diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,54774 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 7820,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00128,
+      "grad_norm": 13.502479553222656,
+      "learning_rate": 0.0,
+      "loss": 10.9785,
+      "step": 1
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 13.624500274658203,
+      "learning_rate": 7.672634271099745e-07,
+      "loss": 10.98,
+      "step": 2
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 13.133468627929688,
+      "learning_rate": 1.534526854219949e-06,
+      "loss": 10.9409,
+      "step": 3
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 13.427982330322266,
+      "learning_rate": 2.301790281329923e-06,
+      "loss": 10.8518,
+      "step": 4
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 12.44352912902832,
+      "learning_rate": 3.069053708439898e-06,
+      "loss": 10.738,
+      "step": 5
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 10.861345291137695,
+      "learning_rate": 3.836317135549872e-06,
+      "loss": 10.6095,
+      "step": 6
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 9.260466575622559,
+      "learning_rate": 4.603580562659846e-06,
+      "loss": 10.4604,
+      "step": 7
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 7.927541255950928,
+      "learning_rate": 5.3708439897698205e-06,
+      "loss": 10.3281,
+      "step": 8
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 6.898108959197998,
+      "learning_rate": 6.138107416879796e-06,
+      "loss": 10.1985,
+      "step": 9
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 5.914000988006592,
+      "learning_rate": 6.905370843989769e-06,
+      "loss": 10.0951,
+      "step": 10
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 5.054530143737793,
+      "learning_rate": 7.672634271099744e-06,
+      "loss": 10.012,
+      "step": 11
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 4.351410865783691,
+      "learning_rate": 8.439897698209718e-06,
+      "loss": 9.9117,
+      "step": 12
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 3.752089500427246,
+      "learning_rate": 9.207161125319692e-06,
+      "loss": 9.8225,
+      "step": 13
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 3.252004623413086,
+      "learning_rate": 9.974424552429668e-06,
+      "loss": 9.7603,
+      "step": 14
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 2.8720457553863525,
+      "learning_rate": 1.0741687979539641e-05,
+      "loss": 9.7285,
+      "step": 15
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 2.666605234146118,
+      "learning_rate": 1.1508951406649615e-05,
+      "loss": 9.6632,
+      "step": 16
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 2.497063398361206,
+      "learning_rate": 1.2276214833759591e-05,
+      "loss": 9.6141,
+      "step": 17
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 2.3066694736480713,
+      "learning_rate": 1.3043478260869564e-05,
+      "loss": 9.6125,
+      "step": 18
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 2.267719268798828,
+      "learning_rate": 1.3810741687979538e-05,
+      "loss": 9.5742,
+      "step": 19
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 2.268841505050659,
+      "learning_rate": 1.4578005115089511e-05,
+      "loss": 9.5505,
+      "step": 20
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 2.2032310962677,
+      "learning_rate": 1.5345268542199487e-05,
+      "loss": 9.5406,
+      "step": 21
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 2.1540212631225586,
+      "learning_rate": 1.611253196930946e-05,
+      "loss": 9.5205,
+      "step": 22
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 2.1648716926574707,
+      "learning_rate": 1.6879795396419436e-05,
+      "loss": 9.5023,
+      "step": 23
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 2.1777994632720947,
+      "learning_rate": 1.764705882352941e-05,
+      "loss": 9.4566,
+      "step": 24
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 2.093733310699463,
+      "learning_rate": 1.8414322250639385e-05,
+      "loss": 9.4797,
+      "step": 25
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 2.14025616645813,
+      "learning_rate": 1.918158567774936e-05,
+      "loss": 9.4326,
+      "step": 26
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 2.243187665939331,
+      "learning_rate": 1.9948849104859337e-05,
+      "loss": 9.3569,
+      "step": 27
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 2.1062655448913574,
+      "learning_rate": 2.0716112531969308e-05,
+      "loss": 9.3849,
+      "step": 28
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 2.0766758918762207,
+      "learning_rate": 2.1483375959079282e-05,
+      "loss": 9.351,
+      "step": 29
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 2.101067304611206,
+      "learning_rate": 2.2250639386189256e-05,
+      "loss": 9.3139,
+      "step": 30
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 2.0506746768951416,
+      "learning_rate": 2.301790281329923e-05,
+      "loss": 9.2755,
+      "step": 31
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 2.015533924102783,
+      "learning_rate": 2.3785166240409205e-05,
+      "loss": 9.2446,
+      "step": 32
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 1.9934455156326294,
+      "learning_rate": 2.4552429667519183e-05,
+      "loss": 9.2283,
+      "step": 33
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 1.9839218854904175,
+      "learning_rate": 2.5319693094629154e-05,
+      "loss": 9.1785,
+      "step": 34
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 2.0591323375701904,
+      "learning_rate": 2.6086956521739128e-05,
+      "loss": 9.1733,
+      "step": 35
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 1.8989123106002808,
+      "learning_rate": 2.6854219948849103e-05,
+      "loss": 9.1775,
+      "step": 36
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 2.2588517665863037,
+      "learning_rate": 2.7621483375959077e-05,
+      "loss": 9.1317,
+      "step": 37
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 1.869653582572937,
+      "learning_rate": 2.838874680306905e-05,
+      "loss": 9.0794,
+      "step": 38
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 2.098316192626953,
+      "learning_rate": 2.9156010230179022e-05,
+      "loss": 9.0853,
+      "step": 39
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 1.8606550693511963,
+      "learning_rate": 2.9923273657289e-05,
+      "loss": 9.0243,
+      "step": 40
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 1.9760617017745972,
+      "learning_rate": 3.0690537084398974e-05,
+      "loss": 8.9874,
+      "step": 41
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.8577723503112793,
+      "learning_rate": 3.145780051150895e-05,
+      "loss": 8.9687,
+      "step": 42
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 1.7823419570922852,
+      "learning_rate": 3.222506393861892e-05,
+      "loss": 8.9302,
+      "step": 43
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 1.8468880653381348,
+      "learning_rate": 3.29923273657289e-05,
+      "loss": 8.8693,
+      "step": 44
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.7553874254226685,
+      "learning_rate": 3.375959079283887e-05,
+      "loss": 8.8835,
+      "step": 45
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 1.8048585653305054,
+      "learning_rate": 3.4526854219948846e-05,
+      "loss": 8.8015,
+      "step": 46
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 1.8145036697387695,
+      "learning_rate": 3.529411764705882e-05,
+      "loss": 8.7847,
+      "step": 47
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 1.6603283882141113,
+      "learning_rate": 3.6061381074168795e-05,
+      "loss": 8.7738,
+      "step": 48
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 1.6837772130966187,
+      "learning_rate": 3.682864450127877e-05,
+      "loss": 8.7088,
+      "step": 49
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 1.686160683631897,
+      "learning_rate": 3.7595907928388744e-05,
+      "loss": 8.6635,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 1.7713919878005981,
+      "learning_rate": 3.836317135549872e-05,
+      "loss": 8.6265,
+      "step": 51
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 1.8491101264953613,
+      "learning_rate": 3.913043478260869e-05,
+      "loss": 8.5384,
+      "step": 52
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 1.6337801218032837,
+      "learning_rate": 3.989769820971867e-05,
+      "loss": 8.5446,
+      "step": 53
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 1.5614173412322998,
+      "learning_rate": 4.066496163682864e-05,
+      "loss": 8.5252,
+      "step": 54
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 1.5374433994293213,
+      "learning_rate": 4.1432225063938615e-05,
+      "loss": 8.4633,
+      "step": 55
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 1.6636866331100464,
+      "learning_rate": 4.219948849104859e-05,
+      "loss": 8.4685,
+      "step": 56
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 1.6900804042816162,
+      "learning_rate": 4.2966751918158564e-05,
+      "loss": 8.4324,
+      "step": 57
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 1.5862479209899902,
+      "learning_rate": 4.373401534526854e-05,
+      "loss": 8.359,
+      "step": 58
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 1.5928101539611816,
+      "learning_rate": 4.450127877237851e-05,
+      "loss": 8.3375,
+      "step": 59
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 1.5766217708587646,
+      "learning_rate": 4.526854219948848e-05,
+      "loss": 8.2851,
+      "step": 60
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 1.5101016759872437,
+      "learning_rate": 4.603580562659846e-05,
+      "loss": 8.2516,
+      "step": 61
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 1.3688981533050537,
+      "learning_rate": 4.6803069053708436e-05,
+      "loss": 8.268,
+      "step": 62
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 1.4474924802780151,
+      "learning_rate": 4.757033248081841e-05,
+      "loss": 8.2099,
+      "step": 63
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 1.4008480310440063,
+      "learning_rate": 4.8337595907928385e-05,
+      "loss": 8.1661,
+      "step": 64
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 1.4576811790466309,
+      "learning_rate": 4.9104859335038366e-05,
+      "loss": 8.1582,
+      "step": 65
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 1.4062941074371338,
+      "learning_rate": 4.987212276214833e-05,
+      "loss": 8.0852,
+      "step": 66
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 1.3305187225341797,
+      "learning_rate": 5.063938618925831e-05,
+      "loss": 8.0746,
+      "step": 67
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 1.3217337131500244,
+      "learning_rate": 5.140664961636828e-05,
+      "loss": 8.0454,
+      "step": 68
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 1.3334351778030396,
+      "learning_rate": 5.2173913043478256e-05,
+      "loss": 7.9983,
+      "step": 69
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 1.2380847930908203,
+      "learning_rate": 5.294117647058824e-05,
+      "loss": 7.9684,
+      "step": 70
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 1.3053793907165527,
+      "learning_rate": 5.3708439897698205e-05,
+      "loss": 7.9135,
+      "step": 71
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 1.3208932876586914,
+      "learning_rate": 5.447570332480817e-05,
+      "loss": 7.9277,
+      "step": 72
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 1.3776912689208984,
+      "learning_rate": 5.5242966751918154e-05,
+      "loss": 7.861,
+      "step": 73
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 1.1503986120224,
+      "learning_rate": 5.601023017902813e-05,
+      "loss": 7.8479,
+      "step": 74
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 1.2197959423065186,
+      "learning_rate": 5.67774936061381e-05,
+      "loss": 7.7553,
+      "step": 75
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 1.0848344564437866,
+      "learning_rate": 5.754475703324808e-05,
+      "loss": 7.7622,
+      "step": 76
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 1.0727125406265259,
+      "learning_rate": 5.8312020460358044e-05,
+      "loss": 7.6749,
+      "step": 77
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 1.0996843576431274,
+      "learning_rate": 5.9079283887468026e-05,
+      "loss": 7.7144,
+      "step": 78
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 1.086287498474121,
+      "learning_rate": 5.9846547314578e-05,
+      "loss": 7.644,
+      "step": 79
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.9513853192329407,
+      "learning_rate": 6.0613810741687974e-05,
+      "loss": 7.6277,
+      "step": 80
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 1.0651500225067139,
+      "learning_rate": 6.138107416879795e-05,
+      "loss": 7.6083,
+      "step": 81
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 1.1370134353637695,
+      "learning_rate": 6.214833759590792e-05,
+      "loss": 7.5705,
+      "step": 82
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 1.0704236030578613,
+      "learning_rate": 6.29156010230179e-05,
+      "loss": 7.5459,
+      "step": 83
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 1.046200156211853,
+      "learning_rate": 6.368286445012787e-05,
+      "loss": 7.5118,
+      "step": 84
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.8698644638061523,
+      "learning_rate": 6.445012787723785e-05,
+      "loss": 7.4862,
+      "step": 85
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 1.1209050416946411,
+      "learning_rate": 6.521739130434782e-05,
+      "loss": 7.485,
+      "step": 86
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 1.3683184385299683,
+      "learning_rate": 6.59846547314578e-05,
+      "loss": 7.4883,
+      "step": 87
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.9742879867553711,
+      "learning_rate": 6.675191815856777e-05,
+      "loss": 7.3843,
+      "step": 88
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 1.3988070487976074,
+      "learning_rate": 6.751918158567774e-05,
+      "loss": 7.394,
+      "step": 89
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 1.1061723232269287,
+      "learning_rate": 6.828644501278772e-05,
+      "loss": 7.3124,
+      "step": 90
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 1.0231846570968628,
+      "learning_rate": 6.905370843989769e-05,
+      "loss": 7.3708,
+      "step": 91
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 1.3509619235992432,
+      "learning_rate": 6.982097186700767e-05,
+      "loss": 7.3974,
+      "step": 92
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.8934109807014465,
+      "learning_rate": 7.058823529411764e-05,
+      "loss": 7.3336,
+      "step": 93
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.7775417566299438,
+      "learning_rate": 7.135549872122762e-05,
+      "loss": 7.3351,
+      "step": 94
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 1.0462260246276855,
+      "learning_rate": 7.212276214833759e-05,
+      "loss": 7.3454,
+      "step": 95
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.9035119414329529,
+      "learning_rate": 7.289002557544756e-05,
+      "loss": 7.2667,
+      "step": 96
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.8907987475395203,
+      "learning_rate": 7.365728900255754e-05,
+      "loss": 7.2154,
+      "step": 97
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.9043131470680237,
+      "learning_rate": 7.442455242966751e-05,
+      "loss": 7.2902,
+      "step": 98
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.9289432764053345,
+      "learning_rate": 7.519181585677749e-05,
+      "loss": 7.2135,
+      "step": 99
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.6835809350013733,
+      "learning_rate": 7.595907928388747e-05,
+      "loss": 7.1967,
+      "step": 100
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 1.0327857732772827,
+      "learning_rate": 7.672634271099744e-05,
+      "loss": 7.1981,
+      "step": 101
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.8814536929130554,
+      "learning_rate": 7.749360613810741e-05,
+      "loss": 7.1923,
+      "step": 102
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.8455113172531128,
+      "learning_rate": 7.826086956521738e-05,
+      "loss": 7.2318,
+      "step": 103
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 1.2715808153152466,
+      "learning_rate": 7.902813299232736e-05,
+      "loss": 7.1884,
+      "step": 104
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.904243528842926,
+      "learning_rate": 7.979539641943735e-05,
+      "loss": 7.1839,
+      "step": 105
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.9426031112670898,
+      "learning_rate": 8.056265984654731e-05,
+      "loss": 7.1849,
+      "step": 106
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.7906021475791931,
+      "learning_rate": 8.132992327365728e-05,
+      "loss": 7.1946,
+      "step": 107
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.8392283320426941,
+      "learning_rate": 8.209718670076726e-05,
+      "loss": 7.1438,
+      "step": 108
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.7050894498825073,
+      "learning_rate": 8.286445012787723e-05,
+      "loss": 7.1547,
+      "step": 109
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.8787212371826172,
+      "learning_rate": 8.363171355498722e-05,
+      "loss": 7.1193,
+      "step": 110
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.5892102122306824,
+      "learning_rate": 8.439897698209718e-05,
+      "loss": 7.0926,
+      "step": 111
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 1.2569445371627808,
+      "learning_rate": 8.516624040920715e-05,
+      "loss": 7.1459,
+      "step": 112
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.6565424799919128,
+      "learning_rate": 8.593350383631713e-05,
+      "loss": 7.0536,
+      "step": 113
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.875479519367218,
+      "learning_rate": 8.670076726342709e-05,
+      "loss": 7.1119,
+      "step": 114
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.6395849585533142,
+      "learning_rate": 8.746803069053708e-05,
+      "loss": 7.0775,
+      "step": 115
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.7110955715179443,
+      "learning_rate": 8.823529411764705e-05,
+      "loss": 7.0741,
+      "step": 116
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.6394131779670715,
+      "learning_rate": 8.900255754475703e-05,
+      "loss": 7.0442,
+      "step": 117
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.9272814393043518,
+      "learning_rate": 8.9769820971867e-05,
+      "loss": 7.0344,
+      "step": 118
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.858862578868866,
+      "learning_rate": 9.053708439897696e-05,
+      "loss": 7.0646,
+      "step": 119
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 1.1841069459915161,
+      "learning_rate": 9.130434782608695e-05,
+      "loss": 7.0339,
+      "step": 120
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.9319404363632202,
+      "learning_rate": 9.207161125319692e-05,
+      "loss": 7.0008,
+      "step": 121
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.8237971067428589,
+      "learning_rate": 9.28388746803069e-05,
+      "loss": 7.0618,
+      "step": 122
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 1.379276156425476,
+      "learning_rate": 9.360613810741687e-05,
+      "loss": 6.9875,
+      "step": 123
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.9294237494468689,
+      "learning_rate": 9.437340153452683e-05,
+      "loss": 6.9856,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9710653424263,
+      "learning_rate": 9.514066496163682e-05,
+      "loss": 6.9368,
+      "step": 125
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 1.0589849948883057,
+      "learning_rate": 9.59079283887468e-05,
+      "loss": 6.9884,
+      "step": 126
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.6673716902732849,
+      "learning_rate": 9.667519181585677e-05,
+      "loss": 6.9719,
+      "step": 127
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.7187034487724304,
+      "learning_rate": 9.744245524296674e-05,
+      "loss": 6.9662,
+      "step": 128
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.7513381242752075,
+      "learning_rate": 9.820971867007673e-05,
+      "loss": 6.8949,
+      "step": 129
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.6384944319725037,
+      "learning_rate": 9.897698209718669e-05,
+      "loss": 6.9296,
+      "step": 130
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.6383824348449707,
+      "learning_rate": 9.974424552429667e-05,
+      "loss": 6.956,
+      "step": 131
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.8002307415008545,
+      "learning_rate": 0.00010051150895140664,
+      "loss": 6.9135,
+      "step": 132
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.6587149500846863,
+      "learning_rate": 0.00010127877237851662,
+      "loss": 6.8886,
+      "step": 133
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.7700603604316711,
+      "learning_rate": 0.0001020460358056266,
+      "loss": 6.9318,
+      "step": 134
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.7758870720863342,
+      "learning_rate": 0.00010281329923273656,
+      "loss": 6.9266,
+      "step": 135
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.818088948726654,
+      "learning_rate": 0.00010358056265984654,
+      "loss": 6.8917,
+      "step": 136
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 1.1521923542022705,
+      "learning_rate": 0.00010434782608695651,
+      "loss": 6.9264,
+      "step": 137
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.7598437070846558,
+      "learning_rate": 0.00010511508951406649,
+      "loss": 6.8756,
+      "step": 138
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.8913196921348572,
+      "learning_rate": 0.00010588235294117647,
+      "loss": 6.8535,
+      "step": 139
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 1.0058693885803223,
+      "learning_rate": 0.00010664961636828644,
+      "loss": 6.88,
+      "step": 140
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.9338283538818359,
+      "learning_rate": 0.00010741687979539641,
+      "loss": 6.8383,
+      "step": 141
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.6229548454284668,
+      "learning_rate": 0.00010818414322250638,
+      "loss": 6.7874,
+      "step": 142
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.7639989256858826,
+      "learning_rate": 0.00010895140664961635,
+      "loss": 6.7625,
+      "step": 143
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.6291145086288452,
+      "learning_rate": 0.00010971867007672633,
+      "loss": 6.7944,
+      "step": 144
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.8313521146774292,
+      "learning_rate": 0.00011048593350383631,
+      "loss": 6.859,
+      "step": 145
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 1.0272287130355835,
+      "learning_rate": 0.00011125319693094628,
+      "loss": 6.8568,
+      "step": 146
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 1.1838618516921997,
+      "learning_rate": 0.00011202046035805626,
+      "loss": 6.7789,
+      "step": 147
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.5190697312355042,
+      "learning_rate": 0.00011278772378516622,
+      "loss": 6.8148,
+      "step": 148
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.7801781892776489,
+      "learning_rate": 0.0001135549872122762,
+      "loss": 6.8162,
+      "step": 149
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.8845781087875366,
+      "learning_rate": 0.00011432225063938618,
+      "loss": 6.7817,
+      "step": 150
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.7235129475593567,
+      "learning_rate": 0.00011508951406649615,
+      "loss": 6.8149,
+      "step": 151
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.827165961265564,
+      "learning_rate": 0.00011585677749360613,
+      "loss": 6.6936,
+      "step": 152
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.6322879791259766,
+      "learning_rate": 0.00011662404092071609,
+      "loss": 6.7753,
+      "step": 153
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.6085387468338013,
+      "learning_rate": 0.00011739130434782608,
+      "loss": 6.7522,
+      "step": 154
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.5771993398666382,
+      "learning_rate": 0.00011815856777493605,
+      "loss": 6.7441,
+      "step": 155
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.6479660272598267,
+      "learning_rate": 0.00011892583120204603,
+      "loss": 6.795,
+      "step": 156
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.7639020085334778,
+      "learning_rate": 0.000119693094629156,
+      "loss": 6.7089,
+      "step": 157
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.7963739037513733,
+      "learning_rate": 0.00012046035805626599,
+      "loss": 6.7321,
+      "step": 158
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.8144316077232361,
+      "learning_rate": 0.00012122762148337595,
+      "loss": 6.7284,
+      "step": 159
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.8595559000968933,
+      "learning_rate": 0.00012199488491048592,
+      "loss": 6.7329,
+      "step": 160
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.8563242554664612,
+      "learning_rate": 0.0001227621483375959,
+      "loss": 6.7072,
+      "step": 161
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.9249243140220642,
+      "learning_rate": 0.00012352941176470587,
+      "loss": 6.7389,
+      "step": 162
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.7681375741958618,
+      "learning_rate": 0.00012429667519181585,
+      "loss": 6.6333,
+      "step": 163
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.8071373105049133,
+      "learning_rate": 0.00012506393861892582,
+      "loss": 6.7132,
+      "step": 164
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.7506120204925537,
+      "learning_rate": 0.0001258312020460358,
+      "loss": 6.6808,
+      "step": 165
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.8336671590805054,
+      "learning_rate": 0.00012659846547314577,
+      "loss": 6.7015,
+      "step": 166
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.9540615677833557,
+      "learning_rate": 0.00012736572890025574,
+      "loss": 6.6975,
+      "step": 167
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.8073275089263916,
+      "learning_rate": 0.00012813299232736572,
+      "loss": 6.6884,
+      "step": 168
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.6824653148651123,
+      "learning_rate": 0.0001289002557544757,
+      "loss": 6.6723,
+      "step": 169
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.5908713340759277,
+      "learning_rate": 0.00012966751918158567,
+      "loss": 6.6807,
+      "step": 170
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.7622809410095215,
+      "learning_rate": 0.00013043478260869564,
+      "loss": 6.6472,
+      "step": 171
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.6623175740242004,
+      "learning_rate": 0.00013120204603580562,
+      "loss": 6.6533,
+      "step": 172
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.7238216400146484,
+      "learning_rate": 0.0001319693094629156,
+      "loss": 6.654,
+      "step": 173
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.6670112013816833,
+      "learning_rate": 0.00013273657289002556,
+      "loss": 6.6262,
+      "step": 174
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.6878064274787903,
+      "learning_rate": 0.00013350383631713554,
+      "loss": 6.6101,
+      "step": 175
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.7931588292121887,
+      "learning_rate": 0.0001342710997442455,
+      "loss": 6.6616,
+      "step": 176
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 1.0632978677749634,
+      "learning_rate": 0.0001350383631713555,
+      "loss": 6.6203,
+      "step": 177
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.9125176668167114,
+      "learning_rate": 0.00013580562659846546,
+      "loss": 6.5786,
+      "step": 178
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.6771340370178223,
+      "learning_rate": 0.00013657289002557544,
+      "loss": 6.6114,
+      "step": 179
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.7708578109741211,
+      "learning_rate": 0.0001373401534526854,
+      "loss": 6.663,
+      "step": 180
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.7952269315719604,
+      "learning_rate": 0.00013810741687979538,
+      "loss": 6.5768,
+      "step": 181
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.9929698705673218,
+      "learning_rate": 0.00013887468030690536,
+      "loss": 6.5578,
+      "step": 182
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 1.016129493713379,
+      "learning_rate": 0.00013964194373401533,
+      "loss": 6.6065,
+      "step": 183
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 1.0715256929397583,
+      "learning_rate": 0.0001404092071611253,
+      "loss": 6.6292,
+      "step": 184
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 1.1549087762832642,
+      "learning_rate": 0.00014117647058823528,
+      "loss": 6.5751,
+      "step": 185
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.736225962638855,
+      "learning_rate": 0.00014194373401534526,
+      "loss": 6.5302,
+      "step": 186
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.6689443588256836,
+      "learning_rate": 0.00014271099744245523,
+      "loss": 6.6393,
+      "step": 187
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.9276636838912964,
+      "learning_rate": 0.0001434782608695652,
+      "loss": 6.5304,
+      "step": 188
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 1.093260645866394,
+      "learning_rate": 0.00014424552429667518,
+      "loss": 6.5826,
+      "step": 189
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.9400092959403992,
+      "learning_rate": 0.00014501278772378515,
+      "loss": 6.5214,
+      "step": 190
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.9401909708976746,
+      "learning_rate": 0.00014578005115089513,
+      "loss": 6.5743,
+      "step": 191
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.8417365550994873,
+      "learning_rate": 0.0001465473145780051,
+      "loss": 6.5711,
+      "step": 192
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.8696411848068237,
+      "learning_rate": 0.00014731457800511508,
+      "loss": 6.5372,
+      "step": 193
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.696698784828186,
+      "learning_rate": 0.00014808184143222505,
+      "loss": 6.5471,
+      "step": 194
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.8627312779426575,
+      "learning_rate": 0.00014884910485933503,
+      "loss": 6.495,
+      "step": 195
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.8102883100509644,
+      "learning_rate": 0.000149616368286445,
+      "loss": 6.5341,
+      "step": 196
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.8614912629127502,
+      "learning_rate": 0.00015038363171355497,
+      "loss": 6.5825,
+      "step": 197
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.6828194260597229,
+      "learning_rate": 0.00015115089514066495,
+      "loss": 6.5386,
+      "step": 198
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.6624521613121033,
+      "learning_rate": 0.00015191815856777495,
+      "loss": 6.4857,
+      "step": 199
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.6489097476005554,
+      "learning_rate": 0.0001526854219948849,
+      "loss": 6.5139,
+      "step": 200
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.6855762004852295,
+      "learning_rate": 0.00015345268542199487,
+      "loss": 6.4764,
+      "step": 201
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.8213603496551514,
+      "learning_rate": 0.00015421994884910485,
+      "loss": 6.5242,
+      "step": 202
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.824782133102417,
+      "learning_rate": 0.00015498721227621482,
+      "loss": 6.5478,
+      "step": 203
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.9565497636795044,
+      "learning_rate": 0.00015575447570332482,
+      "loss": 6.5082,
+      "step": 204
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.9202224612236023,
+      "learning_rate": 0.00015652173913043477,
+      "loss": 6.4114,
+      "step": 205
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.992260217666626,
+      "learning_rate": 0.00015728900255754474,
+      "loss": 6.4625,
+      "step": 206
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.8465138077735901,
+      "learning_rate": 0.00015805626598465472,
+      "loss": 6.439,
+      "step": 207
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.7151897549629211,
+      "learning_rate": 0.0001588235294117647,
+      "loss": 6.4956,
+      "step": 208
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.8685120940208435,
+      "learning_rate": 0.0001595907928388747,
+      "loss": 6.4803,
+      "step": 209
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.81340491771698,
+      "learning_rate": 0.00016035805626598464,
+      "loss": 6.4193,
+      "step": 210
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.7921631932258606,
+      "learning_rate": 0.00016112531969309462,
+      "loss": 6.4279,
+      "step": 211
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.6297151446342468,
+      "learning_rate": 0.0001618925831202046,
+      "loss": 6.5347,
+      "step": 212
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.6811320781707764,
+      "learning_rate": 0.00016265984654731456,
+      "loss": 6.4414,
+      "step": 213
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.6986665725708008,
+      "learning_rate": 0.00016342710997442457,
+      "loss": 6.4284,
+      "step": 214
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.6655412316322327,
+      "learning_rate": 0.0001641943734015345,
+      "loss": 6.4398,
+      "step": 215
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.6471274495124817,
+      "learning_rate": 0.0001649616368286445,
+      "loss": 6.4229,
+      "step": 216
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.7184582948684692,
+      "learning_rate": 0.00016572890025575446,
+      "loss": 6.4255,
+      "step": 217
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.7616591453552246,
+      "learning_rate": 0.00016649616368286444,
+      "loss": 6.4472,
+      "step": 218
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.6204221248626709,
+      "learning_rate": 0.00016726342710997444,
+      "loss": 6.3404,
+      "step": 219
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.7307862639427185,
+      "learning_rate": 0.00016803069053708438,
+      "loss": 6.4221,
+      "step": 220
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.6093372702598572,
+      "learning_rate": 0.00016879795396419436,
+      "loss": 6.3769,
+      "step": 221
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.7051405906677246,
+      "learning_rate": 0.00016956521739130433,
+      "loss": 6.4473,
+      "step": 222
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.6887394785881042,
+      "learning_rate": 0.0001703324808184143,
+      "loss": 6.4528,
+      "step": 223
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.7406412959098816,
+      "learning_rate": 0.00017109974424552428,
+      "loss": 6.375,
+      "step": 224
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.6635006666183472,
+      "learning_rate": 0.00017186700767263426,
+      "loss": 6.4649,
+      "step": 225
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.7995834946632385,
+      "learning_rate": 0.00017263427109974423,
+      "loss": 6.3737,
+      "step": 226
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.9514179825782776,
+      "learning_rate": 0.00017340153452685418,
+      "loss": 6.3449,
+      "step": 227
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 1.0529365539550781,
+      "learning_rate": 0.00017416879795396418,
+      "loss": 6.3666,
+      "step": 228
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 1.0295612812042236,
+      "learning_rate": 0.00017493606138107415,
+      "loss": 6.4025,
+      "step": 229
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.07936692237854,
+      "learning_rate": 0.00017570332480818413,
+      "loss": 6.3812,
+      "step": 230
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.9472998976707458,
+      "learning_rate": 0.0001764705882352941,
+      "loss": 6.3867,
+      "step": 231
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.6875970959663391,
+      "learning_rate": 0.00017723785166240405,
+      "loss": 6.4047,
+      "step": 232
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.8679647445678711,
+      "learning_rate": 0.00017800511508951405,
+      "loss": 6.3743,
+      "step": 233
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.9203290939331055,
+      "learning_rate": 0.00017877237851662403,
+      "loss": 6.3606,
+      "step": 234
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.8777357935905457,
+      "learning_rate": 0.000179539641943734,
+      "loss": 6.3564,
+      "step": 235
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.8760378956794739,
+      "learning_rate": 0.00018030690537084397,
+      "loss": 6.3576,
+      "step": 236
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.7797200083732605,
+      "learning_rate": 0.00018107416879795392,
+      "loss": 6.2981,
+      "step": 237
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.7287745475769043,
+      "learning_rate": 0.00018184143222506392,
+      "loss": 6.342,
+      "step": 238
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.7142657041549683,
+      "learning_rate": 0.0001826086956521739,
+      "loss": 6.4097,
+      "step": 239
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.8947247266769409,
+      "learning_rate": 0.00018337595907928387,
+      "loss": 6.4198,
+      "step": 240
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.7610101103782654,
+      "learning_rate": 0.00018414322250639385,
+      "loss": 6.3638,
+      "step": 241
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.7212010622024536,
+      "learning_rate": 0.0001849104859335038,
+      "loss": 6.3509,
+      "step": 242
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.711169421672821,
+      "learning_rate": 0.0001856777493606138,
+      "loss": 6.2656,
+      "step": 243
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.6636462807655334,
+      "learning_rate": 0.00018644501278772377,
+      "loss": 6.3338,
+      "step": 244
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.6644899249076843,
+      "learning_rate": 0.00018721227621483374,
+      "loss": 6.3888,
+      "step": 245
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.8558899760246277,
+      "learning_rate": 0.00018797953964194372,
+      "loss": 6.383,
+      "step": 246
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.8236832618713379,
+      "learning_rate": 0.00018874680306905366,
+      "loss": 6.3142,
+      "step": 247
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.6856957674026489,
+      "learning_rate": 0.00018951406649616367,
+      "loss": 6.3231,
+      "step": 248
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.8850679993629456,
+      "learning_rate": 0.00019028132992327364,
+      "loss": 6.334,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9059402942657471,
+      "learning_rate": 0.00019104859335038361,
+      "loss": 6.3826,
+      "step": 250
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.7600975036621094,
+      "learning_rate": 0.0001918158567774936,
+      "loss": 6.3067,
+      "step": 251
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.73809814453125,
+      "learning_rate": 0.00019258312020460354,
+      "loss": 6.3276,
+      "step": 252
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.7253278493881226,
+      "learning_rate": 0.00019335038363171354,
+      "loss": 6.282,
+      "step": 253
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.6514863967895508,
+      "learning_rate": 0.0001941176470588235,
+      "loss": 6.2885,
+      "step": 254
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.8405912518501282,
+      "learning_rate": 0.0001948849104859335,
+      "loss": 6.3166,
+      "step": 255
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 1.0162445306777954,
+      "learning_rate": 0.00019565217391304346,
+      "loss": 6.2927,
+      "step": 256
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 1.0011142492294312,
+      "learning_rate": 0.00019641943734015346,
+      "loss": 6.2595,
+      "step": 257
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.8047503232955933,
+      "learning_rate": 0.0001971867007672634,
+      "loss": 6.333,
+      "step": 258
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.7660624384880066,
+      "learning_rate": 0.00019795396419437338,
+      "loss": 6.2664,
+      "step": 259
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.7502520680427551,
+      "learning_rate": 0.00019872122762148336,
+      "loss": 6.2721,
+      "step": 260
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.9337821006774902,
+      "learning_rate": 0.00019948849104859333,
+      "loss": 6.3069,
+      "step": 261
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.6733538508415222,
+      "learning_rate": 0.00020025575447570333,
+      "loss": 6.221,
+      "step": 262
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.7917484045028687,
+      "learning_rate": 0.00020102301790281328,
+      "loss": 6.2338,
+      "step": 263
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.7609951496124268,
+      "learning_rate": 0.00020179028132992326,
+      "loss": 6.3031,
+      "step": 264
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.6764228940010071,
+      "learning_rate": 0.00020255754475703323,
+      "loss": 6.2349,
+      "step": 265
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.7090582847595215,
+      "learning_rate": 0.0002033248081841432,
+      "loss": 6.244,
+      "step": 266
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.7496779561042786,
+      "learning_rate": 0.0002040920716112532,
+      "loss": 6.2315,
+      "step": 267
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.987930417060852,
+      "learning_rate": 0.00020485933503836315,
+      "loss": 6.2223,
+      "step": 268
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.8948729634284973,
+      "learning_rate": 0.00020562659846547313,
+      "loss": 6.2722,
+      "step": 269
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7875823974609375,
+      "learning_rate": 0.0002063938618925831,
+      "loss": 6.228,
+      "step": 270
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.7225289940834045,
+      "learning_rate": 0.00020716112531969308,
+      "loss": 6.1819,
+      "step": 271
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.9579319953918457,
+      "learning_rate": 0.00020792838874680308,
+      "loss": 6.2365,
+      "step": 272
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 1.02460777759552,
+      "learning_rate": 0.00020869565217391303,
+      "loss": 6.2907,
+      "step": 273
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 1.275744915008545,
+      "learning_rate": 0.000209462915601023,
+      "loss": 6.259,
+      "step": 274
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.8093612194061279,
+      "learning_rate": 0.00021023017902813297,
+      "loss": 6.1701,
+      "step": 275
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.9731954336166382,
+      "learning_rate": 0.00021099744245524295,
+      "loss": 6.259,
+      "step": 276
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 1.0986355543136597,
+      "learning_rate": 0.00021176470588235295,
+      "loss": 6.1786,
+      "step": 277
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 1.1325825452804565,
+      "learning_rate": 0.0002125319693094629,
+      "loss": 6.253,
+      "step": 278
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.7239522337913513,
+      "learning_rate": 0.00021329923273657287,
+      "loss": 6.2677,
+      "step": 279
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.9211587905883789,
+      "learning_rate": 0.00021406649616368285,
+      "loss": 6.1859,
+      "step": 280
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.8542487621307373,
+      "learning_rate": 0.00021483375959079282,
+      "loss": 6.1534,
+      "step": 281
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.7646416425704956,
+      "learning_rate": 0.0002156010230179028,
+      "loss": 6.1181,
+      "step": 282
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.7755529880523682,
+      "learning_rate": 0.00021636828644501277,
+      "loss": 6.1676,
+      "step": 283
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.6631358861923218,
+      "learning_rate": 0.00021713554987212274,
+      "loss": 6.2041,
+      "step": 284
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.676986038684845,
+      "learning_rate": 0.0002179028132992327,
+      "loss": 6.178,
+      "step": 285
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.7331904172897339,
+      "learning_rate": 0.0002186700767263427,
+      "loss": 6.1775,
+      "step": 286
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.7227631211280823,
+      "learning_rate": 0.00021943734015345267,
+      "loss": 6.1277,
+      "step": 287
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.7674806118011475,
+      "learning_rate": 0.00022020460358056264,
+      "loss": 6.1654,
+      "step": 288
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.7369276881217957,
+      "learning_rate": 0.00022097186700767261,
+      "loss": 6.1768,
+      "step": 289
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.7614607214927673,
+      "learning_rate": 0.00022173913043478256,
+      "loss": 6.175,
+      "step": 290
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.6262795329093933,
+      "learning_rate": 0.00022250639386189256,
+      "loss": 6.1554,
+      "step": 291
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.7323906421661377,
+      "learning_rate": 0.00022327365728900254,
+      "loss": 6.1748,
+      "step": 292
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.6152296662330627,
+      "learning_rate": 0.0002240409207161125,
+      "loss": 6.1391,
+      "step": 293
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.672359824180603,
+      "learning_rate": 0.0002248081841432225,
+      "loss": 6.1529,
+      "step": 294
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.8039406538009644,
+      "learning_rate": 0.00022557544757033243,
+      "loss": 6.1391,
+      "step": 295
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.9391714334487915,
+      "learning_rate": 0.00022634271099744244,
+      "loss": 6.1234,
+      "step": 296
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.6732305884361267,
+      "learning_rate": 0.0002271099744245524,
+      "loss": 6.1335,
+      "step": 297
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.8230670094490051,
+      "learning_rate": 0.00022787723785166238,
+      "loss": 6.136,
+      "step": 298
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.5934028625488281,
+      "learning_rate": 0.00022864450127877236,
+      "loss": 6.1398,
+      "step": 299
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.8017310500144958,
+      "learning_rate": 0.0002294117647058823,
+      "loss": 6.1337,
+      "step": 300
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.7498769760131836,
+      "learning_rate": 0.0002301790281329923,
+      "loss": 6.1117,
+      "step": 301
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.727734386920929,
+      "learning_rate": 0.00023094629156010228,
+      "loss": 6.1219,
+      "step": 302
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.7669496536254883,
+      "learning_rate": 0.00023171355498721226,
+      "loss": 6.1272,
+      "step": 303
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.7476457357406616,
+      "learning_rate": 0.00023248081841432223,
+      "loss": 6.0301,
+      "step": 304
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.7540501952171326,
+      "learning_rate": 0.00023324808184143218,
+      "loss": 6.1143,
+      "step": 305
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 1.032804012298584,
+      "learning_rate": 0.00023401534526854218,
+      "loss": 6.1371,
+      "step": 306
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 1.1408971548080444,
+      "learning_rate": 0.00023478260869565215,
+      "loss": 6.0916,
+      "step": 307
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 1.027319312095642,
+      "learning_rate": 0.00023554987212276213,
+      "loss": 6.1407,
+      "step": 308
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.8574281930923462,
+      "learning_rate": 0.0002363171355498721,
+      "loss": 6.0626,
+      "step": 309
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.937127411365509,
+      "learning_rate": 0.0002370843989769821,
+      "loss": 6.1034,
+      "step": 310
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.7400676012039185,
+      "learning_rate": 0.00023785166240409205,
+      "loss": 6.0996,
+      "step": 311
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.7361345291137695,
+      "learning_rate": 0.00023861892583120203,
+      "loss": 6.047,
+      "step": 312
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.7439408898353577,
+      "learning_rate": 0.000239386189258312,
+      "loss": 6.0906,
+      "step": 313
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.7513951659202576,
+      "learning_rate": 0.00024015345268542197,
+      "loss": 6.0956,
+      "step": 314
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.7911155223846436,
+      "learning_rate": 0.00024092071611253198,
+      "loss": 6.0411,
+      "step": 315
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.7218300700187683,
+      "learning_rate": 0.00024168797953964192,
+      "loss": 6.1029,
+      "step": 316
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.8403254151344299,
+      "learning_rate": 0.0002424552429667519,
+      "loss": 6.0944,
+      "step": 317
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.676937460899353,
+      "learning_rate": 0.00024322250639386187,
+      "loss": 6.037,
+      "step": 318
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.8323513865470886,
+      "learning_rate": 0.00024398976982097185,
+      "loss": 6.0856,
+      "step": 319
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.9101656675338745,
+      "learning_rate": 0.0002447570332480818,
+      "loss": 6.089,
+      "step": 320
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.7529237270355225,
+      "learning_rate": 0.0002455242966751918,
+      "loss": 6.0525,
+      "step": 321
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.6977245211601257,
+      "learning_rate": 0.00024629156010230177,
+      "loss": 6.0561,
+      "step": 322
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.9390827417373657,
+      "learning_rate": 0.00024705882352941174,
+      "loss": 6.0536,
+      "step": 323
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.895576536655426,
+      "learning_rate": 0.0002478260869565217,
+      "loss": 6.0667,
+      "step": 324
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.9296811819076538,
+      "learning_rate": 0.0002485933503836317,
+      "loss": 6.0086,
+      "step": 325
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.9841163754463196,
+      "learning_rate": 0.00024936061381074167,
+      "loss": 6.0587,
+      "step": 326
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.7630248665809631,
+      "learning_rate": 0.00025012787723785164,
+      "loss": 5.9773,
+      "step": 327
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.8875618577003479,
+      "learning_rate": 0.0002508951406649616,
+      "loss": 6.0148,
+      "step": 328
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.8923934102058411,
+      "learning_rate": 0.0002516624040920716,
+      "loss": 6.0467,
+      "step": 329
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.8564996719360352,
+      "learning_rate": 0.00025242966751918156,
+      "loss": 6.0135,
+      "step": 330
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.803805410861969,
+      "learning_rate": 0.00025319693094629154,
+      "loss": 5.9922,
+      "step": 331
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.6739373803138733,
+      "learning_rate": 0.0002539641943734015,
+      "loss": 6.0465,
+      "step": 332
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.7415125966072083,
+      "learning_rate": 0.0002547314578005115,
+      "loss": 6.0081,
+      "step": 333
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.6618335247039795,
+      "learning_rate": 0.00025549872122762146,
+      "loss": 5.9794,
+      "step": 334
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.7585832476615906,
+      "learning_rate": 0.00025626598465473144,
+      "loss": 5.9738,
+      "step": 335
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.7203150987625122,
+      "learning_rate": 0.0002570332480818414,
+      "loss": 6.0329,
+      "step": 336
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.7902894020080566,
+      "learning_rate": 0.0002578005115089514,
+      "loss": 5.9299,
+      "step": 337
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.725581705570221,
+      "learning_rate": 0.00025856777493606136,
+      "loss": 5.9855,
+      "step": 338
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.8299371004104614,
+      "learning_rate": 0.00025933503836317133,
+      "loss": 6.0628,
+      "step": 339
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.7858306765556335,
+      "learning_rate": 0.0002601023017902813,
+      "loss": 6.0055,
+      "step": 340
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.8693034052848816,
+      "learning_rate": 0.0002608695652173913,
+      "loss": 6.032,
+      "step": 341
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.8078804612159729,
+      "learning_rate": 0.00026163682864450126,
+      "loss": 5.9117,
+      "step": 342
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.6977505683898926,
+      "learning_rate": 0.00026240409207161123,
+      "loss": 6.0052,
+      "step": 343
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.7459436058998108,
+      "learning_rate": 0.0002631713554987212,
+      "loss": 6.0064,
+      "step": 344
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.8525255918502808,
+      "learning_rate": 0.0002639386189258312,
+      "loss": 5.9702,
+      "step": 345
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.8468485474586487,
+      "learning_rate": 0.00026470588235294115,
+      "loss": 6.0015,
+      "step": 346
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.8215516805648804,
+      "learning_rate": 0.00026547314578005113,
+      "loss": 5.9744,
+      "step": 347
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.7234410643577576,
+      "learning_rate": 0.0002662404092071611,
+      "loss": 5.9851,
+      "step": 348
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.7426914572715759,
+      "learning_rate": 0.0002670076726342711,
+      "loss": 5.9737,
+      "step": 349
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.7684586048126221,
+      "learning_rate": 0.00026777493606138105,
+      "loss": 5.9324,
+      "step": 350
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.8180778622627258,
+      "learning_rate": 0.000268542199488491,
+      "loss": 5.8851,
+      "step": 351
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.7742370367050171,
+      "learning_rate": 0.000269309462915601,
+      "loss": 5.9474,
+      "step": 352
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.9160585403442383,
+      "learning_rate": 0.000270076726342711,
+      "loss": 5.958,
+      "step": 353
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.9427086710929871,
+      "learning_rate": 0.00027084398976982095,
+      "loss": 5.9811,
+      "step": 354
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 1.0058586597442627,
+      "learning_rate": 0.0002716112531969309,
+      "loss": 5.9625,
+      "step": 355
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 1.0322132110595703,
+      "learning_rate": 0.0002723785166240409,
+      "loss": 5.8985,
+      "step": 356
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.8881152868270874,
+      "learning_rate": 0.00027314578005115087,
+      "loss": 6.0119,
+      "step": 357
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.7439785599708557,
+      "learning_rate": 0.00027391304347826085,
+      "loss": 5.9378,
+      "step": 358
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.8855685591697693,
+      "learning_rate": 0.0002746803069053708,
+      "loss": 5.9244,
+      "step": 359
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.6977657079696655,
+      "learning_rate": 0.0002754475703324808,
+      "loss": 5.9455,
+      "step": 360
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.6967872381210327,
+      "learning_rate": 0.00027621483375959077,
+      "loss": 5.9463,
+      "step": 361
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.6706664562225342,
+      "learning_rate": 0.00027698209718670074,
+      "loss": 5.9024,
+      "step": 362
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 1.084336757659912,
+      "learning_rate": 0.0002777493606138107,
+      "loss": 5.864,
+      "step": 363
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 1.0996636152267456,
+      "learning_rate": 0.0002785166240409207,
+      "loss": 5.9009,
+      "step": 364
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.9186033606529236,
+      "learning_rate": 0.00027928388746803067,
+      "loss": 5.8365,
+      "step": 365
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.8409336805343628,
+      "learning_rate": 0.00028005115089514064,
+      "loss": 5.908,
+      "step": 366
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.8734238147735596,
+      "learning_rate": 0.0002808184143222506,
+      "loss": 5.8922,
+      "step": 367
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 1.0424097776412964,
+      "learning_rate": 0.0002815856777493606,
+      "loss": 5.975,
+      "step": 368
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 1.0325592756271362,
+      "learning_rate": 0.00028235294117647056,
+      "loss": 5.8882,
+      "step": 369
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.8004043102264404,
+      "learning_rate": 0.00028312020460358054,
+      "loss": 5.8663,
+      "step": 370
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.6563780307769775,
+      "learning_rate": 0.0002838874680306905,
+      "loss": 5.8704,
+      "step": 371
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.7133100032806396,
+      "learning_rate": 0.0002846547314578005,
+      "loss": 5.8959,
+      "step": 372
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.656963050365448,
+      "learning_rate": 0.00028542199488491046,
+      "loss": 5.866,
+      "step": 373
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.761816143989563,
+      "learning_rate": 0.00028618925831202044,
+      "loss": 5.8776,
+      "step": 374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.900486171245575,
+      "learning_rate": 0.0002869565217391304,
+      "loss": 5.9083,
+      "step": 375
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 1.0122328996658325,
+      "learning_rate": 0.0002877237851662404,
+      "loss": 5.8945,
+      "step": 376
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.9674193859100342,
+      "learning_rate": 0.00028849104859335036,
+      "loss": 5.8972,
+      "step": 377
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.7003923058509827,
+      "learning_rate": 0.00028925831202046033,
+      "loss": 5.8883,
+      "step": 378
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.7299737930297852,
+      "learning_rate": 0.0002900255754475703,
+      "loss": 5.8642,
+      "step": 379
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.68450927734375,
+      "learning_rate": 0.0002907928388746803,
+      "loss": 5.8051,
+      "step": 380
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.744691789150238,
+      "learning_rate": 0.00029156010230179026,
+      "loss": 5.8283,
+      "step": 381
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.8168903589248657,
+      "learning_rate": 0.00029232736572890023,
+      "loss": 5.8316,
+      "step": 382
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 1.2333744764328003,
+      "learning_rate": 0.0002930946291560102,
+      "loss": 5.8244,
+      "step": 383
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.9197435975074768,
+      "learning_rate": 0.0002938618925831202,
+      "loss": 5.8464,
+      "step": 384
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 1.0070711374282837,
+      "learning_rate": 0.00029462915601023015,
+      "loss": 5.8809,
+      "step": 385
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.9315182566642761,
+      "learning_rate": 0.00029539641943734013,
+      "loss": 5.8207,
+      "step": 386
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.7878130078315735,
+      "learning_rate": 0.0002961636828644501,
+      "loss": 5.8888,
+      "step": 387
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.8333479762077332,
+      "learning_rate": 0.0002969309462915601,
+      "loss": 5.8616,
+      "step": 388
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.8158892393112183,
+      "learning_rate": 0.00029769820971867005,
+      "loss": 5.8188,
+      "step": 389
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.6871780157089233,
+      "learning_rate": 0.00029846547314578,
+      "loss": 5.8525,
+      "step": 390
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.7237544059753418,
+      "learning_rate": 0.00029923273657289,
+      "loss": 5.8758,
+      "step": 391
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.6453325152397156,
+      "learning_rate": 0.0003,
+      "loss": 5.7895,
+      "step": 392
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.5974975824356079,
+      "learning_rate": 0.0002999596177143626,
+      "loss": 5.8014,
+      "step": 393
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.6222047209739685,
+      "learning_rate": 0.00029991923542872523,
+      "loss": 5.7825,
+      "step": 394
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.6509471535682678,
+      "learning_rate": 0.00029987885314308786,
+      "loss": 5.806,
+      "step": 395
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.6112677454948425,
+      "learning_rate": 0.0002998384708574505,
+      "loss": 5.7888,
+      "step": 396
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.67631596326828,
+      "learning_rate": 0.0002997980885718131,
+      "loss": 5.7826,
+      "step": 397
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.6522372961044312,
+      "learning_rate": 0.00029975770628617575,
+      "loss": 5.7987,
+      "step": 398
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.5814371109008789,
+      "learning_rate": 0.00029971732400053844,
+      "loss": 5.7732,
+      "step": 399
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.7316078543663025,
+      "learning_rate": 0.00029967694171490107,
+      "loss": 5.7616,
+      "step": 400
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.817568302154541,
+      "learning_rate": 0.00029963655942926364,
+      "loss": 5.8326,
+      "step": 401
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.6369678378105164,
+      "learning_rate": 0.0002995961771436263,
+      "loss": 5.7762,
+      "step": 402
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.6471033096313477,
+      "learning_rate": 0.00029955579485798896,
+      "loss": 5.7027,
+      "step": 403
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.6128319501876831,
+      "learning_rate": 0.0002995154125723516,
+      "loss": 5.754,
+      "step": 404
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.7475748062133789,
+      "learning_rate": 0.0002994750302867142,
+      "loss": 5.7301,
+      "step": 405
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.6675015091896057,
+      "learning_rate": 0.00029943464800107685,
+      "loss": 5.7793,
+      "step": 406
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.640458345413208,
+      "learning_rate": 0.0002993942657154395,
+      "loss": 5.735,
+      "step": 407
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.6345043778419495,
+      "learning_rate": 0.0002993538834298021,
+      "loss": 5.6872,
+      "step": 408
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.6826564073562622,
+      "learning_rate": 0.00029931350114416474,
+      "loss": 5.691,
+      "step": 409
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.576991856098175,
+      "learning_rate": 0.00029927311885852737,
+      "loss": 5.7254,
+      "step": 410
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.572139322757721,
+      "learning_rate": 0.00029923273657289,
+      "loss": 5.7618,
+      "step": 411
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.7463775277137756,
+      "learning_rate": 0.00029919235428725263,
+      "loss": 5.7217,
+      "step": 412
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.7591479420661926,
+      "learning_rate": 0.00029915197200161526,
+      "loss": 5.745,
+      "step": 413
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.7695474624633789,
+      "learning_rate": 0.0002991115897159779,
+      "loss": 5.7294,
+      "step": 414
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.704517126083374,
+      "learning_rate": 0.0002990712074303405,
+      "loss": 5.7123,
+      "step": 415
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.7229870557785034,
+      "learning_rate": 0.0002990308251447032,
+      "loss": 5.6996,
+      "step": 416
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.7659012079238892,
+      "learning_rate": 0.0002989904428590658,
+      "loss": 5.7188,
+      "step": 417
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.8571960926055908,
+      "learning_rate": 0.0002989500605734284,
+      "loss": 5.7277,
+      "step": 418
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.9707953929901123,
+      "learning_rate": 0.00029890967828779104,
+      "loss": 5.668,
+      "step": 419
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.9887204766273499,
+      "learning_rate": 0.00029886929600215367,
+      "loss": 5.7147,
+      "step": 420
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.8890637159347534,
+      "learning_rate": 0.00029882891371651635,
+      "loss": 5.6983,
+      "step": 421
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.9526740908622742,
+      "learning_rate": 0.000298788531430879,
+      "loss": 5.7069,
+      "step": 422
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 1.1278778314590454,
+      "learning_rate": 0.0002987481491452416,
+      "loss": 5.6393,
+      "step": 423
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 1.0044233798980713,
+      "learning_rate": 0.0002987077668596042,
+      "loss": 5.7143,
+      "step": 424
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.9827817678451538,
+      "learning_rate": 0.0002986673845739669,
+      "loss": 5.7981,
+      "step": 425
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.8233329057693481,
+      "learning_rate": 0.0002986270022883295,
+      "loss": 5.7221,
+      "step": 426
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.7597877383232117,
+      "learning_rate": 0.00029858662000269213,
+      "loss": 5.698,
+      "step": 427
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.8371834754943848,
+      "learning_rate": 0.00029854623771705476,
+      "loss": 5.6721,
+      "step": 428
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.6564275026321411,
+      "learning_rate": 0.0002985058554314174,
+      "loss": 5.6973,
+      "step": 429
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.7076083421707153,
+      "learning_rate": 0.00029846547314578,
+      "loss": 5.6448,
+      "step": 430
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.7080224752426147,
+      "learning_rate": 0.00029842509086014266,
+      "loss": 5.6874,
+      "step": 431
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.792473554611206,
+      "learning_rate": 0.0002983847085745053,
+      "loss": 5.7225,
+      "step": 432
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 1.0753412246704102,
+      "learning_rate": 0.0002983443262888679,
+      "loss": 5.6673,
+      "step": 433
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 1.0193465948104858,
+      "learning_rate": 0.00029830394400323055,
+      "loss": 5.7159,
+      "step": 434
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 1.0127590894699097,
+      "learning_rate": 0.0002982635617175932,
+      "loss": 5.6806,
+      "step": 435
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 1.2227166891098022,
+      "learning_rate": 0.0002982231794319558,
+      "loss": 5.66,
+      "step": 436
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.7891542315483093,
+      "learning_rate": 0.00029818279714631844,
+      "loss": 5.6076,
+      "step": 437
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.7889094948768616,
+      "learning_rate": 0.0002981424148606811,
+      "loss": 5.6662,
+      "step": 438
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.7303580045700073,
+      "learning_rate": 0.00029810203257504375,
+      "loss": 5.6123,
+      "step": 439
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.7027185559272766,
+      "learning_rate": 0.0002980616502894063,
+      "loss": 5.6627,
+      "step": 440
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.6738660931587219,
+      "learning_rate": 0.00029802126800376896,
+      "loss": 5.6118,
+      "step": 441
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.7443022727966309,
+      "learning_rate": 0.00029798088571813164,
+      "loss": 5.629,
+      "step": 442
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.685570478439331,
+      "learning_rate": 0.00029794050343249427,
+      "loss": 5.6223,
+      "step": 443
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.672305703163147,
+      "learning_rate": 0.0002979001211468569,
+      "loss": 5.6039,
+      "step": 444
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.7415111064910889,
+      "learning_rate": 0.00029785973886121953,
+      "loss": 5.6167,
+      "step": 445
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.7740527987480164,
+      "learning_rate": 0.00029781935657558216,
+      "loss": 5.5913,
+      "step": 446
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.7420479655265808,
+      "learning_rate": 0.0002977789742899448,
+      "loss": 5.6269,
+      "step": 447
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.7080816626548767,
+      "learning_rate": 0.0002977385920043074,
+      "loss": 5.7041,
+      "step": 448
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.7158259153366089,
+      "learning_rate": 0.00029769820971867005,
+      "loss": 5.6587,
+      "step": 449
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.7243630290031433,
+      "learning_rate": 0.0002976578274330327,
+      "loss": 5.6178,
+      "step": 450
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.6919866800308228,
+      "learning_rate": 0.0002976174451473953,
+      "loss": 5.6674,
+      "step": 451
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.7445707321166992,
+      "learning_rate": 0.00029757706286175794,
+      "loss": 5.6193,
+      "step": 452
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.7470800876617432,
+      "learning_rate": 0.00029753668057612057,
+      "loss": 5.6655,
+      "step": 453
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.7181954383850098,
+      "learning_rate": 0.0002974962982904832,
+      "loss": 5.607,
+      "step": 454
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.7144619822502136,
+      "learning_rate": 0.0002974559160048459,
+      "loss": 5.6176,
+      "step": 455
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.6728463768959045,
+      "learning_rate": 0.00029741553371920846,
+      "loss": 5.5554,
+      "step": 456
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.7108640074729919,
+      "learning_rate": 0.0002973751514335711,
+      "loss": 5.5815,
+      "step": 457
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.5689119696617126,
+      "learning_rate": 0.0002973347691479337,
+      "loss": 5.5097,
+      "step": 458
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.6072651743888855,
+      "learning_rate": 0.0002972943868622964,
+      "loss": 5.5147,
+      "step": 459
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.5563207268714905,
+      "learning_rate": 0.00029725400457665904,
+      "loss": 5.5191,
+      "step": 460
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.6030489802360535,
+      "learning_rate": 0.00029721362229102167,
+      "loss": 5.5896,
+      "step": 461
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.6141897439956665,
+      "learning_rate": 0.0002971732400053843,
+      "loss": 5.6277,
+      "step": 462
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.6721770167350769,
+      "learning_rate": 0.00029713285771974687,
+      "loss": 5.5537,
+      "step": 463
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.6960611343383789,
+      "learning_rate": 0.00029709247543410956,
+      "loss": 5.569,
+      "step": 464
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.6458708047866821,
+      "learning_rate": 0.0002970520931484722,
+      "loss": 5.5418,
+      "step": 465
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.7103924751281738,
+      "learning_rate": 0.0002970117108628348,
+      "loss": 5.5574,
+      "step": 466
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.677963376045227,
+      "learning_rate": 0.00029697132857719745,
+      "loss": 5.5791,
+      "step": 467
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.6947761178016663,
+      "learning_rate": 0.0002969309462915601,
+      "loss": 5.5745,
+      "step": 468
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.6401717662811279,
+      "learning_rate": 0.0002968905640059227,
+      "loss": 5.5222,
+      "step": 469
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.6908291578292847,
+      "learning_rate": 0.00029685018172028534,
+      "loss": 5.5118,
+      "step": 470
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.7326011061668396,
+      "learning_rate": 0.00029680979943464797,
+      "loss": 5.5127,
+      "step": 471
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.8172990679740906,
+      "learning_rate": 0.00029676941714901065,
+      "loss": 5.5237,
+      "step": 472
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.7394744157791138,
+      "learning_rate": 0.0002967290348633732,
+      "loss": 5.4849,
+      "step": 473
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.6962910294532776,
+      "learning_rate": 0.00029668865257773586,
+      "loss": 5.5167,
+      "step": 474
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.6705038547515869,
+      "learning_rate": 0.0002966482702920985,
+      "loss": 5.5084,
+      "step": 475
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.8199055790901184,
+      "learning_rate": 0.0002966078880064611,
+      "loss": 5.5338,
+      "step": 476
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.833018958568573,
+      "learning_rate": 0.0002965675057208238,
+      "loss": 5.5654,
+      "step": 477
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.7582036852836609,
+      "learning_rate": 0.00029652712343518643,
+      "loss": 5.5814,
+      "step": 478
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.8647993803024292,
+      "learning_rate": 0.000296486741149549,
+      "loss": 5.5089,
+      "step": 479
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.1151838302612305,
+      "learning_rate": 0.00029644635886391164,
+      "loss": 5.5816,
+      "step": 480
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.8047844767570496,
+      "learning_rate": 0.0002964059765782743,
+      "loss": 5.5261,
+      "step": 481
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.7434077858924866,
+      "learning_rate": 0.00029636559429263695,
+      "loss": 5.5063,
+      "step": 482
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.7304471135139465,
+      "learning_rate": 0.0002963252120069996,
+      "loss": 5.5246,
+      "step": 483
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.686316728591919,
+      "learning_rate": 0.0002962848297213622,
+      "loss": 5.4783,
+      "step": 484
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.6505491733551025,
+      "learning_rate": 0.00029624444743572484,
+      "loss": 5.4989,
+      "step": 485
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.6548778414726257,
+      "learning_rate": 0.00029620406515008747,
+      "loss": 5.4819,
+      "step": 486
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.6001530289649963,
+      "learning_rate": 0.0002961636828644501,
+      "loss": 5.4859,
+      "step": 487
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.6683222055435181,
+      "learning_rate": 0.00029612330057881273,
+      "loss": 5.4872,
+      "step": 488
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.7493375539779663,
+      "learning_rate": 0.00029608291829317536,
+      "loss": 5.5236,
+      "step": 489
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.7252801060676575,
+      "learning_rate": 0.000296042536007538,
+      "loss": 5.4432,
+      "step": 490
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.6479570269584656,
+      "learning_rate": 0.0002960021537219006,
+      "loss": 5.4957,
+      "step": 491
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.6929423809051514,
+      "learning_rate": 0.00029596177143626325,
+      "loss": 5.4412,
+      "step": 492
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.6217756271362305,
+      "learning_rate": 0.0002959213891506259,
+      "loss": 5.4922,
+      "step": 493
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.7749668955802917,
+      "learning_rate": 0.00029588100686498857,
+      "loss": 5.5026,
+      "step": 494
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.7131249308586121,
+      "learning_rate": 0.0002958406245793512,
+      "loss": 5.4615,
+      "step": 495
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.6029372215270996,
+      "learning_rate": 0.00029580024229371377,
+      "loss": 5.5491,
+      "step": 496
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.8320589661598206,
+      "learning_rate": 0.0002957598600080764,
+      "loss": 5.4513,
+      "step": 497
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.8968410491943359,
+      "learning_rate": 0.0002957194777224391,
+      "loss": 5.5575,
+      "step": 498
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.8323341012001038,
+      "learning_rate": 0.0002956790954368017,
+      "loss": 5.4464,
+      "step": 499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7660730481147766,
+      "learning_rate": 0.00029563871315116435,
+      "loss": 5.5199,
+      "step": 500
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.8838502764701843,
+      "learning_rate": 0.000295598330865527,
+      "loss": 5.4954,
+      "step": 501
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.8735381960868835,
+      "learning_rate": 0.0002955579485798896,
+      "loss": 5.4527,
+      "step": 502
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.7608069777488708,
