      "loss": 1.6537,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8821438337435056,
+      "grad_norm": 0.16706202924251556,
+      "learning_rate": 5e-05,
+      "loss": 1.5556,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8824172819250752,
+      "grad_norm": 0.1704121083021164,
+      "learning_rate": 5e-05,
+      "loss": 1.6341,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8826907301066448,
+      "grad_norm": 0.17939800024032593,
+      "learning_rate": 5e-05,
+      "loss": 1.4981,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8829641782882144,
+      "grad_norm": 0.17186492681503296,
+      "learning_rate": 5e-05,
+      "loss": 1.5591,
+      "step": 3229
+    },
+    {
+      "epoch": 0.883237626469784,
+      "grad_norm": 0.160826176404953,
+      "learning_rate": 5e-05,
+      "loss": 1.5478,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8835110746513536,
+      "grad_norm": 0.190667062997818,
+      "learning_rate": 5e-05,
+      "loss": 1.5104,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8837845228329232,
+      "grad_norm": 0.18037305772304535,
+      "learning_rate": 5e-05,
+      "loss": 1.6009,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8840579710144928,
+      "grad_norm": 0.15924717485904694,
+      "learning_rate": 5e-05,
+      "loss": 1.5045,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8843314191960624,
+      "grad_norm": 0.19811680912971497,
+      "learning_rate": 5e-05,
+      "loss": 1.5899,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8846048673776319,
+      "grad_norm": 0.1798917055130005,
+      "learning_rate": 5e-05,
+      "loss": 1.5403,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8848783155592015,
+      "grad_norm": 0.17577169835567474,
+      "learning_rate": 5e-05,
+      "loss": 1.5191,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8851517637407711,
+      "grad_norm": 0.18266037106513977,
+      "learning_rate": 5e-05,
+      "loss": 1.5779,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8854252119223407,
+      "grad_norm": 0.1692240834236145,
+      "learning_rate": 5e-05,
+      "loss": 1.5353,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8856986601039103,
+      "grad_norm": 0.1703694760799408,
+      "learning_rate": 5e-05,
+      "loss": 1.5582,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8859721082854799,
+      "grad_norm": 0.1754840612411499,
+      "learning_rate": 5e-05,
+      "loss": 1.5573,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8862455564670495,
+      "grad_norm": 0.16203297674655914,
+      "learning_rate": 5e-05,
+      "loss": 1.5276,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8865190046486191,
+      "grad_norm": 0.15817001461982727,
+      "learning_rate": 5e-05,
+      "loss": 1.583,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8867924528301887,
+      "grad_norm": 0.1767439991235733,
+      "learning_rate": 5e-05,
+      "loss": 1.6644,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8870659010117583,
+      "grad_norm": 0.16701658070087433,
+      "learning_rate": 5e-05,
+      "loss": 1.5825,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8873393491933279,
+      "grad_norm": 0.16390787065029144,
+      "learning_rate": 5e-05,
+      "loss": 1.6174,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8876127973748975,
+      "grad_norm": 0.1741890013217926,
+      "learning_rate": 5e-05,
+      "loss": 1.5714,
+      "step": 3246
+    },
+    {
+      "epoch": 0.887886245556467,
+      "grad_norm": 0.1701374500989914,
+      "learning_rate": 5e-05,
+      "loss": 1.6041,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8881596937380366,
+      "grad_norm": 0.18417386710643768,
+      "learning_rate": 5e-05,
+      "loss": 1.6007,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8884331419196062,
+      "grad_norm": 0.18053127825260162,
+      "learning_rate": 5e-05,
+      "loss": 1.5415,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8887065901011758,
+      "grad_norm": 0.16399511694908142,
+      "learning_rate": 5e-05,
+      "loss": 1.5805,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8889800382827454,
+      "grad_norm": 0.1606212705373764,
+      "learning_rate": 5e-05,
+      "loss": 1.5433,
+      "step": 3251
+    },
+    {
+      "epoch": 0.889253486464315,
+      "grad_norm": 0.16091781854629517,
+      "learning_rate": 5e-05,
+      "loss": 1.5801,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8895269346458846,
+      "grad_norm": 0.17079661786556244,
+      "learning_rate": 5e-05,
+      "loss": 1.6032,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8898003828274542,
+      "grad_norm": 0.16236819326877594,
+      "learning_rate": 5e-05,
+      "loss": 1.5309,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8900738310090238,
+      "grad_norm": 0.16366977989673615,
+      "learning_rate": 5e-05,
+      "loss": 1.4854,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8903472791905934,
+      "grad_norm": 0.1692305952310562,
+      "learning_rate": 5e-05,
+      "loss": 1.5546,
+      "step": 3256
+    },
+    {
+      "epoch": 0.890620727372163,
+      "grad_norm": 0.1591336876153946,
+      "learning_rate": 5e-05,
+      "loss": 1.5724,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8908941755537325,
+      "grad_norm": 0.1661178022623062,
+      "learning_rate": 5e-05,
+      "loss": 1.6642,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8911676237353021,
+      "grad_norm": 0.1741829365491867,
+      "learning_rate": 5e-05,
+      "loss": 1.5808,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8914410719168717,
+      "grad_norm": 0.16046129167079926,
+      "learning_rate": 5e-05,
+      "loss": 1.6136,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8917145200984413,
+      "grad_norm": 0.16220787167549133,
+      "learning_rate": 5e-05,
+      "loss": 1.5145,
+      "step": 3261
+    },
+    {
+      "epoch": 0.8919879682800109,
+      "grad_norm": 0.17300353944301605,
+      "learning_rate": 5e-05,
+      "loss": 1.5805,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8922614164615805,
+      "grad_norm": 0.1636408567428589,
+      "learning_rate": 5e-05,
+      "loss": 1.5679,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8925348646431501,
+      "grad_norm": 0.1789669394493103,
+      "learning_rate": 5e-05,
+      "loss": 1.7185,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8928083128247197,
+      "grad_norm": 0.161798357963562,
+      "learning_rate": 5e-05,
+      "loss": 1.4734,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8930817610062893,
+      "grad_norm": 0.1663116067647934,
+      "learning_rate": 5e-05,
+      "loss": 1.559,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8933552091878589,
+      "grad_norm": 0.16649827361106873,
+      "learning_rate": 5e-05,
+      "loss": 1.5498,
+      "step": 3267
+    },
+    {
+      "epoch": 0.8936286573694285,
+      "grad_norm": 0.1723203808069229,
+      "learning_rate": 5e-05,
+      "loss": 1.623,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8939021055509981,
+      "grad_norm": 0.16643695533275604,
+      "learning_rate": 5e-05,
+      "loss": 1.6505,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8941755537325676,
+      "grad_norm": 0.16589291393756866,
+      "learning_rate": 5e-05,
+      "loss": 1.6633,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8944490019141372,
+      "grad_norm": 0.1624186635017395,
+      "learning_rate": 5e-05,
+      "loss": 1.5022,
+      "step": 3271
+    },
+    {
+      "epoch": 0.8947224500957068,
+      "grad_norm": 0.1685740202665329,
+      "learning_rate": 5e-05,
+      "loss": 1.6225,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8949958982772764,
+      "grad_norm": 0.16466566920280457,
+      "learning_rate": 5e-05,
+      "loss": 1.6597,
+      "step": 3273
+    },
+    {
+      "epoch": 0.895269346458846,
+      "grad_norm": 0.16794006526470184,
+      "learning_rate": 5e-05,
+      "loss": 1.5795,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8955427946404156,
+      "grad_norm": 0.1675807386636734,
+      "learning_rate": 5e-05,
+      "loss": 1.6751,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8958162428219852,
+      "grad_norm": 0.1637728214263916,
+      "learning_rate": 5e-05,
+      "loss": 1.5746,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8960896910035548,
+      "grad_norm": 0.16383953392505646,
+      "learning_rate": 5e-05,
+      "loss": 1.5741,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8963631391851244,
+      "grad_norm": 0.16199614107608795,
+      "learning_rate": 5e-05,
+      "loss": 1.5629,
+      "step": 3278
+    },
+    {
+      "epoch": 0.896636587366694,
+      "grad_norm": 0.17193222045898438,
+      "learning_rate": 5e-05,
+      "loss": 1.6165,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8969100355482637,
+      "grad_norm": 0.16081291437149048,
+      "learning_rate": 5e-05,
+      "loss": 1.5044,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8971834837298331,
+      "grad_norm": 0.16551733016967773,
+      "learning_rate": 5e-05,
+      "loss": 1.6052,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8974569319114027,
+      "grad_norm": 0.1634800136089325,
+      "learning_rate": 5e-05,
+      "loss": 1.5435,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8977303800929723,
+      "grad_norm": 0.17315998673439026,
+      "learning_rate": 5e-05,
+      "loss": 1.6515,
+      "step": 3283
+    },
+    {
+      "epoch": 0.898003828274542,
+      "grad_norm": 0.17361579835414886,
+      "learning_rate": 5e-05,
+      "loss": 1.6037,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8982772764561116,
+      "grad_norm": 0.17537641525268555,
+      "learning_rate": 5e-05,
+      "loss": 1.6171,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8985507246376812,
+      "grad_norm": 0.17007243633270264,
+      "learning_rate": 5e-05,
+      "loss": 1.6756,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8988241728192508,
+      "grad_norm": 0.1736474186182022,
+      "learning_rate": 5e-05,
+      "loss": 1.5852,
+      "step": 3287
+    },
+    {
+      "epoch": 0.8990976210008204,
+      "grad_norm": 0.1645202934741974,
+      "learning_rate": 5e-05,
+      "loss": 1.6534,
+      "step": 3288
+    },
+    {
+      "epoch": 0.89937106918239,
+      "grad_norm": 0.16424302756786346,
+      "learning_rate": 5e-05,
+      "loss": 1.5194,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8996445173639596,
+      "grad_norm": 0.17903803288936615,
+      "learning_rate": 5e-05,
+      "loss": 1.5683,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8999179655455292,
+      "grad_norm": 0.16927549242973328,
+      "learning_rate": 5e-05,
+      "loss": 1.6283,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9001914137270988,
+      "grad_norm": 0.16738973557949066,
+      "learning_rate": 5e-05,
+      "loss": 1.6196,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9004648619086683,
+      "grad_norm": 0.16482320427894592,
+      "learning_rate": 5e-05,
+      "loss": 1.5135,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9007383100902379,
+      "grad_norm": 0.1613280475139618,
+      "learning_rate": 5e-05,
+      "loss": 1.6036,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9010117582718075,
+      "grad_norm": 0.16160857677459717,
+      "learning_rate": 5e-05,
+      "loss": 1.5348,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9012852064533771,
+      "grad_norm": 0.1576218158006668,
+      "learning_rate": 5e-05,
+      "loss": 1.5171,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9015586546349467,
+      "grad_norm": 0.1682513952255249,
+      "learning_rate": 5e-05,
+      "loss": 1.5429,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9018321028165163,
+      "grad_norm": 0.16176675260066986,
+      "learning_rate": 5e-05,
+      "loss": 1.5041,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9021055509980859,
+      "grad_norm": 0.16141022741794586,
+      "learning_rate": 5e-05,
+      "loss": 1.5655,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9023789991796555,
+      "grad_norm": 0.16335152089595795,
+      "learning_rate": 5e-05,
+      "loss": 1.6064,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9026524473612251,
+      "grad_norm": 0.16691720485687256,
+      "learning_rate": 5e-05,
+      "loss": 1.6316,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9029258955427947,
+      "grad_norm": 0.16636385023593903,
+      "learning_rate": 5e-05,
+      "loss": 1.59,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9031993437243643,
+      "grad_norm": 0.1710832118988037,
+      "learning_rate": 5e-05,
+      "loss": 1.5199,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9034727919059339,
+      "grad_norm": 0.169814333319664,
+      "learning_rate": 5e-05,
+      "loss": 1.5754,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9037462400875034,
+      "grad_norm": 0.1706944704055786,
+      "learning_rate": 5e-05,
+      "loss": 1.6101,
+      "step": 3305
+    },
+    {
+      "epoch": 0.904019688269073,
+      "grad_norm": 0.15784287452697754,
+      "learning_rate": 5e-05,
+      "loss": 1.5232,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9042931364506426,
+      "grad_norm": 0.1703675389289856,
+      "learning_rate": 5e-05,
+      "loss": 1.6001,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9045665846322122,
+      "grad_norm": 0.16346101462841034,
+      "learning_rate": 5e-05,
+      "loss": 1.5885,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9048400328137818,
+      "grad_norm": 0.18370665609836578,
+      "learning_rate": 5e-05,
+      "loss": 1.6754,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9051134809953514,
+      "grad_norm": 0.16520695388317108,
+      "learning_rate": 5e-05,
+      "loss": 1.5961,
+      "step": 3310
+    },
+    {
+      "epoch": 0.905386929176921,
+      "grad_norm": 0.17497234046459198,
+      "learning_rate": 5e-05,
+      "loss": 1.5912,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.16902846097946167,
+      "learning_rate": 5e-05,
+      "loss": 1.6057,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9059338255400602,
+      "grad_norm": 0.16770271956920624,
+      "learning_rate": 5e-05,
+      "loss": 1.6274,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9062072737216298,
+      "grad_norm": 0.16241686046123505,
+      "learning_rate": 5e-05,
+      "loss": 1.5494,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9064807219031994,
+      "grad_norm": 0.16672220826148987,
+      "learning_rate": 5e-05,
+      "loss": 1.6654,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9067541700847689,
+      "grad_norm": 0.1626524180173874,
+      "learning_rate": 5e-05,
+      "loss": 1.5578,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9070276182663385,
+      "grad_norm": 0.16794568300247192,
+      "learning_rate": 5e-05,
+      "loss": 1.6463,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9073010664479081,
+      "grad_norm": 0.16510622203350067,
+      "learning_rate": 5e-05,
+      "loss": 1.597,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9075745146294777,
+      "grad_norm": 0.16373200714588165,
+      "learning_rate": 5e-05,
+      "loss": 1.6206,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9078479628110473,
+      "grad_norm": 0.16739732027053833,
+      "learning_rate": 5e-05,
+      "loss": 1.6041,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9081214109926169,
+      "grad_norm": 0.17469698190689087,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9083948591741865,
+      "grad_norm": 0.1681319773197174,
+      "learning_rate": 5e-05,
+      "loss": 1.5485,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9086683073557561,
+      "grad_norm": 0.1650519222021103,
+      "learning_rate": 5e-05,
+      "loss": 1.5211,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9089417555373257,
+      "grad_norm": 0.167753666639328,
+      "learning_rate": 5e-05,
+      "loss": 1.6338,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9092152037188953,
+      "grad_norm": 0.164119154214859,
+      "learning_rate": 5e-05,
+      "loss": 1.5542,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9094886519004649,
+      "grad_norm": 0.16457271575927734,
+      "learning_rate": 5e-05,
+      "loss": 1.6337,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9097621000820345,
+      "grad_norm": 0.17210492491722107,
+      "learning_rate": 5e-05,
+      "loss": 1.6361,
+      "step": 3327
+    },
+    {
+      "epoch": 0.910035548263604,
+      "grad_norm": 0.17178331315517426,
+      "learning_rate": 5e-05,
+      "loss": 1.6849,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9103089964451736,
+      "grad_norm": 0.1700570285320282,
+      "learning_rate": 5e-05,
+      "loss": 1.6171,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9105824446267432,
+      "grad_norm": 0.159864142537117,
+      "learning_rate": 5e-05,
+      "loss": 1.5302,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9108558928083128,
+      "grad_norm": 0.16526418924331665,
+      "learning_rate": 5e-05,
+      "loss": 1.5488,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9111293409898824,
+      "grad_norm": 0.16879980266094208,
+      "learning_rate": 5e-05,
+      "loss": 1.6115,
+      "step": 3332
+    },
+    {
+      "epoch": 0.911402789171452,
+      "grad_norm": 0.1696682572364807,
+      "learning_rate": 5e-05,
+      "loss": 1.6198,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9116762373530216,