+      "learning_rate": 0.00029551756629425224,
+      "loss": 5.4398,
+      "step": 503
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.7376892566680908,
+      "learning_rate": 0.00029547718400861487,
+      "loss": 5.4521,
+      "step": 504
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.6925538778305054,
+      "learning_rate": 0.0002954368017229775,
+      "loss": 5.504,
+      "step": 505
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.6140308380126953,
+      "learning_rate": 0.00029539641943734013,
+      "loss": 5.4147,
+      "step": 506
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.6824973225593567,
+      "learning_rate": 0.00029535603715170276,
+      "loss": 5.4963,
+      "step": 507
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.736219584941864,
+      "learning_rate": 0.0002953156548660654,
+      "loss": 5.4566,
+      "step": 508
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.6932424902915955,
+      "learning_rate": 0.000295275272580428,
+      "loss": 5.3826,
+      "step": 509
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.6957204341888428,
+      "learning_rate": 0.00029523489029479065,
+      "loss": 5.4521,
+      "step": 510
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.5883374810218811,
+      "learning_rate": 0.00029519450800915333,
+      "loss": 5.4681,
+      "step": 511
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.6524292826652527,
+      "learning_rate": 0.0002951541257235159,
+      "loss": 5.3794,
+      "step": 512
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.7906764149665833,
+      "learning_rate": 0.00029511374343787854,
+      "loss": 5.443,
+      "step": 513
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.7420827746391296,
+      "learning_rate": 0.00029507336115224117,
+      "loss": 5.4216,
+      "step": 514
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.7884359955787659,
+      "learning_rate": 0.0002950329788666038,
+      "loss": 5.4701,
+      "step": 515
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.7779830694198608,
+      "learning_rate": 0.0002949925965809665,
+      "loss": 5.4779,
+      "step": 516
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.7066437602043152,
+      "learning_rate": 0.0002949522142953291,
+      "loss": 5.4453,
+      "step": 517
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.8100753426551819,
+      "learning_rate": 0.00029491183200969174,
+      "loss": 5.4251,
+      "step": 518
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.7143018245697021,
+      "learning_rate": 0.0002948714497240543,
+      "loss": 5.4786,
+      "step": 519
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.7291150689125061,
+      "learning_rate": 0.000294831067438417,
+      "loss": 5.4265,
+      "step": 520
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.6859623193740845,
+      "learning_rate": 0.00029479068515277963,
+      "loss": 5.4724,
+      "step": 521
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.6909458637237549,
+      "learning_rate": 0.00029475030286714226,
+      "loss": 5.3719,
+      "step": 522
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.6806208491325378,
+      "learning_rate": 0.0002947099205815049,
+      "loss": 5.3872,
+      "step": 523
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.6667022109031677,
+      "learning_rate": 0.0002946695382958675,
+      "loss": 5.3614,
+      "step": 524
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.6555402278900146,
+      "learning_rate": 0.00029462915601023015,
+      "loss": 5.4534,
+      "step": 525
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.685303270816803,
+      "learning_rate": 0.0002945887737245928,
+      "loss": 5.426,
+      "step": 526
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.7246394753456116,
+      "learning_rate": 0.0002945483914389554,
+      "loss": 5.4041,
+      "step": 527
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.765101969242096,
+      "learning_rate": 0.00029450800915331804,
+      "loss": 5.386,
+      "step": 528
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.7084211111068726,
+      "learning_rate": 0.0002944676268676807,
+      "loss": 5.3894,
+      "step": 529
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.6788263916969299,
+      "learning_rate": 0.0002944272445820433,
+      "loss": 5.375,
+      "step": 530
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.7256125807762146,
+      "learning_rate": 0.00029438686229640593,
+      "loss": 5.3544,
+      "step": 531
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.8374799489974976,
+      "learning_rate": 0.00029434648001076856,
+      "loss": 5.3758,
+      "step": 532
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.8988571166992188,
+      "learning_rate": 0.00029430609772513125,
+      "loss": 5.4003,
+      "step": 533
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.7412604093551636,
+      "learning_rate": 0.0002942657154394939,
+      "loss": 5.3558,
+      "step": 534
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.7604132890701294,
+      "learning_rate": 0.00029422533315385645,
+      "loss": 5.4506,
+      "step": 535
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.9959605932235718,
+      "learning_rate": 0.0002941849508682191,
+      "loss": 5.3877,
+      "step": 536
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.834484875202179,
+      "learning_rate": 0.00029414456858258177,
+      "loss": 5.3407,
+      "step": 537
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.6473014950752258,
+      "learning_rate": 0.0002941041862969444,
+      "loss": 5.3553,
+      "step": 538
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.8236899375915527,
+      "learning_rate": 0.00029406380401130703,
+      "loss": 5.3946,
+      "step": 539
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.7361433506011963,
+      "learning_rate": 0.00029402342172566966,
+      "loss": 5.3825,
+      "step": 540
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.5942022800445557,
+      "learning_rate": 0.0002939830394400323,
+      "loss": 5.3754,
+      "step": 541
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.6257084608078003,
+      "learning_rate": 0.0002939426571543949,
+      "loss": 5.3383,
+      "step": 542
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.5751544237136841,
+      "learning_rate": 0.00029390227486875755,
+      "loss": 5.3087,
+      "step": 543
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.6479839086532593,
+      "learning_rate": 0.0002938618925831202,
+      "loss": 5.4463,
+      "step": 544
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.6127558350563049,
+      "learning_rate": 0.0002938215102974828,
+      "loss": 5.4066,
+      "step": 545
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.6831675171852112,
+      "learning_rate": 0.00029378112801184544,
+      "loss": 5.3043,
+      "step": 546
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.69155353307724,
+      "learning_rate": 0.00029374074572620807,
+      "loss": 5.3776,
+      "step": 547
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.6620556116104126,
+      "learning_rate": 0.0002937003634405707,
+      "loss": 5.3669,
+      "step": 548
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.5906518697738647,
+      "learning_rate": 0.00029365998115493333,
+      "loss": 5.3356,
+      "step": 549
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.6667118668556213,
+      "learning_rate": 0.000293619598869296,
+      "loss": 5.327,
+      "step": 550
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.7207005023956299,
+      "learning_rate": 0.0002935792165836586,
+      "loss": 5.3675,
+      "step": 551
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.8886672854423523,
+      "learning_rate": 0.0002935388342980212,
+      "loss": 5.3172,
+      "step": 552
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.9381269216537476,
+      "learning_rate": 0.00029349845201238385,
+      "loss": 5.3404,
+      "step": 553
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.790340006351471,
+      "learning_rate": 0.0002934580697267465,
+      "loss": 5.3174,
+      "step": 554
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.636177659034729,
+      "learning_rate": 0.00029341768744110916,
+      "loss": 5.3767,
+      "step": 555
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.7448100447654724,
+      "learning_rate": 0.0002933773051554718,
+      "loss": 5.3201,
+      "step": 556
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.7365016341209412,
+      "learning_rate": 0.0002933369228698344,
+      "loss": 5.318,
+      "step": 557
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.6620506048202515,
+      "learning_rate": 0.000293296540584197,
+      "loss": 5.3417,
+      "step": 558
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.5934759378433228,
+      "learning_rate": 0.0002932561582985597,
+      "loss": 5.3751,
+      "step": 559
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.6874362230300903,
+      "learning_rate": 0.0002932157760129223,
+      "loss": 5.3089,
+      "step": 560
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.6296939253807068,
+      "learning_rate": 0.00029317539372728494,
+      "loss": 5.3057,
+      "step": 561
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.5342876315116882,
+      "learning_rate": 0.0002931350114416476,
+      "loss": 5.3055,
+      "step": 562
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.7313459515571594,
+      "learning_rate": 0.0002930946291560102,
+      "loss": 5.3002,
+      "step": 563
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.7934847474098206,
+      "learning_rate": 0.00029305424687037283,
+      "loss": 5.394,
+      "step": 564
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.6675227284431458,
+      "learning_rate": 0.00029301386458473546,
+      "loss": 5.2447,
+      "step": 565
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.6422837972640991,
+      "learning_rate": 0.0002929734822990981,
+      "loss": 5.3177,
+      "step": 566
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.6145399212837219,
+      "learning_rate": 0.0002929331000134607,
+      "loss": 5.3537,
+      "step": 567
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.5638757944107056,
+      "learning_rate": 0.00029289271772782335,
+      "loss": 5.2738,
+      "step": 568
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.6802723407745361,
+      "learning_rate": 0.000292852335442186,
+      "loss": 5.3411,
+      "step": 569
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.8496166467666626,
+      "learning_rate": 0.0002928119531565486,
+      "loss": 5.3137,
+      "step": 570
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.8245171904563904,
+      "learning_rate": 0.00029277157087091125,
+      "loss": 5.3141,
+      "step": 571
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.6584154963493347,
+      "learning_rate": 0.00029273118858527393,
+      "loss": 5.3264,
+      "step": 572
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.6775988936424255,
+      "learning_rate": 0.00029269080629963656,
+      "loss": 5.2959,
+      "step": 573
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.6198107600212097,
+      "learning_rate": 0.0002926504240139992,
+      "loss": 5.2696,
+      "step": 574
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.7053045034408569,
+      "learning_rate": 0.00029261004172836177,
+      "loss": 5.3243,
+      "step": 575
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.7388597130775452,
+      "learning_rate": 0.00029256965944272445,
+      "loss": 5.3298,
+      "step": 576
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.7207250595092773,
+      "learning_rate": 0.0002925292771570871,
+      "loss": 5.3113,
+      "step": 577
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.7020829319953918,
+      "learning_rate": 0.0002924888948714497,
+      "loss": 5.2813,
+      "step": 578
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.8059638738632202,
+      "learning_rate": 0.00029244851258581234,
+      "loss": 5.2436,
+      "step": 579
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 1.0253353118896484,
+      "learning_rate": 0.00029240813030017497,
+      "loss": 5.2739,
+      "step": 580
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.963039755821228,
+      "learning_rate": 0.0002923677480145376,
+      "loss": 5.2639,
+      "step": 581
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.7356200814247131,
+      "learning_rate": 0.00029232736572890023,
+      "loss": 5.2874,
+      "step": 582
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.7162445187568665,
+      "learning_rate": 0.00029228698344326286,
+      "loss": 5.3408,
+      "step": 583
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.708344042301178,
+      "learning_rate": 0.0002922466011576255,
+      "loss": 5.2798,
+      "step": 584
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.5786714553833008,
+      "learning_rate": 0.0002922062188719881,
+      "loss": 5.285,
+      "step": 585
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.6987271904945374,
+      "learning_rate": 0.00029216583658635075,
+      "loss": 5.2781,
+      "step": 586
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.6231822967529297,
+      "learning_rate": 0.0002921254543007134,
+      "loss": 5.2242,
+      "step": 587
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.679735004901886,
+      "learning_rate": 0.000292085072015076,
+      "loss": 5.3149,
+      "step": 588
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.6461718678474426,
+      "learning_rate": 0.0002920446897294387,
+      "loss": 5.2661,
+      "step": 589
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.6281011700630188,
+      "learning_rate": 0.0002920043074438013,
+      "loss": 5.1899,
+      "step": 590
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.6488833427429199,
+      "learning_rate": 0.0002919639251581639,
+      "loss": 5.2635,
+      "step": 591
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.7168968319892883,
+      "learning_rate": 0.00029192354287252653,
+      "loss": 5.1908,
+      "step": 592
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.7377144694328308,
+      "learning_rate": 0.0002918831605868892,
+      "loss": 5.2676,
+      "step": 593
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.5983268022537231,
+      "learning_rate": 0.00029184277830125185,
+      "loss": 5.26,
+      "step": 594
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.6470866203308105,
+      "learning_rate": 0.0002918023960156145,
+      "loss": 5.228,
+      "step": 595
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.7147027850151062,
+      "learning_rate": 0.0002917620137299771,
+      "loss": 5.2475,
+      "step": 596
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.686434268951416,
+      "learning_rate": 0.00029172163144433974,
+      "loss": 5.2174,
+      "step": 597
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.7053865194320679,
+      "learning_rate": 0.00029168124915870237,
+      "loss": 5.2822,
+      "step": 598
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.8161548972129822,
+      "learning_rate": 0.000291640866873065,
+      "loss": 5.1864,
+      "step": 599
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6877032518386841,
+      "learning_rate": 0.0002916004845874276,
+      "loss": 5.281,
+      "step": 600
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.5830395817756653,
+      "learning_rate": 0.00029156010230179026,
+      "loss": 5.2073,
+      "step": 601
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.7007346749305725,
+      "learning_rate": 0.0002915197200161529,
+      "loss": 5.3173,
+      "step": 602
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.7271237969398499,
+      "learning_rate": 0.0002914793377305155,
+      "loss": 5.235,
+      "step": 603
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.5913999080657959,
+      "learning_rate": 0.00029143895544487815,
+      "loss": 5.2552,
+      "step": 604
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.6722851991653442,
+      "learning_rate": 0.0002913985731592408,
+      "loss": 5.2112,
+      "step": 605
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.6968867182731628,
+      "learning_rate": 0.00029135819087360346,
+      "loss": 5.2505,
+      "step": 606
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.6645686030387878,
+      "learning_rate": 0.00029131780858796604,
+      "loss": 5.1846,
+      "step": 607
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.6209181547164917,
+      "learning_rate": 0.00029127742630232867,
+      "loss": 5.2269,
+      "step": 608
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.7518525123596191,
+      "learning_rate": 0.0002912370440166913,
+      "loss": 5.1761,
+      "step": 609
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.6878477334976196,
+      "learning_rate": 0.0002911966617310539,
+      "loss": 5.2348,
+      "step": 610
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.7975960969924927,
+      "learning_rate": 0.0002911562794454166,
+      "loss": 5.2519,
+      "step": 611
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.7792649269104004,
+      "learning_rate": 0.00029111589715977924,
+      "loss": 5.1812,
+      "step": 612
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.7435747385025024,
+      "learning_rate": 0.00029107551487414187,
+      "loss": 5.1505,
+      "step": 613
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.7414288520812988,
+      "learning_rate": 0.00029103513258850445,
+      "loss": 5.1972,
+      "step": 614
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.7069399356842041,
+      "learning_rate": 0.00029099475030286713,
+      "loss": 5.1766,
+      "step": 615
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.5848967432975769,
+      "learning_rate": 0.00029095436801722976,
+      "loss": 5.1791,
+      "step": 616
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.6382315158843994,
+      "learning_rate": 0.0002909139857315924,
+      "loss": 5.2266,
+      "step": 617
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.6295589208602905,
+      "learning_rate": 0.000290873603445955,
+      "loss": 5.2489,
+      "step": 618
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.6399298310279846,
+      "learning_rate": 0.00029083322116031765,
+      "loss": 5.2295,
+      "step": 619
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.6969920992851257,
+      "learning_rate": 0.0002907928388746803,
+      "loss": 5.1619,
+      "step": 620
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.6793097853660583,
+      "learning_rate": 0.0002907524565890429,
+      "loss": 5.2213,
+      "step": 621
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.6430673003196716,
+      "learning_rate": 0.00029071207430340554,
+      "loss": 5.2848,
+      "step": 622
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.7029622793197632,
+      "learning_rate": 0.00029067169201776817,
+      "loss": 5.2335,
+      "step": 623
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.5936407446861267,
+      "learning_rate": 0.0002906313097321308,
+      "loss": 5.2245,
+      "step": 624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7349820733070374,
+      "learning_rate": 0.00029059092744649343,
+      "loss": 5.2356,
+      "step": 625
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.747024655342102,
+      "learning_rate": 0.00029055054516085606,
+      "loss": 5.1537,
+      "step": 626
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.5645571351051331,
+      "learning_rate": 0.0002905101628752187,
+      "loss": 5.1635,
+      "step": 627
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.7009657621383667,
+      "learning_rate": 0.0002904697805895814,
+      "loss": 5.2236,
+      "step": 628
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 0.7518696188926697,
+      "learning_rate": 0.000290429398303944,
+      "loss": 5.132,
+      "step": 629
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.7385541796684265,
+      "learning_rate": 0.0002903890160183066,
+      "loss": 5.184,
+      "step": 630
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 0.6494706869125366,
+      "learning_rate": 0.0002903486337326692,
+      "loss": 5.2385,
+      "step": 631
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 0.7356922626495361,
+      "learning_rate": 0.0002903082514470319,
+      "loss": 5.1194,
+      "step": 632
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.5700814723968506,
+      "learning_rate": 0.0002902678691613945,
+      "loss": 5.1652,
+      "step": 633
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 0.5895199775695801,
+      "learning_rate": 0.00029022748687575716,
+      "loss": 5.1429,
+      "step": 634
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.5886778831481934,
+      "learning_rate": 0.0002901871045901198,
+      "loss": 5.1932,
+      "step": 635
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.6429824829101562,
+      "learning_rate": 0.0002901467223044824,
+      "loss": 5.1488,
+      "step": 636
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 0.5817999243736267,
+      "learning_rate": 0.00029010634001884505,
+      "loss": 5.1306,
+      "step": 637
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 0.6249139308929443,
+      "learning_rate": 0.0002900659577332077,
+      "loss": 5.1539,
+      "step": 638
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.6901640295982361,
+      "learning_rate": 0.0002900255754475703,
+      "loss": 5.1751,
+      "step": 639
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.7093073725700378,
+      "learning_rate": 0.00028998519316193294,
+      "loss": 5.1692,
+      "step": 640
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 0.6962242722511292,
+      "learning_rate": 0.00028994481087629557,
+      "loss": 5.1768,
+      "step": 641
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.712927520275116,
+      "learning_rate": 0.0002899044285906582,
+      "loss": 5.1236,
+      "step": 642
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 0.600985586643219,
+      "learning_rate": 0.00028986404630502083,
+      "loss": 5.2116,
+      "step": 643
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 0.6715438365936279,
+      "learning_rate": 0.00028982366401938346,
+      "loss": 5.1825,
+      "step": 644
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.854724645614624,
+      "learning_rate": 0.00028978328173374614,
+      "loss": 5.1579,
+      "step": 645
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 0.8737854957580566,
+      "learning_rate": 0.00028974289944810877,
+      "loss": 5.1176,
+      "step": 646
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 0.8141425251960754,
+      "learning_rate": 0.00028970251716247135,
+      "loss": 5.172,
+      "step": 647
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.6891692280769348,
+      "learning_rate": 0.000289662134876834,
+      "loss": 5.2035,
+      "step": 648
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 0.6787595748901367,
+      "learning_rate": 0.0002896217525911966,
+      "loss": 5.1818,
+      "step": 649
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.8094693422317505,
+      "learning_rate": 0.0002895813703055593,
+      "loss": 5.0946,
+      "step": 650
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.6562574505805969,
+      "learning_rate": 0.0002895409880199219,
+      "loss": 5.093,
+      "step": 651
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 0.5972064733505249,
+      "learning_rate": 0.00028950060573428455,
+      "loss": 5.2078,
+      "step": 652
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 0.6744362115859985,
+      "learning_rate": 0.00028946022344864713,
+      "loss": 5.0914,
+      "step": 653
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.7293115258216858,
+      "learning_rate": 0.0002894198411630098,
+      "loss": 5.0954,
+      "step": 654
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.6300185918807983,
+      "learning_rate": 0.00028937945887737244,
+      "loss": 5.2061,
+      "step": 655
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 0.729161262512207,
+      "learning_rate": 0.00028933907659173507,
+      "loss": 5.1129,
+      "step": 656
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.7147690653800964,
+      "learning_rate": 0.0002892986943060977,
+      "loss": 5.1773,
+      "step": 657
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 0.6832327842712402,
+      "learning_rate": 0.00028925831202046033,
+      "loss": 5.1379,
+      "step": 658
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 0.7403486371040344,
+      "learning_rate": 0.00028921792973482296,
+      "loss": 5.1094,
+      "step": 659
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.7538831830024719,
+      "learning_rate": 0.0002891775474491856,
+      "loss": 5.0802,
+      "step": 660
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 0.7270022630691528,
+      "learning_rate": 0.0002891371651635482,
+      "loss": 5.1342,
+      "step": 661
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 0.6888076066970825,
+      "learning_rate": 0.00028909678287791085,
+      "loss": 5.118,
+      "step": 662
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.6924152374267578,
+      "learning_rate": 0.0002890564005922735,
+      "loss": 5.1728,
+      "step": 663
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 0.6232258081436157,
+      "learning_rate": 0.0002890160183066361,
+      "loss": 5.1357,
+      "step": 664
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.7234542369842529,
+      "learning_rate": 0.00028897563602099874,
+      "loss": 5.0607,
+      "step": 665
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.7320601940155029,
+      "learning_rate": 0.0002889352537353614,
+      "loss": 5.1074,
+      "step": 666
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 0.6856744289398193,
+      "learning_rate": 0.00028889487144972406,
+      "loss": 5.0398,
+      "step": 667
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 0.6123003363609314,
+      "learning_rate": 0.0002888544891640867,
+      "loss": 5.178,
+      "step": 668
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.6712738871574402,
+      "learning_rate": 0.0002888141068784493,
+      "loss": 5.1151,
+      "step": 669
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.6770631670951843,
+      "learning_rate": 0.0002887737245928119,
+      "loss": 5.1691,
+      "step": 670
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 0.603068470954895,
+      "learning_rate": 0.0002887333423071746,
+      "loss": 5.0652,
+      "step": 671
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.7057216167449951,
+      "learning_rate": 0.0002886929600215372,
+      "loss": 5.1038,
+      "step": 672
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 0.708411693572998,
+      "learning_rate": 0.00028865257773589984,
+      "loss": 5.1423,
+      "step": 673
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 0.634673535823822,
+      "learning_rate": 0.00028861219545026247,
+      "loss": 5.0765,
+      "step": 674
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.6029425263404846,
+      "learning_rate": 0.0002885718131646251,
+      "loss": 5.0668,
+      "step": 675
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 0.6047025918960571,
+      "learning_rate": 0.00028853143087898773,
+      "loss": 5.0427,
+      "step": 676
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 0.6101155281066895,
+      "learning_rate": 0.00028849104859335036,
+      "loss": 5.1077,
+      "step": 677
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.6415575742721558,
+      "learning_rate": 0.000288450666307713,
+      "loss": 5.1074,
+      "step": 678
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 0.6284694671630859,
+      "learning_rate": 0.0002884102840220756,
+      "loss": 5.1415,
+      "step": 679
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.6168042421340942,
+      "learning_rate": 0.00028836990173643825,
+      "loss": 5.06,
+      "step": 680
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.6429285407066345,
+      "learning_rate": 0.0002883295194508009,
+      "loss": 5.1369,
+      "step": 681
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 0.6893006563186646,
+      "learning_rate": 0.0002882891371651635,
+      "loss": 5.0639,
+      "step": 682
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 0.5782660841941833,
+      "learning_rate": 0.00028824875487952614,
+      "loss": 5.0679,
+      "step": 683
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.6747514009475708,
+      "learning_rate": 0.0002882083725938888,
+      "loss": 5.097,
+      "step": 684
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.7318249940872192,
+      "learning_rate": 0.00028816799030825145,
+      "loss": 5.1338,
+      "step": 685
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 0.6769218444824219,
+      "learning_rate": 0.00028812760802261403,
+      "loss": 5.1185,
+      "step": 686
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.6341337561607361,
+      "learning_rate": 0.00028808722573697666,
+      "loss": 5.1188,
+      "step": 687
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 0.6396981477737427,
+      "learning_rate": 0.0002880468434513393,
+      "loss": 5.0718,
+      "step": 688
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 0.703349232673645,
+      "learning_rate": 0.000288006461165702,
+      "loss": 5.1783,
+      "step": 689
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.7219293713569641,
+      "learning_rate": 0.0002879660788800646,
+      "loss": 5.0716,
+      "step": 690
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 0.6729134321212769,
+      "learning_rate": 0.00028792569659442723,
+      "loss": 5.0406,
+      "step": 691
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 0.6320032477378845,
+      "learning_rate": 0.00028788531430878986,
+      "loss": 5.0302,
+      "step": 692
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.662146806716919,
+      "learning_rate": 0.0002878449320231525,
+      "loss": 5.0877,
+      "step": 693
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 0.6499378085136414,
+      "learning_rate": 0.0002878045497375151,
+      "loss": 4.983,
+      "step": 694
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.600250780582428,
+      "learning_rate": 0.00028776416745187775,
+      "loss": 5.0135,
+      "step": 695
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.6061322689056396,
+      "learning_rate": 0.0002877237851662404,
+      "loss": 5.0913,
+      "step": 696
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 0.6367159485816956,
+      "learning_rate": 0.000287683402880603,
+      "loss": 5.0589,
+      "step": 697
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 0.6972000598907471,
+      "learning_rate": 0.00028764302059496564,
+      "loss": 5.057,
+      "step": 698
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.7460048794746399,
+      "learning_rate": 0.0002876026383093283,
+      "loss": 4.9804,
+      "step": 699
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.7880580425262451,
+      "learning_rate": 0.0002875622560236909,
+      "loss": 5.0366,
+      "step": 700
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 0.857755720615387,
+      "learning_rate": 0.00028752187373805353,
+      "loss": 5.1399,
+      "step": 701
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.8949594497680664,
+      "learning_rate": 0.00028748149145241616,
+      "loss": 5.0537,
+      "step": 702
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 0.6883854866027832,
+      "learning_rate": 0.0002874411091667788,
+      "loss": 5.0771,
+      "step": 703
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 0.7915716171264648,
+      "learning_rate": 0.0002874007268811414,
+      "loss": 5.1028,
+      "step": 704
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.6309002041816711,
+      "learning_rate": 0.00028736034459550405,
+      "loss": 5.079,
+      "step": 705
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 0.639817476272583,
+      "learning_rate": 0.00028731996230986674,
+      "loss": 5.1016,
+      "step": 706
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 0.6250396966934204,
+      "learning_rate": 0.00028727958002422937,
+      "loss": 4.9772,
+      "step": 707
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.5922684669494629,
+      "learning_rate": 0.000287239197738592,
+      "loss": 5.0849,
+      "step": 708
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 0.6167359352111816,
+      "learning_rate": 0.0002871988154529546,
+      "loss": 5.0408,
+      "step": 709
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.6019271612167358,
+      "learning_rate": 0.00028715843316731726,
+      "loss": 5.0312,
+      "step": 710
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.6929312348365784,
+      "learning_rate": 0.0002871180508816799,
+      "loss": 5.0345,
+      "step": 711
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 0.6549073457717896,
+      "learning_rate": 0.0002870776685960425,
+      "loss": 5.0344,
+      "step": 712
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 0.6558994054794312,
+      "learning_rate": 0.00028703728631040515,
+      "loss": 4.9554,
+      "step": 713
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.5860869884490967,
+      "learning_rate": 0.0002869969040247678,
+      "loss": 4.9838,
+      "step": 714
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.6493767499923706,
+      "learning_rate": 0.0002869565217391304,
+      "loss": 5.0245,
+      "step": 715
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 0.6401461958885193,
+      "learning_rate": 0.00028691613945349304,
+      "loss": 5.0256,
+      "step": 716
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.8375266194343567,
+      "learning_rate": 0.00028687575716785567,
+      "loss": 4.9672,
+      "step": 717
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 0.7304956912994385,
+      "learning_rate": 0.0002868353748822183,
+      "loss": 5.024,
+      "step": 718
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 0.7382602095603943,
+      "learning_rate": 0.00028679499259658093,
+      "loss": 5.0483,
+      "step": 719
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.7157966494560242,
+      "learning_rate": 0.00028675461031094356,
+      "loss": 4.9645,
+      "step": 720
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 0.7719393372535706,
+      "learning_rate": 0.0002867142280253062,
+      "loss": 5.0018,
+      "step": 721
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 0.6828843355178833,
+      "learning_rate": 0.0002866738457396688,
+      "loss": 4.9844,
+      "step": 722
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.6229972243309021,
+      "learning_rate": 0.0002866334634540315,
+      "loss": 4.9949,
+      "step": 723
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 0.6336907744407654,
+      "learning_rate": 0.00028659308116839413,
+      "loss": 5.0498,
+      "step": 724
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.6664116978645325,
+      "learning_rate": 0.0002865526988827567,
+      "loss": 5.0202,
+      "step": 725
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.6601686477661133,
+      "learning_rate": 0.00028651231659711934,
+      "loss": 4.892,
+      "step": 726
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 0.713094174861908,
+      "learning_rate": 0.000286471934311482,
+      "loss": 4.9817,
+      "step": 727
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 0.5439857840538025,
+      "learning_rate": 0.00028643155202584465,
+      "loss": 5.0493,
+      "step": 728
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.6119688749313354,
+      "learning_rate": 0.0002863911697402073,
+      "loss": 5.0149,
+      "step": 729
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.6848023533821106,
+      "learning_rate": 0.0002863507874545699,
+      "loss": 5.0056,
+      "step": 730
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 0.6950737833976746,
+      "learning_rate": 0.00028631040516893255,
+      "loss": 5.0121,
+      "step": 731
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.825659453868866,
+      "learning_rate": 0.0002862700228832952,
+      "loss": 4.9852,
+      "step": 732
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 0.7373353242874146,
+      "learning_rate": 0.0002862296405976578,
+      "loss": 4.943,
+      "step": 733
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 0.6202784776687622,
+      "learning_rate": 0.00028618925831202044,
+      "loss": 5.0669,
+      "step": 734
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.6890573501586914,
+      "learning_rate": 0.00028614887602638307,
+      "loss": 4.9571,
+      "step": 735
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 0.6824938654899597,
+      "learning_rate": 0.0002861084937407457,
+      "loss": 4.9799,
+      "step": 736
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 0.693109393119812,
+      "learning_rate": 0.0002860681114551083,
+      "loss": 5.0417,
+      "step": 737
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.5777782201766968,
+      "learning_rate": 0.00028602772916947096,
+      "loss": 5.0035,
+      "step": 738
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 0.6748936772346497,
+      "learning_rate": 0.0002859873468838336,
+      "loss": 4.9764,
+      "step": 739
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.7863640785217285,
+      "learning_rate": 0.00028594696459819627,
+      "loss": 4.9554,
+      "step": 740
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.7226070761680603,
+      "learning_rate": 0.0002859065823125589,
+      "loss": 4.9605,
+      "step": 741
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 0.6418464779853821,
+      "learning_rate": 0.0002858662000269215,
+      "loss": 4.9418,
+      "step": 742
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 0.8660367727279663,
+      "learning_rate": 0.0002858258177412841,
+      "loss": 5.0473,
+      "step": 743
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.7218934297561646,
+      "learning_rate": 0.00028578543545564674,
+      "loss": 5.0287,
+      "step": 744
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.6135517954826355,
+      "learning_rate": 0.0002857450531700094,
+      "loss": 4.9141,
+      "step": 745
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 0.7062588334083557,
+      "learning_rate": 0.00028570467088437205,
+      "loss": 4.974,
+      "step": 746
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.6527339220046997,
+      "learning_rate": 0.0002856642885987347,
+      "loss": 4.9275,
+      "step": 747
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 0.601574718952179,
+      "learning_rate": 0.0002856239063130973,
+      "loss": 5.0131,
+      "step": 748
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 0.6972563862800598,
+      "learning_rate": 0.00028558352402745994,
+      "loss": 5.0037,
+      "step": 749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6463833451271057,
+      "learning_rate": 0.00028554314174182257,
+      "loss": 5.0173,
+      "step": 750
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 0.7086387872695923,
+      "learning_rate": 0.0002855027594561852,
+      "loss": 5.0418,
+      "step": 751
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 0.6756019592285156,
+      "learning_rate": 0.00028546237717054783,
+      "loss": 4.9527,
+      "step": 752
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.7218304872512817,
+      "learning_rate": 0.00028542199488491046,
+      "loss": 4.99,
+      "step": 753
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 0.6159854531288147,
+      "learning_rate": 0.0002853816125992731,
+      "loss": 4.9612,
+      "step": 754
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.7188109159469604,
+      "learning_rate": 0.0002853412303136357,
+      "loss": 4.957,
+      "step": 755
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.7729025483131409,
+      "learning_rate": 0.00028530084802799835,
+      "loss": 5.03,
+      "step": 756
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 0.6761441230773926,
+      "learning_rate": 0.000285260465742361,
+      "loss": 4.9148,
+      "step": 757
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 0.6675111055374146,
+      "learning_rate": 0.0002852200834567236,
+      "loss": 4.9832,
+      "step": 758
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.6335251927375793,
+      "learning_rate": 0.00028517970117108624,
+      "loss": 4.9268,
+      "step": 759
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.633287787437439,
+      "learning_rate": 0.00028513931888544887,
+      "loss": 4.9717,
+      "step": 760
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 0.5922399163246155,
+      "learning_rate": 0.0002850989365998115,
+      "loss": 4.9945,
+      "step": 761
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.5817521214485168,
+      "learning_rate": 0.0002850585543141742,
+      "loss": 4.9261,
+      "step": 762
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 0.6280835270881653,
+      "learning_rate": 0.0002850181720285368,
+      "loss": 4.9935,
+      "step": 763
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 0.6975805163383484,
+      "learning_rate": 0.00028497778974289945,
+      "loss": 4.9161,
+      "step": 764
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.63359135389328,
+      "learning_rate": 0.000284937407457262,
+      "loss": 4.9866,
+      "step": 765
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 0.5709059238433838,
+      "learning_rate": 0.0002848970251716247,
+      "loss": 5.0262,
+      "step": 766
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 0.5760507583618164,
+      "learning_rate": 0.00028485664288598734,
+      "loss": 4.9359,
+      "step": 767
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.5946138501167297,
+      "learning_rate": 0.00028481626060034997,
+      "loss": 4.9843,
+      "step": 768
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 0.6102321743965149,
+      "learning_rate": 0.0002847758783147126,
+      "loss": 4.9541,
+      "step": 769
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.7592500448226929,
+      "learning_rate": 0.0002847354960290752,
+      "loss": 4.9617,
+      "step": 770
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.7633534669876099,
+      "learning_rate": 0.00028469511374343786,
+      "loss": 4.9587,
+      "step": 771
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 0.778260350227356,
+      "learning_rate": 0.0002846547314578005,
+      "loss": 4.9701,
+      "step": 772
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 0.6937357187271118,
+      "learning_rate": 0.0002846143491721631,
+      "loss": 4.8862,
+      "step": 773
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.7559827566146851,
+      "learning_rate": 0.00028457396688652575,
+      "loss": 4.9598,
+      "step": 774
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.699320375919342,
+      "learning_rate": 0.0002845335846008884,
+      "loss": 4.9923,
+      "step": 775
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 0.6152313947677612,
+      "learning_rate": 0.000284493202315251,
+      "loss": 4.9211,
+      "step": 776
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.6833812594413757,
+      "learning_rate": 0.00028445282002961364,
+      "loss": 4.9163,
+      "step": 777
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 0.6250735521316528,
+      "learning_rate": 0.00028441243774397627,
+      "loss": 4.8906,
+      "step": 778
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 0.592545747756958,
+      "learning_rate": 0.00028437205545833895,
+      "loss": 4.8642,
+      "step": 779
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.6293628811836243,
+      "learning_rate": 0.0002843316731727016,
+      "loss": 4.899,
+      "step": 780
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 0.8118374943733215,
+      "learning_rate": 0.00028429129088706416,
+      "loss": 4.9125,
+      "step": 781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0649632215499878,
+      "learning_rate": 0.0002842509086014268,
+      "loss": 4.8757,
+      "step": 782
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 0.8830782175064087,
+      "learning_rate": 0.0002842105263157894,
+      "loss": 4.9041,
+      "step": 783
+    },
+    {
+      "epoch": 1.00256,
+      "grad_norm": 0.9660894870758057,
+      "learning_rate": 0.0002841701440301521,
+      "loss": 4.9646,
+      "step": 784
+    },
+    {
+      "epoch": 1.00384,
+      "grad_norm": 0.8279880285263062,
+      "learning_rate": 0.00028412976174451473,
+      "loss": 4.9266,
+      "step": 785
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.8470652103424072,
+      "learning_rate": 0.00028408937945887736,
+      "loss": 4.8514,
+      "step": 786
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 0.7699068784713745,
+      "learning_rate": 0.00028404899717324,
+      "loss": 4.809,
+      "step": 787
+    },
+    {
+      "epoch": 1.00768,
+      "grad_norm": 0.7924918532371521,
+      "learning_rate": 0.0002840086148876026,
+      "loss": 4.8487,
+      "step": 788
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.7347840070724487,
+      "learning_rate": 0.00028396823260196525,
+      "loss": 4.8208,
+      "step": 789
+    },
+    {
+      "epoch": 1.01024,
+      "grad_norm": 0.6861421465873718,
+      "learning_rate": 0.0002839278503163279,
+      "loss": 4.8625,
+      "step": 790
+    },
+    {
+      "epoch": 1.01152,
+      "grad_norm": 0.6445014476776123,
+      "learning_rate": 0.0002838874680306905,
+      "loss": 4.916,
+      "step": 791
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.6200584769248962,
+      "learning_rate": 0.00028384708574505314,
+      "loss": 4.955,
+      "step": 792
+    },
+    {
+      "epoch": 1.01408,
+      "grad_norm": 0.5568555593490601,
+      "learning_rate": 0.00028380670345941577,
+      "loss": 4.9134,
+      "step": 793
+    },
+    {
+      "epoch": 1.01536,
+      "grad_norm": 0.6153519749641418,
+      "learning_rate": 0.0002837663211737784,
+      "loss": 4.8438,
+      "step": 794
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.7319142818450928,
+      "learning_rate": 0.00028372593888814103,
+      "loss": 4.919,
+      "step": 795
+    },
+    {
+      "epoch": 1.01792,
+      "grad_norm": 0.7483083605766296,
+      "learning_rate": 0.00028368555660250366,
+      "loss": 4.8444,
+      "step": 796
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 0.7571709156036377,
+      "learning_rate": 0.0002836451743168663,
+      "loss": 4.907,
+      "step": 797
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.7250765562057495,
+      "learning_rate": 0.0002836047920312289,
+      "loss": 4.9101,
+      "step": 798
+    },
+    {
+      "epoch": 1.02176,
+      "grad_norm": 0.7267122864723206,
+      "learning_rate": 0.00028356440974559155,
+      "loss": 4.8398,
+      "step": 799
+    },
+    {
+      "epoch": 1.02304,
+      "grad_norm": 0.7116459608078003,
+      "learning_rate": 0.0002835240274599542,
+      "loss": 4.8418,
+      "step": 800
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.6556388735771179,
+      "learning_rate": 0.00028348364517431687,
+      "loss": 4.8538,
+      "step": 801
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 0.6552951335906982,
+      "learning_rate": 0.0002834432628886795,
+      "loss": 4.8909,
+      "step": 802
+    },
+    {
+      "epoch": 1.02688,
+      "grad_norm": 0.7316877841949463,
+      "learning_rate": 0.00028340288060304213,
+      "loss": 4.8016,
+      "step": 803
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.7442478537559509,
+      "learning_rate": 0.0002833624983174047,
+      "loss": 4.8387,
+      "step": 804
+    },
+    {
+      "epoch": 1.02944,
+      "grad_norm": 0.6588963866233826,
+      "learning_rate": 0.0002833221160317674,
+      "loss": 4.8561,
+      "step": 805
+    },
+    {
+      "epoch": 1.03072,
+      "grad_norm": 0.6899812817573547,
+      "learning_rate": 0.00028328173374613,
+      "loss": 4.914,
+      "step": 806
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.7177838087081909,
+      "learning_rate": 0.00028324135146049265,
+      "loss": 4.7727,
+      "step": 807
+    },
+    {
+      "epoch": 1.03328,
+      "grad_norm": 0.7071174383163452,
+      "learning_rate": 0.0002832009691748553,
+      "loss": 4.9105,
+      "step": 808
+    },
+    {
+      "epoch": 1.03456,
+      "grad_norm": 0.6587151288986206,
+      "learning_rate": 0.0002831605868892179,
+      "loss": 4.8851,
+      "step": 809
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.6153483390808105,
+      "learning_rate": 0.00028312020460358054,
+      "loss": 4.8638,
+      "step": 810
+    },
+    {
+      "epoch": 1.03712,
+      "grad_norm": 0.594373345375061,
+      "learning_rate": 0.00028307982231794317,
+      "loss": 4.865,
+      "step": 811
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 0.6585094332695007,
+      "learning_rate": 0.0002830394400323058,
+      "loss": 4.8542,
+      "step": 812
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.6621379852294922,
+      "learning_rate": 0.00028299905774666843,
+      "loss": 4.8349,
+      "step": 813
+    },
+    {
+      "epoch": 1.04096,
+      "grad_norm": 0.5880494713783264,
+      "learning_rate": 0.00028295867546103106,
+      "loss": 4.8509,
+      "step": 814
+    },
+    {
+      "epoch": 1.04224,
+      "grad_norm": 0.6204310059547424,
+      "learning_rate": 0.0002829182931753937,
+      "loss": 4.8359,
+      "step": 815
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.6215956211090088,
+      "learning_rate": 0.0002828779108897563,
+      "loss": 4.8127,
+      "step": 816
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 0.6148485541343689,
+      "learning_rate": 0.00028283752860411895,
+      "loss": 4.8714,
+      "step": 817
+    },
+    {
+      "epoch": 1.04608,
+      "grad_norm": 0.6803502440452576,
+      "learning_rate": 0.00028279714631848163,
+      "loss": 4.8163,
+      "step": 818
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.6470843553543091,
+      "learning_rate": 0.00028275676403284426,
+      "loss": 4.8039,
+      "step": 819
+    },
+    {
+      "epoch": 1.04864,
+      "grad_norm": 0.5897905230522156,
+      "learning_rate": 0.0002827163817472069,
+      "loss": 4.8399,
+      "step": 820
+    },
+    {
+      "epoch": 1.04992,
+      "grad_norm": 0.6008836030960083,
+      "learning_rate": 0.00028267599946156947,
+      "loss": 4.8523,
+      "step": 821
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.5722599029541016,
+      "learning_rate": 0.0002826356171759321,
+      "loss": 4.8057,
+      "step": 822
+    },
+    {
+      "epoch": 1.05248,
+      "grad_norm": 0.5825332403182983,
+      "learning_rate": 0.0002825952348902948,
+      "loss": 4.8341,
+      "step": 823
+    },
+    {
+      "epoch": 1.05376,
+      "grad_norm": 0.6503397822380066,
+      "learning_rate": 0.0002825548526046574,
+      "loss": 4.8925,
+      "step": 824
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.6947764158248901,
+      "learning_rate": 0.00028251447031902004,
+      "loss": 4.7832,
+      "step": 825
+    },
+    {
+      "epoch": 1.05632,
+      "grad_norm": 0.6167904734611511,
+      "learning_rate": 0.0002824740880333827,
+      "loss": 4.7814,
+      "step": 826
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 0.624954342842102,
+      "learning_rate": 0.0002824337057477453,
+      "loss": 4.7072,
+      "step": 827
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.7460287809371948,
+      "learning_rate": 0.00028239332346210793,
+      "loss": 4.7652,
+      "step": 828
+    },
+    {
+      "epoch": 1.06016,
+      "grad_norm": 0.6951721906661987,
+      "learning_rate": 0.00028235294117647056,
+      "loss": 4.8089,
+      "step": 829
+    },
+    {
+      "epoch": 1.06144,
+      "grad_norm": 0.7647305130958557,
+      "learning_rate": 0.0002823125588908332,
+      "loss": 4.8414,
+      "step": 830
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.7322429418563843,
+      "learning_rate": 0.0002822721766051958,
+      "loss": 4.787,
+      "step": 831
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.6130174398422241,
+      "learning_rate": 0.00028223179431955845,
+      "loss": 4.8232,
+      "step": 832
+    },
+    {
+      "epoch": 1.06528,
+      "grad_norm": 0.7291455268859863,
+      "learning_rate": 0.0002821914120339211,
+      "loss": 4.9358,
+      "step": 833
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.7091724872589111,
+      "learning_rate": 0.0002821510297482837,
+      "loss": 4.7688,
+      "step": 834
+    },
+    {
+      "epoch": 1.06784,
+      "grad_norm": 0.6038631200790405,
+      "learning_rate": 0.0002821106474626464,
+      "loss": 4.8498,
+      "step": 835
+    },
+    {
+      "epoch": 1.06912,
+      "grad_norm": 0.5576080083847046,
+      "learning_rate": 0.00028207026517700903,
+      "loss": 4.7541,
+      "step": 836
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.6607750058174133,
+      "learning_rate": 0.0002820298828913716,
+      "loss": 4.7774,
+      "step": 837
+    },
+    {
+      "epoch": 1.07168,
+      "grad_norm": 0.6005157232284546,
+      "learning_rate": 0.00028198950060573423,
+      "loss": 4.8224,
+      "step": 838
+    },
+    {
+      "epoch": 1.07296,
+      "grad_norm": 0.7134353518486023,
+      "learning_rate": 0.00028194911832009686,
+      "loss": 4.7846,
+      "step": 839
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.8062997460365295,
+      "learning_rate": 0.00028190873603445955,
+      "loss": 4.8548,
+      "step": 840
+    },
+    {
+      "epoch": 1.07552,
+      "grad_norm": 0.7981602549552917,
+      "learning_rate": 0.0002818683537488222,
+      "loss": 4.8362,
+      "step": 841
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 0.7718906998634338,
+      "learning_rate": 0.0002818279714631848,
+      "loss": 4.8407,
+      "step": 842
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.7952736020088196,
+      "learning_rate": 0.00028178758917754744,
+      "loss": 4.8551,
+      "step": 843
+    },
+    {
+      "epoch": 1.07936,
+      "grad_norm": 0.7064546346664429,
+      "learning_rate": 0.00028174720689191007,
+      "loss": 4.755,
+      "step": 844
+    },
+    {
+      "epoch": 1.08064,
+      "grad_norm": 0.6656110882759094,
+      "learning_rate": 0.0002817068246062727,
+      "loss": 4.7786,
+      "step": 845
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.6304865479469299,
+      "learning_rate": 0.00028166644232063533,
+      "loss": 4.8153,
+      "step": 846
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 0.6220903992652893,
+      "learning_rate": 0.00028162606003499796,
+      "loss": 4.8118,
+      "step": 847
+    },
+    {
+      "epoch": 1.08448,
+      "grad_norm": 0.7010989189147949,
+      "learning_rate": 0.0002815856777493606,
+      "loss": 4.8627,
+      "step": 848
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.6893423795700073,
+      "learning_rate": 0.0002815452954637232,
+      "loss": 4.7906,
+      "step": 849
+    },
+    {
+      "epoch": 1.08704,
+      "grad_norm": 0.6807447075843811,
+      "learning_rate": 0.00028150491317808585,
+      "loss": 4.7955,
+      "step": 850
+    },
+    {
+      "epoch": 1.08832,
+      "grad_norm": 0.6272568702697754,
+      "learning_rate": 0.0002814645308924485,
+      "loss": 4.6878,
+      "step": 851
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.6356240510940552,
+      "learning_rate": 0.0002814241486068111,
+      "loss": 4.7682,
+      "step": 852
+    },
+    {
+      "epoch": 1.09088,
+      "grad_norm": 0.7134535312652588,
+      "learning_rate": 0.00028138376632117374,
+      "loss": 4.7629,
+      "step": 853
+    },
+    {
+      "epoch": 1.09216,
+      "grad_norm": 0.6460220813751221,
+      "learning_rate": 0.00028134338403553637,
+      "loss": 4.7132,
+      "step": 854
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.746244490146637,
+      "learning_rate": 0.000281303001749899,
+      "loss": 4.8273,
+      "step": 855
+    },
+    {
+      "epoch": 1.09472,
+      "grad_norm": 0.7077825665473938,
+      "learning_rate": 0.00028126261946426163,
+      "loss": 4.7677,
+      "step": 856
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 0.652271568775177,
+      "learning_rate": 0.0002812222371786243,
+      "loss": 4.812,
+      "step": 857
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.6102200150489807,
+      "learning_rate": 0.00028118185489298694,
+      "loss": 4.7964,
+      "step": 858
+    },
+    {
+      "epoch": 1.09856,
+      "grad_norm": 0.7953206896781921,
+      "learning_rate": 0.0002811414726073496,
+      "loss": 4.7903,
+      "step": 859
+    },
+    {
+      "epoch": 1.09984,
+      "grad_norm": 0.641959547996521,
+      "learning_rate": 0.00028110109032171215,
+      "loss": 4.8177,
+      "step": 860
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.7065843939781189,
+      "learning_rate": 0.00028106070803607483,
+      "loss": 4.7971,
+      "step": 861
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 0.6949476003646851,
+      "learning_rate": 0.00028102032575043746,
+      "loss": 4.836,
+      "step": 862
+    },
+    {
+      "epoch": 1.10368,
+      "grad_norm": 0.7160281538963318,
+      "learning_rate": 0.0002809799434648001,
+      "loss": 4.7041,
+      "step": 863
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.6318966150283813,
+      "learning_rate": 0.0002809395611791627,
+      "loss": 4.7332,
+      "step": 864
+    },
+    {
+      "epoch": 1.1062400000000001,
+      "grad_norm": 0.6665629148483276,
+      "learning_rate": 0.00028089917889352535,
+      "loss": 4.8316,
+      "step": 865
+    },
+    {
+      "epoch": 1.10752,
+      "grad_norm": 0.6924526691436768,
+      "learning_rate": 0.000280858796607888,
+      "loss": 4.7881,
+      "step": 866
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.6507744789123535,
+      "learning_rate": 0.0002808184143222506,
+      "loss": 4.8198,
+      "step": 867
+    },
+    {
+      "epoch": 1.11008,
+      "grad_norm": 0.7218488454818726,
+      "learning_rate": 0.00028077803203661324,
+      "loss": 4.8512,
+      "step": 868
+    },
+    {
+      "epoch": 1.11136,
+      "grad_norm": 0.7143361568450928,
+      "learning_rate": 0.0002807376497509759,
+      "loss": 4.795,
+      "step": 869
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.6624107956886292,
+      "learning_rate": 0.0002806972674653385,
+      "loss": 4.7502,
+      "step": 870
+    },
+    {
+      "epoch": 1.11392,
+      "grad_norm": 0.6978575587272644,
+      "learning_rate": 0.00028065688517970114,
+      "loss": 4.8148,
+      "step": 871
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 0.6641087532043457,
+      "learning_rate": 0.00028061650289406377,
+      "loss": 4.7369,
+      "step": 872
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.7032644748687744,
+      "learning_rate": 0.0002805761206084264,
+      "loss": 4.7203,
+      "step": 873
+    },
+    {
+      "epoch": 1.11776,
+      "grad_norm": 0.6251493096351624,
+      "learning_rate": 0.0002805357383227891,
+      "loss": 4.8037,
+      "step": 874
+    },
+    {
+      "epoch": 1.11904,
+      "grad_norm": 0.6594140529632568,
+      "learning_rate": 0.0002804953560371517,
+      "loss": 4.7187,
+      "step": 875
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.6820217370986938,
+      "learning_rate": 0.0002804549737515143,
+      "loss": 4.7954,
+      "step": 876
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 0.7188774347305298,
+      "learning_rate": 0.0002804145914658769,
+      "loss": 4.7619,
+      "step": 877
+    },
+    {
+      "epoch": 1.12288,
+      "grad_norm": 0.7069806456565857,
+      "learning_rate": 0.00028037420918023955,
+      "loss": 4.7879,
+      "step": 878
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.5844770669937134,
+      "learning_rate": 0.00028033382689460223,
+      "loss": 4.7458,
+      "step": 879
+    },
+    {
+      "epoch": 1.12544,
+      "grad_norm": 0.6429082751274109,
+      "learning_rate": 0.00028029344460896486,
+      "loss": 4.6233,
+      "step": 880
+    },
+    {
+      "epoch": 1.12672,
+      "grad_norm": 0.736958384513855,
+      "learning_rate": 0.0002802530623233275,
+      "loss": 4.745,
+      "step": 881
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.6151041984558105,
+      "learning_rate": 0.0002802126800376901,
+      "loss": 4.6771,
+      "step": 882
+    },
+    {
+      "epoch": 1.12928,
+      "grad_norm": 0.6079396605491638,
+      "learning_rate": 0.00028017229775205275,
+      "loss": 4.8183,
+      "step": 883
+    },
+    {
+      "epoch": 1.13056,
+      "grad_norm": 0.6472486853599548,
+      "learning_rate": 0.0002801319154664154,
+      "loss": 4.713,
+      "step": 884
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.6157152056694031,
+      "learning_rate": 0.000280091533180778,
+      "loss": 4.727,
+      "step": 885
+    },
+    {
+      "epoch": 1.13312,
+      "grad_norm": 0.5990360975265503,
+      "learning_rate": 0.00028005115089514064,
+      "loss": 4.8425,
+      "step": 886
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 0.6508432030677795,
+      "learning_rate": 0.00028001076860950327,
+      "loss": 4.7481,
+      "step": 887
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.6805848479270935,
+      "learning_rate": 0.0002799703863238659,
+      "loss": 4.7301,
+      "step": 888
+    },
+    {
+      "epoch": 1.13696,
+      "grad_norm": 0.6594050526618958,
+      "learning_rate": 0.00027993000403822853,
+      "loss": 4.726,
+      "step": 889
+    },
+    {
+      "epoch": 1.13824,
+      "grad_norm": 0.5740160942077637,
+      "learning_rate": 0.00027988962175259116,
+      "loss": 4.7209,
+      "step": 890
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.6788250207901001,
+      "learning_rate": 0.0002798492394669538,
+      "loss": 4.7201,
+      "step": 891
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 0.6584210395812988,
+      "learning_rate": 0.0002798088571813165,
+      "loss": 4.7145,
+      "step": 892
+    },
+    {
+      "epoch": 1.14208,
+      "grad_norm": 0.6468678116798401,
+      "learning_rate": 0.00027976847489567905,
+      "loss": 4.6967,
+      "step": 893
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.6821480989456177,
+      "learning_rate": 0.0002797280926100417,
+      "loss": 4.7826,
+      "step": 894
+    },
+    {
+      "epoch": 1.1446399999999999,
+      "grad_norm": 0.6273735165596008,
+      "learning_rate": 0.0002796877103244043,
+      "loss": 4.7037,
+      "step": 895
+    },
+    {
+      "epoch": 1.14592,
+      "grad_norm": 0.6939446926116943,
+      "learning_rate": 0.000279647328038767,
+      "loss": 4.7438,
+      "step": 896
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.7489894032478333,
+      "learning_rate": 0.0002796069457531296,
+      "loss": 4.6979,
+      "step": 897
+    },
+    {
+      "epoch": 1.14848,
+      "grad_norm": 0.7742366194725037,
+      "learning_rate": 0.00027956656346749226,
+      "loss": 4.691,
+      "step": 898
+    },
+    {
+      "epoch": 1.1497600000000001,
+      "grad_norm": 0.7484803199768066,
+      "learning_rate": 0.00027952618118185483,
+      "loss": 4.8254,
+      "step": 899
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.8308618068695068,
+      "learning_rate": 0.0002794857988962175,
+      "loss": 4.7098,
+      "step": 900
+    },
+    {
+      "epoch": 1.15232,
+      "grad_norm": 0.6708440780639648,
+      "learning_rate": 0.00027944541661058015,
+      "loss": 4.8018,
+      "step": 901
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 0.6915727853775024,
+      "learning_rate": 0.0002794050343249428,
+      "loss": 4.6875,
+      "step": 902
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.6630460619926453,
+      "learning_rate": 0.0002793646520393054,
+      "loss": 4.7815,
+      "step": 903
+    },
+    {
+      "epoch": 1.15616,
+      "grad_norm": 0.6760913133621216,
+      "learning_rate": 0.00027932426975366804,
+      "loss": 4.7237,
+      "step": 904
+    },
+    {
+      "epoch": 1.15744,
+      "grad_norm": 0.6769623160362244,
+      "learning_rate": 0.00027928388746803067,
+      "loss": 4.814,
+      "step": 905
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.6976802945137024,
+      "learning_rate": 0.0002792435051823933,
+      "loss": 4.7954,
+      "step": 906
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5799634456634521,
+      "learning_rate": 0.0002792031228967559,
+      "loss": 4.6929,
+      "step": 907
+    },
+    {
+      "epoch": 1.16128,
+      "grad_norm": 0.6125800609588623,
+      "learning_rate": 0.00027916274061111856,
+      "loss": 4.6535,
+      "step": 908
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.6879367828369141,
+      "learning_rate": 0.0002791223583254812,
+      "loss": 4.7521,
+      "step": 909
+    },
+    {
+      "epoch": 1.16384,
+      "grad_norm": 0.737657368183136,
+      "learning_rate": 0.0002790819760398438,
+      "loss": 4.7687,
+      "step": 910
+    },
+    {
+      "epoch": 1.16512,
+      "grad_norm": 0.6375575065612793,
+      "learning_rate": 0.00027904159375420645,
+      "loss": 4.8383,
+      "step": 911
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.5931581258773804,
+      "learning_rate": 0.0002790012114685691,
+      "loss": 4.7189,
+      "step": 912
+    },
+    {
+      "epoch": 1.16768,
+      "grad_norm": 0.7057203054428101,
+      "learning_rate": 0.00027896082918293176,
+      "loss": 4.6494,
+      "step": 913
+    },
+    {
+      "epoch": 1.16896,
+      "grad_norm": 0.7094667553901672,
+      "learning_rate": 0.0002789204468972944,
+      "loss": 4.7491,
+      "step": 914
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.6365569233894348,
+      "learning_rate": 0.000278880064611657,
+      "loss": 4.7078,
+      "step": 915
+    },
+    {
+      "epoch": 1.1715200000000001,
+      "grad_norm": 0.7238393425941467,
+      "learning_rate": 0.0002788396823260196,
+      "loss": 4.6681,
+      "step": 916
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 0.7257333993911743,
+      "learning_rate": 0.0002787993000403822,
+      "loss": 4.759,
+      "step": 917
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.644520103931427,
+      "learning_rate": 0.0002787589177547449,
+      "loss": 4.7318,
+      "step": 918
+    },
+    {
+      "epoch": 1.17536,
+      "grad_norm": 0.6309005618095398,
+      "learning_rate": 0.00027871853546910754,
+      "loss": 4.7544,
+      "step": 919
+    },
+    {
+      "epoch": 1.17664,
+      "grad_norm": 0.715380847454071,
+      "learning_rate": 0.00027867815318347017,
+      "loss": 4.7209,
+      "step": 920
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.7284442186355591,
+      "learning_rate": 0.0002786377708978328,
+      "loss": 4.7191,
+      "step": 921
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 0.6391010284423828,
+      "learning_rate": 0.00027859738861219543,
+      "loss": 4.7282,
+      "step": 922
+    },
+    {
+      "epoch": 1.18048,
+      "grad_norm": 0.6188117861747742,
+      "learning_rate": 0.00027855700632655806,
+      "loss": 4.6801,
+      "step": 923
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.6303733587265015,
+      "learning_rate": 0.0002785166240409207,
+      "loss": 4.7246,
+      "step": 924
+    },
+    {
+      "epoch": 1.18304,
+      "grad_norm": 0.5999323129653931,
+      "learning_rate": 0.0002784762417552833,
+      "loss": 4.749,
+      "step": 925
+    },
+    {
+      "epoch": 1.18432,
+      "grad_norm": 0.6512132883071899,
+      "learning_rate": 0.00027843585946964595,
+      "loss": 4.6749,
+      "step": 926
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.6530429124832153,
+      "learning_rate": 0.0002783954771840086,
+      "loss": 4.7483,
+      "step": 927
+    },
+    {
+      "epoch": 1.18688,
+      "grad_norm": 0.6326019763946533,
+      "learning_rate": 0.0002783550948983712,
+      "loss": 4.6818,
+      "step": 928
+    },
+    {
+      "epoch": 1.1881599999999999,
+      "grad_norm": 0.6044508218765259,
+      "learning_rate": 0.00027831471261273384,
+      "loss": 4.737,
+      "step": 929
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.6268877983093262,
+      "learning_rate": 0.00027827433032709647,
+      "loss": 4.6278,
+      "step": 930
+    },
+    {
+      "epoch": 1.19072,
+      "grad_norm": 0.6803573966026306,
+      "learning_rate": 0.00027823394804145916,
+      "loss": 4.6907,
+      "step": 931
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 0.7021913528442383,
+      "learning_rate": 0.00027819356575582173,
+      "loss": 4.683,
+      "step": 932
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.6417155861854553,
+      "learning_rate": 0.00027815318347018436,
+      "loss": 4.7198,
+      "step": 933
+    },
+    {
+      "epoch": 1.19456,
+      "grad_norm": 0.6091512441635132,
+      "learning_rate": 0.000278112801184547,
+      "loss": 4.694,
+      "step": 934
+    },
+    {
+      "epoch": 1.19584,
+      "grad_norm": 0.7351169586181641,
+      "learning_rate": 0.0002780724188989097,
+      "loss": 4.697,
+      "step": 935
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.7421384453773499,
+      "learning_rate": 0.0002780320366132723,
+      "loss": 4.6373,
+      "step": 936
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 0.745575487613678,
+      "learning_rate": 0.00027799165432763494,
+      "loss": 4.6504,
+      "step": 937
+    },
+    {
+      "epoch": 1.19968,
+      "grad_norm": 0.650227963924408,
+      "learning_rate": 0.00027795127204199757,
+      "loss": 4.6666,
+      "step": 938
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.6333515048027039,
+      "learning_rate": 0.0002779108897563602,
+      "loss": 4.6865,
+      "step": 939
+    },
+    {
+      "epoch": 1.20224,
+      "grad_norm": 0.6844230890274048,
+      "learning_rate": 0.00027787050747072283,
+      "loss": 4.6709,
+      "step": 940
+    },
+    {
+      "epoch": 1.20352,
+      "grad_norm": 0.6432576775550842,
+      "learning_rate": 0.00027783012518508546,
+      "loss": 4.7449,
+      "step": 941
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.7197453379631042,
+      "learning_rate": 0.0002777897428994481,
+      "loss": 4.6855,
+      "step": 942
+    },
+    {
+      "epoch": 1.20608,
+      "grad_norm": 0.8134915232658386,
+      "learning_rate": 0.0002777493606138107,
+      "loss": 4.7429,
+      "step": 943
+    },
+    {
+      "epoch": 1.20736,
+      "grad_norm": 0.7857049107551575,
+      "learning_rate": 0.00027770897832817335,
+      "loss": 4.7009,
+      "step": 944
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.6796789765357971,
+      "learning_rate": 0.000277668596042536,
+      "loss": 4.7261,
+      "step": 945
+    },
+    {
+      "epoch": 1.2099199999999999,
+      "grad_norm": 0.7167194485664368,
+      "learning_rate": 0.0002776282137568986,
+      "loss": 4.6099,
+      "step": 946
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 0.7006614804267883,
+      "learning_rate": 0.00027758783147126124,
+      "loss": 4.6235,
+      "step": 947
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.7764577865600586,
+      "learning_rate": 0.00027754744918562387,
+      "loss": 4.667,
+      "step": 948
+    },
+    {
+      "epoch": 1.21376,
+      "grad_norm": 0.6722946166992188,
+      "learning_rate": 0.0002775070668999865,
+      "loss": 4.7208,
+      "step": 949
+    },
+    {
+      "epoch": 1.2150400000000001,
+      "grad_norm": 0.6938252449035645,
+      "learning_rate": 0.00027746668461434913,
+      "loss": 4.6394,
+      "step": 950
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.7050387859344482,
+      "learning_rate": 0.00027742630232871176,
+      "loss": 4.6333,
+      "step": 951
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 0.7311280965805054,
+      "learning_rate": 0.00027738592004307444,
+      "loss": 4.6614,
+      "step": 952
+    },
+    {
+      "epoch": 1.21888,
+      "grad_norm": 0.6674593091011047,
+      "learning_rate": 0.00027734553775743707,
+      "loss": 4.739,
+      "step": 953
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.698934018611908,
+      "learning_rate": 0.0002773051554717997,
+      "loss": 4.6763,
+      "step": 954
+    },
+    {
+      "epoch": 1.22144,
+      "grad_norm": 0.7208249568939209,
+      "learning_rate": 0.0002772647731861623,
+      "loss": 4.6478,
+      "step": 955
+    },
+    {
+      "epoch": 1.22272,
+      "grad_norm": 0.6473603248596191,
+      "learning_rate": 0.00027722439090052496,
+      "loss": 4.6642,
+      "step": 956
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.7811574339866638,
+      "learning_rate": 0.0002771840086148876,
+      "loss": 4.7188,
+      "step": 957
+    },
+    {
+      "epoch": 1.22528,
+      "grad_norm": 0.7508848309516907,
+      "learning_rate": 0.0002771436263292502,
+      "loss": 4.6873,
+      "step": 958
+    },
+    {
+      "epoch": 1.22656,
+      "grad_norm": 0.6528134942054749,
+      "learning_rate": 0.00027710324404361285,
+      "loss": 4.6341,
+      "step": 959
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.7064421772956848,
+      "learning_rate": 0.0002770628617579755,
+      "loss": 4.6409,
+      "step": 960
+    },
+    {
+      "epoch": 1.22912,
+      "grad_norm": 0.7718419432640076,
+      "learning_rate": 0.0002770224794723381,
+      "loss": 4.7507,
+      "step": 961
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 0.6676574945449829,
+      "learning_rate": 0.00027698209718670074,
+      "loss": 4.6491,
+      "step": 962
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.7909667491912842,
+      "learning_rate": 0.0002769417149010634,
+      "loss": 4.6478,
+      "step": 963
+    },
+    {
+      "epoch": 1.23296,
+      "grad_norm": 0.8206857442855835,
+      "learning_rate": 0.000276901332615426,
+      "loss": 4.6306,
+      "step": 964
+    },
+    {
+      "epoch": 1.23424,
+      "grad_norm": 0.6872743964195251,
+      "learning_rate": 0.00027686095032978863,
+      "loss": 4.6992,
+      "step": 965
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.659874439239502,
+      "learning_rate": 0.00027682056804415126,
+      "loss": 4.6612,
+      "step": 966
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 0.7639236450195312,
+      "learning_rate": 0.0002767801857585139,
+      "loss": 4.6523,
+      "step": 967
+    },
+    {
+      "epoch": 1.23808,
+      "grad_norm": 0.6640061140060425,
+      "learning_rate": 0.0002767398034728765,
+      "loss": 4.6604,
+      "step": 968
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.6309702396392822,
+      "learning_rate": 0.0002766994211872392,
+      "loss": 4.6036,
+      "step": 969
+    },
+    {
+      "epoch": 1.24064,
+      "grad_norm": 0.6865962147712708,
+      "learning_rate": 0.00027665903890160184,
+      "loss": 4.6821,
+      "step": 970
+    },
+    {
+      "epoch": 1.24192,
+      "grad_norm": 0.7177154421806335,
+      "learning_rate": 0.0002766186566159644,
+      "loss": 4.5896,
+      "step": 971
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.7251400351524353,
+      "learning_rate": 0.00027657827433032704,
+      "loss": 4.567,
+      "step": 972
+    },
+    {
+      "epoch": 1.24448,
+      "grad_norm": 0.6386240124702454,
+      "learning_rate": 0.0002765378920446897,
+      "loss": 4.5937,
+      "step": 973
+    },
+    {
+      "epoch": 1.24576,
+      "grad_norm": 0.666797935962677,
+      "learning_rate": 0.00027649750975905236,
+      "loss": 4.6854,
+      "step": 974
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.6357764601707458,
+      "learning_rate": 0.000276457127473415,
+      "loss": 4.623,
+      "step": 975
+    },
+    {
+      "epoch": 1.24832,
+      "grad_norm": 0.6851757168769836,
+      "learning_rate": 0.0002764167451877776,
+      "loss": 4.6517,
+      "step": 976
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 0.6265616416931152,
+      "learning_rate": 0.00027637636290214025,
+      "loss": 4.6597,
+      "step": 977
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.698387861251831,
+      "learning_rate": 0.0002763359806165029,
+      "loss": 4.5611,
+      "step": 978
+    },
+    {
+      "epoch": 1.25216,
+      "grad_norm": 0.6262152194976807,
+      "learning_rate": 0.0002762955983308655,
+      "loss": 4.6248,
+      "step": 979
+    },
+    {
+      "epoch": 1.2534399999999999,
+      "grad_norm": 0.675564706325531,
+      "learning_rate": 0.00027625521604522814,
+      "loss": 4.6137,
+      "step": 980
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.5722922682762146,
+      "learning_rate": 0.00027621483375959077,
+      "loss": 4.6744,
+      "step": 981
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 0.6412468552589417,
+      "learning_rate": 0.0002761744514739534,
+      "loss": 4.5666,
+      "step": 982
+    },
+    {
+      "epoch": 1.25728,
+      "grad_norm": 0.7287204265594482,
+      "learning_rate": 0.00027613406918831603,
+      "loss": 4.6355,
+      "step": 983
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.742250382900238,
+      "learning_rate": 0.00027609368690267866,
+      "loss": 4.667,
+      "step": 984
+    },
+    {
+      "epoch": 1.25984,
+      "grad_norm": 0.6279268860816956,
+      "learning_rate": 0.0002760533046170413,
+      "loss": 4.5517,
+      "step": 985
+    },
+    {
+      "epoch": 1.26112,
+      "grad_norm": 0.6083707213401794,
+      "learning_rate": 0.0002760129223314039,
+      "loss": 4.5803,
+      "step": 986
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.6249653697013855,
+      "learning_rate": 0.0002759725400457666,
+      "loss": 4.6421,
+      "step": 987
+    },
+    {
+      "epoch": 1.26368,
+      "grad_norm": 0.5808601975440979,
+      "learning_rate": 0.0002759321577601292,
+      "loss": 4.6447,
+      "step": 988
+    },
+    {
+      "epoch": 1.2649599999999999,
+      "grad_norm": 0.6520289778709412,
+      "learning_rate": 0.0002758917754744918,
+      "loss": 4.7274,
+      "step": 989
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.7377476692199707,
+      "learning_rate": 0.00027585139318885444,
+      "loss": 4.625,
+      "step": 990
+    },
+    {
+      "epoch": 1.26752,
+      "grad_norm": 0.6483809947967529,
+      "learning_rate": 0.0002758110109032171,
+      "loss": 4.6238,
+      "step": 991
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 0.625088632106781,
+      "learning_rate": 0.00027577062861757975,
+      "loss": 4.6473,
+      "step": 992
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.6513147950172424,
+      "learning_rate": 0.0002757302463319424,
+      "loss": 4.5971,
+      "step": 993
+    },
+    {
+      "epoch": 1.27136,
+      "grad_norm": 0.6792516112327576,
+      "learning_rate": 0.000275689864046305,
+      "loss": 4.6855,
+      "step": 994
+    },
+    {
+      "epoch": 1.27264,
+      "grad_norm": 0.6334472894668579,
+      "learning_rate": 0.00027564948176066764,
+      "loss": 4.6,
+      "step": 995
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.6187386512756348,
+      "learning_rate": 0.0002756090994750303,
+      "loss": 4.6733,
+      "step": 996
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 0.6294391751289368,
+      "learning_rate": 0.0002755687171893929,
+      "loss": 4.6067,
+      "step": 997
+    },
+    {
+      "epoch": 1.27648,
+      "grad_norm": 0.6575160026550293,
+      "learning_rate": 0.00027552833490375553,
+      "loss": 4.662,
+      "step": 998
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.6640904545783997,
+      "learning_rate": 0.00027548795261811816,
+      "loss": 4.6074,