+      "grad_norm": 0.1723857969045639,
+      "learning_rate": 5e-05,
+      "loss": 1.4767,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9119496855345912,
+      "grad_norm": 0.16016238927841187,
+      "learning_rate": 5e-05,
+      "loss": 1.5293,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9122231337161608,
+      "grad_norm": 0.1660720705986023,
+      "learning_rate": 5e-05,
+      "loss": 1.5294,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9124965818977304,
+      "grad_norm": 0.16862016916275024,
+      "learning_rate": 5e-05,
+      "loss": 1.5987,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9127700300793,
+      "grad_norm": 0.17382635176181793,
+      "learning_rate": 5e-05,
+      "loss": 1.6834,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.18816977739334106,
+      "learning_rate": 5e-05,
+      "loss": 1.5269,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9133169264424391,
+      "grad_norm": 0.17092393338680267,
+      "learning_rate": 5e-05,
+      "loss": 1.5092,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9135903746240087,
+      "grad_norm": 0.1623528152704239,
+      "learning_rate": 5e-05,
+      "loss": 1.4633,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9138638228055783,
+      "grad_norm": 0.16653479635715485,
+      "learning_rate": 5e-05,
+      "loss": 1.5334,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9141372709871479,
+      "grad_norm": 0.1750710904598236,
+      "learning_rate": 5e-05,
+      "loss": 1.6056,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9144107191687175,
+      "grad_norm": 0.1836736500263214,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9146841673502871,
+      "grad_norm": 0.1607818603515625,
+      "learning_rate": 5e-05,
+      "loss": 1.5236,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9149576155318567,
+      "grad_norm": 0.1604374796152115,
+      "learning_rate": 5e-05,
+      "loss": 1.4829,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9152310637134263,
+      "grad_norm": 0.16865043342113495,
+      "learning_rate": 5e-05,
+      "loss": 1.6105,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9155045118949959,
+      "grad_norm": 0.17235668003559113,
+      "learning_rate": 5e-05,
+      "loss": 1.6007,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9157779600765655,
+      "grad_norm": 0.16054973006248474,
+      "learning_rate": 5e-05,
+      "loss": 1.5466,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9160514082581351,
+      "grad_norm": 0.1764586865901947,
+      "learning_rate": 5e-05,
+      "loss": 1.6551,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9163248564397046,
+      "grad_norm": 0.15826088190078735,
+      "learning_rate": 5e-05,
+      "loss": 1.538,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9165983046212742,
+      "grad_norm": 0.16789792478084564,
+      "learning_rate": 5e-05,
+      "loss": 1.5845,
+      "step": 3352
+    },
+    {
+      "epoch": 0.9168717528028438,
+      "grad_norm": 0.1768852025270462,
+      "learning_rate": 5e-05,
+      "loss": 1.5847,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9171452009844134,
+      "grad_norm": 0.16304756700992584,
+      "learning_rate": 5e-05,
+      "loss": 1.5508,
+      "step": 3354
+    },
+    {
+      "epoch": 0.917418649165983,
+      "grad_norm": 0.16250407695770264,
+      "learning_rate": 5e-05,
+      "loss": 1.5004,
+      "step": 3355
+    },
+    {
+      "epoch": 0.9176920973475526,
+      "grad_norm": 0.18279612064361572,
+      "learning_rate": 5e-05,
+      "loss": 1.6446,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9179655455291222,
+      "grad_norm": 0.16867072880268097,
+      "learning_rate": 5e-05,
+      "loss": 1.5235,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9182389937106918,
+      "grad_norm": 0.16388383507728577,
+      "learning_rate": 5e-05,
+      "loss": 1.5635,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9185124418922614,
+      "grad_norm": 0.16723370552062988,
+      "learning_rate": 5e-05,
+      "loss": 1.6193,
+      "step": 3359
+    },
+    {
+      "epoch": 0.918785890073831,
+      "grad_norm": 0.1766999363899231,
+      "learning_rate": 5e-05,
+      "loss": 1.6716,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9190593382554006,
+      "grad_norm": 0.16779550909996033,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9193327864369701,
+      "grad_norm": 0.16861297190189362,
+      "learning_rate": 5e-05,
+      "loss": 1.6456,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9196062346185397,
+      "grad_norm": 0.18011167645454407,
+      "learning_rate": 5e-05,
+      "loss": 1.6536,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9198796828001093,
+      "grad_norm": 0.1704559475183487,
+      "learning_rate": 5e-05,
+      "loss": 1.6436,
+      "step": 3364
+    },
+    {
+      "epoch": 0.920153130981679,
+      "grad_norm": 0.16970516741275787,
+      "learning_rate": 5e-05,
+      "loss": 1.5132,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9204265791632485,
+      "grad_norm": 0.1771494746208191,
+      "learning_rate": 5e-05,
+      "loss": 1.569,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9207000273448182,
+      "grad_norm": 0.17308154702186584,
+      "learning_rate": 5e-05,
+      "loss": 1.6102,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9209734755263878,
+      "grad_norm": 0.1670798808336258,
+      "learning_rate": 5e-05,
+      "loss": 1.5677,
+      "step": 3368
+    },
+    {
+      "epoch": 0.9212469237079574,
+      "grad_norm": 0.1702473908662796,
+      "learning_rate": 5e-05,
+      "loss": 1.56,
+      "step": 3369
+    },
+    {
+      "epoch": 0.921520371889527,
+      "grad_norm": 0.1635974645614624,
+      "learning_rate": 5e-05,
+      "loss": 1.5166,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9217938200710966,
+      "grad_norm": 0.16810756921768188,
+      "learning_rate": 5e-05,
+      "loss": 1.5151,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9220672682526662,
+      "grad_norm": 0.15942378342151642,
+      "learning_rate": 5e-05,
+      "loss": 1.5737,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9223407164342358,
+      "grad_norm": 0.17249250411987305,
+      "learning_rate": 5e-05,
+      "loss": 1.5527,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9226141646158053,
+      "grad_norm": 0.16102191805839539,
+      "learning_rate": 5e-05,
+      "loss": 1.5104,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9228876127973749,
+      "grad_norm": 0.17065097391605377,
+      "learning_rate": 5e-05,
+      "loss": 1.5749,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9231610609789445,
+      "grad_norm": 0.17183524370193481,
+      "learning_rate": 5e-05,
+      "loss": 1.5149,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9234345091605141,
+      "grad_norm": 0.17372307181358337,
+      "learning_rate": 5e-05,
+      "loss": 1.6725,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9237079573420837,
+      "grad_norm": 0.17314964532852173,
+      "learning_rate": 5e-05,
+      "loss": 1.5872,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9239814055236533,
+      "grad_norm": 0.17013640701770782,
+      "learning_rate": 5e-05,
+      "loss": 1.6004,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9242548537052229,
+      "grad_norm": 0.16398966312408447,
+      "learning_rate": 5e-05,
+      "loss": 1.5362,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9245283018867925,
+      "grad_norm": 0.16314879059791565,
+      "learning_rate": 5e-05,
+      "loss": 1.473,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9248017500683621,
+      "grad_norm": 0.164910227060318,
+      "learning_rate": 5e-05,
+      "loss": 1.5161,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9250751982499317,
+      "grad_norm": 0.16402222216129303,
+      "learning_rate": 5e-05,
+      "loss": 1.5689,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9253486464315013,
+      "grad_norm": 0.1658565104007721,
+      "learning_rate": 5e-05,
+      "loss": 1.5275,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9256220946130709,
+      "grad_norm": 0.1655322015285492,
+      "learning_rate": 5e-05,
+      "loss": 1.5571,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9258955427946404,
+      "grad_norm": 0.16581609845161438,
+      "learning_rate": 5e-05,
+      "loss": 1.6255,
+      "step": 3386
+    },
+    {
+      "epoch": 0.92616899097621,
+      "grad_norm": 0.1806444674730301,
+      "learning_rate": 5e-05,
+      "loss": 1.6025,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9264424391577796,
+      "grad_norm": 0.1640556901693344,
+      "learning_rate": 5e-05,
+      "loss": 1.5615,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9267158873393492,
+      "grad_norm": 0.15630288422107697,
+      "learning_rate": 5e-05,
+      "loss": 1.5421,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9269893355209188,
+      "grad_norm": 0.16430017352104187,
+      "learning_rate": 5e-05,
+      "loss": 1.594,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9272627837024884,
+      "grad_norm": 0.17217886447906494,
+      "learning_rate": 5e-05,
+      "loss": 1.5472,
+      "step": 3391
+    },
+    {
+      "epoch": 0.927536231884058,
+      "grad_norm": 0.17046159505844116,
+      "learning_rate": 5e-05,
+      "loss": 1.5272,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9278096800656276,
+      "grad_norm": 0.16787946224212646,
+      "learning_rate": 5e-05,
+      "loss": 1.5965,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9280831282471972,
+      "grad_norm": 0.1614540070295334,
+      "learning_rate": 5e-05,
+      "loss": 1.4654,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9283565764287668,
+      "grad_norm": 0.16552212834358215,
+      "learning_rate": 5e-05,
+      "loss": 1.5901,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9286300246103364,
+      "grad_norm": 0.1740752011537552,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9289034727919059,
+      "grad_norm": 0.16223378479480743,
+      "learning_rate": 5e-05,
+      "loss": 1.5006,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9291769209734755,
+      "grad_norm": 0.18281352519989014,
+      "learning_rate": 5e-05,
+      "loss": 1.6221,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9294503691550451,
+      "grad_norm": 0.17504733800888062,
+      "learning_rate": 5e-05,
+      "loss": 1.5975,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9297238173366147,
+      "grad_norm": 0.1651788204908371,
+      "learning_rate": 5e-05,
+      "loss": 1.5912,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9299972655181843,
+      "grad_norm": 0.172093003988266,
+      "learning_rate": 5e-05,
+      "loss": 1.6516,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9302707136997539,
+      "grad_norm": 0.17026016116142273,
+      "learning_rate": 5e-05,
+      "loss": 1.5742,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9305441618813235,
+      "grad_norm": 0.17041485011577606,
+      "learning_rate": 5e-05,
+      "loss": 1.6175,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9308176100628931,
+      "grad_norm": 0.17498993873596191,
+      "learning_rate": 5e-05,
+      "loss": 1.554,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9310910582444627,
+      "grad_norm": 0.1785575896501541,
+      "learning_rate": 5e-05,
+      "loss": 1.5705,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9313645064260323,
+      "grad_norm": 0.17882578074932098,
+      "learning_rate": 5e-05,
+      "loss": 1.5676,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9316379546076019,
+      "grad_norm": 0.16166472434997559,
+      "learning_rate": 5e-05,
+      "loss": 1.5654,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9319114027891715,
+      "grad_norm": 0.1717950403690338,
+      "learning_rate": 5e-05,
+      "loss": 1.5881,
+      "step": 3408
+    },
+    {
+      "epoch": 0.932184850970741,
+      "grad_norm": 0.17605219781398773,
+      "learning_rate": 5e-05,
+      "loss": 1.5802,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9324582991523106,
+      "grad_norm": 0.16096574068069458,
+      "learning_rate": 5e-05,
+      "loss": 1.51,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9327317473338802,
+      "grad_norm": 0.16242116689682007,
+      "learning_rate": 5e-05,
+      "loss": 1.5665,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9330051955154498,
+      "grad_norm": 0.17469222843647003,
+      "learning_rate": 5e-05,
+      "loss": 1.601,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9332786436970194,
+      "grad_norm": 0.16520720720291138,
+      "learning_rate": 5e-05,
+      "loss": 1.5964,
+      "step": 3413
+    },
+    {
+      "epoch": 0.933552091878589,
+      "grad_norm": 0.17878501117229462,
+      "learning_rate": 5e-05,
+      "loss": 1.7176,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9338255400601586,
+      "grad_norm": 0.16792459785938263,
+      "learning_rate": 5e-05,
+      "loss": 1.5655,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9340989882417282,
+      "grad_norm": 0.1701999306678772,
+      "learning_rate": 5e-05,
+      "loss": 1.5766,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9343724364232978,
+      "grad_norm": 0.17062143981456757,
+      "learning_rate": 5e-05,
+      "loss": 1.6155,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9346458846048674,
+      "grad_norm": 0.17067895829677582,
+      "learning_rate": 5e-05,
+      "loss": 1.5856,
+      "step": 3418
+    },
+    {
+      "epoch": 0.934919332786437,
+      "grad_norm": 0.17343777418136597,
+      "learning_rate": 5e-05,
+      "loss": 1.5841,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9351927809680065,
+      "grad_norm": 0.16160957515239716,
+      "learning_rate": 5e-05,
+      "loss": 1.5255,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9354662291495761,
+      "grad_norm": 0.15786494314670563,
+      "learning_rate": 5e-05,
+      "loss": 1.507,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9357396773311457,
+      "grad_norm": 0.18263064324855804,
+      "learning_rate": 5e-05,
+      "loss": 1.7325,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9360131255127153,
+      "grad_norm": 0.16316089034080505,
+      "learning_rate": 5e-05,
+      "loss": 1.421,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9362865736942849,
+      "grad_norm": 0.16691453754901886,
+      "learning_rate": 5e-05,
+      "loss": 1.6236,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9365600218758545,
+      "grad_norm": 0.16807349026203156,
+      "learning_rate": 5e-05,
+      "loss": 1.5508,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9368334700574241,
+      "grad_norm": 0.16728545725345612,
+      "learning_rate": 5e-05,
+      "loss": 1.4871,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9371069182389937,
+      "grad_norm": 0.1841699630022049,
+      "learning_rate": 5e-05,
+      "loss": 1.5571,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9373803664205633,
+      "grad_norm": 0.17346476018428802,
+      "learning_rate": 5e-05,
+      "loss": 1.5599,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9376538146021329,
+      "grad_norm": 0.17243662476539612,
+      "learning_rate": 5e-05,
+      "loss": 1.6099,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9379272627837025,
+      "grad_norm": 0.16930198669433594,
+      "learning_rate": 5e-05,
+      "loss": 1.6239,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9382007109652721,
+      "grad_norm": 0.17017267644405365,
+      "learning_rate": 5e-05,
+      "loss": 1.6719,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9384741591468416,
+      "grad_norm": 0.16782057285308838,
+      "learning_rate": 5e-05,
+      "loss": 1.6005,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9387476073284112,
+      "grad_norm": 0.16345928609371185,
+      "learning_rate": 5e-05,
+      "loss": 1.6027,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9390210555099808,
+      "grad_norm": 0.16896604001522064,
+      "learning_rate": 5e-05,
+      "loss": 1.6702,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9392945036915504,
+      "grad_norm": 0.16361366212368011,
+      "learning_rate": 5e-05,
+      "loss": 1.6016,
+      "step": 3435
+    },
+    {
+      "epoch": 0.93956795187312,
+      "grad_norm": 0.1642320603132248,
+      "learning_rate": 5e-05,
+      "loss": 1.5547,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9398414000546896,
+      "grad_norm": 0.1652129888534546,
+      "learning_rate": 5e-05,
+      "loss": 1.5415,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9401148482362592,
+      "grad_norm": 0.16417425870895386,
+      "learning_rate": 5e-05,
+      "loss": 1.5715,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9403882964178288,
+      "grad_norm": 0.16755367815494537,
+      "learning_rate": 5e-05,
+      "loss": 1.4629,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9406617445993984,
+      "grad_norm": 0.1799638420343399,
+      "learning_rate": 5e-05,
+      "loss": 1.5797,
+      "step": 3440
+    },
+    {
+      "epoch": 0.940935192780968,
+      "grad_norm": 0.17268924415111542,
+      "learning_rate": 5e-05,
+      "loss": 1.4751,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9412086409625376,
+      "grad_norm": 0.16716521978378296,
+      "learning_rate": 5e-05,
+      "loss": 1.5751,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9414820891441071,
+      "grad_norm": 0.17508552968502045,