+      "step": 999
+    },
+    {
+      "epoch": 1.27904,
+      "grad_norm": 0.630664050579071,
+      "learning_rate": 0.0002754475703324808,
+      "loss": 4.6227,
+      "step": 1000
+    },
+    {
+      "epoch": 1.2803200000000001,
+      "grad_norm": 0.7079930901527405,
+      "learning_rate": 0.0002754071880468434,
+      "loss": 4.6341,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.7556226253509521,
+      "learning_rate": 0.00027536680576120605,
+      "loss": 4.6352,
+      "step": 1002
+    },
+    {
+      "epoch": 1.28288,
+      "grad_norm": 0.5955154895782471,
+      "learning_rate": 0.0002753264234755687,
+      "loss": 4.5975,
+      "step": 1003
+    },
+    {
+      "epoch": 1.28416,
+      "grad_norm": 0.7106191515922546,
+      "learning_rate": 0.0002752860411899313,
+      "loss": 4.6302,
+      "step": 1004
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.7254298329353333,
+      "learning_rate": 0.00027524565890429394,
+      "loss": 4.6382,
+      "step": 1005
+    },
+    {
+      "epoch": 1.2867199999999999,
+      "grad_norm": 0.6350436210632324,
+      "learning_rate": 0.0002752052766186566,
+      "loss": 4.6083,
+      "step": 1006
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 0.7333962321281433,
+      "learning_rate": 0.0002751648943330192,
+      "loss": 4.6096,
+      "step": 1007
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.5859498977661133,
+      "learning_rate": 0.0002751245120473819,
+      "loss": 4.6632,
+      "step": 1008
+    },
+    {
+      "epoch": 1.29056,
+      "grad_norm": 0.7196768522262573,
+      "learning_rate": 0.0002750841297617445,
+      "loss": 4.7089,
+      "step": 1009
+    },
+    {
+      "epoch": 1.29184,
+      "grad_norm": 0.6264676451683044,
+      "learning_rate": 0.00027504374747610715,
+      "loss": 4.609,
+      "step": 1010
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.6764931082725525,
+      "learning_rate": 0.0002750033651904697,
+      "loss": 4.5912,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 0.6790111064910889,
+      "learning_rate": 0.00027496298290483236,
+      "loss": 4.6433,
+      "step": 1012
+    },
+    {
+      "epoch": 1.29568,
+      "grad_norm": 0.6634168028831482,
+      "learning_rate": 0.00027492260061919504,
+      "loss": 4.6097,
+      "step": 1013
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.63548344373703,
+      "learning_rate": 0.00027488221833355767,
+      "loss": 4.5944,
+      "step": 1014
+    },
+    {
+      "epoch": 1.29824,
+      "grad_norm": 0.6369843482971191,
+      "learning_rate": 0.0002748418360479203,
+      "loss": 4.5632,
+      "step": 1015
+    },
+    {
+      "epoch": 1.29952,
+      "grad_norm": 0.6853580474853516,
+      "learning_rate": 0.00027480145376228293,
+      "loss": 4.6141,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.648949921131134,
+      "learning_rate": 0.00027476107147664556,
+      "loss": 4.5685,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3020800000000001,
+      "grad_norm": 0.5883113145828247,
+      "learning_rate": 0.0002747206891910082,
+      "loss": 4.6188,
+      "step": 1018
+    },
+    {
+      "epoch": 1.30336,
+      "grad_norm": 0.6754709482192993,
+      "learning_rate": 0.0002746803069053708,
+      "loss": 4.6567,
+      "step": 1019
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.6051660776138306,
+      "learning_rate": 0.00027463992461973345,
+      "loss": 4.6101,
+      "step": 1020
+    },
+    {
+      "epoch": 1.30592,
+      "grad_norm": 0.609302282333374,
+      "learning_rate": 0.0002745995423340961,
+      "loss": 4.5448,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 0.6241468191146851,
+      "learning_rate": 0.0002745591600484587,
+      "loss": 4.6168,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.6044094562530518,
+      "learning_rate": 0.00027451877776282134,
+      "loss": 4.6363,
+      "step": 1023
+    },
+    {
+      "epoch": 1.30976,
+      "grad_norm": 0.6301230788230896,
+      "learning_rate": 0.00027447839547718397,
+      "loss": 4.5699,
+      "step": 1024
+    },
+    {
+      "epoch": 1.31104,
+      "grad_norm": 0.6775204539299011,
+      "learning_rate": 0.0002744380131915466,
+      "loss": 4.582,
+      "step": 1025
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.631717324256897,
+      "learning_rate": 0.0002743976309059093,
+      "loss": 4.559,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 0.631300687789917,
+      "learning_rate": 0.00027435724862027186,
+      "loss": 4.5394,
+      "step": 1027
+    },
+    {
+      "epoch": 1.31488,
+      "grad_norm": 0.6998112201690674,
+      "learning_rate": 0.0002743168663346345,
+      "loss": 4.5337,
+      "step": 1028
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.6501991748809814,
+      "learning_rate": 0.0002742764840489971,
+      "loss": 4.5545,
+      "step": 1029
+    },
+    {
+      "epoch": 1.31744,
+      "grad_norm": 0.6726256608963013,
+      "learning_rate": 0.0002742361017633598,
+      "loss": 4.5374,
+      "step": 1030
+    },
+    {
+      "epoch": 1.31872,
+      "grad_norm": 0.6180437803268433,
+      "learning_rate": 0.00027419571947772244,
+      "loss": 4.5881,
+      "step": 1031
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6200023889541626,
+      "learning_rate": 0.00027415533719208507,
+      "loss": 4.5298,
+      "step": 1032
+    },
+    {
+      "epoch": 1.32128,
+      "grad_norm": 0.6141354441642761,
+      "learning_rate": 0.0002741149549064477,
+      "loss": 4.4838,
+      "step": 1033
+    },
+    {
+      "epoch": 1.32256,
+      "grad_norm": 0.600661039352417,
+      "learning_rate": 0.0002740745726208103,
+      "loss": 4.5947,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.7037731409072876,
+      "learning_rate": 0.00027403419033517296,
+      "loss": 4.544,
+      "step": 1035
+    },
+    {
+      "epoch": 1.32512,
+      "grad_norm": 0.6471102237701416,
+      "learning_rate": 0.0002739938080495356,
+      "loss": 4.5557,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 0.6341505646705627,
+      "learning_rate": 0.0002739534257638982,
+      "loss": 4.662,
+      "step": 1037
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.660749614238739,
+      "learning_rate": 0.00027391304347826085,
+      "loss": 4.628,
+      "step": 1038
+    },
+    {
+      "epoch": 1.32896,
+      "grad_norm": 0.6301653385162354,
+      "learning_rate": 0.0002738726611926235,
+      "loss": 4.5595,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3302399999999999,
+      "grad_norm": 0.6750319004058838,
+      "learning_rate": 0.0002738322789069861,
+      "loss": 4.6398,
+      "step": 1040
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.6259351372718811,
+      "learning_rate": 0.00027379189662134874,
+      "loss": 4.6368,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 0.6357682347297668,
+      "learning_rate": 0.00027375151433571137,
+      "loss": 4.5332,
+      "step": 1042
+    },
+    {
+      "epoch": 1.33408,
+      "grad_norm": 0.6520000100135803,
+      "learning_rate": 0.000273711132050074,
+      "loss": 4.6516,
+      "step": 1043
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.7159348130226135,
+      "learning_rate": 0.0002736707497644366,
+      "loss": 4.5656,
+      "step": 1044
+    },
+    {
+      "epoch": 1.33664,
+      "grad_norm": 0.7513425946235657,
+      "learning_rate": 0.00027363036747879926,
+      "loss": 4.6232,
+      "step": 1045
+    },
+    {
+      "epoch": 1.33792,
+      "grad_norm": 0.7019796967506409,
+      "learning_rate": 0.0002735899851931619,
+      "loss": 4.516,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.5741701722145081,
+      "learning_rate": 0.00027354960290752457,
+      "loss": 4.5341,
+      "step": 1047
+    },
+    {
+      "epoch": 1.34048,
+      "grad_norm": 0.6854413151741028,
+      "learning_rate": 0.0002735092206218872,
+      "loss": 4.5648,
+      "step": 1048
+    },
+    {
+      "epoch": 1.34176,
+      "grad_norm": 0.599707305431366,
+      "learning_rate": 0.00027346883833624983,
+      "loss": 4.5988,
+      "step": 1049
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.6394950151443481,
+      "learning_rate": 0.0002734284560506124,
+      "loss": 4.5769,
+      "step": 1050
+    },
+    {
+      "epoch": 1.34432,
+      "grad_norm": 0.5611634850502014,
+      "learning_rate": 0.00027338807376497504,
+      "loss": 4.5538,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 0.6633642315864563,
+      "learning_rate": 0.0002733476914793377,
+      "loss": 4.4661,
+      "step": 1052
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.6434425711631775,
+      "learning_rate": 0.00027330730919370035,
+      "loss": 4.5914,
+      "step": 1053
+    },
+    {
+      "epoch": 1.34816,
+      "grad_norm": 0.6126766204833984,
+      "learning_rate": 0.000273266926908063,
+      "loss": 4.5852,
+      "step": 1054
+    },
+    {
+      "epoch": 1.34944,
+      "grad_norm": 0.6653432250022888,
+      "learning_rate": 0.0002732265446224256,
+      "loss": 4.603,
+      "step": 1055
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.7112525701522827,
+      "learning_rate": 0.00027318616233678824,
+      "loss": 4.5842,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.6654365062713623,
+      "learning_rate": 0.00027314578005115087,
+      "loss": 4.5841,
+      "step": 1057
+    },
+    {
+      "epoch": 1.35328,
+      "grad_norm": 0.6465290784835815,
+      "learning_rate": 0.0002731053977655135,
+      "loss": 4.5549,
+      "step": 1058
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.6904200911521912,
+      "learning_rate": 0.00027306501547987613,
+      "loss": 4.6345,
+      "step": 1059
+    },
+    {
+      "epoch": 1.35584,
+      "grad_norm": 0.5807105302810669,
+      "learning_rate": 0.00027302463319423876,
+      "loss": 4.4827,
+      "step": 1060
+    },
+    {
+      "epoch": 1.35712,
+      "grad_norm": 0.6408985257148743,
+      "learning_rate": 0.0002729842509086014,
+      "loss": 4.5378,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.5737316608428955,
+      "learning_rate": 0.000272943868622964,
+      "loss": 4.5322,
+      "step": 1062
+    },
+    {
+      "epoch": 1.35968,
+      "grad_norm": 0.6886394023895264,
+      "learning_rate": 0.00027290348633732665,
+      "loss": 4.5859,
+      "step": 1063
+    },
+    {
+      "epoch": 1.36096,
+      "grad_norm": 0.6456865072250366,
+      "learning_rate": 0.0002728631040516893,
+      "loss": 4.6009,
+      "step": 1064
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.6285141110420227,
+      "learning_rate": 0.00027282272176605197,
+      "loss": 4.6286,
+      "step": 1065
+    },
+    {
+      "epoch": 1.36352,
+      "grad_norm": 0.5843459963798523,
+      "learning_rate": 0.00027278233948041454,
+      "loss": 4.5497,
+      "step": 1066
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 0.6061887145042419,
+      "learning_rate": 0.00027274195719477717,
+      "loss": 4.6066,
+      "step": 1067
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.668018639087677,
+      "learning_rate": 0.0002727015749091398,
+      "loss": 4.5314,
+      "step": 1068
+    },
+    {
+      "epoch": 1.3673600000000001,
+      "grad_norm": 0.6002050042152405,
+      "learning_rate": 0.0002726611926235025,
+      "loss": 4.5271,
+      "step": 1069
+    },
+    {
+      "epoch": 1.36864,
+      "grad_norm": 0.6185527443885803,
+      "learning_rate": 0.0002726208103378651,
+      "loss": 4.5401,
+      "step": 1070
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.6066594123840332,
+      "learning_rate": 0.00027258042805222775,
+      "loss": 4.5988,
+      "step": 1071
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 0.6096265912055969,
+      "learning_rate": 0.0002725400457665904,
+      "loss": 4.5043,
+      "step": 1072
+    },
+    {
+      "epoch": 1.37248,
+      "grad_norm": 0.6521106958389282,
+      "learning_rate": 0.000272499663480953,
+      "loss": 4.5546,
+      "step": 1073
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.582575798034668,
+      "learning_rate": 0.00027245928119531564,
+      "loss": 4.4859,
+      "step": 1074
+    },
+    {
+      "epoch": 1.37504,
+      "grad_norm": 0.5959002375602722,
+      "learning_rate": 0.00027241889890967827,
+      "loss": 4.5375,
+      "step": 1075
+    },
+    {
+      "epoch": 1.37632,
+      "grad_norm": 0.7106321454048157,
+      "learning_rate": 0.0002723785166240409,
+      "loss": 4.53,
+      "step": 1076
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.648299515247345,
+      "learning_rate": 0.0002723381343384035,
+      "loss": 4.6495,
+      "step": 1077
+    },
+    {
+      "epoch": 1.37888,
+      "grad_norm": 0.7060708403587341,
+      "learning_rate": 0.00027229775205276616,
+      "loss": 4.5456,
+      "step": 1078
+    },
+    {
+      "epoch": 1.38016,
+      "grad_norm": 0.655817985534668,
+      "learning_rate": 0.0002722573697671288,
+      "loss": 4.5239,
+      "step": 1079
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.625787079334259,
+      "learning_rate": 0.0002722169874814914,
+      "loss": 4.5517,
+      "step": 1080
+    },
+    {
+      "epoch": 1.38272,
+      "grad_norm": 0.6610313057899475,
+      "learning_rate": 0.00027217660519585405,
+      "loss": 4.4966,
+      "step": 1081
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 0.6236038208007812,
+      "learning_rate": 0.00027213622291021673,
+      "loss": 4.5368,
+      "step": 1082
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.7284337878227234,
+      "learning_rate": 0.0002720958406245793,
+      "loss": 4.5835,
+      "step": 1083
+    },
+    {
+      "epoch": 1.38656,
+      "grad_norm": 0.6679523587226868,
+      "learning_rate": 0.00027205545833894194,
+      "loss": 4.4663,
+      "step": 1084
+    },
+    {
+      "epoch": 1.38784,
+      "grad_norm": 0.6354249715805054,
+      "learning_rate": 0.00027201507605330457,
+      "loss": 4.4995,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.5772947669029236,
+      "learning_rate": 0.00027197469376766725,
+      "loss": 4.4817,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 0.6011262536048889,
+      "learning_rate": 0.0002719343114820299,
+      "loss": 4.5965,
+      "step": 1087
+    },
+    {
+      "epoch": 1.39168,
+      "grad_norm": 0.6319146752357483,
+      "learning_rate": 0.0002718939291963925,
+      "loss": 4.4057,
+      "step": 1088
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.6585641503334045,
+      "learning_rate": 0.00027185354691075514,
+      "loss": 4.4783,
+      "step": 1089
+    },
+    {
+      "epoch": 1.39424,
+      "grad_norm": 0.6011718511581421,
+      "learning_rate": 0.00027181316462511777,
+      "loss": 4.5192,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3955199999999999,
+      "grad_norm": 0.6210798621177673,
+      "learning_rate": 0.0002717727823394804,
+      "loss": 4.5047,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.591758131980896,
+      "learning_rate": 0.00027173240005384303,
+      "loss": 4.5768,
+      "step": 1092
+    },
+    {
+      "epoch": 1.39808,
+      "grad_norm": 0.636916995048523,
+      "learning_rate": 0.00027169201776820566,
+      "loss": 4.5244,
+      "step": 1093
+    },
+    {
+      "epoch": 1.39936,
+      "grad_norm": 0.697331428527832,
+      "learning_rate": 0.0002716516354825683,
+      "loss": 4.5199,
+      "step": 1094
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.6302866339683533,
+      "learning_rate": 0.0002716112531969309,
+      "loss": 4.5211,
+      "step": 1095
+    },
+    {
+      "epoch": 1.40192,
+      "grad_norm": 0.7289062738418579,
+      "learning_rate": 0.00027157087091129355,
+      "loss": 4.4961,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 0.6740280985832214,
+      "learning_rate": 0.0002715304886256562,
+      "loss": 4.5848,
+      "step": 1097
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.6113433241844177,
+      "learning_rate": 0.0002714901063400188,
+      "loss": 4.482,
+      "step": 1098
+    },
+    {
+      "epoch": 1.40576,
+      "grad_norm": 0.6321492195129395,
+      "learning_rate": 0.00027144972405438144,
+      "loss": 4.463,
+      "step": 1099
+    },
+    {
+      "epoch": 1.40704,
+      "grad_norm": 0.6910666227340698,
+      "learning_rate": 0.00027140934176874407,
+      "loss": 4.53,
+      "step": 1100
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.6177242994308472,
+      "learning_rate": 0.0002713689594831067,
+      "loss": 4.5152,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 0.6206321716308594,
+      "learning_rate": 0.00027132857719746933,
+      "loss": 4.5393,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4108800000000001,
+      "grad_norm": 0.6209645867347717,
+      "learning_rate": 0.000271288194911832,
+      "loss": 4.4624,
+      "step": 1103
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.7263031601905823,
+      "learning_rate": 0.00027124781262619465,
+      "loss": 4.5301,
+      "step": 1104
+    },
+    {
+      "epoch": 1.41344,
+      "grad_norm": 0.6254947781562805,
+      "learning_rate": 0.0002712074303405573,
+      "loss": 4.5234,
+      "step": 1105
+    },
+    {
+      "epoch": 1.41472,
+      "grad_norm": 0.6119322180747986,
+      "learning_rate": 0.00027116704805491985,
+      "loss": 4.4959,
+      "step": 1106
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.6006280183792114,
+      "learning_rate": 0.0002711266657692825,
+      "loss": 4.5361,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4172799999999999,
+      "grad_norm": 0.6143077611923218,
+      "learning_rate": 0.00027108628348364517,
+      "loss": 4.4948,
+      "step": 1108
+    },
+    {
+      "epoch": 1.41856,
+      "grad_norm": 0.6156766414642334,
+      "learning_rate": 0.0002710459011980078,
+      "loss": 4.502,
+      "step": 1109
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.6217614412307739,
+      "learning_rate": 0.00027100551891237043,
+      "loss": 4.4685,
+      "step": 1110
+    },
+    {
+      "epoch": 1.42112,
+      "grad_norm": 0.592921793460846,
+      "learning_rate": 0.00027096513662673306,
+      "loss": 4.5341,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 0.5901917815208435,
+      "learning_rate": 0.0002709247543410957,
+      "loss": 4.4953,
+      "step": 1112
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.7119625210762024,
+      "learning_rate": 0.0002708843720554583,
+      "loss": 4.5269,
+      "step": 1113
+    },
+    {
+      "epoch": 1.42496,
+      "grad_norm": 0.7488095164299011,
+      "learning_rate": 0.00027084398976982095,
+      "loss": 4.5553,
+      "step": 1114
+    },
+    {
+      "epoch": 1.42624,
+      "grad_norm": 0.6430398225784302,
+      "learning_rate": 0.0002708036074841836,
+      "loss": 4.524,
+      "step": 1115
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.6690952777862549,
+      "learning_rate": 0.0002707632251985462,
+      "loss": 4.5503,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 0.7325178980827332,
+      "learning_rate": 0.00027072284291290884,
+      "loss": 4.4675,
+      "step": 1117
+    },
+    {
+      "epoch": 1.43008,
+      "grad_norm": 0.6277335286140442,
+      "learning_rate": 0.00027068246062727147,
+      "loss": 4.5683,
+      "step": 1118
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.6613750457763672,
+      "learning_rate": 0.0002706420783416341,
+      "loss": 4.5421,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4326400000000001,
+      "grad_norm": 0.6405130624771118,
+      "learning_rate": 0.00027060169605599673,
+      "loss": 4.4973,
+      "step": 1120
+    },
+    {
+      "epoch": 1.43392,
+      "grad_norm": 0.7388433218002319,
+      "learning_rate": 0.0002705613137703594,
+      "loss": 4.4663,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.7195430397987366,
+      "learning_rate": 0.000270520931484722,
+      "loss": 4.4404,
+      "step": 1122
+    },
+    {
+      "epoch": 1.43648,
+      "grad_norm": 0.5988063812255859,
+      "learning_rate": 0.0002704805491990846,
+      "loss": 4.4924,
+      "step": 1123
+    },
+    {
+      "epoch": 1.43776,
+      "grad_norm": 0.6605175733566284,
+      "learning_rate": 0.00027044016691344725,
+      "loss": 4.4393,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.6142287850379944,
+      "learning_rate": 0.00027039978462780993,
+      "loss": 4.4968,
+      "step": 1125
+    },
+    {
+      "epoch": 1.44032,
+      "grad_norm": 0.6107270121574402,
+      "learning_rate": 0.00027035940234217256,
+      "loss": 4.5411,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 0.5510809421539307,
+      "learning_rate": 0.0002703190200565352,
+      "loss": 4.5815,
+      "step": 1127
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.6231082081794739,
+      "learning_rate": 0.0002702786377708978,
+      "loss": 4.4505,
+      "step": 1128
+    },
+    {
+      "epoch": 1.44416,
+      "grad_norm": 0.6307341456413269,
+      "learning_rate": 0.00027023825548526045,
+      "loss": 4.5774,
+      "step": 1129
+    },
+    {
+      "epoch": 1.44544,
+      "grad_norm": 0.5655580759048462,
+      "learning_rate": 0.0002701978731996231,
+      "loss": 4.451,
+      "step": 1130
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.6464011669158936,
+      "learning_rate": 0.0002701574909139857,
+      "loss": 4.446,
+      "step": 1131
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.6590990424156189,
+      "learning_rate": 0.00027011710862834834,
+      "loss": 4.5092,
+      "step": 1132
+    },
+    {
+      "epoch": 1.44928,
+      "grad_norm": 0.5793240666389465,
+      "learning_rate": 0.000270076726342711,
+      "loss": 4.4777,
+      "step": 1133
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.6022570729255676,
+      "learning_rate": 0.0002700363440570736,
+      "loss": 4.4337,
+      "step": 1134
+    },
+    {
+      "epoch": 1.45184,
+      "grad_norm": 0.6188995838165283,
+      "learning_rate": 0.00026999596177143623,
+      "loss": 4.5174,
+      "step": 1135
+    },
+    {
+      "epoch": 1.45312,
+      "grad_norm": 0.6144684553146362,
+      "learning_rate": 0.00026995557948579886,
+      "loss": 4.4565,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.6763806939125061,
+      "learning_rate": 0.0002699151972001615,
+      "loss": 4.4995,
+      "step": 1137
+    },
+    {
+      "epoch": 1.45568,
+      "grad_norm": 0.7390356063842773,
+      "learning_rate": 0.0002698748149145241,
+      "loss": 4.5674,
+      "step": 1138
+    },
+    {
+      "epoch": 1.45696,
+      "grad_norm": 0.6634504199028015,
+      "learning_rate": 0.00026983443262888675,
+      "loss": 4.5301,
+      "step": 1139
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.6377415060997009,
+      "learning_rate": 0.0002697940503432494,
+      "loss": 4.4919,
+      "step": 1140
+    },
+    {
+      "epoch": 1.45952,
+      "grad_norm": 0.6241118907928467,
+      "learning_rate": 0.000269753668057612,
+      "loss": 4.5093,
+      "step": 1141
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 0.676541805267334,
+      "learning_rate": 0.0002697132857719747,
+      "loss": 4.5714,
+      "step": 1142
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.6231991648674011,
+      "learning_rate": 0.00026967290348633733,
+      "loss": 4.494,
+      "step": 1143
+    },
+    {
+      "epoch": 1.46336,
+      "grad_norm": 0.6391655802726746,
+      "learning_rate": 0.00026963252120069996,
+      "loss": 4.5113,
+      "step": 1144
+    },
+    {
+      "epoch": 1.46464,
+      "grad_norm": 0.6239742636680603,
+      "learning_rate": 0.00026959213891506253,
+      "loss": 4.4657,
+      "step": 1145
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.6046785116195679,
+      "learning_rate": 0.00026955175662942516,
+      "loss": 4.5132,
+      "step": 1146
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 0.663732647895813,
+      "learning_rate": 0.00026951137434378785,
+      "loss": 4.5261,
+      "step": 1147
+    },
+    {
+      "epoch": 1.46848,
+      "grad_norm": 0.6262268424034119,
+      "learning_rate": 0.0002694709920581505,
+      "loss": 4.5257,
+      "step": 1148
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.552179217338562,
+      "learning_rate": 0.0002694306097725131,
+      "loss": 4.5232,
+      "step": 1149
+    },
+    {
+      "epoch": 1.47104,
+      "grad_norm": 0.6553817987442017,
+      "learning_rate": 0.00026939022748687574,
+      "loss": 4.442,
+      "step": 1150
+    },
+    {
+      "epoch": 1.47232,
+      "grad_norm": 0.6651473641395569,
+      "learning_rate": 0.00026934984520123837,
+      "loss": 4.5369,
+      "step": 1151
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.6147942543029785,
+      "learning_rate": 0.000269309462915601,
+      "loss": 4.5399,
+      "step": 1152
+    },
+    {
+      "epoch": 1.47488,
+      "grad_norm": 0.662200927734375,
+      "learning_rate": 0.00026926908062996363,
+      "loss": 4.5464,
+      "step": 1153
+    },
+    {
+      "epoch": 1.4761600000000001,
+      "grad_norm": 0.6553072929382324,
+      "learning_rate": 0.00026922869834432626,
+      "loss": 4.5623,
+      "step": 1154
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.5930311679840088,
+      "learning_rate": 0.0002691883160586889,
+      "loss": 4.4286,
+      "step": 1155
+    },
+    {
+      "epoch": 1.47872,
+      "grad_norm": 0.6711944341659546,
+      "learning_rate": 0.0002691479337730515,
+      "loss": 4.4298,
+      "step": 1156
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5974631905555725,
+      "learning_rate": 0.00026910755148741415,
+      "loss": 4.4365,
+      "step": 1157
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.6224148273468018,
+      "learning_rate": 0.0002690671692017768,
+      "loss": 4.4888,
+      "step": 1158
+    },
+    {
+      "epoch": 1.4825599999999999,
+      "grad_norm": 0.5823886394500732,
+      "learning_rate": 0.0002690267869161394,
+      "loss": 4.48,
+      "step": 1159
+    },
+    {
+      "epoch": 1.48384,
+      "grad_norm": 0.6274778842926025,
+      "learning_rate": 0.0002689864046305021,
+      "loss": 4.5036,
+      "step": 1160
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.6259602904319763,
+      "learning_rate": 0.0002689460223448647,
+      "loss": 4.5225,
+      "step": 1161
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 0.5894110798835754,
+      "learning_rate": 0.0002689056400592273,
+      "loss": 4.5004,
+      "step": 1162
+    },
+    {
+      "epoch": 1.4876800000000001,
+      "grad_norm": 0.6239898204803467,
+      "learning_rate": 0.00026886525777358993,
+      "loss": 4.5111,
+      "step": 1163
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.6786986589431763,
+      "learning_rate": 0.0002688248754879526,
+      "loss": 4.5008,
+      "step": 1164
+    },
+    {
+      "epoch": 1.49024,
+      "grad_norm": 0.5871908664703369,
+      "learning_rate": 0.00026878449320231524,
+      "loss": 4.5222,
+      "step": 1165
+    },
+    {
+      "epoch": 1.49152,
+      "grad_norm": 0.7175498604774475,
+      "learning_rate": 0.0002687441109166779,
+      "loss": 4.5113,
+      "step": 1166
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.6610713005065918,
+      "learning_rate": 0.0002687037286310405,
+      "loss": 4.489,
+      "step": 1167
+    },
+    {
+      "epoch": 1.49408,
+      "grad_norm": 0.6492409110069275,
+      "learning_rate": 0.00026866334634540313,
+      "loss": 4.5002,
+      "step": 1168
+    },
+    {
+      "epoch": 1.49536,
+      "grad_norm": 0.637315034866333,
+      "learning_rate": 0.00026862296405976576,
+      "loss": 4.503,
+      "step": 1169
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.6214714646339417,
+      "learning_rate": 0.0002685825817741284,
+      "loss": 4.5694,
+      "step": 1170
+    },
+    {
+      "epoch": 1.49792,
+      "grad_norm": 0.6326239109039307,
+      "learning_rate": 0.000268542199488491,
+      "loss": 4.4157,
+      "step": 1171
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 0.726486086845398,
+      "learning_rate": 0.00026850181720285366,
+      "loss": 4.445,
+      "step": 1172
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.5670982599258423,
+      "learning_rate": 0.0002684614349172163,
+      "loss": 4.5223,
+      "step": 1173
+    },
+    {
+      "epoch": 1.50176,
+      "grad_norm": 0.6748657822608948,
+      "learning_rate": 0.0002684210526315789,
+      "loss": 4.4798,
+      "step": 1174
+    },
+    {
+      "epoch": 1.50304,
+      "grad_norm": 0.6664227843284607,
+      "learning_rate": 0.00026838067034594155,
+      "loss": 4.5394,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.6093580722808838,
+      "learning_rate": 0.0002683402880603042,
+      "loss": 4.4771,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 0.593794047832489,
+      "learning_rate": 0.00026829990577466686,
+      "loss": 4.4568,
+      "step": 1177
+    },
+    {
+      "epoch": 1.50688,
+      "grad_norm": 0.5962923765182495,
+      "learning_rate": 0.00026825952348902944,
+      "loss": 4.5111,
+      "step": 1178
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.5938080549240112,
+      "learning_rate": 0.00026821914120339207,
+      "loss": 4.4462,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5094400000000001,
+      "grad_norm": 0.600837767124176,
+      "learning_rate": 0.0002681787589177547,
+      "loss": 4.5177,
+      "step": 1180
+    },
+    {
+      "epoch": 1.51072,
+      "grad_norm": 0.601696252822876,
+      "learning_rate": 0.0002681383766321174,
+      "loss": 4.4331,
+      "step": 1181
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.5914311408996582,
+      "learning_rate": 0.00026809799434648,
+      "loss": 4.4859,
+      "step": 1182
+    },
+    {
+      "epoch": 1.51328,
+      "grad_norm": 0.6770086288452148,
+      "learning_rate": 0.00026805761206084264,
+      "loss": 4.4447,
+      "step": 1183
+    },
+    {
+      "epoch": 1.51456,
+      "grad_norm": 0.6460147500038147,
+      "learning_rate": 0.00026801722977520527,
+      "loss": 4.4399,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.5961685180664062,
+      "learning_rate": 0.00026797684748956785,
+      "loss": 4.4354,
+      "step": 1185
+    },
+    {
+      "epoch": 1.51712,
+      "grad_norm": 0.6054593324661255,
+      "learning_rate": 0.00026793646520393053,
+      "loss": 4.3969,
+      "step": 1186
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 0.6508775949478149,
+      "learning_rate": 0.00026789608291829316,
+      "loss": 4.4683,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.7032009363174438,
+      "learning_rate": 0.0002678557006326558,
+      "loss": 4.4864,
+      "step": 1188
+    },
+    {
+      "epoch": 1.52096,
+      "grad_norm": 0.6030545234680176,
+      "learning_rate": 0.0002678153183470184,
+      "loss": 4.389,
+      "step": 1189
+    },
+    {
+      "epoch": 1.52224,
+      "grad_norm": 0.5884782075881958,
+      "learning_rate": 0.00026777493606138105,
+      "loss": 4.5031,
+      "step": 1190
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.6099117994308472,
+      "learning_rate": 0.0002677345537757437,
+      "loss": 4.4668,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 0.5954748392105103,
+      "learning_rate": 0.0002676941714901063,
+      "loss": 4.4554,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5260799999999999,
+      "grad_norm": 0.5383332371711731,
+      "learning_rate": 0.00026765378920446894,
+      "loss": 4.5648,
+      "step": 1193
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.5823171138763428,
+      "learning_rate": 0.00026761340691883157,
+      "loss": 4.5046,
+      "step": 1194
+    },
+    {
+      "epoch": 1.52864,
+      "grad_norm": 0.6584584712982178,
+      "learning_rate": 0.0002675730246331942,
+      "loss": 4.4529,
+      "step": 1195
+    },
+    {
+      "epoch": 1.52992,
+      "grad_norm": 0.5978401899337769,
+      "learning_rate": 0.00026753264234755683,
+      "loss": 4.4333,
+      "step": 1196
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.591083288192749,
+      "learning_rate": 0.00026749226006191946,
+      "loss": 4.4279,
+      "step": 1197
+    },
+    {
+      "epoch": 1.53248,
+      "grad_norm": 0.5882396101951599,
+      "learning_rate": 0.00026745187777628215,
+      "loss": 4.3951,
+      "step": 1198
+    },
+    {
+      "epoch": 1.53376,
+      "grad_norm": 0.571337103843689,
+      "learning_rate": 0.0002674114954906448,
+      "loss": 4.4579,
+      "step": 1199
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.5543642044067383,
+      "learning_rate": 0.0002673711132050074,
+      "loss": 4.4323,
+      "step": 1200
+    },
+    {
+      "epoch": 1.53632,
+      "grad_norm": 0.5652990341186523,
+      "learning_rate": 0.00026733073091937,
+      "loss": 4.4284,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5375999999999999,
+      "grad_norm": 0.5995742678642273,
+      "learning_rate": 0.0002672903486337326,
+      "loss": 4.4789,
+      "step": 1202
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 0.5706351399421692,
+      "learning_rate": 0.0002672499663480953,
+      "loss": 4.4814,
+      "step": 1203
+    },
+    {
+      "epoch": 1.54016,
+      "grad_norm": 0.5830731391906738,
+      "learning_rate": 0.0002672095840624579,
+      "loss": 4.4525,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5414400000000001,
+      "grad_norm": 0.6036829352378845,
+      "learning_rate": 0.00026716920177682056,
+      "loss": 4.3816,
+      "step": 1205
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 0.5913065671920776,
+      "learning_rate": 0.0002671288194911832,
+      "loss": 4.4997,
+      "step": 1206
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 0.6033707857131958,
+      "learning_rate": 0.0002670884372055458,
+      "loss": 4.4725,
+      "step": 1207
+    },
+    {
+      "epoch": 1.54528,
+      "grad_norm": 0.6108120679855347,
+      "learning_rate": 0.00026704805491990845,
+      "loss": 4.4673,
+      "step": 1208
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 0.6634685397148132,
+      "learning_rate": 0.0002670076726342711,
+      "loss": 4.4737,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5478399999999999,
+      "grad_norm": 0.5890428423881531,
+      "learning_rate": 0.0002669672903486337,
+      "loss": 4.4499,
+      "step": 1210
+    },
+    {
+      "epoch": 1.54912,
+      "grad_norm": 0.6377815008163452,
+      "learning_rate": 0.00026692690806299634,
+      "loss": 4.4546,
+      "step": 1211
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.6385360956192017,
+      "learning_rate": 0.00026688652577735897,
+      "loss": 4.4226,
+      "step": 1212
+    },
+    {
+      "epoch": 1.55168,
+      "grad_norm": 0.5952944159507751,
+      "learning_rate": 0.0002668461434917216,
+      "loss": 4.4762,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5529600000000001,
+      "grad_norm": 0.5970460772514343,
+      "learning_rate": 0.0002668057612060842,
+      "loss": 4.4617,
+      "step": 1214
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 0.6275203227996826,
+      "learning_rate": 0.00026676537892044686,
+      "loss": 4.4252,
+      "step": 1215
+    },
+    {
+      "epoch": 1.55552,
+      "grad_norm": 0.6631754636764526,
+      "learning_rate": 0.00026672499663480954,
+      "loss": 4.5075,
+      "step": 1216
+    },
+    {
+      "epoch": 1.5568,
+      "grad_norm": 0.6074349880218506,
+      "learning_rate": 0.0002666846143491721,
+      "loss": 4.442,
+      "step": 1217
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 0.6577631235122681,
+      "learning_rate": 0.00026664423206353475,
+      "loss": 4.4871,
+      "step": 1218
+    },
+    {
+      "epoch": 1.5593599999999999,
+      "grad_norm": 0.6167892217636108,
+      "learning_rate": 0.0002666038497778974,
+      "loss": 4.4119,
+      "step": 1219
+    },
+    {
+      "epoch": 1.56064,
+      "grad_norm": 0.6109944581985474,
+      "learning_rate": 0.00026656346749226006,
+      "loss": 4.439,
+      "step": 1220
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 0.6446990966796875,
+      "learning_rate": 0.0002665230852066227,
+      "loss": 4.4472,
+      "step": 1221
+    },
+    {
+      "epoch": 1.5632000000000001,
+      "grad_norm": 0.5861752033233643,
+      "learning_rate": 0.0002664827029209853,
+      "loss": 4.4121,
+      "step": 1222
+    },
+    {
+      "epoch": 1.56448,
+      "grad_norm": 0.620345413684845,
+      "learning_rate": 0.00026644232063534795,
+      "loss": 4.5141,
+      "step": 1223
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 0.6211066842079163,
+      "learning_rate": 0.0002664019383497106,
+      "loss": 4.4676,
+      "step": 1224
+    },
+    {
+      "epoch": 1.56704,
+      "grad_norm": 0.5987910628318787,
+      "learning_rate": 0.0002663615560640732,
+      "loss": 4.4051,
+      "step": 1225
+    },
+    {
+      "epoch": 1.56832,
+      "grad_norm": 0.6385419368743896,
+      "learning_rate": 0.00026632117377843584,
+      "loss": 4.457,
+      "step": 1226
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.6010638475418091,
+      "learning_rate": 0.00026628079149279847,
+      "loss": 4.4322,
+      "step": 1227
+    },
+    {
+      "epoch": 1.57088,
+      "grad_norm": 0.6236041784286499,
+      "learning_rate": 0.0002662404092071611,
+      "loss": 4.4491,
+      "step": 1228
+    },
+    {
+      "epoch": 1.57216,
+      "grad_norm": 0.6467570066452026,
+      "learning_rate": 0.00026620002692152373,
+      "loss": 4.4418,
+      "step": 1229
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 0.5972265601158142,
+      "learning_rate": 0.00026615964463588636,
+      "loss": 4.4185,
+      "step": 1230
+    },
+    {
+      "epoch": 1.5747200000000001,
+      "grad_norm": 0.6430701017379761,
+      "learning_rate": 0.000266119262350249,
+      "loss": 4.4144,
+      "step": 1231
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 0.5837542414665222,
+      "learning_rate": 0.0002660788800646116,
+      "loss": 4.4138,
+      "step": 1232
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 0.6243807077407837,
+      "learning_rate": 0.0002660384977789743,
+      "loss": 4.4507,
+      "step": 1233
+    },
+    {
+      "epoch": 1.57856,
+      "grad_norm": 0.5919942259788513,
+      "learning_rate": 0.0002659981154933369,
+      "loss": 4.4508,
+      "step": 1234
+    },
+    {
+      "epoch": 1.57984,
+      "grad_norm": 0.540010929107666,
+      "learning_rate": 0.0002659577332076995,
+      "loss": 4.3369,
+      "step": 1235
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 0.568360447883606,
+      "learning_rate": 0.00026591735092206214,
+      "loss": 4.4326,
+      "step": 1236
+    },
+    {
+      "epoch": 1.5824,
+      "grad_norm": 0.5942397117614746,
+      "learning_rate": 0.0002658769686364248,
+      "loss": 4.412,
+      "step": 1237
+    },
+    {
+      "epoch": 1.58368,
+      "grad_norm": 0.5995660424232483,
+      "learning_rate": 0.00026583658635078746,
+      "loss": 4.4086,
+      "step": 1238
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 0.620974063873291,
+      "learning_rate": 0.0002657962040651501,
+      "loss": 4.3333,
+      "step": 1239
+    },
+    {
+      "epoch": 1.58624,
+      "grad_norm": 0.6151325702667236,
+      "learning_rate": 0.00026575582177951266,
+      "loss": 4.4844,
+      "step": 1240
+    },
+    {
+      "epoch": 1.58752,
+      "grad_norm": 0.640434741973877,
+      "learning_rate": 0.0002657154394938753,
+      "loss": 4.3976,
+      "step": 1241
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.5995258688926697,
+      "learning_rate": 0.000265675057208238,
+      "loss": 4.3342,
+      "step": 1242
+    },
+    {
+      "epoch": 1.59008,
+      "grad_norm": 0.5680806636810303,
+      "learning_rate": 0.0002656346749226006,
+      "loss": 4.4304,
+      "step": 1243
+    },
+    {
+      "epoch": 1.5913599999999999,
+      "grad_norm": 0.6433593034744263,
+      "learning_rate": 0.00026559429263696324,
+      "loss": 4.375,
+      "step": 1244
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 0.5471886992454529,
+      "learning_rate": 0.00026555391035132587,
+      "loss": 4.3626,
+      "step": 1245
+    },
+    {
+      "epoch": 1.59392,
+      "grad_norm": 0.6657853126525879,
+      "learning_rate": 0.0002655135280656885,
+      "loss": 4.3994,
+      "step": 1246
+    },
+    {
+      "epoch": 1.5952,
+      "grad_norm": 0.6484869122505188,
+      "learning_rate": 0.00026547314578005113,
+      "loss": 4.4166,
+      "step": 1247
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 0.6245217323303223,
+      "learning_rate": 0.00026543276349441376,
+      "loss": 4.3975,
+      "step": 1248
+    },
+    {
+      "epoch": 1.59776,
+      "grad_norm": 0.6242031455039978,
+      "learning_rate": 0.0002653923812087764,
+      "loss": 4.4047,
+      "step": 1249
+    },
+    {
+      "epoch": 1.59904,
+      "grad_norm": 0.6540752649307251,
+      "learning_rate": 0.000265351998923139,
+      "loss": 4.3767,
+      "step": 1250
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 0.5985410213470459,
+      "learning_rate": 0.00026531161663750165,
+      "loss": 4.4277,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6016,
+      "grad_norm": 0.7160242795944214,
+      "learning_rate": 0.0002652712343518643,
+      "loss": 4.3891,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6028799999999999,
+      "grad_norm": 0.5682446360588074,
+      "learning_rate": 0.0002652308520662269,
+      "loss": 4.3845,
+      "step": 1253
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 0.6273159980773926,
+      "learning_rate": 0.00026519046978058954,
+      "loss": 4.3962,
+      "step": 1254
+    },
+    {
+      "epoch": 1.60544,
+      "grad_norm": 0.631803035736084,
+      "learning_rate": 0.0002651500874949522,
+      "loss": 4.4692,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6067200000000001,
+      "grad_norm": 0.5755056142807007,
+      "learning_rate": 0.00026510970520931485,
+      "loss": 4.4311,
+      "step": 1256
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.5929430723190308,
+      "learning_rate": 0.00026506932292367743,
+      "loss": 4.4624,
+      "step": 1257
+    },
+    {
+      "epoch": 1.60928,
+      "grad_norm": 0.6523311138153076,
+      "learning_rate": 0.00026502894063804006,
+      "loss": 4.417,
+      "step": 1258
+    },
+    {
+      "epoch": 1.61056,
+      "grad_norm": 0.5619366765022278,
+      "learning_rate": 0.00026498855835240274,
+      "loss": 4.3603,
+      "step": 1259
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 0.60685795545578,
+      "learning_rate": 0.00026494817606676537,
+      "loss": 4.4113,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6131199999999999,
+      "grad_norm": 0.5795391201972961,
+      "learning_rate": 0.000264907793781128,
+      "loss": 4.3658,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6143999999999998,
+      "grad_norm": 0.6668011546134949,
+      "learning_rate": 0.00026486741149549063,
+      "loss": 4.3773,
+      "step": 1262
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 0.7039081454277039,
+      "learning_rate": 0.00026482702920985326,
+      "loss": 4.4388,
+      "step": 1263
+    },
+    {
+      "epoch": 1.61696,
+      "grad_norm": 0.6980475187301636,
+      "learning_rate": 0.0002647866469242159,
+      "loss": 4.4047,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6182400000000001,
+      "grad_norm": 0.6274135112762451,
+      "learning_rate": 0.0002647462646385785,
+      "loss": 4.3472,
+      "step": 1265
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 0.5991054773330688,
+      "learning_rate": 0.00026470588235294115,
+      "loss": 4.402,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6208,
+      "grad_norm": 0.5963894724845886,
+      "learning_rate": 0.0002646655000673038,
+      "loss": 4.4285,
+      "step": 1267
+    },
+    {
+      "epoch": 1.62208,
+      "grad_norm": 0.6448376178741455,
+      "learning_rate": 0.0002646251177816664,
+      "loss": 4.3825,
+      "step": 1268
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 0.5770543217658997,
+      "learning_rate": 0.00026458473549602904,
+      "loss": 4.3484,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6246399999999999,
+      "grad_norm": 0.6031239032745361,
+      "learning_rate": 0.0002645443532103917,
+      "loss": 4.3538,
+      "step": 1270
+    },
+    {
+      "epoch": 1.62592,
+      "grad_norm": 0.5922113060951233,
+      "learning_rate": 0.0002645039709247543,
+      "loss": 4.3903,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.5830681324005127,
+      "learning_rate": 0.000264463588639117,
+      "loss": 4.4233,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6284800000000001,
+      "grad_norm": 0.5209598541259766,
+      "learning_rate": 0.00026442320635347956,
+      "loss": 4.404,
+      "step": 1273
+    },
+    {
+      "epoch": 1.62976,
+      "grad_norm": 0.6153559684753418,
+      "learning_rate": 0.0002643828240678422,
+      "loss": 4.3666,
+      "step": 1274
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 0.566469132900238,
+      "learning_rate": 0.0002643424417822048,
+      "loss": 4.3731,
+      "step": 1275
+    },
+    {
+      "epoch": 1.63232,
+      "grad_norm": 0.5851961374282837,
+      "learning_rate": 0.0002643020594965675,
+      "loss": 4.3827,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6336,
+      "grad_norm": 0.5738173127174377,
+      "learning_rate": 0.00026426167721093014,
+      "loss": 4.3779,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 0.5467177629470825,
+      "learning_rate": 0.00026422129492529277,
+      "loss": 4.3595,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6361599999999998,
+      "grad_norm": 0.5850498080253601,
+      "learning_rate": 0.0002641809126396554,
+      "loss": 4.3581,
+      "step": 1279
+    },
+    {
+      "epoch": 1.63744,
+      "grad_norm": 0.5361942052841187,
+      "learning_rate": 0.000264140530354018,
+      "loss": 4.4323,
+      "step": 1280
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 0.5595226883888245,
+      "learning_rate": 0.00026410014806838066,
+      "loss": 4.4528,
+      "step": 1281
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.6067069172859192,
+      "learning_rate": 0.0002640597657827433,
+      "loss": 4.3867,
+      "step": 1282
+    },
+    {
+      "epoch": 1.64128,
+      "grad_norm": 0.5635828971862793,
+      "learning_rate": 0.0002640193834971059,
+      "loss": 4.3671,
+      "step": 1283
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 0.6461728811264038,
+      "learning_rate": 0.00026397900121146855,
+      "loss": 4.3887,
+      "step": 1284
+    },
+    {
+      "epoch": 1.64384,
+      "grad_norm": 0.5938706398010254,
+      "learning_rate": 0.0002639386189258312,
+      "loss": 4.4002,
+      "step": 1285
+    },
+    {
+      "epoch": 1.64512,
+      "grad_norm": 0.6693472862243652,
+      "learning_rate": 0.0002638982366401938,
+      "loss": 4.3767,
+      "step": 1286
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.6100233197212219,
+      "learning_rate": 0.00026385785435455644,
+      "loss": 4.4038,
+      "step": 1287
+    },
+    {
+      "epoch": 1.64768,
+      "grad_norm": 0.5637868642807007,
+      "learning_rate": 0.00026381747206891907,
+      "loss": 4.3205,
+      "step": 1288
+    },
+    {
+      "epoch": 1.64896,
+      "grad_norm": 0.6211172342300415,
+      "learning_rate": 0.0002637770897832817,
+      "loss": 4.3709,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 0.5223026871681213,
+      "learning_rate": 0.00026373670749764433,
+      "loss": 4.381,
+      "step": 1290
+    },
+    {
+      "epoch": 1.65152,
+      "grad_norm": 0.5736677646636963,
+      "learning_rate": 0.00026369632521200696,
+      "loss": 4.3631,
+      "step": 1291
+    },
+    {
+      "epoch": 1.6528,
+      "grad_norm": 0.5660174489021301,
+      "learning_rate": 0.0002636559429263696,
+      "loss": 4.4151,
+      "step": 1292
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 0.5827959775924683,
+      "learning_rate": 0.0002636155606407322,
+      "loss": 4.2954,
+      "step": 1293
+    },
+    {
+      "epoch": 1.65536,
+      "grad_norm": 0.6205708384513855,
+      "learning_rate": 0.0002635751783550949,
+      "loss": 4.3476,
+      "step": 1294
+    },
+    {
+      "epoch": 1.65664,
+      "grad_norm": 0.599597692489624,
+      "learning_rate": 0.00026353479606945753,
+      "loss": 4.3624,
+      "step": 1295
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 0.6274026036262512,
+      "learning_rate": 0.0002634944137838201,
+      "loss": 4.4197,
+      "step": 1296
+    },
+    {
+      "epoch": 1.6592,
+      "grad_norm": 0.570179283618927,
+      "learning_rate": 0.00026345403149818274,
+      "loss": 4.4012,
+      "step": 1297
+    },
+    {
+      "epoch": 1.66048,
+      "grad_norm": 0.618751049041748,
+      "learning_rate": 0.0002634136492125454,
+      "loss": 4.3122,
+      "step": 1298
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 0.6431691646575928,
+      "learning_rate": 0.00026337326692690805,
+      "loss": 4.4175,
+      "step": 1299
+    },
+    {
+      "epoch": 1.66304,
+      "grad_norm": 0.6071829199790955,
+      "learning_rate": 0.0002633328846412707,
+      "loss": 4.394,
+      "step": 1300
+    },
+    {
+      "epoch": 1.66432,
+      "grad_norm": 0.5877499580383301,
+      "learning_rate": 0.0002632925023556333,
+      "loss": 4.3989,
+      "step": 1301
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.5928570628166199,
+      "learning_rate": 0.00026325212006999594,
+      "loss": 4.4153,
+      "step": 1302
+    },
+    {
+      "epoch": 1.66688,
+      "grad_norm": 0.5851948261260986,
+      "learning_rate": 0.0002632117377843586,
+      "loss": 4.4008,
+      "step": 1303
+    },
+    {
+      "epoch": 1.6681599999999999,
+      "grad_norm": 0.5977253317832947,
+      "learning_rate": 0.0002631713554987212,
+      "loss": 4.3256,
+      "step": 1304
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 0.5974704027175903,
+      "learning_rate": 0.00026313097321308383,
+      "loss": 4.3814,
+      "step": 1305
+    },
+    {
+      "epoch": 1.67072,
+      "grad_norm": 0.631009578704834,
+      "learning_rate": 0.00026309059092744646,
+      "loss": 4.3496,
+      "step": 1306
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 0.5566964149475098,
+      "learning_rate": 0.0002630502086418091,
+      "loss": 4.4002,
+      "step": 1307
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 0.5787959694862366,
+      "learning_rate": 0.0002630098263561717,
+      "loss": 4.3317,
+      "step": 1308
+    },
+    {
+      "epoch": 1.67456,
+      "grad_norm": 0.6714434623718262,
+      "learning_rate": 0.00026296944407053435,
+      "loss": 4.3374,
+      "step": 1309
+    },
+    {
+      "epoch": 1.67584,
+      "grad_norm": 0.6594877243041992,
+      "learning_rate": 0.000262929061784897,
+      "loss": 4.4183,
+      "step": 1310
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 0.5786787867546082,
+      "learning_rate": 0.00026288867949925967,
+      "loss": 4.3543,
+      "step": 1311
+    },
+    {
+      "epoch": 1.6784,
+      "grad_norm": 0.5720179677009583,
+      "learning_rate": 0.00026284829721362225,
+      "loss": 4.3702,
+      "step": 1312
+    },
+    {
+      "epoch": 1.6796799999999998,
+      "grad_norm": 0.6159886121749878,
+      "learning_rate": 0.0002628079149279849,
+      "loss": 4.3829,
+      "step": 1313
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 0.6541562080383301,
+      "learning_rate": 0.0002627675326423475,
+      "loss": 4.3588,
+      "step": 1314
+    },
+    {
+      "epoch": 1.68224,
+      "grad_norm": 0.6164893507957458,
+      "learning_rate": 0.0002627271503567102,
+      "loss": 4.3731,
+      "step": 1315
+    },
+    {
+      "epoch": 1.6835200000000001,
+      "grad_norm": 0.611280083656311,
+      "learning_rate": 0.0002626867680710728,
+      "loss": 4.454,
+      "step": 1316
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.6023463606834412,
+      "learning_rate": 0.00026264638578543545,
+      "loss": 4.3007,
+      "step": 1317
+    },
+    {
+      "epoch": 1.68608,
+      "grad_norm": 0.6427450180053711,
+      "learning_rate": 0.0002626060034997981,
+      "loss": 4.3692,
+      "step": 1318
+    },
+    {
+      "epoch": 1.68736,
+      "grad_norm": 0.5956165194511414,
+      "learning_rate": 0.00026256562121416066,
+      "loss": 4.352,
+      "step": 1319
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 0.5930037498474121,
+      "learning_rate": 0.00026252523892852334,
+      "loss": 4.3185,
+      "step": 1320
+    },
+    {
+      "epoch": 1.6899199999999999,
+      "grad_norm": 0.5842078328132629,
+      "learning_rate": 0.00026248485664288597,
+      "loss": 4.3912,
+      "step": 1321
+    },
+    {
+      "epoch": 1.6912,
+      "grad_norm": 0.5779310464859009,
+      "learning_rate": 0.0002624444743572486,
+      "loss": 4.4048,
+      "step": 1322
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 0.5934463739395142,
+      "learning_rate": 0.00026240409207161123,
+      "loss": 4.3657,
+      "step": 1323
+    },
+    {
+      "epoch": 1.6937600000000002,
+      "grad_norm": 0.636505663394928,
+      "learning_rate": 0.00026236370978597386,
+      "loss": 4.4074,
+      "step": 1324
+    },
+    {
+      "epoch": 1.69504,
+      "grad_norm": 0.6304707527160645,
+      "learning_rate": 0.0002623233275003365,
+      "loss": 4.4129,
+      "step": 1325
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 0.660848081111908,
+      "learning_rate": 0.0002622829452146991,
+      "loss": 4.3167,
+      "step": 1326
+    },
+    {
+      "epoch": 1.6976,
+      "grad_norm": 0.5828097462654114,
+      "learning_rate": 0.00026224256292906175,
+      "loss": 4.3415,
+      "step": 1327
+    },
+    {
+      "epoch": 1.69888,
+      "grad_norm": 0.6500957608222961,
+      "learning_rate": 0.00026220218064342443,
+      "loss": 4.3257,
+      "step": 1328
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 0.5696218013763428,
+      "learning_rate": 0.000262161798357787,
+      "loss": 4.423,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7014399999999998,
+      "grad_norm": 0.6378457546234131,
+      "learning_rate": 0.00026212141607214964,
+      "loss": 4.371,
+      "step": 1330
+    },
+    {
+      "epoch": 1.70272,
+      "grad_norm": 0.5619131326675415,
+      "learning_rate": 0.00026208103378651227,
+      "loss": 4.3918,
+      "step": 1331
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.607213020324707,
+      "learning_rate": 0.00026204065150087496,
+      "loss": 4.2955,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7052800000000001,
+      "grad_norm": 0.6598789095878601,
+      "learning_rate": 0.0002620002692152376,
+      "loss": 4.4082,
+      "step": 1333
+    },
+    {
+      "epoch": 1.70656,
+      "grad_norm": 0.6173972487449646,
+      "learning_rate": 0.0002619598869296002,
+      "loss": 4.3991,
+      "step": 1334
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 0.709816575050354,
+      "learning_rate": 0.00026191950464396285,
+      "loss": 4.3563,
+      "step": 1335
+    },
+    {
+      "epoch": 1.70912,
+      "grad_norm": 0.5555682182312012,
+      "learning_rate": 0.0002618791223583254,
+      "loss": 4.386,
+      "step": 1336
+    },
+    {
+      "epoch": 1.7104,
+      "grad_norm": 0.6640469431877136,
+      "learning_rate": 0.0002618387400726881,
+      "loss": 4.3813,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 0.5649694800376892,
+      "learning_rate": 0.00026179835778705074,
+      "loss": 4.3737,
+      "step": 1338
+    },
+    {
+      "epoch": 1.71296,
+      "grad_norm": 0.5696978569030762,
+      "learning_rate": 0.00026175797550141337,
+      "loss": 4.3254,
+      "step": 1339
+    },
+    {
+      "epoch": 1.71424,
+      "grad_norm": 0.5677264928817749,
+      "learning_rate": 0.000261717593215776,
+      "loss": 4.3328,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 0.5705046057701111,
+      "learning_rate": 0.0002616772109301386,
+      "loss": 4.2976,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7168,
+      "grad_norm": 0.5434849858283997,
+      "learning_rate": 0.00026163682864450126,
+      "loss": 4.4098,
+      "step": 1342
+    },
+    {
+      "epoch": 1.71808,
+      "grad_norm": 0.6120153665542603,
+      "learning_rate": 0.0002615964463588639,
+      "loss": 4.3524,
+      "step": 1343
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 0.5820379853248596,
+      "learning_rate": 0.0002615560640732265,
+      "loss": 4.3321,
+      "step": 1344
+    },
+    {
+      "epoch": 1.72064,
+      "grad_norm": 0.586432695388794,
+      "learning_rate": 0.00026151568178758915,
+      "loss": 4.3806,
+      "step": 1345
+    },
+    {
+      "epoch": 1.72192,
+      "grad_norm": 0.6412893533706665,
+      "learning_rate": 0.0002614752995019518,
+      "loss": 4.3795,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.6333811283111572,
+      "learning_rate": 0.0002614349172163144,
+      "loss": 4.3555,
+      "step": 1347
+    },
+    {
+      "epoch": 1.72448,
+      "grad_norm": 0.5543333292007446,
+      "learning_rate": 0.00026139453493067704,
+      "loss": 4.3285,
+      "step": 1348
+    },
+    {
+      "epoch": 1.72576,
+      "grad_norm": 0.5972537398338318,
+      "learning_rate": 0.00026135415264503967,
+      "loss": 4.3328,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 0.5956273078918457,
+      "learning_rate": 0.00026131377035940235,
+      "loss": 4.3246,
+      "step": 1350
+    },
+    {
+      "epoch": 1.72832,
+      "grad_norm": 0.6021464467048645,
+      "learning_rate": 0.000261273388073765,
+      "loss": 4.4009,
+      "step": 1351
+    },
+    {
+      "epoch": 1.7296,
+      "grad_norm": 0.6117057800292969,
+      "learning_rate": 0.00026123300578812756,
+      "loss": 4.3637,
+      "step": 1352
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 0.5697115659713745,
+      "learning_rate": 0.0002611926235024902,
+      "loss": 4.3032,
+      "step": 1353
+    },
+    {
+      "epoch": 1.73216,
+      "grad_norm": 0.6455910205841064,
+      "learning_rate": 0.00026115224121685287,
+      "loss": 4.2995,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7334399999999999,
+      "grad_norm": 0.5969467759132385,
+      "learning_rate": 0.0002611118589312155,
+      "loss": 4.3408,
+      "step": 1355
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 0.6078411340713501,
+      "learning_rate": 0.00026107147664557813,
+      "loss": 4.3799,
+      "step": 1356
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 0.5551348328590393,
+      "learning_rate": 0.00026103109435994076,
+      "loss": 4.3688,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7372800000000002,
+      "grad_norm": 0.6095216870307922,
+      "learning_rate": 0.0002609907120743034,
+      "loss": 4.3867,
+      "step": 1358
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 0.5706983208656311,
+      "learning_rate": 0.000260950329788666,
+      "loss": 4.3618,
+      "step": 1359
+    },
+    {
+      "epoch": 1.73984,
+      "grad_norm": 0.5468519926071167,
+      "learning_rate": 0.00026090994750302865,
+      "loss": 4.361,
+      "step": 1360
+    },
+    {
+      "epoch": 1.74112,
+      "grad_norm": 0.5678277611732483,
+      "learning_rate": 0.0002608695652173913,
+      "loss": 4.3545,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.5359609723091125,
+      "learning_rate": 0.0002608291829317539,
+      "loss": 4.4574,
+      "step": 1362
+    },
+    {
+      "epoch": 1.74368,
+      "grad_norm": 0.5697743892669678,
+      "learning_rate": 0.00026078880064611654,
+      "loss": 4.295,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7449599999999998,
+      "grad_norm": 0.605635404586792,
+      "learning_rate": 0.00026074841836047917,
+      "loss": 4.3309,
+      "step": 1364
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 0.5920169353485107,
+      "learning_rate": 0.0002607080360748418,
+      "loss": 4.3717,
+      "step": 1365
+    },
+    {
+      "epoch": 1.74752,
+      "grad_norm": 0.6027970910072327,
+      "learning_rate": 0.00026066765378920443,
+      "loss": 4.3445,
+      "step": 1366
+    },
+    {
+      "epoch": 1.7488000000000001,
+      "grad_norm": 0.5497733354568481,
+      "learning_rate": 0.0002606272715035671,
+      "loss": 4.3752,
+      "step": 1367
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 0.6217249631881714,
+      "learning_rate": 0.0002605868892179297,
+      "loss": 4.3425,
+      "step": 1368
+    },
+    {
+      "epoch": 1.75136,
+      "grad_norm": 0.5949388742446899,
+      "learning_rate": 0.0002605465069322923,
+      "loss": 4.3937,
+      "step": 1369
+    },
+    {
+      "epoch": 1.75264,
+      "grad_norm": 0.5935553908348083,
+      "learning_rate": 0.00026050612464665495,
+      "loss": 4.318,
+      "step": 1370
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 0.5247051119804382,
+      "learning_rate": 0.00026046574236101764,
+      "loss": 4.305,
+      "step": 1371
+    },
+    {
+      "epoch": 1.7551999999999999,
+      "grad_norm": 0.5983899235725403,
+      "learning_rate": 0.00026042536007538027,
+      "loss": 4.3383,
+      "step": 1372
+    },
+    {
+      "epoch": 1.75648,
+      "grad_norm": 0.5618011951446533,
+      "learning_rate": 0.0002603849777897429,
+      "loss": 4.3788,
+      "step": 1373
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 0.5628573298454285,
+      "learning_rate": 0.0002603445955041055,
+      "loss": 4.2755,
+      "step": 1374
+    },
+    {
+      "epoch": 1.7590400000000002,
+      "grad_norm": 0.580451488494873,
+      "learning_rate": 0.0002603042132184681,
+      "loss": 4.3533,
+      "step": 1375
+    },
+    {
+      "epoch": 1.76032,
+      "grad_norm": 0.5849772691726685,
+      "learning_rate": 0.0002602638309328308,
+      "loss": 4.3453,
+      "step": 1376
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.6029077172279358,
+      "learning_rate": 0.0002602234486471934,
+      "loss": 4.3619,
+      "step": 1377
+    },
+    {
+      "epoch": 1.76288,
+      "grad_norm": 0.5800113677978516,
+      "learning_rate": 0.00026018306636155605,
+      "loss": 4.3204,
+      "step": 1378
+    },
+    {
+      "epoch": 1.76416,
+      "grad_norm": 0.5955139994621277,
+      "learning_rate": 0.0002601426840759187,
+      "loss": 4.3915,
+      "step": 1379
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 0.5562100410461426,
+      "learning_rate": 0.0002601023017902813,
+      "loss": 4.2759,
+      "step": 1380
+    },
+    {
+      "epoch": 1.7667199999999998,
+      "grad_norm": 0.61802738904953,
+      "learning_rate": 0.00026006191950464394,
+      "loss": 4.2538,
+      "step": 1381
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.5850224494934082,
+      "learning_rate": 0.00026002153721900657,
+      "loss": 4.3395,
+      "step": 1382
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 0.5995482802391052,
+      "learning_rate": 0.0002599811549333692,
+      "loss": 4.2534,
+      "step": 1383
+    },
+    {
+      "epoch": 1.7705600000000001,
+      "grad_norm": 0.5256519913673401,
+      "learning_rate": 0.00025994077264773183,
+      "loss": 4.2427,
+      "step": 1384
+    },
+    {
+      "epoch": 1.77184,
+      "grad_norm": 0.5336461067199707,
+      "learning_rate": 0.00025990039036209446,
+      "loss": 4.3061,
+      "step": 1385
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 0.5198113322257996,
+      "learning_rate": 0.0002598600080764571,
+      "loss": 4.3705,
+      "step": 1386
+    },
+    {
+      "epoch": 1.7744,
+      "grad_norm": 0.5634206533432007,
+      "learning_rate": 0.0002598196257908197,
+      "loss": 4.3167,
+      "step": 1387
+    },
+    {
+      "epoch": 1.77568,
+      "grad_norm": 0.598160445690155,
+      "learning_rate": 0.00025977924350518235,
+      "loss": 4.2889,
+      "step": 1388
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 0.6371156573295593,
+      "learning_rate": 0.00025973886121954503,
+      "loss": 4.3066,
+      "step": 1389
+    },
+    {
+      "epoch": 1.77824,
+      "grad_norm": 0.5342087149620056,
+      "learning_rate": 0.00025969847893390766,
+      "loss": 4.2872,
+      "step": 1390
+    },
+    {
+      "epoch": 1.77952,
+      "grad_norm": 0.6202256083488464,
+      "learning_rate": 0.00025965809664827024,
+      "loss": 4.3316,
+      "step": 1391
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.5649963021278381,
+      "learning_rate": 0.00025961771436263287,
+      "loss": 4.3691,
+      "step": 1392
+    },
+    {
+      "epoch": 1.78208,
+      "grad_norm": 0.5771881341934204,
+      "learning_rate": 0.00025957733207699555,
+      "loss": 4.3228,
+      "step": 1393
+    },
+    {
+      "epoch": 1.78336,
+      "grad_norm": 0.5943997502326965,
+      "learning_rate": 0.0002595369497913582,
+      "loss": 4.4159,
+      "step": 1394
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 0.5805171132087708,
+      "learning_rate": 0.0002594965675057208,
+      "loss": 4.287,
+      "step": 1395
+    },
+    {
+      "epoch": 1.78592,
+      "grad_norm": 0.5803609490394592,
+      "learning_rate": 0.00025945618522008344,
+      "loss": 4.2801,
+      "step": 1396
+    },
+    {
+      "epoch": 1.7872,
+      "grad_norm": 0.5380381345748901,
+      "learning_rate": 0.00025941580293444607,
+      "loss": 4.3378,
+      "step": 1397
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 0.5976554155349731,
+      "learning_rate": 0.0002593754206488087,
+      "loss": 4.3923,
+      "step": 1398
+    },
+    {
+      "epoch": 1.78976,
+      "grad_norm": 0.6023809909820557,
+      "learning_rate": 0.00025933503836317133,
+      "loss": 4.3821,
+      "step": 1399
+    },
+    {
+      "epoch": 1.79104,
+      "grad_norm": 0.4953809976577759,
+      "learning_rate": 0.00025929465607753396,
+      "loss": 4.3558,
+      "step": 1400
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 0.6279889345169067,
+      "learning_rate": 0.0002592542737918966,
+      "loss": 4.319,
+      "step": 1401
+    },
+    {
+      "epoch": 1.7936,
+      "grad_norm": 0.5510340929031372,
+      "learning_rate": 0.0002592138915062592,
+      "loss": 4.3578,
+      "step": 1402
+    },
+    {
+      "epoch": 1.79488,
+      "grad_norm": 0.5811865925788879,
+      "learning_rate": 0.00025917350922062185,
+      "loss": 4.3256,
+      "step": 1403
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 0.5557198524475098,
+      "learning_rate": 0.0002591331269349845,
+      "loss": 4.2975,
+      "step": 1404
+    },
+    {
+      "epoch": 1.79744,
+      "grad_norm": 0.5800006985664368,
+      "learning_rate": 0.0002590927446493471,
+      "loss": 4.3556,
+      "step": 1405
+    },
+    {
+      "epoch": 1.7987199999999999,
+      "grad_norm": 0.5096418857574463,
+      "learning_rate": 0.0002590523623637098,
+      "loss": 4.2946,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5896133780479431,
+      "learning_rate": 0.00025901198007807243,
+      "loss": 4.3412,
+      "step": 1407
+    },
+    {
+      "epoch": 1.80128,
+      "grad_norm": 0.5542340874671936,
+      "learning_rate": 0.000258971597792435,
+      "loss": 4.3673,
+      "step": 1408
+    },
+    {
+      "epoch": 1.8025600000000002,
+      "grad_norm": 0.5914231538772583,
+      "learning_rate": 0.00025893121550679763,
+      "loss": 4.3599,
+      "step": 1409
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 0.6253953576087952,
+      "learning_rate": 0.0002588908332211603,
+      "loss": 4.2901,
+      "step": 1410
+    },
+    {
+      "epoch": 1.80512,
+      "grad_norm": 0.5754669904708862,
+      "learning_rate": 0.00025885045093552295,
+      "loss": 4.3587,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8064,
+      "grad_norm": 0.5589056015014648,
+      "learning_rate": 0.0002588100686498856,
+      "loss": 4.2848,
+      "step": 1412
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 0.6159608960151672,
+      "learning_rate": 0.0002587696863642482,
+      "loss": 4.2993,
+      "step": 1413
+    },
+    {
+      "epoch": 1.80896,
+      "grad_norm": 0.5682271718978882,
+      "learning_rate": 0.0002587293040786108,
+      "loss": 4.3867,
+      "step": 1414
+    },
+    {
+      "epoch": 1.8102399999999998,
+      "grad_norm": 0.5960492491722107,
+      "learning_rate": 0.00025868892179297347,
+      "loss": 4.2558,
+      "step": 1415
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 0.6439085006713867,
+      "learning_rate": 0.0002586485395073361,
+      "loss": 4.3216,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8128,
+      "grad_norm": 0.5476023554801941,
+      "learning_rate": 0.00025860815722169873,
+      "loss": 4.34,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8140800000000001,
+      "grad_norm": 0.6070132851600647,
+      "learning_rate": 0.00025856777493606136,
+      "loss": 4.3153,
+      "step": 1418
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 0.6208356618881226,
+      "learning_rate": 0.000258527392650424,
+      "loss": 4.4003,
+      "step": 1419
+    },
+    {
+      "epoch": 1.81664,
+      "grad_norm": 0.553810715675354,
+      "learning_rate": 0.0002584870103647866,
+      "loss": 4.321,
+      "step": 1420
+    },
+    {
+      "epoch": 1.81792,
+      "grad_norm": 0.6032571196556091,
+      "learning_rate": 0.00025844662807914925,
+      "loss": 4.3312,
+      "step": 1421
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.5861594676971436,
+      "learning_rate": 0.0002584062457935119,
+      "loss": 4.3138,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8204799999999999,
+      "grad_norm": 0.5738282203674316,
+      "learning_rate": 0.00025836586350787456,
+      "loss": 4.2894,
+      "step": 1423
+    },
+    {
+      "epoch": 1.82176,
+      "grad_norm": 0.5541486144065857,
+      "learning_rate": 0.00025832548122223714,
+      "loss": 4.3835,
+      "step": 1424
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 0.5869741439819336,
+      "learning_rate": 0.00025828509893659977,
+      "loss": 4.2246,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8243200000000002,
+      "grad_norm": 0.5501008033752441,
+      "learning_rate": 0.0002582447166509624,
+      "loss": 4.3119,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8256000000000001,
+      "grad_norm": 0.5797874927520752,
+      "learning_rate": 0.00025820433436532503,
+      "loss": 4.2885,
+      "step": 1427
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 0.5502588748931885,
+      "learning_rate": 0.0002581639520796877,
+      "loss": 4.2423,
+      "step": 1428
+    },
+    {
+      "epoch": 1.82816,
+      "grad_norm": 0.6173840761184692,
+      "learning_rate": 0.00025812356979405034,
+      "loss": 4.259,
+      "step": 1429
+    },
+    {
+      "epoch": 1.82944,
+      "grad_norm": 0.5949298143386841,
+      "learning_rate": 0.000258083187508413,
+      "loss": 4.3307,
+      "step": 1430
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 0.5942235589027405,
+      "learning_rate": 0.00025804280522277555,
+      "loss": 4.279,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 0.6103411912918091,
+      "learning_rate": 0.00025800242293713823,
+      "loss": 4.2624,
+      "step": 1432
+    },
+    {
+      "epoch": 1.83328,
+      "grad_norm": 0.5956135392189026,
+      "learning_rate": 0.00025796204065150086,
+      "loss": 4.3028,
+      "step": 1433
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 0.5850538015365601,
+      "learning_rate": 0.0002579216583658635,
+      "loss": 4.2829,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8358400000000001,
+      "grad_norm": 0.5809298157691956,
+      "learning_rate": 0.0002578812760802261,
+      "loss": 4.332,
+      "step": 1435
+    },
+    {
+      "epoch": 1.83712,
+      "grad_norm": 0.8085721135139465,
+      "learning_rate": 0.00025784089379458875,
+      "loss": 4.3069,
+      "step": 1436
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.5805991888046265,
+      "learning_rate": 0.0002578005115089514,
+      "loss": 4.3018,
+      "step": 1437
+    },
+    {
+      "epoch": 1.83968,
+      "grad_norm": 0.5863122344017029,
+      "learning_rate": 0.000257760129223314,
+      "loss": 4.275,
+      "step": 1438
+    },
+    {
+      "epoch": 1.84096,
+      "grad_norm": 0.573593258857727,
+      "learning_rate": 0.00025771974693767664,
+      "loss": 4.2953,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 0.6307305097579956,
+      "learning_rate": 0.0002576793646520393,
+      "loss": 4.3356,
+      "step": 1440
+    },
+    {
+      "epoch": 1.84352,
+      "grad_norm": 0.5647289752960205,
+      "learning_rate": 0.0002576389823664019,
+      "loss": 4.3151,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8448,
+      "grad_norm": 0.640838086605072,
+      "learning_rate": 0.00025759860008076453,
+      "loss": 4.3466,
+      "step": 1442
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 0.5609426498413086,
+      "learning_rate": 0.00025755821779512716,
+      "loss": 4.3062,
+      "step": 1443
+    },
+    {
+      "epoch": 1.8473600000000001,
+      "grad_norm": 0.6041327118873596,
+      "learning_rate": 0.0002575178355094898,
+      "loss": 4.3406,
+      "step": 1444
+    },
+    {
+      "epoch": 1.84864,
+      "grad_norm": 0.5379366874694824,
+      "learning_rate": 0.0002574774532238525,
+      "loss": 4.2807,
+      "step": 1445
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 0.582633376121521,
+      "learning_rate": 0.0002574370709382151,
+      "loss": 4.316,
+      "step": 1446
+    },
+    {
+      "epoch": 1.8512,
+      "grad_norm": 0.6131902933120728,
+      "learning_rate": 0.0002573966886525777,
+      "loss": 4.2406,
+      "step": 1447
+    },
+    {
+      "epoch": 1.85248,
+      "grad_norm": 0.6240634322166443,
+      "learning_rate": 0.0002573563063669403,
+      "loss": 4.307,
+      "step": 1448
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 0.5122919678688049,
+      "learning_rate": 0.000257315924081303,
+      "loss": 4.253,
+      "step": 1449
+    },
+    {
+      "epoch": 1.85504,
+      "grad_norm": 0.6418374180793762,
+      "learning_rate": 0.00025727554179566563,
+      "loss": 4.3222,
+      "step": 1450
+    },
+    {
+      "epoch": 1.85632,
+      "grad_norm": 0.5671072602272034,
+      "learning_rate": 0.00025723515951002826,
+      "loss": 4.3006,
+      "step": 1451
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.5645801424980164,
+      "learning_rate": 0.0002571947772243909,
+      "loss": 4.3623,
+      "step": 1452
+    },
+    {
+      "epoch": 1.85888,
+      "grad_norm": 0.5495478510856628,
+      "learning_rate": 0.0002571543949387535,
+      "loss": 4.2478,
+      "step": 1453
+    },
+    {
+      "epoch": 1.86016,
+      "grad_norm": 0.5628487467765808,
+      "learning_rate": 0.00025711401265311615,
+      "loss": 4.2659,
+      "step": 1454
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 0.5409569144248962,
+      "learning_rate": 0.0002570736303674788,
+      "loss": 4.2707,
+      "step": 1455
+    },
+    {
+      "epoch": 1.86272,
+      "grad_norm": 0.5503236651420593,
+      "learning_rate": 0.0002570332480818414,
+      "loss": 4.2997,
+      "step": 1456
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 0.5647298693656921,
+      "learning_rate": 0.00025699286579620404,
+      "loss": 4.3171,
+      "step": 1457
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 0.5661958456039429,
+      "learning_rate": 0.00025695248351056667,
+      "loss": 4.2884,
+      "step": 1458
+    },
+    {
+      "epoch": 1.86656,
+      "grad_norm": 0.5522183775901794,
+      "learning_rate": 0.0002569121012249293,
+      "loss": 4.1818,
+      "step": 1459
+    },
+    {
+      "epoch": 1.86784,
+      "grad_norm": 0.5519942045211792,
+      "learning_rate": 0.00025687171893929193,
+      "loss": 4.3356,
+      "step": 1460
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 0.5674711465835571,
+      "learning_rate": 0.00025683133665365456,
+      "loss": 4.3287,
+      "step": 1461
+    },
+    {
+      "epoch": 1.8704,
+      "grad_norm": 0.6130083799362183,
+      "learning_rate": 0.00025679095436801724,
+      "loss": 4.3046,
+      "step": 1462
+    },
+    {
+      "epoch": 1.87168,
+      "grad_norm": 0.5580954551696777,
+      "learning_rate": 0.0002567505720823798,
+      "loss": 4.2846,
+      "step": 1463
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 0.6269332766532898,
+      "learning_rate": 0.00025671018979674245,
+      "loss": 4.3118,
+      "step": 1464
+    },
+    {
+      "epoch": 1.87424,
+      "grad_norm": 0.5813364386558533,
+      "learning_rate": 0.0002566698075111051,
+      "loss": 4.2828,
+      "step": 1465
+    },
+    {
+      "epoch": 1.8755199999999999,
+      "grad_norm": 0.584498941898346,
+      "learning_rate": 0.00025662942522546776,
+      "loss": 4.2625,
+      "step": 1466
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.6057561635971069,
+      "learning_rate": 0.0002565890429398304,
+      "loss": 4.3507,
+      "step": 1467
+    },
+    {
+      "epoch": 1.87808,
+      "grad_norm": 0.5964322686195374,
+      "learning_rate": 0.000256548660654193,
+      "loss": 4.3682,
+      "step": 1468
+    },
+    {
+      "epoch": 1.8793600000000001,
+      "grad_norm": 0.6251275539398193,
+      "learning_rate": 0.00025650827836855565,
+      "loss": 4.3605,
+      "step": 1469
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 0.6410034894943237,
+      "learning_rate": 0.00025646789608291823,
+      "loss": 4.3105,
+      "step": 1470
+    },
+    {
+      "epoch": 1.88192,
+      "grad_norm": 0.6155179738998413,
+      "learning_rate": 0.0002564275137972809,
+      "loss": 4.2392,
+      "step": 1471
+    },
+    {
+      "epoch": 1.8832,
+      "grad_norm": 0.5981379747390747,
+      "learning_rate": 0.00025638713151164355,
+      "loss": 4.3841,
+      "step": 1472
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 0.637157142162323,
+      "learning_rate": 0.0002563467492260062,
+      "loss": 4.2601,
+      "step": 1473
+    },
+    {
+      "epoch": 1.8857599999999999,
+      "grad_norm": 0.5592889189720154,
+      "learning_rate": 0.0002563063669403688,
+      "loss": 4.3205,
+      "step": 1474
+    },
+    {
+      "epoch": 1.88704,
+      "grad_norm": 0.6076372265815735,
+      "learning_rate": 0.00025626598465473144,
+      "loss": 4.2291,
+      "step": 1475
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 0.5905406475067139,
+      "learning_rate": 0.00025622560236909407,
+      "loss": 4.2718,
+      "step": 1476
+    },
+    {
+      "epoch": 1.8896,
+      "grad_norm": 0.5950998663902283,
+      "learning_rate": 0.0002561852200834567,
+      "loss": 4.3082,
+      "step": 1477
+    },
+    {
+      "epoch": 1.8908800000000001,
+      "grad_norm": 0.6349307298660278,
+      "learning_rate": 0.0002561448377978193,
+      "loss": 4.3038,
+      "step": 1478
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 0.5779610276222229,
+      "learning_rate": 0.000256104455512182,
+      "loss": 4.2575,
+      "step": 1479
+    },
+    {
+      "epoch": 1.89344,
+      "grad_norm": 0.5922044515609741,
+      "learning_rate": 0.0002560640732265446,
+      "loss": 4.2998,
+      "step": 1480
+    },
+    {
+      "epoch": 1.89472,
+      "grad_norm": 0.6293798089027405,
+      "learning_rate": 0.0002560236909409072,
+      "loss": 4.2816,
+      "step": 1481
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.542755126953125,
+      "learning_rate": 0.00025598330865526985,
+      "loss": 4.3644,
+      "step": 1482
+    },
+    {
+      "epoch": 1.8972799999999999,
+      "grad_norm": 0.591120719909668,
+      "learning_rate": 0.0002559429263696325,
+      "loss": 4.277,
+      "step": 1483
+    },
+    {
+      "epoch": 1.89856,
+      "grad_norm": 0.5428628325462341,
+      "learning_rate": 0.00025590254408399516,
+      "loss": 4.1826,
+      "step": 1484
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 0.5935090780258179,
+      "learning_rate": 0.0002558621617983578,
+      "loss": 4.265,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9011200000000001,
+      "grad_norm": 0.532969057559967,