+      "learning_rate": 5e-05,
+      "loss": 1.5959,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9417555373256767,
+      "grad_norm": 0.1688762605190277,
+      "learning_rate": 5e-05,
+      "loss": 1.6451,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9420289855072463,
+      "grad_norm": 0.17425693571567535,
+      "learning_rate": 5e-05,
+      "loss": 1.6965,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9423024336888159,
+      "grad_norm": 0.16022168099880219,
+      "learning_rate": 5e-05,
+      "loss": 1.5167,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9425758818703855,
+      "grad_norm": 0.16879145801067352,
+      "learning_rate": 5e-05,
+      "loss": 1.5844,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9428493300519551,
+      "grad_norm": 0.17180216312408447,
+      "learning_rate": 5e-05,
+      "loss": 1.5252,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9431227782335248,
+      "grad_norm": 0.17670823633670807,
+      "learning_rate": 5e-05,
+      "loss": 1.5545,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.1751013547182083,
+      "learning_rate": 5e-05,
+      "loss": 1.6204,
+      "step": 3450
+    },
+    {
+      "epoch": 0.943669674596664,
+      "grad_norm": 0.16705964505672455,
+      "learning_rate": 5e-05,
+      "loss": 1.5458,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9439431227782336,
+      "grad_norm": 0.17305395007133484,
+      "learning_rate": 5e-05,
+      "loss": 1.5924,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9442165709598032,
+      "grad_norm": 0.17383334040641785,
+      "learning_rate": 5e-05,
+      "loss": 1.6049,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9444900191413728,
+      "grad_norm": 0.16880303621292114,
+      "learning_rate": 5e-05,
+      "loss": 1.5726,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9447634673229423,
+      "grad_norm": 0.17519088089466095,
+      "learning_rate": 5e-05,
+      "loss": 1.5866,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9450369155045119,
+      "grad_norm": 0.1687891036272049,
+      "learning_rate": 5e-05,
+      "loss": 1.5832,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9453103636860815,
+      "grad_norm": 0.16175159811973572,
+      "learning_rate": 5e-05,
+      "loss": 1.6019,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9455838118676511,
+      "grad_norm": 0.17194457352161407,
+      "learning_rate": 5e-05,
+      "loss": 1.6486,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9458572600492207,
+      "grad_norm": 0.17335595190525055,
+      "learning_rate": 5e-05,
+      "loss": 1.6298,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9461307082307903,
+      "grad_norm": 0.17388460040092468,
+      "learning_rate": 5e-05,
+      "loss": 1.6029,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9464041564123599,
+      "grad_norm": 0.15883868932724,
+      "learning_rate": 5e-05,
+      "loss": 1.5193,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9466776045939295,
+      "grad_norm": 0.16564515233039856,
+      "learning_rate": 5e-05,
+      "loss": 1.5726,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9469510527754991,
+      "grad_norm": 0.1713998019695282,
+      "learning_rate": 5e-05,
+      "loss": 1.619,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9472245009570687,
+      "grad_norm": 0.16872616112232208,
+      "learning_rate": 5e-05,
+      "loss": 1.6437,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9474979491386383,
+      "grad_norm": 0.17497317492961884,
+      "learning_rate": 5e-05,
+      "loss": 1.533,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9477713973202079,
+      "grad_norm": 0.15996572375297546,
+      "learning_rate": 5e-05,
+      "loss": 1.4727,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9480448455017774,
+      "grad_norm": 0.1631600260734558,
+      "learning_rate": 5e-05,
+      "loss": 1.5921,
+      "step": 3467
+    },
+    {
+      "epoch": 0.948318293683347,
+      "grad_norm": 0.1694975644350052,
+      "learning_rate": 5e-05,
+      "loss": 1.6251,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9485917418649166,
+      "grad_norm": 0.1703629493713379,
+      "learning_rate": 5e-05,
+      "loss": 1.6061,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9488651900464862,
+      "grad_norm": 0.16327685117721558,
+      "learning_rate": 5e-05,
+      "loss": 1.5616,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9491386382280558,
+      "grad_norm": 0.16655370593070984,
+      "learning_rate": 5e-05,
+      "loss": 1.6223,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9494120864096254,
+      "grad_norm": 0.18266244232654572,
+      "learning_rate": 5e-05,
+      "loss": 1.5366,
+      "step": 3472
+    },
+    {
+      "epoch": 0.949685534591195,
+      "grad_norm": 0.1686059832572937,
+      "learning_rate": 5e-05,
+      "loss": 1.6063,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9499589827727646,
+      "grad_norm": 0.17475081980228424,
+      "learning_rate": 5e-05,
+      "loss": 1.5283,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9502324309543342,
+      "grad_norm": 0.17627473175525665,
+      "learning_rate": 5e-05,
+      "loss": 1.5529,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9505058791359038,
+      "grad_norm": 0.1728144884109497,
+      "learning_rate": 5e-05,
+      "loss": 1.4832,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9507793273174734,
+      "grad_norm": 0.19535605609416962,
+      "learning_rate": 5e-05,
+      "loss": 1.5747,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9510527754990429,
+      "grad_norm": 0.2011018991470337,
+      "learning_rate": 5e-05,
+      "loss": 1.5601,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9513262236806125,
+      "grad_norm": 0.17405074834823608,
+      "learning_rate": 5e-05,
+      "loss": 1.6053,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9515996718621821,
+      "grad_norm": 0.20217041671276093,
+      "learning_rate": 5e-05,
+      "loss": 1.607,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9518731200437517,
+      "grad_norm": 0.17352132499217987,
+      "learning_rate": 5e-05,
+      "loss": 1.5501,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9521465682253213,
+      "grad_norm": 0.17558707296848297,
+      "learning_rate": 5e-05,
+      "loss": 1.6479,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9524200164068909,
+      "grad_norm": 0.17117784917354584,
+      "learning_rate": 5e-05,
+      "loss": 1.5656,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9526934645884605,
+      "grad_norm": 0.16838552057743073,
+      "learning_rate": 5e-05,
+      "loss": 1.5381,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9529669127700301,
+      "grad_norm": 0.17733819782733917,
+      "learning_rate": 5e-05,
+      "loss": 1.6453,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9532403609515997,
+      "grad_norm": 0.1712433099746704,
+      "learning_rate": 5e-05,
+      "loss": 1.5944,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9535138091331693,
+      "grad_norm": 0.19022230803966522,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9537872573147389,
+      "grad_norm": 0.17896117269992828,
+      "learning_rate": 5e-05,
+      "loss": 1.6098,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9540607054963085,
+      "grad_norm": 0.17199327051639557,
+      "learning_rate": 5e-05,
+      "loss": 1.589,
+      "step": 3489
+    },
+    {
+      "epoch": 0.954334153677878,
+      "grad_norm": 0.19092747569084167,
+      "learning_rate": 5e-05,
+      "loss": 1.522,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9546076018594476,
+      "grad_norm": 0.17024657130241394,
+      "learning_rate": 5e-05,
+      "loss": 1.6758,
+      "step": 3491
+    },
+    {
+      "epoch": 0.9548810500410172,
+      "grad_norm": 0.17147701978683472,
+      "learning_rate": 5e-05,
+      "loss": 1.5168,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9551544982225868,
+      "grad_norm": 0.18577666580677032,
+      "learning_rate": 5e-05,
+      "loss": 1.5943,
+      "step": 3493
+    },
+    {
+      "epoch": 0.9554279464041564,
+      "grad_norm": 0.16658198833465576,
+      "learning_rate": 5e-05,
+      "loss": 1.6217,
+      "step": 3494
+    },
+    {
+      "epoch": 0.955701394585726,
+      "grad_norm": 0.17051103711128235,
+      "learning_rate": 5e-05,
+      "loss": 1.521,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9559748427672956,
+      "grad_norm": 0.1750299036502838,
+      "learning_rate": 5e-05,
+      "loss": 1.541,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9562482909488652,
+      "grad_norm": 0.185097336769104,
+      "learning_rate": 5e-05,
+      "loss": 1.664,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.17204828560352325,
+      "learning_rate": 5e-05,
+      "loss": 1.6225,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9567951873120044,
+      "grad_norm": 0.17451748251914978,
+      "learning_rate": 5e-05,
+      "loss": 1.5786,
+      "step": 3499
+    },
+    {
+      "epoch": 0.957068635493574,
+      "grad_norm": 0.20204168558120728,
+      "learning_rate": 5e-05,
+      "loss": 1.7499,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9573420836751435,
+      "grad_norm": 0.17379069328308105,
+      "learning_rate": 5e-05,
+      "loss": 1.585,
+      "step": 3501
+    },
+    {
+      "epoch": 0.9576155318567131,
+      "grad_norm": 0.1748073250055313,
+      "learning_rate": 5e-05,
+      "loss": 1.5567,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9578889800382827,
+      "grad_norm": 0.20730933547019958,
+      "learning_rate": 5e-05,
+      "loss": 1.7386,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9581624282198523,
+      "grad_norm": 0.1670350432395935,
+      "learning_rate": 5e-05,
+      "loss": 1.5471,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9584358764014219,
+      "grad_norm": 0.16766297817230225,
+      "learning_rate": 5e-05,
+      "loss": 1.4913,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9587093245829915,
+      "grad_norm": 0.18351885676383972,
+      "learning_rate": 5e-05,
+      "loss": 1.5318,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9589827727645611,
+      "grad_norm": 0.15731576085090637,
+      "learning_rate": 5e-05,
+      "loss": 1.4561,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9592562209461307,
+      "grad_norm": 0.18964198231697083,
+      "learning_rate": 5e-05,
+      "loss": 1.6603,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9595296691277003,
+      "grad_norm": 0.19028975069522858,
+      "learning_rate": 5e-05,
+      "loss": 1.6721,
+      "step": 3509
+    },
+    {
+      "epoch": 0.9598031173092699,
+      "grad_norm": 0.16938328742980957,
+      "learning_rate": 5e-05,
+      "loss": 1.5652,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9600765654908395,
+      "grad_norm": 0.17705866694450378,
+      "learning_rate": 5e-05,
+      "loss": 1.5832,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9603500136724091,
+      "grad_norm": 0.1780325025320053,
+      "learning_rate": 5e-05,
+      "loss": 1.5572,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9606234618539786,
+      "grad_norm": 0.18230192363262177,
+      "learning_rate": 5e-05,
+      "loss": 1.6136,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9608969100355482,
+      "grad_norm": 0.18078622221946716,
+      "learning_rate": 5e-05,
+      "loss": 1.5788,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9611703582171178,
+      "grad_norm": 0.17788712680339813,
+      "learning_rate": 5e-05,
+      "loss": 1.4601,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9614438063986874,
+      "grad_norm": 0.17222706973552704,
+      "learning_rate": 5e-05,
+      "loss": 1.586,
+      "step": 3516
+    },
+    {
+      "epoch": 0.961717254580257,
+      "grad_norm": 0.1748836636543274,
+      "learning_rate": 5e-05,
+      "loss": 1.6527,
+      "step": 3517
+    },
+    {
+      "epoch": 0.9619907027618266,
+      "grad_norm": 0.18419338762760162,
+      "learning_rate": 5e-05,
+      "loss": 1.6447,
+      "step": 3518
+    },
+    {
+      "epoch": 0.9622641509433962,
+      "grad_norm": 0.17671184241771698,
+      "learning_rate": 5e-05,
+      "loss": 1.6371,
+      "step": 3519
+    },
+    {
+      "epoch": 0.9625375991249658,
+      "grad_norm": 0.1626741737127304,
+      "learning_rate": 5e-05,
+      "loss": 1.6034,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9628110473065354,
+      "grad_norm": 0.17750827968120575,
+      "learning_rate": 5e-05,
+      "loss": 1.5709,
+      "step": 3521
+    },
+    {
+      "epoch": 0.963084495488105,
+      "grad_norm": 0.18092156946659088,
+      "learning_rate": 5e-05,
+      "loss": 1.5488,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9633579436696746,
+      "grad_norm": 0.20844420790672302,
+      "learning_rate": 5e-05,
+      "loss": 1.6209,
+      "step": 3523
+    },
+    {
+      "epoch": 0.9636313918512441,
+      "grad_norm": 0.1704307496547699,
+      "learning_rate": 5e-05,
+      "loss": 1.6154,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9639048400328137,
+      "grad_norm": 0.17518380284309387,
+      "learning_rate": 5e-05,
+      "loss": 1.4978,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9641782882143833,
+      "grad_norm": 0.18488885462284088,
+      "learning_rate": 5e-05,
+      "loss": 1.5973,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9644517363959529,
+      "grad_norm": 0.1840328723192215,
+      "learning_rate": 5e-05,
+      "loss": 1.6942,
+      "step": 3527
+    },
+    {
+      "epoch": 0.9647251845775225,
+      "grad_norm": 0.16851232945919037,
+      "learning_rate": 5e-05,
+      "loss": 1.5614,
+      "step": 3528
+    },
+    {
+      "epoch": 0.9649986327590921,
+      "grad_norm": 0.1788729429244995,
+      "learning_rate": 5e-05,
+      "loss": 1.5971,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9652720809406617,
+      "grad_norm": 0.17521923780441284,
+      "learning_rate": 5e-05,
+      "loss": 1.5597,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9655455291222313,
+      "grad_norm": 0.16558045148849487,
+      "learning_rate": 5e-05,
+      "loss": 1.5802,
+      "step": 3531
+    },
+    {
+      "epoch": 0.965818977303801,
+      "grad_norm": 0.16922442615032196,
+      "learning_rate": 5e-05,
+      "loss": 1.5096,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9660924254853706,
+      "grad_norm": 0.16690833866596222,
+      "learning_rate": 5e-05,
+      "loss": 1.5354,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9663658736669402,
+      "grad_norm": 0.17629151046276093,
+      "learning_rate": 5e-05,
+      "loss": 1.5951,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9666393218485098,
+      "grad_norm": 0.1715892106294632,
+      "learning_rate": 5e-05,
+      "loss": 1.6381,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9669127700300792,
+      "grad_norm": 0.17085449397563934,
+      "learning_rate": 5e-05,
+      "loss": 1.5276,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9671862182116489,
+      "grad_norm": 0.17155250906944275,
+      "learning_rate": 5e-05,
+      "loss": 1.5824,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9674596663932185,
+      "grad_norm": 0.16368617117404938,
+      "learning_rate": 5e-05,
+      "loss": 1.5595,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9677331145747881,
+      "grad_norm": 0.17515048384666443,
+      "learning_rate": 5e-05,
+      "loss": 1.5234,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9680065627563577,
+      "grad_norm": 0.17045453190803528,
+      "learning_rate": 5e-05,
+      "loss": 1.5847,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9682800109379273,
+      "grad_norm": 0.17592334747314453,
+      "learning_rate": 5e-05,
+      "loss": 1.681,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9685534591194969,
+      "grad_norm": 0.174262136220932,
+      "learning_rate": 5e-05,
+      "loss": 1.6834,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9688269073010665,
+      "grad_norm": 0.16702409088611603,
+      "learning_rate": 5e-05,
+      "loss": 1.5888,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9691003554826361,
+      "grad_norm": 0.16639576852321625,
+      "learning_rate": 5e-05,
+      "loss": 1.57,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9693738036642057,
+      "grad_norm": 0.16943658888339996,
+      "learning_rate": 5e-05,
+      "loss": 1.605,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9696472518457753,
+      "grad_norm": 0.16538041830062866,
+      "learning_rate": 5e-05,
+      "loss": 1.562,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9699207000273449,
+      "grad_norm": 0.16645492613315582,
+      "learning_rate": 5e-05,
+      "loss": 1.5321,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9701941482089144,
+      "grad_norm": 0.17388178408145905,
+      "learning_rate": 5e-05,
+      "loss": 1.576,
+      "step": 3548
+    },
+    {
+      "epoch": 0.970467596390484,
+      "grad_norm": 0.17088045179843903,
+      "learning_rate": 5e-05,
+      "loss": 1.5195,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9707410445720536,
+      "grad_norm": 0.17386609315872192,
+      "learning_rate": 5e-05,
+      "loss": 1.5194,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9710144927536232,
+      "grad_norm": 0.1669374406337738,
+      "learning_rate": 5e-05,
+      "loss": 1.4916,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9712879409351928,