+      "learning_rate": 0.00025582177951272037,
+      "loss": 4.3217,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9024,
+      "grad_norm": 0.5747160315513611,
+      "learning_rate": 0.000255781397227083,
+      "loss": 4.2958,
+      "step": 1487
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 0.5506500005722046,
+      "learning_rate": 0.0002557410149414457,
+      "loss": 4.3303,
+      "step": 1488
+    },
+    {
+      "epoch": 1.90496,
+      "grad_norm": 0.5252817869186401,
+      "learning_rate": 0.0002557006326558083,
+      "loss": 4.2845,
+      "step": 1489
+    },
+    {
+      "epoch": 1.90624,
+      "grad_norm": 0.5319386124610901,
+      "learning_rate": 0.00025566025037017094,
+      "loss": 4.2375,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 0.5286985635757446,
+      "learning_rate": 0.00025561986808453357,
+      "loss": 4.3033,
+      "step": 1491
+    },
+    {
+      "epoch": 1.9088,
+      "grad_norm": 0.6041566133499146,
+      "learning_rate": 0.0002555794857988962,
+      "loss": 4.2996,
+      "step": 1492
+    },
+    {
+      "epoch": 1.91008,
+      "grad_norm": 0.563486635684967,
+      "learning_rate": 0.00025553910351325883,
+      "loss": 4.3267,
+      "step": 1493
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 0.5622044801712036,
+      "learning_rate": 0.00025549872122762146,
+      "loss": 4.2366,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9126400000000001,
+      "grad_norm": 0.5659383535385132,
+      "learning_rate": 0.0002554583389419841,
+      "loss": 4.3028,
+      "step": 1495
+    },
+    {
+      "epoch": 1.91392,
+      "grad_norm": 0.6131373047828674,
+      "learning_rate": 0.0002554179566563467,
+      "loss": 4.2772,
+      "step": 1496
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.5460529923439026,
+      "learning_rate": 0.00025537757437070935,
+      "loss": 4.2903,
+      "step": 1497
+    },
+    {
+      "epoch": 1.91648,
+      "grad_norm": 0.6126199960708618,
+      "learning_rate": 0.000255337192085072,
+      "loss": 4.3442,
+      "step": 1498
+    },
+    {
+      "epoch": 1.91776,
+      "grad_norm": 0.5673563480377197,
+      "learning_rate": 0.0002552968097994346,
+      "loss": 4.2208,
+      "step": 1499
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 0.5467031002044678,
+      "learning_rate": 0.00025525642751379724,
+      "loss": 4.287,
+      "step": 1500
+    },
+    {
+      "epoch": 1.92032,
+      "grad_norm": 0.5507038235664368,
+      "learning_rate": 0.0002552160452281599,
+      "loss": 4.2693,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9216,
+      "grad_norm": 0.5434231758117676,
+      "learning_rate": 0.00025517566294252256,
+      "loss": 4.2913,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 0.5921497941017151,
+      "learning_rate": 0.00025513528065688513,
+      "loss": 4.3204,
+      "step": 1503
+    },
+    {
+      "epoch": 1.92416,
+      "grad_norm": 0.5684370398521423,
+      "learning_rate": 0.00025509489837124776,
+      "loss": 4.2581,
+      "step": 1504
+    },
+    {
+      "epoch": 1.92544,
+      "grad_norm": 0.5858400464057922,
+      "learning_rate": 0.00025505451608561045,
+      "loss": 4.2407,
+      "step": 1505
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 0.5300204157829285,
+      "learning_rate": 0.0002550141337999731,
+      "loss": 4.3043,
+      "step": 1506
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.5366335511207581,
+      "learning_rate": 0.0002549737515143357,
+      "loss": 4.2903,
+      "step": 1507
+    },
+    {
+      "epoch": 1.9292799999999999,
+      "grad_norm": 0.5699880719184875,
+      "learning_rate": 0.00025493336922869834,
+      "loss": 4.2495,
+      "step": 1508
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 0.5679522752761841,
+      "learning_rate": 0.00025489298694306097,
+      "loss": 4.2577,
+      "step": 1509
+    },
+    {
+      "epoch": 1.93184,
+      "grad_norm": 0.5857032537460327,
+      "learning_rate": 0.0002548526046574236,
+      "loss": 4.273,
+      "step": 1510
+    },
+    {
+      "epoch": 1.93312,
+      "grad_norm": 0.574712872505188,
+      "learning_rate": 0.0002548122223717862,
+      "loss": 4.3117,
+      "step": 1511
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.5547901391983032,
+      "learning_rate": 0.00025477184008614886,
+      "loss": 4.3168,
+      "step": 1512
+    },
+    {
+      "epoch": 1.93568,
+      "grad_norm": 0.557061493396759,
+      "learning_rate": 0.0002547314578005115,
+      "loss": 4.2803,
+      "step": 1513
+    },
+    {
+      "epoch": 1.93696,
+      "grad_norm": 0.5565205216407776,
+      "learning_rate": 0.0002546910755148741,
+      "loss": 4.2377,
+      "step": 1514
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 0.5655121803283691,
+      "learning_rate": 0.00025465069322923675,
+      "loss": 4.2526,
+      "step": 1515
+    },
+    {
+      "epoch": 1.93952,
+      "grad_norm": 0.6136040091514587,
+      "learning_rate": 0.0002546103109435994,
+      "loss": 4.3138,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9407999999999999,
+      "grad_norm": 0.5623380541801453,
+      "learning_rate": 0.000254569928657962,
+      "loss": 4.1775,
+      "step": 1517
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 0.5378047823905945,
+      "learning_rate": 0.0002545295463723247,
+      "loss": 4.2809,
+      "step": 1518
+    },
+    {
+      "epoch": 1.94336,
+      "grad_norm": 0.6011268496513367,
+      "learning_rate": 0.00025448916408668727,
+      "loss": 4.2017,
+      "step": 1519
+    },
+    {
+      "epoch": 1.9446400000000001,
+      "grad_norm": 0.5707088708877563,
+      "learning_rate": 0.0002544487818010499,
+      "loss": 4.22,
+      "step": 1520
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 0.5897967219352722,
+      "learning_rate": 0.00025440839951541253,
+      "loss": 4.2578,
+      "step": 1521
+    },
+    {
+      "epoch": 1.9472,
+      "grad_norm": 0.5613348484039307,
+      "learning_rate": 0.00025436801722977516,
+      "loss": 4.2774,
+      "step": 1522
+    },
+    {
+      "epoch": 1.94848,
+      "grad_norm": 0.5693078637123108,
+      "learning_rate": 0.00025432763494413784,
+      "loss": 4.2862,
+      "step": 1523
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 0.6159646511077881,
+      "learning_rate": 0.00025428725265850047,
+      "loss": 4.3357,
+      "step": 1524
+    },
+    {
+      "epoch": 1.9510399999999999,
+      "grad_norm": 0.5490036010742188,
+      "learning_rate": 0.0002542468703728631,
+      "loss": 4.2645,
+      "step": 1525
+    },
+    {
+      "epoch": 1.95232,
+      "grad_norm": 0.6109630465507507,
+      "learning_rate": 0.0002542064880872257,
+      "loss": 4.2504,
+      "step": 1526
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.5659199953079224,
+      "learning_rate": 0.00025416610580158836,
+      "loss": 4.2713,
+      "step": 1527
+    },
+    {
+      "epoch": 1.95488,
+      "grad_norm": 0.5810695886611938,
+      "learning_rate": 0.000254125723515951,
+      "loss": 4.2639,
+      "step": 1528
+    },
+    {
+      "epoch": 1.9561600000000001,
+      "grad_norm": 0.5711521506309509,
+      "learning_rate": 0.0002540853412303136,
+      "loss": 4.2478,
+      "step": 1529
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 0.571063756942749,
+      "learning_rate": 0.00025404495894467625,
+      "loss": 4.2389,
+      "step": 1530
+    },
+    {
+      "epoch": 1.95872,
+      "grad_norm": 0.5734847784042358,
+      "learning_rate": 0.0002540045766590389,
+      "loss": 4.216,
+      "step": 1531
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5617849230766296,
+      "learning_rate": 0.0002539641943734015,
+      "loss": 4.2903,
+      "step": 1532
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 0.5797948241233826,
+      "learning_rate": 0.00025392381208776414,
+      "loss": 4.2572,
+      "step": 1533
+    },
+    {
+      "epoch": 1.9625599999999999,
+      "grad_norm": 0.5373856425285339,
+      "learning_rate": 0.00025388342980212677,
+      "loss": 4.2827,
+      "step": 1534
+    },
+    {
+      "epoch": 1.96384,
+      "grad_norm": 0.5615948438644409,
+      "learning_rate": 0.0002538430475164894,
+      "loss": 4.2297,
+      "step": 1535
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 0.5433390140533447,
+      "learning_rate": 0.00025380266523085203,
+      "loss": 4.3284,
+      "step": 1536
+    },
+    {
+      "epoch": 1.9664000000000001,
+      "grad_norm": 0.604446530342102,
+      "learning_rate": 0.00025376228294521466,
+      "loss": 4.2762,
+      "step": 1537
+    },
+    {
+      "epoch": 1.96768,
+      "grad_norm": 0.5676113367080688,
+      "learning_rate": 0.0002537219006595773,
+      "loss": 4.174,
+      "step": 1538
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 0.5773786902427673,
+      "learning_rate": 0.0002536815183739399,
+      "loss": 4.227,
+      "step": 1539
+    },
+    {
+      "epoch": 1.97024,
+      "grad_norm": 0.5377183556556702,
+      "learning_rate": 0.0002536411360883026,
+      "loss": 4.232,
+      "step": 1540
+    },
+    {
+      "epoch": 1.97152,
+      "grad_norm": 0.5951273441314697,
+      "learning_rate": 0.00025360075380266524,
+      "loss": 4.2514,
+      "step": 1541
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.5194257497787476,
+      "learning_rate": 0.0002535603715170278,
+      "loss": 4.2341,
+      "step": 1542
+    },
+    {
+      "epoch": 1.9740799999999998,
+      "grad_norm": 0.5489000678062439,
+      "learning_rate": 0.00025351998923139044,
+      "loss": 4.2023,
+      "step": 1543
+    },
+    {
+      "epoch": 1.97536,
+      "grad_norm": 0.5540454983711243,
+      "learning_rate": 0.00025347960694575313,
+      "loss": 4.2174,
+      "step": 1544
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 0.5399606823921204,
+      "learning_rate": 0.00025343922466011576,
+      "loss": 4.335,
+      "step": 1545
+    },
+    {
+      "epoch": 1.9779200000000001,
+      "grad_norm": 0.5686156749725342,
+      "learning_rate": 0.0002533988423744784,
+      "loss": 4.2687,
+      "step": 1546
+    },
+    {
+      "epoch": 1.9792,
+      "grad_norm": 0.528323769569397,
+      "learning_rate": 0.000253358460088841,
+      "loss": 4.3328,
+      "step": 1547
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 0.5400658845901489,
+      "learning_rate": 0.00025331807780320365,
+      "loss": 4.2065,
+      "step": 1548
+    },
+    {
+      "epoch": 1.98176,
+      "grad_norm": 0.5493596792221069,
+      "learning_rate": 0.0002532776955175663,
+      "loss": 4.2173,
+      "step": 1549
+    },
+    {
+      "epoch": 1.98304,
+      "grad_norm": 0.5500946044921875,
+      "learning_rate": 0.0002532373132319289,
+      "loss": 4.2876,
+      "step": 1550
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 0.5746954679489136,
+      "learning_rate": 0.00025319693094629154,
+      "loss": 4.3206,
+      "step": 1551
+    },
+    {
+      "epoch": 1.9856,
+      "grad_norm": 0.5554184913635254,
+      "learning_rate": 0.00025315654866065417,
+      "loss": 4.2171,
+      "step": 1552
+    },
+    {
+      "epoch": 1.98688,
+      "grad_norm": 0.5702386498451233,
+      "learning_rate": 0.0002531161663750168,
+      "loss": 4.2557,
+      "step": 1553
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 0.5968025922775269,
+      "learning_rate": 0.00025307578408937943,
+      "loss": 4.2913,
+      "step": 1554
+    },
+    {
+      "epoch": 1.98944,
+      "grad_norm": 0.5453073382377625,
+      "learning_rate": 0.00025303540180374206,
+      "loss": 4.2114,
+      "step": 1555
+    },
+    {
+      "epoch": 1.99072,
+      "grad_norm": 0.5790675282478333,
+      "learning_rate": 0.0002529950195181047,
+      "loss": 4.2186,
+      "step": 1556
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.5652514100074768,
+      "learning_rate": 0.00025295463723246737,
+      "loss": 4.2595,
+      "step": 1557
+    },
+    {
+      "epoch": 1.99328,
+      "grad_norm": 0.5572407245635986,
+      "learning_rate": 0.00025291425494682995,
+      "loss": 4.2544,
+      "step": 1558
+    },
+    {
+      "epoch": 1.9945599999999999,
+      "grad_norm": 0.5312286019325256,
+      "learning_rate": 0.0002528738726611926,
+      "loss": 4.2369,
+      "step": 1559
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 0.5568103194236755,
+      "learning_rate": 0.0002528334903755552,
+      "loss": 4.3173,
+      "step": 1560
+    },
+    {
+      "epoch": 1.99712,
+      "grad_norm": 0.5829612612724304,
+      "learning_rate": 0.00025279310808991784,
+      "loss": 4.2247,
+      "step": 1561
+    },
+    {
+      "epoch": 1.9984,
+      "grad_norm": 0.5639691352844238,
+      "learning_rate": 0.0002527527258042805,
+      "loss": 4.2355,
+      "step": 1562
+    },
+    {
+      "epoch": 1.9996800000000001,
+      "grad_norm": 0.5624785423278809,
+      "learning_rate": 0.00025271234351864315,
+      "loss": 4.2893,
+      "step": 1563
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9181849360466003,
+      "learning_rate": 0.0002526719612330058,
+      "loss": 4.2339,
+      "step": 1564
+    },
+    {
+      "epoch": 2.00128,
+      "grad_norm": 0.6469334959983826,
+      "learning_rate": 0.00025263157894736836,
+      "loss": 4.1552,
+      "step": 1565
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 0.607097327709198,
+      "learning_rate": 0.00025259119666173104,
+      "loss": 4.1764,
+      "step": 1566
+    },
+    {
+      "epoch": 2.00384,
+      "grad_norm": 0.5565678477287292,
+      "learning_rate": 0.0002525508143760937,
+      "loss": 4.1168,
+      "step": 1567
+    },
+    {
+      "epoch": 2.00512,
+      "grad_norm": 0.5955238342285156,
+      "learning_rate": 0.0002525104320904563,
+      "loss": 4.1852,
+      "step": 1568
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.5889109969139099,
+      "learning_rate": 0.00025247004980481893,
+      "loss": 4.1784,
+      "step": 1569
+    },
+    {
+      "epoch": 2.00768,
+      "grad_norm": 0.6503939032554626,
+      "learning_rate": 0.00025242966751918156,
+      "loss": 4.1564,
+      "step": 1570
+    },
+    {
+      "epoch": 2.00896,
+      "grad_norm": 0.5937455296516418,
+      "learning_rate": 0.0002523892852335442,
+      "loss": 4.2143,
+      "step": 1571
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 0.6147340536117554,
+      "learning_rate": 0.0002523489029479068,
+      "loss": 4.1123,
+      "step": 1572
+    },
+    {
+      "epoch": 2.01152,
+      "grad_norm": 0.5893435478210449,
+      "learning_rate": 0.00025230852066226945,
+      "loss": 4.0739,
+      "step": 1573
+    },
+    {
+      "epoch": 2.0128,
+      "grad_norm": 0.584883451461792,
+      "learning_rate": 0.00025226813837663214,
+      "loss": 4.2348,
+      "step": 1574
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 0.5527332425117493,
+      "learning_rate": 0.0002522277560909947,
+      "loss": 4.205,
+      "step": 1575
+    },
+    {
+      "epoch": 2.01536,
+      "grad_norm": 0.5769275426864624,
+      "learning_rate": 0.00025218737380535734,
+      "loss": 4.1406,
+      "step": 1576
+    },
+    {
+      "epoch": 2.01664,
+      "grad_norm": 0.5587900280952454,
+      "learning_rate": 0.00025214699151972,
+      "loss": 4.1601,
+      "step": 1577
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 0.5883750319480896,
+      "learning_rate": 0.0002521066092340826,
+      "loss": 4.1505,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0192,
+      "grad_norm": 0.5383011102676392,
+      "learning_rate": 0.0002520662269484453,
+      "loss": 4.1801,
+      "step": 1579
+    },
+    {
+      "epoch": 2.02048,
+      "grad_norm": 0.5970503687858582,
+      "learning_rate": 0.0002520258446628079,
+      "loss": 4.1597,
+      "step": 1580
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 0.5630191564559937,
+      "learning_rate": 0.00025198546237717055,
+      "loss": 4.1925,
+      "step": 1581
+    },
+    {
+      "epoch": 2.02304,
+      "grad_norm": 0.5271741151809692,
+      "learning_rate": 0.0002519450800915331,
+      "loss": 4.2025,
+      "step": 1582
+    },
+    {
+      "epoch": 2.02432,
+      "grad_norm": 0.5406798124313354,
+      "learning_rate": 0.0002519046978058958,
+      "loss": 4.1666,
+      "step": 1583
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.5316964387893677,
+      "learning_rate": 0.00025186431552025844,
+      "loss": 4.1439,
+      "step": 1584
+    },
+    {
+      "epoch": 2.02688,
+      "grad_norm": 0.5242227911949158,
+      "learning_rate": 0.00025182393323462107,
+      "loss": 4.1849,
+      "step": 1585
+    },
+    {
+      "epoch": 2.02816,
+      "grad_norm": 0.516384482383728,
+      "learning_rate": 0.0002517835509489837,
+      "loss": 4.1493,
+      "step": 1586
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 0.5419257879257202,
+      "learning_rate": 0.00025174316866334633,
+      "loss": 4.1466,
+      "step": 1587
+    },
+    {
+      "epoch": 2.03072,
+      "grad_norm": 0.5190739631652832,
+      "learning_rate": 0.00025170278637770896,
+      "loss": 4.2076,
+      "step": 1588
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 0.5474783778190613,
+      "learning_rate": 0.0002516624040920716,
+      "loss": 4.1463,
+      "step": 1589
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 0.5086979269981384,
+      "learning_rate": 0.0002516220218064342,
+      "loss": 4.1831,
+      "step": 1590
+    },
+    {
+      "epoch": 2.03456,
+      "grad_norm": 0.5824074149131775,
+      "learning_rate": 0.00025158163952079685,
+      "loss": 4.1523,
+      "step": 1591
+    },
+    {
+      "epoch": 2.03584,
+      "grad_norm": 0.523669958114624,
+      "learning_rate": 0.0002515412572351595,
+      "loss": 4.1735,
+      "step": 1592
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 0.5100351572036743,
+      "learning_rate": 0.0002515008749495221,
+      "loss": 4.1265,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0384,
+      "grad_norm": 0.5400765538215637,
+      "learning_rate": 0.00025146049266388474,
+      "loss": 4.1038,
+      "step": 1594
+    },
+    {
+      "epoch": 2.03968,
+      "grad_norm": 0.5364828705787659,
+      "learning_rate": 0.00025142011037824737,
+      "loss": 4.1539,
+      "step": 1595
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 0.5193732976913452,
+      "learning_rate": 0.00025137972809261005,
+      "loss": 4.1766,
+      "step": 1596
+    },
+    {
+      "epoch": 2.04224,
+      "grad_norm": 0.5233401656150818,
+      "learning_rate": 0.0002513393458069727,
+      "loss": 4.1535,
+      "step": 1597
+    },
+    {
+      "epoch": 2.04352,
+      "grad_norm": 0.5625421404838562,
+      "learning_rate": 0.00025129896352133526,
+      "loss": 4.1345,
+      "step": 1598
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.5566902756690979,
+      "learning_rate": 0.0002512585812356979,
+      "loss": 4.1438,
+      "step": 1599
+    },
+    {
+      "epoch": 2.04608,
+      "grad_norm": 0.5623239874839783,
+      "learning_rate": 0.0002512181989500606,
+      "loss": 4.1251,
+      "step": 1600
+    },
+    {
+      "epoch": 2.04736,
+      "grad_norm": 0.578046441078186,
+      "learning_rate": 0.0002511778166644232,
+      "loss": 4.0756,
+      "step": 1601
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 0.5461819171905518,
+      "learning_rate": 0.00025113743437878583,
+      "loss": 4.1906,
+      "step": 1602
+    },
+    {
+      "epoch": 2.04992,
+      "grad_norm": 0.5973131656646729,
+      "learning_rate": 0.00025109705209314846,
+      "loss": 4.1781,
+      "step": 1603
+    },
+    {
+      "epoch": 2.0512,
+      "grad_norm": 0.5380710959434509,
+      "learning_rate": 0.0002510566698075111,
+      "loss": 4.1321,
+      "step": 1604
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 0.5838267207145691,
+      "learning_rate": 0.0002510162875218737,
+      "loss": 4.1575,
+      "step": 1605
+    },
+    {
+      "epoch": 2.05376,
+      "grad_norm": 0.5240848660469055,
+      "learning_rate": 0.00025097590523623635,
+      "loss": 4.0436,
+      "step": 1606
+    },
+    {
+      "epoch": 2.05504,
+      "grad_norm": 0.5829635858535767,
+      "learning_rate": 0.000250935522950599,
+      "loss": 4.1312,
+      "step": 1607
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 0.5419280529022217,
+      "learning_rate": 0.0002508951406649616,
+      "loss": 4.1518,
+      "step": 1608
+    },
+    {
+      "epoch": 2.0576,
+      "grad_norm": 0.5682242512702942,
+      "learning_rate": 0.00025085475837932424,
+      "loss": 4.2161,
+      "step": 1609
+    },
+    {
+      "epoch": 2.05888,
+      "grad_norm": 0.5263741612434387,
+      "learning_rate": 0.0002508143760936869,
+      "loss": 4.1819,
+      "step": 1610
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 0.5568137168884277,
+      "learning_rate": 0.0002507739938080495,
+      "loss": 4.0907,
+      "step": 1611
+    },
+    {
+      "epoch": 2.06144,
+      "grad_norm": 0.5448589324951172,
+      "learning_rate": 0.00025073361152241214,
+      "loss": 4.171,
+      "step": 1612
+    },
+    {
+      "epoch": 2.06272,
+      "grad_norm": 0.5299394726753235,
+      "learning_rate": 0.0002506932292367748,
+      "loss": 4.1997,
+      "step": 1613
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.5272650122642517,
+      "learning_rate": 0.0002506528469511374,
+      "loss": 4.0966,
+      "step": 1614
+    },
+    {
+      "epoch": 2.06528,
+      "grad_norm": 0.5520989894866943,
+      "learning_rate": 0.0002506124646655,
+      "loss": 4.1154,
+      "step": 1615
+    },
+    {
+      "epoch": 2.06656,
+      "grad_norm": 0.5448061227798462,
+      "learning_rate": 0.00025057208237986266,
+      "loss": 4.096,
+      "step": 1616
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 0.5410912036895752,
+      "learning_rate": 0.0002505317000942253,
+      "loss": 4.1071,
+      "step": 1617
+    },
+    {
+      "epoch": 2.06912,
+      "grad_norm": 0.5440801978111267,
+      "learning_rate": 0.00025049131780858797,
+      "loss": 4.2481,
+      "step": 1618
+    },
+    {
+      "epoch": 2.0704,
+      "grad_norm": 0.5587443709373474,
+      "learning_rate": 0.0002504509355229506,
+      "loss": 4.1345,
+      "step": 1619
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 0.5543466210365295,
+      "learning_rate": 0.00025041055323731323,
+      "loss": 4.2023,
+      "step": 1620
+    },
+    {
+      "epoch": 2.07296,
+      "grad_norm": 0.5711597800254822,
+      "learning_rate": 0.0002503701709516758,
+      "loss": 4.1222,
+      "step": 1621
+    },
+    {
+      "epoch": 2.07424,
+      "grad_norm": 0.5779085159301758,
+      "learning_rate": 0.0002503297886660385,
+      "loss": 4.1628,
+      "step": 1622
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 0.5895453691482544,
+      "learning_rate": 0.0002502894063804011,
+      "loss": 4.1175,
+      "step": 1623
+    },
+    {
+      "epoch": 2.0768,
+      "grad_norm": 0.5698716640472412,
+      "learning_rate": 0.00025024902409476375,
+      "loss": 4.0683,
+      "step": 1624
+    },
+    {
+      "epoch": 2.07808,
+      "grad_norm": 0.6177151799201965,
+      "learning_rate": 0.0002502086418091264,
+      "loss": 4.2269,
+      "step": 1625
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 0.5702541470527649,
+      "learning_rate": 0.000250168259523489,
+      "loss": 4.1249,
+      "step": 1626
+    },
+    {
+      "epoch": 2.08064,
+      "grad_norm": 0.5631515383720398,
+      "learning_rate": 0.00025012787723785164,
+      "loss": 4.0755,
+      "step": 1627
+    },
+    {
+      "epoch": 2.08192,
+      "grad_norm": 0.5638665556907654,
+      "learning_rate": 0.00025008749495221427,
+      "loss": 4.1912,
+      "step": 1628
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.5572354793548584,
+      "learning_rate": 0.0002500471126665769,
+      "loss": 4.1421,
+      "step": 1629
+    },
+    {
+      "epoch": 2.08448,
+      "grad_norm": 0.5218389630317688,
+      "learning_rate": 0.00025000673038093953,
+      "loss": 4.1606,
+      "step": 1630
+    },
+    {
+      "epoch": 2.08576,
+      "grad_norm": 0.543541669845581,
+      "learning_rate": 0.00024996634809530216,
+      "loss": 4.2001,
+      "step": 1631
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 0.5522968173027039,
+      "learning_rate": 0.0002499259658096648,
+      "loss": 4.1279,
+      "step": 1632
+    },
+    {
+      "epoch": 2.08832,
+      "grad_norm": 0.5614569783210754,
+      "learning_rate": 0.0002498855835240274,
+      "loss": 4.057,
+      "step": 1633
+    },
+    {
+      "epoch": 2.0896,
+      "grad_norm": 0.5415524244308472,
+      "learning_rate": 0.00024984520123839005,
+      "loss": 4.2062,
+      "step": 1634
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 0.5276180505752563,
+      "learning_rate": 0.00024980481895275274,
+      "loss": 4.1219,
+      "step": 1635
+    },
+    {
+      "epoch": 2.09216,
+      "grad_norm": 0.5661503076553345,
+      "learning_rate": 0.00024976443666711537,
+      "loss": 4.0845,
+      "step": 1636
+    },
+    {
+      "epoch": 2.09344,
+      "grad_norm": 0.5094005465507507,
+      "learning_rate": 0.00024972405438147794,
+      "loss": 4.1161,
+      "step": 1637
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 0.5565866231918335,
+      "learning_rate": 0.00024968367209584057,
+      "loss": 4.1302,
+      "step": 1638
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 0.5100904107093811,
+      "learning_rate": 0.00024964328981020326,
+      "loss": 4.189,
+      "step": 1639
+    },
+    {
+      "epoch": 2.09728,
+      "grad_norm": 0.5307742953300476,
+      "learning_rate": 0.0002496029075245659,
+      "loss": 4.0909,
+      "step": 1640
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 0.5522825717926025,
+      "learning_rate": 0.0002495625252389285,
+      "loss": 4.086,
+      "step": 1641
+    },
+    {
+      "epoch": 2.09984,
+      "grad_norm": 0.5281661152839661,
+      "learning_rate": 0.00024952214295329115,
+      "loss": 4.0938,
+      "step": 1642
+    },
+    {
+      "epoch": 2.10112,
+      "grad_norm": 0.5300430655479431,
+      "learning_rate": 0.0002494817606676538,
+      "loss": 4.2113,
+      "step": 1643
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.5836507081985474,
+      "learning_rate": 0.0002494413783820164,
+      "loss": 4.1544,
+      "step": 1644
+    },
+    {
+      "epoch": 2.1036799999999998,
+      "grad_norm": 0.5055437088012695,
+      "learning_rate": 0.00024940099609637904,
+      "loss": 4.1888,
+      "step": 1645
+    },
+    {
+      "epoch": 2.10496,
+      "grad_norm": 0.5788672566413879,
+      "learning_rate": 0.00024936061381074167,
+      "loss": 4.1254,
+      "step": 1646
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 0.5161530375480652,
+      "learning_rate": 0.0002493202315251043,
+      "loss": 4.1339,
+      "step": 1647
+    },
+    {
+      "epoch": 2.10752,
+      "grad_norm": 0.526464581489563,
+      "learning_rate": 0.0002492798492394669,
+      "loss": 4.0877,
+      "step": 1648
+    },
+    {
+      "epoch": 2.1088,
+      "grad_norm": 0.5108904242515564,
+      "learning_rate": 0.00024923946695382956,
+      "loss": 4.1818,
+      "step": 1649
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 0.5350533723831177,
+      "learning_rate": 0.0002491990846681922,
+      "loss": 4.1912,
+      "step": 1650
+    },
+    {
+      "epoch": 2.11136,
+      "grad_norm": 0.552305281162262,
+      "learning_rate": 0.0002491587023825548,
+      "loss": 4.0919,
+      "step": 1651
+    },
+    {
+      "epoch": 2.11264,
+      "grad_norm": 0.5402622818946838,
+      "learning_rate": 0.0002491183200969175,
+      "loss": 4.1347,
+      "step": 1652
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 0.5491945147514343,
+      "learning_rate": 0.00024907793781128013,
+      "loss": 4.1458,
+      "step": 1653
+    },
+    {
+      "epoch": 2.1152,
+      "grad_norm": 0.5369943380355835,
+      "learning_rate": 0.0002490375555256427,
+      "loss": 4.1845,
+      "step": 1654
+    },
+    {
+      "epoch": 2.11648,
+      "grad_norm": 0.5386082530021667,
+      "learning_rate": 0.00024899717324000534,
+      "loss": 4.1702,
+      "step": 1655
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 0.5493932366371155,
+      "learning_rate": 0.00024895679095436797,
+      "loss": 4.1743,
+      "step": 1656
+    },
+    {
+      "epoch": 2.11904,
+      "grad_norm": 0.5019392967224121,
+      "learning_rate": 0.00024891640866873065,
+      "loss": 4.0747,
+      "step": 1657
+    },
+    {
+      "epoch": 2.12032,
+      "grad_norm": 0.556978166103363,
+      "learning_rate": 0.0002488760263830933,
+      "loss": 4.1214,
+      "step": 1658
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.5479657053947449,
+      "learning_rate": 0.0002488356440974559,
+      "loss": 4.1133,
+      "step": 1659
+    },
+    {
+      "epoch": 2.12288,
+      "grad_norm": 0.5306862592697144,
+      "learning_rate": 0.0002487952618118185,
+      "loss": 4.1571,
+      "step": 1660
+    },
+    {
+      "epoch": 2.12416,
+      "grad_norm": 0.5123226046562195,
+      "learning_rate": 0.00024875487952618117,
+      "loss": 4.1351,
+      "step": 1661
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 0.5251739621162415,
+      "learning_rate": 0.0002487144972405438,
+      "loss": 4.1093,
+      "step": 1662
+    },
+    {
+      "epoch": 2.12672,
+      "grad_norm": 0.5215139389038086,
+      "learning_rate": 0.00024867411495490643,
+      "loss": 4.1659,
+      "step": 1663
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 0.579883873462677,
+      "learning_rate": 0.00024863373266926906,
+      "loss": 4.153,
+      "step": 1664
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 0.5173061490058899,
+      "learning_rate": 0.0002485933503836317,
+      "loss": 4.1132,
+      "step": 1665
+    },
+    {
+      "epoch": 2.13056,
+      "grad_norm": 0.5922846794128418,
+      "learning_rate": 0.0002485529680979943,
+      "loss": 4.149,
+      "step": 1666
+    },
+    {
+      "epoch": 2.13184,
+      "grad_norm": 0.5453259348869324,
+      "learning_rate": 0.00024851258581235695,
+      "loss": 4.1019,
+      "step": 1667
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 0.5758420825004578,
+      "learning_rate": 0.0002484722035267196,
+      "loss": 4.1267,
+      "step": 1668
+    },
+    {
+      "epoch": 2.1344,
+      "grad_norm": 0.5560485124588013,
+      "learning_rate": 0.0002484318212410822,
+      "loss": 4.1299,
+      "step": 1669
+    },
+    {
+      "epoch": 2.13568,
+      "grad_norm": 0.5476799011230469,
+      "learning_rate": 0.00024839143895544484,
+      "loss": 4.0798,
+      "step": 1670
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 0.5895335078239441,
+      "learning_rate": 0.00024835105666980747,
+      "loss": 4.1217,
+      "step": 1671
+    },
+    {
+      "epoch": 2.13824,
+      "grad_norm": 0.5773458480834961,
+      "learning_rate": 0.0002483106743841701,
+      "loss": 4.1645,
+      "step": 1672
+    },
+    {
+      "epoch": 2.13952,
+      "grad_norm": 0.5620774626731873,
+      "learning_rate": 0.00024827029209853273,
+      "loss": 4.1691,
+      "step": 1673
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.5326398015022278,
+      "learning_rate": 0.0002482299098128954,
+      "loss": 4.1493,
+      "step": 1674
+    },
+    {
+      "epoch": 2.14208,
+      "grad_norm": 0.5294104218482971,
+      "learning_rate": 0.00024818952752725805,
+      "loss": 4.0135,
+      "step": 1675
+    },
+    {
+      "epoch": 2.14336,
+      "grad_norm": 0.6189349293708801,
+      "learning_rate": 0.0002481491452416207,
+      "loss": 4.1086,
+      "step": 1676
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 0.5574718713760376,
+      "learning_rate": 0.00024810876295598325,
+      "loss": 4.0684,
+      "step": 1677
+    },
+    {
+      "epoch": 2.14592,
+      "grad_norm": 0.5645029544830322,
+      "learning_rate": 0.00024806838067034594,
+      "loss": 4.0929,
+      "step": 1678
+    },
+    {
+      "epoch": 2.1471999999999998,
+      "grad_norm": 0.5400739312171936,
+      "learning_rate": 0.00024802799838470857,
+      "loss": 4.0444,
+      "step": 1679
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 0.5390148162841797,
+      "learning_rate": 0.0002479876160990712,
+      "loss": 4.1553,
+      "step": 1680
+    },
+    {
+      "epoch": 2.14976,
+      "grad_norm": 0.520582377910614,
+      "learning_rate": 0.00024794723381343383,
+      "loss": 4.0592,
+      "step": 1681
+    },
+    {
+      "epoch": 2.15104,
+      "grad_norm": 0.5372869968414307,
+      "learning_rate": 0.00024790685152779646,
+      "loss": 4.1065,
+      "step": 1682
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 0.510082483291626,
+      "learning_rate": 0.0002478664692421591,
+      "loss": 4.1625,
+      "step": 1683
+    },
+    {
+      "epoch": 2.1536,
+      "grad_norm": 0.5741236805915833,
+      "learning_rate": 0.0002478260869565217,
+      "loss": 4.0973,
+      "step": 1684
+    },
+    {
+      "epoch": 2.15488,
+      "grad_norm": 0.5604896545410156,
+      "learning_rate": 0.00024778570467088435,
+      "loss": 4.0886,
+      "step": 1685
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 0.5670208930969238,
+      "learning_rate": 0.000247745322385247,
+      "loss": 4.1643,
+      "step": 1686
+    },
+    {
+      "epoch": 2.15744,
+      "grad_norm": 0.5674409866333008,
+      "learning_rate": 0.0002477049400996096,
+      "loss": 4.1747,
+      "step": 1687
+    },
+    {
+      "epoch": 2.15872,
+      "grad_norm": 0.5394083857536316,
+      "learning_rate": 0.00024766455781397224,
+      "loss": 4.1235,
+      "step": 1688
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5426709651947021,
+      "learning_rate": 0.00024762417552833487,
+      "loss": 4.1005,
+      "step": 1689
+    },
+    {
+      "epoch": 2.16128,
+      "grad_norm": 0.5474469661712646,
+      "learning_rate": 0.0002475837932426975,
+      "loss": 4.1152,
+      "step": 1690
+    },
+    {
+      "epoch": 2.16256,
+      "grad_norm": 0.6519491076469421,
+      "learning_rate": 0.0002475434109570602,
+      "loss": 4.0955,
+      "step": 1691
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 0.5159724354743958,
+      "learning_rate": 0.0002475030286714228,
+      "loss": 4.1734,
+      "step": 1692
+    },
+    {
+      "epoch": 2.16512,
+      "grad_norm": 0.5642454028129578,
+      "learning_rate": 0.0002474626463857854,
+      "loss": 4.0931,
+      "step": 1693
+    },
+    {
+      "epoch": 2.1664,
+      "grad_norm": 0.5765841603279114,
+      "learning_rate": 0.000247422264100148,
+      "loss": 4.1322,
+      "step": 1694
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 0.5622481107711792,
+      "learning_rate": 0.0002473818818145107,
+      "loss": 4.1613,
+      "step": 1695
+    },
+    {
+      "epoch": 2.16896,
+      "grad_norm": 0.5255911350250244,
+      "learning_rate": 0.00024734149952887333,
+      "loss": 4.1129,
+      "step": 1696
+    },
+    {
+      "epoch": 2.17024,
+      "grad_norm": 0.5299469232559204,
+      "learning_rate": 0.00024730111724323596,
+      "loss": 4.0843,
+      "step": 1697
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 0.5808050632476807,
+      "learning_rate": 0.0002472607349575986,
+      "loss": 4.1286,
+      "step": 1698
+    },
+    {
+      "epoch": 2.1728,
+      "grad_norm": 0.563703179359436,
+      "learning_rate": 0.0002472203526719612,
+      "loss": 4.1928,
+      "step": 1699
+    },
+    {
+      "epoch": 2.17408,
+      "grad_norm": 0.5682193636894226,
+      "learning_rate": 0.00024717997038632385,
+      "loss": 4.1669,
+      "step": 1700
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 0.5745776295661926,
+      "learning_rate": 0.0002471395881006865,
+      "loss": 4.1129,
+      "step": 1701
+    },
+    {
+      "epoch": 2.17664,
+      "grad_norm": 0.5572461485862732,
+      "learning_rate": 0.0002470992058150491,
+      "loss": 4.1546,
+      "step": 1702
+    },
+    {
+      "epoch": 2.17792,
+      "grad_norm": 0.5437954664230347,
+      "learning_rate": 0.00024705882352941174,
+      "loss": 4.159,
+      "step": 1703
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.536284327507019,
+      "learning_rate": 0.0002470184412437744,
+      "loss": 4.1498,
+      "step": 1704
+    },
+    {
+      "epoch": 2.18048,
+      "grad_norm": 0.5438897609710693,
+      "learning_rate": 0.000246978058958137,
+      "loss": 4.0687,
+      "step": 1705
+    },
+    {
+      "epoch": 2.18176,
+      "grad_norm": 0.5237944722175598,
+      "learning_rate": 0.00024693767667249963,
+      "loss": 4.1393,
+      "step": 1706
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 0.5314120054244995,
+      "learning_rate": 0.00024689729438686226,
+      "loss": 4.1041,
+      "step": 1707
+    },
+    {
+      "epoch": 2.18432,
+      "grad_norm": 0.5380268096923828,
+      "learning_rate": 0.00024685691210122495,
+      "loss": 4.1046,
+      "step": 1708
+    },
+    {
+      "epoch": 2.1856,
+      "grad_norm": 0.5044991970062256,
+      "learning_rate": 0.0002468165298155875,
+      "loss": 4.0293,
+      "step": 1709
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 0.5603588223457336,
+      "learning_rate": 0.00024677614752995015,
+      "loss": 4.1426,
+      "step": 1710
+    },
+    {
+      "epoch": 2.18816,
+      "grad_norm": 0.5293896794319153,
+      "learning_rate": 0.0002467357652443128,
+      "loss": 4.0825,
+      "step": 1711
+    },
+    {
+      "epoch": 2.18944,
+      "grad_norm": 0.5398738980293274,
+      "learning_rate": 0.0002466953829586754,
+      "loss": 4.0774,
+      "step": 1712
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 0.5111704468727112,
+      "learning_rate": 0.0002466550006730381,
+      "loss": 4.0992,
+      "step": 1713
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 0.5483059287071228,
+      "learning_rate": 0.00024661461838740073,
+      "loss": 4.0721,
+      "step": 1714
+    },
+    {
+      "epoch": 2.19328,
+      "grad_norm": 0.5132558345794678,
+      "learning_rate": 0.00024657423610176336,
+      "loss": 4.1077,
+      "step": 1715
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 0.5515888929367065,
+      "learning_rate": 0.00024653385381612593,
+      "loss": 4.0439,
+      "step": 1716
+    },
+    {
+      "epoch": 2.19584,
+      "grad_norm": 0.5087373852729797,
+      "learning_rate": 0.0002464934715304886,
+      "loss": 4.1587,
+      "step": 1717
+    },
+    {
+      "epoch": 2.19712,
+      "grad_norm": 0.5192369222640991,
+      "learning_rate": 0.00024645308924485125,
+      "loss": 4.0419,
+      "step": 1718
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.5411146283149719,
+      "learning_rate": 0.0002464127069592139,
+      "loss": 4.1674,
+      "step": 1719
+    },
+    {
+      "epoch": 2.19968,
+      "grad_norm": 0.5327039957046509,
+      "learning_rate": 0.0002463723246735765,
+      "loss": 4.1486,
+      "step": 1720
+    },
+    {
+      "epoch": 2.20096,
+      "grad_norm": 0.5581017732620239,
+      "learning_rate": 0.00024633194238793914,
+      "loss": 4.1585,
+      "step": 1721
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 0.5399036407470703,
+      "learning_rate": 0.00024629156010230177,
+      "loss": 4.1631,
+      "step": 1722
+    },
+    {
+      "epoch": 2.20352,
+      "grad_norm": 0.5582420229911804,
+      "learning_rate": 0.0002462511778166644,
+      "loss": 4.0919,
+      "step": 1723
+    },
+    {
+      "epoch": 2.2048,
+      "grad_norm": 0.5208436846733093,
+      "learning_rate": 0.00024621079553102703,
+      "loss": 4.1291,
+      "step": 1724
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 0.5868030190467834,
+      "learning_rate": 0.00024617041324538966,
+      "loss": 4.0925,
+      "step": 1725
+    },
+    {
+      "epoch": 2.20736,
+      "grad_norm": 0.5274373292922974,
+      "learning_rate": 0.0002461300309597523,
+      "loss": 4.1289,
+      "step": 1726
+    },
+    {
+      "epoch": 2.20864,
+      "grad_norm": 0.5727094411849976,
+      "learning_rate": 0.0002460896486741149,
+      "loss": 4.1407,
+      "step": 1727
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 0.5434361100196838,
+      "learning_rate": 0.00024604926638847755,
+      "loss": 4.1471,
+      "step": 1728
+    },
+    {
+      "epoch": 2.2112,
+      "grad_norm": 0.5813825726509094,
+      "learning_rate": 0.0002460088841028402,
+      "loss": 4.1329,
+      "step": 1729
+    },
+    {
+      "epoch": 2.2124800000000002,
+      "grad_norm": 0.5424474477767944,
+      "learning_rate": 0.00024596850181720286,
+      "loss": 4.0815,
+      "step": 1730
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 0.546615481376648,
+      "learning_rate": 0.0002459281195315655,
+      "loss": 4.1526,
+      "step": 1731
+    },
+    {
+      "epoch": 2.21504,
+      "grad_norm": 0.5413877964019775,
+      "learning_rate": 0.00024588773724592807,
+      "loss": 4.104,
+      "step": 1732
+    },
+    {
+      "epoch": 2.21632,
+      "grad_norm": 0.5641849040985107,
+      "learning_rate": 0.0002458473549602907,
+      "loss": 4.0903,
+      "step": 1733
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.6042072176933289,
+      "learning_rate": 0.0002458069726746534,
+      "loss": 4.1078,
+      "step": 1734
+    },
+    {
+      "epoch": 2.21888,
+      "grad_norm": 0.5650026202201843,
+      "learning_rate": 0.000245766590389016,
+      "loss": 4.1688,
+      "step": 1735
+    },
+    {
+      "epoch": 2.22016,
+      "grad_norm": 0.5781028866767883,
+      "learning_rate": 0.00024572620810337864,
+      "loss": 4.1266,
+      "step": 1736
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 0.5524300932884216,
+      "learning_rate": 0.0002456858258177413,
+      "loss": 4.1045,
+      "step": 1737
+    },
+    {
+      "epoch": 2.22272,
+      "grad_norm": 0.5217704772949219,
+      "learning_rate": 0.0002456454435321039,
+      "loss": 4.1377,
+      "step": 1738
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 0.5524937510490417,
+      "learning_rate": 0.00024560506124646653,
+      "loss": 4.1003,
+      "step": 1739
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 0.5031275153160095,
+      "learning_rate": 0.00024556467896082916,
+      "loss": 4.1322,
+      "step": 1740
+    },
+    {
+      "epoch": 2.22656,
+      "grad_norm": 0.5842592716217041,
+      "learning_rate": 0.0002455242966751918,
+      "loss": 4.1308,
+      "step": 1741
+    },
+    {
+      "epoch": 2.22784,
+      "grad_norm": 0.5045456290245056,
+      "learning_rate": 0.0002454839143895544,
+      "loss": 4.1172,
+      "step": 1742
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 0.5572609901428223,
+      "learning_rate": 0.00024544353210391705,
+      "loss": 4.0625,
+      "step": 1743
+    },
+    {
+      "epoch": 2.2304,
+      "grad_norm": 0.5589746236801147,
+      "learning_rate": 0.0002454031498182797,
+      "loss": 4.0665,
+      "step": 1744
+    },
+    {
+      "epoch": 2.23168,
+      "grad_norm": 0.5695862174034119,
+      "learning_rate": 0.0002453627675326423,
+      "loss": 4.0537,
+      "step": 1745
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 0.5234766006469727,
+      "learning_rate": 0.00024532238524700494,
+      "loss": 4.1759,
+      "step": 1746
+    },
+    {
+      "epoch": 2.23424,
+      "grad_norm": 0.5754182934761047,
+      "learning_rate": 0.00024528200296136763,
+      "loss": 4.1404,
+      "step": 1747
+    },
+    {
+      "epoch": 2.23552,
+      "grad_norm": 0.5176681280136108,
+      "learning_rate": 0.00024524162067573026,
+      "loss": 4.1085,
+      "step": 1748
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.5456019043922424,
+      "learning_rate": 0.00024520123839009283,
+      "loss": 4.1433,
+      "step": 1749
+    },
+    {
+      "epoch": 2.23808,
+      "grad_norm": 0.5013742446899414,
+      "learning_rate": 0.00024516085610445547,
+      "loss": 4.1624,
+      "step": 1750
+    },
+    {
+      "epoch": 2.23936,
+      "grad_norm": 0.5228570103645325,
+      "learning_rate": 0.0002451204738188181,
+      "loss": 4.0994,
+      "step": 1751
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 0.5208329558372498,
+      "learning_rate": 0.0002450800915331808,
+      "loss": 4.1814,
+      "step": 1752
+    },
+    {
+      "epoch": 2.24192,
+      "grad_norm": 0.5297362804412842,
+      "learning_rate": 0.0002450397092475434,
+      "loss": 4.147,
+      "step": 1753
+    },
+    {
+      "epoch": 2.2432,
+      "grad_norm": 0.5192136764526367,
+      "learning_rate": 0.00024499932696190604,
+      "loss": 4.175,
+      "step": 1754
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 0.5594687461853027,
+      "learning_rate": 0.00024495894467626867,
+      "loss": 4.1391,
+      "step": 1755
+    },
+    {
+      "epoch": 2.24576,
+      "grad_norm": 0.5374172329902649,
+      "learning_rate": 0.0002449185623906313,
+      "loss": 4.1404,
+      "step": 1756
+    },
+    {
+      "epoch": 2.24704,
+      "grad_norm": 0.5575613975524902,
+      "learning_rate": 0.00024487818010499393,
+      "loss": 4.0406,
+      "step": 1757
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 0.507201075553894,
+      "learning_rate": 0.00024483779781935656,
+      "loss": 4.1444,
+      "step": 1758
+    },
+    {
+      "epoch": 2.2496,
+      "grad_norm": 0.5673658847808838,
+      "learning_rate": 0.0002447974155337192,
+      "loss": 4.0135,
+      "step": 1759
+    },
+    {
+      "epoch": 2.25088,
+      "grad_norm": 0.5208572149276733,
+      "learning_rate": 0.0002447570332480818,
+      "loss": 4.1486,
+      "step": 1760
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 0.5534422397613525,
+      "learning_rate": 0.00024471665096244445,
+      "loss": 4.0913,
+      "step": 1761
+    },
+    {
+      "epoch": 2.25344,
+      "grad_norm": 0.5177516937255859,
+      "learning_rate": 0.0002446762686768071,
+      "loss": 4.131,
+      "step": 1762
+    },
+    {
+      "epoch": 2.25472,
+      "grad_norm": 0.5717141032218933,
+      "learning_rate": 0.0002446358863911697,
+      "loss": 4.0954,
+      "step": 1763
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.5581514835357666,
+      "learning_rate": 0.00024459550410553234,
+      "loss": 4.1365,
+      "step": 1764
+    },
+    {
+      "epoch": 2.25728,
+      "grad_norm": 0.5282269716262817,
+      "learning_rate": 0.00024455512181989497,
+      "loss": 4.0941,
+      "step": 1765
+    },
+    {
+      "epoch": 2.25856,
+      "grad_norm": 0.5452956557273865,
+      "learning_rate": 0.0002445147395342576,
+      "loss": 4.1019,
+      "step": 1766
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 0.5826310515403748,
+      "learning_rate": 0.00024447435724862023,
+      "loss": 4.0641,
+      "step": 1767
+    },
+    {
+      "epoch": 2.26112,
+      "grad_norm": 0.5480028986930847,
+      "learning_rate": 0.00024443397496298286,
+      "loss": 4.096,
+      "step": 1768
+    },
+    {
+      "epoch": 2.2624,
+      "grad_norm": 0.5349941849708557,
+      "learning_rate": 0.00024439359267734554,
+      "loss": 4.088,
+      "step": 1769
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 0.5315197706222534,
+      "learning_rate": 0.0002443532103917082,
+      "loss": 4.1428,
+      "step": 1770
+    },
+    {
+      "epoch": 2.26496,
+      "grad_norm": 0.5485361218452454,
+      "learning_rate": 0.0002443128281060708,
+      "loss": 4.0701,
+      "step": 1771
+    },
+    {
+      "epoch": 2.26624,
+      "grad_norm": 0.49806830286979675,
+      "learning_rate": 0.0002442724458204334,
+      "loss": 4.0907,
+      "step": 1772
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 0.5614639520645142,
+      "learning_rate": 0.00024423206353479607,
+      "loss": 4.1116,
+      "step": 1773
+    },
+    {
+      "epoch": 2.2688,
+      "grad_norm": 0.5616679191589355,
+      "learning_rate": 0.0002441916812491587,
+      "loss": 4.1148,
+      "step": 1774
+    },
+    {
+      "epoch": 2.27008,
+      "grad_norm": 0.5605160593986511,
+      "learning_rate": 0.0002441512989635213,
+      "loss": 4.0161,
+      "step": 1775
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 0.5229089856147766,
+      "learning_rate": 0.00024411091667788393,
+      "loss": 4.1171,
+      "step": 1776
+    },
+    {
+      "epoch": 2.27264,
+      "grad_norm": 0.5960175395011902,
+      "learning_rate": 0.00024407053439224656,
+      "loss": 4.1207,
+      "step": 1777
+    },
+    {
+      "epoch": 2.27392,
+      "grad_norm": 0.5144199132919312,
+      "learning_rate": 0.00024403015210660922,
+      "loss": 4.162,
+      "step": 1778
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.5757052302360535,
+      "learning_rate": 0.00024398976982097185,
+      "loss": 4.129,
+      "step": 1779
+    },
+    {
+      "epoch": 2.27648,
+      "grad_norm": 0.5588791966438293,
+      "learning_rate": 0.00024394938753533448,
+      "loss": 4.0564,
+      "step": 1780
+    },
+    {
+      "epoch": 2.27776,
+      "grad_norm": 0.5604344606399536,
+      "learning_rate": 0.0002439090052496971,
+      "loss": 4.0195,
+      "step": 1781
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 0.5186293721199036,
+      "learning_rate": 0.00024386862296405976,
+      "loss": 4.0465,
+      "step": 1782
+    },
+    {
+      "epoch": 2.28032,
+      "grad_norm": 0.5566515922546387,
+      "learning_rate": 0.0002438282406784224,
+      "loss": 4.1026,
+      "step": 1783
+    },
+    {
+      "epoch": 2.2816,
+      "grad_norm": 0.5155622363090515,
+      "learning_rate": 0.000243787858392785,
+      "loss": 4.096,
+      "step": 1784
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 0.5197412967681885,
+      "learning_rate": 0.00024374747610714763,
+      "loss": 4.0739,
+      "step": 1785
+    },
+    {
+      "epoch": 2.28416,
+      "grad_norm": 0.5129163861274719,
+      "learning_rate": 0.00024370709382151028,
+      "loss": 4.054,
+      "step": 1786
+    },
+    {
+      "epoch": 2.28544,
+      "grad_norm": 0.4941207468509674,
+      "learning_rate": 0.0002436667115358729,
+      "loss": 4.0845,
+      "step": 1787
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 0.5023811459541321,
+      "learning_rate": 0.00024362632925023554,
+      "loss": 4.1381,
+      "step": 1788
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 0.507765531539917,
+      "learning_rate": 0.00024358594696459817,
+      "loss": 4.0451,
+      "step": 1789
+    },
+    {
+      "epoch": 2.2892799999999998,
+      "grad_norm": 0.4993181526660919,
+      "learning_rate": 0.0002435455646789608,
+      "loss": 4.0173,
+      "step": 1790
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 0.5167925953865051,
+      "learning_rate": 0.00024350518239332346,
+      "loss": 4.1821,
+      "step": 1791
+    },
+    {
+      "epoch": 2.29184,
+      "grad_norm": 0.4934568703174591,
+      "learning_rate": 0.00024346480010768606,
+      "loss": 4.1292,
+      "step": 1792
+    },
+    {
+      "epoch": 2.29312,
+      "grad_norm": 0.5314911007881165,
+      "learning_rate": 0.0002434244178220487,
+      "loss": 4.1175,
+      "step": 1793
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.5263243913650513,
+      "learning_rate": 0.00024338403553641132,
+      "loss": 4.0788,
+      "step": 1794
+    },
+    {
+      "epoch": 2.29568,
+      "grad_norm": 0.5136982202529907,
+      "learning_rate": 0.00024334365325077398,
+      "loss": 4.0407,
+      "step": 1795
+    },
+    {
+      "epoch": 2.29696,
+      "grad_norm": 0.53715580701828,
+      "learning_rate": 0.0002433032709651366,
+      "loss": 4.0982,
+      "step": 1796
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 0.537679135799408,
+      "learning_rate": 0.00024326288867949924,
+      "loss": 4.0186,
+      "step": 1797
+    },
+    {
+      "epoch": 2.2995200000000002,
+      "grad_norm": 0.5566760301589966,
+      "learning_rate": 0.00024322250639386187,
+      "loss": 4.0081,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3008,
+      "grad_norm": 0.5105803608894348,
+      "learning_rate": 0.00024318212410822453,
+      "loss": 4.1103,
+      "step": 1799
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 0.5302501916885376,
+      "learning_rate": 0.00024314174182258713,
+      "loss": 4.0884,
+      "step": 1800
+    },
+    {
+      "epoch": 2.30336,
+      "grad_norm": 0.5275909900665283,
+      "learning_rate": 0.00024310135953694976,
+      "loss": 4.1443,
+      "step": 1801
+    },
+    {
+      "epoch": 2.30464,
+      "grad_norm": 0.529685378074646,
+      "learning_rate": 0.0002430609772513124,
+      "loss": 4.0931,
+      "step": 1802
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 0.5178261995315552,
+      "learning_rate": 0.00024302059496567502,
+      "loss": 4.0161,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3072,
+      "grad_norm": 0.5614741444587708,
+      "learning_rate": 0.00024298021268003768,
+      "loss": 4.0476,
+      "step": 1804
+    },
+    {
+      "epoch": 2.30848,
+      "grad_norm": 0.5194922685623169,
+      "learning_rate": 0.0002429398303944003,
+      "loss": 4.0889,
+      "step": 1805
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 0.5637295842170715,
+      "learning_rate": 0.00024289944810876294,
+      "loss": 4.1037,
+      "step": 1806
+    },
+    {
+      "epoch": 2.31104,
+      "grad_norm": 0.5420023202896118,
+      "learning_rate": 0.00024285906582312554,
+      "loss": 4.1075,
+      "step": 1807
+    },
+    {
+      "epoch": 2.31232,
+      "grad_norm": 0.5864022374153137,
+      "learning_rate": 0.0002428186835374882,
+      "loss": 4.0976,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.5702569484710693,
+      "learning_rate": 0.00024277830125185083,
+      "loss": 4.0542,
+      "step": 1809
+    },
+    {
+      "epoch": 2.31488,
+      "grad_norm": 0.5524796843528748,
+      "learning_rate": 0.00024273791896621346,
+      "loss": 4.0888,
+      "step": 1810
+    },
+    {
+      "epoch": 2.31616,
+      "grad_norm": 0.5307409167289734,
+      "learning_rate": 0.0002426975366805761,
+      "loss": 4.1252,
+      "step": 1811
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 0.543082594871521,
+      "learning_rate": 0.00024265715439493875,
+      "loss": 4.1532,
+      "step": 1812
+    },
+    {
+      "epoch": 2.31872,
+      "grad_norm": 0.5330712199211121,
+      "learning_rate": 0.00024261677210930138,
+      "loss": 4.1552,
+      "step": 1813
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5282273292541504,
+      "learning_rate": 0.000242576389823664,
+      "loss": 4.1571,
+      "step": 1814
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 0.5447540283203125,
+      "learning_rate": 0.0002425360075380266,
+      "loss": 4.1582,
+      "step": 1815
+    },
+    {
+      "epoch": 2.32256,
+      "grad_norm": 0.5365498065948486,
+      "learning_rate": 0.0002424956252523893,
+      "loss": 4.1033,
+      "step": 1816
+    },
+    {
+      "epoch": 2.32384,
+      "grad_norm": 0.5478320121765137,
+      "learning_rate": 0.0002424552429667519,
+      "loss": 4.1227,
+      "step": 1817
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 0.5289247035980225,
+      "learning_rate": 0.00024241486068111453,
+      "loss": 4.1199,
+      "step": 1818
+    },
+    {
+      "epoch": 2.3264,
+      "grad_norm": 0.5543386340141296,
+      "learning_rate": 0.00024237447839547716,
+      "loss": 4.0492,
+      "step": 1819
+    },
+    {
+      "epoch": 2.32768,
+      "grad_norm": 0.5105769634246826,
+      "learning_rate": 0.0002423340961098398,
+      "loss": 4.0887,
+      "step": 1820
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 0.5353228449821472,
+      "learning_rate": 0.00024229371382420244,
+      "loss": 3.9982,
+      "step": 1821
+    },
+    {
+      "epoch": 2.33024,
+      "grad_norm": 0.4944192171096802,
+      "learning_rate": 0.00024225333153856507,
+      "loss": 4.0972,
+      "step": 1822
+    },
+    {
+      "epoch": 2.33152,
+      "grad_norm": 0.5060207843780518,
+      "learning_rate": 0.00024221294925292768,
+      "loss": 4.0681,
+      "step": 1823
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.503966212272644,
+      "learning_rate": 0.0002421725669672903,
+      "loss": 4.095,
+      "step": 1824
+    },
+    {
+      "epoch": 2.33408,
+      "grad_norm": 0.5177589654922485,
+      "learning_rate": 0.00024213218468165296,
+      "loss": 4.048,
+      "step": 1825
+    },
+    {
+      "epoch": 2.33536,
+      "grad_norm": 0.513486921787262,
+      "learning_rate": 0.0002420918023960156,
+      "loss": 4.1617,
+      "step": 1826
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 0.5109826326370239,
+      "learning_rate": 0.00024205142011037822,
+      "loss": 4.1553,
+      "step": 1827
+    },
+    {
+      "epoch": 2.33792,
+      "grad_norm": 0.5415542125701904,
+      "learning_rate": 0.00024201103782474085,
+      "loss": 4.0252,
+      "step": 1828
+    },
+    {
+      "epoch": 2.3392,
+      "grad_norm": 0.5250722169876099,
+      "learning_rate": 0.0002419706555391035,
+      "loss": 4.1262,
+      "step": 1829
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 0.490961492061615,
+      "learning_rate": 0.00024193027325346614,
+      "loss": 4.1085,
+      "step": 1830
+    },
+    {
+      "epoch": 2.34176,
+      "grad_norm": 0.5075300931930542,
+      "learning_rate": 0.00024188989096782877,
+      "loss": 4.0679,
+      "step": 1831
+    },
+    {
+      "epoch": 2.3430400000000002,
+      "grad_norm": 0.5052638053894043,
+      "learning_rate": 0.00024184950868219138,
+      "loss": 4.0838,
+      "step": 1832
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 0.5473994016647339,
+      "learning_rate": 0.000241809126396554,
+      "loss": 4.0862,
+      "step": 1833
+    },
+    {
+      "epoch": 2.3456,
+      "grad_norm": 0.502493143081665,
+      "learning_rate": 0.00024176874411091666,
+      "loss": 4.0654,
+      "step": 1834
+    },
+    {
+      "epoch": 2.34688,
+      "grad_norm": 0.5568253397941589,
+      "learning_rate": 0.0002417283618252793,
+      "loss": 4.072,
+      "step": 1835
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 0.535667359828949,
+      "learning_rate": 0.00024168797953964192,
+      "loss": 4.1611,
+      "step": 1836
+    },
+    {
+      "epoch": 2.34944,
+      "grad_norm": 0.5129615068435669,
+      "learning_rate": 0.00024164759725400455,
+      "loss": 4.0566,
+      "step": 1837
+    },
+    {
+      "epoch": 2.35072,
+      "grad_norm": 0.5144100785255432,
+      "learning_rate": 0.0002416072149683672,
+      "loss": 4.2091,
+      "step": 1838
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.500097393989563,
+      "learning_rate": 0.00024156683268272984,
+      "loss": 4.1066,
+      "step": 1839
+    },
+    {
+      "epoch": 2.35328,
+      "grad_norm": 0.520469069480896,
+      "learning_rate": 0.00024152645039709244,
+      "loss": 4.0778,
+      "step": 1840
+    },
+    {
+      "epoch": 2.35456,
+      "grad_norm": 0.5127030611038208,
+      "learning_rate": 0.00024148606811145507,
+      "loss": 4.1135,
+      "step": 1841
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 0.540470540523529,
+      "learning_rate": 0.00024144568582581773,
+      "loss": 4.0279,
+      "step": 1842
+    },
+    {
+      "epoch": 2.35712,
+      "grad_norm": 0.5427557826042175,
+      "learning_rate": 0.00024140530354018036,
+      "loss": 4.0734,
+      "step": 1843
+    },
+    {
+      "epoch": 2.3584,
+      "grad_norm": 0.5099870562553406,
+      "learning_rate": 0.000241364921254543,
+      "loss": 4.0705,
+      "step": 1844
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 0.5234814882278442,
+      "learning_rate": 0.00024132453896890562,
+      "loss": 4.0138,
+      "step": 1845
+    },
+    {
+      "epoch": 2.36096,
+      "grad_norm": 0.5056787729263306,
+      "learning_rate": 0.00024128415668326825,
+      "loss": 4.1004,
+      "step": 1846
+    },
+    {
+      "epoch": 2.36224,
+      "grad_norm": 0.5264895558357239,
+      "learning_rate": 0.0002412437743976309,
+      "loss": 4.0975,
+      "step": 1847
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 0.5280450582504272,
+      "learning_rate": 0.0002412033921119935,
+      "loss": 4.0805,
+      "step": 1848
+    },
+    {
+      "epoch": 2.3648,
+      "grad_norm": 0.509524941444397,
+      "learning_rate": 0.00024116300982635614,
+      "loss": 4.0108,
+      "step": 1849
+    },
+    {
+      "epoch": 2.36608,
+      "grad_norm": 0.5179848074913025,
+      "learning_rate": 0.00024112262754071877,
+      "loss": 4.0785,
+      "step": 1850
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 0.5557752847671509,
+      "learning_rate": 0.00024108224525508143,
+      "loss": 4.1493,
+      "step": 1851
+    },
+    {
+      "epoch": 2.36864,
+      "grad_norm": 0.5200245976448059,
+      "learning_rate": 0.00024104186296944406,
+      "loss": 4.0801,
+      "step": 1852
+    },
+    {
+      "epoch": 2.36992,
+      "grad_norm": 0.5246480703353882,
+      "learning_rate": 0.0002410014806838067,
+      "loss": 4.0722,
+      "step": 1853
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.5433456301689148,
+      "learning_rate": 0.00024096109839816932,
+      "loss": 4.1251,
+      "step": 1854
+    },
+    {
+      "epoch": 2.37248,
+      "grad_norm": 0.5170203447341919,
+      "learning_rate": 0.00024092071611253198,
+      "loss": 4.0856,
+      "step": 1855
+    },
+    {
+      "epoch": 2.37376,
+      "grad_norm": 0.5600740313529968,
+      "learning_rate": 0.00024088033382689458,
+      "loss": 4.0519,
+      "step": 1856
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 0.549588680267334,
+      "learning_rate": 0.0002408399515412572,
+      "loss": 4.0641,
+      "step": 1857
+    },
+    {
+      "epoch": 2.3763199999999998,
+      "grad_norm": 0.5169657468795776,
+      "learning_rate": 0.00024079956925561984,
+      "loss": 4.0794,
+      "step": 1858
+    },
+    {
+      "epoch": 2.3776,
+      "grad_norm": 0.5614109039306641,
+      "learning_rate": 0.00024075918696998247,
+      "loss": 4.1284,
+      "step": 1859
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 0.5183959603309631,
+      "learning_rate": 0.00024071880468434513,
+      "loss": 4.133,
+      "step": 1860
+    },
+    {
+      "epoch": 2.38016,
+      "grad_norm": 0.584951639175415,
+      "learning_rate": 0.00024067842239870776,
+      "loss": 4.0667,
+      "step": 1861
+    },
+    {
+      "epoch": 2.38144,
+      "grad_norm": 0.5184288620948792,
+      "learning_rate": 0.00024063804011307039,
+      "loss": 4.0535,
+      "step": 1862
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 0.558650016784668,
+      "learning_rate": 0.000240597657827433,
+      "loss": 4.0802,
+      "step": 1863
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 0.5584487318992615,
+      "learning_rate": 0.00024055727554179565,
+      "loss": 4.1414,
+      "step": 1864
+    },
+    {
+      "epoch": 2.38528,
+      "grad_norm": 0.5305871963500977,
+      "learning_rate": 0.00024051689325615828,
+      "loss": 4.0945,
+      "step": 1865
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 0.5361427664756775,
+      "learning_rate": 0.0002404765109705209,
+      "loss": 4.0935,
+      "step": 1866
+    },
+    {
+      "epoch": 2.38784,
+      "grad_norm": 0.49007391929626465,
+      "learning_rate": 0.00024043612868488354,
+      "loss": 4.1032,
+      "step": 1867
+    },
+    {
+      "epoch": 2.38912,
+      "grad_norm": 0.5409270524978638,
+      "learning_rate": 0.0002403957463992462,
+      "loss": 4.0758,
+      "step": 1868
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.5215279459953308,
+      "learning_rate": 0.00024035536411360882,
+      "loss": 4.0599,
+      "step": 1869
+    },
+    {
+      "epoch": 2.39168,
+      "grad_norm": 0.5199880599975586,
+      "learning_rate": 0.00024031498182797145,
+      "loss": 4.0554,
+      "step": 1870
+    },
+    {
+      "epoch": 2.39296,
+      "grad_norm": 0.5253363251686096,
+      "learning_rate": 0.00024027459954233406,
+      "loss": 4.1072,
+      "step": 1871
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 0.5469797253608704,
+      "learning_rate": 0.0002402342172566967,
+      "loss": 4.0496,
+      "step": 1872
+    },
+    {
+      "epoch": 2.39552,
+      "grad_norm": 0.5029854774475098,
+      "learning_rate": 0.00024019383497105934,
+      "loss": 4.0267,
+      "step": 1873
+    },
+    {
+      "epoch": 2.3968,
+      "grad_norm": 0.5808607339859009,
+      "learning_rate": 0.00024015345268542197,
+      "loss": 4.0903,
+      "step": 1874
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 0.4929969906806946,
+      "learning_rate": 0.0002401130703997846,
+      "loss": 4.0763,
+      "step": 1875
+    },
+    {
+      "epoch": 2.39936,
+      "grad_norm": 0.5704619288444519,
+      "learning_rate": 0.00024007268811414723,
+      "loss": 4.0923,
+      "step": 1876
+    },
+    {
+      "epoch": 2.40064,
+      "grad_norm": 0.5273481607437134,
+      "learning_rate": 0.0002400323058285099,
+      "loss": 4.0345,
+      "step": 1877
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 0.5336358547210693,
+      "learning_rate": 0.00023999192354287252,
+      "loss": 4.1127,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4032,
+      "grad_norm": 0.5384494662284851,
+      "learning_rate": 0.00023995154125723512,
+      "loss": 4.131,
+      "step": 1879
+    },
+    {
+      "epoch": 2.40448,
+      "grad_norm": 0.5182468891143799,
+      "learning_rate": 0.00023991115897159775,
+      "loss": 4.088,
+      "step": 1880
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 0.54859459400177,
+      "learning_rate": 0.0002398707766859604,
+      "loss": 4.0991,
+      "step": 1881
+    },
+    {
+      "epoch": 2.40704,
+      "grad_norm": 0.5565007925033569,
+      "learning_rate": 0.00023983039440032304,
+      "loss": 4.0784,
+      "step": 1882
+    },
+    {
+      "epoch": 2.40832,
+      "grad_norm": 0.5592086315155029,
+      "learning_rate": 0.00023979001211468567,
+      "loss": 4.126,
+      "step": 1883
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.5493931174278259,
+      "learning_rate": 0.0002397496298290483,
+      "loss": 4.1278,
+      "step": 1884
+    },
+    {
+      "epoch": 2.41088,
+      "grad_norm": 0.5680795311927795,
+      "learning_rate": 0.00023970924754341093,
+      "loss": 4.1366,
+      "step": 1885
+    },
+    {
+      "epoch": 2.41216,
+      "grad_norm": 0.5174584984779358,
+      "learning_rate": 0.0002396688652577736,
+      "loss": 4.0928,
+      "step": 1886
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 0.5714235305786133,
+      "learning_rate": 0.0002396284829721362,
+      "loss": 4.1146,
+      "step": 1887
+    },
+    {
+      "epoch": 2.41472,
+      "grad_norm": 0.5287262201309204,
+      "learning_rate": 0.00023958810068649882,
+      "loss": 4.1175,
+      "step": 1888
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 0.5177103877067566,
+      "learning_rate": 0.00023954771840086145,
+      "loss": 4.0145,
+      "step": 1889
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 0.5252452492713928,
+      "learning_rate": 0.0002395073361152241,
+      "loss": 4.0529,
+      "step": 1890
+    },
+    {
+      "epoch": 2.41856,
+      "grad_norm": 0.5262645483016968,
+      "learning_rate": 0.00023946695382958674,
+      "loss": 4.0403,
+      "step": 1891
+    },
+    {
+      "epoch": 2.4198399999999998,
+      "grad_norm": 0.5197089910507202,
+      "learning_rate": 0.00023942657154394937,
+      "loss": 4.0744,
+      "step": 1892
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 0.512195885181427,
+      "learning_rate": 0.000239386189258312,
+      "loss": 4.0726,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4224,
+      "grad_norm": 0.5221033692359924,
+      "learning_rate": 0.00023934580697267466,
+      "loss": 4.1091,
+      "step": 1894
+    },
+    {
+      "epoch": 2.42368,
+      "grad_norm": 0.5204744338989258,
+      "learning_rate": 0.00023930542468703726,
+      "loss": 4.0276,
+      "step": 1895
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 0.5527137517929077,
+      "learning_rate": 0.0002392650424013999,
+      "loss": 4.0472,
+      "step": 1896
+    },
+    {
+      "epoch": 2.42624,
+      "grad_norm": 0.5223075151443481,
+      "learning_rate": 0.00023922466011576252,
+      "loss": 4.0678,
+      "step": 1897
+    },
+    {
+      "epoch": 2.42752,
+      "grad_norm": 0.5136579871177673,
+      "learning_rate": 0.00023918427783012515,
+      "loss": 3.968,
+      "step": 1898
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.5277352929115295,
+      "learning_rate": 0.0002391438955444878,
+      "loss": 4.0798,
+      "step": 1899
+    },
+    {
+      "epoch": 2.4300800000000002,
+      "grad_norm": 0.524479329586029,
+      "learning_rate": 0.00023910351325885044,
+      "loss": 4.0096,
+      "step": 1900
+    },
+    {
+      "epoch": 2.43136,
+      "grad_norm": 0.5202826261520386,
+      "learning_rate": 0.00023906313097321307,
+      "loss": 4.0303,
+      "step": 1901
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 0.5311558842658997,
+      "learning_rate": 0.00023902274868757567,
+      "loss": 4.1015,
+      "step": 1902
+    },
+    {
+      "epoch": 2.43392,
+      "grad_norm": 0.5515623092651367,
+      "learning_rate": 0.00023898236640193835,
+      "loss": 4.0423,
+      "step": 1903
+    },
+    {
+      "epoch": 2.4352,
+      "grad_norm": 0.55174320936203,
+      "learning_rate": 0.00023894198411630096,
+      "loss": 4.0474,
+      "step": 1904
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 0.5991604924201965,
+      "learning_rate": 0.0002389016018306636,
+      "loss": 4.1282,
+      "step": 1905
+    },
+    {
+      "epoch": 2.43776,
+      "grad_norm": 0.5626751780509949,
+      "learning_rate": 0.00023886121954502622,
+      "loss": 4.0855,
+      "step": 1906
+    },
+    {
+      "epoch": 2.43904,
+      "grad_norm": 0.5392490029335022,
+      "learning_rate": 0.00023882083725938887,
+      "loss": 4.1173,
+      "step": 1907
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 0.5552417635917664,
+      "learning_rate": 0.0002387804549737515,
+      "loss": 4.123,
+      "step": 1908
+    },
+    {
+      "epoch": 2.4416,
+      "grad_norm": 0.5797650218009949,
+      "learning_rate": 0.00023874007268811413,
+      "loss": 4.0547,
+      "step": 1909
+    },
+    {
+      "epoch": 2.44288,
+      "grad_norm": 0.5046030282974243,
+      "learning_rate": 0.00023869969040247674,
+      "loss": 4.1031,
+      "step": 1910
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 0.6118117570877075,
+      "learning_rate": 0.00023865930811683937,
+      "loss": 4.0797,
+      "step": 1911
+    },
+    {
+      "epoch": 2.44544,
+      "grad_norm": 0.5165206789970398,
+      "learning_rate": 0.00023861892583120203,
+      "loss": 4.0243,
+      "step": 1912
+    },
+    {
+      "epoch": 2.44672,
+      "grad_norm": 0.5483618378639221,
+      "learning_rate": 0.00023857854354556466,
+      "loss": 4.0234,
+      "step": 1913
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.5128400325775146,
+      "learning_rate": 0.00023853816125992729,
+      "loss": 3.9816,
+      "step": 1914
+    },
+    {
+      "epoch": 2.44928,
+      "grad_norm": 0.5472554564476013,
+      "learning_rate": 0.00023849777897428992,
+      "loss": 4.0127,
+      "step": 1915
+    },
+    {
+      "epoch": 2.45056,
+      "grad_norm": 0.5363740921020508,
+      "learning_rate": 0.00023845739668865257,
+      "loss": 4.1075,
+      "step": 1916
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 0.5734500288963318,
+      "learning_rate": 0.0002384170144030152,
+      "loss": 4.0291,
+      "step": 1917
+    },
+    {
+      "epoch": 2.45312,
+      "grad_norm": 0.5291422009468079,
+      "learning_rate": 0.00023837663211737783,
+      "loss": 4.119,
+      "step": 1918
+    },
+    {
+      "epoch": 2.4544,
+      "grad_norm": 0.541768491268158,
+      "learning_rate": 0.00023833624983174044,
+      "loss": 4.0881,
+      "step": 1919
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 0.5272345542907715,
+      "learning_rate": 0.0002382958675461031,
+      "loss": 4.0666,
+      "step": 1920
+    },
+    {
+      "epoch": 2.45696,
+      "grad_norm": 0.5372911095619202,
+      "learning_rate": 0.00023825548526046572,
+      "loss": 4.0557,
+      "step": 1921
+    },
+    {
+      "epoch": 2.45824,
+      "grad_norm": 0.5213048458099365,
+      "learning_rate": 0.00023821510297482835,
+      "loss": 4.0735,
+      "step": 1922
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 0.5202589631080627,
+      "learning_rate": 0.00023817472068919098,
+      "loss": 4.0209,
+      "step": 1923
+    },
+    {
+      "epoch": 2.4608,
+      "grad_norm": 0.4860571026802063,
+      "learning_rate": 0.0002381343384035536,
+      "loss": 4.1663,
+      "step": 1924
+    },
+    {
+      "epoch": 2.46208,
+      "grad_norm": 0.5281109809875488,
+      "learning_rate": 0.00023809395611791627,
+      "loss": 4.0593,
+      "step": 1925
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 0.508752167224884,
+      "learning_rate": 0.0002380535738322789,
+      "loss": 4.0568,
+      "step": 1926
+    },
+    {
+      "epoch": 2.46464,
+      "grad_norm": 0.5302470326423645,
+      "learning_rate": 0.0002380131915466415,
+      "loss": 4.0717,
+      "step": 1927
+    },
+    {
+      "epoch": 2.46592,
+      "grad_norm": 0.5096002817153931,
+      "learning_rate": 0.00023797280926100413,
+      "loss": 4.0652,
+      "step": 1928
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.5053885579109192,
+      "learning_rate": 0.0002379324269753668,
+      "loss": 4.1062,
+      "step": 1929
+    },
+    {
+      "epoch": 2.46848,
+      "grad_norm": 0.5277340412139893,
+      "learning_rate": 0.00023789204468972942,
+      "loss": 4.1013,
+      "step": 1930
+    },
+    {
+      "epoch": 2.46976,
+      "grad_norm": 0.5312148928642273,
+      "learning_rate": 0.00023785166240409205,
+      "loss": 4.0802,
+      "step": 1931
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 0.5116835832595825,
+      "learning_rate": 0.00023781128011845468,
+      "loss": 3.9968,
+      "step": 1932
+    },
+    {
+      "epoch": 2.47232,
+      "grad_norm": 0.5169728398323059,
+      "learning_rate": 0.00023777089783281734,
+      "loss": 4.0566,
+      "step": 1933
+    },
+    {
+      "epoch": 2.4736000000000002,
+      "grad_norm": 0.5306958556175232,
+      "learning_rate": 0.00023773051554717997,
+      "loss": 4.0178,
+      "step": 1934
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 0.5176283717155457,
+      "learning_rate": 0.00023769013326154257,
+      "loss": 4.0897,
+      "step": 1935
+    },
+    {
+      "epoch": 2.47616,
+      "grad_norm": 0.5138524174690247,
+      "learning_rate": 0.0002376497509759052,
+      "loss": 4.0544,
+      "step": 1936
+    },
+    {
+      "epoch": 2.47744,
+      "grad_norm": 0.5273920297622681,
+      "learning_rate": 0.00023760936869026783,
+      "loss": 4.0539,
+      "step": 1937
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 0.5098421573638916,
+      "learning_rate": 0.0002375689864046305,
+      "loss": 4.0386,
+      "step": 1938
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.5550551414489746,
+      "learning_rate": 0.00023752860411899312,
+      "loss": 4.0598,
+      "step": 1939
+    },
+    {
+      "epoch": 2.48128,
+      "grad_norm": 0.5031646490097046,
+      "learning_rate": 0.00023748822183335575,
+      "loss": 4.1084,
+      "step": 1940
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 0.5071708559989929,
+      "learning_rate": 0.00023744783954771838,
+      "loss": 4.07,
+      "step": 1941
+    },
+    {
+      "epoch": 2.48384,
+      "grad_norm": 0.5383012890815735,
+      "learning_rate": 0.00023740745726208104,
+      "loss": 4.0044,
+      "step": 1942
+    },
+    {
+      "epoch": 2.48512,
+      "grad_norm": 0.5164763927459717,
+      "learning_rate": 0.00023736707497644364,
+      "loss": 4.1217,
+      "step": 1943
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.531086266040802,
+      "learning_rate": 0.00023732669269080627,
+      "loss": 4.055,
+      "step": 1944
+    },
+    {
+      "epoch": 2.48768,
+      "grad_norm": 0.5626212358474731,
+      "learning_rate": 0.0002372863104051689,
+      "loss": 3.9839,
+      "step": 1945
+    },
+    {
+      "epoch": 2.48896,
+      "grad_norm": 0.48967134952545166,
+      "learning_rate": 0.00023724592811953156,
+      "loss": 4.0099,
+      "step": 1946
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 0.4974314272403717,
+      "learning_rate": 0.00023720554583389419,
+      "loss": 4.0544,
+      "step": 1947
+    },
+    {
+      "epoch": 2.49152,
+      "grad_norm": 0.5294041037559509,
+      "learning_rate": 0.00023716516354825682,
+      "loss": 4.0783,
+      "step": 1948
+    },
+    {
+      "epoch": 2.4928,
+      "grad_norm": 0.5068480372428894,
+      "learning_rate": 0.00023712478126261945,
+      "loss": 4.0757,
+      "step": 1949
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 0.4977387487888336,
+      "learning_rate": 0.0002370843989769821,
+      "loss": 4.1196,
+      "step": 1950
+    },
+    {
+      "epoch": 2.49536,
+      "grad_norm": 0.4864615797996521,
+      "learning_rate": 0.0002370440166913447,
+      "loss": 4.0737,
+      "step": 1951
+    },
+    {
+      "epoch": 2.49664,
+      "grad_norm": 0.5214549899101257,
+      "learning_rate": 0.00023700363440570734,
+      "loss": 4.0846,
+      "step": 1952
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 0.538516640663147,
+      "learning_rate": 0.00023696325212006997,
+      "loss": 4.0236,
+      "step": 1953
+    },
+    {
+      "epoch": 2.4992,
+      "grad_norm": 0.5184758901596069,
+      "learning_rate": 0.0002369228698344326,
+      "loss": 4.0424,
+      "step": 1954
+    },
+    {
+      "epoch": 2.50048,
+      "grad_norm": 0.512209415435791,
+      "learning_rate": 0.00023688248754879525,
+      "loss": 4.0921,
+      "step": 1955
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 0.5255154371261597,
+      "learning_rate": 0.00023684210526315788,
+      "loss": 4.1313,
+      "step": 1956
+    },
+    {
+      "epoch": 2.50304,
+      "grad_norm": 0.5283589959144592,
+      "learning_rate": 0.00023680172297752051,
+      "loss": 4.0266,
+      "step": 1957
+    },
+    {
+      "epoch": 2.50432,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.00023676134069188312,
+      "loss": 4.1258,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.5446499586105347,
+      "learning_rate": 0.00023672095840624577,
+      "loss": 4.0665,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5068799999999998,
+      "grad_norm": 0.4942466914653778,
+      "learning_rate": 0.0002366805761206084,
+      "loss": 4.0474,
+      "step": 1960
+    },
+    {
+      "epoch": 2.50816,
+      "grad_norm": 0.5173544883728027,
+      "learning_rate": 0.00023664019383497103,
+      "loss": 3.9846,
+      "step": 1961
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 0.5310766100883484,
+      "learning_rate": 0.00023659981154933366,
+      "loss": 4.0044,
+      "step": 1962
+    },
+    {
+      "epoch": 2.51072,
+      "grad_norm": 0.5134363770484924,
+      "learning_rate": 0.00023655942926369632,
+      "loss": 4.0734,
+      "step": 1963
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 0.5086606740951538,
+      "learning_rate": 0.00023651904697805895,
+      "loss": 4.0758,
+      "step": 1964
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 0.5497402548789978,
+      "learning_rate": 0.00023647866469242158,
+      "loss": 4.075,
+      "step": 1965
+    },
+    {
+      "epoch": 2.51456,
+      "grad_norm": 0.5068183541297913,
+      "learning_rate": 0.00023643828240678418,
+      "loss": 4.0666,
+      "step": 1966
+    },
+    {
+      "epoch": 2.51584,
+      "grad_norm": 0.5550326704978943,
+      "learning_rate": 0.00023639790012114681,
+      "loss": 4.0539,
+      "step": 1967
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 0.49992436170578003,
+      "learning_rate": 0.00023635751783550947,
+      "loss": 4.0962,
+      "step": 1968
+    },
+    {
+      "epoch": 2.5183999999999997,
+      "grad_norm": 0.5428693294525146,
+      "learning_rate": 0.0002363171355498721,
+      "loss": 3.9748,
+      "step": 1969
+    },
+    {
+      "epoch": 2.51968,
+      "grad_norm": 0.5194831490516663,
+      "learning_rate": 0.00023627675326423473,
+      "loss": 4.1121,
+      "step": 1970
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 0.5650715827941895,
+      "learning_rate": 0.00023623637097859736,
+      "loss": 4.0938,
+      "step": 1971
+    },
+    {
+      "epoch": 2.52224,
+      "grad_norm": 0.5177878141403198,
+      "learning_rate": 0.00023619598869296002,
+      "loss": 4.0946,
+      "step": 1972
+    },
+    {
+      "epoch": 2.52352,
+      "grad_norm": 0.5797203183174133,
+      "learning_rate": 0.00023615560640732265,
+      "loss": 4.0559,
+      "step": 1973
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.5099804997444153,
+      "learning_rate": 0.00023611522412168525,
+      "loss": 4.0196,
+      "step": 1974
+    },
+    {
+      "epoch": 2.52608,
+      "grad_norm": 0.5663504004478455,
+      "learning_rate": 0.00023607484183604788,
+      "loss": 4.0268,
+      "step": 1975
+    },
+    {
+      "epoch": 2.52736,
+      "grad_norm": 0.4963954985141754,
+      "learning_rate": 0.00023603445955041054,
+      "loss": 4.0723,
+      "step": 1976
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 0.5366328954696655,
+      "learning_rate": 0.00023599407726477317,
+      "loss": 4.0549,
+      "step": 1977
+    },
+    {
+      "epoch": 2.5299199999999997,
+      "grad_norm": 0.515116274356842,
+      "learning_rate": 0.0002359536949791358,
+      "loss": 4.0863,
+      "step": 1978
+    },
+    {
+      "epoch": 2.5312,
+      "grad_norm": 0.5633923411369324,
+      "learning_rate": 0.00023591331269349843,
+      "loss": 4.0656,
+      "step": 1979
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 0.515352725982666,
+      "learning_rate": 0.00023587293040786106,
+      "loss": 4.0995,
+      "step": 1980
+    },
+    {
+      "epoch": 2.53376,
+      "grad_norm": 0.5454360842704773,
+      "learning_rate": 0.00023583254812222372,
+      "loss": 4.0187,
+      "step": 1981
+    },
+    {
+      "epoch": 2.53504,
+      "grad_norm": 0.5400582551956177,
+      "learning_rate": 0.00023579216583658632,
+      "loss": 4.0149,
+      "step": 1982
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 0.5306532382965088,
+      "learning_rate": 0.00023575178355094895,
+      "loss": 4.0082,
+      "step": 1983
+    },
+    {
+      "epoch": 2.5376,
+      "grad_norm": 0.535999059677124,
+      "learning_rate": 0.00023571140126531158,
+      "loss": 4.0181,
+      "step": 1984
+    },
+    {
+      "epoch": 2.53888,
+      "grad_norm": 0.5338032245635986,
+      "learning_rate": 0.00023567101897967424,
+      "loss": 3.9664,
+      "step": 1985
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 0.5602062344551086,
+      "learning_rate": 0.00023563063669403687,
+      "loss": 4.0308,
+      "step": 1986
+    },
+    {
+      "epoch": 2.54144,
+      "grad_norm": 0.5280570387840271,
+      "learning_rate": 0.0002355902544083995,
+      "loss": 4.0465,
+      "step": 1987
+    },
+    {
+      "epoch": 2.54272,
+      "grad_norm": 0.5154047608375549,
+      "learning_rate": 0.00023554987212276213,
+      "loss": 4.0587,
+      "step": 1988
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.551994800567627,
+      "learning_rate": 0.00023550948983712478,
+      "loss": 4.0683,
+      "step": 1989
+    },
+    {
+      "epoch": 2.54528,
+      "grad_norm": 0.5081275105476379,
+      "learning_rate": 0.00023546910755148741,
+      "loss": 4.0972,
+      "step": 1990
+    },
+    {
+      "epoch": 2.54656,
+      "grad_norm": 0.5598514676094055,
+      "learning_rate": 0.00023542872526585002,
+      "loss": 4.0323,
+      "step": 1991
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 0.48178520798683167,
+      "learning_rate": 0.00023538834298021265,
+      "loss": 3.9906,
+      "step": 1992
+    },
+    {
+      "epoch": 2.5491200000000003,
+      "grad_norm": 0.558735191822052,
+      "learning_rate": 0.00023534796069457528,
+      "loss": 4.0245,
+      "step": 1993
+    },
+    {
+      "epoch": 2.5504,
+      "grad_norm": 0.540560781955719,
+      "learning_rate": 0.00023530757840893794,
+      "loss": 4.1077,
+      "step": 1994
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 0.5557510852813721,
+      "learning_rate": 0.00023526719612330057,
+      "loss": 4.0786,
+      "step": 1995
+    },
+    {
+      "epoch": 2.55296,
+      "grad_norm": 0.5289320945739746,
+      "learning_rate": 0.0002352268138376632,
+      "loss": 4.0445,
+      "step": 1996
+    },
+    {
+      "epoch": 2.55424,
+      "grad_norm": 0.5117323398590088,
+      "learning_rate": 0.0002351864315520258,
+      "loss": 4.0146,
+      "step": 1997
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 0.525393009185791,
+      "learning_rate": 0.00023514604926638848,
+      "loss": 4.1132,
+      "step": 1998
+    },
+    {
+      "epoch": 2.5568,
+      "grad_norm": 0.5251143574714661,
+      "learning_rate": 0.00023510566698075109,
+      "loss": 4.0507,
+      "step": 1999
+    },
+    {
+      "epoch": 2.55808,
+      "grad_norm": 0.5183683633804321,
+      "learning_rate": 0.00023506528469511372,
+      "loss": 4.0713,
+      "step": 2000
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 0.5445199012756348,
+      "learning_rate": 0.00023502490240947635,
+      "loss": 4.0467,
+      "step": 2001
+    },
+    {
+      "epoch": 2.5606400000000002,
+      "grad_norm": 0.5163613557815552,
+      "learning_rate": 0.000234984520123839,
+      "loss": 4.0313,
+      "step": 2002
+    },
+    {
+      "epoch": 2.5619199999999998,
+      "grad_norm": 0.5514837503433228,
+      "learning_rate": 0.00023494413783820163,
+      "loss": 4.03,
+      "step": 2003
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.527066707611084,
+      "learning_rate": 0.00023490375555256426,
+      "loss": 4.097,
+      "step": 2004
+    },
+    {
+      "epoch": 2.56448,
+      "grad_norm": 0.5349636673927307,
+      "learning_rate": 0.0002348633732669269,
+      "loss": 4.019,
+      "step": 2005
+    },
+    {
+      "epoch": 2.56576,
+      "grad_norm": 0.5117310285568237,
+      "learning_rate": 0.0002348229909812895,
+      "loss": 3.9889,
+      "step": 2006
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 0.5110161900520325,
+      "learning_rate": 0.00023478260869565215,
+      "loss": 4.0419,
+      "step": 2007
+    },
+    {
+      "epoch": 2.56832,
+      "grad_norm": 0.5449533462524414,
+      "learning_rate": 0.00023474222641001478,
+      "loss": 4.1393,
+      "step": 2008
+    },
+    {
+      "epoch": 2.5696,
+      "grad_norm": 0.5280770063400269,
+      "learning_rate": 0.0002347018441243774,
+      "loss": 4.0109,
+      "step": 2009
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 0.5067214369773865,
+      "learning_rate": 0.00023466146183874004,
+      "loss": 4.0415,
+      "step": 2010
+    },
+    {
+      "epoch": 2.5721600000000002,
+      "grad_norm": 0.5153365135192871,
+      "learning_rate": 0.0002346210795531027,
+      "loss": 4.0349,
+      "step": 2011
+    },
+    {
+      "epoch": 2.5734399999999997,
+      "grad_norm": 0.509453296661377,
+      "learning_rate": 0.00023458069726746533,
+      "loss": 4.0768,
+      "step": 2012
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 0.5531152486801147,
+      "learning_rate": 0.00023454031498182796,
+      "loss": 4.0643,
+      "step": 2013
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 0.5219248533248901,
+      "learning_rate": 0.00023449993269619056,
+      "loss": 4.0365,
+      "step": 2014
+    },
+    {
+      "epoch": 2.57728,
+      "grad_norm": 0.5232629776000977,
+      "learning_rate": 0.00023445955041055322,
+      "loss": 4.0989,
+      "step": 2015
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 0.5008588433265686,
+      "learning_rate": 0.00023441916812491585,
+      "loss": 3.9987,
+      "step": 2016
+    },
+    {
+      "epoch": 2.57984,
+      "grad_norm": 0.5122629404067993,
+      "learning_rate": 0.00023437878583927848,
+      "loss": 4.0479,
+      "step": 2017
+    },
+    {
+      "epoch": 2.58112,
+      "grad_norm": 0.48956623673439026,
+      "learning_rate": 0.0002343384035536411,
+      "loss": 4.1008,
+      "step": 2018
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.5428857207298279,
+      "learning_rate": 0.00023429802126800374,
+      "loss": 4.0692,
+      "step": 2019
+    },
+    {
+      "epoch": 2.58368,
+      "grad_norm": 0.5066841840744019,
+      "learning_rate": 0.0002342576389823664,
+      "loss": 4.0896,
+      "step": 2020
+    },
+    {
+      "epoch": 2.58496,
+      "grad_norm": 0.5314661860466003,
+      "learning_rate": 0.00023421725669672903,
+      "loss": 3.9449,
+      "step": 2021
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 0.5035257935523987,
+      "learning_rate": 0.00023417687441109163,
+      "loss": 4.0175,
+      "step": 2022
+    },
+    {
+      "epoch": 2.58752,
+      "grad_norm": 0.4997243583202362,
+      "learning_rate": 0.00023413649212545426,
+      "loss": 4.1032,
+      "step": 2023
+    },
+    {
+      "epoch": 2.5888,
+      "grad_norm": 0.4923264980316162,
+      "learning_rate": 0.00023409610983981692,
+      "loss": 4.0363,
+      "step": 2024
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 0.5122047662734985,
+      "learning_rate": 0.00023405572755417955,
+      "loss": 4.0297,
+      "step": 2025
+    },
+    {
+      "epoch": 2.59136,
+      "grad_norm": 0.5043531060218811,
+      "learning_rate": 0.00023401534526854218,
+      "loss": 4.0732,
+      "step": 2026
+    },
+    {
+      "epoch": 2.59264,
+      "grad_norm": 0.5018818974494934,
+      "learning_rate": 0.0002339749629829048,
+      "loss": 4.0188,
+      "step": 2027
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 0.4968951940536499,
+      "learning_rate": 0.00023393458069726747,
+      "loss": 4.0574,
+      "step": 2028
+    },
+    {
+      "epoch": 2.5952,
+      "grad_norm": 0.5588295459747314,
+      "learning_rate": 0.0002338941984116301,
+      "loss": 4.0049,
+      "step": 2029
+    },
+    {
+      "epoch": 2.59648,
+      "grad_norm": 0.5243644714355469,
+      "learning_rate": 0.0002338538161259927,
+      "loss": 3.8995,
+      "step": 2030
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 0.5623689293861389,
+      "learning_rate": 0.00023381343384035533,
+      "loss": 4.0505,
+      "step": 2031
+    },
+    {
+      "epoch": 2.59904,
+      "grad_norm": 0.5318677425384521,
+      "learning_rate": 0.00023377305155471796,
+      "loss": 3.9892,
+      "step": 2032
+    },
+    {
+      "epoch": 2.60032,
+      "grad_norm": 0.5533915758132935,
+      "learning_rate": 0.00023373266926908062,
+      "loss": 4.0111,
+      "step": 2033
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.5214724540710449,
+      "learning_rate": 0.00023369228698344325,
+      "loss": 4.0176,
+      "step": 2034
+    },
+    {
+      "epoch": 2.60288,
+      "grad_norm": 0.550571620464325,
+      "learning_rate": 0.00023365190469780588,
+      "loss": 4.0881,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6041600000000003,
+      "grad_norm": 0.5152320265769958,
+      "learning_rate": 0.0002336115224121685,
+      "loss": 4.043,
+      "step": 2036
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 0.5005114078521729,
+      "learning_rate": 0.00023357114012653116,
+      "loss": 3.9851,
+      "step": 2037
+    },
+    {
+      "epoch": 2.60672,
+      "grad_norm": 0.5171644687652588,
+      "learning_rate": 0.00023353075784089377,
+      "loss": 3.9649,
+      "step": 2038
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.5362170338630676,
+      "learning_rate": 0.0002334903755552564,
+      "loss": 4.0048,
+      "step": 2039
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 0.5056260824203491,
+      "learning_rate": 0.00023344999326961903,
+      "loss": 4.056,
+      "step": 2040
+    },
+    {
+      "epoch": 2.61056,
+      "grad_norm": 0.5133628249168396,
+      "learning_rate": 0.00023340961098398168,
+      "loss": 4.0981,
+      "step": 2041
+    },
+    {
+      "epoch": 2.61184,
+      "grad_norm": 0.5374043583869934,
+      "learning_rate": 0.00023336922869834431,
+      "loss": 3.9999,
+      "step": 2042
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 0.4886053800582886,
+      "learning_rate": 0.00023332884641270694,
+      "loss": 4.1064,
+      "step": 2043
+    },
+    {
+      "epoch": 2.6144,
+      "grad_norm": 0.5053584575653076,
+      "learning_rate": 0.00023328846412706957,
+      "loss": 3.9454,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6156800000000002,
+      "grad_norm": 0.4969666004180908,
+      "learning_rate": 0.00023324808184143218,
+      "loss": 4.146,
+      "step": 2045
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 0.5076320767402649,
+      "learning_rate": 0.00023320769955579483,
+      "loss": 4.0381,
+      "step": 2046
+    },
+    {
+      "epoch": 2.61824,
+      "grad_norm": 0.4875280261039734,
+      "learning_rate": 0.00023316731727015746,
+      "loss": 4.0423,
+      "step": 2047
+    },
+    {
+      "epoch": 2.61952,
+      "grad_norm": 0.5112111568450928,
+      "learning_rate": 0.0002331269349845201,
+      "loss": 4.0637,
+      "step": 2048
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.5074511170387268,
+      "learning_rate": 0.00023308655269888272,
+      "loss": 3.9859,
+      "step": 2049
+    },
+    {
+      "epoch": 2.62208,
+      "grad_norm": 0.5409026145935059,
+      "learning_rate": 0.00023304617041324538,
+      "loss": 4.1341,
+      "step": 2050
+    },
+    {
+      "epoch": 2.62336,
+      "grad_norm": 0.5368846654891968,
+      "learning_rate": 0.000233005788127608,
+      "loss": 3.9908,
+      "step": 2051
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 0.5240885615348816,
+      "learning_rate": 0.00023296540584197064,
+      "loss": 4.0777,
+      "step": 2052
+    },
+    {
+      "epoch": 2.62592,
+      "grad_norm": 0.5103102326393127,
+      "learning_rate": 0.00023292502355633325,
+      "loss": 4.0188,
+      "step": 2053
+    },
+    {
+      "epoch": 2.6272,
+      "grad_norm": 0.5250452756881714,
+      "learning_rate": 0.0002328846412706959,
+      "loss": 4.0489,
+      "step": 2054
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 0.5160833597183228,
+      "learning_rate": 0.00023284425898505853,
+      "loss": 4.0773,
+      "step": 2055
+    },
+    {
+      "epoch": 2.62976,
+      "grad_norm": 0.5143999457359314,
+      "learning_rate": 0.00023280387669942116,
+      "loss": 4.0191,
+      "step": 2056
+    },
+    {
+      "epoch": 2.63104,
+      "grad_norm": 0.5295541286468506,
+      "learning_rate": 0.0002327634944137838,
+      "loss": 4.0769,
+      "step": 2057
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 0.5218490958213806,
+      "learning_rate": 0.00023272311212814642,
+      "loss": 3.9913,
+      "step": 2058
+    },
+    {
+      "epoch": 2.6336,
+      "grad_norm": 0.488608717918396,
+      "learning_rate": 0.00023268272984250908,
+      "loss": 4.0126,
+      "step": 2059
+    },
+    {
+      "epoch": 2.63488,
+      "grad_norm": 0.5450034141540527,
+      "learning_rate": 0.0002326423475568717,
+      "loss": 4.0593,
+      "step": 2060
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 0.4820656478404999,
+      "learning_rate": 0.0002326019652712343,
+      "loss": 4.0532,
+      "step": 2061
+    },
+    {
+      "epoch": 2.63744,
+      "grad_norm": 0.5224117636680603,
+      "learning_rate": 0.00023256158298559694,
+      "loss": 4.0053,
+      "step": 2062
+    },
+    {
+      "epoch": 2.63872,
+      "grad_norm": 0.4709533154964447,
+      "learning_rate": 0.0002325212006999596,
+      "loss": 3.9811,
+      "step": 2063
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5281003713607788,
+      "learning_rate": 0.00023248081841432223,
+      "loss": 4.0737,
+      "step": 2064
+    },
+    {
+      "epoch": 2.64128,
+      "grad_norm": 0.49592503905296326,
+      "learning_rate": 0.00023244043612868486,
+      "loss": 4.0537,
+      "step": 2065
+    },
+    {
+      "epoch": 2.64256,
+      "grad_norm": 0.5004292726516724,
+      "learning_rate": 0.0002324000538430475,
+      "loss": 4.0797,
+      "step": 2066
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 0.5057252049446106,
+      "learning_rate": 0.00023235967155741015,
+      "loss": 4.0881,
+      "step": 2067
+    },
+    {
+      "epoch": 2.64512,
+      "grad_norm": 0.511310875415802,
+      "learning_rate": 0.00023231928927177278,
+      "loss": 4.0383,
+      "step": 2068
+    },
+    {
+      "epoch": 2.6464,
+      "grad_norm": 0.5321905016899109,
+      "learning_rate": 0.00023227890698613538,
+      "loss": 4.0814,
+      "step": 2069
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 0.5654548406600952,
+      "learning_rate": 0.000232238524700498,
+      "loss": 4.1224,
+      "step": 2070
+    },
+    {
+      "epoch": 2.6489599999999998,
+      "grad_norm": 0.527220606803894,
+      "learning_rate": 0.00023219814241486067,
+      "loss": 4.0622,
+      "step": 2071
+    },
+    {
+      "epoch": 2.65024,
+      "grad_norm": 0.5580291152000427,
+      "learning_rate": 0.0002321577601292233,
+      "loss": 4.0797,
+      "step": 2072
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 0.545524001121521,
+      "learning_rate": 0.00023211737784358593,
+      "loss": 4.0666,
+      "step": 2073
+    },
+    {
+      "epoch": 2.6528,
+      "grad_norm": 0.5084550380706787,
+      "learning_rate": 0.00023207699555794856,
+      "loss": 4.0051,
+      "step": 2074
+    },
+    {
+      "epoch": 2.65408,
+      "grad_norm": 0.5336723923683167,
+      "learning_rate": 0.0002320366132723112,
+      "loss": 4.0471,
+      "step": 2075
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 0.5522205829620361,
+      "learning_rate": 0.00023199623098667385,
+      "loss": 4.0846,
+      "step": 2076
+    },
+    {
+      "epoch": 2.65664,
+      "grad_norm": 0.5465154051780701,
+      "learning_rate": 0.00023195584870103645,
+      "loss": 3.9559,
+      "step": 2077
+    },
+    {
+      "epoch": 2.65792,
+      "grad_norm": 0.5187036395072937,
+      "learning_rate": 0.00023191546641539908,
+      "loss": 3.9902,
+      "step": 2078
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.5408727526664734,
+      "learning_rate": 0.0002318750841297617,
+      "loss": 4.0133,
+      "step": 2079
+    },
+    {
+      "epoch": 2.6604799999999997,
+      "grad_norm": 0.5757235288619995,
+      "learning_rate": 0.00023183470184412437,
+      "loss": 4.0698,
+      "step": 2080
+    },
+    {
+      "epoch": 2.66176,
+      "grad_norm": 0.5221036076545715,
+      "learning_rate": 0.000231794319558487,
+      "loss": 4.1101,
+      "step": 2081
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 0.5259095430374146,
+      "learning_rate": 0.00023175393727284963,
+      "loss": 3.9804,
+      "step": 2082
+    },
+    {
+      "epoch": 2.66432,
+      "grad_norm": 0.520976185798645,
+      "learning_rate": 0.00023171355498721226,
+      "loss": 4.02,
+      "step": 2083
+    },
+    {
+      "epoch": 2.6656,
+      "grad_norm": 0.5386399626731873,
+      "learning_rate": 0.0002316731727015749,
+      "loss": 4.1377,
+      "step": 2084
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 0.5121170282363892,
+      "learning_rate": 0.00023163279041593754,
+      "loss": 4.0627,
+      "step": 2085
+    },
+    {
+      "epoch": 2.66816,
+      "grad_norm": 0.5606420636177063,
+      "learning_rate": 0.00023159240813030015,
+      "loss": 4.0281,
+      "step": 2086
+    },
+    {
+      "epoch": 2.66944,
+      "grad_norm": 0.5451529622077942,
+      "learning_rate": 0.00023155202584466278,
+      "loss": 3.9724,
+      "step": 2087
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 0.5385946035385132,
+      "learning_rate": 0.0002315116435590254,
+      "loss": 4.0462,
+      "step": 2088
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.5405113101005554,
+      "learning_rate": 0.00023147126127338806,
+      "loss": 3.9622,
+      "step": 2089
+    },
+    {
+      "epoch": 2.67328,
+      "grad_norm": 0.5446247458457947,
+      "learning_rate": 0.0002314308789877507,
+      "loss": 4.0747,
+      "step": 2090
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 0.5345394611358643,
+      "learning_rate": 0.00023139049670211332,
+      "loss": 3.9045,
+      "step": 2091
+    },
+    {
+      "epoch": 2.67584,
+      "grad_norm": 0.5287266969680786,
+      "learning_rate": 0.00023135011441647595,
+      "loss": 4.0296,
+      "step": 2092
+    },
+    {
+      "epoch": 2.67712,
+      "grad_norm": 0.4985159933567047,
+      "learning_rate": 0.0002313097321308386,
+      "loss": 4.0478,
+      "step": 2093
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.5355061292648315,
+      "learning_rate": 0.00023126934984520121,
+      "loss": 4.0843,
+      "step": 2094
+    },
+    {
+      "epoch": 2.67968,
+      "grad_norm": 0.4830259382724762,
+      "learning_rate": 0.00023122896755956384,
+      "loss": 4.0583,
+      "step": 2095
+    },
+    {
+      "epoch": 2.68096,
+      "grad_norm": 0.5162714719772339,
+      "learning_rate": 0.00023118858527392647,
+      "loss": 3.967,
+      "step": 2096
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 0.5130695700645447,
+      "learning_rate": 0.00023114820298828913,
+      "loss": 4.0423,
+      "step": 2097
+    },
+    {
+      "epoch": 2.68352,
+      "grad_norm": 0.507447361946106,
+      "learning_rate": 0.00023110782070265176,
+      "loss": 4.0212,
+      "step": 2098
+    },
+    {
+      "epoch": 2.6848,
+      "grad_norm": 0.6723561882972717,
+      "learning_rate": 0.0002310674384170144,
+      "loss": 4.0209,
+      "step": 2099
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 0.49744725227355957,
+      "learning_rate": 0.00023102705613137702,
+      "loss": 3.9865,
+      "step": 2100
+    },
+    {
+      "epoch": 2.68736,
+      "grad_norm": 0.5632631778717041,
+      "learning_rate": 0.00023098667384573962,
+      "loss": 4.0426,
+      "step": 2101
+    },
+    {
+      "epoch": 2.68864,
+      "grad_norm": 0.5552526712417603,
+      "learning_rate": 0.00023094629156010228,
+      "loss": 4.0527,
+      "step": 2102
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 0.5476700663566589,
+      "learning_rate": 0.0002309059092744649,
+      "loss": 4.0647,
+      "step": 2103
+    },
+    {
+      "epoch": 2.6912000000000003,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.00023086552698882754,
+      "loss": 4.0247,
+      "step": 2104
+    },
+    {
+      "epoch": 2.6924799999999998,
+      "grad_norm": 0.5172504186630249,
+      "learning_rate": 0.00023082514470319017,
+      "loss": 4.0629,
+      "step": 2105
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 0.5142618417739868,
+      "learning_rate": 0.00023078476241755283,
+      "loss": 3.9725,
+      "step": 2106
+    },
+    {
+      "epoch": 2.69504,
+      "grad_norm": 0.49800601601600647,
+      "learning_rate": 0.00023074438013191546,
+      "loss": 3.9891,
+      "step": 2107
+    },
+    {
+      "epoch": 2.69632,
+      "grad_norm": 0.5199533700942993,
+      "learning_rate": 0.0002307039978462781,
+      "loss": 4.0479,
+      "step": 2108
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.4997861981391907,
+      "learning_rate": 0.0002306636155606407,
+      "loss": 3.9831,
+      "step": 2109
+    },
+    {
+      "epoch": 2.69888,
+      "grad_norm": 0.5149181485176086,
+      "learning_rate": 0.00023062323327500335,
+      "loss": 4.0014,
+      "step": 2110
+    },
+    {
+      "epoch": 2.70016,
+      "grad_norm": 0.5101584792137146,
+      "learning_rate": 0.00023058285098936598,
+      "loss": 4.0213,
+      "step": 2111
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 0.4832100570201874,
+      "learning_rate": 0.0002305424687037286,
+      "loss": 4.0012,
+      "step": 2112
+    },
+    {
+      "epoch": 2.7027200000000002,
+      "grad_norm": 0.527621328830719,
+      "learning_rate": 0.00023050208641809124,
+      "loss": 4.0418,
+      "step": 2113
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.4823199510574341,
+      "learning_rate": 0.00023046170413245387,
+      "loss": 4.0175,
+      "step": 2114
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 0.5674161911010742,
+      "learning_rate": 0.00023042132184681653,
+      "loss": 4.0703,
+      "step": 2115
+    },
+    {
+      "epoch": 2.70656,
+      "grad_norm": 0.5498639345169067,
+      "learning_rate": 0.00023038093956117916,
+      "loss": 4.0389,
+      "step": 2116
+    },
+    {
+      "epoch": 2.70784,
+      "grad_norm": 0.5504740476608276,
+      "learning_rate": 0.00023034055727554176,
+      "loss": 4.0138,
+      "step": 2117
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 0.5479692816734314,
+      "learning_rate": 0.0002303001749899044,
+      "loss": 4.0722,
+      "step": 2118
+    },
+    {
+      "epoch": 2.7104,
+      "grad_norm": 0.4745061993598938,
+      "learning_rate": 0.00023025979270426705,
+      "loss": 4.0672,
+      "step": 2119
+    },
+    {
+      "epoch": 2.71168,
+      "grad_norm": 0.5192096829414368,
+      "learning_rate": 0.00023021941041862968,
+      "loss": 3.9691,
+      "step": 2120
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 0.5102022290229797,
+      "learning_rate": 0.0002301790281329923,
+      "loss": 4.0599,
+      "step": 2121
+    },
+    {
+      "epoch": 2.71424,
+      "grad_norm": 0.49399974942207336,
+      "learning_rate": 0.00023013864584735494,
+      "loss": 3.9948,
+      "step": 2122
+    },
+    {
+      "epoch": 2.71552,
+      "grad_norm": 0.5214704871177673,
+      "learning_rate": 0.0002300982635617176,
+      "loss": 4.0376,
+      "step": 2123
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.5198361873626709,
+      "learning_rate": 0.00023005788127608022,
+      "loss": 4.0227,
+      "step": 2124
+    },
+    {
+      "epoch": 2.71808,
+      "grad_norm": 0.5009457468986511,
+      "learning_rate": 0.00023001749899044283,
+      "loss": 4.0152,
+      "step": 2125
+    },
+    {
+      "epoch": 2.71936,
+      "grad_norm": 0.5245611667633057,
+      "learning_rate": 0.00022997711670480546,
+      "loss": 4.056,
+      "step": 2126
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 0.4951135814189911,
+      "learning_rate": 0.0002299367344191681,
+      "loss": 4.0599,
+      "step": 2127
+    },
+    {
+      "epoch": 2.72192,
+      "grad_norm": 0.5234236717224121,
+      "learning_rate": 0.00022989635213353074,
+      "loss": 4.0412,
+      "step": 2128
+    },
+    {
+      "epoch": 2.7232,
+      "grad_norm": 0.5335028171539307,
+      "learning_rate": 0.00022985596984789337,
+      "loss": 3.9789,
+      "step": 2129
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 0.527600109577179,
+      "learning_rate": 0.000229815587562256,
+      "loss": 4.0591,
+      "step": 2130
+    },
+    {
+      "epoch": 2.72576,
+      "grad_norm": 0.5804891586303711,
+      "learning_rate": 0.00022977520527661864,
+      "loss": 4.029,
+      "step": 2131
+    },
+    {
+      "epoch": 2.72704,
+      "grad_norm": 0.5268876552581787,
+      "learning_rate": 0.0002297348229909813,
+      "loss": 3.9655,
+      "step": 2132
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 0.5200521349906921,
+      "learning_rate": 0.0002296944407053439,
+      "loss": 3.9752,
+      "step": 2133
+    },
+    {
+      "epoch": 2.7296,
+      "grad_norm": 0.5403656363487244,
+      "learning_rate": 0.00022965405841970653,
+      "loss": 4.0228,
+      "step": 2134
+    },
+    {
+      "epoch": 2.73088,
+      "grad_norm": 0.528051495552063,
+      "learning_rate": 0.00022961367613406916,
+      "loss": 3.9931,
+      "step": 2135
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 0.5134397745132446,
+      "learning_rate": 0.0002295732938484318,
+      "loss": 4.0045,
+      "step": 2136
+    },
+    {
+      "epoch": 2.73344,
+      "grad_norm": 0.5310769081115723,
+      "learning_rate": 0.00022953291156279444,
+      "loss": 4.0,
+      "step": 2137
+    },
+    {
+      "epoch": 2.7347200000000003,
+      "grad_norm": 0.5168265700340271,
+      "learning_rate": 0.00022949252927715707,
+      "loss": 4.0438,
+      "step": 2138
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.5118173360824585,
+      "learning_rate": 0.0002294521469915197,
+      "loss": 4.0294,
+      "step": 2139
+    },
+    {
+      "epoch": 2.73728,
+      "grad_norm": 0.5167641043663025,
+      "learning_rate": 0.0002294117647058823,
+      "loss": 3.9984,
+      "step": 2140
+    },
+    {
+      "epoch": 2.73856,
+      "grad_norm": 0.48330435156822205,
+      "learning_rate": 0.00022937138242024496,
+      "loss": 4.076,
+      "step": 2141
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 0.5215317606925964,
+      "learning_rate": 0.0002293310001346076,
+      "loss": 4.0358,
+      "step": 2142
+    },
+    {
+      "epoch": 2.74112,
+      "grad_norm": 0.5229592323303223,
+      "learning_rate": 0.00022929061784897022,
+      "loss": 3.9641,
+      "step": 2143
+    },
+    {
+      "epoch": 2.7424,
+      "grad_norm": 0.49107488989830017,
+      "learning_rate": 0.00022925023556333285,
+      "loss": 3.9869,
+      "step": 2144
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 0.5166721343994141,
+      "learning_rate": 0.0002292098532776955,
+      "loss": 3.9876,
+      "step": 2145
+    },
+    {
+      "epoch": 2.74496,
+      "grad_norm": 0.5030099749565125,
+      "learning_rate": 0.00022916947099205814,
+      "loss": 3.9987,
+      "step": 2146
+    },
+    {
+      "epoch": 2.7462400000000002,
+      "grad_norm": 0.5043811202049255,
+      "learning_rate": 0.00022912908870642077,
+      "loss": 4.0329,
+      "step": 2147
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 0.5001031756401062,
+      "learning_rate": 0.00022908870642078337,
+      "loss": 4.0079,
+      "step": 2148
+    },
+    {
+      "epoch": 2.7488,
+      "grad_norm": 0.5411626100540161,
+      "learning_rate": 0.00022904832413514603,
+      "loss": 4.0499,
+      "step": 2149
+    },
+    {
+      "epoch": 2.75008,
+      "grad_norm": 0.529230535030365,
+      "learning_rate": 0.00022900794184950866,
+      "loss": 4.0719,
+      "step": 2150
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 0.5309909582138062,
+      "learning_rate": 0.0002289675595638713,
+      "loss": 3.9798,
+      "step": 2151
+    },
+    {
+      "epoch": 2.75264,
+      "grad_norm": 0.5642079710960388,
+      "learning_rate": 0.00022892717727823392,
+      "loss": 4.0106,
+      "step": 2152
+    },
+    {
+      "epoch": 2.75392,
+      "grad_norm": 0.5178594589233398,
+      "learning_rate": 0.00022888679499259655,
+      "loss": 4.057,
+      "step": 2153
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.5369943380355835,
+      "learning_rate": 0.0002288464127069592,
+      "loss": 3.996,
+      "step": 2154
+    },
+    {
+      "epoch": 2.75648,
+      "grad_norm": 0.5157840251922607,
+      "learning_rate": 0.00022880603042132184,
+      "loss": 4.038,
+      "step": 2155
+    },
+    {
+      "epoch": 2.75776,
+      "grad_norm": 0.5400508642196655,
+      "learning_rate": 0.00022876564813568444,
+      "loss": 3.9986,
+      "step": 2156
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 0.49471211433410645,
+      "learning_rate": 0.00022872526585004707,
+      "loss": 4.0016,
+      "step": 2157
+    },
+    {
+      "epoch": 2.76032,
+      "grad_norm": 0.5369953513145447,
+      "learning_rate": 0.00022868488356440973,
+      "loss": 4.0411,
+      "step": 2158
+    },
+    {
+      "epoch": 2.7616,
+      "grad_norm": 0.5281562209129333,
+      "learning_rate": 0.00022864450127877236,
+      "loss": 4.0303,
+      "step": 2159
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 0.5228774547576904,
+      "learning_rate": 0.000228604118993135,
+      "loss": 4.0643,
+      "step": 2160
+    },
+    {
+      "epoch": 2.76416,
+      "grad_norm": 0.5000044107437134,
+      "learning_rate": 0.00022856373670749762,
+      "loss": 4.0538,
+      "step": 2161
+    },
+    {
+      "epoch": 2.76544,
+      "grad_norm": 0.505082905292511,
+      "learning_rate": 0.00022852335442186028,
+      "loss": 4.0106,
+      "step": 2162
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 0.5144949555397034,
+      "learning_rate": 0.0002284829721362229,
+      "loss": 3.9746,
+      "step": 2163
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 0.5088104605674744,
+      "learning_rate": 0.0002284425898505855,
+      "loss": 3.9718,
+      "step": 2164
+    },
+    {
+      "epoch": 2.76928,
+      "grad_norm": 0.5328196287155151,
+      "learning_rate": 0.00022840220756494814,
+      "loss": 4.0054,
+      "step": 2165
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 0.5345327854156494,
+      "learning_rate": 0.00022836182527931077,
+      "loss": 4.0231,
+      "step": 2166
+    },
+    {
+      "epoch": 2.77184,
+      "grad_norm": 0.49523457884788513,
+      "learning_rate": 0.00022832144299367343,
+      "loss": 3.9831,
+      "step": 2167
+    },
+    {
+      "epoch": 2.77312,
+      "grad_norm": 0.5508283376693726,
+      "learning_rate": 0.00022828106070803606,
+      "loss": 3.9685,
+      "step": 2168
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.5366639494895935,
+      "learning_rate": 0.00022824067842239869,
+      "loss": 4.0006,
+      "step": 2169
+    },
+    {
+      "epoch": 2.77568,
+      "grad_norm": 0.5522692799568176,
+      "learning_rate": 0.00022820029613676132,
+      "loss": 3.9667,
+      "step": 2170
+    },
+    {
+      "epoch": 2.77696,
+      "grad_norm": 0.4968757927417755,
+      "learning_rate": 0.00022815991385112397,
+      "loss": 4.0107,
+      "step": 2171
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 0.5238058567047119,
+      "learning_rate": 0.0002281195315654866,
+      "loss": 3.9716,
+      "step": 2172
+    },
+    {
+      "epoch": 2.7795199999999998,
+      "grad_norm": 0.5009492039680481,
+      "learning_rate": 0.0002280791492798492,
+      "loss": 4.0209,
+      "step": 2173
+    },
+    {
+      "epoch": 2.7808,
+      "grad_norm": 0.5323073863983154,
+      "learning_rate": 0.00022803876699421184,
+      "loss": 4.0487,
+      "step": 2174
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 0.5168647170066833,
+      "learning_rate": 0.0002279983847085745,
+      "loss": 4.0385,
+      "step": 2175
+    },
+    {
+      "epoch": 2.78336,
+      "grad_norm": 0.5620954632759094,
+      "learning_rate": 0.00022795800242293712,
+      "loss": 4.0084,
+      "step": 2176
+    },
+    {
+      "epoch": 2.78464,
+      "grad_norm": 0.5150801539421082,
+      "learning_rate": 0.00022791762013729975,
+      "loss": 3.9903,
+      "step": 2177
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 0.5225955843925476,
+      "learning_rate": 0.00022787723785166238,
+      "loss": 3.9493,
+      "step": 2178
+    },
+    {
+      "epoch": 2.7872,
+      "grad_norm": 0.5136468410491943,
+      "learning_rate": 0.000227836855566025,
+      "loss": 4.0297,
+      "step": 2179
+    },
+    {
+      "epoch": 2.78848,
+      "grad_norm": 0.5070092082023621,
+      "learning_rate": 0.00022779647328038767,
+      "loss": 3.9916,
+      "step": 2180
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 0.48506057262420654,
+      "learning_rate": 0.00022775609099475027,
+      "loss": 4.049,
+      "step": 2181
+    },
+    {
+      "epoch": 2.7910399999999997,
+      "grad_norm": 0.5056405067443848,
+      "learning_rate": 0.0002277157087091129,
+      "loss": 4.0344,
+      "step": 2182
+    },
+    {
+      "epoch": 2.79232,
+      "grad_norm": 0.5065452456474304,
+      "learning_rate": 0.00022767532642347553,
+      "loss": 4.0575,
+      "step": 2183
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.4931611716747284,
+      "learning_rate": 0.0002276349441378382,
+      "loss": 3.9662,
+      "step": 2184
+    },
+    {
+      "epoch": 2.79488,
+      "grad_norm": 0.5401006937026978,
+      "learning_rate": 0.00022759456185220082,
+      "loss": 3.9968,
+      "step": 2185
+    },
+    {
+      "epoch": 2.79616,
+      "grad_norm": 0.48308640718460083,
+      "learning_rate": 0.00022755417956656345,
+      "loss": 3.9867,
+      "step": 2186
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 0.5414093732833862,
+      "learning_rate": 0.00022751379728092608,
+      "loss": 3.938,
+      "step": 2187
+    },
+    {
+      "epoch": 2.79872,
+      "grad_norm": 0.47990182042121887,
+      "learning_rate": 0.00022747341499528874,
+      "loss": 4.1135,
+      "step": 2188
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.5447584986686707,
+      "learning_rate": 0.00022743303270965134,
+      "loss": 4.1674,
+      "step": 2189
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 0.4829598665237427,
+      "learning_rate": 0.00022739265042401397,
+      "loss": 4.0011,
+      "step": 2190
+    },
+    {
+      "epoch": 2.80256,
+      "grad_norm": 0.5174542665481567,
+      "learning_rate": 0.0002273522681383766,
+      "loss": 4.0109,
+      "step": 2191
+    },
+    {
+      "epoch": 2.80384,
+      "grad_norm": 0.5266432166099548,
+      "learning_rate": 0.00022731188585273926,
+      "loss": 3.9463,
+      "step": 2192
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 0.5114743709564209,
+      "learning_rate": 0.0002272715035671019,
+      "loss": 4.0119,
+      "step": 2193
+    },
+    {
+      "epoch": 2.8064,
+      "grad_norm": 0.5330787301063538,
+      "learning_rate": 0.00022723112128146452,
+      "loss": 4.0184,
+      "step": 2194
+    },
+    {
+      "epoch": 2.80768,
+      "grad_norm": 0.49550044536590576,
+      "learning_rate": 0.00022719073899582715,
+      "loss": 4.0103,
+      "step": 2195
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 0.5248696804046631,
+      "learning_rate": 0.00022715035671018975,
+      "loss": 4.0211,
+      "step": 2196
+    },
+    {
+      "epoch": 2.81024,
+      "grad_norm": 0.528124213218689,
+      "learning_rate": 0.0002271099744245524,
+      "loss": 3.9816,
+      "step": 2197
+    },
+    {
+      "epoch": 2.81152,
+      "grad_norm": 0.498807817697525,
+      "learning_rate": 0.00022706959213891504,
+      "loss": 3.9876,
+      "step": 2198
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.5002254247665405,
+      "learning_rate": 0.00022702920985327767,
+      "loss": 3.9442,
+      "step": 2199
+    },
+    {
+      "epoch": 2.81408,
+      "grad_norm": 0.5136492848396301,
+      "learning_rate": 0.0002269888275676403,
+      "loss": 4.0369,
+      "step": 2200
+    },
+    {
+      "epoch": 2.81536,
+      "grad_norm": 0.5032553672790527,
+      "learning_rate": 0.00022694844528200296,
+      "loss": 4.0838,
+      "step": 2201
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 0.4921877086162567,
+      "learning_rate": 0.0002269080629963656,
+      "loss": 4.0399,
+      "step": 2202
+    },
+    {
+      "epoch": 2.81792,
+      "grad_norm": 0.48802658915519714,
+      "learning_rate": 0.00022686768071072822,
+      "loss": 3.9631,
+      "step": 2203
+    },
+    {
+      "epoch": 2.8192,
+      "grad_norm": 0.5272952914237976,
+      "learning_rate": 0.00022682729842509082,
+      "loss": 4.0163,
+      "step": 2204
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 0.5049785375595093,
+      "learning_rate": 0.00022678691613945348,
+      "loss": 3.9296,
+      "step": 2205
+    },
+    {
+      "epoch": 2.8217600000000003,
+      "grad_norm": 0.512064516544342,
+      "learning_rate": 0.0002267465338538161,
+      "loss": 3.9826,
+      "step": 2206
+    },
+    {
+      "epoch": 2.8230399999999998,
+      "grad_norm": 0.49538886547088623,
+      "learning_rate": 0.00022670615156817874,
+      "loss": 3.9296,
+      "step": 2207
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 0.5314378142356873,
+      "learning_rate": 0.00022666576928254137,
+      "loss": 4.0114,
+      "step": 2208
+    },
+    {
+      "epoch": 2.8256,
+      "grad_norm": 0.5182934403419495,
+      "learning_rate": 0.000226625386996904,
+      "loss": 4.0499,
+      "step": 2209
+    },
+    {
+      "epoch": 2.82688,
+      "grad_norm": 0.5282546877861023,
+      "learning_rate": 0.00022658500471126665,
+      "loss": 4.0127,
+      "step": 2210
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 0.49509358406066895,
+      "learning_rate": 0.00022654462242562929,
+      "loss": 3.9841,
+      "step": 2211
+    },
+    {
+      "epoch": 2.82944,
+      "grad_norm": 0.5452840328216553,
+      "learning_rate": 0.0002265042401399919,
+      "loss": 4.0112,
+      "step": 2212
+    },
+    {
+      "epoch": 2.83072,
+      "grad_norm": 0.5101568102836609,
+      "learning_rate": 0.00022646385785435452,
+      "loss": 3.9766,
+      "step": 2213
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.5204622149467468,
+      "learning_rate": 0.00022642347556871718,
+      "loss": 4.0391,
+      "step": 2214
+    },
+    {
+      "epoch": 2.8332800000000002,
+      "grad_norm": 0.49549055099487305,
+      "learning_rate": 0.0002263830932830798,
+      "loss": 3.9876,
+      "step": 2215
+    },
+    {
+      "epoch": 2.8345599999999997,
+      "grad_norm": 0.5057578682899475,
+      "learning_rate": 0.00022634271099744244,
+      "loss": 4.0252,
+      "step": 2216
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 0.5117456912994385,
+      "learning_rate": 0.00022630232871180507,
+      "loss": 4.0438,
+      "step": 2217
+    },
+    {
+      "epoch": 2.83712,
+      "grad_norm": 0.49402594566345215,
+      "learning_rate": 0.00022626194642616772,
+      "loss": 4.0064,
+      "step": 2218
+    },
+    {
+      "epoch": 2.8384,
+      "grad_norm": 0.49484384059906006,
+      "learning_rate": 0.00022622156414053035,
+      "loss": 3.9281,
+      "step": 2219
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 0.49207979440689087,
+      "learning_rate": 0.00022618118185489296,
+      "loss": 4.0203,
+      "step": 2220
+    },
+    {
+      "epoch": 2.84096,
+      "grad_norm": 0.5029460191726685,
+      "learning_rate": 0.00022614079956925559,
+      "loss": 4.0214,
+      "step": 2221
+    },
+    {
+      "epoch": 2.84224,
+      "grad_norm": 0.49335840344429016,
+      "learning_rate": 0.00022610041728361822,
+      "loss": 4.0127,
+      "step": 2222
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 0.5236724019050598,
+      "learning_rate": 0.00022606003499798087,
+      "loss": 4.0602,
+      "step": 2223
+    },
+    {
+      "epoch": 2.8448,
+      "grad_norm": 0.5051988363265991,
+      "learning_rate": 0.0002260196527123435,
+      "loss": 3.9739,
+      "step": 2224
+    },
+    {
+      "epoch": 2.84608,
+      "grad_norm": 0.48310649394989014,
+      "learning_rate": 0.00022597927042670613,
+      "loss": 4.0336,
+      "step": 2225
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 0.49196314811706543,
+      "learning_rate": 0.00022593888814106876,
+      "loss": 4.068,
+      "step": 2226
+    },
+    {
+      "epoch": 2.84864,
+      "grad_norm": 0.48503825068473816,
+      "learning_rate": 0.00022589850585543142,
+      "loss": 4.0305,
+      "step": 2227
+    },
+    {
+      "epoch": 2.84992,
+      "grad_norm": 0.5028344988822937,
+      "learning_rate": 0.00022585812356979402,
+      "loss": 3.9993,
+      "step": 2228
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.4741714298725128,
+      "learning_rate": 0.00022581774128415665,
+      "loss": 4.0975,
+      "step": 2229
+    },
+    {
+      "epoch": 2.85248,
+      "grad_norm": 0.5247095227241516,
+      "learning_rate": 0.00022577735899851928,
+      "loss": 4.0058,
+      "step": 2230
+    },
+    {
+      "epoch": 2.85376,
+      "grad_norm": 0.5030754804611206,
+      "learning_rate": 0.00022573697671288194,
+      "loss": 3.968,
+      "step": 2231
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 0.5085668563842773,
+      "learning_rate": 0.00022569659442724457,
+      "loss": 4.0359,
+      "step": 2232
+    },
+    {
+      "epoch": 2.85632,
+      "grad_norm": 0.5207344889640808,
+      "learning_rate": 0.0002256562121416072,
+      "loss": 3.9844,
+      "step": 2233
+    },
+    {
+      "epoch": 2.8576,
+      "grad_norm": 0.5112427473068237,
+      "learning_rate": 0.00022561582985596983,
+      "loss": 4.0718,
+      "step": 2234
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 0.49677979946136475,
+      "learning_rate": 0.00022557544757033243,
+      "loss": 3.9815,
+      "step": 2235
+    },
+    {
+      "epoch": 2.86016,
+      "grad_norm": 0.5005208253860474,
+      "learning_rate": 0.0002255350652846951,
+      "loss": 4.0531,
+      "step": 2236
+    },
+    {
+      "epoch": 2.86144,
+      "grad_norm": 0.4932452142238617,
+      "learning_rate": 0.00022549468299905772,
+      "loss": 4.0379,
+      "step": 2237
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 0.510642945766449,
+      "learning_rate": 0.00022545430071342035,
+      "loss": 4.0408,
+      "step": 2238
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.49646908044815063,
+      "learning_rate": 0.00022541391842778298,
+      "loss": 4.0044,
+      "step": 2239
+    },
+    {
+      "epoch": 2.8652800000000003,
+      "grad_norm": 0.5001757144927979,
+      "learning_rate": 0.00022537353614214564,
+      "loss": 4.0446,
+      "step": 2240
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 0.5238614678382874,
+      "learning_rate": 0.00022533315385650827,
+      "loss": 3.9808,
+      "step": 2241
+    },
+    {
+      "epoch": 2.86784,
+      "grad_norm": 0.5164071917533875,
+      "learning_rate": 0.0002252927715708709,
+      "loss": 4.0207,
+      "step": 2242
+    },
+    {
+      "epoch": 2.86912,
+      "grad_norm": 0.5213832855224609,
+      "learning_rate": 0.0002252523892852335,
+      "loss": 4.0449,
+      "step": 2243
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.4973851144313812,
+      "learning_rate": 0.00022521200699959619,
+      "loss": 3.9746,
+      "step": 2244
+    },
+    {
+      "epoch": 2.87168,
+      "grad_norm": 0.479270875453949,
+      "learning_rate": 0.0002251716247139588,
+      "loss": 3.9801,
+      "step": 2245
+    },
+    {
+      "epoch": 2.87296,
+      "grad_norm": 0.48888400197029114,
+      "learning_rate": 0.00022513124242832142,
+      "loss": 4.0178,
+      "step": 2246
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 0.48972752690315247,
+      "learning_rate": 0.00022509086014268405,
+      "loss": 4.0051,
+      "step": 2247
+    },
+    {
+      "epoch": 2.87552,
+      "grad_norm": 0.4787149429321289,
+      "learning_rate": 0.00022505047785704668,
+      "loss": 3.8699,
+      "step": 2248
+    },
+    {
+      "epoch": 2.8768000000000002,
+      "grad_norm": 0.49900394678115845,
+      "learning_rate": 0.00022501009557140934,
+      "loss": 3.9537,
+      "step": 2249
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 0.48665642738342285,
+      "learning_rate": 0.00022496971328577197,
+      "loss": 3.9841,
+      "step": 2250
+    },
+    {
+      "epoch": 2.87936,
+      "grad_norm": 0.48270756006240845,
+      "learning_rate": 0.00022492933100013457,
+      "loss": 4.0203,
+      "step": 2251
+    },
+    {
+      "epoch": 2.88064,
+      "grad_norm": 0.48800358176231384,
+      "learning_rate": 0.0002248889487144972,
+      "loss": 4.0187,
+      "step": 2252
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 0.5024043321609497,
+      "learning_rate": 0.00022484856642885986,
+      "loss": 3.968,
+      "step": 2253
+    },
+    {
+      "epoch": 2.8832,
+      "grad_norm": 0.5062508583068848,
+      "learning_rate": 0.0002248081841432225,
+      "loss": 4.0056,
+      "step": 2254
+    },
+    {
+      "epoch": 2.88448,
+      "grad_norm": 0.4831932485103607,
+      "learning_rate": 0.00022476780185758512,
+      "loss": 4.0342,
+      "step": 2255
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 0.4784030020236969,
+      "learning_rate": 0.00022472741957194775,
+      "loss": 4.0064,
+      "step": 2256
+    },
+    {
+      "epoch": 2.88704,
+      "grad_norm": 0.49033915996551514,
+      "learning_rate": 0.0002246870372863104,
+      "loss": 4.0344,
+      "step": 2257
+    },
+    {
+      "epoch": 2.88832,
+      "grad_norm": 0.505535364151001,
+      "learning_rate": 0.00022464665500067303,
+      "loss": 3.9952,
+      "step": 2258
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.5267712473869324,
+      "learning_rate": 0.00022460627271503566,
+      "loss": 3.9862,
+      "step": 2259
+    },
+    {
+      "epoch": 2.89088,
+      "grad_norm": 0.5031368136405945,
+      "learning_rate": 0.00022456589042939827,
+      "loss": 4.0297,
+      "step": 2260
+    },
+    {
+      "epoch": 2.89216,
+      "grad_norm": 0.5366697907447815,
+      "learning_rate": 0.0002245255081437609,
+      "loss": 4.037,
+      "step": 2261
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 0.5078505873680115,
+      "learning_rate": 0.00022448512585812355,
+      "loss": 3.9513,
+      "step": 2262
+    },
+    {
+      "epoch": 2.89472,
+      "grad_norm": 0.5012052655220032,
+      "learning_rate": 0.00022444474357248618,
+      "loss": 3.9266,
+      "step": 2263
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 0.5153883695602417,
+      "learning_rate": 0.00022440436128684881,
+      "loss": 3.9948,
+      "step": 2264
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 0.5152749419212341,
+      "learning_rate": 0.00022436397900121144,
+      "loss": 3.9879,
+      "step": 2265
+    },
+    {
+      "epoch": 2.89856,
+      "grad_norm": 0.5094638466835022,
+      "learning_rate": 0.0002243235967155741,
+      "loss": 3.9949,
+      "step": 2266
+    },
+    {
+      "epoch": 2.89984,
+      "grad_norm": 0.489305317401886,
+      "learning_rate": 0.00022428321442993673,
+      "loss": 3.9818,
+      "step": 2267
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 0.4954513907432556,
+      "learning_rate": 0.00022424283214429933,
+      "loss": 3.9221,
+      "step": 2268
+    },
+    {
+      "epoch": 2.9024,
+      "grad_norm": 0.5154023170471191,
+      "learning_rate": 0.00022420244985866196,
+      "loss": 4.0049,
+      "step": 2269
+    },
+    {
+      "epoch": 2.90368,
+      "grad_norm": 0.5139758586883545,
+      "learning_rate": 0.00022416206757302462,
+      "loss": 3.9232,
+      "step": 2270
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 0.5243357419967651,
+      "learning_rate": 0.00022412168528738725,
+      "loss": 4.0215,
+      "step": 2271
+    },
+    {
+      "epoch": 2.90624,
+      "grad_norm": 0.5350946187973022,
+      "learning_rate": 0.00022408130300174988,
+      "loss": 3.9502,
+      "step": 2272
+    },
+    {
+      "epoch": 2.90752,
+      "grad_norm": 0.5469184517860413,
+      "learning_rate": 0.0002240409207161125,
+      "loss": 3.9617,
+      "step": 2273
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.5040557980537415,
+      "learning_rate": 0.00022400053843047514,
+      "loss": 3.9336,
+      "step": 2274
+    },
+    {
+      "epoch": 2.91008,
+      "grad_norm": 0.5210381150245667,
+      "learning_rate": 0.0002239601561448378,
+      "loss": 3.9707,
+      "step": 2275
+    },
+    {
+      "epoch": 2.91136,
+      "grad_norm": 0.5240786075592041,
+      "learning_rate": 0.0002239197738592004,
+      "loss": 3.9758,
+      "step": 2276
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 0.5405933260917664,
+      "learning_rate": 0.00022387939157356303,
+      "loss": 4.036,
+      "step": 2277
+    },
+    {
+      "epoch": 2.91392,
+      "grad_norm": 0.5046157240867615,
+      "learning_rate": 0.00022383900928792566,
+      "loss": 3.9727,
+      "step": 2278
+    },
+    {
+      "epoch": 2.9152,
+      "grad_norm": 0.49960970878601074,
+      "learning_rate": 0.00022379862700228832,
+      "loss": 4.0635,
+      "step": 2279
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 0.514678418636322,
+      "learning_rate": 0.00022375824471665095,
+      "loss": 4.01,
+      "step": 2280
+    },
+    {
+      "epoch": 2.91776,
+      "grad_norm": 0.5296341180801392,
+      "learning_rate": 0.00022371786243101358,
+      "loss": 3.9588,
+      "step": 2281
+    },
+    {
+      "epoch": 2.91904,
+      "grad_norm": 0.49709784984588623,
+      "learning_rate": 0.0002236774801453762,
+      "loss": 3.9068,
+      "step": 2282
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 0.5464451909065247,
+      "learning_rate": 0.00022363709785973887,
+      "loss": 4.0025,
+      "step": 2283
+    },
+    {
+      "epoch": 2.9215999999999998,
+      "grad_norm": 0.5068522095680237,
+      "learning_rate": 0.00022359671557410147,
+      "loss": 3.9793,
+      "step": 2284
+    },
+    {
+      "epoch": 2.92288,
+      "grad_norm": 0.5234406590461731,
+      "learning_rate": 0.0002235563332884641,
+      "loss": 3.9384,
+      "step": 2285
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 0.5124624967575073,
+      "learning_rate": 0.00022351595100282673,
+      "loss": 3.9735,
+      "step": 2286
+    },
+    {
+      "epoch": 2.92544,
+      "grad_norm": 0.5246328115463257,
+      "learning_rate": 0.00022347556871718936,
+      "loss": 4.0739,
+      "step": 2287
+    },
+    {
+      "epoch": 2.92672,
+      "grad_norm": 0.5275157690048218,
+      "learning_rate": 0.00022343518643155202,
+      "loss": 3.9339,
+      "step": 2288
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.5206854939460754,
+      "learning_rate": 0.00022339480414591465,
+      "loss": 3.9908,
+      "step": 2289
+    },
+    {
+      "epoch": 2.92928,
+      "grad_norm": 0.50496506690979,
+      "learning_rate": 0.00022335442186027728,
+      "loss": 4.0184,
+      "step": 2290
+    },
+    {
+      "epoch": 2.93056,
+      "grad_norm": 0.5281472206115723,
+      "learning_rate": 0.00022331403957463988,
+      "loss": 4.0205,
+      "step": 2291
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 0.5035421252250671,
+      "learning_rate": 0.00022327365728900254,
+      "loss": 4.016,
+      "step": 2292
+    },
+    {
+      "epoch": 2.9331199999999997,
+      "grad_norm": 0.5101969242095947,
+      "learning_rate": 0.00022323327500336517,
+      "loss": 3.9902,
+      "step": 2293
+    },
+    {
+      "epoch": 2.9344,
+      "grad_norm": 0.49704429507255554,
+      "learning_rate": 0.0002231928927177278,
+      "loss": 3.9911,
+      "step": 2294
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 0.5134301781654358,
+      "learning_rate": 0.00022315251043209043,
+      "loss": 3.9861,
+      "step": 2295
+    },
+    {
+      "epoch": 2.93696,
+      "grad_norm": 0.5616884231567383,
+      "learning_rate": 0.00022311212814645309,
+      "loss": 4.0493,
+      "step": 2296
+    },
+    {
+      "epoch": 2.93824,
+      "grad_norm": 0.4961213767528534,
+      "learning_rate": 0.00022307174586081572,
+      "loss": 3.9555,
+      "step": 2297
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 0.5437341928482056,
+      "learning_rate": 0.00022303136357517835,
+      "loss": 3.9674,
+      "step": 2298
+    },
+    {
+      "epoch": 2.9408,
+      "grad_norm": 0.5309537053108215,
+      "learning_rate": 0.00022299098128954095,
+      "loss": 4.0694,
+      "step": 2299
+    },
+    {
+      "epoch": 2.94208,
+      "grad_norm": 0.5304505825042725,
+      "learning_rate": 0.00022295059900390358,
+      "loss": 3.9685,
+      "step": 2300
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 0.5457247495651245,
+      "learning_rate": 0.00022291021671826624,
+      "loss": 3.9616,
+      "step": 2301
+    },
+    {
+      "epoch": 2.94464,
+      "grad_norm": 0.5207186341285706,
+      "learning_rate": 0.00022286983443262887,
+      "loss": 3.9636,
+      "step": 2302
+    },
+    {
+      "epoch": 2.94592,
+      "grad_norm": 0.511628270149231,
+      "learning_rate": 0.0002228294521469915,
+      "loss": 4.0361,
+      "step": 2303
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.5089471340179443,
+      "learning_rate": 0.00022278906986135413,
+      "loss": 3.9737,
+      "step": 2304
+    },
+    {
+      "epoch": 2.94848,
+      "grad_norm": 0.5251920819282532,
+      "learning_rate": 0.00022274868757571678,
+      "loss": 4.0431,
+      "step": 2305
+    },
+    {
+      "epoch": 2.94976,
+      "grad_norm": 0.528783917427063,
+      "learning_rate": 0.0002227083052900794,
+      "loss": 3.9471,
+      "step": 2306
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 0.507645845413208,
+      "learning_rate": 0.00022266792300444202,
+      "loss": 4.0272,
+      "step": 2307
+    },
+    {
+      "epoch": 2.9523200000000003,
+      "grad_norm": 0.5167163014411926,
+      "learning_rate": 0.00022262754071880465,
+      "loss": 3.9864,
+      "step": 2308
+    },
+    {
+      "epoch": 2.9536,
+      "grad_norm": 0.48903942108154297,
+      "learning_rate": 0.0002225871584331673,
+      "loss": 4.0455,
+      "step": 2309
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 0.5110606551170349,
+      "learning_rate": 0.00022254677614752993,
+      "loss": 3.9581,
+      "step": 2310
+    },
+    {
+      "epoch": 2.95616,
+      "grad_norm": 0.5128504633903503,
+      "learning_rate": 0.00022250639386189256,
+      "loss": 3.9707,
+      "step": 2311
+    },
+    {
+      "epoch": 2.95744,
+      "grad_norm": 0.5019478797912598,
+      "learning_rate": 0.0002224660115762552,
+      "loss": 4.0202,
+      "step": 2312
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 0.4747721254825592,
+      "learning_rate": 0.00022242562929061785,
+      "loss": 4.0257,
+      "step": 2313
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.4858660399913788,
+      "learning_rate": 0.00022238524700498048,
+      "loss": 3.9312,
+      "step": 2314
+    },
+    {
+      "epoch": 2.96128,
+      "grad_norm": 0.49668505787849426,
+      "learning_rate": 0.00022234486471934308,
+      "loss": 4.0515,
+      "step": 2315
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 0.5372991561889648,
+      "learning_rate": 0.00022230448243370571,
+      "loss": 3.8684,
+      "step": 2316
+    },
+    {
+      "epoch": 2.9638400000000003,
+      "grad_norm": 0.49556052684783936,
+      "learning_rate": 0.00022226410014806834,
+      "loss": 4.0073,
+      "step": 2317
+    },
+    {
+      "epoch": 2.9651199999999998,
+      "grad_norm": 0.5234814882278442,
+      "learning_rate": 0.000222223717862431,
+      "loss": 3.9878,
+      "step": 2318
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.49563363194465637,
+      "learning_rate": 0.00022218333557679363,
+      "loss": 4.0238,
+      "step": 2319
+    },
+    {
+      "epoch": 2.96768,
+      "grad_norm": 0.5069983005523682,
+      "learning_rate": 0.00022214295329115626,
+      "loss": 3.9518,
+      "step": 2320
+    },
+    {
+      "epoch": 2.96896,
+      "grad_norm": 0.5253263711929321,
+      "learning_rate": 0.0002221025710055189,
+      "loss": 4.012,
+      "step": 2321
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 0.5037320256233215,
+      "learning_rate": 0.00022206218871988155,
+      "loss": 3.965,
+      "step": 2322
+    },
+    {
+      "epoch": 2.97152,
+      "grad_norm": 0.5014638900756836,
+      "learning_rate": 0.00022202180643424415,
+      "loss": 3.9803,
+      "step": 2323
+    },
+    {
+      "epoch": 2.9728,
+      "grad_norm": 0.5158342719078064,
+      "learning_rate": 0.00022198142414860678,
+      "loss": 3.9949,
+      "step": 2324
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 0.5213910937309265,
+      "learning_rate": 0.0002219410418629694,
+      "loss": 3.9926,
+      "step": 2325
+    },
+    {
+      "epoch": 2.9753600000000002,
+      "grad_norm": 0.5123266577720642,
+      "learning_rate": 0.00022190065957733207,
+      "loss": 3.957,
+      "step": 2326
+    },
+    {
+      "epoch": 2.9766399999999997,
+      "grad_norm": 0.5103859305381775,
+      "learning_rate": 0.0002218602772916947,
+      "loss": 4.0097,
+      "step": 2327
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 0.5117716789245605,
+      "learning_rate": 0.00022181989500605733,
+      "loss": 3.9832,
+      "step": 2328
+    },
+    {
+      "epoch": 2.9792,
+      "grad_norm": 0.48839670419692993,
+      "learning_rate": 0.00022177951272041996,
+      "loss": 4.0224,
+      "step": 2329
+    },
+    {
+      "epoch": 2.98048,
+      "grad_norm": 0.5360726714134216,
+      "learning_rate": 0.00022173913043478256,
+      "loss": 3.9817,
+      "step": 2330
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 0.4825299382209778,
+      "learning_rate": 0.00022169874814914525,
+      "loss": 3.9989,
+      "step": 2331
+    },
+    {
+      "epoch": 2.98304,
+      "grad_norm": 0.5100297331809998,
+      "learning_rate": 0.00022165836586350785,
+      "loss": 3.9841,
+      "step": 2332
+    },
+    {
+      "epoch": 2.98432,
+      "grad_norm": 0.4855094850063324,
+      "learning_rate": 0.00022161798357787048,
+      "loss": 3.963,
+      "step": 2333
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.48381808400154114,
+      "learning_rate": 0.0002215776012922331,
+      "loss": 3.9805,
+      "step": 2334
+    },
+    {
+      "epoch": 2.98688,
+      "grad_norm": 0.4909382462501526,
+      "learning_rate": 0.00022153721900659577,
+      "loss": 3.9692,
+      "step": 2335
+    },
+    {
+      "epoch": 2.98816,
+      "grad_norm": 0.4872201085090637,
+      "learning_rate": 0.0002214968367209584,
+      "loss": 3.9689,
+      "step": 2336
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 0.470969557762146,
+      "learning_rate": 0.00022145645443532103,
+      "loss": 3.9348,
+      "step": 2337
+    },
+    {
+      "epoch": 2.99072,
+      "grad_norm": 0.4978778064250946,
+      "learning_rate": 0.00022141607214968363,
+      "loss": 3.9148,
+      "step": 2338
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.48822498321533203,
+      "learning_rate": 0.00022137568986404631,
+      "loss": 3.858,
+      "step": 2339
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 0.4724681079387665,
+      "learning_rate": 0.00022133530757840892,
+      "loss": 3.9882,
+      "step": 2340
+    },
+    {
+      "epoch": 2.99456,
+      "grad_norm": 0.5068411231040955,
+      "learning_rate": 0.00022129492529277155,
+      "loss": 3.9354,
+      "step": 2341
+    },
+    {
+      "epoch": 2.99584,
+      "grad_norm": 0.49231788516044617,
+      "learning_rate": 0.00022125454300713418,
+      "loss": 3.9973,
+      "step": 2342
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 0.5040780901908875,
+      "learning_rate": 0.0002212141607214968,
+      "loss": 3.9624,
+      "step": 2343
+    },
+    {
+      "epoch": 2.9984,
+      "grad_norm": 0.5048054456710815,
+      "learning_rate": 0.00022117377843585946,
+      "loss": 4.0559,
+      "step": 2344
+    },
+    {
+      "epoch": 2.99968,
+      "grad_norm": 0.5000688433647156,
+      "learning_rate": 0.0002211333961502221,
+      "loss": 4.0228,
+      "step": 2345
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9295437932014465,
+      "learning_rate": 0.00022109301386458472,
+      "loss": 3.9067,
+      "step": 2346
+    },
+    {
+      "epoch": 3.00128,
+      "grad_norm": 0.5954660177230835,
+      "learning_rate": 0.00022105263157894733,
+      "loss": 3.7776,
+      "step": 2347
+    },
+    {
+      "epoch": 3.00256,
+      "grad_norm": 0.5306954979896545,
+      "learning_rate": 0.00022101224929330998,
+      "loss": 3.8486,
+      "step": 2348
+    },
+    {
+      "epoch": 3.00384,
+      "grad_norm": 0.5421395897865295,
+      "learning_rate": 0.00022097186700767261,
+      "loss": 3.8924,
+      "step": 2349
+    },
+    {
+      "epoch": 3.00512,
+      "grad_norm": 0.5273167490959167,
+      "learning_rate": 0.00022093148472203524,
+      "loss": 3.8585,
+      "step": 2350
+    },
+    {
+      "epoch": 3.0064,
+      "grad_norm": 0.5256027579307556,
+      "learning_rate": 0.00022089110243639788,
+      "loss": 3.8755,
+      "step": 2351
+    },
+    {
+      "epoch": 3.00768,
+      "grad_norm": 0.5198791027069092,
+      "learning_rate": 0.00022085072015076053,
+      "loss": 3.8962,
+      "step": 2352
+    },
+    {
+      "epoch": 3.00896,
+      "grad_norm": 0.5103099346160889,
+      "learning_rate": 0.00022081033786512316,
+      "loss": 3.8984,
+      "step": 2353
+    },
+    {
+      "epoch": 3.01024,
+      "grad_norm": 0.48842713236808777,
+      "learning_rate": 0.0002207699555794858,
+      "loss": 3.9612,
+      "step": 2354
+    },
+    {
+      "epoch": 3.01152,
+      "grad_norm": 0.5094363689422607,
+      "learning_rate": 0.0002207295732938484,
+      "loss": 3.8044,
+      "step": 2355
+    },
+    {
+      "epoch": 3.0128,
+      "grad_norm": 0.537132740020752,
+      "learning_rate": 0.00022068919100821103,
+      "loss": 3.9496,
+      "step": 2356
+    },
+    {
+      "epoch": 3.01408,
+      "grad_norm": 0.5116694569587708,
+      "learning_rate": 0.00022064880872257368,
+      "loss": 3.8491,
+      "step": 2357
+    },
+    {
+      "epoch": 3.01536,
+      "grad_norm": 0.5091819763183594,
+      "learning_rate": 0.0002206084264369363,
+      "loss": 3.8916,
+      "step": 2358
+    },
+    {
+      "epoch": 3.01664,
+      "grad_norm": 0.5170096158981323,
+      "learning_rate": 0.00022056804415129894,
+      "loss": 3.9251,
+      "step": 2359
+    },
+    {
+      "epoch": 3.01792,
+      "grad_norm": 0.4943462014198303,
+      "learning_rate": 0.00022052766186566157,
+      "loss": 3.9045,
+      "step": 2360
+    },
+    {
+      "epoch": 3.0192,
+      "grad_norm": 0.4762275815010071,
+      "learning_rate": 0.00022048727958002423,
+      "loss": 3.8835,
+      "step": 2361
+    },
+    {
+      "epoch": 3.02048,
+      "grad_norm": 0.5517547130584717,
+      "learning_rate": 0.00022044689729438686,
+      "loss": 3.8825,
+      "step": 2362
+    },
+    {
+      "epoch": 3.02176,
+      "grad_norm": 0.5428676605224609,
+      "learning_rate": 0.00022040651500874946,
+      "loss": 3.8153,
+      "step": 2363
+    },
+    {
+      "epoch": 3.02304,
+      "grad_norm": 0.5272209644317627,
+      "learning_rate": 0.0002203661327231121,
+      "loss": 3.8748,
+      "step": 2364
+    },
+    {
+      "epoch": 3.02432,
+      "grad_norm": 0.5124238729476929,
+      "learning_rate": 0.00022032575043747475,
+      "loss": 3.9815,
+      "step": 2365
+    },
+    {
+      "epoch": 3.0256,
+      "grad_norm": 0.5289603471755981,
+      "learning_rate": 0.00022028536815183738,
+      "loss": 3.8756,
+      "step": 2366
+    },
+    {
+      "epoch": 3.02688,
+      "grad_norm": 0.5256252884864807,
+      "learning_rate": 0.0002202449858662,
+      "loss": 3.8492,
+      "step": 2367
+    },
+    {
+      "epoch": 3.02816,
+      "grad_norm": 0.5126022696495056,
+      "learning_rate": 0.00022020460358056264,
+      "loss": 3.814,
+      "step": 2368
+    },
+    {
+      "epoch": 3.02944,
+      "grad_norm": 0.5239128470420837,
+      "learning_rate": 0.00022016422129492527,
+      "loss": 3.8932,
+      "step": 2369
+    },
+    {
+      "epoch": 3.03072,
+      "grad_norm": 0.49754399061203003,
+      "learning_rate": 0.00022012383900928793,
+      "loss": 3.8759,
+      "step": 2370
+    },
+    {
+      "epoch": 3.032,
+      "grad_norm": 0.49979376792907715,
+      "learning_rate": 0.00022008345672365053,
+      "loss": 3.9152,
+      "step": 2371
+    },
+    {
+      "epoch": 3.03328,
+      "grad_norm": 0.48074454069137573,
+      "learning_rate": 0.00022004307443801316,
+      "loss": 3.8479,
+      "step": 2372
+    },
+    {
+      "epoch": 3.03456,
+      "grad_norm": 0.5178848505020142,
+      "learning_rate": 0.0002200026921523758,
+      "loss": 3.8689,
+      "step": 2373
+    },
+    {
+      "epoch": 3.03584,
+      "grad_norm": 0.5065923929214478,
+      "learning_rate": 0.00021996230986673845,
+      "loss": 3.868,
+      "step": 2374
+    },
+    {
+      "epoch": 3.03712,
+      "grad_norm": 0.4748089015483856,
+      "learning_rate": 0.00021992192758110108,
+      "loss": 3.8268,
+      "step": 2375
+    },
+    {
+      "epoch": 3.0384,
+      "grad_norm": 0.4894319474697113,
+      "learning_rate": 0.0002198815452954637,
+      "loss": 3.7976,
+      "step": 2376
+    },
+    {
+      "epoch": 3.03968,
+      "grad_norm": 0.5088182687759399,
+      "learning_rate": 0.00021984116300982634,
+      "loss": 3.8785,
+      "step": 2377
+    },
+    {
+      "epoch": 3.04096,
+      "grad_norm": 0.48660656809806824,
+      "learning_rate": 0.000219800780724189,
+      "loss": 3.8136,
+      "step": 2378
+    },
+    {
+      "epoch": 3.04224,
+      "grad_norm": 0.5052852034568787,
+      "learning_rate": 0.0002197603984385516,
+      "loss": 3.8922,
+      "step": 2379
+    },
+    {
+      "epoch": 3.04352,
+      "grad_norm": 0.49681487679481506,
+      "learning_rate": 0.00021972001615291423,
+      "loss": 3.8612,
+      "step": 2380
+    },
+    {
+      "epoch": 3.0448,
+      "grad_norm": 0.4781433045864105,
+      "learning_rate": 0.00021967963386727686,
+      "loss": 3.844,
+      "step": 2381
+    },
+    {
+      "epoch": 3.04608,
+      "grad_norm": 0.5147704482078552,
+      "learning_rate": 0.0002196392515816395,
+      "loss": 3.8779,
+      "step": 2382
+    },
+    {
+      "epoch": 3.04736,
+      "grad_norm": 0.49647682905197144,
+      "learning_rate": 0.00021959886929600215,
+      "loss": 3.8326,
+      "step": 2383
+    },
+    {
+      "epoch": 3.04864,
+      "grad_norm": 0.4942707121372223,
+      "learning_rate": 0.00021955848701036478,
+      "loss": 3.8618,
+      "step": 2384
+    },
+    {
+      "epoch": 3.04992,
+      "grad_norm": 0.5220282077789307,
+      "learning_rate": 0.0002195181047247274,
+      "loss": 3.9195,
+      "step": 2385
+    },
+    {
+      "epoch": 3.0512,
+      "grad_norm": 0.4788862466812134,
+      "learning_rate": 0.00021947772243909,
+      "loss": 3.8758,
+      "step": 2386
+    },
+    {
+      "epoch": 3.05248,
+      "grad_norm": 0.4899081289768219,
+      "learning_rate": 0.00021943734015345267,
+      "loss": 3.8233,
+      "step": 2387
+    },
+    {
+      "epoch": 3.05376,
+      "grad_norm": 0.4979023337364197,
+      "learning_rate": 0.0002193969578678153,
+      "loss": 3.8196,
+      "step": 2388
+    },
+    {
+      "epoch": 3.05504,
+      "grad_norm": 0.48769399523735046,
+      "learning_rate": 0.00021935657558217793,
+      "loss": 3.85,
+      "step": 2389
+    },
+    {
+      "epoch": 3.05632,
+      "grad_norm": 0.5051479935646057,
+      "learning_rate": 0.00021931619329654056,
+      "loss": 3.8659,
+      "step": 2390
+    },
+    {
+      "epoch": 3.0576,
+      "grad_norm": 0.508083701133728,
+      "learning_rate": 0.00021927581101090321,
+      "loss": 3.8542,
+      "step": 2391
+    },
+    {
+      "epoch": 3.05888,
+      "grad_norm": 0.5128031373023987,
+      "learning_rate": 0.00021923542872526584,
+      "loss": 3.9097,
+      "step": 2392
+    },
+    {
+      "epoch": 3.06016,
+      "grad_norm": 0.49602705240249634,
+      "learning_rate": 0.00021919504643962847,
+      "loss": 3.8487,
+      "step": 2393
+    },
+    {
+      "epoch": 3.06144,
+      "grad_norm": 0.5159900188446045,
+      "learning_rate": 0.00021915466415399108,
+      "loss": 3.7835,
+      "step": 2394
+    },
+    {
+      "epoch": 3.06272,
+      "grad_norm": 0.4902365803718567,
+      "learning_rate": 0.0002191142818683537,
+      "loss": 3.8462,
+      "step": 2395
+    },
+    {
+      "epoch": 3.064,
+      "grad_norm": 0.5062222480773926,
+      "learning_rate": 0.00021907389958271636,
+      "loss": 3.8156,
+      "step": 2396
+    },
+    {
+      "epoch": 3.06528,
+      "grad_norm": 0.5004411339759827,
+      "learning_rate": 0.000219033517297079,
+      "loss": 3.7797,
+      "step": 2397
+    },
+    {
+      "epoch": 3.06656,
+      "grad_norm": 0.5420650839805603,
+      "learning_rate": 0.00021899313501144162,
+      "loss": 3.9092,
+      "step": 2398
+    },
+    {
+      "epoch": 3.06784,
+      "grad_norm": 0.5353574752807617,
+      "learning_rate": 0.00021895275272580425,
+      "loss": 3.8596,
+      "step": 2399
+    },
+    {
+      "epoch": 3.06912,
+      "grad_norm": 0.4925222396850586,
+      "learning_rate": 0.0002189123704401669,
+      "loss": 3.8678,
+      "step": 2400
+    },
+    {
+      "epoch": 3.0704,
+      "grad_norm": 0.54100501537323,
+      "learning_rate": 0.00021887198815452954,
+      "loss": 3.8494,
+      "step": 2401
+    },
+    {
+      "epoch": 3.07168,
+      "grad_norm": 0.5081486105918884,
+      "learning_rate": 0.00021883160586889214,
+      "loss": 3.8965,
+      "step": 2402
+    },
+    {
+      "epoch": 3.07296,
+      "grad_norm": 0.5015511512756348,
+      "learning_rate": 0.00021879122358325477,
+      "loss": 3.9272,
+      "step": 2403
+    },
+    {
+      "epoch": 3.07424,
+      "grad_norm": 0.5019526481628418,
+      "learning_rate": 0.00021875084129761743,
+      "loss": 3.8459,
+      "step": 2404
+    },
+    {
+      "epoch": 3.07552,
+      "grad_norm": 0.491059273481369,
+      "learning_rate": 0.00021871045901198006,
+      "loss": 3.8815,
+      "step": 2405
+    },
+    {
+      "epoch": 3.0768,
+      "grad_norm": 0.5351020097732544,
+      "learning_rate": 0.0002186700767263427,
+      "loss": 3.8489,
+      "step": 2406
+    },
+    {
+      "epoch": 3.07808,
+      "grad_norm": 0.5229183435440063,
+      "learning_rate": 0.00021862969444070532,
+      "loss": 3.8421,
+      "step": 2407
+    },
+    {
+      "epoch": 3.07936,
+      "grad_norm": 0.507849395275116,
+      "learning_rate": 0.00021858931215506795,
+      "loss": 3.8573,
+      "step": 2408
+    },
+    {
+      "epoch": 3.08064,
+      "grad_norm": 0.5348861813545227,
+      "learning_rate": 0.0002185489298694306,
+      "loss": 3.8958,
+      "step": 2409
+    },
+    {
+      "epoch": 3.08192,
+      "grad_norm": 0.5243686437606812,
+      "learning_rate": 0.0002185085475837932,
+      "loss": 3.8338,
+      "step": 2410
+    },
+    {
+      "epoch": 3.0832,
+      "grad_norm": 0.5294802784919739,
+      "learning_rate": 0.00021846816529815584,
+      "loss": 3.9149,
+      "step": 2411
+    },
+    {
+      "epoch": 3.08448,
+      "grad_norm": 0.5236945748329163,
+      "learning_rate": 0.00021842778301251847,
+      "loss": 3.848,
+      "step": 2412
+    },
+    {
+      "epoch": 3.08576,
+      "grad_norm": 0.5086913704872131,
+      "learning_rate": 0.00021838740072688113,
+      "loss": 3.794,
+      "step": 2413
+    },
+    {
+      "epoch": 3.08704,
+      "grad_norm": 0.5362405776977539,
+      "learning_rate": 0.00021834701844124376,
+      "loss": 3.9174,
+      "step": 2414
+    },
+    {
+      "epoch": 3.08832,
+      "grad_norm": 0.5077964067459106,
+      "learning_rate": 0.0002183066361556064,
+      "loss": 3.8253,
+      "step": 2415
+    },
+    {
+      "epoch": 3.0896,
+      "grad_norm": 0.5096865296363831,
+      "learning_rate": 0.00021826625386996902,
+      "loss": 3.8887,
+      "step": 2416
+    },
+    {
+      "epoch": 3.09088,
+      "grad_norm": 0.5185666084289551,
+      "learning_rate": 0.00021822587158433168,
+      "loss": 3.8292,
+      "step": 2417
+    },
+    {
+      "epoch": 3.09216,
+      "grad_norm": 0.4858289659023285,
+      "learning_rate": 0.0002181854892986943,
+      "loss": 3.8525,
+      "step": 2418
+    },
+    {
+      "epoch": 3.09344,
+      "grad_norm": 0.5283902287483215,
+      "learning_rate": 0.0002181451070130569,
+      "loss": 3.8476,
+      "step": 2419
+    },
+    {
+      "epoch": 3.09472,
+      "grad_norm": 0.5762797594070435,
+      "learning_rate": 0.00021810472472741954,
+      "loss": 3.8591,
+      "step": 2420
+    },
+    {
+      "epoch": 3.096,
+      "grad_norm": 0.5110985636711121,
+      "learning_rate": 0.00021806434244178217,
+      "loss": 3.7624,
+      "step": 2421
+    },
+    {
+      "epoch": 3.09728,
+      "grad_norm": 0.5029181241989136,
+      "learning_rate": 0.00021802396015614483,
+      "loss": 3.8236,
+      "step": 2422
+    },
+    {
+      "epoch": 3.09856,
+      "grad_norm": 0.5185674428939819,
+      "learning_rate": 0.00021798357787050746,
+      "loss": 3.8156,
+      "step": 2423
+    },
+    {
+      "epoch": 3.09984,
+      "grad_norm": 0.5150001645088196,
+      "learning_rate": 0.0002179431955848701,
+      "loss": 3.8182,
+      "step": 2424
+    },
+    {
+      "epoch": 3.10112,
+      "grad_norm": 0.4945181608200073,
+      "learning_rate": 0.0002179028132992327,
+      "loss": 3.8906,
+      "step": 2425
+    },
+    {
+      "epoch": 3.1024,
+      "grad_norm": 0.5029820799827576,
+      "learning_rate": 0.00021786243101359537,
+      "loss": 3.8907,
+      "step": 2426
+    },
+    {
+      "epoch": 3.1036799999999998,
+      "grad_norm": 0.5240580439567566,
+      "learning_rate": 0.00021782204872795798,
+      "loss": 3.911,
+      "step": 2427
+    },
+    {
+      "epoch": 3.10496,
+      "grad_norm": 0.526394248008728,
+      "learning_rate": 0.0002177816664423206,
+      "loss": 3.928,
+      "step": 2428
+    },
+    {
+      "epoch": 3.10624,
+      "grad_norm": 0.517666757106781,
+      "learning_rate": 0.00021774128415668324,
+      "loss": 3.8165,
+      "step": 2429
+    },
+    {
+      "epoch": 3.10752,
+      "grad_norm": 0.5141002535820007,
+      "learning_rate": 0.0002177009018710459,
+      "loss": 3.7981,
+      "step": 2430
+    },
+    {
+      "epoch": 3.1088,
+      "grad_norm": 0.4856560230255127,
+      "learning_rate": 0.00021766051958540853,
+      "loss": 3.8666,
+      "step": 2431
+    },
+    {
+      "epoch": 3.11008,
+      "grad_norm": 0.5310648083686829,
+      "learning_rate": 0.00021762013729977116,
+      "loss": 3.8851,
+      "step": 2432
+    },
+    {
+      "epoch": 3.11136,
+      "grad_norm": 0.5003724098205566,
+      "learning_rate": 0.00021757975501413379,
+      "loss": 3.8379,
+      "step": 2433
+    },
+    {
+      "epoch": 3.11264,
+      "grad_norm": 0.5054803490638733,
+      "learning_rate": 0.0002175393727284964,
+      "loss": 3.7736,
+      "step": 2434
+    },
+    {
+      "epoch": 3.11392,
+      "grad_norm": 0.4943746328353882,
+      "learning_rate": 0.00021749899044285905,
+      "loss": 3.8826,
+      "step": 2435
+    },
+    {
+      "epoch": 3.1152,
+      "grad_norm": 0.49656054377555847,
+      "learning_rate": 0.00021745860815722168,
+      "loss": 3.838,
+      "step": 2436
+    },
+    {
+      "epoch": 3.11648,
+      "grad_norm": 0.5456673502922058,
+      "learning_rate": 0.0002174182258715843,
+      "loss": 3.9265,
+      "step": 2437
+    },
+    {
+      "epoch": 3.11776,
+      "grad_norm": 0.4923163056373596,
+      "learning_rate": 0.00021737784358594694,
+      "loss": 3.9262,
+      "step": 2438
+    },
+    {
+      "epoch": 3.11904,
+      "grad_norm": 0.5190243124961853,
+      "learning_rate": 0.0002173374613003096,
+      "loss": 3.8364,
+      "step": 2439
+    },
+    {
+      "epoch": 3.12032,
+      "grad_norm": 0.47707489132881165,
+      "learning_rate": 0.00021729707901467222,
+      "loss": 3.8502,
+      "step": 2440
+    },
+    {
+      "epoch": 3.1216,
+      "grad_norm": 0.5326244235038757,
+      "learning_rate": 0.00021725669672903485,
+      "loss": 3.8078,
+      "step": 2441
+    },
+    {
+      "epoch": 3.12288,
+      "grad_norm": 0.5211552381515503,
+      "learning_rate": 0.00021721631444339746,
+      "loss": 3.8304,
+      "step": 2442
+    },
+    {
+      "epoch": 3.12416,
+      "grad_norm": 0.524936854839325,
+      "learning_rate": 0.0002171759321577601,
+      "loss": 3.7785,
+      "step": 2443
+    },
+    {
+      "epoch": 3.12544,
+      "grad_norm": 0.5139725804328918,
+      "learning_rate": 0.00021713554987212274,
+      "loss": 3.8666,
+      "step": 2444
+    },
+    {
+      "epoch": 3.12672,
+      "grad_norm": 0.5270793437957764,
+      "learning_rate": 0.00021709516758648537,
+      "loss": 3.8508,
+      "step": 2445
+    },
+    {
+      "epoch": 3.128,
+      "grad_norm": 0.5018444657325745,
+      "learning_rate": 0.000217054785300848,
+      "loss": 3.8951,
+      "step": 2446
+    },
+    {
+      "epoch": 3.12928,
+      "grad_norm": 0.5170176029205322,
+      "learning_rate": 0.00021701440301521066,
+      "loss": 3.8977,
+      "step": 2447
+    },
+    {
+      "epoch": 3.13056,
+      "grad_norm": 0.49703750014305115,
+      "learning_rate": 0.0002169740207295733,