+      "grad_norm": 0.18806956708431244,
+      "learning_rate": 5e-05,
+      "loss": 1.4877,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9715613891167624,
+      "grad_norm": 0.1629197746515274,
+      "learning_rate": 5e-05,
+      "loss": 1.5365,
+      "step": 3553
+    },
+    {
+      "epoch": 0.971834837298332,
+      "grad_norm": 0.16897207498550415,
+      "learning_rate": 5e-05,
+      "loss": 1.556,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9721082854799016,
+      "grad_norm": 0.16960465908050537,
+      "learning_rate": 5e-05,
+      "loss": 1.5432,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9723817336614712,
+      "grad_norm": 0.16694605350494385,
+      "learning_rate": 5e-05,
+      "loss": 1.6413,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9726551818430408,
+      "grad_norm": 0.162724107503891,
+      "learning_rate": 5e-05,
+      "loss": 1.491,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9729286300246104,
+      "grad_norm": 0.1615595519542694,
+      "learning_rate": 5e-05,
+      "loss": 1.5386,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9732020782061799,
+      "grad_norm": 0.15930253267288208,
+      "learning_rate": 5e-05,
+      "loss": 1.5991,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9734755263877495,
+      "grad_norm": 0.16046123206615448,
+      "learning_rate": 5e-05,
+      "loss": 1.5796,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9737489745693191,
+      "grad_norm": 0.16169172525405884,
+      "learning_rate": 5e-05,
+      "loss": 1.5435,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9740224227508887,
+      "grad_norm": 0.17044523358345032,
+      "learning_rate": 5e-05,
+      "loss": 1.5045,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9742958709324583,
+      "grad_norm": 0.16750258207321167,
+      "learning_rate": 5e-05,
+      "loss": 1.6184,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9745693191140279,
+      "grad_norm": 0.16423483192920685,
+      "learning_rate": 5e-05,
+      "loss": 1.5574,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9748427672955975,
+      "grad_norm": 0.16018086671829224,
+      "learning_rate": 5e-05,
+      "loss": 1.4814,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9751162154771671,
+      "grad_norm": 0.16692084074020386,
+      "learning_rate": 5e-05,
+      "loss": 1.565,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9753896636587367,
+      "grad_norm": 0.16623759269714355,
+      "learning_rate": 5e-05,
+      "loss": 1.5177,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9756631118403063,
+      "grad_norm": 0.1683354675769806,
+      "learning_rate": 5e-05,
+      "loss": 1.5655,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9759365600218759,
+      "grad_norm": 0.17353872954845428,
+      "learning_rate": 5e-05,
+      "loss": 1.5703,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9762100082034455,
+      "grad_norm": 0.1665719896554947,
+      "learning_rate": 5e-05,
+      "loss": 1.5936,
+      "step": 3570
+    },
+    {
+      "epoch": 0.976483456385015,
+      "grad_norm": 0.17849120497703552,
+      "learning_rate": 5e-05,
+      "loss": 1.6144,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9767569045665846,
+      "grad_norm": 0.17340973019599915,
+      "learning_rate": 5e-05,
+      "loss": 1.6221,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9770303527481542,
+      "grad_norm": 0.17012323439121246,
+      "learning_rate": 5e-05,
+      "loss": 1.5603,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9773038009297238,
+      "grad_norm": 0.1630171239376068,
+      "learning_rate": 5e-05,
+      "loss": 1.4353,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9775772491112934,
+      "grad_norm": 0.16621048748493195,
+      "learning_rate": 5e-05,
+      "loss": 1.5738,
+      "step": 3575
+    },
+    {
+      "epoch": 0.977850697292863,
+      "grad_norm": 0.16702908277511597,
+      "learning_rate": 5e-05,
+      "loss": 1.5689,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9781241454744326,
+      "grad_norm": 0.18123649060726166,
+      "learning_rate": 5e-05,
+      "loss": 1.687,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9783975936560022,
+      "grad_norm": 0.1809684783220291,
+      "learning_rate": 5e-05,
+      "loss": 1.5811,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9786710418375718,
+      "grad_norm": 0.18062925338745117,
+      "learning_rate": 5e-05,
+      "loss": 1.6317,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9789444900191414,
+      "grad_norm": 0.1767573207616806,
+      "learning_rate": 5e-05,
+      "loss": 1.6907,
+      "step": 3580
+    },
+    {
+      "epoch": 0.979217938200711,
+      "grad_norm": 0.17363345623016357,
+      "learning_rate": 5e-05,
+      "loss": 1.6019,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9794913863822805,
+      "grad_norm": 0.173218235373497,
+      "learning_rate": 5e-05,
+      "loss": 1.5115,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9797648345638501,
+      "grad_norm": 0.16840875148773193,
+      "learning_rate": 5e-05,
+      "loss": 1.5806,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9800382827454197,
+      "grad_norm": 0.17666743695735931,
+      "learning_rate": 5e-05,
+      "loss": 1.6649,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9803117309269893,
+      "grad_norm": 0.1744329184293747,
+      "learning_rate": 5e-05,
+      "loss": 1.5703,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9805851791085589,
+      "grad_norm": 0.1757236123085022,
+      "learning_rate": 5e-05,
+      "loss": 1.6247,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9808586272901285,
+      "grad_norm": 0.17235401272773743,
+      "learning_rate": 5e-05,
+      "loss": 1.6508,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.1712830811738968,
+      "learning_rate": 5e-05,
+      "loss": 1.6403,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9814055236532677,
+      "grad_norm": 0.16565677523612976,
+      "learning_rate": 5e-05,
+      "loss": 1.5304,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9816789718348373,
+      "grad_norm": 0.1652953326702118,
+      "learning_rate": 5e-05,
+      "loss": 1.553,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9819524200164069,
+      "grad_norm": 0.17286698520183563,
+      "learning_rate": 5e-05,
+      "loss": 1.5573,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9822258681979765,
+      "grad_norm": 0.17360328137874603,
+      "learning_rate": 5e-05,
+      "loss": 1.6319,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9824993163795461,
+      "grad_norm": 0.1673014909029007,
+      "learning_rate": 5e-05,
+      "loss": 1.5778,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9827727645611156,
+      "grad_norm": 0.17071346938610077,
+      "learning_rate": 5e-05,
+      "loss": 1.6138,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9830462127426852,
+      "grad_norm": 0.1741405427455902,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9833196609242548,
+      "grad_norm": 0.17240159213542938,
+      "learning_rate": 5e-05,
+      "loss": 1.6181,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9835931091058244,
+      "grad_norm": 0.1678202599287033,
+      "learning_rate": 5e-05,
+      "loss": 1.5552,
+      "step": 3597
+    },
+    {
+      "epoch": 0.983866557287394,
+      "grad_norm": 0.17836597561836243,
+      "learning_rate": 5e-05,
+      "loss": 1.5859,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9841400054689636,
+      "grad_norm": 0.17026908695697784,
+      "learning_rate": 5e-05,
+      "loss": 1.6616,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9844134536505332,
+      "grad_norm": 0.17007942497730255,
+      "learning_rate": 5e-05,
+      "loss": 1.48,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9846869018321028,
+      "grad_norm": 0.1805422455072403,
+      "learning_rate": 5e-05,
+      "loss": 1.6306,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9849603500136724,
+      "grad_norm": 0.16335126757621765,
+      "learning_rate": 5e-05,
+      "loss": 1.5206,
+      "step": 3602
+    },
+    {
+      "epoch": 0.985233798195242,
+      "grad_norm": 0.1635698527097702,
+      "learning_rate": 5e-05,
+      "loss": 1.5101,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9855072463768116,
+      "grad_norm": 0.18003027141094208,
+      "learning_rate": 5e-05,
+      "loss": 1.643,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9857806945583811,
+      "grad_norm": 0.17599767446517944,
+      "learning_rate": 5e-05,
+      "loss": 1.6374,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9860541427399507,
+      "grad_norm": 0.1730010062456131,
+      "learning_rate": 5e-05,