+      "loss": 3.7843,
+      "step": 2448
+    },
+    {
+      "epoch": 3.13184,
+      "grad_norm": 0.5318528413772583,
+      "learning_rate": 0.00021693363844393592,
+      "loss": 3.9147,
+      "step": 2449
+    },
+    {
+      "epoch": 3.13312,
+      "grad_norm": 0.5124686360359192,
+      "learning_rate": 0.00021689325615829852,
+      "loss": 3.8174,
+      "step": 2450
+    },
+    {
+      "epoch": 3.1344,
+      "grad_norm": 0.5229965448379517,
+      "learning_rate": 0.00021685287387266115,
+      "loss": 3.9171,
+      "step": 2451
+    },
+    {
+      "epoch": 3.13568,
+      "grad_norm": 0.5190241932868958,
+      "learning_rate": 0.0002168124915870238,
+      "loss": 3.8315,
+      "step": 2452
+    },
+    {
+      "epoch": 3.13696,
+      "grad_norm": 0.5506916642189026,
+      "learning_rate": 0.00021677210930138644,
+      "loss": 3.8351,
+      "step": 2453
+    },
+    {
+      "epoch": 3.13824,
+      "grad_norm": 0.5469547510147095,
+      "learning_rate": 0.00021673172701574907,
+      "loss": 3.9098,
+      "step": 2454
+    },
+    {
+      "epoch": 3.13952,
+      "grad_norm": 0.5336132049560547,
+      "learning_rate": 0.0002166913447301117,
+      "loss": 3.8963,
+      "step": 2455
+    },
+    {
+      "epoch": 3.1408,
+      "grad_norm": 0.549925684928894,
+      "learning_rate": 0.00021665096244447436,
+      "loss": 3.8541,
+      "step": 2456
+    },
+    {
+      "epoch": 3.14208,
+      "grad_norm": 0.5148488283157349,
+      "learning_rate": 0.000216610580158837,
+      "loss": 3.8654,
+      "step": 2457
+    },
+    {
+      "epoch": 3.14336,
+      "grad_norm": 0.5574371218681335,
+      "learning_rate": 0.0002165701978731996,
+      "loss": 3.9102,
+      "step": 2458
+    },
+    {
+      "epoch": 3.14464,
+      "grad_norm": 0.5164331793785095,
+      "learning_rate": 0.00021652981558756222,
+      "loss": 3.7154,
+      "step": 2459
+    },
+    {
+      "epoch": 3.14592,
+      "grad_norm": 0.5179708003997803,
+      "learning_rate": 0.00021648943330192488,
+      "loss": 3.9097,
+      "step": 2460
+    },
+    {
+      "epoch": 3.1471999999999998,
+      "grad_norm": 0.5360696315765381,
+      "learning_rate": 0.0002164490510162875,
+      "loss": 3.8359,
+      "step": 2461
+    },
+    {
+      "epoch": 3.14848,
+      "grad_norm": 0.5113580822944641,
+      "learning_rate": 0.00021640866873065014,
+      "loss": 3.911,
+      "step": 2462
+    },
+    {
+      "epoch": 3.14976,
+      "grad_norm": 0.5048407316207886,
+      "learning_rate": 0.00021636828644501277,
+      "loss": 3.8197,
+      "step": 2463
+    },
+    {
+      "epoch": 3.15104,
+      "grad_norm": 0.5186136960983276,
+      "learning_rate": 0.0002163279041593754,
+      "loss": 3.8374,
+      "step": 2464
+    },
+    {
+      "epoch": 3.15232,
+      "grad_norm": 0.4925592243671417,
+      "learning_rate": 0.00021628752187373806,
+      "loss": 3.8343,
+      "step": 2465
+    },
+    {
+      "epoch": 3.1536,
+      "grad_norm": 0.49732711911201477,
+      "learning_rate": 0.00021624713958810066,
+      "loss": 3.8915,
+      "step": 2466
+    },
+    {
+      "epoch": 3.15488,
+      "grad_norm": 0.5022438764572144,
+      "learning_rate": 0.0002162067573024633,
+      "loss": 3.8251,
+      "step": 2467
+    },
+    {
+      "epoch": 3.15616,
+      "grad_norm": 0.5367256999015808,
+      "learning_rate": 0.00021616637501682592,
+      "loss": 3.8802,
+      "step": 2468
+    },
+    {
+      "epoch": 3.15744,
+      "grad_norm": 0.4930475950241089,
+      "learning_rate": 0.00021612599273118858,
+      "loss": 3.8364,
+      "step": 2469
+    },
+    {
+      "epoch": 3.15872,
+      "grad_norm": 0.5071558356285095,
+      "learning_rate": 0.0002160856104455512,
+      "loss": 3.828,
+      "step": 2470
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.5163344740867615,
+      "learning_rate": 0.00021604522815991384,
+      "loss": 3.7945,
+      "step": 2471
+    },
+    {
+      "epoch": 3.16128,
+      "grad_norm": 0.5042238831520081,
+      "learning_rate": 0.00021600484587427647,
+      "loss": 3.8724,
+      "step": 2472
+    },
+    {
+      "epoch": 3.16256,
+      "grad_norm": 0.5206556916236877,
+      "learning_rate": 0.00021596446358863912,
+      "loss": 3.8445,
+      "step": 2473
+    },
+    {
+      "epoch": 3.16384,
+      "grad_norm": 0.4869347810745239,
+      "learning_rate": 0.00021592408130300173,
+      "loss": 3.8982,
+      "step": 2474
+    },
+    {
+      "epoch": 3.16512,
+      "grad_norm": 0.4939110279083252,
+      "learning_rate": 0.00021588369901736436,
+      "loss": 3.8215,
+      "step": 2475
+    },
+    {
+      "epoch": 3.1664,
+      "grad_norm": 0.4792523980140686,
+      "learning_rate": 0.000215843316731727,
+      "loss": 3.8195,
+      "step": 2476
+    },
+    {
+      "epoch": 3.16768,
+      "grad_norm": 0.4833280146121979,
+      "learning_rate": 0.00021580293444608962,
+      "loss": 3.878,
+      "step": 2477
+    },
+    {
+      "epoch": 3.16896,
+      "grad_norm": 0.48897233605384827,
+      "learning_rate": 0.00021576255216045227,
+      "loss": 3.8418,
+      "step": 2478
+    },
+    {
+      "epoch": 3.17024,
+      "grad_norm": 0.5081768035888672,
+      "learning_rate": 0.0002157221698748149,
+      "loss": 3.8296,
+      "step": 2479
+    },
+    {
+      "epoch": 3.17152,
+      "grad_norm": 0.499881386756897,
+      "learning_rate": 0.00021568178758917753,
+      "loss": 3.9065,
+      "step": 2480
+    },
+    {
+      "epoch": 3.1728,
+      "grad_norm": 0.5089867115020752,
+      "learning_rate": 0.00021564140530354014,
+      "loss": 3.9007,
+      "step": 2481
+    },
+    {
+      "epoch": 3.17408,
+      "grad_norm": 0.490253746509552,
+      "learning_rate": 0.0002156010230179028,
+      "loss": 3.9089,
+      "step": 2482
+    },
+    {
+      "epoch": 3.17536,
+      "grad_norm": 0.49601680040359497,
+      "learning_rate": 0.00021556064073226542,
+      "loss": 3.8876,
+      "step": 2483
+    },
+    {
+      "epoch": 3.17664,
+      "grad_norm": 0.5113343000411987,
+      "learning_rate": 0.00021552025844662805,
+      "loss": 3.8393,
+      "step": 2484
+    },
+    {
+      "epoch": 3.17792,
+      "grad_norm": 0.47494059801101685,
+      "learning_rate": 0.00021547987616099068,
+      "loss": 3.8461,
+      "step": 2485
+    },
+    {
+      "epoch": 3.1792,
+      "grad_norm": 0.5105437636375427,
+      "learning_rate": 0.00021543949387535334,
+      "loss": 3.8837,
+      "step": 2486
+    },
+    {
+      "epoch": 3.18048,
+      "grad_norm": 0.5032381415367126,
+      "learning_rate": 0.00021539911158971597,
+      "loss": 3.8457,
+      "step": 2487
+    },
+    {
+      "epoch": 3.18176,
+      "grad_norm": 0.49342939257621765,
+      "learning_rate": 0.0002153587293040786,
+      "loss": 3.819,
+      "step": 2488
+    },
+    {
+      "epoch": 3.18304,
+      "grad_norm": 0.49922072887420654,
+      "learning_rate": 0.0002153183470184412,
+      "loss": 3.792,
+      "step": 2489
+    },
+    {
+      "epoch": 3.18432,
+      "grad_norm": 0.49167224764823914,
+      "learning_rate": 0.00021527796473280384,
+      "loss": 3.8155,
+      "step": 2490
+    },
+    {
+      "epoch": 3.1856,
+      "grad_norm": 0.4955747425556183,
+      "learning_rate": 0.0002152375824471665,
+      "loss": 3.9052,
+      "step": 2491
+    },
+    {
+      "epoch": 3.18688,
+      "grad_norm": 0.513163149356842,
+      "learning_rate": 0.00021519720016152912,
+      "loss": 3.946,
+      "step": 2492
+    },
+    {
+      "epoch": 3.18816,
+      "grad_norm": 0.4796956181526184,
+      "learning_rate": 0.00021515681787589175,
+      "loss": 3.8418,
+      "step": 2493
+    },
+    {
+      "epoch": 3.18944,
+      "grad_norm": 0.5129488110542297,
+      "learning_rate": 0.00021511643559025438,
+      "loss": 3.9219,
+      "step": 2494
+    },
+    {
+      "epoch": 3.19072,
+      "grad_norm": 0.49274012446403503,
+      "learning_rate": 0.00021507605330461704,
+      "loss": 3.9179,
+      "step": 2495
+    },
+    {
+      "epoch": 3.192,
+      "grad_norm": 0.5269827246665955,
+      "learning_rate": 0.00021503567101897967,
+      "loss": 3.8845,
+      "step": 2496
+    },
+    {
+      "epoch": 3.19328,
+      "grad_norm": 0.5039204955101013,
+      "learning_rate": 0.00021499528873334227,
+      "loss": 3.8432,
+      "step": 2497
+    },
+    {
+      "epoch": 3.19456,
+      "grad_norm": 0.5350068807601929,
+      "learning_rate": 0.0002149549064477049,
+      "loss": 3.7202,
+      "step": 2498
+    },
+    {
+      "epoch": 3.19584,
+      "grad_norm": 0.49248334765434265,
+      "learning_rate": 0.00021491452416206756,
+      "loss": 3.848,
+      "step": 2499
+    },
+    {
+      "epoch": 3.19712,
+      "grad_norm": 0.5159714221954346,
+      "learning_rate": 0.0002148741418764302,
+      "loss": 3.8145,
+      "step": 2500
+    },
+    {
+      "epoch": 3.1984,
+      "grad_norm": 0.4729326367378235,
+      "learning_rate": 0.00021483375959079282,
+      "loss": 3.9304,
+      "step": 2501
+    },
+    {
+      "epoch": 3.19968,
+      "grad_norm": 0.5187593698501587,
+      "learning_rate": 0.00021479337730515545,
+      "loss": 3.8462,
+      "step": 2502
+    },
+    {
+      "epoch": 3.20096,
+      "grad_norm": 0.521251380443573,
+      "learning_rate": 0.00021475299501951808,
+      "loss": 3.8268,
+      "step": 2503
+    },
+    {
+      "epoch": 3.20224,
+      "grad_norm": 0.4812679886817932,
+      "learning_rate": 0.00021471261273388074,
+      "loss": 3.8409,
+      "step": 2504
+    },
+    {
+      "epoch": 3.20352,
+      "grad_norm": 0.4940718114376068,
+      "learning_rate": 0.00021467223044824337,
+      "loss": 3.8029,
+      "step": 2505
+    },
+    {
+      "epoch": 3.2048,
+      "grad_norm": 0.5103253126144409,
+      "learning_rate": 0.00021463184816260597,
+      "loss": 3.8354,
+      "step": 2506
+    },
+    {
+      "epoch": 3.20608,
+      "grad_norm": 0.5237296223640442,
+      "learning_rate": 0.0002145914658769686,
+      "loss": 3.8567,
+      "step": 2507
+    },
+    {
+      "epoch": 3.20736,
+      "grad_norm": 0.5030867457389832,
+      "learning_rate": 0.00021455108359133126,
+      "loss": 3.8715,
+      "step": 2508
+    },
+    {
+      "epoch": 3.20864,
+      "grad_norm": 0.49784404039382935,
+      "learning_rate": 0.0002145107013056939,
+      "loss": 3.7889,
+      "step": 2509
+    },
+    {
+      "epoch": 3.20992,
+      "grad_norm": 0.5086187124252319,
+      "learning_rate": 0.00021447031902005652,
+      "loss": 3.9156,
+      "step": 2510
+    },
+    {
+      "epoch": 3.2112,
+      "grad_norm": 0.5107704401016235,
+      "learning_rate": 0.00021442993673441915,
+      "loss": 3.7905,
+      "step": 2511
+    },
+    {
+      "epoch": 3.2124800000000002,
+      "grad_norm": 0.4935329258441925,
+      "learning_rate": 0.0002143895544487818,
+      "loss": 3.8257,
+      "step": 2512
+    },
+    {
+      "epoch": 3.21376,
+      "grad_norm": 0.5265173316001892,
+      "learning_rate": 0.00021434917216314444,
+      "loss": 3.8725,
+      "step": 2513
+    },
+    {
+      "epoch": 3.21504,
+      "grad_norm": 0.5278881192207336,
+      "learning_rate": 0.00021430878987750704,
+      "loss": 3.8997,
+      "step": 2514
+    },
+    {
+      "epoch": 3.21632,
+      "grad_norm": 0.5073565244674683,
+      "learning_rate": 0.00021426840759186967,
+      "loss": 3.8788,
+      "step": 2515
+    },
+    {
+      "epoch": 3.2176,
+      "grad_norm": 0.5007202625274658,
+      "learning_rate": 0.0002142280253062323,
+      "loss": 3.8289,
+      "step": 2516
+    },
+    {
+      "epoch": 3.21888,
+      "grad_norm": 0.5268493294715881,
+      "learning_rate": 0.00021418764302059496,
+      "loss": 3.9057,
+      "step": 2517
+    },
+    {
+      "epoch": 3.22016,
+      "grad_norm": 0.5133515000343323,
+      "learning_rate": 0.00021414726073495759,
+      "loss": 3.9227,
+      "step": 2518
+    },
+    {
+      "epoch": 3.22144,
+      "grad_norm": 0.5341963171958923,
+      "learning_rate": 0.00021410687844932022,
+      "loss": 3.8918,
+      "step": 2519
+    },
+    {
+      "epoch": 3.22272,
+      "grad_norm": 0.488156795501709,
+      "learning_rate": 0.00021406649616368285,
+      "loss": 3.9192,
+      "step": 2520
+    },
+    {
+      "epoch": 3.224,
+      "grad_norm": 0.5290320515632629,
+      "learning_rate": 0.0002140261138780455,
+      "loss": 3.7908,
+      "step": 2521
+    },
+    {
+      "epoch": 3.22528,
+      "grad_norm": 0.510662853717804,
+      "learning_rate": 0.0002139857315924081,
+      "loss": 3.837,
+      "step": 2522
+    },
+    {
+      "epoch": 3.22656,
+      "grad_norm": 0.5320321321487427,
+      "learning_rate": 0.00021394534930677074,
+      "loss": 3.8889,
+      "step": 2523
+    },
+    {
+      "epoch": 3.22784,
+      "grad_norm": 0.5298739075660706,
+      "learning_rate": 0.00021390496702113337,
+      "loss": 3.9002,
+      "step": 2524
+    },
+    {
+      "epoch": 3.22912,
+      "grad_norm": 0.5062413215637207,
+      "learning_rate": 0.00021386458473549602,
+      "loss": 3.8557,
+      "step": 2525
+    },
+    {
+      "epoch": 3.2304,
+      "grad_norm": 0.5112326145172119,
+      "learning_rate": 0.00021382420244985865,
+      "loss": 3.8719,
+      "step": 2526
+    },
+    {
+      "epoch": 3.23168,
+      "grad_norm": 0.48387032747268677,
+      "learning_rate": 0.00021378382016422128,
+      "loss": 3.8151,
+      "step": 2527
+    },
+    {
+      "epoch": 3.23296,
+      "grad_norm": 0.49385952949523926,
+      "learning_rate": 0.0002137434378785839,
+      "loss": 3.8642,
+      "step": 2528
+    },
+    {
+      "epoch": 3.23424,
+      "grad_norm": 0.5138410925865173,
+      "learning_rate": 0.00021370305559294652,
+      "loss": 3.9133,
+      "step": 2529
+    },
+    {
+      "epoch": 3.23552,
+      "grad_norm": 0.5019318461418152,
+      "learning_rate": 0.00021366267330730917,
+      "loss": 3.8908,
+      "step": 2530
+    },
+    {
+      "epoch": 3.2368,
+      "grad_norm": 0.4930397868156433,
+      "learning_rate": 0.0002136222910216718,
+      "loss": 3.8062,
+      "step": 2531
+    },
+    {
+      "epoch": 3.23808,
+      "grad_norm": 0.5121169090270996,
+      "learning_rate": 0.00021358190873603443,
+      "loss": 3.7471,
+      "step": 2532
+    },
+    {
+      "epoch": 3.23936,
+      "grad_norm": 0.4978577494621277,
+      "learning_rate": 0.00021354152645039706,
+      "loss": 3.9114,
+      "step": 2533
+    },
+    {
+      "epoch": 3.24064,
+      "grad_norm": 0.5160923004150391,
+      "learning_rate": 0.00021350114416475972,
+      "loss": 3.832,
+      "step": 2534
+    },
+    {
+      "epoch": 3.24192,
+      "grad_norm": 0.49270156025886536,
+      "learning_rate": 0.00021346076187912235,
+      "loss": 3.8092,
+      "step": 2535
+    },
+    {
+      "epoch": 3.2432,
+      "grad_norm": 0.5121193528175354,
+      "learning_rate": 0.00021342037959348498,
+      "loss": 3.7557,
+      "step": 2536
+    },
+    {
+      "epoch": 3.24448,
+      "grad_norm": 0.518182098865509,
+      "learning_rate": 0.00021337999730784758,
+      "loss": 3.8582,
+      "step": 2537
+    },
+    {
+      "epoch": 3.24576,
+      "grad_norm": 0.5109360814094543,
+      "learning_rate": 0.00021333961502221024,
+      "loss": 3.8122,
+      "step": 2538
+    },
+    {
+      "epoch": 3.24704,
+      "grad_norm": 0.5071660280227661,
+      "learning_rate": 0.00021329923273657287,
+      "loss": 3.8727,
+      "step": 2539
+    },
+    {
+      "epoch": 3.24832,
+      "grad_norm": 0.5436646342277527,
+      "learning_rate": 0.0002132588504509355,
+      "loss": 3.903,
+      "step": 2540
+    },
+    {
+      "epoch": 3.2496,
+      "grad_norm": 0.5585988163948059,
+      "learning_rate": 0.00021321846816529813,
+      "loss": 3.8499,
+      "step": 2541
+    },
+    {
+      "epoch": 3.25088,
+      "grad_norm": 0.5021224617958069,
+      "learning_rate": 0.00021317808587966076,
+      "loss": 3.8658,
+      "step": 2542
+    },
+    {
+      "epoch": 3.25216,
+      "grad_norm": 0.532137393951416,
+      "learning_rate": 0.00021313770359402342,
+      "loss": 3.8951,
+      "step": 2543
+    },
+    {
+      "epoch": 3.25344,
+      "grad_norm": 0.5324034690856934,
+      "learning_rate": 0.00021309732130838605,
+      "loss": 3.8472,
+      "step": 2544
+    },
+    {
+      "epoch": 3.25472,
+      "grad_norm": 0.5090122222900391,
+      "learning_rate": 0.00021305693902274865,
+      "loss": 3.8011,
+      "step": 2545
+    },
+    {
+      "epoch": 3.2560000000000002,
+      "grad_norm": 0.5446542501449585,
+      "learning_rate": 0.00021301655673711128,
+      "loss": 3.8029,
+      "step": 2546
+    },
+    {
+      "epoch": 3.25728,
+      "grad_norm": 0.5329357385635376,
+      "learning_rate": 0.00021297617445147394,
+      "loss": 3.7843,
+      "step": 2547
+    },
+    {
+      "epoch": 3.25856,
+      "grad_norm": 0.552749752998352,
+      "learning_rate": 0.00021293579216583657,
+      "loss": 3.8847,
+      "step": 2548
+    },
+    {
+      "epoch": 3.25984,
+      "grad_norm": 0.529541552066803,
+      "learning_rate": 0.0002128954098801992,
+      "loss": 3.9156,
+      "step": 2549
+    },
+    {
+      "epoch": 3.26112,
+      "grad_norm": 0.5321997404098511,
+      "learning_rate": 0.00021285502759456183,
+      "loss": 3.8645,
+      "step": 2550
+    },
+    {
+      "epoch": 3.2624,
+      "grad_norm": 0.5043097138404846,
+      "learning_rate": 0.0002128146453089245,
+      "loss": 3.8073,
+      "step": 2551
+    },
+    {
+      "epoch": 3.26368,
+      "grad_norm": 0.5317241549491882,
+      "learning_rate": 0.00021277426302328712,
+      "loss": 3.8946,
+      "step": 2552
+    },
+    {
+      "epoch": 3.26496,
+      "grad_norm": 0.5115827918052673,
+      "learning_rate": 0.00021273388073764972,
+      "loss": 3.8361,
+      "step": 2553
+    },
+    {
+      "epoch": 3.26624,
+      "grad_norm": 0.5349282622337341,
+      "learning_rate": 0.00021269349845201235,
+      "loss": 3.8482,
+      "step": 2554
+    },
+    {
+      "epoch": 3.26752,
+      "grad_norm": 0.5121451616287231,
+      "learning_rate": 0.00021265311616637498,
+      "loss": 3.8416,
+      "step": 2555
+    },
+    {
+      "epoch": 3.2688,
+      "grad_norm": 0.4995376765727997,
+      "learning_rate": 0.00021261273388073764,
+      "loss": 3.848,
+      "step": 2556
+    },
+    {
+      "epoch": 3.27008,
+      "grad_norm": 0.5239928960800171,
+      "learning_rate": 0.00021257235159510027,
+      "loss": 3.8595,
+      "step": 2557
+    },
+    {
+      "epoch": 3.27136,
+      "grad_norm": 0.5018921494483948,
+      "learning_rate": 0.0002125319693094629,
+      "loss": 3.8748,
+      "step": 2558
+    },
+    {
+      "epoch": 3.27264,
+      "grad_norm": 0.5001853108406067,
+      "learning_rate": 0.00021249158702382553,
+      "loss": 3.7935,
+      "step": 2559
+    },
+    {
+      "epoch": 3.27392,
+      "grad_norm": 0.5090051889419556,
+      "learning_rate": 0.00021245120473818818,
+      "loss": 3.8594,
+      "step": 2560
+    },
+    {
+      "epoch": 3.2752,
+      "grad_norm": 0.491327702999115,
+      "learning_rate": 0.0002124108224525508,
+      "loss": 3.8575,
+      "step": 2561
+    },
+    {
+      "epoch": 3.27648,
+      "grad_norm": 0.516203761100769,
+      "learning_rate": 0.00021237044016691342,
+      "loss": 3.8331,
+      "step": 2562
+    },
+    {
+      "epoch": 3.27776,
+      "grad_norm": 0.5098548531532288,
+      "learning_rate": 0.00021233005788127605,
+      "loss": 3.8881,
+      "step": 2563
+    },
+    {
+      "epoch": 3.27904,
+      "grad_norm": 0.5232280492782593,
+      "learning_rate": 0.0002122896755956387,
+      "loss": 3.8788,
+      "step": 2564
+    },
+    {
+      "epoch": 3.28032,
+      "grad_norm": 0.5226551294326782,
+      "learning_rate": 0.00021224929331000133,
+      "loss": 3.7725,
+      "step": 2565
+    },
+    {
+      "epoch": 3.2816,
+      "grad_norm": 0.5192921161651611,
+      "learning_rate": 0.00021220891102436396,
+      "loss": 3.9056,
+      "step": 2566
+    },
+    {
+      "epoch": 3.28288,
+      "grad_norm": 0.5096680521965027,
+      "learning_rate": 0.0002121685287387266,
+      "loss": 3.8403,
+      "step": 2567
+    },
+    {
+      "epoch": 3.28416,
+      "grad_norm": 0.4832972586154938,
+      "learning_rate": 0.00021212814645308925,
+      "loss": 3.8747,
+      "step": 2568
+    },
+    {
+      "epoch": 3.28544,
+      "grad_norm": 0.5360098481178284,
+      "learning_rate": 0.00021208776416745185,
+      "loss": 3.8489,
+      "step": 2569
+    },
+    {
+      "epoch": 3.28672,
+      "grad_norm": 0.5098423361778259,
+      "learning_rate": 0.00021204738188181448,
+      "loss": 3.8063,
+      "step": 2570
+    },
+    {
+      "epoch": 3.288,
+      "grad_norm": 0.5014640092849731,
+      "learning_rate": 0.00021200699959617712,
+      "loss": 3.8698,
+      "step": 2571
+    },
+    {
+      "epoch": 3.2892799999999998,
+      "grad_norm": 0.531622052192688,
+      "learning_rate": 0.00021196661731053975,
+      "loss": 4.001,
+      "step": 2572
+    },
+    {
+      "epoch": 3.29056,
+      "grad_norm": 0.5045267939567566,
+      "learning_rate": 0.0002119262350249024,
+      "loss": 3.8148,
+      "step": 2573
+    },
+    {
+      "epoch": 3.29184,
+      "grad_norm": 0.5068945288658142,
+      "learning_rate": 0.00021188585273926503,
+      "loss": 3.8387,
+      "step": 2574
+    },
+    {
+      "epoch": 3.29312,
+      "grad_norm": 0.5441416501998901,
+      "learning_rate": 0.00021184547045362766,
+      "loss": 3.8937,
+      "step": 2575
+    },
+    {
+      "epoch": 3.2944,
+      "grad_norm": 0.5135897397994995,
+      "learning_rate": 0.00021180508816799027,
+      "loss": 3.8456,
+      "step": 2576
+    },
+    {
+      "epoch": 3.29568,
+      "grad_norm": 0.5423394441604614,
+      "learning_rate": 0.00021176470588235295,
+      "loss": 3.9034,
+      "step": 2577
+    },
+    {
+      "epoch": 3.29696,
+      "grad_norm": 0.5269873738288879,
+      "learning_rate": 0.00021172432359671555,
+      "loss": 3.8488,
+      "step": 2578
+    },
+    {
+      "epoch": 3.29824,
+      "grad_norm": 0.57085782289505,
+      "learning_rate": 0.00021168394131107818,
+      "loss": 3.8606,
+      "step": 2579
+    },
+    {
+      "epoch": 3.2995200000000002,
+      "grad_norm": 0.5277544260025024,
+      "learning_rate": 0.0002116435590254408,
+      "loss": 3.8475,
+      "step": 2580
+    },
+    {
+      "epoch": 3.3008,
+      "grad_norm": 0.5475723743438721,
+      "learning_rate": 0.00021160317673980347,
+      "loss": 3.8538,
+      "step": 2581
+    },
+    {
+      "epoch": 3.30208,
+      "grad_norm": 0.5339505076408386,
+      "learning_rate": 0.0002115627944541661,
+      "loss": 3.8168,
+      "step": 2582
+    },
+    {
+      "epoch": 3.30336,
+      "grad_norm": 0.561708927154541,
+      "learning_rate": 0.00021152241216852873,
+      "loss": 3.8303,
+      "step": 2583
+    },
+    {
+      "epoch": 3.30464,
+      "grad_norm": 0.507378339767456,
+      "learning_rate": 0.00021148202988289133,
+      "loss": 3.8375,
+      "step": 2584
+    },
+    {
+      "epoch": 3.30592,
+      "grad_norm": 0.5297276377677917,
+      "learning_rate": 0.00021144164759725396,
+      "loss": 3.8747,
+      "step": 2585
+    },
+    {
+      "epoch": 3.3072,
+      "grad_norm": 0.5358178615570068,
+      "learning_rate": 0.00021140126531161662,
+      "loss": 3.8599,
+      "step": 2586
+    },
+    {
+      "epoch": 3.30848,
+      "grad_norm": 0.5373551845550537,
+      "learning_rate": 0.00021136088302597925,
+      "loss": 3.8097,
+      "step": 2587
+    },
+    {
+      "epoch": 3.30976,
+      "grad_norm": 0.5402244925498962,
+      "learning_rate": 0.00021132050074034188,
+      "loss": 3.8853,
+      "step": 2588
+    },
+    {
+      "epoch": 3.31104,
+      "grad_norm": 0.5271474123001099,
+      "learning_rate": 0.0002112801184547045,
+      "loss": 3.7922,
+      "step": 2589
+    },
+    {
+      "epoch": 3.31232,
+      "grad_norm": 0.5495550632476807,
+      "learning_rate": 0.00021123973616906717,
+      "loss": 3.8268,
+      "step": 2590
+    },
+    {
+      "epoch": 3.3136,
+      "grad_norm": 0.5179243683815002,
+      "learning_rate": 0.0002111993538834298,
+      "loss": 3.9258,
+      "step": 2591
+    },
+    {
+      "epoch": 3.31488,
+      "grad_norm": 0.5357790589332581,
+      "learning_rate": 0.00021115897159779243,
+      "loss": 3.8901,
+      "step": 2592
+    },
+    {
+      "epoch": 3.31616,
+      "grad_norm": 0.4818624258041382,
+      "learning_rate": 0.00021111858931215503,
+      "loss": 3.8431,
+      "step": 2593
+    },
+    {
+      "epoch": 3.31744,
+      "grad_norm": 0.5085586905479431,
+      "learning_rate": 0.0002110782070265177,
+      "loss": 3.8345,
+      "step": 2594
+    },
+    {
+      "epoch": 3.31872,
+      "grad_norm": 0.518665611743927,
+      "learning_rate": 0.00021103782474088032,
+      "loss": 3.7475,
+      "step": 2595
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.5138918161392212,
+      "learning_rate": 0.00021099744245524295,
+      "loss": 3.8736,
+      "step": 2596
+    },
+    {
+      "epoch": 3.32128,
+      "grad_norm": 0.5055665373802185,
+      "learning_rate": 0.00021095706016960558,
+      "loss": 3.8456,
+      "step": 2597
+    },
+    {
+      "epoch": 3.32256,
+      "grad_norm": 0.4695318937301636,
+      "learning_rate": 0.0002109166778839682,
+      "loss": 3.8635,
+      "step": 2598
+    },
+    {
+      "epoch": 3.32384,
+      "grad_norm": 0.47867250442504883,
+      "learning_rate": 0.00021087629559833087,
+      "loss": 3.8418,
+      "step": 2599
+    },
+    {
+      "epoch": 3.32512,
+      "grad_norm": 0.5042849779129028,
+      "learning_rate": 0.0002108359133126935,
+      "loss": 3.8351,
+      "step": 2600
+    },
+    {
+      "epoch": 3.3264,
+      "grad_norm": 0.5028448700904846,
+      "learning_rate": 0.0002107955310270561,
+      "loss": 3.8349,
+      "step": 2601
+    },
+    {
+      "epoch": 3.32768,
+      "grad_norm": 0.5034447312355042,
+      "learning_rate": 0.00021075514874141873,
+      "loss": 3.8095,
+      "step": 2602
+    },
+    {
+      "epoch": 3.32896,
+      "grad_norm": 0.4787198007106781,
+      "learning_rate": 0.00021071476645578139,
+      "loss": 3.7805,
+      "step": 2603
+    },
+    {
+      "epoch": 3.33024,
+      "grad_norm": 0.535429835319519,
+      "learning_rate": 0.00021067438417014402,
+      "loss": 3.8945,
+      "step": 2604
+    },
+    {
+      "epoch": 3.33152,
+      "grad_norm": 0.5033429265022278,
+      "learning_rate": 0.00021063400188450665,
+      "loss": 3.7983,
+      "step": 2605
+    },
+    {
+      "epoch": 3.3327999999999998,
+      "grad_norm": 0.4938407242298126,
+      "learning_rate": 0.00021059361959886928,
+      "loss": 3.8798,
+      "step": 2606
+    },
+    {
+      "epoch": 3.33408,
+      "grad_norm": 0.5288311839103699,
+      "learning_rate": 0.00021055323731323193,
+      "loss": 3.833,
+      "step": 2607
+    },
+    {
+      "epoch": 3.33536,
+      "grad_norm": 0.5072165131568909,
+      "learning_rate": 0.00021051285502759456,
+      "loss": 3.8302,
+      "step": 2608
+    },
+    {
+      "epoch": 3.33664,
+      "grad_norm": 0.5025217533111572,
+      "learning_rate": 0.00021047247274195717,
+      "loss": 3.8571,
+      "step": 2609
+    },
+    {
+      "epoch": 3.33792,
+      "grad_norm": 0.5042093396186829,
+      "learning_rate": 0.0002104320904563198,
+      "loss": 3.8965,
+      "step": 2610
+    },
+    {
+      "epoch": 3.3392,
+      "grad_norm": 0.5024438500404358,
+      "learning_rate": 0.00021039170817068243,
+      "loss": 3.8112,
+      "step": 2611
+    },
+    {
+      "epoch": 3.34048,
+      "grad_norm": 0.4909932613372803,
+      "learning_rate": 0.00021035132588504508,
+      "loss": 3.8429,
+      "step": 2612
+    },
+    {
+      "epoch": 3.34176,
+      "grad_norm": 0.505427896976471,
+      "learning_rate": 0.00021031094359940771,
+      "loss": 3.9004,
+      "step": 2613
+    },
+    {
+      "epoch": 3.3430400000000002,
+      "grad_norm": 0.5072449445724487,
+      "learning_rate": 0.00021027056131377034,
+      "loss": 3.824,
+      "step": 2614
+    },
+    {
+      "epoch": 3.34432,
+      "grad_norm": 0.5105025172233582,
+      "learning_rate": 0.00021023017902813297,
+      "loss": 3.8643,
+      "step": 2615
+    },
+    {
+      "epoch": 3.3456,
+      "grad_norm": 0.5096602439880371,
+      "learning_rate": 0.00021018979674249563,
+      "loss": 3.8363,
+      "step": 2616
+    },
+    {
+      "epoch": 3.34688,
+      "grad_norm": 0.4964136779308319,
+      "learning_rate": 0.00021014941445685823,
+      "loss": 3.8228,
+      "step": 2617
+    },
+    {
+      "epoch": 3.34816,
+      "grad_norm": 0.510119616985321,
+      "learning_rate": 0.00021010903217122086,
+      "loss": 3.8846,
+      "step": 2618
+    },
+    {
+      "epoch": 3.34944,
+      "grad_norm": 0.504747211933136,
+      "learning_rate": 0.0002100686498855835,
+      "loss": 3.8775,
+      "step": 2619
+    },
+    {
+      "epoch": 3.35072,
+      "grad_norm": 0.496619313955307,
+      "learning_rate": 0.00021002826759994615,
+      "loss": 3.8405,
+      "step": 2620
+    },
+    {
+      "epoch": 3.352,
+      "grad_norm": 0.5143598914146423,
+      "learning_rate": 0.00020998788531430878,
+      "loss": 3.8538,
+      "step": 2621
+    },
+    {
+      "epoch": 3.35328,
+      "grad_norm": 0.49931085109710693,
+      "learning_rate": 0.0002099475030286714,
+      "loss": 3.8484,
+      "step": 2622
+    },
+    {
+      "epoch": 3.35456,
+      "grad_norm": 0.49670740962028503,
+      "learning_rate": 0.00020990712074303404,
+      "loss": 3.8549,
+      "step": 2623
+    },
+    {
+      "epoch": 3.35584,
+      "grad_norm": 0.5194510221481323,
+      "learning_rate": 0.00020986673845739664,
+      "loss": 3.9321,
+      "step": 2624
+    },
+    {
+      "epoch": 3.35712,
+      "grad_norm": 0.4933635890483856,
+      "learning_rate": 0.0002098263561717593,
+      "loss": 3.9203,
+      "step": 2625
+    },
+    {
+      "epoch": 3.3584,
+      "grad_norm": 0.5265275835990906,
+      "learning_rate": 0.00020978597388612193,
+      "loss": 3.7854,
+      "step": 2626
+    },
+    {
+      "epoch": 3.35968,
+      "grad_norm": 0.5067412853240967,
+      "learning_rate": 0.00020974559160048456,
+      "loss": 3.8297,
+      "step": 2627
+    },
+    {
+      "epoch": 3.36096,
+      "grad_norm": 0.5101473927497864,
+      "learning_rate": 0.0002097052093148472,
+      "loss": 3.8358,
+      "step": 2628
+    },
+    {
+      "epoch": 3.36224,
+      "grad_norm": 0.5028688311576843,
+      "learning_rate": 0.00020966482702920985,
+      "loss": 3.8475,
+      "step": 2629
+    },
+    {
+      "epoch": 3.36352,
+      "grad_norm": 0.5051252245903015,
+      "learning_rate": 0.00020962444474357248,
+      "loss": 3.8307,
+      "step": 2630
+    },
+    {
+      "epoch": 3.3648,
+      "grad_norm": 0.490156888961792,
+      "learning_rate": 0.0002095840624579351,
+      "loss": 3.843,
+      "step": 2631
+    },
+    {
+      "epoch": 3.36608,
+      "grad_norm": 0.49856409430503845,
+      "learning_rate": 0.0002095436801722977,
+      "loss": 3.8606,
+      "step": 2632
+    },
+    {
+      "epoch": 3.36736,
+      "grad_norm": 0.47945141792297363,
+      "learning_rate": 0.00020950329788666037,
+      "loss": 3.8623,
+      "step": 2633
+    },
+    {
+      "epoch": 3.36864,
+      "grad_norm": 0.5224535465240479,
+      "learning_rate": 0.000209462915601023,
+      "loss": 3.8351,
+      "step": 2634
+    },
+    {
+      "epoch": 3.36992,
+      "grad_norm": 0.49152296781539917,
+      "learning_rate": 0.00020942253331538563,
+      "loss": 3.812,
+      "step": 2635
+    },
+    {
+      "epoch": 3.3712,
+      "grad_norm": 0.4920552968978882,
+      "learning_rate": 0.00020938215102974826,
+      "loss": 3.8568,
+      "step": 2636
+    },
+    {
+      "epoch": 3.37248,
+      "grad_norm": 0.4960016906261444,
+      "learning_rate": 0.0002093417687441109,
+      "loss": 3.8487,
+      "step": 2637
+    },
+    {
+      "epoch": 3.37376,
+      "grad_norm": 0.5091875195503235,
+      "learning_rate": 0.00020930138645847355,
+      "loss": 3.8627,
+      "step": 2638
+    },
+    {
+      "epoch": 3.37504,
+      "grad_norm": 0.4994186758995056,
+      "learning_rate": 0.00020926100417283618,
+      "loss": 3.8548,
+      "step": 2639
+    },
+    {
+      "epoch": 3.3763199999999998,
+      "grad_norm": 0.49067172408103943,
+      "learning_rate": 0.00020922062188719878,
+      "loss": 3.8619,
+      "step": 2640
+    },
+    {
+      "epoch": 3.3776,
+      "grad_norm": 0.4815825819969177,
+      "learning_rate": 0.0002091802396015614,
+      "loss": 3.8202,
+      "step": 2641
+    },
+    {
+      "epoch": 3.37888,
+      "grad_norm": 0.501998245716095,
+      "learning_rate": 0.00020913985731592407,
+      "loss": 3.8243,
+      "step": 2642
+    },
+    {
+      "epoch": 3.38016,
+      "grad_norm": 0.49832189083099365,
+      "learning_rate": 0.0002090994750302867,
+      "loss": 3.8018,
+      "step": 2643
+    },
+    {
+      "epoch": 3.38144,
+      "grad_norm": 0.5382563471794128,
+      "learning_rate": 0.00020905909274464933,
+      "loss": 3.8121,
+      "step": 2644
+    },
+    {
+      "epoch": 3.38272,
+      "grad_norm": 0.5151340961456299,
+      "learning_rate": 0.00020901871045901196,
+      "loss": 3.8461,
+      "step": 2645
+    },
+    {
+      "epoch": 3.384,
+      "grad_norm": 0.5148212909698486,
+      "learning_rate": 0.00020897832817337461,
+      "loss": 3.9021,
+      "step": 2646
+    },
+    {
+      "epoch": 3.38528,
+      "grad_norm": 0.49888908863067627,
+      "learning_rate": 0.00020893794588773724,
+      "loss": 3.9323,
+      "step": 2647
+    },
+    {
+      "epoch": 3.3865600000000002,
+      "grad_norm": 0.5344513058662415,
+      "learning_rate": 0.00020889756360209985,
+      "loss": 3.8421,
+      "step": 2648
+    },
+    {
+      "epoch": 3.38784,
+      "grad_norm": 0.5026503205299377,
+      "learning_rate": 0.00020885718131646248,
+      "loss": 3.7643,
+      "step": 2649
+    },
+    {
+      "epoch": 3.38912,
+      "grad_norm": 0.4816747009754181,
+      "learning_rate": 0.0002088167990308251,
+      "loss": 3.7415,
+      "step": 2650
+    },
+    {
+      "epoch": 3.3904,
+      "grad_norm": 0.5043998956680298,
+      "learning_rate": 0.00020877641674518777,
+      "loss": 3.8523,
+      "step": 2651
+    },
+    {
+      "epoch": 3.39168,
+      "grad_norm": 0.5302547812461853,
+      "learning_rate": 0.0002087360344595504,
+      "loss": 3.8868,
+      "step": 2652
+    },
+    {
+      "epoch": 3.39296,
+      "grad_norm": 0.47889089584350586,
+      "learning_rate": 0.00020869565217391303,
+      "loss": 3.7999,
+      "step": 2653
+    },
+    {
+      "epoch": 3.39424,
+      "grad_norm": 0.5710118412971497,
+      "learning_rate": 0.00020865526988827566,
+      "loss": 3.8922,
+      "step": 2654
+    },
+    {
+      "epoch": 3.39552,
+      "grad_norm": 0.504432201385498,
+      "learning_rate": 0.0002086148876026383,
+      "loss": 3.8663,
+      "step": 2655
+    },
+    {
+      "epoch": 3.3968,
+      "grad_norm": 0.5317485928535461,
+      "learning_rate": 0.00020857450531700092,
+      "loss": 3.8964,
+      "step": 2656
+    },
+    {
+      "epoch": 3.39808,
+      "grad_norm": 0.5189566612243652,
+      "learning_rate": 0.00020853412303136355,
+      "loss": 3.8165,
+      "step": 2657
+    },
+    {
+      "epoch": 3.39936,
+      "grad_norm": 0.5216357111930847,
+      "learning_rate": 0.00020849374074572618,
+      "loss": 3.8114,
+      "step": 2658
+    },
+    {
+      "epoch": 3.40064,
+      "grad_norm": 0.517992377281189,
+      "learning_rate": 0.00020845335846008883,
+      "loss": 3.8755,
+      "step": 2659
+    },
+    {
+      "epoch": 3.40192,
+      "grad_norm": 0.5476480722427368,
+      "learning_rate": 0.00020841297617445146,
+      "loss": 3.8006,
+      "step": 2660
+    },
+    {
+      "epoch": 3.4032,
+      "grad_norm": 0.5155659914016724,
+      "learning_rate": 0.0002083725938888141,
+      "loss": 3.7955,
+      "step": 2661
+    },
+    {
+      "epoch": 3.40448,
+      "grad_norm": 0.5250527262687683,
+      "learning_rate": 0.00020833221160317672,
+      "loss": 3.9009,
+      "step": 2662
+    },
+    {
+      "epoch": 3.40576,
+      "grad_norm": 0.5005857944488525,
+      "learning_rate": 0.00020829182931753933,
+      "loss": 3.8746,
+      "step": 2663
+    },
+    {
+      "epoch": 3.40704,
+      "grad_norm": 0.5127919316291809,
+      "learning_rate": 0.000208251447031902,
+      "loss": 3.8685,
+      "step": 2664
+    },
+    {
+      "epoch": 3.40832,
+      "grad_norm": 0.49947747588157654,
+      "learning_rate": 0.0002082110647462646,
+      "loss": 3.8362,
+      "step": 2665
+    },
+    {
+      "epoch": 3.4096,
+      "grad_norm": 0.48780733346939087,
+      "learning_rate": 0.00020817068246062724,
+      "loss": 3.821,
+      "step": 2666
+    },
+    {
+      "epoch": 3.41088,
+      "grad_norm": 0.5010076761245728,
+      "learning_rate": 0.00020813030017498987,
+      "loss": 3.7895,
+      "step": 2667
+    },
+    {
+      "epoch": 3.41216,
+      "grad_norm": 0.5000413060188293,
+      "learning_rate": 0.00020808991788935253,
+      "loss": 3.8729,
+      "step": 2668
+    },
+    {
+      "epoch": 3.41344,
+      "grad_norm": 0.5171912312507629,
+      "learning_rate": 0.00020804953560371516,
+      "loss": 3.8008,
+      "step": 2669
+    },
+    {
+      "epoch": 3.41472,
+      "grad_norm": 0.49489569664001465,
+      "learning_rate": 0.0002080091533180778,
+      "loss": 3.8762,
+      "step": 2670
+    },
+    {
+      "epoch": 3.416,
+      "grad_norm": 0.49014776945114136,
+      "learning_rate": 0.0002079687710324404,
+      "loss": 3.7536,
+      "step": 2671
+    },
+    {
+      "epoch": 3.41728,
+      "grad_norm": 0.5062114596366882,
+      "learning_rate": 0.00020792838874680308,
+      "loss": 3.8479,
+      "step": 2672
+    },
+    {
+      "epoch": 3.41856,
+      "grad_norm": 0.49555784463882446,
+      "learning_rate": 0.00020788800646116568,
+      "loss": 3.851,
+      "step": 2673
+    },
+    {
+      "epoch": 3.4198399999999998,
+      "grad_norm": 0.48227742314338684,
+      "learning_rate": 0.0002078476241755283,
+      "loss": 3.8026,
+      "step": 2674
+    },
+    {
+      "epoch": 3.42112,
+      "grad_norm": 0.49328896403312683,
+      "learning_rate": 0.00020780724188989094,
+      "loss": 3.8565,
+      "step": 2675
+    },
+    {
+      "epoch": 3.4224,
+      "grad_norm": 0.4922971725463867,
+      "learning_rate": 0.00020776685960425357,
+      "loss": 3.8081,
+      "step": 2676
+    },
+    {
+      "epoch": 3.42368,
+      "grad_norm": 0.5213683843612671,
+      "learning_rate": 0.00020772647731861623,
+      "loss": 3.7913,
+      "step": 2677
+    },
+    {
+      "epoch": 3.42496,
+      "grad_norm": 0.4958494007587433,
+      "learning_rate": 0.00020768609503297886,
+      "loss": 3.8678,
+      "step": 2678
+    },
+    {
+      "epoch": 3.42624,
+      "grad_norm": 0.4969683885574341,
+      "learning_rate": 0.0002076457127473415,
+      "loss": 3.7767,
+      "step": 2679
+    },
+    {
+      "epoch": 3.42752,
+      "grad_norm": 0.5043670535087585,
+      "learning_rate": 0.0002076053304617041,
+      "loss": 3.8217,
+      "step": 2680
+    },
+    {
+      "epoch": 3.4288,
+      "grad_norm": 0.4834843575954437,
+      "learning_rate": 0.00020756494817606675,
+      "loss": 3.7935,
+      "step": 2681
+    },
+    {
+      "epoch": 3.4300800000000002,
+      "grad_norm": 0.5033546686172485,
+      "learning_rate": 0.00020752456589042938,
+      "loss": 3.8507,
+      "step": 2682
+    },
+    {
+      "epoch": 3.43136,
+      "grad_norm": 0.4869885742664337,
+      "learning_rate": 0.000207484183604792,
+      "loss": 3.7742,
+      "step": 2683
+    },
+    {
+      "epoch": 3.43264,
+      "grad_norm": 0.5016533732414246,
+      "learning_rate": 0.00020744380131915464,
+      "loss": 3.8161,
+      "step": 2684
+    },
+    {
+      "epoch": 3.43392,
+      "grad_norm": 0.4721163809299469,
+      "learning_rate": 0.0002074034190335173,
+      "loss": 3.8743,
+      "step": 2685
+    },
+    {
+      "epoch": 3.4352,
+      "grad_norm": 0.5129477381706238,
+      "learning_rate": 0.00020736303674787993,
+      "loss": 3.9012,
+      "step": 2686
+    },
+    {
+      "epoch": 3.43648,
+      "grad_norm": 0.5066404938697815,
+      "learning_rate": 0.00020732265446224256,
+      "loss": 3.8459,
+      "step": 2687
+    },
+    {
+      "epoch": 3.43776,
+      "grad_norm": 0.5318638682365417,
+      "learning_rate": 0.00020728227217660516,
+      "loss": 3.744,
+      "step": 2688
+    },
+    {
+      "epoch": 3.43904,
+      "grad_norm": 0.497341513633728,
+      "learning_rate": 0.00020724188989096782,
+      "loss": 3.8856,
+      "step": 2689
+    },
+    {
+      "epoch": 3.44032,
+      "grad_norm": 0.5572753548622131,
+      "learning_rate": 0.00020720150760533045,
+      "loss": 3.7982,
+      "step": 2690
+    },
+    {
+      "epoch": 3.4416,
+      "grad_norm": 0.4999849498271942,
+      "learning_rate": 0.00020716112531969308,
+      "loss": 3.7728,
+      "step": 2691
+    },
+    {
+      "epoch": 3.44288,
+      "grad_norm": 0.5322354435920715,
+      "learning_rate": 0.0002071207430340557,
+      "loss": 3.9,
+      "step": 2692
+    },
+    {
+      "epoch": 3.44416,
+      "grad_norm": 0.5226873755455017,
+      "learning_rate": 0.00020708036074841834,
+      "loss": 3.8428,
+      "step": 2693
+    },
+    {
+      "epoch": 3.44544,
+      "grad_norm": 0.5086526870727539,
+      "learning_rate": 0.000207039978462781,
+      "loss": 3.8667,
+      "step": 2694
+    },
+    {
+      "epoch": 3.44672,
+      "grad_norm": 0.4990271329879761,
+      "learning_rate": 0.00020699959617714362,
+      "loss": 3.8685,
+      "step": 2695
+    },
+    {
+      "epoch": 3.448,
+      "grad_norm": 0.5120396018028259,
+      "learning_rate": 0.00020695921389150623,
+      "loss": 3.8694,
+      "step": 2696
+    },
+    {
+      "epoch": 3.44928,
+      "grad_norm": 0.5174608826637268,
+      "learning_rate": 0.00020691883160586886,
+      "loss": 3.8495,
+      "step": 2697
+    },
+    {
+      "epoch": 3.45056,
+      "grad_norm": 0.5169567465782166,
+      "learning_rate": 0.00020687844932023151,
+      "loss": 3.8418,
+      "step": 2698
+    },
+    {
+      "epoch": 3.45184,
+      "grad_norm": 0.5117883682250977,
+      "learning_rate": 0.00020683806703459414,
+      "loss": 3.8707,
+      "step": 2699
+    },
+    {
+      "epoch": 3.45312,
+      "grad_norm": 0.5186121463775635,
+      "learning_rate": 0.00020679768474895677,
+      "loss": 3.8601,
+      "step": 2700
+    },
+    {
+      "epoch": 3.4544,
+      "grad_norm": 0.5014731884002686,
+      "learning_rate": 0.0002067573024633194,
+      "loss": 3.8657,
+      "step": 2701
+    },
+    {
+      "epoch": 3.45568,
+      "grad_norm": 0.4928348958492279,
+      "learning_rate": 0.00020671692017768206,
+      "loss": 3.7956,
+      "step": 2702
+    },
+    {
+      "epoch": 3.45696,
+      "grad_norm": 0.4971044063568115,
+      "learning_rate": 0.0002066765378920447,
+      "loss": 3.9533,
+      "step": 2703
+    },
+    {
+      "epoch": 3.45824,
+      "grad_norm": 0.5187544226646423,
+      "learning_rate": 0.0002066361556064073,
+      "loss": 3.847,
+      "step": 2704
+    },
+    {
+      "epoch": 3.45952,
+      "grad_norm": 0.4827615022659302,
+      "learning_rate": 0.00020659577332076992,
+      "loss": 3.8399,
+      "step": 2705
+    },
+    {
+      "epoch": 3.4608,
+      "grad_norm": 0.5062766671180725,
+      "learning_rate": 0.00020655539103513255,
+      "loss": 3.9054,
+      "step": 2706
+    },
+    {
+      "epoch": 3.46208,
+      "grad_norm": 0.49423202872276306,
+      "learning_rate": 0.0002065150087494952,
+      "loss": 3.8656,
+      "step": 2707
+    },
+    {
+      "epoch": 3.4633599999999998,
+      "grad_norm": 0.5079230666160583,
+      "learning_rate": 0.00020647462646385784,
+      "loss": 3.7847,
+      "step": 2708
+    },
+    {
+      "epoch": 3.46464,
+      "grad_norm": 0.5124461650848389,
+      "learning_rate": 0.00020643424417822047,
+      "loss": 3.7327,
+      "step": 2709
+    },
+    {
+      "epoch": 3.46592,
+      "grad_norm": 0.5108417868614197,
+      "learning_rate": 0.0002063938618925831,
+      "loss": 3.848,
+      "step": 2710
+    },
+    {
+      "epoch": 3.4672,
+      "grad_norm": 0.5090741515159607,
+      "learning_rate": 0.00020635347960694576,
+      "loss": 3.7348,
+      "step": 2711
+    },
+    {
+      "epoch": 3.46848,
+      "grad_norm": 0.4850289523601532,
+      "learning_rate": 0.00020631309732130836,
+      "loss": 3.8364,
+      "step": 2712
+    },
+    {
+      "epoch": 3.46976,
+      "grad_norm": 0.5068538188934326,
+      "learning_rate": 0.000206272715035671,
+      "loss": 3.8469,
+      "step": 2713
+    },
+    {
+      "epoch": 3.47104,
+      "grad_norm": 0.5119503140449524,
+      "learning_rate": 0.00020623233275003362,
+      "loss": 3.9369,
+      "step": 2714
+    },
+    {
+      "epoch": 3.47232,
+      "grad_norm": 0.49882522225379944,
+      "learning_rate": 0.00020619195046439628,
+      "loss": 3.8124,
+      "step": 2715
+    },
+    {
+      "epoch": 3.4736000000000002,
+      "grad_norm": 0.5284246802330017,
+      "learning_rate": 0.0002061515681787589,
+      "loss": 3.7402,
+      "step": 2716
+    },
+    {
+      "epoch": 3.47488,
+      "grad_norm": 0.5059674382209778,
+      "learning_rate": 0.00020611118589312154,
+      "loss": 3.8813,
+      "step": 2717
+    },
+    {
+      "epoch": 3.47616,
+      "grad_norm": 0.5152750015258789,
+      "learning_rate": 0.00020607080360748417,
+      "loss": 3.8808,
+      "step": 2718
+    },
+    {
+      "epoch": 3.47744,
+      "grad_norm": 0.5104813575744629,
+      "learning_rate": 0.00020603042132184677,
+      "loss": 3.7924,
+      "step": 2719
+    },
+    {
+      "epoch": 3.47872,
+      "grad_norm": 0.500645637512207,
+      "learning_rate": 0.00020599003903620943,
+      "loss": 3.8189,
+      "step": 2720
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.511396050453186,
+      "learning_rate": 0.00020594965675057206,
+      "loss": 3.8472,
+      "step": 2721
+    },
+    {
+      "epoch": 3.48128,
+      "grad_norm": 0.5173667073249817,
+      "learning_rate": 0.0002059092744649347,
+      "loss": 3.8637,
+      "step": 2722
+    },
+    {
+      "epoch": 3.48256,
+      "grad_norm": 0.49966612458229065,
+      "learning_rate": 0.00020586889217929732,
+      "loss": 3.8889,
+      "step": 2723
+    },
+    {
+      "epoch": 3.48384,
+      "grad_norm": 0.4898408055305481,
+      "learning_rate": 0.00020582850989365998,
+      "loss": 3.8718,
+      "step": 2724
+    },
+    {
+      "epoch": 3.48512,
+      "grad_norm": 0.5065445303916931,
+      "learning_rate": 0.0002057881276080226,
+      "loss": 3.8627,
+      "step": 2725
+    },
+    {
+      "epoch": 3.4864,
+      "grad_norm": 0.485765278339386,
+      "learning_rate": 0.00020574774532238524,
+      "loss": 3.7733,
+      "step": 2726
+    },
+    {
+      "epoch": 3.48768,
+      "grad_norm": 0.5096103549003601,
+      "learning_rate": 0.00020570736303674784,
+      "loss": 3.7589,
+      "step": 2727
+    },
+    {
+      "epoch": 3.48896,
+      "grad_norm": 0.5073655843734741,
+      "learning_rate": 0.0002056669807511105,
+      "loss": 3.8593,
+      "step": 2728
+    },
+    {
+      "epoch": 3.49024,
+      "grad_norm": 0.48420706391334534,
+      "learning_rate": 0.00020562659846547313,
+      "loss": 3.7242,
+      "step": 2729
+    },
+    {
+      "epoch": 3.49152,
+      "grad_norm": 0.5061640739440918,
+      "learning_rate": 0.00020558621617983576,
+      "loss": 3.9017,
+      "step": 2730
+    },
+    {
+      "epoch": 3.4928,
+      "grad_norm": 0.5040037035942078,
+      "learning_rate": 0.0002055458338941984,
+      "loss": 3.8566,
+      "step": 2731
+    },
+    {
+      "epoch": 3.49408,
+      "grad_norm": 0.4816730320453644,
+      "learning_rate": 0.00020550545160856102,
+      "loss": 3.8164,
+      "step": 2732
+    },
+    {
+      "epoch": 3.49536,
+      "grad_norm": 0.5050880908966064,
+      "learning_rate": 0.00020546506932292368,
+      "loss": 3.899,
+      "step": 2733
+    },
+    {
+      "epoch": 3.49664,
+      "grad_norm": 0.47876450419425964,
+      "learning_rate": 0.0002054246870372863,
+      "loss": 3.8351,
+      "step": 2734
+    },
+    {
+      "epoch": 3.49792,
+      "grad_norm": 0.5011029243469238,
+      "learning_rate": 0.0002053843047516489,
+      "loss": 3.7863,
+      "step": 2735
+    },
+    {
+      "epoch": 3.4992,
+      "grad_norm": 0.509458065032959,
+      "learning_rate": 0.00020534392246601154,
+      "loss": 3.889,
+      "step": 2736
+    },
+    {
+      "epoch": 3.50048,
+      "grad_norm": 0.4975719749927521,
+      "learning_rate": 0.0002053035401803742,
+      "loss": 3.9014,
+      "step": 2737
+    },
+    {
+      "epoch": 3.50176,
+      "grad_norm": 0.49265387654304504,
+      "learning_rate": 0.00020526315789473683,
+      "loss": 3.839,
+      "step": 2738
+    },
+    {
+      "epoch": 3.50304,
+      "grad_norm": 0.5019585490226746,
+      "learning_rate": 0.00020522277560909946,
+      "loss": 3.8466,
+      "step": 2739
+    },
+    {
+      "epoch": 3.50432,
+      "grad_norm": 0.49570876359939575,
+      "learning_rate": 0.00020518239332346209,
+      "loss": 3.8707,
+      "step": 2740
+    },
+    {
+      "epoch": 3.5056000000000003,
+      "grad_norm": 0.5000244975090027,
+      "learning_rate": 0.00020514201103782474,
+      "loss": 3.7987,
+      "step": 2741
+    },
+    {
+      "epoch": 3.5068799999999998,
+      "grad_norm": 0.5056788325309753,
+      "learning_rate": 0.00020510162875218737,
+      "loss": 3.8735,
+      "step": 2742
+    },
+    {
+      "epoch": 3.50816,
+      "grad_norm": 0.5059694647789001,
+      "learning_rate": 0.00020506124646654998,
+      "loss": 3.8705,
+      "step": 2743
+    },
+    {
+      "epoch": 3.50944,
+      "grad_norm": 0.47248658537864685,
+      "learning_rate": 0.0002050208641809126,
+      "loss": 3.8288,
+      "step": 2744
+    },
+    {
+      "epoch": 3.51072,
+      "grad_norm": 0.5001699328422546,
+      "learning_rate": 0.00020498048189527524,
+      "loss": 3.8767,
+      "step": 2745
+    },
+    {
+      "epoch": 3.512,
+      "grad_norm": 0.5233591794967651,
+      "learning_rate": 0.0002049400996096379,
+      "loss": 3.8299,
+      "step": 2746
+    },
+    {
+      "epoch": 3.51328,
+      "grad_norm": 0.5596291422843933,
+      "learning_rate": 0.00020489971732400052,
+      "loss": 3.9301,
+      "step": 2747
+    },
+    {
+      "epoch": 3.51456,
+      "grad_norm": 0.5503137111663818,
+      "learning_rate": 0.00020485933503836315,
+      "loss": 3.8602,
+      "step": 2748
+    },
+    {
+      "epoch": 3.51584,
+      "grad_norm": 0.48577359318733215,
+      "learning_rate": 0.00020481895275272578,
+      "loss": 3.7755,
+      "step": 2749
+    },
+    {
+      "epoch": 3.5171200000000002,
+      "grad_norm": 0.5109073519706726,
+      "learning_rate": 0.00020477857046708844,
+      "loss": 3.8155,
+      "step": 2750
+    },
+    {
+      "epoch": 3.5183999999999997,
+      "grad_norm": 0.5062686800956726,
+      "learning_rate": 0.00020473818818145104,
+      "loss": 3.819,
+      "step": 2751
+    },
+    {
+      "epoch": 3.51968,
+      "grad_norm": 0.5061745047569275,
+      "learning_rate": 0.00020469780589581367,
+      "loss": 3.892,
+      "step": 2752
+    },
+    {
+      "epoch": 3.52096,
+      "grad_norm": 0.49665218591690063,
+      "learning_rate": 0.0002046574236101763,
+      "loss": 3.8608,
+      "step": 2753
+    },
+    {
+      "epoch": 3.52224,
+      "grad_norm": 0.511926531791687,
+      "learning_rate": 0.00020461704132453896,
+      "loss": 3.7848,
+      "step": 2754
+    },
+    {
+      "epoch": 3.52352,
+      "grad_norm": 0.4888019859790802,
+      "learning_rate": 0.0002045766590389016,
+      "loss": 3.7409,
+      "step": 2755
+    },
+    {
+      "epoch": 3.5248,
+      "grad_norm": 0.4757629334926605,
+      "learning_rate": 0.00020453627675326422,
+      "loss": 3.8083,
+      "step": 2756
+    },
+    {
+      "epoch": 3.52608,
+      "grad_norm": 0.48579922318458557,
+      "learning_rate": 0.00020449589446762685,
+      "loss": 3.8436,
+      "step": 2757
+    },
+    {
+      "epoch": 3.52736,
+      "grad_norm": 0.47741061449050903,
+      "learning_rate": 0.00020445551218198945,
+      "loss": 3.7938,
+      "step": 2758
+    },
+    {
+      "epoch": 3.52864,
+      "grad_norm": 0.5141867995262146,
+      "learning_rate": 0.00020441512989635214,
+      "loss": 3.8169,
+      "step": 2759
+    },
+    {
+      "epoch": 3.5299199999999997,
+      "grad_norm": 0.4805345833301544,
+      "learning_rate": 0.00020437474761071474,
+      "loss": 3.8487,
+      "step": 2760
+    },
+    {
+      "epoch": 3.5312,
+      "grad_norm": 0.5061795711517334,
+      "learning_rate": 0.00020433436532507737,
+      "loss": 3.811,
+      "step": 2761
+    },
+    {
+      "epoch": 3.53248,
+      "grad_norm": 0.48661288619041443,
+      "learning_rate": 0.00020429398303944,
+      "loss": 3.7974,
+      "step": 2762
+    },
+    {
+      "epoch": 3.53376,
+      "grad_norm": 0.5097336173057556,
+      "learning_rate": 0.00020425360075380266,
+      "loss": 3.8154,
+      "step": 2763
+    },
+    {
+      "epoch": 3.53504,
+      "grad_norm": 0.4881674349308014,
+      "learning_rate": 0.0002042132184681653,
+      "loss": 3.8714,
+      "step": 2764
+    },
+    {
+      "epoch": 3.53632,
+      "grad_norm": 0.5308511853218079,
+      "learning_rate": 0.00020417283618252792,
+      "loss": 3.786,
+      "step": 2765
+    },
+    {
+      "epoch": 3.5376,
+      "grad_norm": 0.48169809579849243,
+      "learning_rate": 0.00020413245389689055,
+      "loss": 3.8127,
+      "step": 2766
+    },
+    {
+      "epoch": 3.53888,
+      "grad_norm": 0.5140031576156616,
+      "learning_rate": 0.0002040920716112532,
+      "loss": 3.8552,
+      "step": 2767
+    },
+    {
+      "epoch": 3.54016,
+      "grad_norm": 0.4895153343677521,
+      "learning_rate": 0.0002040516893256158,
+      "loss": 3.8319,
+      "step": 2768
+    },
+    {
+      "epoch": 3.54144,
+      "grad_norm": 0.5244706869125366,
+      "learning_rate": 0.00020401130703997844,
+      "loss": 3.7663,
+      "step": 2769
+    },
+    {
+      "epoch": 3.54272,
+      "grad_norm": 0.5048360228538513,
+      "learning_rate": 0.00020397092475434107,
+      "loss": 3.8417,
+      "step": 2770
+    },
+    {
+      "epoch": 3.544,
+      "grad_norm": 0.515129029750824,
+      "learning_rate": 0.0002039305424687037,
+      "loss": 3.9146,
+      "step": 2771
+    },
+    {
+      "epoch": 3.54528,
+      "grad_norm": 0.5083556771278381,
+      "learning_rate": 0.00020389016018306636,
+      "loss": 3.8344,
+      "step": 2772
+    },
+    {
+      "epoch": 3.54656,
+      "grad_norm": 0.4877746105194092,
+      "learning_rate": 0.000203849777897429,
+      "loss": 3.8346,
+      "step": 2773
+    },
+    {
+      "epoch": 3.54784,
+      "grad_norm": 0.508374035358429,
+      "learning_rate": 0.00020380939561179162,
+      "loss": 3.7713,
+      "step": 2774
+    },
+    {
+      "epoch": 3.5491200000000003,
+      "grad_norm": 0.496378093957901,
+      "learning_rate": 0.00020376901332615422,
+      "loss": 3.8809,
+      "step": 2775
+    },
+    {
+      "epoch": 3.5504,
+      "grad_norm": 0.5039408206939697,
+      "learning_rate": 0.00020372863104051688,
+      "loss": 3.8309,
+      "step": 2776
+    },
+    {
+      "epoch": 3.55168,
+      "grad_norm": 0.496198445558548,
+      "learning_rate": 0.0002036882487548795,
+      "loss": 3.8634,
+      "step": 2777
+    },
+    {
+      "epoch": 3.55296,
+      "grad_norm": 0.48462727665901184,
+      "learning_rate": 0.00020364786646924214,
+      "loss": 3.8348,
+      "step": 2778
+    },
+    {
+      "epoch": 3.55424,
+      "grad_norm": 0.48560038208961487,
+      "learning_rate": 0.00020360748418360477,
+      "loss": 3.7579,
+      "step": 2779
+    },
+    {
+      "epoch": 3.55552,
+      "grad_norm": 0.502811074256897,
+      "learning_rate": 0.00020356710189796742,
+      "loss": 3.8567,
+      "step": 2780
+    },
+    {
+      "epoch": 3.5568,
+      "grad_norm": 0.4968845546245575,
+      "learning_rate": 0.00020352671961233005,
+      "loss": 3.8431,
+      "step": 2781
+    },
+    {
+      "epoch": 3.55808,
+      "grad_norm": 0.48642468452453613,
+      "learning_rate": 0.00020348633732669268,
+      "loss": 3.7647,
+      "step": 2782
+    },
+    {
+      "epoch": 3.55936,
+      "grad_norm": 0.4881822168827057,
+      "learning_rate": 0.0002034459550410553,
+      "loss": 3.792,
+      "step": 2783
+    },
+    {
+      "epoch": 3.5606400000000002,
+      "grad_norm": 0.49411314725875854,
+      "learning_rate": 0.00020340557275541792,
+      "loss": 3.794,
+      "step": 2784
+    },
+    {
+      "epoch": 3.5619199999999998,
+      "grad_norm": 0.48321542143821716,
+      "learning_rate": 0.00020336519046978057,
+      "loss": 3.8337,
+      "step": 2785
+    },
+    {
+      "epoch": 3.5632,
+      "grad_norm": 0.4860212504863739,
+      "learning_rate": 0.0002033248081841432,
+      "loss": 3.8145,
+      "step": 2786
+    },
+    {
+      "epoch": 3.56448,
+      "grad_norm": 0.4979793429374695,
+      "learning_rate": 0.00020328442589850583,
+      "loss": 3.8073,
+      "step": 2787
+    },
+    {
+      "epoch": 3.56576,
+      "grad_norm": 0.4866310954093933,
+      "learning_rate": 0.00020324404361286846,
+      "loss": 3.7945,
+      "step": 2788
+    },
+    {
+      "epoch": 3.56704,
+      "grad_norm": 0.516071617603302,
+      "learning_rate": 0.00020320366132723112,
+      "loss": 3.8237,
+      "step": 2789
+    },
+    {
+      "epoch": 3.56832,
+      "grad_norm": 0.5022051930427551,
+      "learning_rate": 0.00020316327904159375,
+      "loss": 3.8451,
+      "step": 2790
+    },
+    {
+      "epoch": 3.5696,
+      "grad_norm": 0.5036980509757996,
+      "learning_rate": 0.00020312289675595636,
+      "loss": 3.7898,
+      "step": 2791
+    },
+    {
+      "epoch": 3.57088,
+      "grad_norm": 0.5079161524772644,
+      "learning_rate": 0.00020308251447031899,
+      "loss": 3.7985,
+      "step": 2792
+    },
+    {
+      "epoch": 3.5721600000000002,
+      "grad_norm": 0.48699140548706055,
+      "learning_rate": 0.00020304213218468164,
+      "loss": 3.8116,
+      "step": 2793
+    },
+    {
+      "epoch": 3.5734399999999997,
+      "grad_norm": 0.5011927485466003,
+      "learning_rate": 0.00020300174989904427,
+      "loss": 3.9555,
+      "step": 2794
+    },
+    {
+      "epoch": 3.57472,
+      "grad_norm": 0.5062665939331055,
+      "learning_rate": 0.0002029613676134069,
+      "loss": 3.8183,
+      "step": 2795
+    },
+    {
+      "epoch": 3.576,
+      "grad_norm": 0.49101898074150085,
+      "learning_rate": 0.00020292098532776953,
+      "loss": 3.8428,
+      "step": 2796
+    },
+    {
+      "epoch": 3.57728,
+      "grad_norm": 0.5135859251022339,
+      "learning_rate": 0.00020288060304213216,
+      "loss": 3.8469,
+      "step": 2797
+    },
+    {
+      "epoch": 3.57856,
+      "grad_norm": 0.5152158141136169,
+      "learning_rate": 0.00020284022075649482,
+      "loss": 3.8385,
+      "step": 2798
+    },
+    {
+      "epoch": 3.57984,
+      "grad_norm": 0.4942494034767151,
+      "learning_rate": 0.00020279983847085742,
+      "loss": 3.9062,
+      "step": 2799
+    },
+    {
+      "epoch": 3.58112,
+      "grad_norm": 0.5052478313446045,
+      "learning_rate": 0.00020275945618522005,
+      "loss": 3.7949,
+      "step": 2800
+    },
+    {
+      "epoch": 3.5824,
+      "grad_norm": 0.5218126177787781,
+      "learning_rate": 0.00020271907389958268,
+      "loss": 3.8128,
+      "step": 2801
+    },
+    {
+      "epoch": 3.58368,
+      "grad_norm": 0.5123757719993591,
+      "learning_rate": 0.00020267869161394534,
+      "loss": 3.8098,
+      "step": 2802
+    },
+    {
+      "epoch": 3.58496,
+      "grad_norm": 0.48139896988868713,
+      "learning_rate": 0.00020263830932830797,
+      "loss": 3.8333,
+      "step": 2803
+    },
+    {
+      "epoch": 3.58624,
+      "grad_norm": 0.494536817073822,
+      "learning_rate": 0.0002025979270426706,
+      "loss": 3.7192,
+      "step": 2804
+    },
+    {
+      "epoch": 3.58752,
+      "grad_norm": 0.4789034128189087,
+      "learning_rate": 0.00020255754475703323,
+      "loss": 3.7481,
+      "step": 2805
+    },
+    {
+      "epoch": 3.5888,
+      "grad_norm": 0.4984082877635956,
+      "learning_rate": 0.0002025171624713959,
+      "loss": 3.8158,
+      "step": 2806
+    },
+    {
+      "epoch": 3.59008,
+      "grad_norm": 0.48809587955474854,
+      "learning_rate": 0.0002024767801857585,
+      "loss": 3.8401,
+      "step": 2807
+    },
+    {
+      "epoch": 3.59136,
+      "grad_norm": 0.5058606863021851,
+      "learning_rate": 0.00020243639790012112,
+      "loss": 3.8788,
+      "step": 2808
+    },
+    {
+      "epoch": 3.59264,
+      "grad_norm": 0.5134665369987488,
+      "learning_rate": 0.00020239601561448375,
+      "loss": 3.8255,
+      "step": 2809
+    },
+    {
+      "epoch": 3.59392,
+      "grad_norm": 0.4707070589065552,
+      "learning_rate": 0.0002023556333288464,
+      "loss": 3.7997,
+      "step": 2810
+    },
+    {
+      "epoch": 3.5952,
+      "grad_norm": 0.4775642454624176,
+      "learning_rate": 0.00020231525104320904,
+      "loss": 3.8048,
+      "step": 2811
+    },
+    {
+      "epoch": 3.59648,
+      "grad_norm": 0.5038156509399414,
+      "learning_rate": 0.00020227486875757167,
+      "loss": 3.8541,
+      "step": 2812
+    },
+    {
+      "epoch": 3.59776,
+      "grad_norm": 0.5030291676521301,
+      "learning_rate": 0.0002022344864719343,
+      "loss": 3.8111,
+      "step": 2813
+    },
+    {
+      "epoch": 3.59904,
+      "grad_norm": 0.5007965564727783,
+      "learning_rate": 0.0002021941041862969,
+      "loss": 3.7762,
+      "step": 2814
+    },
+    {
+      "epoch": 3.60032,
+      "grad_norm": 0.49890565872192383,
+      "learning_rate": 0.00020215372190065956,
+      "loss": 3.7919,
+      "step": 2815
+    },
+    {
+      "epoch": 3.6016,
+      "grad_norm": 0.48519954085350037,
+      "learning_rate": 0.0002021133396150222,
+      "loss": 3.8374,
+      "step": 2816
+    },
+    {
+      "epoch": 3.60288,
+      "grad_norm": 0.5019754767417908,
+      "learning_rate": 0.00020207295732938482,
+      "loss": 3.7988,
+      "step": 2817
+    },
+    {
+      "epoch": 3.6041600000000003,
+      "grad_norm": 0.4958573877811432,
+      "learning_rate": 0.00020203257504374745,
+      "loss": 3.8074,
+      "step": 2818
+    },
+    {
+      "epoch": 3.6054399999999998,
+      "grad_norm": 0.5017514228820801,
+      "learning_rate": 0.0002019921927581101,
+      "loss": 3.819,
+      "step": 2819
+    },
+    {
+      "epoch": 3.60672,
+      "grad_norm": 0.5000422596931458,
+      "learning_rate": 0.00020195181047247274,
+      "loss": 3.8453,
+      "step": 2820
+    },
+    {
+      "epoch": 3.608,
+      "grad_norm": 0.4973451793193817,
+      "learning_rate": 0.00020191142818683537,
+      "loss": 3.7835,
+      "step": 2821
+    },
+    {
+      "epoch": 3.60928,
+      "grad_norm": 0.4728100001811981,
+      "learning_rate": 0.00020187104590119797,
+      "loss": 3.8969,
+      "step": 2822
+    },
+    {
+      "epoch": 3.61056,
+      "grad_norm": 0.5011451840400696,
+      "learning_rate": 0.00020183066361556063,
+      "loss": 3.8835,
+      "step": 2823
+    },
+    {
+      "epoch": 3.61184,
+      "grad_norm": 0.5048329830169678,
+      "learning_rate": 0.00020179028132992326,
+      "loss": 3.8225,
+      "step": 2824
+    },
+    {
+      "epoch": 3.61312,
+      "grad_norm": 0.5037258863449097,
+      "learning_rate": 0.00020174989904428589,
+      "loss": 3.7936,
+      "step": 2825
+    },
+    {
+      "epoch": 3.6144,
+      "grad_norm": 0.4970305263996124,
+      "learning_rate": 0.00020170951675864852,
+      "loss": 3.8651,
+      "step": 2826
+    },
+    {
+      "epoch": 3.6156800000000002,
+      "grad_norm": 0.5086222290992737,
+      "learning_rate": 0.00020166913447301115,
+      "loss": 3.7828,
+      "step": 2827
+    },
+    {
+      "epoch": 3.6169599999999997,
+      "grad_norm": 0.4862697720527649,
+      "learning_rate": 0.0002016287521873738,
+      "loss": 3.7793,
+      "step": 2828
+    },
+    {
+      "epoch": 3.61824,
+      "grad_norm": 0.5067240595817566,
+      "learning_rate": 0.00020158836990173643,
+      "loss": 3.8363,
+      "step": 2829
+    },
+    {
+      "epoch": 3.61952,
+      "grad_norm": 0.5017095804214478,
+      "learning_rate": 0.00020154798761609904,
+      "loss": 3.8429,
+      "step": 2830
+    },
+    {
+      "epoch": 3.6208,
+      "grad_norm": 0.4909166991710663,
+      "learning_rate": 0.00020150760533046167,
+      "loss": 3.8017,
+      "step": 2831
+    },
+    {
+      "epoch": 3.62208,
+      "grad_norm": 0.5125616192817688,
+      "learning_rate": 0.00020146722304482432,
+      "loss": 3.7553,
+      "step": 2832
+    },
+    {
+      "epoch": 3.62336,
+      "grad_norm": 0.48391321301460266,
+      "learning_rate": 0.00020142684075918695,
+      "loss": 3.8041,
+      "step": 2833
+    },
+    {
+      "epoch": 3.62464,
+      "grad_norm": 0.4965222477912903,
+      "learning_rate": 0.00020138645847354958,
+      "loss": 3.8016,
+      "step": 2834
+    },
+    {
+      "epoch": 3.62592,
+      "grad_norm": 0.4799792170524597,
+      "learning_rate": 0.00020134607618791221,
+      "loss": 3.8301,
+      "step": 2835
+    },
+    {
+      "epoch": 3.6272,
+      "grad_norm": 0.5186593532562256,
+      "learning_rate": 0.00020130569390227487,
+      "loss": 3.7594,
+      "step": 2836
+    },
+    {
+      "epoch": 3.62848,
+      "grad_norm": 0.4939570426940918,
+      "learning_rate": 0.0002012653116166375,
+      "loss": 3.8343,
+      "step": 2837
+    },
+    {
+      "epoch": 3.62976,
+      "grad_norm": 0.4945909082889557,
+      "learning_rate": 0.0002012249293310001,
+      "loss": 3.8158,
+      "step": 2838
+    },
+    {
+      "epoch": 3.63104,
+      "grad_norm": 0.5133264660835266,
+      "learning_rate": 0.00020118454704536273,
+      "loss": 3.8507,
+      "step": 2839
+    },
+    {
+      "epoch": 3.63232,
+      "grad_norm": 0.4962891936302185,
+      "learning_rate": 0.00020114416475972536,
+      "loss": 3.8135,
+      "step": 2840
+    },
+    {
+      "epoch": 3.6336,
+      "grad_norm": 0.5103831887245178,
+      "learning_rate": 0.00020110378247408802,
+      "loss": 3.7948,
+      "step": 2841
+    },
+    {
+      "epoch": 3.63488,
+      "grad_norm": 0.5109889507293701,
+      "learning_rate": 0.00020106340018845065,
+      "loss": 3.8751,
+      "step": 2842
+    },
+    {
+      "epoch": 3.63616,
+      "grad_norm": 0.4803555905818939,
+      "learning_rate": 0.00020102301790281328,
+      "loss": 3.7989,
+      "step": 2843
+    },
+    {
+      "epoch": 3.63744,
+      "grad_norm": 0.494994193315506,
+      "learning_rate": 0.0002009826356171759,
+      "loss": 3.8933,
+      "step": 2844
+    },
+    {
+      "epoch": 3.63872,
+      "grad_norm": 0.4815881848335266,
+      "learning_rate": 0.00020094225333153857,
+      "loss": 3.7667,
+      "step": 2845
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.5187839865684509,
+      "learning_rate": 0.0002009018710459012,
+      "loss": 3.8322,
+      "step": 2846
+    },
+    {
+      "epoch": 3.64128,
+      "grad_norm": 0.5170251727104187,
+      "learning_rate": 0.0002008614887602638,
+      "loss": 3.8624,
+      "step": 2847
+    },
+    {
+      "epoch": 3.64256,
+      "grad_norm": 0.4987940490245819,
+      "learning_rate": 0.00020082110647462643,
+      "loss": 3.8021,
+      "step": 2848
+    },
+    {
+      "epoch": 3.64384,
+      "grad_norm": 0.5137864351272583,
+      "learning_rate": 0.0002007807241889891,
+      "loss": 3.8342,
+      "step": 2849
+    },
+    {
+      "epoch": 3.64512,
+      "grad_norm": 0.4848984479904175,
+      "learning_rate": 0.00020074034190335172,
+      "loss": 3.7829,
+      "step": 2850
+    },
+    {
+      "epoch": 3.6464,
+      "grad_norm": 0.5238636136054993,
+      "learning_rate": 0.00020069995961771435,
+      "loss": 3.8299,
+      "step": 2851
+    },
+    {
+      "epoch": 3.6476800000000003,
+      "grad_norm": 0.49876147508621216,
+      "learning_rate": 0.00020065957733207698,
+      "loss": 3.7739,
+      "step": 2852
+    },
+    {
+      "epoch": 3.6489599999999998,
+      "grad_norm": 0.5346967577934265,
+      "learning_rate": 0.00020061919504643958,
+      "loss": 3.8934,
+      "step": 2853
+    },
+    {
+      "epoch": 3.65024,
+      "grad_norm": 0.4966239631175995,
+      "learning_rate": 0.00020057881276080227,
+      "loss": 3.9085,
+      "step": 2854
+    },
+    {
+      "epoch": 3.65152,
+      "grad_norm": 0.5036845207214355,
+      "learning_rate": 0.00020053843047516487,
+      "loss": 3.7837,
+      "step": 2855
+    },
+    {
+      "epoch": 3.6528,
+      "grad_norm": 0.49921709299087524,
+      "learning_rate": 0.0002004980481895275,
+      "loss": 3.8029,
+      "step": 2856
+    },
+    {
+      "epoch": 3.65408,
+      "grad_norm": 0.5502278804779053,
+      "learning_rate": 0.00020045766590389013,
+      "loss": 3.8748,
+      "step": 2857
+    },
+    {
+      "epoch": 3.65536,
+      "grad_norm": 0.4963222146034241,
+      "learning_rate": 0.0002004172836182528,
+      "loss": 3.7893,
+      "step": 2858
+    },
+    {
+      "epoch": 3.65664,
+      "grad_norm": 0.5209027528762817,
+      "learning_rate": 0.00020037690133261542,
+      "loss": 3.7979,
+      "step": 2859
+    },
+    {
+      "epoch": 3.65792,
+      "grad_norm": 0.4983651340007782,
+      "learning_rate": 0.00020033651904697805,
+      "loss": 3.8321,
+      "step": 2860
+    },
+    {
+      "epoch": 3.6592000000000002,
+      "grad_norm": 0.6087284088134766,
+      "learning_rate": 0.00020029613676134068,
+      "loss": 3.8507,
+      "step": 2861
+    },
+    {
+      "epoch": 3.6604799999999997,
+      "grad_norm": 0.5027658343315125,
+      "learning_rate": 0.00020025575447570333,
+      "loss": 3.8565,
+      "step": 2862
+    },
+    {
+      "epoch": 3.66176,
+      "grad_norm": 0.5117068886756897,
+      "learning_rate": 0.00020021537219006594,
+      "loss": 3.825,
+      "step": 2863
+    },
+    {
+      "epoch": 3.66304,
+      "grad_norm": 0.5000763535499573,
+      "learning_rate": 0.00020017498990442857,
+      "loss": 3.8242,
+      "step": 2864
+    },
+    {
+      "epoch": 3.66432,
+      "grad_norm": 0.5197402834892273,
+      "learning_rate": 0.0002001346076187912,
+      "loss": 3.8651,
+      "step": 2865
+    },
+    {
+      "epoch": 3.6656,
+      "grad_norm": 0.5055534839630127,
+      "learning_rate": 0.00020009422533315383,
+      "loss": 3.8355,
+      "step": 2866
+    },
+    {
+      "epoch": 3.66688,
+      "grad_norm": 0.5054891705513,
+      "learning_rate": 0.00020005384304751648,
+      "loss": 3.8619,
+      "step": 2867
+    },
+    {
+      "epoch": 3.66816,
+      "grad_norm": 0.47256743907928467,
+      "learning_rate": 0.00020001346076187911,
+      "loss": 3.7752,
+      "step": 2868
+    },
+    {
+      "epoch": 3.66944,
+      "grad_norm": 0.48075199127197266,
+      "learning_rate": 0.00019997307847624174,
+      "loss": 3.8054,
+      "step": 2869
+    },
+    {
+      "epoch": 3.67072,
+      "grad_norm": 0.4974772036075592,
+      "learning_rate": 0.00019993269619060435,
+      "loss": 3.7665,
+      "step": 2870
+    },
+    {
+      "epoch": 3.672,
+      "grad_norm": 0.48151201009750366,
+      "learning_rate": 0.000199892313904967,
+      "loss": 3.8581,
+      "step": 2871
+    },
+    {
+      "epoch": 3.67328,
+      "grad_norm": 0.5086482763290405,
+      "learning_rate": 0.00019985193161932964,
+      "loss": 3.8571,
+      "step": 2872
+    },
+    {
+      "epoch": 3.67456,
+      "grad_norm": 0.4769311845302582,
+      "learning_rate": 0.00019981154933369227,
+      "loss": 3.8258,
+      "step": 2873
+    },
+    {
+      "epoch": 3.67584,
+      "grad_norm": 0.5033696889877319,
+      "learning_rate": 0.0001997711670480549,
+      "loss": 3.8217,
+      "step": 2874
+    },
+    {
+      "epoch": 3.67712,
+      "grad_norm": 0.4959774613380432,
+      "learning_rate": 0.00019973078476241755,
+      "loss": 3.8151,
+      "step": 2875
+    },
+    {
+      "epoch": 3.6784,
+      "grad_norm": 0.4797199070453644,
+      "learning_rate": 0.00019969040247678018,
+      "loss": 3.8164,
+      "step": 2876
+    },
+    {
+      "epoch": 3.67968,
+      "grad_norm": 0.5271828770637512,
+      "learning_rate": 0.0001996500201911428,
+      "loss": 3.7692,
+      "step": 2877
+    },
+    {
+      "epoch": 3.68096,
+      "grad_norm": 0.489942729473114,
+      "learning_rate": 0.00019960963790550542,
+      "loss": 3.8024,
+      "step": 2878
+    },
+    {
+      "epoch": 3.68224,
+      "grad_norm": 0.5125555992126465,
+      "learning_rate": 0.00019956925561986805,
+      "loss": 3.7912,
+      "step": 2879
+    },
+    {
+      "epoch": 3.68352,
+      "grad_norm": 0.4915968179702759,
+      "learning_rate": 0.0001995288733342307,
+      "loss": 3.7815,
+      "step": 2880
+    },
+    {
+      "epoch": 3.6848,
+      "grad_norm": 0.49331218004226685,
+      "learning_rate": 0.00019948849104859333,
+      "loss": 3.878,
+      "step": 2881
+    },
+    {
+      "epoch": 3.68608,
+      "grad_norm": 0.5256069898605347,
+      "learning_rate": 0.00019944810876295596,
+      "loss": 3.8137,
+      "step": 2882
+    },
+    {
+      "epoch": 3.68736,
+      "grad_norm": 0.4750528335571289,
+      "learning_rate": 0.0001994077264773186,
+      "loss": 3.7832,
+      "step": 2883
+    },
+    {
+      "epoch": 3.68864,
+      "grad_norm": 0.5187920928001404,
+      "learning_rate": 0.00019936734419168125,
+      "loss": 3.8111,
+      "step": 2884
+    },
+    {
+      "epoch": 3.68992,
+      "grad_norm": 0.4905349016189575,
+      "learning_rate": 0.00019932696190604388,
+      "loss": 3.8054,
+      "step": 2885
+    },
+    {
+      "epoch": 3.6912000000000003,
+      "grad_norm": 0.5346660614013672,
+      "learning_rate": 0.00019928657962040648,
+      "loss": 3.8365,
+      "step": 2886
+    },
+    {
+      "epoch": 3.6924799999999998,
+      "grad_norm": 0.5014554262161255,
+      "learning_rate": 0.0001992461973347691,
+      "loss": 3.8105,
+      "step": 2887
+    },
+    {
+      "epoch": 3.69376,
+      "grad_norm": 0.4906976521015167,
+      "learning_rate": 0.00019920581504913177,
+      "loss": 3.7636,
+      "step": 2888
+    },
+    {
+      "epoch": 3.69504,
+      "grad_norm": 0.5077499747276306,
+      "learning_rate": 0.0001991654327634944,
+      "loss": 3.7638,
+      "step": 2889
+    },
+    {
+      "epoch": 3.69632,
+      "grad_norm": 0.48475882411003113,
+      "learning_rate": 0.00019912505047785703,
+      "loss": 3.8723,
+      "step": 2890
+    },
+    {
+      "epoch": 3.6976,
+      "grad_norm": 0.509909451007843,
+      "learning_rate": 0.00019908466819221966,
+      "loss": 3.8396,
+      "step": 2891
+    },
+    {
+      "epoch": 3.69888,
+      "grad_norm": 0.49699631333351135,
+      "learning_rate": 0.0001990442859065823,
+      "loss": 3.7865,
+      "step": 2892
+    },
+    {
+      "epoch": 3.70016,
+      "grad_norm": 0.4960545301437378,
+      "learning_rate": 0.00019900390362094495,
+      "loss": 3.819,
+      "step": 2893
+    },
+    {
+      "epoch": 3.70144,
+      "grad_norm": 0.5059762597084045,
+      "learning_rate": 0.00019896352133530755,
+      "loss": 3.8588,
+      "step": 2894
+    },
+    {
+      "epoch": 3.7027200000000002,
+      "grad_norm": 0.503450870513916,
+      "learning_rate": 0.00019892313904967018,
+      "loss": 3.8109,
+      "step": 2895
+    },
+    {
+      "epoch": 3.7039999999999997,
+      "grad_norm": 0.505183219909668,
+      "learning_rate": 0.0001988827567640328,
+      "loss": 3.7826,
+      "step": 2896
+    },
+    {
+      "epoch": 3.70528,
+      "grad_norm": 0.5047440528869629,
+      "learning_rate": 0.00019884237447839547,
+      "loss": 3.8386,
+      "step": 2897
+    },
+    {
+      "epoch": 3.70656,
+      "grad_norm": 0.5242089033126831,
+      "learning_rate": 0.0001988019921927581,
+      "loss": 3.7865,
+      "step": 2898
+    },
+    {
+      "epoch": 3.70784,
+      "grad_norm": 0.48960578441619873,
+      "learning_rate": 0.00019876160990712073,
+      "loss": 3.8107,
+      "step": 2899
+    },
+    {
+      "epoch": 3.70912,
+      "grad_norm": 0.5148612260818481,
+      "learning_rate": 0.00019872122762148336,
+      "loss": 3.7782,
+      "step": 2900
+    },
+    {
+      "epoch": 3.7104,
+      "grad_norm": 0.4915827810764313,
+      "learning_rate": 0.00019868084533584602,
+      "loss": 3.8166,
+      "step": 2901
+    },
+    {
+      "epoch": 3.71168,
+      "grad_norm": 0.4962410032749176,
+      "learning_rate": 0.00019864046305020862,
+      "loss": 3.7864,
+      "step": 2902
+    },
+    {
+      "epoch": 3.71296,
+      "grad_norm": 0.50472092628479,
+      "learning_rate": 0.00019860008076457125,
+      "loss": 3.7607,
+      "step": 2903
+    },
+    {
+      "epoch": 3.71424,
+      "grad_norm": 0.5142050981521606,
+      "learning_rate": 0.00019855969847893388,
+      "loss": 3.835,
+      "step": 2904
+    },
+    {
+      "epoch": 3.71552,
+      "grad_norm": 0.5082858204841614,
+      "learning_rate": 0.0001985193161932965,
+      "loss": 3.903,
+      "step": 2905
+    },
+    {
+      "epoch": 3.7168,
+      "grad_norm": 0.529228150844574,
+      "learning_rate": 0.00019847893390765917,
+      "loss": 3.7935,
+      "step": 2906
+    },
+    {
+      "epoch": 3.71808,
+      "grad_norm": 0.515741765499115,
+      "learning_rate": 0.0001984385516220218,
+      "loss": 3.863,
+      "step": 2907
+    },
+    {
+      "epoch": 3.71936,
+      "grad_norm": 0.5094571113586426,
+      "learning_rate": 0.00019839816933638443,
+      "loss": 3.8557,
+      "step": 2908
+    },
+    {
+      "epoch": 3.72064,
+      "grad_norm": 0.5035596489906311,
+      "learning_rate": 0.00019835778705074703,
+      "loss": 3.8019,
+      "step": 2909
+    },
+    {
+      "epoch": 3.72192,
+      "grad_norm": 0.5284061431884766,
+      "learning_rate": 0.00019831740476510969,
+      "loss": 3.8547,
+      "step": 2910
+    },
+    {
+      "epoch": 3.7232,
+      "grad_norm": 0.5245686173439026,
+      "learning_rate": 0.00019827702247947232,
+      "loss": 3.7454,
+      "step": 2911
+    },
+    {
+      "epoch": 3.72448,
+      "grad_norm": 0.514725923538208,
+      "learning_rate": 0.00019823664019383495,
+      "loss": 3.8237,
+      "step": 2912
+    },
+    {
+      "epoch": 3.72576,
+      "grad_norm": 0.5170477628707886,
+      "learning_rate": 0.00019819625790819758,
+      "loss": 3.8264,
+      "step": 2913
+    },
+    {
+      "epoch": 3.72704,
+      "grad_norm": 0.5244784951210022,
+      "learning_rate": 0.00019815587562256023,
+      "loss": 3.7872,
+      "step": 2914
+    },
+    {
+      "epoch": 3.72832,
+      "grad_norm": 0.510220468044281,
+      "learning_rate": 0.00019811549333692286,
+      "loss": 3.8574,
+      "step": 2915
+    },
+    {
+      "epoch": 3.7296,
+      "grad_norm": 0.5242193937301636,
+      "learning_rate": 0.0001980751110512855,
+      "loss": 3.8547,
+      "step": 2916
+    },
+    {
+      "epoch": 3.73088,
+      "grad_norm": 0.48602530360221863,
+      "learning_rate": 0.0001980347287656481,
+      "loss": 3.7948,
+      "step": 2917
+    },
+    {
+      "epoch": 3.73216,
+      "grad_norm": 0.4796367585659027,
+      "learning_rate": 0.00019799434648001073,
+      "loss": 3.793,
+      "step": 2918
+    },
+    {
+      "epoch": 3.73344,
+      "grad_norm": 0.4868846535682678,
+      "learning_rate": 0.00019795396419437338,
+      "loss": 3.8138,
+      "step": 2919
+    },
+    {
+      "epoch": 3.7347200000000003,
+      "grad_norm": 0.48275643587112427,
+      "learning_rate": 0.00019791358190873601,
+      "loss": 3.8185,
+      "step": 2920
+    },
+    {
+      "epoch": 3.7359999999999998,
+      "grad_norm": 0.4956841468811035,
+      "learning_rate": 0.00019787319962309864,
+      "loss": 3.8374,
+      "step": 2921
+    },
+    {
+      "epoch": 3.73728,
+      "grad_norm": 0.5024325251579285,
+      "learning_rate": 0.00019783281733746127,
+      "loss": 3.9032,
+      "step": 2922
+    },