+      "loss": 1.5686,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9863275909215203,
+      "grad_norm": 0.17399701476097107,
+      "learning_rate": 5e-05,
+      "loss": 1.6065,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9866010391030899,
+      "grad_norm": 0.18367905914783478,
+      "learning_rate": 5e-05,
+      "loss": 1.7215,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9868744872846595,
+      "grad_norm": 0.18548168241977692,
+      "learning_rate": 5e-05,
+      "loss": 1.5337,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9871479354662291,
+      "grad_norm": 0.1609538048505783,
+      "learning_rate": 5e-05,
+      "loss": 1.5586,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9874213836477987,
+      "grad_norm": 0.17349039018154144,
+      "learning_rate": 5e-05,
+      "loss": 1.659,
+      "step": 3611
+    },
+    {
+      "epoch": 0.9876948318293683,
+      "grad_norm": 0.17284166812896729,
+      "learning_rate": 5e-05,
+      "loss": 1.6581,
+      "step": 3612
+    },
+    {
+      "epoch": 0.987968280010938,
+      "grad_norm": 0.16577011346817017,
+      "learning_rate": 5e-05,
+      "loss": 1.4734,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9882417281925076,
+      "grad_norm": 0.17548039555549622,
+      "learning_rate": 5e-05,
+      "loss": 1.664,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9885151763740772,
+      "grad_norm": 0.16056503355503082,
+      "learning_rate": 5e-05,
+      "loss": 1.5385,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9887886245556468,
+      "grad_norm": 0.1729920357465744,
+      "learning_rate": 5e-05,
+      "loss": 1.6822,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9890620727372162,
+      "grad_norm": 0.1658102124929428,
+      "learning_rate": 5e-05,
+      "loss": 1.5311,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9893355209187858,
+      "grad_norm": 0.15829437971115112,
+      "learning_rate": 5e-05,
+      "loss": 1.5528,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9896089691003555,
+      "grad_norm": 0.160248264670372,
+      "learning_rate": 5e-05,
+      "loss": 1.5386,
+      "step": 3619
+    },
+    {
+      "epoch": 0.989882417281925,
+      "grad_norm": 0.1627732813358307,
+      "learning_rate": 5e-05,
+      "loss": 1.5654,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9901558654634947,
+      "grad_norm": 0.17312705516815186,
+      "learning_rate": 5e-05,
+      "loss": 1.4819,
+      "step": 3621
+    },
+    {
+      "epoch": 0.9904293136450643,
+      "grad_norm": 0.1666928231716156,
+      "learning_rate": 5e-05,
+      "loss": 1.5008,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9907027618266339,
+      "grad_norm": 0.16598451137542725,
+      "learning_rate": 5e-05,
+      "loss": 1.4918,
+      "step": 3623
+    },
+    {
+      "epoch": 0.9909762100082035,
+      "grad_norm": 0.17471113801002502,
+      "learning_rate": 5e-05,
+      "loss": 1.6568,
+      "step": 3624
+    },
+    {
+      "epoch": 0.9912496581897731,
+      "grad_norm": 0.16479487717151642,
+      "learning_rate": 5e-05,
+      "loss": 1.5205,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9915231063713427,
+      "grad_norm": 0.18188925087451935,
+      "learning_rate": 5e-05,
+      "loss": 1.6185,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9917965545529123,
+      "grad_norm": 0.1632193922996521,
+      "learning_rate": 5e-05,
+      "loss": 1.5691,
+      "step": 3627
+    },
+    {
+      "epoch": 0.9920700027344819,
+      "grad_norm": 0.16891783475875854,
+      "learning_rate": 5e-05,
+      "loss": 1.53,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9923434509160514,
+      "grad_norm": 0.17035579681396484,
+      "learning_rate": 5e-05,
+      "loss": 1.5631,
+      "step": 3629
+    },
+    {
+      "epoch": 0.992616899097621,
+      "grad_norm": 0.16007426381111145,
+      "learning_rate": 5e-05,
+      "loss": 1.414,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9928903472791906,
+      "grad_norm": 0.17161479592323303,
+      "learning_rate": 5e-05,
+      "loss": 1.5988,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9931637954607602,
+      "grad_norm": 0.16341786086559296,
+      "learning_rate": 5e-05,
+      "loss": 1.5492,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9934372436423298,
+      "grad_norm": 0.17597921192646027,
+      "learning_rate": 5e-05,
+      "loss": 1.6547,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9937106918238994,
+      "grad_norm": 0.18249720335006714,
+      "learning_rate": 5e-05,
+      "loss": 1.7218,
+      "step": 3634
+    },
+    {
+      "epoch": 0.993984140005469,
+      "grad_norm": 0.17179979383945465,
+      "learning_rate": 5e-05,
+      "loss": 1.6261,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9942575881870386,
+      "grad_norm": 0.16686037182807922,
+      "learning_rate": 5e-05,
+      "loss": 1.5381,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9945310363686082,
+      "grad_norm": 0.17167183756828308,
+      "learning_rate": 5e-05,
+      "loss": 1.6132,
+      "step": 3637
+    },
+    {
+      "epoch": 0.9948044845501778,
+      "grad_norm": 0.1685621589422226,
+      "learning_rate": 5e-05,
+      "loss": 1.5806,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9950779327317474,
+      "grad_norm": 0.17752127349376678,
+      "learning_rate": 5e-05,
+      "loss": 1.5842,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9953513809133169,
+      "grad_norm": 0.18407252430915833,
+      "learning_rate": 5e-05,
+      "loss": 1.5769,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9956248290948865,
+      "grad_norm": 0.17489740252494812,
+      "learning_rate": 5e-05,
+      "loss": 1.6766,
+      "step": 3641
+    },
+    {
+      "epoch": 0.9958982772764561,
+      "grad_norm": 0.1652628481388092,
+      "learning_rate": 5e-05,
+      "loss": 1.5431,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9961717254580257,
+      "grad_norm": 0.16765595972537994,
+      "learning_rate": 5e-05,
+      "loss": 1.5277,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9964451736395953,
+      "grad_norm": 0.18051624298095703,
+      "learning_rate": 5e-05,
+      "loss": 1.7006,
+      "step": 3644
+    },
+    {
+      "epoch": 0.9967186218211649,
+      "grad_norm": 0.16528740525245667,
+      "learning_rate": 5e-05,
+      "loss": 1.5304,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9969920700027345,
+      "grad_norm": 0.18011535704135895,
+      "learning_rate": 5e-05,
+      "loss": 1.6645,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9972655181843041,
+      "grad_norm": 0.17780528962612152,
+      "learning_rate": 5e-05,
+      "loss": 1.5679,
+      "step": 3647
+    },
+    {
+      "epoch": 0.9975389663658737,
+      "grad_norm": 0.16724787652492523,
+      "learning_rate": 5e-05,
+      "loss": 1.5967,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9978124145474433,
+      "grad_norm": 0.1816672682762146,
+      "learning_rate": 5e-05,
+      "loss": 1.5915,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9980858627290129,
+      "grad_norm": 0.1704689860343933,
+      "learning_rate": 5e-05,
+      "loss": 1.4793,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9983593109105825,
+      "grad_norm": 0.16653746366500854,
+      "learning_rate": 5e-05,
+      "loss": 1.5062,
+      "step": 3651
+    },
+    {
+      "epoch": 0.998632759092152,
+      "grad_norm": 0.17958907783031464,
+      "learning_rate": 5e-05,
+      "loss": 1.5776,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9989062072737216,
+      "grad_norm": 0.1751994490623474,
+      "learning_rate": 5e-05,
+      "loss": 1.5787,
+      "step": 3653
+    },
+    {
+      "epoch": 0.9991796554552912,
+      "grad_norm": 0.16929645836353302,
+      "learning_rate": 5e-05,
+      "loss": 1.5965,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9994531036368608,
+      "grad_norm": 0.19185209274291992,
+      "learning_rate": 5e-05,
+      "loss": 1.5234,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9997265518184304,
+      "grad_norm": 0.17126961052417755,
+      "learning_rate": 5e-05,
+      "loss": 1.5285,
+      "step": 3656
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17371578514575958,
+      "learning_rate": 5e-05,
+      "loss": 1.5768,
+      "step": 3657
+    },
+    {
+      "epoch": 1.0,
+      "step": 3657,
+      "total_flos": 2.81863078907845e+18,
+      "train_loss": 1.6374119961558387,
+      "train_runtime": 72812.9252,
+      "train_samples_per_second": 0.402,
+      "train_steps_per_second": 0.05
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3657,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.81863078907845e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}