+    {
+      "epoch": 3.73856,
+      "grad_norm": 0.4799942970275879,
+      "learning_rate": 0.00019779243505182393,
+      "loss": 3.8317,
+      "step": 2923
+    },
+    {
+      "epoch": 3.73984,
+      "grad_norm": 0.5100151300430298,
+      "learning_rate": 0.00019775205276618656,
+      "loss": 3.8394,
+      "step": 2924
+    },
+    {
+      "epoch": 3.74112,
+      "grad_norm": 0.4840579926967621,
+      "learning_rate": 0.00019771167048054916,
+      "loss": 3.7803,
+      "step": 2925
+    },
+    {
+      "epoch": 3.7424,
+      "grad_norm": 0.495714008808136,
+      "learning_rate": 0.0001976712881949118,
+      "loss": 3.7849,
+      "step": 2926
+    },
+    {
+      "epoch": 3.74368,
+      "grad_norm": 0.49256429076194763,
+      "learning_rate": 0.00019763090590927445,
+      "loss": 3.7815,
+      "step": 2927
+    },
+    {
+      "epoch": 3.74496,
+      "grad_norm": 0.4639676809310913,
+      "learning_rate": 0.00019759052362363708,
+      "loss": 3.8497,
+      "step": 2928
+    },
+    {
+      "epoch": 3.7462400000000002,
+      "grad_norm": 0.486234575510025,
+      "learning_rate": 0.0001975501413379997,
+      "loss": 3.866,
+      "step": 2929
+    },
+    {
+      "epoch": 3.7475199999999997,
+      "grad_norm": 0.4823131263256073,
+      "learning_rate": 0.00019750975905236234,
+      "loss": 3.7977,
+      "step": 2930
+    },
+    {
+      "epoch": 3.7488,
+      "grad_norm": 0.4761921465396881,
+      "learning_rate": 0.000197469376766725,
+      "loss": 3.7816,
+      "step": 2931
+    },
+    {
+      "epoch": 3.75008,
+      "grad_norm": 0.4939184784889221,
+      "learning_rate": 0.00019742899448108763,
+      "loss": 3.8268,
+      "step": 2932
+    },
+    {
+      "epoch": 3.75136,
+      "grad_norm": 0.4557759761810303,
+      "learning_rate": 0.00019738861219545026,
+      "loss": 3.7831,
+      "step": 2933
+    },
+    {
+      "epoch": 3.75264,
+      "grad_norm": 0.4687483012676239,
+      "learning_rate": 0.00019734822990981286,
+      "loss": 3.8796,
+      "step": 2934
+    },
+    {
+      "epoch": 3.75392,
+      "grad_norm": 0.4676748514175415,
+      "learning_rate": 0.0001973078476241755,
+      "loss": 3.8334,
+      "step": 2935
+    },
+    {
+      "epoch": 3.7552,
+      "grad_norm": 0.47445160150527954,
+      "learning_rate": 0.00019726746533853815,
+      "loss": 3.7532,
+      "step": 2936
+    },
+    {
+      "epoch": 3.75648,
+      "grad_norm": 0.4851106107234955,
+      "learning_rate": 0.00019722708305290078,
+      "loss": 3.7783,
+      "step": 2937
+    },
+    {
+      "epoch": 3.75776,
+      "grad_norm": 0.4669123888015747,
+      "learning_rate": 0.0001971867007672634,
+      "loss": 3.8144,
+      "step": 2938
+    },
+    {
+      "epoch": 3.75904,
+      "grad_norm": 0.4675630033016205,
+      "learning_rate": 0.00019714631848162604,
+      "loss": 3.7909,
+      "step": 2939
+    },
+    {
+      "epoch": 3.76032,
+      "grad_norm": 0.4890853762626648,
+      "learning_rate": 0.0001971059361959887,
+      "loss": 3.803,
+      "step": 2940
+    },
+    {
+      "epoch": 3.7616,
+      "grad_norm": 0.4596586227416992,
+      "learning_rate": 0.00019706555391035133,
+      "loss": 3.7987,
+      "step": 2941
+    },
+    {
+      "epoch": 3.76288,
+      "grad_norm": 0.5085187554359436,
+      "learning_rate": 0.00019702517162471393,
+      "loss": 3.7938,
+      "step": 2942
+    },
+    {
+      "epoch": 3.76416,
+      "grad_norm": 0.4674232006072998,
+      "learning_rate": 0.00019698478933907656,
+      "loss": 3.9139,
+      "step": 2943
+    },
+    {
+      "epoch": 3.76544,
+      "grad_norm": 0.4964526891708374,
+      "learning_rate": 0.00019694440705343922,
+      "loss": 3.805,
+      "step": 2944
+    },
+    {
+      "epoch": 3.76672,
+      "grad_norm": 0.4870041310787201,
+      "learning_rate": 0.00019690402476780185,
+      "loss": 3.7573,
+      "step": 2945
+    },
+    {
+      "epoch": 3.768,
+      "grad_norm": 0.4788266718387604,
+      "learning_rate": 0.00019686364248216448,
+      "loss": 3.8082,
+      "step": 2946
+    },
+    {
+      "epoch": 3.76928,
+      "grad_norm": 0.5005785822868347,
+      "learning_rate": 0.0001968232601965271,
+      "loss": 3.7935,
+      "step": 2947
+    },
+    {
+      "epoch": 3.77056,
+      "grad_norm": 0.48432236909866333,
+      "learning_rate": 0.00019678287791088974,
+      "loss": 3.8643,
+      "step": 2948
+    },
+    {
+      "epoch": 3.77184,
+      "grad_norm": 0.4844403862953186,
+      "learning_rate": 0.0001967424956252524,
+      "loss": 3.8046,
+      "step": 2949
+    },
+    {
+      "epoch": 3.77312,
+      "grad_norm": 0.49328166246414185,
+      "learning_rate": 0.000196702113339615,
+      "loss": 3.8041,
+      "step": 2950
+    },
+    {
+      "epoch": 3.7744,
+      "grad_norm": 0.48709914088249207,
+      "learning_rate": 0.00019666173105397763,
+      "loss": 3.7822,
+      "step": 2951
+    },
+    {
+      "epoch": 3.77568,
+      "grad_norm": 0.4914034903049469,
+      "learning_rate": 0.00019662134876834026,
+      "loss": 3.866,
+      "step": 2952
+    },
+    {
+      "epoch": 3.77696,
+      "grad_norm": 0.5005134344100952,
+      "learning_rate": 0.00019658096648270292,
+      "loss": 3.7386,
+      "step": 2953
+    },
+    {
+      "epoch": 3.7782400000000003,
+      "grad_norm": 0.4764963388442993,
+      "learning_rate": 0.00019654058419706555,
+      "loss": 3.8049,
+      "step": 2954
+    },
+    {
+      "epoch": 3.7795199999999998,
+      "grad_norm": 0.4586517810821533,
+      "learning_rate": 0.00019650020191142818,
+      "loss": 3.8072,
+      "step": 2955
+    },
+    {
+      "epoch": 3.7808,
+      "grad_norm": 0.4838949739933014,
+      "learning_rate": 0.0001964598196257908,
+      "loss": 3.8139,
+      "step": 2956
+    },
+    {
+      "epoch": 3.78208,
+      "grad_norm": 0.48450782895088196,
+      "learning_rate": 0.00019641943734015346,
+      "loss": 3.8376,
+      "step": 2957
+    },
+    {
+      "epoch": 3.78336,
+      "grad_norm": 0.5044925808906555,
+      "learning_rate": 0.00019637905505451607,
+      "loss": 3.7702,
+      "step": 2958
+    },
+    {
+      "epoch": 3.78464,
+      "grad_norm": 0.4797843396663666,
+      "learning_rate": 0.0001963386727688787,
+      "loss": 3.7628,
+      "step": 2959
+    },
+    {
+      "epoch": 3.78592,
+      "grad_norm": 0.4827995300292969,
+      "learning_rate": 0.00019629829048324133,
+      "loss": 3.7901,
+      "step": 2960
+    },
+    {
+      "epoch": 3.7872,
+      "grad_norm": 0.4926610291004181,
+      "learning_rate": 0.00019625790819760396,
+      "loss": 3.8232,
+      "step": 2961
+    },
+    {
+      "epoch": 3.78848,
+      "grad_norm": 0.46942880749702454,
+      "learning_rate": 0.0001962175259119666,
+      "loss": 3.7819,
+      "step": 2962
+    },
+    {
+      "epoch": 3.7897600000000002,
+      "grad_norm": 0.5101065039634705,
+      "learning_rate": 0.00019617714362632924,
+      "loss": 3.8026,
+      "step": 2963
+    },
+    {
+      "epoch": 3.7910399999999997,
+      "grad_norm": 0.4860995411872864,
+      "learning_rate": 0.00019613676134069187,
+      "loss": 3.8669,
+      "step": 2964
+    },
+    {
+      "epoch": 3.79232,
+      "grad_norm": 0.49931204319000244,
+      "learning_rate": 0.00019609637905505448,
+      "loss": 3.8031,
+      "step": 2965
+    },
+    {
+      "epoch": 3.7936,
+      "grad_norm": 0.4776234030723572,
+      "learning_rate": 0.00019605599676941713,
+      "loss": 3.8091,
+      "step": 2966
+    },
+    {
+      "epoch": 3.79488,
+      "grad_norm": 0.5147337317466736,
+      "learning_rate": 0.00019601561448377976,
+      "loss": 3.8127,
+      "step": 2967
+    },
+    {
+      "epoch": 3.79616,
+      "grad_norm": 0.49513697624206543,
+      "learning_rate": 0.0001959752321981424,
+      "loss": 3.9379,
+      "step": 2968
+    },
+    {
+      "epoch": 3.79744,
+      "grad_norm": 0.5062114596366882,
+      "learning_rate": 0.00019593484991250502,
+      "loss": 3.833,
+      "step": 2969
+    },
+    {
+      "epoch": 3.79872,
+      "grad_norm": 0.48605409264564514,
+      "learning_rate": 0.00019589446762686768,
+      "loss": 3.8156,
+      "step": 2970
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.48607662320137024,
+      "learning_rate": 0.0001958540853412303,
+      "loss": 3.8235,
+      "step": 2971
+    },
+    {
+      "epoch": 3.80128,
+      "grad_norm": 0.48669230937957764,
+      "learning_rate": 0.00019581370305559294,
+      "loss": 3.7372,
+      "step": 2972
+    },
+    {
+      "epoch": 3.80256,
+      "grad_norm": 0.48195433616638184,
+      "learning_rate": 0.00019577332076995554,
+      "loss": 3.798,
+      "step": 2973
+    },
+    {
+      "epoch": 3.80384,
+      "grad_norm": 0.4917437434196472,
+      "learning_rate": 0.00019573293848431817,
+      "loss": 3.8529,
+      "step": 2974
+    },
+    {
+      "epoch": 3.80512,
+      "grad_norm": 0.5031919479370117,
+      "learning_rate": 0.00019569255619868083,
+      "loss": 3.7061,
+      "step": 2975
+    },
+    {
+      "epoch": 3.8064,
+      "grad_norm": 0.5153105854988098,
+      "learning_rate": 0.00019565217391304346,
+      "loss": 3.7935,
+      "step": 2976
+    },
+    {
+      "epoch": 3.80768,
+      "grad_norm": 0.5166071057319641,
+      "learning_rate": 0.0001956117916274061,
+      "loss": 3.8835,
+      "step": 2977
+    },
+    {
+      "epoch": 3.80896,
+      "grad_norm": 0.47903141379356384,
+      "learning_rate": 0.00019557140934176872,
+      "loss": 3.7476,
+      "step": 2978
+    },
+    {
+      "epoch": 3.81024,
+      "grad_norm": 0.4941449761390686,
+      "learning_rate": 0.00019553102705613138,
+      "loss": 3.7796,
+      "step": 2979
+    },
+    {
+      "epoch": 3.81152,
+      "grad_norm": 0.5048912763595581,
+      "learning_rate": 0.000195490644770494,
+      "loss": 3.7708,
+      "step": 2980
+    },
+    {
+      "epoch": 3.8128,
+      "grad_norm": 0.4860134422779083,
+      "learning_rate": 0.0001954502624848566,
+      "loss": 3.8097,
+      "step": 2981
+    },
+    {
+      "epoch": 3.81408,
+      "grad_norm": 0.48185980319976807,
+      "learning_rate": 0.00019540988019921924,
+      "loss": 3.8444,
+      "step": 2982
+    },
+    {
+      "epoch": 3.81536,
+      "grad_norm": 0.4978376030921936,
+      "learning_rate": 0.0001953694979135819,
+      "loss": 3.7791,
+      "step": 2983
+    },
+    {
+      "epoch": 3.81664,
+      "grad_norm": 0.5320491194725037,
+      "learning_rate": 0.00019532911562794453,
+      "loss": 3.7774,
+      "step": 2984
+    },
+    {
+      "epoch": 3.81792,
+      "grad_norm": 0.4956910014152527,
+      "learning_rate": 0.00019528873334230716,
+      "loss": 3.8242,
+      "step": 2985
+    },
+    {
+      "epoch": 3.8192,
+      "grad_norm": 0.4940122663974762,
+      "learning_rate": 0.0001952483510566698,
+      "loss": 3.7981,
+      "step": 2986
+    },
+    {
+      "epoch": 3.82048,
+      "grad_norm": 0.49585092067718506,
+      "learning_rate": 0.00019520796877103242,
+      "loss": 3.7541,
+      "step": 2987
+    },
+    {
+      "epoch": 3.8217600000000003,
+      "grad_norm": 0.49807533621788025,
+      "learning_rate": 0.00019516758648539508,
+      "loss": 3.7866,
+      "step": 2988
+    },
+    {
+      "epoch": 3.8230399999999998,
+      "grad_norm": 0.5053405165672302,
+      "learning_rate": 0.00019512720419975768,
+      "loss": 3.8625,
+      "step": 2989
+    },
+    {
+      "epoch": 3.82432,
+      "grad_norm": 0.5178548693656921,
+      "learning_rate": 0.0001950868219141203,
+      "loss": 3.7998,
+      "step": 2990
+    },
+    {
+      "epoch": 3.8256,
+      "grad_norm": 0.5026320815086365,
+      "learning_rate": 0.00019504643962848294,
+      "loss": 3.815,
+      "step": 2991
+    },
+    {
+      "epoch": 3.82688,
+      "grad_norm": 0.5041228532791138,
+      "learning_rate": 0.0001950060573428456,
+      "loss": 3.7673,
+      "step": 2992
+    },
+    {
+      "epoch": 3.82816,
+      "grad_norm": 0.5110290050506592,
+      "learning_rate": 0.00019496567505720823,
+      "loss": 3.8342,
+      "step": 2993
+    },
+    {
+      "epoch": 3.82944,
+      "grad_norm": 0.5149263739585876,
+      "learning_rate": 0.00019492529277157086,
+      "loss": 3.8922,
+      "step": 2994
+    },
+    {
+      "epoch": 3.83072,
+      "grad_norm": 0.49787768721580505,
+      "learning_rate": 0.0001948849104859335,
+      "loss": 3.8788,
+      "step": 2995
+    },
+    {
+      "epoch": 3.832,
+      "grad_norm": 0.48449409008026123,
+      "learning_rate": 0.00019484452820029614,
+      "loss": 3.7207,
+      "step": 2996
+    },
+    {
+      "epoch": 3.8332800000000002,
+      "grad_norm": 0.49689674377441406,
+      "learning_rate": 0.00019480414591465875,
+      "loss": 3.7686,
+      "step": 2997
+    },
+    {
+      "epoch": 3.8345599999999997,
+      "grad_norm": 0.533508837223053,
+      "learning_rate": 0.00019476376362902138,
+      "loss": 3.8948,
+      "step": 2998
+    },
+    {
+      "epoch": 3.83584,
+      "grad_norm": 0.48604533076286316,
+      "learning_rate": 0.000194723381343384,
+      "loss": 3.7144,
+      "step": 2999
+    },
+    {
+      "epoch": 3.83712,
+      "grad_norm": 0.5123369693756104,
+      "learning_rate": 0.00019468299905774664,
+      "loss": 3.7716,
+      "step": 3000
+    },
+    {
+      "epoch": 3.8384,
+      "grad_norm": 0.4925461709499359,
+      "learning_rate": 0.0001946426167721093,
+      "loss": 3.862,
+      "step": 3001
+    },
+    {
+      "epoch": 3.83968,
+      "grad_norm": 0.5041381120681763,
+      "learning_rate": 0.00019460223448647192,
+      "loss": 3.7986,
+      "step": 3002
+    },
+    {
+      "epoch": 3.84096,
+      "grad_norm": 0.4902060031890869,
+      "learning_rate": 0.00019456185220083455,
+      "loss": 3.8608,
+      "step": 3003
+    },
+    {
+      "epoch": 3.84224,
+      "grad_norm": 0.5032182931900024,
+      "learning_rate": 0.00019452146991519716,
+      "loss": 3.8192,
+      "step": 3004
+    },
+    {
+      "epoch": 3.84352,
+      "grad_norm": 0.5030463337898254,
+      "learning_rate": 0.00019448108762955984,
+      "loss": 3.8128,
+      "step": 3005
+    },
+    {
+      "epoch": 3.8448,
+      "grad_norm": 0.49824386835098267,
+      "learning_rate": 0.00019444070534392244,
+      "loss": 3.7264,
+      "step": 3006
+    },
+    {
+      "epoch": 3.84608,
+      "grad_norm": 0.5038865208625793,
+      "learning_rate": 0.00019440032305828507,
+      "loss": 3.8521,
+      "step": 3007
+    },
+    {
+      "epoch": 3.84736,
+      "grad_norm": 0.5093699097633362,
+      "learning_rate": 0.0001943599407726477,
+      "loss": 3.7095,
+      "step": 3008
+    },
+    {
+      "epoch": 3.84864,
+      "grad_norm": 0.49875956773757935,
+      "learning_rate": 0.00019431955848701036,
+      "loss": 3.868,
+      "step": 3009
+    },
+    {
+      "epoch": 3.84992,
+      "grad_norm": 0.5082076787948608,
+      "learning_rate": 0.000194279176201373,
+      "loss": 3.8264,
+      "step": 3010
+    },
+    {
+      "epoch": 3.8512,
+      "grad_norm": 0.5088693499565125,
+      "learning_rate": 0.00019423879391573562,
+      "loss": 3.8094,
+      "step": 3011
+    },
+    {
+      "epoch": 3.85248,
+      "grad_norm": 0.49538499116897583,
+      "learning_rate": 0.00019419841163009823,
+      "loss": 3.8007,
+      "step": 3012
+    },
+    {
+      "epoch": 3.85376,
+      "grad_norm": 0.5211043953895569,
+      "learning_rate": 0.00019415802934446086,
+      "loss": 3.7566,
+      "step": 3013
+    },
+    {
+      "epoch": 3.85504,
+      "grad_norm": 0.4987725019454956,
+      "learning_rate": 0.0001941176470588235,
+      "loss": 3.8489,
+      "step": 3014
+    },
+    {
+      "epoch": 3.85632,
+      "grad_norm": 0.4974265992641449,
+      "learning_rate": 0.00019407726477318614,
+      "loss": 3.8121,
+      "step": 3015
+    },
+    {
+      "epoch": 3.8576,
+      "grad_norm": 0.5079092383384705,
+      "learning_rate": 0.00019403688248754877,
+      "loss": 3.8206,
+      "step": 3016
+    },
+    {
+      "epoch": 3.85888,
+      "grad_norm": 0.4906683564186096,
+      "learning_rate": 0.0001939965002019114,
+      "loss": 3.7863,
+      "step": 3017
+    },
+    {
+      "epoch": 3.86016,
+      "grad_norm": 0.5003571510314941,
+      "learning_rate": 0.00019395611791627406,
+      "loss": 3.7905,
+      "step": 3018
+    },
+    {
+      "epoch": 3.86144,
+      "grad_norm": 0.5029097199440002,
+      "learning_rate": 0.0001939157356306367,
+      "loss": 3.8112,
+      "step": 3019
+    },
+    {
+      "epoch": 3.86272,
+      "grad_norm": 0.494642972946167,
+      "learning_rate": 0.00019387535334499932,
+      "loss": 3.7935,
+      "step": 3020
+    },
+    {
+      "epoch": 3.864,
+      "grad_norm": 0.48063820600509644,
+      "learning_rate": 0.00019383497105936192,
+      "loss": 3.8126,
+      "step": 3021
+    },
+    {
+      "epoch": 3.8652800000000003,
+      "grad_norm": 0.4792153835296631,
+      "learning_rate": 0.00019379458877372458,
+      "loss": 3.7361,
+      "step": 3022
+    },
+    {
+      "epoch": 3.8665599999999998,
+      "grad_norm": 0.46592941880226135,
+      "learning_rate": 0.0001937542064880872,
+      "loss": 3.8774,
+      "step": 3023
+    },
+    {
+      "epoch": 3.86784,
+      "grad_norm": 0.46901074051856995,
+      "learning_rate": 0.00019371382420244984,
+      "loss": 3.8471,
+      "step": 3024
+    },
+    {
+      "epoch": 3.86912,
+      "grad_norm": 0.4709526300430298,
+      "learning_rate": 0.00019367344191681247,
+      "loss": 3.8575,
+      "step": 3025
+    },
+    {
+      "epoch": 3.8704,
+      "grad_norm": 0.4765014946460724,
+      "learning_rate": 0.0001936330596311751,
+      "loss": 3.7684,
+      "step": 3026
+    },
+    {
+      "epoch": 3.87168,
+      "grad_norm": 0.48558369278907776,
+      "learning_rate": 0.00019359267734553776,
+      "loss": 3.8208,
+      "step": 3027
+    },
+    {
+      "epoch": 3.87296,
+      "grad_norm": 0.47557520866394043,
+      "learning_rate": 0.0001935522950599004,
+      "loss": 3.755,
+      "step": 3028
+    },
+    {
+      "epoch": 3.87424,
+      "grad_norm": 0.4816844165325165,
+      "learning_rate": 0.000193511912774263,
+      "loss": 3.8242,
+      "step": 3029
+    },
+    {
+      "epoch": 3.87552,
+      "grad_norm": 0.464916467666626,
+      "learning_rate": 0.00019347153048862562,
+      "loss": 3.8323,
+      "step": 3030
+    },
+    {
+      "epoch": 3.8768000000000002,
+      "grad_norm": 0.4912608563899994,
+      "learning_rate": 0.00019343114820298828,
+      "loss": 3.833,
+      "step": 3031
+    },
+    {
+      "epoch": 3.8780799999999997,
+      "grad_norm": 0.4688814580440521,
+      "learning_rate": 0.0001933907659173509,
+      "loss": 3.8551,
+      "step": 3032
+    },
+    {
+      "epoch": 3.87936,
+      "grad_norm": 0.4833512306213379,
+      "learning_rate": 0.00019335038363171354,
+      "loss": 3.8419,
+      "step": 3033
+    },
+    {
+      "epoch": 3.88064,
+      "grad_norm": 0.49540048837661743,
+      "learning_rate": 0.00019331000134607617,
+      "loss": 3.8368,
+      "step": 3034
+    },
+    {
+      "epoch": 3.88192,
+      "grad_norm": 0.4662631154060364,
+      "learning_rate": 0.00019326961906043883,
+      "loss": 3.9145,
+      "step": 3035
+    },
+    {
+      "epoch": 3.8832,
+      "grad_norm": 0.4789687693119049,
+      "learning_rate": 0.00019322923677480146,
+      "loss": 3.809,
+      "step": 3036
+    },
+    {
+      "epoch": 3.88448,
+      "grad_norm": 0.47663864493370056,
+      "learning_rate": 0.00019318885448916406,
+      "loss": 3.7629,
+      "step": 3037
+    },
+    {
+      "epoch": 3.88576,
+      "grad_norm": 0.46691396832466125,
+      "learning_rate": 0.0001931484722035267,
+      "loss": 3.7381,
+      "step": 3038
+    },
+    {
+      "epoch": 3.88704,
+      "grad_norm": 0.5005871057510376,
+      "learning_rate": 0.00019310808991788932,
+      "loss": 3.8393,
+      "step": 3039
+    },
+    {
+      "epoch": 3.88832,
+      "grad_norm": 0.4863739311695099,
+      "learning_rate": 0.00019306770763225198,
+      "loss": 3.7096,
+      "step": 3040
+    },
+    {
+      "epoch": 3.8895999999999997,
+      "grad_norm": 0.49686795473098755,
+      "learning_rate": 0.0001930273253466146,
+      "loss": 3.74,
+      "step": 3041
+    },
+    {
+      "epoch": 3.89088,
+      "grad_norm": 0.4751260578632355,
+      "learning_rate": 0.00019298694306097724,
+      "loss": 3.8243,
+      "step": 3042
+    },
+    {
+      "epoch": 3.89216,
+      "grad_norm": 0.4915493428707123,
+      "learning_rate": 0.00019294656077533987,
+      "loss": 3.8116,
+      "step": 3043
+    },
+    {
+      "epoch": 3.89344,
+      "grad_norm": 0.4819605350494385,
+      "learning_rate": 0.00019290617848970252,
+      "loss": 3.843,
+      "step": 3044
+    },
+    {
+      "epoch": 3.89472,
+      "grad_norm": 0.4787873923778534,
+      "learning_rate": 0.00019286579620406513,
+      "loss": 3.7245,
+      "step": 3045
+    },
+    {
+      "epoch": 3.896,
+      "grad_norm": 0.4805007874965668,
+      "learning_rate": 0.00019282541391842776,
+      "loss": 3.7573,
+      "step": 3046
+    },
+    {
+      "epoch": 3.89728,
+      "grad_norm": 0.4945317804813385,
+      "learning_rate": 0.00019278503163279039,
+      "loss": 3.9226,
+      "step": 3047
+    },
+    {
+      "epoch": 3.89856,
+      "grad_norm": 0.49421226978302,
+      "learning_rate": 0.00019274464934715304,
+      "loss": 3.7792,
+      "step": 3048
+    },
+    {
+      "epoch": 3.89984,
+      "grad_norm": 0.47549110651016235,
+      "learning_rate": 0.00019270426706151567,
+      "loss": 3.8229,
+      "step": 3049
+    },
+    {
+      "epoch": 3.90112,
+      "grad_norm": 0.5038889050483704,
+      "learning_rate": 0.0001926638847758783,
+      "loss": 3.8079,
+      "step": 3050
+    },
+    {
+      "epoch": 3.9024,
+      "grad_norm": 0.4997271001338959,
+      "learning_rate": 0.00019262350249024093,
+      "loss": 3.8135,
+      "step": 3051
+    },
+    {
+      "epoch": 3.90368,
+      "grad_norm": 0.4927690625190735,
+      "learning_rate": 0.00019258312020460354,
+      "loss": 3.8534,
+      "step": 3052
+    },
+    {
+      "epoch": 3.90496,
+      "grad_norm": 0.4786875545978546,
+      "learning_rate": 0.0001925427379189662,
+      "loss": 3.7229,
+      "step": 3053
+    },
+    {
+      "epoch": 3.90624,
+      "grad_norm": 0.5047836899757385,
+      "learning_rate": 0.00019250235563332882,
+      "loss": 3.8432,
+      "step": 3054
+    },
+    {
+      "epoch": 3.90752,
+      "grad_norm": 0.5120929479598999,
+      "learning_rate": 0.00019246197334769145,
+      "loss": 3.777,
+      "step": 3055
+    },
+    {
+      "epoch": 3.9088000000000003,
+      "grad_norm": 0.49564608931541443,
+      "learning_rate": 0.00019242159106205408,
+      "loss": 3.8209,
+      "step": 3056
+    },
+    {
+      "epoch": 3.91008,
+      "grad_norm": 0.4866943657398224,
+      "learning_rate": 0.00019238120877641674,
+      "loss": 3.7899,
+      "step": 3057
+    },
+    {
+      "epoch": 3.91136,
+      "grad_norm": 0.487051784992218,
+      "learning_rate": 0.00019234082649077937,
+      "loss": 3.799,
+      "step": 3058
+    },
+    {
+      "epoch": 3.91264,
+      "grad_norm": 0.4829874634742737,
+      "learning_rate": 0.000192300444205142,
+      "loss": 3.7346,
+      "step": 3059
+    },
+    {
+      "epoch": 3.91392,
+      "grad_norm": 0.4827929735183716,
+      "learning_rate": 0.0001922600619195046,
+      "loss": 3.7442,
+      "step": 3060
+    },
+    {
+      "epoch": 3.9152,
+      "grad_norm": 0.4668867588043213,
+      "learning_rate": 0.00019221967963386726,
+      "loss": 3.7373,
+      "step": 3061
+    },
+    {
+      "epoch": 3.91648,
+      "grad_norm": 0.4852744936943054,
+      "learning_rate": 0.0001921792973482299,
+      "loss": 3.7558,
+      "step": 3062
+    },
+    {
+      "epoch": 3.91776,
+      "grad_norm": 0.4621356129646301,
+      "learning_rate": 0.00019213891506259252,
+      "loss": 3.7306,
+      "step": 3063
+    },
+    {
+      "epoch": 3.91904,
+      "grad_norm": 0.47092723846435547,
+      "learning_rate": 0.00019209853277695515,
+      "loss": 3.7813,
+      "step": 3064
+    },
+    {
+      "epoch": 3.9203200000000002,
+      "grad_norm": 0.48093530535697937,
+      "learning_rate": 0.0001920581504913178,
+      "loss": 3.7827,
+      "step": 3065
+    },
+    {
+      "epoch": 3.9215999999999998,
+      "grad_norm": 0.4793148338794708,
+      "learning_rate": 0.00019201776820568044,
+      "loss": 3.7267,
+      "step": 3066
+    },
+    {
+      "epoch": 3.92288,
+      "grad_norm": 0.4646039605140686,
+      "learning_rate": 0.00019197738592004307,
+      "loss": 3.7712,
+      "step": 3067
+    },
+    {
+      "epoch": 3.92416,
+      "grad_norm": 0.47865545749664307,
+      "learning_rate": 0.00019193700363440567,
+      "loss": 3.7797,
+      "step": 3068
+    },
+    {
+      "epoch": 3.92544,
+      "grad_norm": 0.4747565984725952,
+      "learning_rate": 0.0001918966213487683,
+      "loss": 3.7688,
+      "step": 3069
+    },
+    {
+      "epoch": 3.92672,
+      "grad_norm": 0.4655320346355438,
+      "learning_rate": 0.00019185623906313096,
+      "loss": 3.8159,
+      "step": 3070
+    },
+    {
+      "epoch": 3.928,
+      "grad_norm": 0.4775558114051819,
+      "learning_rate": 0.0001918158567774936,
+      "loss": 3.7575,
+      "step": 3071
+    },
+    {
+      "epoch": 3.92928,
+      "grad_norm": 0.4783431589603424,
+      "learning_rate": 0.00019177547449185622,
+      "loss": 3.6739,
+      "step": 3072
+    },
+    {
+      "epoch": 3.93056,
+      "grad_norm": 0.48486626148223877,
+      "learning_rate": 0.00019173509220621885,
+      "loss": 3.7762,
+      "step": 3073
+    },
+    {
+      "epoch": 3.9318400000000002,
+      "grad_norm": 0.4990038275718689,
+      "learning_rate": 0.0001916947099205815,
+      "loss": 3.821,
+      "step": 3074
+    },
+    {
+      "epoch": 3.9331199999999997,
+      "grad_norm": 0.4997273087501526,
+      "learning_rate": 0.00019165432763494414,
+      "loss": 3.8262,
+      "step": 3075
+    },
+    {
+      "epoch": 3.9344,
+      "grad_norm": 0.48248130083084106,
+      "learning_rate": 0.00019161394534930674,
+      "loss": 3.8229,
+      "step": 3076
+    },
+    {
+      "epoch": 3.93568,
+      "grad_norm": 0.4947473406791687,
+      "learning_rate": 0.00019157356306366937,
+      "loss": 3.8035,
+      "step": 3077
+    },
+    {
+      "epoch": 3.93696,
+      "grad_norm": 0.49929845333099365,
+      "learning_rate": 0.00019153318077803203,
+      "loss": 3.7907,
+      "step": 3078
+    },
+    {
+      "epoch": 3.93824,
+      "grad_norm": 0.5050824880599976,
+      "learning_rate": 0.00019149279849239466,
+      "loss": 3.7588,
+      "step": 3079
+    },
+    {
+      "epoch": 3.93952,
+      "grad_norm": 0.5087640881538391,
+      "learning_rate": 0.0001914524162067573,
+      "loss": 3.8331,
+      "step": 3080
+    },
+    {
+      "epoch": 3.9408,
+      "grad_norm": 0.4996209442615509,
+      "learning_rate": 0.00019141203392111992,
+      "loss": 3.7626,
+      "step": 3081
+    },
+    {
+      "epoch": 3.94208,
+      "grad_norm": 0.49905964732170105,
+      "learning_rate": 0.00019137165163548255,
+      "loss": 3.7356,
+      "step": 3082
+    },
+    {
+      "epoch": 3.94336,
+      "grad_norm": 0.5181260704994202,
+      "learning_rate": 0.0001913312693498452,
+      "loss": 3.8077,
+      "step": 3083
+    },
+    {
+      "epoch": 3.94464,
+      "grad_norm": 0.5053640604019165,
+      "learning_rate": 0.0001912908870642078,
+      "loss": 3.8205,
+      "step": 3084
+    },
+    {
+      "epoch": 3.94592,
+      "grad_norm": 0.5048377513885498,
+      "learning_rate": 0.00019125050477857044,
+      "loss": 3.7736,
+      "step": 3085
+    },
+    {
+      "epoch": 3.9472,
+      "grad_norm": 0.4769393503665924,
+      "learning_rate": 0.00019121012249293307,
+      "loss": 3.7224,
+      "step": 3086
+    },
+    {
+      "epoch": 3.94848,
+      "grad_norm": 0.5042276382446289,
+      "learning_rate": 0.00019116974020729572,
+      "loss": 3.7301,
+      "step": 3087
+    },
+    {
+      "epoch": 3.94976,
+      "grad_norm": 0.4883654713630676,
+      "learning_rate": 0.00019112935792165835,
+      "loss": 3.7873,
+      "step": 3088
+    },
+    {
+      "epoch": 3.95104,
+      "grad_norm": 0.46280887722969055,
+      "learning_rate": 0.00019108897563602098,
+      "loss": 3.8232,
+      "step": 3089
+    },
+    {
+      "epoch": 3.9523200000000003,
+      "grad_norm": 0.47715169191360474,
+      "learning_rate": 0.00019104859335038361,
+      "loss": 3.8285,
+      "step": 3090
+    },
+    {
+      "epoch": 3.9536,
+      "grad_norm": 0.4663173258304596,
+      "learning_rate": 0.00019100821106474627,
+      "loss": 3.7407,
+      "step": 3091
+    },
+    {
+      "epoch": 3.95488,
+      "grad_norm": 0.4840726852416992,
+      "learning_rate": 0.0001909678287791089,
+      "loss": 3.7961,
+      "step": 3092
+    },
+    {
+      "epoch": 3.95616,
+      "grad_norm": 0.4775456190109253,
+      "learning_rate": 0.0001909274464934715,
+      "loss": 3.7243,
+      "step": 3093
+    },
+    {
+      "epoch": 3.95744,
+      "grad_norm": 0.506077229976654,
+      "learning_rate": 0.00019088706420783414,
+      "loss": 3.832,
+      "step": 3094
+    },
+    {
+      "epoch": 3.95872,
+      "grad_norm": 0.4873894155025482,
+      "learning_rate": 0.00019084668192219677,
+      "loss": 3.8327,
+      "step": 3095
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.4892643690109253,
+      "learning_rate": 0.00019080629963655942,
+      "loss": 3.8131,
+      "step": 3096
+    },
+    {
+      "epoch": 3.96128,
+      "grad_norm": 0.503563642501831,
+      "learning_rate": 0.00019076591735092205,
+      "loss": 3.8291,
+      "step": 3097
+    },
+    {
+      "epoch": 3.96256,
+      "grad_norm": 0.4653085768222809,
+      "learning_rate": 0.00019072553506528468,
+      "loss": 3.7782,
+      "step": 3098
+    },
+    {
+      "epoch": 3.9638400000000003,
+      "grad_norm": 0.46706753969192505,
+      "learning_rate": 0.00019068515277964729,
+      "loss": 3.7926,
+      "step": 3099
+    },
+    {
+      "epoch": 3.9651199999999998,
+      "grad_norm": 0.4738360345363617,
+      "learning_rate": 0.00019064477049400997,
+      "loss": 3.7901,
+      "step": 3100
+    },
+    {
+      "epoch": 3.9664,
+      "grad_norm": 0.48858094215393066,
+      "learning_rate": 0.00019060438820837257,
+      "loss": 3.7761,
+      "step": 3101
+    },
+    {
+      "epoch": 3.96768,
+      "grad_norm": 0.4789169430732727,
+      "learning_rate": 0.0001905640059227352,
+      "loss": 3.8362,
+      "step": 3102
+    },
+    {
+      "epoch": 3.96896,
+      "grad_norm": 0.4818405508995056,
+      "learning_rate": 0.00019052362363709783,
+      "loss": 3.7566,
+      "step": 3103
+    },
+    {
+      "epoch": 3.97024,
+      "grad_norm": 0.4910163879394531,
+      "learning_rate": 0.0001904832413514605,
+      "loss": 3.8183,
+      "step": 3104
+    },
+    {
+      "epoch": 3.97152,
+      "grad_norm": 0.47668853402137756,
+      "learning_rate": 0.00019044285906582312,
+      "loss": 3.7694,
+      "step": 3105
+    },
+    {
+      "epoch": 3.9728,
+      "grad_norm": 0.4871366024017334,
+      "learning_rate": 0.00019040247678018575,
+      "loss": 3.8093,
+      "step": 3106
+    },
+    {
+      "epoch": 3.97408,
+      "grad_norm": 0.4852496087551117,
+      "learning_rate": 0.00019036209449454838,
+      "loss": 3.8424,
+      "step": 3107
+    },
+    {
+      "epoch": 3.9753600000000002,
+      "grad_norm": 0.49890244007110596,
+      "learning_rate": 0.00019032171220891098,
+      "loss": 3.7696,
+      "step": 3108
+    },
+    {
+      "epoch": 3.9766399999999997,
+      "grad_norm": 0.49646714329719543,
+      "learning_rate": 0.00019028132992327364,
+      "loss": 3.8346,
+      "step": 3109
+    },
+    {
+      "epoch": 3.97792,
+      "grad_norm": 0.48989424109458923,
+      "learning_rate": 0.00019024094763763627,
+      "loss": 3.7555,
+      "step": 3110
+    },
+    {
+      "epoch": 3.9792,
+      "grad_norm": 0.4914839267730713,
+      "learning_rate": 0.0001902005653519989,
+      "loss": 3.8055,
+      "step": 3111
+    },
+    {
+      "epoch": 3.98048,
+      "grad_norm": 0.48345229029655457,
+      "learning_rate": 0.00019016018306636153,
+      "loss": 3.8296,
+      "step": 3112
+    },
+    {
+      "epoch": 3.98176,
+      "grad_norm": 0.4965314269065857,
+      "learning_rate": 0.0001901198007807242,
+      "loss": 3.8099,
+      "step": 3113
+    },
+    {
+      "epoch": 3.98304,
+      "grad_norm": 0.47324666380882263,
+      "learning_rate": 0.00019007941849508682,
+      "loss": 3.8194,
+      "step": 3114
+    },
+    {
+      "epoch": 3.98432,
+      "grad_norm": 0.48489847779273987,
+      "learning_rate": 0.00019003903620944945,
+      "loss": 3.792,
+      "step": 3115
+    },
+    {
+      "epoch": 3.9856,
+      "grad_norm": 0.4774338901042938,
+      "learning_rate": 0.00018999865392381205,
+      "loss": 3.8118,
+      "step": 3116
+    },
+    {
+      "epoch": 3.98688,
+      "grad_norm": 0.4876146912574768,
+      "learning_rate": 0.0001899582716381747,
+      "loss": 3.8224,
+      "step": 3117
+    },
+    {
+      "epoch": 3.98816,
+      "grad_norm": 0.4737136662006378,
+      "learning_rate": 0.00018991788935253734,
+      "loss": 3.8342,
+      "step": 3118
+    },
+    {
+      "epoch": 3.98944,
+      "grad_norm": 0.46729955077171326,
+      "learning_rate": 0.00018987750706689997,
+      "loss": 3.7307,
+      "step": 3119
+    },
+    {
+      "epoch": 3.99072,
+      "grad_norm": 0.46377983689308167,
+      "learning_rate": 0.0001898371247812626,
+      "loss": 3.8138,
+      "step": 3120
+    },
+    {
+      "epoch": 3.992,
+      "grad_norm": 0.4843078851699829,
+      "learning_rate": 0.00018979674249562523,
+      "loss": 3.8357,
+      "step": 3121
+    },
+    {
+      "epoch": 3.99328,
+      "grad_norm": 0.47113025188446045,
+      "learning_rate": 0.00018975636020998789,
+      "loss": 3.7578,
+      "step": 3122
+    },
+    {
+      "epoch": 3.99456,
+      "grad_norm": 0.5183811187744141,
+      "learning_rate": 0.00018971597792435052,
+      "loss": 3.7808,
+      "step": 3123
+    },
+    {
+      "epoch": 3.99584,
+      "grad_norm": 0.4774457812309265,
+      "learning_rate": 0.00018967559563871312,
+      "loss": 3.7492,
+      "step": 3124
+    },
+    {
+      "epoch": 3.99712,
+      "grad_norm": 0.5077742338180542,
+      "learning_rate": 0.00018963521335307575,
+      "loss": 3.6986,
+      "step": 3125
+    },
+    {
+      "epoch": 3.9984,
+      "grad_norm": 0.489953875541687,
+      "learning_rate": 0.0001895948310674384,
+      "loss": 3.7662,
+      "step": 3126
+    },
+    {
+      "epoch": 3.99968,
+      "grad_norm": 0.5075393319129944,
+      "learning_rate": 0.00018955444878180104,
+      "loss": 3.7844,
+      "step": 3127
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.8455784916877747,
+      "learning_rate": 0.00018951406649616367,
+      "loss": 3.6669,
+      "step": 3128
+    },
+    {
+      "epoch": 4.00128,
+      "grad_norm": 0.5457375645637512,
+      "learning_rate": 0.0001894736842105263,
+      "loss": 3.696,
+      "step": 3129
+    },
+    {
+      "epoch": 4.00256,
+      "grad_norm": 0.5012986063957214,
+      "learning_rate": 0.00018943330192488895,
+      "loss": 3.7483,
+      "step": 3130
+    },
+    {
+      "epoch": 4.00384,
+      "grad_norm": 0.5168102979660034,
+      "learning_rate": 0.00018939291963925158,
+      "loss": 3.6833,
+      "step": 3131
+    },
+    {
+      "epoch": 4.00512,
+      "grad_norm": 0.5085991024971008,
+      "learning_rate": 0.0001893525373536142,
+      "loss": 3.7756,
+      "step": 3132
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 0.5034124255180359,
+      "learning_rate": 0.00018931215506797682,
+      "loss": 3.6705,
+      "step": 3133
+    },
+    {
+      "epoch": 4.00768,
+      "grad_norm": 0.5143512487411499,
+      "learning_rate": 0.00018927177278233945,
+      "loss": 3.7017,
+      "step": 3134
+    },
+    {
+      "epoch": 4.00896,
+      "grad_norm": 0.49328911304473877,
+      "learning_rate": 0.0001892313904967021,
+      "loss": 3.618,
+      "step": 3135
+    },
+    {
+      "epoch": 4.01024,
+      "grad_norm": 0.49207302927970886,
+      "learning_rate": 0.00018919100821106473,
+      "loss": 3.6588,
+      "step": 3136
+    },
+    {
+      "epoch": 4.01152,
+      "grad_norm": 0.5474900603294373,
+      "learning_rate": 0.00018915062592542736,
+      "loss": 3.7265,
+      "step": 3137
+    },
+    {
+      "epoch": 4.0128,
+      "grad_norm": 0.48699188232421875,
+      "learning_rate": 0.00018911024363979,
+      "loss": 3.7186,
+      "step": 3138
+    },
+    {
+      "epoch": 4.01408,
+      "grad_norm": 0.5149625539779663,
+      "learning_rate": 0.00018906986135415265,
+      "loss": 3.6609,
+      "step": 3139
+    },
+    {
+      "epoch": 4.01536,
+      "grad_norm": 0.5144675374031067,
+      "learning_rate": 0.00018902947906851525,
+      "loss": 3.6982,
+      "step": 3140
+    },
+    {
+      "epoch": 4.01664,
+      "grad_norm": 0.4917832612991333,
+      "learning_rate": 0.00018898909678287788,
+      "loss": 3.6243,
+      "step": 3141
+    },
+    {
+      "epoch": 4.01792,
+      "grad_norm": 0.5480942726135254,
+      "learning_rate": 0.00018894871449724051,
+      "loss": 3.7372,
+      "step": 3142
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.49381646513938904,
+      "learning_rate": 0.00018890833221160317,
+      "loss": 3.6939,
+      "step": 3143
+    },
+    {
+      "epoch": 4.02048,
+      "grad_norm": 0.5019636750221252,
+      "learning_rate": 0.0001888679499259658,
+      "loss": 3.7502,
+      "step": 3144
+    },
+    {
+      "epoch": 4.0217600000000004,
+      "grad_norm": 0.49695998430252075,
+      "learning_rate": 0.00018882756764032843,
+      "loss": 3.7423,
+      "step": 3145
+    },
+    {
+      "epoch": 4.02304,
+      "grad_norm": 0.5167843103408813,
+      "learning_rate": 0.00018878718535469106,
+      "loss": 3.7029,
+      "step": 3146
+    },
+    {
+      "epoch": 4.02432,
+      "grad_norm": 0.49102476239204407,
+      "learning_rate": 0.00018874680306905366,
+      "loss": 3.6505,
+      "step": 3147
+    },
+    {
+      "epoch": 4.0256,
+      "grad_norm": 0.5086076855659485,
+      "learning_rate": 0.00018870642078341632,
+      "loss": 3.654,
+      "step": 3148
+    },
+    {
+      "epoch": 4.02688,
+      "grad_norm": 0.5053331851959229,
+      "learning_rate": 0.00018866603849777895,
+      "loss": 3.6908,
+      "step": 3149
+    },
+    {
+      "epoch": 4.02816,
+      "grad_norm": 0.5181117653846741,
+      "learning_rate": 0.00018862565621214158,
+      "loss": 3.6742,
+      "step": 3150
+    },
+    {
+      "epoch": 4.02944,
+      "grad_norm": 0.4862511157989502,
+      "learning_rate": 0.0001885852739265042,
+      "loss": 3.6927,
+      "step": 3151
+    },
+    {
+      "epoch": 4.03072,
+      "grad_norm": 0.5201453566551208,
+      "learning_rate": 0.00018854489164086687,
+      "loss": 3.7295,
+      "step": 3152
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.4780125319957733,
+      "learning_rate": 0.0001885045093552295,
+      "loss": 3.6679,
+      "step": 3153
+    },
+    {
+      "epoch": 4.03328,
+      "grad_norm": 0.5376641750335693,
+      "learning_rate": 0.00018846412706959213,
+      "loss": 3.7114,
+      "step": 3154
+    },
+    {
+      "epoch": 4.03456,
+      "grad_norm": 0.4939403235912323,
+      "learning_rate": 0.00018842374478395473,
+      "loss": 3.6443,
+      "step": 3155
+    },
+    {
+      "epoch": 4.03584,
+      "grad_norm": 0.502256453037262,
+      "learning_rate": 0.0001883833624983174,
+      "loss": 3.7188,
+      "step": 3156
+    },
+    {
+      "epoch": 4.03712,
+      "grad_norm": 0.49722784757614136,
+      "learning_rate": 0.00018834298021268002,
+      "loss": 3.6703,
+      "step": 3157
+    },
+    {
+      "epoch": 4.0384,
+      "grad_norm": 0.4928574562072754,
+      "learning_rate": 0.00018830259792704265,
+      "loss": 3.6398,
+      "step": 3158
+    },
+    {
+      "epoch": 4.03968,
+      "grad_norm": 0.5021003484725952,
+      "learning_rate": 0.00018826221564140528,
+      "loss": 3.6147,
+      "step": 3159
+    },
+    {
+      "epoch": 4.04096,
+      "grad_norm": 0.49602240324020386,
+      "learning_rate": 0.0001882218333557679,
+      "loss": 3.6448,
+      "step": 3160
+    },
+    {
+      "epoch": 4.04224,
+      "grad_norm": 0.5034336447715759,
+      "learning_rate": 0.00018818145107013057,
+      "loss": 3.7424,
+      "step": 3161
+    },
+    {
+      "epoch": 4.04352,
+      "grad_norm": 0.4804299473762512,
+      "learning_rate": 0.0001881410687844932,
+      "loss": 3.6461,
+      "step": 3162
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.509303092956543,
+      "learning_rate": 0.0001881006864988558,
+      "loss": 3.6735,
+      "step": 3163
+    },
+    {
+      "epoch": 4.04608,
+      "grad_norm": 0.5042724609375,
+      "learning_rate": 0.00018806030421321843,
+      "loss": 3.6822,
+      "step": 3164
+    },
+    {
+      "epoch": 4.04736,
+      "grad_norm": 0.49135807156562805,
+      "learning_rate": 0.0001880199219275811,
+      "loss": 3.6458,
+      "step": 3165
+    },
+    {
+      "epoch": 4.04864,
+      "grad_norm": 0.486905038356781,
+      "learning_rate": 0.00018797953964194372,
+      "loss": 3.6209,
+      "step": 3166
+    },
+    {
+      "epoch": 4.04992,
+      "grad_norm": 0.4827145040035248,
+      "learning_rate": 0.00018793915735630635,
+      "loss": 3.6738,
+      "step": 3167
+    },
+    {
+      "epoch": 4.0512,
+      "grad_norm": 0.4779025912284851,
+      "learning_rate": 0.00018789877507066898,
+      "loss": 3.6753,
+      "step": 3168
+    },
+    {
+      "epoch": 4.05248,
+      "grad_norm": 0.49523454904556274,
+      "learning_rate": 0.00018785839278503163,
+      "loss": 3.6973,
+      "step": 3169
+    },
+    {
+      "epoch": 4.05376,
+      "grad_norm": 0.49853113293647766,
+      "learning_rate": 0.00018781801049939426,
+      "loss": 3.7392,
+      "step": 3170
+    },
+    {
+      "epoch": 4.05504,
+      "grad_norm": 0.500675618648529,
+      "learning_rate": 0.00018777762821375687,
+      "loss": 3.6922,
+      "step": 3171
+    },
+    {
+      "epoch": 4.05632,
+      "grad_norm": 0.49067091941833496,
+      "learning_rate": 0.0001877372459281195,
+      "loss": 3.6531,
+      "step": 3172
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.4818088710308075,
+      "learning_rate": 0.00018769686364248213,
+      "loss": 3.7128,
+      "step": 3173
+    },
+    {
+      "epoch": 4.05888,
+      "grad_norm": 0.5128709673881531,
+      "learning_rate": 0.00018765648135684479,
+      "loss": 3.7044,
+      "step": 3174
+    },
+    {
+      "epoch": 4.06016,
+      "grad_norm": 0.5045835375785828,
+      "learning_rate": 0.00018761609907120742,
+      "loss": 3.6572,
+      "step": 3175
+    },
+    {
+      "epoch": 4.06144,
+      "grad_norm": 0.4782851040363312,
+      "learning_rate": 0.00018757571678557005,
+      "loss": 3.7127,
+      "step": 3176
+    },
+    {
+      "epoch": 4.06272,
+      "grad_norm": 0.4840090274810791,
+      "learning_rate": 0.00018753533449993268,
+      "loss": 3.6702,
+      "step": 3177
+    },
+    {
+      "epoch": 4.064,
+      "grad_norm": 0.49672845005989075,
+      "learning_rate": 0.00018749495221429533,
+      "loss": 3.6537,
+      "step": 3178
+    },
+    {
+      "epoch": 4.06528,
+      "grad_norm": 0.4983557462692261,
+      "learning_rate": 0.00018745456992865796,
+      "loss": 3.6794,
+      "step": 3179
+    },
+    {
+      "epoch": 4.06656,
+      "grad_norm": 0.5004916191101074,
+      "learning_rate": 0.00018741418764302057,
+      "loss": 3.6771,
+      "step": 3180
+    },
+    {
+      "epoch": 4.06784,
+      "grad_norm": 0.5043333768844604,
+      "learning_rate": 0.0001873738053573832,
+      "loss": 3.7056,
+      "step": 3181
+    },
+    {
+      "epoch": 4.06912,
+      "grad_norm": 0.4966271221637726,
+      "learning_rate": 0.00018733342307174585,
+      "loss": 3.6367,
+      "step": 3182
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.5312650203704834,
+      "learning_rate": 0.00018729304078610848,
+      "loss": 3.6806,
+      "step": 3183
+    },
+    {
+      "epoch": 4.07168,
+      "grad_norm": 0.48910966515541077,
+      "learning_rate": 0.0001872526585004711,
+      "loss": 3.726,
+      "step": 3184
+    },
+    {
+      "epoch": 4.07296,
+      "grad_norm": 0.5237113237380981,
+      "learning_rate": 0.00018721227621483374,
+      "loss": 3.6479,
+      "step": 3185
+    },
+    {
+      "epoch": 4.07424,
+      "grad_norm": 0.48384127020835876,
+      "learning_rate": 0.0001871718939291964,
+      "loss": 3.7127,
+      "step": 3186
+    },
+    {
+      "epoch": 4.07552,
+      "grad_norm": 0.5084925889968872,
+      "learning_rate": 0.00018713151164355903,
+      "loss": 3.6814,
+      "step": 3187
+    },
+    {
+      "epoch": 4.0768,
+      "grad_norm": 0.5011091828346252,
+      "learning_rate": 0.00018709112935792163,
+      "loss": 3.6603,
+      "step": 3188
+    },
+    {
+      "epoch": 4.07808,
+      "grad_norm": 0.4861987829208374,
+      "learning_rate": 0.00018705074707228426,
+      "loss": 3.7693,
+      "step": 3189
+    },
+    {
+      "epoch": 4.07936,
+      "grad_norm": 0.5180191993713379,
+      "learning_rate": 0.0001870103647866469,
+      "loss": 3.6892,
+      "step": 3190
+    },
+    {
+      "epoch": 4.08064,
+      "grad_norm": 0.48718804121017456,
+      "learning_rate": 0.00018696998250100955,
+      "loss": 3.6065,
+      "step": 3191
+    },
+    {
+      "epoch": 4.08192,
+      "grad_norm": 0.520859956741333,
+      "learning_rate": 0.00018692960021537218,
+      "loss": 3.656,
+      "step": 3192
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.49004432559013367,
+      "learning_rate": 0.0001868892179297348,
+      "loss": 3.6136,
+      "step": 3193
+    },
+    {
+      "epoch": 4.08448,
+      "grad_norm": 0.5408604741096497,
+      "learning_rate": 0.00018684883564409744,
+      "loss": 3.6482,
+      "step": 3194
+    },
+    {
+      "epoch": 4.08576,
+      "grad_norm": 0.4856225252151489,
+      "learning_rate": 0.0001868084533584601,
+      "loss": 3.6174,
+      "step": 3195
+    },
+    {
+      "epoch": 4.08704,
+      "grad_norm": 0.5170305371284485,
+      "learning_rate": 0.0001867680710728227,
+      "loss": 3.7183,
+      "step": 3196
+    },
+    {
+      "epoch": 4.08832,
+      "grad_norm": 0.5067542791366577,
+      "learning_rate": 0.00018672768878718533,
+      "loss": 3.6856,
+      "step": 3197
+    },
+    {
+      "epoch": 4.0896,
+      "grad_norm": 0.4944978952407837,
+      "learning_rate": 0.00018668730650154796,
+      "loss": 3.6034,
+      "step": 3198
+    },
+    {
+      "epoch": 4.09088,
+      "grad_norm": 0.5182904601097107,
+      "learning_rate": 0.00018664692421591062,
+      "loss": 3.7634,
+      "step": 3199
+    },
+    {
+      "epoch": 4.09216,
+      "grad_norm": 0.5076016187667847,
+      "learning_rate": 0.00018660654193027325,
+      "loss": 3.6408,
+      "step": 3200
+    },
+    {
+      "epoch": 4.09344,
+      "grad_norm": 0.4935440123081207,
+      "learning_rate": 0.00018656615964463588,
+      "loss": 3.6573,
+      "step": 3201
+    },
+    {
+      "epoch": 4.09472,
+      "grad_norm": 0.5223529934883118,
+      "learning_rate": 0.0001865257773589985,
+      "loss": 3.6398,
+      "step": 3202
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.4967701733112335,
+      "learning_rate": 0.0001864853950733611,
+      "loss": 3.6774,
+      "step": 3203
+    },
+    {
+      "epoch": 4.09728,
+      "grad_norm": 0.5154727697372437,
+      "learning_rate": 0.00018644501278772377,
+      "loss": 3.7467,
+      "step": 3204
+    },
+    {
+      "epoch": 4.09856,
+      "grad_norm": 0.5095906257629395,
+      "learning_rate": 0.0001864046305020864,
+      "loss": 3.7121,
+      "step": 3205
+    },
+    {
+      "epoch": 4.09984,
+      "grad_norm": 0.5027210116386414,
+      "learning_rate": 0.00018636424821644903,
+      "loss": 3.671,
+      "step": 3206
+    },
+    {
+      "epoch": 4.10112,
+      "grad_norm": 0.537636399269104,
+      "learning_rate": 0.00018632386593081166,
+      "loss": 3.7246,
+      "step": 3207
+    },
+    {
+      "epoch": 4.1024,
+      "grad_norm": 0.4799419641494751,
+      "learning_rate": 0.00018628348364517432,
+      "loss": 3.7003,
+      "step": 3208
+    },
+    {
+      "epoch": 4.10368,
+      "grad_norm": 0.5242260098457336,
+      "learning_rate": 0.00018624310135953695,
+      "loss": 3.6924,
+      "step": 3209
+    },
+    {
+      "epoch": 4.10496,
+      "grad_norm": 0.5360248684883118,
+      "learning_rate": 0.00018620271907389958,
+      "loss": 3.6041,
+      "step": 3210
+    },
+    {
+      "epoch": 4.10624,
+      "grad_norm": 0.5119033455848694,
+      "learning_rate": 0.00018616233678826218,
+      "loss": 3.6326,
+      "step": 3211
+    },
+    {
+      "epoch": 4.10752,
+      "grad_norm": 0.5700018405914307,
+      "learning_rate": 0.00018612195450262484,
+      "loss": 3.7075,
+      "step": 3212
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.5085955262184143,
+      "learning_rate": 0.00018608157221698747,
+      "loss": 3.6672,
+      "step": 3213
+    },
+    {
+      "epoch": 4.11008,
+      "grad_norm": 0.5086491703987122,
+      "learning_rate": 0.0001860411899313501,
+      "loss": 3.6875,
+      "step": 3214
+    },
+    {
+      "epoch": 4.11136,
+      "grad_norm": 0.5247465968132019,
+      "learning_rate": 0.00018600080764571273,
+      "loss": 3.6448,
+      "step": 3215
+    },
+    {
+      "epoch": 4.11264,
+      "grad_norm": 0.5041964054107666,
+      "learning_rate": 0.00018596042536007536,
+      "loss": 3.6232,
+      "step": 3216
+    },
+    {
+      "epoch": 4.11392,
+      "grad_norm": 0.5098519325256348,
+      "learning_rate": 0.00018592004307443801,
+      "loss": 3.6918,
+      "step": 3217
+    },
+    {
+      "epoch": 4.1152,
+      "grad_norm": 0.5236632227897644,
+      "learning_rate": 0.00018587966078880064,
+      "loss": 3.6658,
+      "step": 3218
+    },
+    {
+      "epoch": 4.11648,
+      "grad_norm": 0.4816301465034485,
+      "learning_rate": 0.00018583927850316325,
+      "loss": 3.7479,
+      "step": 3219
+    },
+    {
+      "epoch": 4.11776,
+      "grad_norm": 0.5393995642662048,
+      "learning_rate": 0.00018579889621752588,
+      "loss": 3.6477,
+      "step": 3220
+    },
+    {
+      "epoch": 4.11904,
+      "grad_norm": 0.4957987666130066,
+      "learning_rate": 0.00018575851393188853,
+      "loss": 3.7106,
+      "step": 3221
+    },
+    {
+      "epoch": 4.12032,
+      "grad_norm": 0.5068425536155701,
+      "learning_rate": 0.00018571813164625116,
+      "loss": 3.6857,
+      "step": 3222
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.5265698432922363,
+      "learning_rate": 0.0001856777493606138,
+      "loss": 3.7724,
+      "step": 3223
+    },
+    {
+      "epoch": 4.12288,
+      "grad_norm": 0.5055711269378662,
+      "learning_rate": 0.00018563736707497642,
+      "loss": 3.7152,
+      "step": 3224
+    },
+    {
+      "epoch": 4.12416,
+      "grad_norm": 0.5185781717300415,
+      "learning_rate": 0.00018559698478933908,
+      "loss": 3.729,
+      "step": 3225
+    },
+    {
+      "epoch": 4.12544,
+      "grad_norm": 0.5067470669746399,
+      "learning_rate": 0.0001855566025037017,
+      "loss": 3.7011,
+      "step": 3226
+    },
+    {
+      "epoch": 4.12672,
+      "grad_norm": 0.5095682144165039,
+      "learning_rate": 0.00018551622021806431,
+      "loss": 3.6616,
+      "step": 3227
+    },
+    {
+      "epoch": 4.128,
+      "grad_norm": 0.5128963589668274,
+      "learning_rate": 0.00018547583793242694,
+      "loss": 3.6418,
+      "step": 3228
+    },
+    {
+      "epoch": 4.12928,
+      "grad_norm": 0.5066363215446472,
+      "learning_rate": 0.00018543545564678957,
+      "loss": 3.6522,
+      "step": 3229
+    },
+    {
+      "epoch": 4.13056,
+      "grad_norm": 0.5100614428520203,
+      "learning_rate": 0.00018539507336115223,
+      "loss": 3.6805,
+      "step": 3230
+    },
+    {
+      "epoch": 4.13184,
+      "grad_norm": 0.5063032507896423,
+      "learning_rate": 0.00018535469107551486,
+      "loss": 3.654,
+      "step": 3231
+    },
+    {
+      "epoch": 4.13312,
+      "grad_norm": 0.5069268345832825,
+      "learning_rate": 0.0001853143087898775,
+      "loss": 3.7731,
+      "step": 3232
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.5150942206382751,
+      "learning_rate": 0.00018527392650424012,
+      "loss": 3.7019,
+      "step": 3233
+    },
+    {
+      "epoch": 4.13568,
+      "grad_norm": 0.4924696087837219,
+      "learning_rate": 0.00018523354421860278,
+      "loss": 3.7097,
+      "step": 3234
+    },
+    {
+      "epoch": 4.13696,
+      "grad_norm": 0.5000399351119995,
+      "learning_rate": 0.00018519316193296538,
+      "loss": 3.6958,
+      "step": 3235
+    },
+    {
+      "epoch": 4.13824,
+      "grad_norm": 0.468902051448822,
+      "learning_rate": 0.000185152779647328,
+      "loss": 3.6186,
+      "step": 3236
+    },
+    {
+      "epoch": 4.13952,
+      "grad_norm": 0.49955734610557556,
+      "learning_rate": 0.00018511239736169064,
+      "loss": 3.7507,
+      "step": 3237
+    },
+    {
+      "epoch": 4.1408,
+      "grad_norm": 0.5048097372055054,
+      "learning_rate": 0.0001850720150760533,
+      "loss": 3.6257,
+      "step": 3238
+    },
+    {
+      "epoch": 4.14208,
+      "grad_norm": 0.4985654056072235,
+      "learning_rate": 0.00018503163279041593,
+      "loss": 3.6897,
+      "step": 3239
+    },
+    {
+      "epoch": 4.14336,
+      "grad_norm": 0.4974263310432434,
+      "learning_rate": 0.00018499125050477856,
+      "loss": 3.7192,
+      "step": 3240
+    },
+    {
+      "epoch": 4.14464,
+      "grad_norm": 0.5002862215042114,
+      "learning_rate": 0.0001849508682191412,
+      "loss": 3.694,
+      "step": 3241
+    },
+    {
+      "epoch": 4.14592,
+      "grad_norm": 0.49395427107810974,
+      "learning_rate": 0.0001849104859335038,
+      "loss": 3.7136,
+      "step": 3242
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.5043503046035767,
+      "learning_rate": 0.00018487010364786645,
+      "loss": 3.6868,
+      "step": 3243
+    },
+    {
+      "epoch": 4.14848,
+      "grad_norm": 0.4990045726299286,
+      "learning_rate": 0.00018482972136222908,
+      "loss": 3.6802,
+      "step": 3244
+    },
+    {
+      "epoch": 4.14976,
+      "grad_norm": 0.5021066069602966,
+      "learning_rate": 0.0001847893390765917,
+      "loss": 3.6486,
+      "step": 3245
+    },
+    {
+      "epoch": 4.15104,
+      "grad_norm": 0.48476442694664,
+      "learning_rate": 0.00018474895679095434,
+      "loss": 3.6201,
+      "step": 3246
+    },
+    {
+      "epoch": 4.15232,
+      "grad_norm": 0.5046278238296509,
+      "learning_rate": 0.000184708574505317,
+      "loss": 3.7327,
+      "step": 3247
+    },
+    {
+      "epoch": 4.1536,
+      "grad_norm": 0.49205055832862854,
+      "learning_rate": 0.00018466819221967963,
+      "loss": 3.7463,
+      "step": 3248
+    },
+    {
+      "epoch": 4.15488,
+      "grad_norm": 0.4804619252681732,
+      "learning_rate": 0.00018462780993404226,
+      "loss": 3.5683,
+      "step": 3249
+    },
+    {
+      "epoch": 4.15616,
+      "grad_norm": 0.4981594979763031,
+      "learning_rate": 0.00018458742764840486,
+      "loss": 3.6402,
+      "step": 3250
+    },
+    {
+      "epoch": 4.15744,
+      "grad_norm": 0.4988977313041687,
+      "learning_rate": 0.00018454704536276755,
+      "loss": 3.6497,
+      "step": 3251
+    },
+    {
+      "epoch": 4.15872,
+      "grad_norm": 0.5172099471092224,
+      "learning_rate": 0.00018450666307713015,
+      "loss": 3.6574,
+      "step": 3252
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.5064343810081482,
+      "learning_rate": 0.00018446628079149278,
+      "loss": 3.6283,
+      "step": 3253
+    },
+    {
+      "epoch": 4.16128,
+      "grad_norm": 0.47614553570747375,
+      "learning_rate": 0.0001844258985058554,
+      "loss": 3.6493,
+      "step": 3254
+    },
+    {
+      "epoch": 4.16256,
+      "grad_norm": 0.48512086272239685,
+      "learning_rate": 0.00018438551622021804,
+      "loss": 3.6168,
+      "step": 3255
+    },
+    {
+      "epoch": 4.16384,
+      "grad_norm": 0.5282117128372192,
+      "learning_rate": 0.0001843451339345807,
+      "loss": 3.7312,
+      "step": 3256
+    },
+    {
+      "epoch": 4.16512,
+      "grad_norm": 0.4799935817718506,
+      "learning_rate": 0.00018430475164894333,
+      "loss": 3.6321,
+      "step": 3257
+    },
+    {
+      "epoch": 4.1664,
+      "grad_norm": 0.5284485816955566,
+      "learning_rate": 0.00018426436936330593,
+      "loss": 3.6389,
+      "step": 3258
+    },
+    {
+      "epoch": 4.16768,
+      "grad_norm": 0.49284806847572327,
+      "learning_rate": 0.00018422398707766856,
+      "loss": 3.7423,
+      "step": 3259
+    },
+    {
+      "epoch": 4.16896,
+      "grad_norm": 0.5154812335968018,
+      "learning_rate": 0.00018418360479203122,
+      "loss": 3.6436,
+      "step": 3260
+    },
+    {
+      "epoch": 4.17024,
+      "grad_norm": 0.49612143635749817,
+      "learning_rate": 0.00018414322250639385,
+      "loss": 3.6282,
+      "step": 3261
+    },
+    {
+      "epoch": 4.17152,
+      "grad_norm": 0.5098751783370972,
+      "learning_rate": 0.00018410284022075648,
+      "loss": 3.7001,
+      "step": 3262
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.5058940649032593,
+      "learning_rate": 0.0001840624579351191,
+      "loss": 3.7325,
+      "step": 3263
+    },
+    {
+      "epoch": 4.17408,
+      "grad_norm": 0.4903821051120758,
+      "learning_rate": 0.00018402207564948176,
+      "loss": 3.7294,
+      "step": 3264
+    },
+    {
+      "epoch": 4.17536,
+      "grad_norm": 0.5205444693565369,
+      "learning_rate": 0.0001839816933638444,
+      "loss": 3.748,
+      "step": 3265
+    },
+    {
+      "epoch": 4.17664,
+      "grad_norm": 0.5151527523994446,
+      "learning_rate": 0.00018394131107820702,
+      "loss": 3.7253,
+      "step": 3266
+    },
+    {
+      "epoch": 4.17792,
+      "grad_norm": 0.5021133422851562,
+      "learning_rate": 0.00018390092879256963,
+      "loss": 3.7138,
+      "step": 3267
+    },
+    {
+      "epoch": 4.1792,
+      "grad_norm": 0.5128991007804871,
+      "learning_rate": 0.00018386054650693226,
+      "loss": 3.695,
+      "step": 3268
+    },
+    {
+      "epoch": 4.18048,
+      "grad_norm": 0.4980866014957428,
+      "learning_rate": 0.0001838201642212949,
+      "loss": 3.6914,
+      "step": 3269
+    },
+    {
+      "epoch": 4.18176,
+      "grad_norm": 0.4788151681423187,
+      "learning_rate": 0.00018377978193565754,
+      "loss": 3.7036,
+      "step": 3270
+    },
+    {
+      "epoch": 4.18304,
+      "grad_norm": 0.5027557611465454,
+      "learning_rate": 0.00018373939965002017,
+      "loss": 3.6309,
+      "step": 3271
+    },
+    {
+      "epoch": 4.18432,
+      "grad_norm": 0.4961322247982025,
+      "learning_rate": 0.0001836990173643828,
+      "loss": 3.7176,
+      "step": 3272
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.48880982398986816,
+      "learning_rate": 0.00018365863507874546,
+      "loss": 3.6192,
+      "step": 3273
+    },
+    {
+      "epoch": 4.18688,
+      "grad_norm": 0.5047913193702698,
+      "learning_rate": 0.0001836182527931081,
+      "loss": 3.6584,
+      "step": 3274
+    },
+    {
+      "epoch": 4.18816,
+      "grad_norm": 0.5131552815437317,
+      "learning_rate": 0.0001835778705074707,
+      "loss": 3.6554,
+      "step": 3275
+    },
+    {
+      "epoch": 4.18944,
+      "grad_norm": 0.4939292073249817,
+      "learning_rate": 0.00018353748822183332,
+      "loss": 3.6149,
+      "step": 3276
+    },
+    {
+      "epoch": 4.19072,
+      "grad_norm": 0.5033706426620483,
+      "learning_rate": 0.00018349710593619598,
+      "loss": 3.704,
+      "step": 3277
+    },
+    {
+      "epoch": 4.192,
+      "grad_norm": 0.4993079602718353,
+      "learning_rate": 0.0001834567236505586,
+      "loss": 3.7489,
+      "step": 3278
+    },
+    {
+      "epoch": 4.19328,
+      "grad_norm": 0.49270492792129517,
+      "learning_rate": 0.00018341634136492124,
+      "loss": 3.6576,
+      "step": 3279
+    },
+    {
+      "epoch": 4.19456,
+      "grad_norm": 0.513554036617279,
+      "learning_rate": 0.00018337595907928387,
+      "loss": 3.6928,
+      "step": 3280
+    },
+    {
+      "epoch": 4.19584,
+      "grad_norm": 0.4733978509902954,
+      "learning_rate": 0.0001833355767936465,
+      "loss": 3.6538,
+      "step": 3281
+    },
+    {
+      "epoch": 4.19712,
+      "grad_norm": 0.5104435086250305,
+      "learning_rate": 0.00018329519450800916,
+      "loss": 3.637,
+      "step": 3282
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.49731189012527466,
+      "learning_rate": 0.00018325481222237176,
+      "loss": 3.7184,
+      "step": 3283
+    },
+    {
+      "epoch": 4.19968,
+      "grad_norm": 0.5013545155525208,
+      "learning_rate": 0.0001832144299367344,
+      "loss": 3.6749,
+      "step": 3284
+    },
+    {
+      "epoch": 4.20096,
+      "grad_norm": 0.5214062929153442,
+      "learning_rate": 0.00018317404765109702,
+      "loss": 3.6357,
+      "step": 3285
+    },
+    {
+      "epoch": 4.20224,
+      "grad_norm": 0.4991886615753174,
+      "learning_rate": 0.00018313366536545968,
+      "loss": 3.6431,
+      "step": 3286
+    },
+    {
+      "epoch": 4.20352,
+      "grad_norm": 0.5190424919128418,
+      "learning_rate": 0.0001830932830798223,
+      "loss": 3.6487,
+      "step": 3287
+    },
+    {
+      "epoch": 4.2048,
+      "grad_norm": 0.4960169792175293,
+      "learning_rate": 0.00018305290079418494,
+      "loss": 3.7144,
+      "step": 3288
+    },
+    {
+      "epoch": 4.20608,
+      "grad_norm": 0.49390795826911926,
+      "learning_rate": 0.00018301251850854757,
+      "loss": 3.6899,
+      "step": 3289
+    },
+    {
+      "epoch": 4.2073599999999995,
+      "grad_norm": 0.5029337406158447,
+      "learning_rate": 0.00018297213622291023,
+      "loss": 3.7521,
+      "step": 3290
+    },
+    {
+      "epoch": 4.20864,
+      "grad_norm": 0.48957720398902893,
+      "learning_rate": 0.00018293175393727283,
+      "loss": 3.7307,
+      "step": 3291
+    },
+    {
+      "epoch": 4.20992,
+      "grad_norm": 0.48470643162727356,
+      "learning_rate": 0.00018289137165163546,
+      "loss": 3.6277,
+      "step": 3292
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.4815894663333893,
+      "learning_rate": 0.0001828509893659981,
+      "loss": 3.7091,
+      "step": 3293
+    },
+    {
+      "epoch": 4.21248,
+      "grad_norm": 0.49646440148353577,
+      "learning_rate": 0.00018281060708036072,
+      "loss": 3.7575,
+      "step": 3294
+    },
+    {
+      "epoch": 4.21376,
+      "grad_norm": 0.4831089675426483,
+      "learning_rate": 0.00018277022479472338,
+      "loss": 3.745,
+      "step": 3295
+    },
+    {
+      "epoch": 4.21504,
+      "grad_norm": 0.5089125633239746,
+      "learning_rate": 0.000182729842509086,
+      "loss": 3.7263,
+      "step": 3296
+    },
+    {
+      "epoch": 4.21632,
+      "grad_norm": 0.4908083975315094,
+      "learning_rate": 0.00018268946022344864,
+      "loss": 3.6735,
+      "step": 3297
+    },
+    {
+      "epoch": 4.2176,
+      "grad_norm": 0.46788400411605835,
+      "learning_rate": 0.00018264907793781124,
+      "loss": 3.6856,
+      "step": 3298
+    },
+    {
+      "epoch": 4.21888,
+      "grad_norm": 0.4864565432071686,
+      "learning_rate": 0.0001826086956521739,
+      "loss": 3.6435,
+      "step": 3299
+    },
+    {
+      "epoch": 4.22016,
+      "grad_norm": 0.5052434802055359,
+      "learning_rate": 0.00018256831336653653,
+      "loss": 3.6445,
+      "step": 3300
+    },
+    {
+      "epoch": 4.22144,
+      "grad_norm": 0.47929632663726807,
+      "learning_rate": 0.00018252793108089916,
+      "loss": 3.6773,
+      "step": 3301
+    },
+    {
+      "epoch": 4.22272,
+      "grad_norm": 0.522189736366272,
+      "learning_rate": 0.0001824875487952618,
+      "loss": 3.6605,
+      "step": 3302
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.5014151334762573,
+      "learning_rate": 0.00018244716650962444,
+      "loss": 3.6866,
+      "step": 3303
+    },
+    {
+      "epoch": 4.22528,
+      "grad_norm": 0.49791914224624634,
+      "learning_rate": 0.00018240678422398707,
+      "loss": 3.7027,
+      "step": 3304
+    },
+    {
+      "epoch": 4.22656,
+      "grad_norm": 0.5259125828742981,
+      "learning_rate": 0.0001823664019383497,
+      "loss": 3.7465,
+      "step": 3305
+    },
+    {
+      "epoch": 4.22784,
+      "grad_norm": 0.49660128355026245,
+      "learning_rate": 0.0001823260196527123,
+      "loss": 3.7207,
+      "step": 3306
+    },
+    {
+      "epoch": 4.22912,
+      "grad_norm": 0.5191150307655334,
+      "learning_rate": 0.00018228563736707496,
+      "loss": 3.8187,
+      "step": 3307
+    },
+    {
+      "epoch": 4.2304,
+      "grad_norm": 0.5008911490440369,
+      "learning_rate": 0.0001822452550814376,
+      "loss": 3.6294,
+      "step": 3308
+    },
+    {
+      "epoch": 4.23168,
+      "grad_norm": 0.4917088449001312,
+      "learning_rate": 0.00018220487279580022,
+      "loss": 3.7032,
+      "step": 3309
+    },
+    {
+      "epoch": 4.23296,
+      "grad_norm": 0.5173110961914062,
+      "learning_rate": 0.00018216449051016285,
+      "loss": 3.7054,
+      "step": 3310
+    },
+    {
+      "epoch": 4.23424,
+      "grad_norm": 0.5040363073348999,
+      "learning_rate": 0.00018212410822452549,
+      "loss": 3.7053,
+      "step": 3311
+    },
+    {
+      "epoch": 4.23552,
+      "grad_norm": 0.48391327261924744,
+      "learning_rate": 0.00018208372593888814,
+      "loss": 3.6792,
+      "step": 3312
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.5244686603546143,
+      "learning_rate": 0.00018204334365325077,
+      "loss": 3.6416,
+      "step": 3313
+    },
+    {
+      "epoch": 4.23808,
+      "grad_norm": 0.5149886608123779,
+      "learning_rate": 0.00018200296136761338,
+      "loss": 3.7686,
+      "step": 3314
+    },
+    {
+      "epoch": 4.23936,
+      "grad_norm": 0.5035302042961121,
+      "learning_rate": 0.000181962579081976,
+      "loss": 3.7187,
+      "step": 3315
+    },
+    {
+      "epoch": 4.24064,
+      "grad_norm": 0.5213099122047424,
+      "learning_rate": 0.00018192219679633866,
+      "loss": 3.7372,
+      "step": 3316
+    },
+    {
+      "epoch": 4.24192,
+      "grad_norm": 0.49687349796295166,
+      "learning_rate": 0.0001818818145107013,
+      "loss": 3.705,
+      "step": 3317
+    },
+    {
+      "epoch": 4.2432,
+      "grad_norm": 0.49955183267593384,
+      "learning_rate": 0.00018184143222506392,
+      "loss": 3.6738,
+      "step": 3318
+    },
+    {
+      "epoch": 4.24448,
+      "grad_norm": 0.49271631240844727,
+      "learning_rate": 0.00018180104993942655,
+      "loss": 3.671,
+      "step": 3319
+    },
+    {
+      "epoch": 4.24576,
+      "grad_norm": 0.4815945625305176,
+      "learning_rate": 0.0001817606676537892,
+      "loss": 3.7131,
+      "step": 3320
+    },
+    {
+      "epoch": 4.24704,
+      "grad_norm": 0.5155565142631531,
+      "learning_rate": 0.00018172028536815184,
+      "loss": 3.6281,
+      "step": 3321
+    },
+    {
+      "epoch": 4.24832,
+      "grad_norm": 0.48677995800971985,
+      "learning_rate": 0.00018167990308251444,
+      "loss": 3.6267,
+      "step": 3322
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.5277391076087952,
+      "learning_rate": 0.00018163952079687707,
+      "loss": 3.6703,
+      "step": 3323
+    },
+    {
+      "epoch": 4.25088,
+      "grad_norm": 0.4805748164653778,
+      "learning_rate": 0.0001815991385112397,
+      "loss": 3.6601,
+      "step": 3324
+    },
+    {
+      "epoch": 4.25216,
+      "grad_norm": 0.527442216873169,
+      "learning_rate": 0.00018155875622560236,
+      "loss": 3.7159,
+      "step": 3325
+    },
+    {
+      "epoch": 4.25344,
+      "grad_norm": 0.48253729939460754,
+      "learning_rate": 0.000181518373939965,
+      "loss": 3.6895,
+      "step": 3326
+    },
+    {
+      "epoch": 4.25472,
+      "grad_norm": 0.49663540720939636,
+      "learning_rate": 0.00018147799165432762,
+      "loss": 3.7617,
+      "step": 3327
+    },
+    {
+      "epoch": 4.256,
+      "grad_norm": 0.48462435603141785,
+      "learning_rate": 0.00018143760936869025,
+      "loss": 3.6659,
+      "step": 3328
+    },
+    {
+      "epoch": 4.25728,
+      "grad_norm": 0.4957021474838257,
+      "learning_rate": 0.0001813972270830529,
+      "loss": 3.6396,
+      "step": 3329
+    },
+    {
+      "epoch": 4.25856,
+      "grad_norm": 0.4980667233467102,
+      "learning_rate": 0.0001813568447974155,
+      "loss": 3.6964,
+      "step": 3330
+    },
+    {
+      "epoch": 4.25984,
+      "grad_norm": 0.47956860065460205,
+      "learning_rate": 0.00018131646251177814,
+      "loss": 3.6877,
+      "step": 3331
+    },
+    {
+      "epoch": 4.26112,
+      "grad_norm": 0.4983195960521698,
+      "learning_rate": 0.00018127608022614077,
+      "loss": 3.6595,
+      "step": 3332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.5003688335418701,
+      "learning_rate": 0.00018123569794050343,
+      "loss": 3.7105,
+      "step": 3333
+    },
+    {
+      "epoch": 4.26368,
+      "grad_norm": 0.4797384738922119,
+      "learning_rate": 0.00018119531565486606,
+      "loss": 3.7471,
+      "step": 3334
+    },
+    {
+      "epoch": 4.26496,
+      "grad_norm": 0.5109444856643677,
+      "learning_rate": 0.0001811549333692287,
+      "loss": 3.6958,
+      "step": 3335
+    },
+    {
+      "epoch": 4.26624,
+      "grad_norm": 0.4882170557975769,
+      "learning_rate": 0.00018111455108359132,
+      "loss": 3.6789,
+      "step": 3336
+    },
+    {
+      "epoch": 4.26752,
+      "grad_norm": 0.5124632120132446,
+      "learning_rate": 0.00018107416879795392,
+      "loss": 3.702,
+      "step": 3337
+    },
+    {
+      "epoch": 4.2688,
+      "grad_norm": 0.500720739364624,
+      "learning_rate": 0.0001810337865123166,
+      "loss": 3.6423,
+      "step": 3338
+    },
+    {
+      "epoch": 4.27008,
+      "grad_norm": 0.49544718861579895,
+      "learning_rate": 0.0001809934042266792,
+      "loss": 3.7047,
+      "step": 3339
+    },
+    {
+      "epoch": 4.27136,
+      "grad_norm": 0.5047063231468201,
+      "learning_rate": 0.00018095302194104184,
+      "loss": 3.7542,
+      "step": 3340
+    },
+    {
+      "epoch": 4.27264,
+      "grad_norm": 0.4655781388282776,
+      "learning_rate": 0.00018091263965540447,
+      "loss": 3.7705,
+      "step": 3341
+    },
+    {
+      "epoch": 4.27392,
+      "grad_norm": 0.4966285228729248,
+      "learning_rate": 0.00018087225736976713,
+      "loss": 3.6768,
+      "step": 3342
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.4942556321620941,
+      "learning_rate": 0.00018083187508412976,
+      "loss": 3.593,
+      "step": 3343
+    },
+    {
+      "epoch": 4.27648,
+      "grad_norm": 0.517810046672821,
+      "learning_rate": 0.00018079149279849239,
+      "loss": 3.6881,
+      "step": 3344
+    },
+    {
+      "epoch": 4.27776,
+      "grad_norm": 0.4855895936489105,
+      "learning_rate": 0.000180751110512855,
+      "loss": 3.6054,
+      "step": 3345
+    },
+    {
+      "epoch": 4.27904,
+      "grad_norm": 0.5209439396858215,
+      "learning_rate": 0.00018071072822721767,
+      "loss": 3.6611,
+      "step": 3346
+    },
+    {
+      "epoch": 4.28032,
+      "grad_norm": 0.5114684700965881,
+      "learning_rate": 0.00018067034594158028,
+      "loss": 3.6896,
+      "step": 3347
+    },
+    {
+      "epoch": 4.2816,
+      "grad_norm": 0.49628081917762756,
+      "learning_rate": 0.0001806299636559429,
+      "loss": 3.7266,
+      "step": 3348
+    },
+    {
+      "epoch": 4.2828800000000005,
+      "grad_norm": 0.4987524747848511,
+      "learning_rate": 0.00018058958137030554,
+      "loss": 3.6737,
+      "step": 3349
+    },
+    {
+      "epoch": 4.28416,
+      "grad_norm": 0.4812726676464081,
+      "learning_rate": 0.00018054919908466817,
+      "loss": 3.6078,
+      "step": 3350
+    },
+    {
+      "epoch": 4.28544,
+      "grad_norm": 0.5009292960166931,
+      "learning_rate": 0.00018050881679903082,
+      "loss": 3.7269,
+      "step": 3351
+    },
+    {
+      "epoch": 4.28672,
+      "grad_norm": 0.5155140161514282,
+      "learning_rate": 0.00018046843451339345,
+      "loss": 3.6942,
+      "step": 3352
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.5101702809333801,
+      "learning_rate": 0.00018042805222775608,
+      "loss": 3.7346,
+      "step": 3353
+    },
+    {
+      "epoch": 4.28928,
+      "grad_norm": 0.49652808904647827,
+      "learning_rate": 0.0001803876699421187,
+      "loss": 3.6923,
+      "step": 3354
+    },
+    {
+      "epoch": 4.29056,
+      "grad_norm": 0.5246519446372986,
+      "learning_rate": 0.00018034728765648134,
+      "loss": 3.6841,
+      "step": 3355
+    },
+    {
+      "epoch": 4.29184,
+      "grad_norm": 0.49491071701049805,
+      "learning_rate": 0.00018030690537084397,
+      "loss": 3.649,
+      "step": 3356
+    },
+    {
+      "epoch": 4.29312,
+      "grad_norm": 0.5071698427200317,
+      "learning_rate": 0.0001802665230852066,
+      "loss": 3.7119,
+      "step": 3357
+    },
+    {
+      "epoch": 4.2943999999999996,
+      "grad_norm": 0.48967668414115906,
+      "learning_rate": 0.00018022614079956923,
+      "loss": 3.6469,
+      "step": 3358
+    },
+    {
+      "epoch": 4.29568,
+      "grad_norm": 0.502606987953186,
+      "learning_rate": 0.0001801857585139319,
+      "loss": 3.7297,
+      "step": 3359
+    },
+    {
+      "epoch": 4.29696,
+      "grad_norm": 0.5163201689720154,
+      "learning_rate": 0.00018014537622829452,
+      "loss": 3.7605,
+      "step": 3360
+    },
+    {
+      "epoch": 4.29824,
+      "grad_norm": 0.500606894493103,
+      "learning_rate": 0.00018010499394265715,
+      "loss": 3.7258,
+      "step": 3361
+    },
+    {
+      "epoch": 4.29952,
+      "grad_norm": 0.4911852180957794,
+      "learning_rate": 0.00018006461165701975,
+      "loss": 3.7299,
+      "step": 3362
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.494475781917572,
+      "learning_rate": 0.00018002422937138238,
+      "loss": 3.6581,
+      "step": 3363
+    },
+    {
+      "epoch": 4.30208,
+      "grad_norm": 0.49037453532218933,
+      "learning_rate": 0.00017998384708574504,
+      "loss": 3.7097,
+      "step": 3364
+    },
+    {
+      "epoch": 4.30336,
+      "grad_norm": 0.4940754175186157,
+      "learning_rate": 0.00017994346480010767,
+      "loss": 3.7747,
+      "step": 3365
+    },
+    {
+      "epoch": 4.30464,
+      "grad_norm": 0.4911755323410034,
+      "learning_rate": 0.0001799030825144703,
+      "loss": 3.7565,
+      "step": 3366
+    },
+    {
+      "epoch": 4.30592,
+      "grad_norm": 0.473459392786026,
+      "learning_rate": 0.00017986270022883293,
+      "loss": 3.7478,
+      "step": 3367
+    },
+    {
+      "epoch": 4.3072,
+      "grad_norm": 0.4972091317176819,
+      "learning_rate": 0.0001798223179431956,
+      "loss": 3.5687,
+      "step": 3368
+    },
+    {
+      "epoch": 4.30848,
+      "grad_norm": 0.475479394197464,
+      "learning_rate": 0.00017978193565755822,
+      "loss": 3.6893,
+      "step": 3369
+    },
+    {
+      "epoch": 4.30976,
+      "grad_norm": 0.4981231987476349,
+      "learning_rate": 0.00017974155337192082,
+      "loss": 3.6605,
+      "step": 3370
+    },
+    {
+      "epoch": 4.31104,
+      "grad_norm": 0.4935183823108673,
+      "learning_rate": 0.00017970117108628345,
+      "loss": 3.6837,
+      "step": 3371
+    },
+    {
+      "epoch": 4.31232,
+      "grad_norm": 0.491040974855423,
+      "learning_rate": 0.0001796607888006461,
+      "loss": 3.6831,
+      "step": 3372
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.48840728402137756,
+      "learning_rate": 0.00017962040651500874,
+      "loss": 3.7186,
+      "step": 3373
+    },
+    {
+      "epoch": 4.31488,
+      "grad_norm": 0.49155256152153015,
+      "learning_rate": 0.00017958002422937137,
+      "loss": 3.7328,
+      "step": 3374
+    },
+    {
+      "epoch": 4.31616,
+      "grad_norm": 0.4962817132472992,
+      "learning_rate": 0.000179539641943734,
+      "loss": 3.7796,
+      "step": 3375
+    },
+    {
+      "epoch": 4.31744,
+      "grad_norm": 0.47991764545440674,
+      "learning_rate": 0.00017949925965809663,
+      "loss": 3.6687,
+      "step": 3376
+    },
+    {
+      "epoch": 4.31872,
+      "grad_norm": 0.4865191876888275,
+      "learning_rate": 0.0001794588773724593,
+      "loss": 3.6614,
+      "step": 3377
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.4849129617214203,
+      "learning_rate": 0.0001794184950868219,
+      "loss": 3.6637,
+      "step": 3378
+    },
+    {
+      "epoch": 4.32128,
+      "grad_norm": 0.49385684728622437,
+      "learning_rate": 0.00017937811280118452,
+      "loss": 3.6891,
+      "step": 3379
+    },
+    {
+      "epoch": 4.32256,
+      "grad_norm": 0.5029520988464355,
+      "learning_rate": 0.00017933773051554715,
+      "loss": 3.6901,
+      "step": 3380
+    },
+    {
+      "epoch": 4.32384,
+      "grad_norm": 0.4933001399040222,
+      "learning_rate": 0.0001792973482299098,
+      "loss": 3.6471,
+      "step": 3381
+    },
+    {
+      "epoch": 4.32512,
+      "grad_norm": 0.5136417746543884,
+      "learning_rate": 0.00017925696594427244,
+      "loss": 3.6926,
+      "step": 3382
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.5117876529693604,
+      "learning_rate": 0.00017921658365863507,
+      "loss": 3.7587,
+      "step": 3383
+    },
+    {
+      "epoch": 4.32768,
+      "grad_norm": 0.5118081569671631,
+      "learning_rate": 0.0001791762013729977,
+      "loss": 3.7112,
+      "step": 3384
+    },
+    {
+      "epoch": 4.32896,
+      "grad_norm": 0.5017783641815186,
+      "learning_rate": 0.00017913581908736035,
+      "loss": 3.6705,
+      "step": 3385
+    },
+    {
+      "epoch": 4.33024,
+      "grad_norm": 0.5022070407867432,
+      "learning_rate": 0.00017909543680172296,
+      "loss": 3.7245,
+      "step": 3386
+    },
+    {
+      "epoch": 4.33152,
+      "grad_norm": 0.4982999265193939,
+      "learning_rate": 0.0001790550545160856,
+      "loss": 3.6153,
+      "step": 3387
+    },
+    {
+      "epoch": 4.3328,
+      "grad_norm": 0.49681100249290466,
+      "learning_rate": 0.00017901467223044822,
+      "loss": 3.7025,
+      "step": 3388
+    },
+    {
+      "epoch": 4.33408,
+      "grad_norm": 0.5171028971672058,
+      "learning_rate": 0.00017897428994481085,
+      "loss": 3.7245,
+      "step": 3389
+    },
+    {
+      "epoch": 4.33536,
+      "grad_norm": 0.5349324345588684,
+      "learning_rate": 0.0001789339076591735,
+      "loss": 3.6531,
+      "step": 3390
+    },
+    {
+      "epoch": 4.33664,
+      "grad_norm": 0.512718677520752,
+      "learning_rate": 0.00017889352537353614,
+      "loss": 3.709,
+      "step": 3391
+    },
+    {
+      "epoch": 4.33792,
+      "grad_norm": 0.4975050687789917,
+      "learning_rate": 0.00017885314308789877,
+      "loss": 3.7275,
+      "step": 3392
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.49178048968315125,
+      "learning_rate": 0.00017881276080226137,
+      "loss": 3.721,
+      "step": 3393
+    },
+    {
+      "epoch": 4.34048,
+      "grad_norm": 0.5306468605995178,
+      "learning_rate": 0.00017877237851662403,
+      "loss": 3.7299,
+      "step": 3394
+    },
+    {
+      "epoch": 4.34176,
+      "grad_norm": 0.5229880213737488,
+      "learning_rate": 0.00017873199623098666,
+      "loss": 3.694,
+      "step": 3395
+    },
+    {
+      "epoch": 4.34304,
+      "grad_norm": 0.4882638156414032,
+      "learning_rate": 0.00017869161394534929,
+      "loss": 3.6635,
+      "step": 3396
+    },
+    {
+      "epoch": 4.34432,
+      "grad_norm": 0.5192221403121948,
+      "learning_rate": 0.00017865123165971192,
+      "loss": 3.6414,
+      "step": 3397
+    },
+    {
+      "epoch": 4.3456,
+      "grad_norm": 0.5011906027793884,
+      "learning_rate": 0.00017861084937407457,
+      "loss": 3.7079,
+      "step": 3398
+    },
+    {
+      "epoch": 4.34688,
+      "grad_norm": 0.49621590971946716,
+      "learning_rate": 0.0001785704670884372,
+      "loss": 3.7085,
+      "step": 3399
+    },
+    {
+      "epoch": 4.34816,
+      "grad_norm": 0.5203524827957153,
+      "learning_rate": 0.00017853008480279983,
+      "loss": 3.6853,
+      "step": 3400
+    },
+    {
+      "epoch": 4.3494399999999995,
+      "grad_norm": 0.5072999596595764,
+      "learning_rate": 0.00017848970251716244,
+      "loss": 3.6291,
+      "step": 3401
+    },
+    {
+      "epoch": 4.35072,
+      "grad_norm": 0.4849739670753479,
+      "learning_rate": 0.00017844932023152507,
+      "loss": 3.6109,
+      "step": 3402
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.510128915309906,
+      "learning_rate": 0.00017840893794588772,
+      "loss": 3.7104,
+      "step": 3403
+    },
+    {
+      "epoch": 4.35328,
+      "grad_norm": 0.4891580045223236,
+      "learning_rate": 0.00017836855566025035,
+      "loss": 3.6278,
+      "step": 3404
+    },
+    {
+      "epoch": 4.35456,
+      "grad_norm": 0.5085670948028564,
+      "learning_rate": 0.00017832817337461298,
+      "loss": 3.7082,
+      "step": 3405
+    },
+    {
+      "epoch": 4.35584,
+      "grad_norm": 0.49909740686416626,
+      "learning_rate": 0.0001782877910889756,
+      "loss": 3.6073,
+      "step": 3406
+    },
+    {
+      "epoch": 4.35712,
+      "grad_norm": 0.49702468514442444,
+      "learning_rate": 0.00017824740880333827,
+      "loss": 3.6564,
+      "step": 3407
+    },
+    {
+      "epoch": 4.3584,
+      "grad_norm": 0.5047982335090637,
+      "learning_rate": 0.0001782070265177009,
+      "loss": 3.7321,
+      "step": 3408
+    },
+    {
+      "epoch": 4.35968,
+      "grad_norm": 0.4968804717063904,
+      "learning_rate": 0.0001781666442320635,
+      "loss": 3.6783,
+      "step": 3409
+    },
+    {
+      "epoch": 4.36096,
+      "grad_norm": 0.5167127251625061,
+      "learning_rate": 0.00017812626194642613,
+      "loss": 3.6759,
+      "step": 3410
+    },
+    {
+      "epoch": 4.36224,
+      "grad_norm": 0.4954107403755188,
+      "learning_rate": 0.0001780858796607888,
+      "loss": 3.7047,
+      "step": 3411
+    },
+    {
+      "epoch": 4.36352,
+      "grad_norm": 0.519900381565094,
+      "learning_rate": 0.00017804549737515142,
+      "loss": 3.629,
+      "step": 3412
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.5000714659690857,
+      "learning_rate": 0.00017800511508951405,
+      "loss": 3.6723,
+      "step": 3413
+    },
+    {
+      "epoch": 4.36608,
+      "grad_norm": 0.49087822437286377,
+      "learning_rate": 0.00017796473280387668,
+      "loss": 3.7102,
+      "step": 3414
+    },
+    {
+      "epoch": 4.36736,
+      "grad_norm": 0.5202825665473938,
+      "learning_rate": 0.0001779243505182393,
+      "loss": 3.6964,
+      "step": 3415
+    },
+    {
+      "epoch": 4.36864,
+      "grad_norm": 0.5085675716400146,
+      "learning_rate": 0.00017788396823260197,
+      "loss": 3.7475,
+      "step": 3416
+    },
+    {
+      "epoch": 4.3699200000000005,
+      "grad_norm": 0.5006064176559448,
+      "learning_rate": 0.00017784358594696457,
+      "loss": 3.7466,
+      "step": 3417
+    },
+    {
+      "epoch": 4.3712,
+      "grad_norm": 0.4939427375793457,
+      "learning_rate": 0.0001778032036613272,
+      "loss": 3.7078,
+      "step": 3418
+    },
+    {
+      "epoch": 4.37248,
+      "grad_norm": 0.4983762204647064,
+      "learning_rate": 0.00017776282137568983,
+      "loss": 3.728,
+      "step": 3419
+    },
+    {
+      "epoch": 4.37376,
+      "grad_norm": 0.48928341269493103,
+      "learning_rate": 0.0001777224390900525,
+      "loss": 3.6844,
+      "step": 3420
+    },
+    {
+      "epoch": 4.37504,
+      "grad_norm": 0.47871965169906616,
+      "learning_rate": 0.00017768205680441512,
+      "loss": 3.7157,
+      "step": 3421
+    },
+    {
+      "epoch": 4.37632,
+      "grad_norm": 0.5062506794929504,
+      "learning_rate": 0.00017764167451877775,
+      "loss": 3.6605,
+      "step": 3422
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.47811704874038696,
+      "learning_rate": 0.00017760129223314038,
+      "loss": 3.6821,
+      "step": 3423
+    },
+    {
+      "epoch": 4.37888,
+      "grad_norm": 0.510657787322998,
+      "learning_rate": 0.00017756090994750304,
+      "loss": 3.6808,
+      "step": 3424
+    },
+    {
+      "epoch": 4.38016,
+      "grad_norm": 0.4923449754714966,
+      "learning_rate": 0.00017752052766186564,
+      "loss": 3.7051,
+      "step": 3425
+    },
+    {
+      "epoch": 4.38144,
+      "grad_norm": 0.5173118710517883,
+      "learning_rate": 0.00017748014537622827,
+      "loss": 3.7312,
+      "step": 3426
+    },
+    {
+      "epoch": 4.38272,
+      "grad_norm": 0.4835803508758545,
+      "learning_rate": 0.0001774397630905909,
+      "loss": 3.5541,
+      "step": 3427
+    },
+    {
+      "epoch": 4.384,
+      "grad_norm": 0.5001106262207031,
+      "learning_rate": 0.00017739938080495356,
+      "loss": 3.6304,
+      "step": 3428
+    },
+    {
+      "epoch": 4.38528,
+      "grad_norm": 0.4953722059726715,
+      "learning_rate": 0.00017735899851931619,
+      "loss": 3.7689,
+      "step": 3429
+    },
+    {
+      "epoch": 4.38656,
+      "grad_norm": 0.4921940863132477,
+      "learning_rate": 0.00017731861623367882,
+      "loss": 3.7265,
+      "step": 3430
+    },
+    {
+      "epoch": 4.38784,
+      "grad_norm": 0.5121296048164368,
+      "learning_rate": 0.00017727823394804145,
+      "loss": 3.7272,
+      "step": 3431
+    },
+    {
+      "epoch": 4.38912,
+      "grad_norm": 0.48463988304138184,
+      "learning_rate": 0.00017723785166240405,
+      "loss": 3.6316,
+      "step": 3432
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.48769861459732056,
+      "learning_rate": 0.00017719746937676673,
+      "loss": 3.6833,
+      "step": 3433
+    },
+    {
+      "epoch": 4.39168,
+      "grad_norm": 0.5136172771453857,
+      "learning_rate": 0.00017715708709112934,
+      "loss": 3.6876,
+      "step": 3434
+    },
+    {
+      "epoch": 4.39296,
+      "grad_norm": 0.4963926672935486,
+      "learning_rate": 0.00017711670480549197,
+      "loss": 3.7335,
+      "step": 3435
+    },
+    {
+      "epoch": 4.39424,
+      "grad_norm": 0.49306657910346985,
+      "learning_rate": 0.0001770763225198546,
+      "loss": 3.7481,
+      "step": 3436
+    },
+    {
+      "epoch": 4.39552,
+      "grad_norm": 0.5056197643280029,
+      "learning_rate": 0.00017703594023421725,
+      "loss": 3.6287,
+      "step": 3437
+    },
+    {
+      "epoch": 4.3968,
+      "grad_norm": 0.47748926281929016,
+      "learning_rate": 0.00017699555794857988,
+      "loss": 3.6423,
+      "step": 3438
+    },
+    {
+      "epoch": 4.39808,
+      "grad_norm": 0.5189681053161621,
+      "learning_rate": 0.00017695517566294251,
+      "loss": 3.6958,
+      "step": 3439
+    },
+    {
+      "epoch": 4.39936,
+      "grad_norm": 0.48089975118637085,
+      "learning_rate": 0.00017691479337730514,
+      "loss": 3.6822,
+      "step": 3440
+    },
+    {
+      "epoch": 4.40064,
+      "grad_norm": 0.4775058925151825,
+      "learning_rate": 0.0001768744110916678,
+      "loss": 3.7351,
+      "step": 3441
+    },
+    {
+      "epoch": 4.40192,
+      "grad_norm": 0.5030752420425415,
+      "learning_rate": 0.0001768340288060304,
+      "loss": 3.6806,
+      "step": 3442
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.4860950708389282,
+      "learning_rate": 0.00017679364652039303,
+      "loss": 3.6407,
+      "step": 3443
+    },
+    {
+      "epoch": 4.40448,
+      "grad_norm": 0.4938212037086487,
+      "learning_rate": 0.00017675326423475566,
+      "loss": 3.6792,
+      "step": 3444
+    },
+    {
+      "epoch": 4.40576,
+      "grad_norm": 0.5040499567985535,
+      "learning_rate": 0.0001767128819491183,
+      "loss": 3.6686,
+      "step": 3445
+    },
+    {
+      "epoch": 4.40704,
+      "grad_norm": 0.4846273362636566,
+      "learning_rate": 0.00017667249966348095,
+      "loss": 3.6957,
+      "step": 3446
+    },
+    {
+      "epoch": 4.40832,
+      "grad_norm": 0.49627983570098877,
+      "learning_rate": 0.00017663211737784358,
+      "loss": 3.7428,
+      "step": 3447
+    },
+    {
+      "epoch": 4.4096,
+      "grad_norm": 0.4737420380115509,
+      "learning_rate": 0.0001765917350922062,
+      "loss": 3.6932,
+      "step": 3448
+    },
+    {
+      "epoch": 4.41088,
+      "grad_norm": 0.5013241171836853,
+      "learning_rate": 0.00017655135280656881,
+      "loss": 3.7328,
+      "step": 3449
+    },
+    {
+      "epoch": 4.41216,
+      "grad_norm": 0.46999281644821167,
+      "learning_rate": 0.00017651097052093147,
+      "loss": 3.7462,
+      "step": 3450
+    },
+    {
+      "epoch": 4.41344,
+      "grad_norm": 0.5018439888954163,
+      "learning_rate": 0.0001764705882352941,
+      "loss": 3.684,
+      "step": 3451
+    },
+    {
+      "epoch": 4.41472,
+      "grad_norm": 0.49331605434417725,
+      "learning_rate": 0.00017643020594965673,
+      "loss": 3.7176,
+      "step": 3452
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.48275622725486755,
+      "learning_rate": 0.00017638982366401936,
+      "loss": 3.6986,
+      "step": 3453
+    },
+    {
+      "epoch": 4.41728,
+      "grad_norm": 0.48251670598983765,
+      "learning_rate": 0.00017634944137838202,
+      "loss": 3.6331,
+      "step": 3454
+    },
+    {
+      "epoch": 4.41856,
+      "grad_norm": 0.487105131149292,
+      "learning_rate": 0.00017630905909274465,
+      "loss": 3.5988,
+      "step": 3455
+    },
+    {
+      "epoch": 4.41984,
+      "grad_norm": 0.4950862526893616,
+      "learning_rate": 0.00017626867680710728,
+      "loss": 3.6718,
+      "step": 3456
+    },
+    {
+      "epoch": 4.42112,
+      "grad_norm": 0.4823586344718933,
+      "learning_rate": 0.00017622829452146988,
+      "loss": 3.6558,
+      "step": 3457
+    },
+    {
+      "epoch": 4.4224,
+      "grad_norm": 0.49396684765815735,
+      "learning_rate": 0.0001761879122358325,
+      "loss": 3.6592,
+      "step": 3458
+    },
+    {
+      "epoch": 4.42368,
+      "grad_norm": 0.5019631385803223,
+      "learning_rate": 0.00017614752995019517,
+      "loss": 3.683,
+      "step": 3459
+    },
+    {
+      "epoch": 4.4249600000000004,
+      "grad_norm": 0.4832041263580322,
+      "learning_rate": 0.0001761071476645578,
+      "loss": 3.6948,
+      "step": 3460
+    },
+    {
+      "epoch": 4.42624,
+      "grad_norm": 0.5026285648345947,
+      "learning_rate": 0.00017606676537892043,
+      "loss": 3.7162,
+      "step": 3461
+    },
+    {
+      "epoch": 4.42752,
+      "grad_norm": 0.478025883436203,
+      "learning_rate": 0.00017602638309328306,
+      "loss": 3.6677,
+      "step": 3462
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.4910399913787842,
+      "learning_rate": 0.00017598600080764572,
+      "loss": 3.7119,
+      "step": 3463
+    },
+    {
+      "epoch": 4.43008,
+      "grad_norm": 0.4960950016975403,
+      "learning_rate": 0.00017594561852200835,
+      "loss": 3.7209,
+      "step": 3464
+    },
+    {
+      "epoch": 4.43136,
+      "grad_norm": 0.4814871549606323,
+      "learning_rate": 0.00017590523623637095,
+      "loss": 3.6908,
+      "step": 3465
+    },
+    {
+      "epoch": 4.43264,
+      "grad_norm": 0.48105382919311523,
+      "learning_rate": 0.00017586485395073358,
+      "loss": 3.742,
+      "step": 3466
+    },
+    {
+      "epoch": 4.43392,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.00017582447166509624,
+      "loss": 3.6408,
+      "step": 3467
+    },
+    {
+      "epoch": 4.4352,
+      "grad_norm": 0.4880666732788086,
+      "learning_rate": 0.00017578408937945887,
+      "loss": 3.6666,
+      "step": 3468
+    },
+    {
+      "epoch": 4.4364799999999995,
+      "grad_norm": 0.4796549081802368,
+      "learning_rate": 0.0001757437070938215,
+      "loss": 3.6802,
+      "step": 3469
+    },
+    {
+      "epoch": 4.43776,
+      "grad_norm": 0.5090510845184326,
+      "learning_rate": 0.00017570332480818413,
+      "loss": 3.727,
+      "step": 3470
+    },
+    {
+      "epoch": 4.43904,
+      "grad_norm": 0.48752468824386597,
+      "learning_rate": 0.00017566294252254676,
+      "loss": 3.694,
+      "step": 3471
+    },
+    {
+      "epoch": 4.44032,
+      "grad_norm": 0.49315810203552246,
+      "learning_rate": 0.00017562256023690942,
+      "loss": 3.6637,
+      "step": 3472
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.4910982549190521,
+      "learning_rate": 0.00017558217795127202,
+      "loss": 3.697,
+      "step": 3473
+    },
+    {
+      "epoch": 4.44288,
+      "grad_norm": 0.48540931940078735,
+      "learning_rate": 0.00017554179566563465,
+      "loss": 3.6457,
+      "step": 3474
+    },
+    {
+      "epoch": 4.44416,
+      "grad_norm": 0.5137031674385071,
+      "learning_rate": 0.00017550141337999728,
+      "loss": 3.6628,
+      "step": 3475
+    },
+    {
+      "epoch": 4.44544,
+      "grad_norm": 0.4871312379837036,
+      "learning_rate": 0.00017546103109435994,
+      "loss": 3.685,
+      "step": 3476
+    },
+    {
+      "epoch": 4.44672,
+      "grad_norm": 0.48285388946533203,
+      "learning_rate": 0.00017542064880872257,
+      "loss": 3.655,
+      "step": 3477
+    },
+    {
+      "epoch": 4.448,
+      "grad_norm": 0.49935588240623474,
+      "learning_rate": 0.0001753802665230852,
+      "loss": 3.6451,
+      "step": 3478
+    },
+    {
+      "epoch": 4.44928,
+      "grad_norm": 0.49114716053009033,
+      "learning_rate": 0.00017533988423744783,
+      "loss": 3.6701,
+      "step": 3479
+    },
+    {
+      "epoch": 4.45056,
+      "grad_norm": 0.4787088632583618,
+      "learning_rate": 0.00017529950195181048,
+      "loss": 3.6956,
+      "step": 3480
+    },
+    {
+      "epoch": 4.45184,
+      "grad_norm": 0.48162198066711426,
+      "learning_rate": 0.00017525911966617309,
+      "loss": 3.6084,
+      "step": 3481
+    },
+    {
+      "epoch": 4.45312,
+      "grad_norm": 0.49390703439712524,
+      "learning_rate": 0.00017521873738053572,
+      "loss": 3.6884,
+      "step": 3482
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.4952353835105896,
+      "learning_rate": 0.00017517835509489835,
+      "loss": 3.6476,
+      "step": 3483
+    },
+    {
+      "epoch": 4.45568,
+      "grad_norm": 0.48062050342559814,
+      "learning_rate": 0.00017513797280926098,
+      "loss": 3.6257,
+      "step": 3484
+    },
+    {
+      "epoch": 4.45696,
+      "grad_norm": 0.5192123055458069,
+      "learning_rate": 0.00017509759052362363,
+      "loss": 3.6835,
+      "step": 3485
+    },
+    {
+      "epoch": 4.45824,
+      "grad_norm": 0.4845411777496338,
+      "learning_rate": 0.00017505720823798626,
+      "loss": 3.6836,
+      "step": 3486
+    },
+    {
+      "epoch": 4.45952,
+      "grad_norm": 0.47390422224998474,
+      "learning_rate": 0.0001750168259523489,
+      "loss": 3.6174,
+      "step": 3487
+    },
+    {
+      "epoch": 4.4608,
+      "grad_norm": 0.5086140036582947,
+      "learning_rate": 0.0001749764436667115,
+      "loss": 3.7345,
+      "step": 3488
+    },
+    {
+      "epoch": 4.46208,
+      "grad_norm": 0.49887514114379883,
+      "learning_rate": 0.00017493606138107415,
+      "loss": 3.7634,
+      "step": 3489
+    },
+    {
+      "epoch": 4.46336,
+      "grad_norm": 0.5016082525253296,
+      "learning_rate": 0.00017489567909543678,
+      "loss": 3.753,
+      "step": 3490
+    },
+    {
+      "epoch": 4.46464,
+      "grad_norm": 0.5016016960144043,
+      "learning_rate": 0.00017485529680979941,
+      "loss": 3.649,
+      "step": 3491
+    },
+    {
+      "epoch": 4.46592,
+      "grad_norm": 0.5083345174789429,
+      "learning_rate": 0.00017481491452416204,
+      "loss": 3.7257,
+      "step": 3492
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.4881742298603058,
+      "learning_rate": 0.0001747745322385247,
+      "loss": 3.7086,
+      "step": 3493
+    },
+    {
+      "epoch": 4.46848,
+      "grad_norm": 0.5043129920959473,
+      "learning_rate": 0.00017473414995288733,
+      "loss": 3.7117,
+      "step": 3494
+    },
+    {
+      "epoch": 4.46976,
+      "grad_norm": 0.5088573098182678,
+      "learning_rate": 0.00017469376766724996,
+      "loss": 3.6652,
+      "step": 3495
+    },
+    {
+      "epoch": 4.47104,
+      "grad_norm": 0.5027455687522888,
+      "learning_rate": 0.00017465338538161256,
+      "loss": 3.7306,
+      "step": 3496
+    },
+    {
+      "epoch": 4.47232,
+      "grad_norm": 0.49364176392555237,
+      "learning_rate": 0.0001746130030959752,
+      "loss": 3.7809,
+      "step": 3497
+    },
+    {
+      "epoch": 4.4736,
+      "grad_norm": 0.4937020242214203,
+      "learning_rate": 0.00017457262081033785,
+      "loss": 3.7053,
+      "step": 3498
+    },
+    {
+      "epoch": 4.47488,
+      "grad_norm": 0.47408318519592285,
+      "learning_rate": 0.00017453223852470048,
+      "loss": 3.6676,
+      "step": 3499
+    },
+    {
+      "epoch": 4.47616,
+      "grad_norm": 0.4941287338733673,
+      "learning_rate": 0.0001744918562390631,
+      "loss": 3.6673,
+      "step": 3500
+    },
+    {
+      "epoch": 4.47744,
+      "grad_norm": 0.4902247190475464,
+      "learning_rate": 0.00017445147395342574,
+      "loss": 3.6797,
+      "step": 3501
+    },
+    {
+      "epoch": 4.47872,
+      "grad_norm": 0.4781825840473175,
+      "learning_rate": 0.0001744110916677884,
+      "loss": 3.767,
+      "step": 3502
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.5159146189689636,
+      "learning_rate": 0.00017437070938215103,
+      "loss": 3.6403,
+      "step": 3503
+    },
+    {
+      "epoch": 4.48128,
+      "grad_norm": 0.49446627497673035,
+      "learning_rate": 0.00017433032709651363,
+      "loss": 3.6994,
+      "step": 3504
+    },
+    {
+      "epoch": 4.48256,
+      "grad_norm": 0.49100860953330994,
+      "learning_rate": 0.00017428994481087626,
+      "loss": 3.6639,
+      "step": 3505
+    },
+    {
+      "epoch": 4.48384,
+      "grad_norm": 0.495257705450058,
+      "learning_rate": 0.00017424956252523892,
+      "loss": 3.6325,
+      "step": 3506
+    },
+    {
+      "epoch": 4.48512,
+      "grad_norm": 0.4940941333770752,
+      "learning_rate": 0.00017420918023960155,
+      "loss": 3.7,
+      "step": 3507
+    },
+    {
+      "epoch": 4.4864,
+      "grad_norm": 0.4889221489429474,
+      "learning_rate": 0.00017416879795396418,
+      "loss": 3.6322,
+      "step": 3508
+    },
+    {
+      "epoch": 4.48768,
+      "grad_norm": 0.49470090866088867,
+      "learning_rate": 0.0001741284156683268,
+      "loss": 3.6749,
+      "step": 3509
+    },
+    {
+      "epoch": 4.48896,
+      "grad_norm": 0.48783206939697266,
+      "learning_rate": 0.00017408803338268944,
+      "loss": 3.6823,
+      "step": 3510
+    },
+    {
+      "epoch": 4.49024,
+      "grad_norm": 0.47950899600982666,
+      "learning_rate": 0.0001740476510970521,
+      "loss": 3.7092,
+      "step": 3511
+    },
+    {
+      "epoch": 4.49152,
+      "grad_norm": 0.49544283747673035,
+      "learning_rate": 0.0001740072688114147,
+      "loss": 3.6669,
+      "step": 3512
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.5024276375770569,
+      "learning_rate": 0.00017396688652577733,
+      "loss": 3.7114,
+      "step": 3513
+    },
+    {
+      "epoch": 4.49408,
+      "grad_norm": 0.49346235394477844,
+      "learning_rate": 0.00017392650424013996,
+      "loss": 3.684,
+      "step": 3514
+    },
+    {
+      "epoch": 4.49536,
+      "grad_norm": 0.4775454103946686,
+      "learning_rate": 0.00017388612195450262,
+      "loss": 3.6177,
+      "step": 3515
+    },
+    {
+      "epoch": 4.49664,
+      "grad_norm": 0.5153406858444214,
+      "learning_rate": 0.00017384573966886525,
+      "loss": 3.6474,
+      "step": 3516
+    },
+    {
+      "epoch": 4.49792,
+      "grad_norm": 0.5022422075271606,
+      "learning_rate": 0.00017380535738322788,
+      "loss": 3.6711,
+      "step": 3517
+    },
+    {
+      "epoch": 4.4992,
+      "grad_norm": 0.49757120013237,
+      "learning_rate": 0.0001737649750975905,
+      "loss": 3.6771,
+      "step": 3518
+    },
+    {
+      "epoch": 4.50048,
+      "grad_norm": 0.49992772936820984,
+      "learning_rate": 0.00017372459281195316,
+      "loss": 3.5962,
+      "step": 3519
+    },
+    {
+      "epoch": 4.50176,
+      "grad_norm": 0.5268425345420837,
+      "learning_rate": 0.0001736842105263158,
+      "loss": 3.7249,
+      "step": 3520
+    },
+    {
+      "epoch": 4.50304,
+      "grad_norm": 0.4990698993206024,
+      "learning_rate": 0.0001736438282406784,
+      "loss": 3.6477,
+      "step": 3521
+    },
+    {
+      "epoch": 4.50432,
+      "grad_norm": 0.49010327458381653,
+      "learning_rate": 0.00017360344595504103,
+      "loss": 3.6864,
+      "step": 3522
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.5165103673934937,
+      "learning_rate": 0.00017356306366940366,
+      "loss": 3.8092,
+      "step": 3523
+    },
+    {
+      "epoch": 4.50688,
+      "grad_norm": 0.48552533984184265,
+      "learning_rate": 0.00017352268138376631,
+      "loss": 3.6701,
+      "step": 3524
+    },
+    {
+      "epoch": 4.50816,
+      "grad_norm": 0.501692533493042,
+      "learning_rate": 0.00017348229909812894,
+      "loss": 3.6105,
+      "step": 3525
+    },
+    {
+      "epoch": 4.50944,
+      "grad_norm": 0.5025333762168884,
+      "learning_rate": 0.00017344191681249157,
+      "loss": 3.6874,
+      "step": 3526
+    },
+    {
+      "epoch": 4.51072,
+      "grad_norm": 0.4903654456138611,
+      "learning_rate": 0.00017340153452685418,
+      "loss": 3.664,
+      "step": 3527
+    },
+    {
+      "epoch": 4.5120000000000005,
+      "grad_norm": 0.4915325343608856,
+      "learning_rate": 0.00017336115224121686,
+      "loss": 3.7199,
+      "step": 3528
+    },
+    {
+      "epoch": 4.51328,
+      "grad_norm": 0.5007879734039307,
+      "learning_rate": 0.00017332076995557946,
+      "loss": 3.6746,
+      "step": 3529
+    },
+    {
+      "epoch": 4.51456,
+      "grad_norm": 0.48502570390701294,
+      "learning_rate": 0.0001732803876699421,
+      "loss": 3.6661,
+      "step": 3530
+    },
+    {
+      "epoch": 4.51584,
+      "grad_norm": 0.5233677625656128,
+      "learning_rate": 0.00017324000538430473,
+      "loss": 3.6716,
+      "step": 3531
+    },
+    {
+      "epoch": 4.51712,
+      "grad_norm": 0.5040356516838074,
+      "learning_rate": 0.00017319962309866738,
+      "loss": 3.656,
+      "step": 3532
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.5123605132102966,
+      "learning_rate": 0.00017315924081303,
+      "loss": 3.6736,
+      "step": 3533
+    },
+    {
+      "epoch": 4.51968,
+      "grad_norm": 0.48310691118240356,
+      "learning_rate": 0.00017311885852739264,
+      "loss": 3.667,
+      "step": 3534
+    },
+    {
+      "epoch": 4.52096,
+      "grad_norm": 0.48786744475364685,
+      "learning_rate": 0.00017307847624175527,
+      "loss": 3.7278,
+      "step": 3535
+    },
+    {
+      "epoch": 4.52224,
+      "grad_norm": 0.4786374270915985,
+      "learning_rate": 0.00017303809395611788,
+      "loss": 3.6351,
+      "step": 3536
+    },
+    {
+      "epoch": 4.5235199999999995,
+      "grad_norm": 0.49251246452331543,
+      "learning_rate": 0.00017299771167048053,
+      "loss": 3.6494,
+      "step": 3537
+    },
+    {
+      "epoch": 4.5248,
+      "grad_norm": 0.4909958243370056,
+      "learning_rate": 0.00017295732938484316,
+      "loss": 3.6081,
+      "step": 3538
+    },
+    {
+      "epoch": 4.52608,
+      "grad_norm": 0.49989423155784607,
+      "learning_rate": 0.0001729169470992058,
+      "loss": 3.6365,
+      "step": 3539
+    },
+    {
+      "epoch": 4.52736,
+      "grad_norm": 0.5115346312522888,
+      "learning_rate": 0.00017287656481356842,
+      "loss": 3.6791,
+      "step": 3540
+    },
+    {
+      "epoch": 4.52864,
+      "grad_norm": 0.49784228205680847,
+      "learning_rate": 0.00017283618252793108,
+      "loss": 3.6783,
+      "step": 3541
+    },
+    {
+      "epoch": 4.52992,
+      "grad_norm": 0.5055098533630371,
+      "learning_rate": 0.0001727958002422937,
+      "loss": 3.7022,
+      "step": 3542
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.5355752110481262,
+      "learning_rate": 0.00017275541795665634,
+      "loss": 3.7565,
+      "step": 3543
+    },
+    {
+      "epoch": 4.53248,
+      "grad_norm": 0.5136594176292419,
+      "learning_rate": 0.00017271503567101894,
+      "loss": 3.6853,
+      "step": 3544
+    },
+    {
+      "epoch": 4.53376,
+      "grad_norm": 0.502863347530365,
+      "learning_rate": 0.0001726746533853816,
+      "loss": 3.7711,
+      "step": 3545
+    },
+    {
+      "epoch": 4.53504,
+      "grad_norm": 0.5095115303993225,
+      "learning_rate": 0.00017263427109974423,
+      "loss": 3.6232,
+      "step": 3546
+    },
+    {
+      "epoch": 4.53632,
+      "grad_norm": 0.5266593098640442,
+      "learning_rate": 0.00017259388881410686,
+      "loss": 3.7058,
+      "step": 3547
+    },
+    {
+      "epoch": 4.5376,
+      "grad_norm": 0.5122666358947754,
+      "learning_rate": 0.0001725535065284695,
+      "loss": 3.6821,
+      "step": 3548
+    },
+    {
+      "epoch": 4.53888,
+      "grad_norm": 0.5304042100906372,
+      "learning_rate": 0.00017251312424283212,
+      "loss": 3.7056,
+      "step": 3549
+    },
+    {
+      "epoch": 4.54016,
+      "grad_norm": 0.495039165019989,
+      "learning_rate": 0.00017247274195719478,
+      "loss": 3.64,
+      "step": 3550
+    },
+    {
+      "epoch": 4.54144,
+      "grad_norm": 0.5002861618995667,
+      "learning_rate": 0.0001724323596715574,
+      "loss": 3.628,
+      "step": 3551
+    },
+    {
+      "epoch": 4.54272,
+      "grad_norm": 0.47562626004219055,
+      "learning_rate": 0.00017239197738592,
+      "loss": 3.6679,
+      "step": 3552
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.4953966438770294,
+      "learning_rate": 0.00017235159510028264,
+      "loss": 3.6984,
+      "step": 3553
+    },
+    {
+      "epoch": 4.54528,
+      "grad_norm": 0.4846183955669403,
+      "learning_rate": 0.0001723112128146453,
+      "loss": 3.7035,
+      "step": 3554
+    },
+    {
+      "epoch": 4.54656,
+      "grad_norm": 0.4903351068496704,
+      "learning_rate": 0.00017227083052900793,
+      "loss": 3.6948,
+      "step": 3555
+    },
+    {
+      "epoch": 4.54784,
+      "grad_norm": 0.5202940702438354,
+      "learning_rate": 0.00017223044824337056,
+      "loss": 3.7083,
+      "step": 3556
+    },
+    {
+      "epoch": 4.54912,
+      "grad_norm": 0.47886836528778076,
+      "learning_rate": 0.0001721900659577332,
+      "loss": 3.7083,
+      "step": 3557
+    },
+    {
+      "epoch": 4.5504,
+      "grad_norm": 0.47739362716674805,
+      "learning_rate": 0.00017214968367209585,
+      "loss": 3.6878,
+      "step": 3558
+    },
+    {
+      "epoch": 4.55168,
+      "grad_norm": 0.4860458970069885,
+      "learning_rate": 0.00017210930138645848,
+      "loss": 3.6605,
+      "step": 3559
+    },
+    {
+      "epoch": 4.55296,
+      "grad_norm": 0.48289328813552856,
+      "learning_rate": 0.00017206891910082108,
+      "loss": 3.681,
+      "step": 3560
+    },
+    {
+      "epoch": 4.55424,
+      "grad_norm": 0.48003146052360535,
+      "learning_rate": 0.0001720285368151837,
+      "loss": 3.5914,
+      "step": 3561
+    },
+    {
+      "epoch": 4.55552,
+      "grad_norm": 0.4859391152858734,
+      "learning_rate": 0.00017198815452954637,
+      "loss": 3.6338,
+      "step": 3562
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.5082067251205444,
+      "learning_rate": 0.000171947772243909,
+      "loss": 3.6709,
+      "step": 3563
+    },
+    {
+      "epoch": 4.55808,
+      "grad_norm": 0.5058284401893616,
+      "learning_rate": 0.00017190738995827163,
+      "loss": 3.7295,
+      "step": 3564
+    },
+    {
+      "epoch": 4.55936,
+      "grad_norm": 0.4948490560054779,
+      "learning_rate": 0.00017186700767263426,
+      "loss": 3.7362,
+      "step": 3565
+    },
+    {
+      "epoch": 4.56064,
+      "grad_norm": 0.48526522517204285,
+      "learning_rate": 0.00017182662538699689,
+      "loss": 3.6609,
+      "step": 3566
+    },
+    {
+      "epoch": 4.56192,
+      "grad_norm": 0.49817943572998047,
+      "learning_rate": 0.00017178624310135954,
+      "loss": 3.7036,
+      "step": 3567
+    },
+    {
+      "epoch": 4.5632,
+      "grad_norm": 0.4790593981742859,
+      "learning_rate": 0.00017174586081572215,
+      "loss": 3.6634,
+      "step": 3568
+    },
+    {
+      "epoch": 4.56448,
+      "grad_norm": 0.47792062163352966,
+      "learning_rate": 0.00017170547853008478,
+      "loss": 3.688,
+      "step": 3569
+    },
+    {
+      "epoch": 4.56576,
+      "grad_norm": 0.5085933208465576,
+      "learning_rate": 0.0001716650962444474,
+      "loss": 3.6739,
+      "step": 3570
+    },
+    {
+      "epoch": 4.56704,
+      "grad_norm": 0.4828382730484009,
+      "learning_rate": 0.00017162471395881006,
+      "loss": 3.6292,
+      "step": 3571
+    },
+    {
+      "epoch": 4.56832,
+      "grad_norm": 0.47902053594589233,
+      "learning_rate": 0.0001715843316731727,
+      "loss": 3.642,
+      "step": 3572
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.48181986808776855,
+      "learning_rate": 0.00017154394938753532,
+      "loss": 3.7166,
+      "step": 3573
+    },
+    {
+      "epoch": 4.57088,
+      "grad_norm": 0.49542367458343506,
+      "learning_rate": 0.00017150356710189795,
+      "loss": 3.6691,
+      "step": 3574
+    },
+    {
+      "epoch": 4.57216,
+      "grad_norm": 0.48702624440193176,
+      "learning_rate": 0.0001714631848162606,
+      "loss": 3.6509,
+      "step": 3575
+    },
+    {
+      "epoch": 4.57344,
+      "grad_norm": 0.4805475175380707,
+      "learning_rate": 0.00017142280253062321,
+      "loss": 3.7377,
+      "step": 3576
+    },
+    {
+      "epoch": 4.57472,
+      "grad_norm": 0.4737749695777893,
+      "learning_rate": 0.00017138242024498584,
+      "loss": 3.7663,
+      "step": 3577
+    },
+    {
+      "epoch": 4.576,
+      "grad_norm": 0.46877261996269226,
+      "learning_rate": 0.00017134203795934847,
+      "loss": 3.7753,
+      "step": 3578
+    },
+    {
+      "epoch": 4.57728,
+      "grad_norm": 0.4928486943244934,
+      "learning_rate": 0.0001713016556737111,
+      "loss": 3.7,
+      "step": 3579
+    },
+    {
+      "epoch": 4.5785599999999995,
+      "grad_norm": 0.4829467833042145,
+      "learning_rate": 0.00017126127338807376,
+      "loss": 3.6577,
+      "step": 3580
+    },
+    {
+      "epoch": 4.57984,
+      "grad_norm": 0.4743864834308624,
+      "learning_rate": 0.0001712208911024364,
+      "loss": 3.7088,
+      "step": 3581
+    },
+    {
+      "epoch": 4.58112,
+      "grad_norm": 0.4854520261287689,
+      "learning_rate": 0.00017118050881679902,
+      "loss": 3.6397,
+      "step": 3582
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.48954036831855774,
+      "learning_rate": 0.00017114012653116162,
+      "loss": 3.7009,
+      "step": 3583
+    },
+    {
+      "epoch": 4.58368,
+      "grad_norm": 0.4740862250328064,
+      "learning_rate": 0.00017109974424552428,
+      "loss": 3.6539,
+      "step": 3584
+    },
+    {
+      "epoch": 4.58496,
+      "grad_norm": 0.5085862278938293,
+      "learning_rate": 0.0001710593619598869,
+      "loss": 3.7184,
+      "step": 3585
+    },
+    {
+      "epoch": 4.58624,
+      "grad_norm": 0.48374059796333313,
+      "learning_rate": 0.00017101897967424954,
+      "loss": 3.6622,
+      "step": 3586
+    },
+    {
+      "epoch": 4.58752,
+      "grad_norm": 0.49264445900917053,
+      "learning_rate": 0.00017097859738861217,
+      "loss": 3.7557,
+      "step": 3587
+    },
+    {
+      "epoch": 4.5888,
+      "grad_norm": 0.5074787735939026,
+      "learning_rate": 0.00017093821510297483,
+      "loss": 3.7262,
+      "step": 3588
+    },
+    {
+      "epoch": 4.59008,
+      "grad_norm": 0.4799599051475525,
+      "learning_rate": 0.00017089783281733746,
+      "loss": 3.642,
+      "step": 3589
+    },
+    {
+      "epoch": 4.59136,
+      "grad_norm": 0.49617815017700195,
+      "learning_rate": 0.0001708574505317001,
+      "loss": 3.7215,
+      "step": 3590
+    },
+    {
+      "epoch": 4.59264,
+      "grad_norm": 0.48311010003089905,
+      "learning_rate": 0.0001708170682460627,
+      "loss": 3.683,
+      "step": 3591
+    },
+    {
+      "epoch": 4.59392,
+      "grad_norm": 0.48525270819664,
+      "learning_rate": 0.00017077668596042532,
+      "loss": 3.6323,
+      "step": 3592
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.48173174262046814,
+      "learning_rate": 0.00017073630367478798,
+      "loss": 3.7176,
+      "step": 3593
+    },
+    {
+      "epoch": 4.59648,
+      "grad_norm": 0.4832940101623535,
+      "learning_rate": 0.0001706959213891506,
+      "loss": 3.6404,
+      "step": 3594
+    },
+    {
+      "epoch": 4.59776,
+      "grad_norm": 0.48117658495903015,
+      "learning_rate": 0.00017065553910351324,
+      "loss": 3.6617,
+      "step": 3595
+    },
+    {
+      "epoch": 4.5990400000000005,
+      "grad_norm": 0.4833637475967407,
+      "learning_rate": 0.00017061515681787587,
+      "loss": 3.7209,
+      "step": 3596
+    },
+    {
+      "epoch": 4.60032,
+      "grad_norm": 0.48473599553108215,
+      "learning_rate": 0.00017057477453223853,
+      "loss": 3.6349,
+      "step": 3597
+    },
+    {
+      "epoch": 4.6016,
+      "grad_norm": 0.475893497467041,
+      "learning_rate": 0.00017053439224660116,
+      "loss": 3.6932,
+      "step": 3598
+    },
+    {
+      "epoch": 4.60288,
+      "grad_norm": 0.49287232756614685,
+      "learning_rate": 0.00017049400996096376,
+      "loss": 3.6374,
+      "step": 3599
+    },
+    {
+      "epoch": 4.60416,
+      "grad_norm": 0.47559598088264465,
+      "learning_rate": 0.0001704536276753264,
+      "loss": 3.6659,
+      "step": 3600
+    },
+    {
+      "epoch": 4.60544,
+      "grad_norm": 0.4943821430206299,
+      "learning_rate": 0.00017041324538968905,
+      "loss": 3.6761,
+      "step": 3601
+    },
+    {
+      "epoch": 4.60672,
+      "grad_norm": 0.5159769058227539,
+      "learning_rate": 0.00017037286310405168,
+      "loss": 3.6661,
+      "step": 3602
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.5051320791244507,
+      "learning_rate": 0.0001703324808184143,
+      "loss": 3.7369,
+      "step": 3603
+    },
+    {
+      "epoch": 4.60928,
+      "grad_norm": 0.48964956402778625,
+      "learning_rate": 0.00017029209853277694,
+      "loss": 3.6809,
+      "step": 3604
+    },
+    {
+      "epoch": 4.6105599999999995,
+      "grad_norm": 0.4939142167568207,
+      "learning_rate": 0.00017025171624713957,
+      "loss": 3.694,
+      "step": 3605
+    },
+    {
+      "epoch": 4.61184,
+      "grad_norm": 0.5127921104431152,
+      "learning_rate": 0.00017021133396150222,
+      "loss": 3.7422,
+      "step": 3606
+    },
+    {
+      "epoch": 4.61312,
+      "grad_norm": 0.5049188733100891,
+      "learning_rate": 0.00017017095167586485,
+      "loss": 3.6797,
+      "step": 3607
+    },
+    {
+      "epoch": 4.6144,
+      "grad_norm": 0.5061559081077576,
+      "learning_rate": 0.00017013056939022746,
+      "loss": 3.6901,
+      "step": 3608
+    },
+    {
+      "epoch": 4.61568,
+      "grad_norm": 0.48725587129592896,
+      "learning_rate": 0.0001700901871045901,
+      "loss": 3.7239,
+      "step": 3609
+    },
+    {
+      "epoch": 4.61696,
+      "grad_norm": 0.48895296454429626,
+      "learning_rate": 0.00017004980481895274,
+      "loss": 3.73,
+      "step": 3610
+    },
+    {
+      "epoch": 4.61824,
+      "grad_norm": 0.5009008646011353,
+      "learning_rate": 0.00017000942253331538,
+      "loss": 3.636,
+      "step": 3611
+    },
+    {
+      "epoch": 4.61952,
+      "grad_norm": 0.4935823678970337,
+      "learning_rate": 0.000169969040247678,
+      "loss": 3.623,
+      "step": 3612
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.4923403263092041,
+      "learning_rate": 0.00016992865796204064,
+      "loss": 3.7194,
+      "step": 3613
+    },
+    {
+      "epoch": 4.62208,
+      "grad_norm": 0.4789236783981323,
+      "learning_rate": 0.0001698882756764033,
+      "loss": 3.7151,
+      "step": 3614
+    },
+    {
+      "epoch": 4.62336,
+      "grad_norm": 0.5072107911109924,
+      "learning_rate": 0.00016984789339076592,
+      "loss": 3.7177,
+      "step": 3615
+    },
+    {
+      "epoch": 4.62464,
+      "grad_norm": 0.4821150302886963,
+      "learning_rate": 0.00016980751110512853,
+      "loss": 3.6819,
+      "step": 3616
+    },
+    {
+      "epoch": 4.62592,
+      "grad_norm": 0.48691630363464355,
+      "learning_rate": 0.00016976712881949116,
+      "loss": 3.6939,
+      "step": 3617
+    },
+    {
+      "epoch": 4.6272,
+      "grad_norm": 0.4875396490097046,
+      "learning_rate": 0.00016972674653385379,
+      "loss": 3.7381,
+      "step": 3618
+    },
+    {
+      "epoch": 4.62848,
+      "grad_norm": 0.4814916253089905,
+      "learning_rate": 0.00016968636424821644,
+      "loss": 3.6668,
+      "step": 3619
+    },
+    {
+      "epoch": 4.62976,
+      "grad_norm": 0.5000630021095276,
+      "learning_rate": 0.00016964598196257907,
+      "loss": 3.7047,
+      "step": 3620
+    },
+    {
+      "epoch": 4.6310400000000005,
+      "grad_norm": 0.4792519211769104,
+      "learning_rate": 0.0001696055996769417,
+      "loss": 3.6889,
+      "step": 3621
+    },
+    {
+      "epoch": 4.63232,
+      "grad_norm": 0.4922233819961548,
+      "learning_rate": 0.00016956521739130433,
+      "loss": 3.6608,
+      "step": 3622
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.49414339661598206,
+      "learning_rate": 0.000169524835105667,
+      "loss": 3.6623,
+      "step": 3623
+    },
+    {
+      "epoch": 4.63488,
+      "grad_norm": 0.48598942160606384,
+      "learning_rate": 0.0001694844528200296,
+      "loss": 3.6867,
+      "step": 3624
+    },
+    {
+      "epoch": 4.63616,
+      "grad_norm": 0.49352526664733887,
+      "learning_rate": 0.00016944407053439222,
+      "loss": 3.6493,
+      "step": 3625
+    },
+    {
+      "epoch": 4.63744,
+      "grad_norm": 0.5058410167694092,
+      "learning_rate": 0.00016940368824875485,
+      "loss": 3.6692,
+      "step": 3626
+    },
+    {
+      "epoch": 4.63872,
+      "grad_norm": 0.4912284016609192,
+      "learning_rate": 0.0001693633059631175,
+      "loss": 3.7055,
+      "step": 3627
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.4703737199306488,
+      "learning_rate": 0.00016932292367748014,
+      "loss": 3.6885,
+      "step": 3628
+    },
+    {
+      "epoch": 4.64128,
+      "grad_norm": 0.4926270842552185,
+      "learning_rate": 0.00016928254139184277,
+      "loss": 3.6547,
+      "step": 3629
+    },
+    {
+      "epoch": 4.64256,
+      "grad_norm": 0.48712384700775146,
+      "learning_rate": 0.0001692421591062054,
+      "loss": 3.7905,
+      "step": 3630
+    },
+    {
+      "epoch": 4.64384,
+      "grad_norm": 0.5001680850982666,
+      "learning_rate": 0.000169201776820568,
+      "loss": 3.6893,
+      "step": 3631
+    },
+    {
+      "epoch": 4.64512,
+      "grad_norm": 0.4809771180152893,
+      "learning_rate": 0.00016916139453493066,
+      "loss": 3.738,
+      "step": 3632
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.4865325093269348,
+      "learning_rate": 0.0001691210122492933,
+      "loss": 3.6982,
+      "step": 3633
+    },
+    {
+      "epoch": 4.64768,
+      "grad_norm": 0.48700863122940063,
+      "learning_rate": 0.00016908062996365592,
+      "loss": 3.681,
+      "step": 3634
+    },
+    {
+      "epoch": 4.64896,
+      "grad_norm": 0.4818150997161865,
+      "learning_rate": 0.00016904024767801855,
+      "loss": 3.6595,
+      "step": 3635
+    },
+    {
+      "epoch": 4.65024,
+      "grad_norm": 0.4853065609931946,
+      "learning_rate": 0.0001689998653923812,
+      "loss": 3.6553,
+      "step": 3636
+    },
+    {
+      "epoch": 4.65152,
+      "grad_norm": 0.47990745306015015,
+      "learning_rate": 0.00016895948310674384,
+      "loss": 3.7326,
+      "step": 3637
+    },
+    {
+      "epoch": 4.6528,
+      "grad_norm": 0.4638812839984894,
+      "learning_rate": 0.00016891910082110647,
+      "loss": 3.6394,
+      "step": 3638
+    },
+    {
+      "epoch": 4.65408,
+      "grad_norm": 0.48503032326698303,
+      "learning_rate": 0.00016887871853546907,
+      "loss": 3.6976,
+      "step": 3639
+    },
+    {
+      "epoch": 4.65536,
+      "grad_norm": 0.48162126541137695,
+      "learning_rate": 0.00016883833624983173,
+      "loss": 3.6568,
+      "step": 3640
+    },
+    {
+      "epoch": 4.65664,
+      "grad_norm": 0.5152369737625122,
+      "learning_rate": 0.00016879795396419436,
+      "loss": 3.6476,
+      "step": 3641
+    },
+    {
+      "epoch": 4.65792,
+      "grad_norm": 0.4773014783859253,
+      "learning_rate": 0.000168757571678557,
+      "loss": 3.6812,
+      "step": 3642
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.5280107855796814,
+      "learning_rate": 0.00016871718939291962,
+      "loss": 3.6671,
+      "step": 3643
+    },
+    {
+      "epoch": 4.66048,
+      "grad_norm": 0.49400994181632996,
+      "learning_rate": 0.00016867680710728225,
+      "loss": 3.6199,
+      "step": 3644
+    },
+    {
+      "epoch": 4.66176,
+      "grad_norm": 0.5342155694961548,
+      "learning_rate": 0.0001686364248216449,
+      "loss": 3.6443,
+      "step": 3645
+    },
+    {
+      "epoch": 4.66304,
+      "grad_norm": 0.500808835029602,
+      "learning_rate": 0.00016859604253600754,
+      "loss": 3.6912,
+      "step": 3646
+    },
+    {
+      "epoch": 4.66432,
+      "grad_norm": 0.511306643486023,
+      "learning_rate": 0.00016855566025037014,
+      "loss": 3.6898,
+      "step": 3647
+    },
+    {
+      "epoch": 4.6655999999999995,
+      "grad_norm": 0.48556506633758545,
+      "learning_rate": 0.00016851527796473277,
+      "loss": 3.6869,
+      "step": 3648
+    },
+    {
+      "epoch": 4.66688,
+      "grad_norm": 0.48522791266441345,
+      "learning_rate": 0.00016847489567909543,
+      "loss": 3.7425,
+      "step": 3649
+    },
+    {
+      "epoch": 4.66816,
+      "grad_norm": 0.4987458288669586,
+      "learning_rate": 0.00016843451339345806,
+      "loss": 3.6531,
+      "step": 3650
+    },
+    {
+      "epoch": 4.66944,
+      "grad_norm": 0.5028944611549377,
+      "learning_rate": 0.0001683941311078207,
+      "loss": 3.6606,
+      "step": 3651
+    },
+    {
+      "epoch": 4.67072,
+      "grad_norm": 0.4878036081790924,
+      "learning_rate": 0.00016835374882218332,
+      "loss": 3.6368,
+      "step": 3652
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.5253325700759888,
+      "learning_rate": 0.00016831336653654597,
+      "loss": 3.7309,
+      "step": 3653
+    },
+    {
+      "epoch": 4.67328,
+      "grad_norm": 0.48149093985557556,
+      "learning_rate": 0.0001682729842509086,
+      "loss": 3.7106,
+      "step": 3654
+    },
+    {
+      "epoch": 4.67456,
+      "grad_norm": 0.5140679478645325,
+      "learning_rate": 0.0001682326019652712,
+      "loss": 3.721,
+      "step": 3655
+    },
+    {
+      "epoch": 4.67584,
+      "grad_norm": 0.49887320399284363,
+      "learning_rate": 0.00016819221967963384,
+      "loss": 3.6681,
+      "step": 3656
+    },
+    {
+      "epoch": 4.67712,
+      "grad_norm": 0.47831347584724426,
+      "learning_rate": 0.00016815183739399647,
+      "loss": 3.6551,
+      "step": 3657
+    },
+    {
+      "epoch": 4.6784,
+      "grad_norm": 0.5047725439071655,
+      "learning_rate": 0.00016811145510835912,
+      "loss": 3.6415,
+      "step": 3658
+    },
+    {
+      "epoch": 4.67968,
+      "grad_norm": 0.4860306978225708,
+      "learning_rate": 0.00016807107282272175,
+      "loss": 3.6685,
+      "step": 3659
+    },
+    {
+      "epoch": 4.68096,
+      "grad_norm": 0.4931819438934326,
+      "learning_rate": 0.00016803069053708438,
+      "loss": 3.6904,
+      "step": 3660
+    },
+    {
+      "epoch": 4.68224,
+      "grad_norm": 0.5166349411010742,
+      "learning_rate": 0.00016799030825144701,
+      "loss": 3.727,
+      "step": 3661
+    },
+    {
+      "epoch": 4.68352,
+      "grad_norm": 0.5045530796051025,
+      "learning_rate": 0.00016794992596580967,
+      "loss": 3.691,
+      "step": 3662
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.49202415347099304,
+      "learning_rate": 0.00016790954368017227,
+      "loss": 3.6231,
+      "step": 3663
+    },
+    {
+      "epoch": 4.6860800000000005,
+      "grad_norm": 0.4810471832752228,
+      "learning_rate": 0.0001678691613945349,
+      "loss": 3.6767,
+      "step": 3664
+    },
+    {
+      "epoch": 4.68736,
+      "grad_norm": 0.514689028263092,
+      "learning_rate": 0.00016782877910889753,
+      "loss": 3.6977,
+      "step": 3665
+    },
+    {
+      "epoch": 4.68864,
+      "grad_norm": 0.4772043526172638,
+      "learning_rate": 0.0001677883968232602,
+      "loss": 3.6731,
+      "step": 3666
+    },
+    {
+      "epoch": 4.68992,
+      "grad_norm": 0.5057224035263062,
+      "learning_rate": 0.00016774801453762282,
+      "loss": 3.7433,
+      "step": 3667
+    },
+    {
+      "epoch": 4.6912,
+      "grad_norm": 0.4916319251060486,
+      "learning_rate": 0.00016770763225198545,
+      "loss": 3.6379,
+      "step": 3668
+    },
+    {
+      "epoch": 4.69248,
+      "grad_norm": 0.5023067593574524,
+      "learning_rate": 0.00016766724996634808,
+      "loss": 3.7392,
+      "step": 3669
+    },
+    {
+      "epoch": 4.69376,
+      "grad_norm": 0.49110403656959534,
+      "learning_rate": 0.00016762686768071069,
+      "loss": 3.6819,
+      "step": 3670
+    },
+    {
+      "epoch": 4.69504,
+      "grad_norm": 0.47794216871261597,
+      "learning_rate": 0.00016758648539507334,
+      "loss": 3.6164,
+      "step": 3671
+    },
+    {
+      "epoch": 4.69632,
+      "grad_norm": 0.5044650435447693,
+      "learning_rate": 0.00016754610310943597,
+      "loss": 3.6418,
+      "step": 3672
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.5018360614776611,
+      "learning_rate": 0.0001675057208237986,
+      "loss": 3.6579,
+      "step": 3673
+    },
+    {
+      "epoch": 4.69888,
+      "grad_norm": 0.513712465763092,
+      "learning_rate": 0.00016746533853816123,
+      "loss": 3.7271,
+      "step": 3674
+    },
+    {
+      "epoch": 4.70016,
+      "grad_norm": 0.5256522297859192,
+      "learning_rate": 0.0001674249562525239,
+      "loss": 3.6982,
+      "step": 3675
+    },
+    {
+      "epoch": 4.70144,
+      "grad_norm": 0.5181000232696533,
+      "learning_rate": 0.00016738457396688652,
+      "loss": 3.6359,
+      "step": 3676
+    },
+    {
+      "epoch": 4.70272,
+      "grad_norm": 0.5020421147346497,
+      "learning_rate": 0.00016734419168124915,
+      "loss": 3.7344,
+      "step": 3677
+    },
+    {
+      "epoch": 4.704,
+      "grad_norm": 0.5058776140213013,
+      "learning_rate": 0.00016730380939561175,
+      "loss": 3.6141,
+      "step": 3678
+    },
+    {
+      "epoch": 4.70528,
+      "grad_norm": 0.5099439024925232,
+      "learning_rate": 0.00016726342710997444,
+      "loss": 3.6955,
+      "step": 3679
+    },
+    {
+      "epoch": 4.70656,
+      "grad_norm": 0.5178961753845215,
+      "learning_rate": 0.00016722304482433704,
+      "loss": 3.6717,
+      "step": 3680
+    },
+    {
+      "epoch": 4.70784,
+      "grad_norm": 0.5064387917518616,
+      "learning_rate": 0.00016718266253869967,
+      "loss": 3.6438,
+      "step": 3681
+    },
+    {
+      "epoch": 4.70912,
+      "grad_norm": 0.48872536420822144,
+      "learning_rate": 0.0001671422802530623,
+      "loss": 3.6893,
+      "step": 3682
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.486896276473999,
+      "learning_rate": 0.00016710189796742496,
+      "loss": 3.6217,
+      "step": 3683
+    },
+    {
+      "epoch": 4.71168,
+      "grad_norm": 0.49064919352531433,
+      "learning_rate": 0.0001670615156817876,
+      "loss": 3.693,
+      "step": 3684
+    },
+    {
+      "epoch": 4.71296,
+      "grad_norm": 0.5264923572540283,
+      "learning_rate": 0.00016702113339615022,
+      "loss": 3.7152,
+      "step": 3685
+    },
+    {
+      "epoch": 4.71424,
+      "grad_norm": 0.5014359354972839,
+      "learning_rate": 0.00016698075111051282,
+      "loss": 3.7828,
+      "step": 3686
+    },
+    {
+      "epoch": 4.71552,
+      "grad_norm": 0.4952867329120636,
+      "learning_rate": 0.00016694036882487545,
+      "loss": 3.7426,
+      "step": 3687
+    },
+    {
+      "epoch": 4.7168,
+      "grad_norm": 0.49848371744155884,
+      "learning_rate": 0.0001668999865392381,
+      "loss": 3.6818,
+      "step": 3688
+    },
+    {
+      "epoch": 4.7180800000000005,
+      "grad_norm": 0.4835038185119629,
+      "learning_rate": 0.00016685960425360074,
+      "loss": 3.6319,
+      "step": 3689
+    },
+    {
+      "epoch": 4.71936,
+      "grad_norm": 0.5079094171524048,
+      "learning_rate": 0.00016681922196796337,
+      "loss": 3.6938,
+      "step": 3690
+    },
+    {
+      "epoch": 4.7206399999999995,
+      "grad_norm": 0.5054168701171875,
+      "learning_rate": 0.000166778839682326,
+      "loss": 3.7248,
+      "step": 3691
+    },
+    {
+      "epoch": 4.72192,
+      "grad_norm": 0.4924614727497101,
+      "learning_rate": 0.00016673845739668866,
+      "loss": 3.7232,
+      "step": 3692
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.5226231813430786,
+      "learning_rate": 0.00016669807511105129,
+      "loss": 3.6702,
+      "step": 3693
+    },
+    {
+      "epoch": 4.72448,
+      "grad_norm": 0.5076504945755005,
+      "learning_rate": 0.00016665769282541392,
+      "loss": 3.7035,
+      "step": 3694
+    },
+    {
+      "epoch": 4.72576,
+      "grad_norm": 0.5075111985206604,
+      "learning_rate": 0.00016661731053977652,
+      "loss": 3.6875,
+      "step": 3695
+    },
+    {
+      "epoch": 4.72704,
+      "grad_norm": 0.5135008096694946,
+      "learning_rate": 0.00016657692825413918,
+      "loss": 3.6545,
+      "step": 3696
+    },
+    {
+      "epoch": 4.72832,
+      "grad_norm": 0.5121733546257019,
+      "learning_rate": 0.0001665365459685018,
+      "loss": 3.6973,
+      "step": 3697
+    },
+    {
+      "epoch": 4.7296,
+      "grad_norm": 0.4708007574081421,
+      "learning_rate": 0.00016649616368286444,
+      "loss": 3.6359,
+      "step": 3698
+    },
+    {
+      "epoch": 4.73088,
+      "grad_norm": 0.5306479334831238,
+      "learning_rate": 0.00016645578139722707,
+      "loss": 3.7159,
+      "step": 3699
+    },
+    {
+      "epoch": 4.73216,
+      "grad_norm": 0.4930853545665741,
+      "learning_rate": 0.0001664153991115897,
+      "loss": 3.6615,
+      "step": 3700
+    },
+    {
+      "epoch": 4.73344,
+      "grad_norm": 0.5061611533164978,
+      "learning_rate": 0.00016637501682595235,
+      "loss": 3.7783,
+      "step": 3701
+    },
+    {
+      "epoch": 4.73472,
+      "grad_norm": 0.4978872239589691,
+      "learning_rate": 0.00016633463454031498,
+      "loss": 3.6809,
+      "step": 3702
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.5093433856964111,
+      "learning_rate": 0.00016629425225467759,
+      "loss": 3.6233,
+      "step": 3703
+    },
+    {
+      "epoch": 4.73728,
+      "grad_norm": 0.4769956171512604,
+      "learning_rate": 0.00016625386996904022,
+      "loss": 3.5826,
+      "step": 3704
+    },
+    {
+      "epoch": 4.73856,
+      "grad_norm": 0.4818449020385742,
+      "learning_rate": 0.00016621348768340287,
+      "loss": 3.6416,
+      "step": 3705
+    },
+    {
+      "epoch": 4.73984,
+      "grad_norm": 0.4861428439617157,
+      "learning_rate": 0.0001661731053977655,
+      "loss": 3.6334,
+      "step": 3706
+    },
+    {
+      "epoch": 4.7411200000000004,
+      "grad_norm": 0.48706743121147156,
+      "learning_rate": 0.00016613272311212813,
+      "loss": 3.6829,
+      "step": 3707
+    },
+    {
+      "epoch": 4.7424,
+      "grad_norm": 0.48839133977890015,
+      "learning_rate": 0.00016609234082649076,
+      "loss": 3.6995,
+      "step": 3708
+    },
+    {
+      "epoch": 4.74368,
+      "grad_norm": 0.49105438590049744,
+      "learning_rate": 0.00016605195854085342,
+      "loss": 3.6665,
+      "step": 3709
+    },
+    {
+      "epoch": 4.74496,
+      "grad_norm": 0.503555417060852,
+      "learning_rate": 0.00016601157625521605,
+      "loss": 3.7246,
+      "step": 3710
+    },
+    {
+      "epoch": 4.74624,
+      "grad_norm": 0.4822165071964264,
+      "learning_rate": 0.00016597119396957865,
+      "loss": 3.6801,
+      "step": 3711
+    },
+    {
+      "epoch": 4.74752,
+      "grad_norm": 0.4817500114440918,
+      "learning_rate": 0.00016593081168394128,
+      "loss": 3.575,
+      "step": 3712
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.48095250129699707,
+      "learning_rate": 0.00016589042939830391,
+      "loss": 3.6716,
+      "step": 3713
+    },
+    {
+      "epoch": 4.75008,
+      "grad_norm": 0.47289326786994934,
+      "learning_rate": 0.00016585004711266657,
+      "loss": 3.7431,
+      "step": 3714
+    },
+    {
+      "epoch": 4.75136,
+      "grad_norm": 0.46789786219596863,
+      "learning_rate": 0.0001658096648270292,
+      "loss": 3.6603,
+      "step": 3715
+    },
+    {
+      "epoch": 4.7526399999999995,
+      "grad_norm": 0.47495347261428833,
+      "learning_rate": 0.00016576928254139183,
+      "loss": 3.6394,
+      "step": 3716
+    },
+    {
+      "epoch": 4.75392,
+      "grad_norm": 0.47441309690475464,
+      "learning_rate": 0.00016572890025575446,
+      "loss": 3.7065,
+      "step": 3717
+    },
+    {
+      "epoch": 4.7552,
+      "grad_norm": 0.48996591567993164,
+      "learning_rate": 0.00016568851797011712,
+      "loss": 3.7313,
+      "step": 3718
+    },
+    {
+      "epoch": 4.75648,
+      "grad_norm": 0.47536635398864746,
+      "learning_rate": 0.00016564813568447972,
+      "loss": 3.7061,
+      "step": 3719
+    },
+    {
+      "epoch": 4.75776,
+      "grad_norm": 0.508357048034668,
+      "learning_rate": 0.00016560775339884235,
+      "loss": 3.7152,
+      "step": 3720
+    },
+    {
+      "epoch": 4.75904,
+      "grad_norm": 0.4819643795490265,
+      "learning_rate": 0.00016556737111320498,
+      "loss": 3.6843,
+      "step": 3721
+    },
+    {
+      "epoch": 4.76032,
+      "grad_norm": 0.48836493492126465,
+      "learning_rate": 0.00016552698882756764,
+      "loss": 3.6814,
+      "step": 3722
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.4822399616241455,
+      "learning_rate": 0.00016548660654193027,
+      "loss": 3.6709,
+      "step": 3723
+    },
+    {
+      "epoch": 4.76288,
+      "grad_norm": 0.4872921407222748,
+      "learning_rate": 0.0001654462242562929,
+      "loss": 3.6587,
+      "step": 3724
+    },
+    {
+      "epoch": 4.76416,
+      "grad_norm": 0.486806184053421,
+      "learning_rate": 0.00016540584197065553,
+      "loss": 3.669,
+      "step": 3725
+    },
+    {
+      "epoch": 4.76544,
+      "grad_norm": 0.49304312467575073,
+      "learning_rate": 0.00016536545968501813,
+      "loss": 3.7521,
+      "step": 3726
+    },
+    {
+      "epoch": 4.76672,
+      "grad_norm": 0.4882862865924835,
+      "learning_rate": 0.0001653250773993808,
+      "loss": 3.7018,
+      "step": 3727
+    },
+    {
+      "epoch": 4.768,
+      "grad_norm": 0.4865057170391083,
+      "learning_rate": 0.00016528469511374342,
+      "loss": 3.6966,
+      "step": 3728
+    },
+    {
+      "epoch": 4.76928,
+      "grad_norm": 0.5051146745681763,
+      "learning_rate": 0.00016524431282810605,
+      "loss": 3.7802,
+      "step": 3729
+    },
+    {
+      "epoch": 4.77056,
+      "grad_norm": 0.5136163830757141,
+      "learning_rate": 0.00016520393054246868,
+      "loss": 3.6855,
+      "step": 3730
+    },
+    {
+      "epoch": 4.77184,
+      "grad_norm": 0.4789270758628845,
+      "learning_rate": 0.00016516354825683134,
+      "loss": 3.6787,
+      "step": 3731
+    },
+    {
+      "epoch": 4.7731200000000005,
+      "grad_norm": 0.5074111223220825,
+      "learning_rate": 0.00016512316597119397,
+      "loss": 3.6951,
+      "step": 3732
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.5275261998176575,
+      "learning_rate": 0.0001650827836855566,
+      "loss": 3.6443,
+      "step": 3733
+    },
+    {
+      "epoch": 4.77568,
+      "grad_norm": 0.49672234058380127,
+      "learning_rate": 0.0001650424013999192,
+      "loss": 3.6661,
+      "step": 3734
+    },
+    {
+      "epoch": 4.77696,
+      "grad_norm": 0.4923790693283081,
+      "learning_rate": 0.00016500201911428186,
+      "loss": 3.6796,
+      "step": 3735
+    },
+    {
+      "epoch": 4.77824,
+      "grad_norm": 0.4953721761703491,
+      "learning_rate": 0.0001649616368286445,
+      "loss": 3.6743,
+      "step": 3736
+    },
+    {
+      "epoch": 4.77952,
+      "grad_norm": 0.4764186441898346,
+      "learning_rate": 0.00016492125454300712,
+      "loss": 3.6672,
+      "step": 3737
+    },
+    {
+      "epoch": 4.7808,
+      "grad_norm": 0.5113012790679932,
+      "learning_rate": 0.00016488087225736975,
+      "loss": 3.629,
+      "step": 3738
+    },
+    {
+      "epoch": 4.78208,
+      "grad_norm": 0.48934149742126465,
+      "learning_rate": 0.00016484048997173238,
+      "loss": 3.7223,
+      "step": 3739
+    },
+    {
+      "epoch": 4.78336,
+      "grad_norm": 0.5140074491500854,
+      "learning_rate": 0.00016480010768609503,
+      "loss": 3.7186,
+      "step": 3740
+    },
+    {
+      "epoch": 4.78464,
+      "grad_norm": 0.4866582751274109,
+      "learning_rate": 0.00016475972540045766,
+      "loss": 3.62,
+      "step": 3741
+    },
+    {
+      "epoch": 4.78592,
+      "grad_norm": 0.48126623034477234,
+      "learning_rate": 0.00016471934311482027,
+      "loss": 3.608,
+      "step": 3742
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.48048514127731323,
+      "learning_rate": 0.0001646789608291829,
+      "loss": 3.6883,
+      "step": 3743
+    },
+    {
+      "epoch": 4.78848,
+      "grad_norm": 0.4765174984931946,
+      "learning_rate": 0.00016463857854354555,
+      "loss": 3.7108,
+      "step": 3744
+    },
+    {
+      "epoch": 4.78976,
+      "grad_norm": 0.4911894202232361,
+      "learning_rate": 0.00016459819625790818,
+      "loss": 3.6355,
+      "step": 3745
+    },
+    {
+      "epoch": 4.79104,
+      "grad_norm": 0.5015087723731995,
+      "learning_rate": 0.00016455781397227081,
+      "loss": 3.6922,
+      "step": 3746
+    },
+    {
+      "epoch": 4.79232,
+      "grad_norm": 0.4827320873737335,
+      "learning_rate": 0.00016451743168663344,
+      "loss": 3.6264,
+      "step": 3747
+    },
+    {
+      "epoch": 4.7936,
+      "grad_norm": 0.4948306083679199,
+      "learning_rate": 0.0001644770494009961,
+      "loss": 3.7218,
+      "step": 3748
+    },
+    {
+      "epoch": 4.79488,
+      "grad_norm": 0.48731568455696106,
+      "learning_rate": 0.00016443666711535873,
+      "loss": 3.6451,
+      "step": 3749
+    },
+    {
+      "epoch": 4.79616,
+      "grad_norm": 0.4918150305747986,
+      "learning_rate": 0.00016439628482972133,
+      "loss": 3.7111,
+      "step": 3750
+    },
+    {
+      "epoch": 4.79744,
+      "grad_norm": 0.4810066521167755,
+      "learning_rate": 0.00016435590254408397,
+      "loss": 3.6458,
+      "step": 3751
+    },
+    {
+      "epoch": 4.79872,
+      "grad_norm": 0.49135395884513855,
+      "learning_rate": 0.0001643155202584466,
+      "loss": 3.6688,
+      "step": 3752
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.4967387020587921,
+      "learning_rate": 0.00016427513797280925,
+      "loss": 3.6786,
+      "step": 3753
+    },
+    {
+      "epoch": 4.80128,
+      "grad_norm": 0.48914146423339844,
+      "learning_rate": 0.00016423475568717188,
+      "loss": 3.7037,
+      "step": 3754
+    },
+    {
+      "epoch": 4.80256,
+      "grad_norm": 0.5048929452896118,
+      "learning_rate": 0.0001641943734015345,
+      "loss": 3.6746,
+      "step": 3755
+    },
+    {
+      "epoch": 4.80384,
+      "grad_norm": 0.4980829954147339,
+      "learning_rate": 0.00016415399111589714,
+      "loss": 3.6226,
+      "step": 3756
+    },
+    {
+      "epoch": 4.80512,
+      "grad_norm": 0.49464380741119385,
+      "learning_rate": 0.0001641136088302598,
+      "loss": 3.6642,
+      "step": 3757
+    },
+    {
+      "epoch": 4.8064,
+      "grad_norm": 0.4930364489555359,
+      "learning_rate": 0.0001640732265446224,
+      "loss": 3.7118,
+      "step": 3758
+    },
+    {
+      "epoch": 4.8076799999999995,
+      "grad_norm": 0.5031747221946716,
+      "learning_rate": 0.00016403284425898503,
+      "loss": 3.6796,
+      "step": 3759
+    },
+    {
+      "epoch": 4.80896,
+      "grad_norm": 0.4856695830821991,
+      "learning_rate": 0.00016399246197334766,
+      "loss": 3.6577,
+      "step": 3760
+    },
+    {
+      "epoch": 4.81024,
+      "grad_norm": 0.5000268816947937,
+      "learning_rate": 0.00016395207968771032,
+      "loss": 3.638,
+      "step": 3761
+    },
+    {
+      "epoch": 4.81152,
+      "grad_norm": 0.5013075470924377,
+      "learning_rate": 0.00016391169740207295,
+      "loss": 3.7025,
+      "step": 3762
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.49570322036743164,
+      "learning_rate": 0.00016387131511643558,
+      "loss": 3.6357,
+      "step": 3763
+    },
+    {
+      "epoch": 4.81408,
+      "grad_norm": 0.4977611005306244,
+      "learning_rate": 0.0001638309328307982,
+      "loss": 3.671,
+      "step": 3764
+    },
+    {
+      "epoch": 4.81536,
+      "grad_norm": 0.5036880970001221,
+      "learning_rate": 0.0001637905505451608,
+      "loss": 3.6263,
+      "step": 3765
+    },
+    {
+      "epoch": 4.81664,
+      "grad_norm": 0.4825877249240875,
+      "learning_rate": 0.0001637501682595235,
+      "loss": 3.6794,
+      "step": 3766
+    },
+    {
+      "epoch": 4.81792,
+      "grad_norm": 0.4834206998348236,
+      "learning_rate": 0.0001637097859738861,
+      "loss": 3.6166,
+      "step": 3767
+    },
+    {
+      "epoch": 4.8192,
+      "grad_norm": 0.5228777527809143,
+      "learning_rate": 0.00016366940368824873,
+      "loss": 3.7228,
+      "step": 3768
+    },
+    {
+      "epoch": 4.82048,
+      "grad_norm": 0.4963701367378235,
+      "learning_rate": 0.00016362902140261136,
+      "loss": 3.6723,
+      "step": 3769
+    },
+    {
+      "epoch": 4.82176,
+      "grad_norm": 0.5132946968078613,
+      "learning_rate": 0.00016358863911697402,
+      "loss": 3.7093,
+      "step": 3770
+    },
+    {
+      "epoch": 4.82304,
+      "grad_norm": 0.5134023427963257,
+      "learning_rate": 0.00016354825683133665,
+      "loss": 3.6616,
+      "step": 3771
+    },
+    {
+      "epoch": 4.82432,
+      "grad_norm": 0.4977473318576813,
+      "learning_rate": 0.00016350787454569928,
+      "loss": 3.6783,
+      "step": 3772
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.49358707666397095,
+      "learning_rate": 0.00016346749226006188,
+      "loss": 3.6623,
+      "step": 3773
+    },
+    {
+      "epoch": 4.82688,
+      "grad_norm": 0.5281035304069519,
+      "learning_rate": 0.00016342710997442457,
+      "loss": 3.7313,
+      "step": 3774
+    },
+    {
+      "epoch": 4.8281600000000005,
+      "grad_norm": 0.4807729125022888,
+      "learning_rate": 0.00016338672768878717,
+      "loss": 3.6286,
+      "step": 3775
+    },
+    {
+      "epoch": 4.82944,
+      "grad_norm": 0.4929489195346832,
+      "learning_rate": 0.0001633463454031498,
+      "loss": 3.7261,
+      "step": 3776
+    },
+    {
+      "epoch": 4.83072,
+      "grad_norm": 0.493528813123703,
+      "learning_rate": 0.00016330596311751243,
+      "loss": 3.6753,
+      "step": 3777
+    },
+    {
+      "epoch": 4.832,
+      "grad_norm": 0.49640294909477234,
+      "learning_rate": 0.00016326558083187506,
+      "loss": 3.6256,
+      "step": 3778
+    },
+    {
+      "epoch": 4.83328,
+      "grad_norm": 0.48569512367248535,
+      "learning_rate": 0.00016322519854623772,
+      "loss": 3.6442,
+      "step": 3779
+    },
+    {
+      "epoch": 4.83456,
+      "grad_norm": 0.498219758272171,
+      "learning_rate": 0.00016318481626060035,
+      "loss": 3.6997,
+      "step": 3780
+    },
+    {
+      "epoch": 4.83584,
+      "grad_norm": 0.5126128792762756,
+      "learning_rate": 0.00016314443397496298,
+      "loss": 3.7796,
+      "step": 3781
+    },
+    {
+      "epoch": 4.83712,
+      "grad_norm": 0.4836713671684265,
+      "learning_rate": 0.00016310405168932558,
+      "loss": 3.6532,
+      "step": 3782
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.5176852345466614,
+      "learning_rate": 0.00016306366940368824,
+      "loss": 3.6887,
+      "step": 3783
+    },
+    {
+      "epoch": 4.8396799999999995,
+      "grad_norm": 0.4919340908527374,
+      "learning_rate": 0.00016302328711805087,
+      "loss": 3.7374,
+      "step": 3784
+    },
+    {
+      "epoch": 4.84096,
+      "grad_norm": 0.5173082947731018,
+      "learning_rate": 0.0001629829048324135,
+      "loss": 3.6584,
+      "step": 3785
+    },
+    {
+      "epoch": 4.84224,
+      "grad_norm": 0.4877385199069977,
+      "learning_rate": 0.00016294252254677613,
+      "loss": 3.7104,
+      "step": 3786
+    },
+    {
+      "epoch": 4.84352,
+      "grad_norm": 0.5282792448997498,
+      "learning_rate": 0.00016290214026113878,
+      "loss": 3.7029,
+      "step": 3787
+    },
+    {
+      "epoch": 4.8448,
+      "grad_norm": 0.4718897044658661,
+      "learning_rate": 0.0001628617579755014,
+      "loss": 3.5776,
+      "step": 3788
+    },
+    {
+      "epoch": 4.84608,
+      "grad_norm": 0.4864949584007263,
+      "learning_rate": 0.00016282137568986404,
+      "loss": 3.6709,
+      "step": 3789
+    },
+    {
+      "epoch": 4.84736,
+      "grad_norm": 0.4912480115890503,
+      "learning_rate": 0.00016278099340422665,
+      "loss": 3.6211,
+      "step": 3790
+    },
+    {
+      "epoch": 4.84864,
+      "grad_norm": 0.5093181729316711,
+      "learning_rate": 0.00016274061111858928,
+      "loss": 3.8126,
+      "step": 3791
+    },
+    {
+      "epoch": 4.84992,
+      "grad_norm": 0.4893922209739685,
+      "learning_rate": 0.00016270022883295193,
+      "loss": 3.6797,
+      "step": 3792
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.4943704903125763,
+      "learning_rate": 0.00016265984654731456,
+      "loss": 3.6392,
+      "step": 3793
+    },
+    {
+      "epoch": 4.85248,
+      "grad_norm": 0.5246706604957581,
+      "learning_rate": 0.0001626194642616772,
+      "loss": 3.685,
+      "step": 3794
+    },
+    {
+      "epoch": 4.85376,
+      "grad_norm": 0.4894337058067322,
+      "learning_rate": 0.00016257908197603982,
+      "loss": 3.6819,
+      "step": 3795
+    },
+    {
+      "epoch": 4.85504,
+      "grad_norm": 0.49564680457115173,
+      "learning_rate": 0.00016253869969040248,
+      "loss": 3.6284,
+      "step": 3796
+    },
+    {
+      "epoch": 4.85632,
+      "grad_norm": 0.49830925464630127,
+      "learning_rate": 0.0001624983174047651,
+      "loss": 3.6549,
+      "step": 3797
+    },
+    {
+      "epoch": 4.8576,
+      "grad_norm": 0.47998228669166565,
+      "learning_rate": 0.00016245793511912771,
+      "loss": 3.6778,
+      "step": 3798
+    },
+    {
+      "epoch": 4.85888,
+      "grad_norm": 0.4880746006965637,
+      "learning_rate": 0.00016241755283349034,
+      "loss": 3.6366,
+      "step": 3799
+    },
+    {
+      "epoch": 4.8601600000000005,
+      "grad_norm": 0.4946172535419464,
+      "learning_rate": 0.000162377170547853,
+      "loss": 3.7328,
+      "step": 3800
+    },
+    {
+      "epoch": 4.86144,
+      "grad_norm": 0.48107582330703735,
+      "learning_rate": 0.00016233678826221563,
+      "loss": 3.6619,
+      "step": 3801
+    },
+    {
+      "epoch": 4.86272,
+      "grad_norm": 0.47235268354415894,
+      "learning_rate": 0.00016229640597657826,
+      "loss": 3.6913,
+      "step": 3802
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.4809187948703766,
+      "learning_rate": 0.0001622560236909409,
+      "loss": 3.7307,
+      "step": 3803
+    },
+    {
+      "epoch": 4.86528,
+      "grad_norm": 0.4874497950077057,
+      "learning_rate": 0.00016221564140530355,
+      "loss": 3.7355,
+      "step": 3804
+    },
+    {
+      "epoch": 4.86656,
+      "grad_norm": 0.49071377515792847,
+      "learning_rate": 0.00016217525911966618,
+      "loss": 3.579,
+      "step": 3805
+    },
+    {
+      "epoch": 4.86784,
+      "grad_norm": 0.4813006818294525,
+      "learning_rate": 0.00016213487683402878,
+      "loss": 3.6804,
+      "step": 3806
+    },
+    {
+      "epoch": 4.86912,
+      "grad_norm": 0.4884811341762543,
+      "learning_rate": 0.0001620944945483914,
+      "loss": 3.7033,
+      "step": 3807
+    },
+    {
+      "epoch": 4.8704,
+      "grad_norm": 0.5217158198356628,
+      "learning_rate": 0.00016205411226275404,
+      "loss": 3.6843,
+      "step": 3808
+    },
+    {
+      "epoch": 4.87168,
+      "grad_norm": 0.48379671573638916,
+      "learning_rate": 0.0001620137299771167,
+      "loss": 3.7468,
+      "step": 3809
+    },
+    {
+      "epoch": 4.87296,
+      "grad_norm": 0.484479159116745,
+      "learning_rate": 0.00016197334769147933,
+      "loss": 3.7231,
+      "step": 3810
+    },
+    {
+      "epoch": 4.87424,
+      "grad_norm": 0.5106998085975647,
+      "learning_rate": 0.00016193296540584196,
+      "loss": 3.6457,
+      "step": 3811
+    },
+    {
+      "epoch": 4.87552,
+      "grad_norm": 0.4962330162525177,
+      "learning_rate": 0.0001618925831202046,
+      "loss": 3.6255,
+      "step": 3812
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.5050908923149109,
+      "learning_rate": 0.00016185220083456725,
+      "loss": 3.7097,
+      "step": 3813
+    },
+    {
+      "epoch": 4.87808,
+      "grad_norm": 0.495358407497406,
+      "learning_rate": 0.00016181181854892985,
+      "loss": 3.7255,
+      "step": 3814
+    },
+    {
+      "epoch": 4.87936,
+      "grad_norm": 0.495797336101532,
+      "learning_rate": 0.00016177143626329248,
+      "loss": 3.6743,
+      "step": 3815
+    },
+    {
+      "epoch": 4.88064,
+      "grad_norm": 0.49159398674964905,
+      "learning_rate": 0.0001617310539776551,
+      "loss": 3.6507,
+      "step": 3816
+    },
+    {
+      "epoch": 4.88192,
+      "grad_norm": 0.4823581874370575,
+      "learning_rate": 0.00016169067169201777,
+      "loss": 3.7253,
+      "step": 3817
+    },
+    {
+      "epoch": 4.8832,
+      "grad_norm": 0.49354973435401917,
+      "learning_rate": 0.0001616502894063804,
+      "loss": 3.6406,
+      "step": 3818
+    },
+    {
+      "epoch": 4.88448,
+      "grad_norm": 0.4922613799571991,
+      "learning_rate": 0.00016160990712074303,
+      "loss": 3.6345,
+      "step": 3819
+    },
+    {
+      "epoch": 4.88576,
+      "grad_norm": 0.5070391893386841,
+      "learning_rate": 0.00016156952483510566,
+      "loss": 3.6943,
+      "step": 3820
+    },
+    {
+      "epoch": 4.88704,
+      "grad_norm": 0.4859599173069,
+      "learning_rate": 0.00016152914254946826,
+      "loss": 3.6488,
+      "step": 3821
+    },
+    {
+      "epoch": 4.88832,
+      "grad_norm": 0.4835364520549774,
+      "learning_rate": 0.00016148876026383092,
+      "loss": 3.6938,
+      "step": 3822
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.4754778742790222,
+      "learning_rate": 0.00016144837797819355,
+      "loss": 3.6802,
+      "step": 3823
+    },
+    {
+      "epoch": 4.89088,
+      "grad_norm": 0.508832573890686,
+      "learning_rate": 0.00016140799569255618,
+      "loss": 3.6323,
+      "step": 3824
+    },
+    {
+      "epoch": 4.89216,
+      "grad_norm": 0.4894411861896515,
+      "learning_rate": 0.0001613676134069188,
+      "loss": 3.7234,
+      "step": 3825
+    },
+    {
+      "epoch": 4.89344,
+      "grad_norm": 0.5034093260765076,
+      "learning_rate": 0.00016132723112128146,
+      "loss": 3.6997,
+      "step": 3826
+    },
+    {
+      "epoch": 4.8947199999999995,
+      "grad_norm": 0.4902549684047699,
+      "learning_rate": 0.0001612868488356441,
+      "loss": 3.6539,
+      "step": 3827
+    },
+    {
+      "epoch": 4.896,
+      "grad_norm": 0.49553415179252625,
+      "learning_rate": 0.00016124646655000672,
+      "loss": 3.6314,
+      "step": 3828
+    },
+    {
+      "epoch": 4.89728,
+      "grad_norm": 0.5032349824905396,
+      "learning_rate": 0.00016120608426436933,
+      "loss": 3.6966,
+      "step": 3829
+    },
+    {
+      "epoch": 4.89856,
+      "grad_norm": 0.5162962079048157,
+      "learning_rate": 0.00016116570197873198,
+      "loss": 3.7436,
+      "step": 3830
+    },
+    {
+      "epoch": 4.89984,
+      "grad_norm": 0.49726447463035583,
+      "learning_rate": 0.00016112531969309462,
+      "loss": 3.6775,
+      "step": 3831
+    },
+    {
+      "epoch": 4.90112,
+      "grad_norm": 0.47404661774635315,
+      "learning_rate": 0.00016108493740745725,
+      "loss": 3.7073,
+      "step": 3832
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.48400676250457764,
+      "learning_rate": 0.00016104455512181988,
+      "loss": 3.6737,
+      "step": 3833
+    },
+    {
+      "epoch": 4.90368,
+      "grad_norm": 0.48407965898513794,
+      "learning_rate": 0.0001610041728361825,
+      "loss": 3.6579,
+      "step": 3834
+    },
+    {
+      "epoch": 4.90496,
+      "grad_norm": 0.47430333495140076,
+      "learning_rate": 0.00016096379055054516,
+      "loss": 3.6103,
+      "step": 3835
+    },
+    {
+      "epoch": 4.90624,
+      "grad_norm": 0.49005916714668274,
+      "learning_rate": 0.0001609234082649078,
+      "loss": 3.5885,
+      "step": 3836
+    },
+    {
+      "epoch": 4.90752,
+      "grad_norm": 0.472774863243103,
+      "learning_rate": 0.0001608830259792704,
+      "loss": 3.7037,
+      "step": 3837
+    },
+    {
+      "epoch": 4.9088,
+      "grad_norm": 0.48804715275764465,
+      "learning_rate": 0.00016084264369363303,
+      "loss": 3.7332,
+      "step": 3838
+    },
+    {
+      "epoch": 4.91008,
+      "grad_norm": 0.4966636002063751,
+      "learning_rate": 0.00016080226140799568,
+      "loss": 3.6174,
+      "step": 3839
+    },
+    {
+      "epoch": 4.91136,
+      "grad_norm": 0.4942854642868042,
+      "learning_rate": 0.0001607618791223583,
+      "loss": 3.6865,
+      "step": 3840
+    },
+    {
+      "epoch": 4.91264,
+      "grad_norm": 0.48685768246650696,
+      "learning_rate": 0.00016072149683672094,
+      "loss": 3.6556,
+      "step": 3841
+    },
+    {
+      "epoch": 4.91392,
+      "grad_norm": 0.5361490845680237,
+      "learning_rate": 0.00016068111455108357,
+      "loss": 3.7219,
+      "step": 3842
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.47551262378692627,
+      "learning_rate": 0.00016064073226544623,
+      "loss": 3.7257,
+      "step": 3843
+    },
+    {
+      "epoch": 4.91648,
+      "grad_norm": 0.48748570680618286,
+      "learning_rate": 0.00016060034997980886,
+      "loss": 3.7161,
+      "step": 3844
+    },
+    {
+      "epoch": 4.91776,
+      "grad_norm": 0.48473209142684937,
+      "learning_rate": 0.00016055996769417146,
+      "loss": 3.6119,
+      "step": 3845
+    },
+    {
+      "epoch": 4.91904,
+      "grad_norm": 0.4792821407318115,
+      "learning_rate": 0.0001605195854085341,
+      "loss": 3.628,
+      "step": 3846
+    },
+    {
+      "epoch": 4.92032,
+      "grad_norm": 0.5115758180618286,
+      "learning_rate": 0.00016047920312289672,
+      "loss": 3.7057,
+      "step": 3847
+    },
+    {
+      "epoch": 4.9216,
+      "grad_norm": 0.4800952970981598,
+      "learning_rate": 0.00016043882083725938,
+      "loss": 3.6729,
+      "step": 3848
+    },
+    {
+      "epoch": 4.92288,
+      "grad_norm": 0.48038557171821594,
+      "learning_rate": 0.000160398438551622,
+      "loss": 3.6507,
+      "step": 3849
+    },
+    {
+      "epoch": 4.92416,
+      "grad_norm": 0.49345579743385315,
+      "learning_rate": 0.00016035805626598464,
+      "loss": 3.6492,
+      "step": 3850
+    },
+    {
+      "epoch": 4.92544,
+      "grad_norm": 0.4855850338935852,
+      "learning_rate": 0.00016031767398034727,
+      "loss": 3.7056,
+      "step": 3851
+    },
+    {
+      "epoch": 4.9267199999999995,
+      "grad_norm": 0.47242411971092224,
+      "learning_rate": 0.00016027729169470993,
+      "loss": 3.7455,
+      "step": 3852
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.4987708330154419,
+      "learning_rate": 0.00016023690940907256,
+      "loss": 3.7361,
+      "step": 3853
+    },
+    {
+      "epoch": 4.92928,
+      "grad_norm": 0.4940935969352722,
+      "learning_rate": 0.00016019652712343516,
+      "loss": 3.6286,
+      "step": 3854
+    },
+    {
+      "epoch": 4.93056,
+      "grad_norm": 0.4955693483352661,
+      "learning_rate": 0.0001601561448377978,
+      "loss": 3.6897,
+      "step": 3855
+    },
+    {
+      "epoch": 4.93184,
+      "grad_norm": 0.4994462728500366,
+      "learning_rate": 0.00016011576255216045,
+      "loss": 3.6699,
+      "step": 3856
+    },
+    {
+      "epoch": 4.93312,
+      "grad_norm": 0.5072110295295715,
+      "learning_rate": 0.00016007538026652308,
+      "loss": 3.7264,
+      "step": 3857
+    },
+    {
+      "epoch": 4.9344,
+      "grad_norm": 0.49533572793006897,
+      "learning_rate": 0.0001600349979808857,
+      "loss": 3.6975,
+      "step": 3858
+    },
+    {
+      "epoch": 4.93568,
+      "grad_norm": 0.512359619140625,
+      "learning_rate": 0.00015999461569524834,
+      "loss": 3.7374,
+      "step": 3859
+    },
+    {
+      "epoch": 4.93696,
+      "grad_norm": 0.4889034628868103,
+      "learning_rate": 0.00015995423340961094,
+      "loss": 3.6486,
+      "step": 3860
+    },
+    {
+      "epoch": 4.93824,
+      "grad_norm": 0.4879732131958008,
+      "learning_rate": 0.00015991385112397363,
+      "loss": 3.6722,
+      "step": 3861
+    },
+    {
+      "epoch": 4.93952,
+      "grad_norm": 0.5002577900886536,
+      "learning_rate": 0.00015987346883833623,
+      "loss": 3.7448,
+      "step": 3862
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.5065283179283142,
+      "learning_rate": 0.00015983308655269886,
+      "loss": 3.7383,
+      "step": 3863
+    },
+    {
+      "epoch": 4.94208,
+      "grad_norm": 0.48911115527153015,
+      "learning_rate": 0.0001597927042670615,
+      "loss": 3.6939,
+      "step": 3864
+    },
+    {
+      "epoch": 4.94336,
+      "grad_norm": 0.500379741191864,
+      "learning_rate": 0.00015975232198142415,
+      "loss": 3.6542,
+      "step": 3865
+    },
+    {
+      "epoch": 4.94464,
+      "grad_norm": 0.47929057478904724,
+      "learning_rate": 0.00015971193969578678,
+      "loss": 3.7623,
+      "step": 3866
+    },
+    {
+      "epoch": 4.94592,
+      "grad_norm": 0.4697692394256592,
+      "learning_rate": 0.0001596715574101494,
+      "loss": 3.6306,
+      "step": 3867
+    },
+    {
+      "epoch": 4.9472000000000005,
+      "grad_norm": 0.4873841106891632,
+      "learning_rate": 0.00015963117512451204,
+      "loss": 3.5985,
+      "step": 3868
+    },
+    {
+      "epoch": 4.94848,
+      "grad_norm": 0.49631455540657043,
+      "learning_rate": 0.0001595907928388747,
+      "loss": 3.7473,
+      "step": 3869
+    },
+    {
+      "epoch": 4.94976,
+      "grad_norm": 0.48060086369514465,
+      "learning_rate": 0.0001595504105532373,
+      "loss": 3.667,
+      "step": 3870
+    },
+    {
+      "epoch": 4.95104,
+      "grad_norm": 0.4914751350879669,
+      "learning_rate": 0.00015951002826759993,
+      "loss": 3.6433,
+      "step": 3871
+    },
+    {
+      "epoch": 4.95232,
+      "grad_norm": 0.5151807069778442,
+      "learning_rate": 0.00015946964598196256,
+      "loss": 3.6871,
+      "step": 3872
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.48747456073760986,
+      "learning_rate": 0.0001594292636963252,
+      "loss": 3.6983,
+      "step": 3873
+    },
+    {
+      "epoch": 4.95488,
+      "grad_norm": 0.49215272068977356,
+      "learning_rate": 0.00015938888141068784,
+      "loss": 3.7184,
+      "step": 3874
+    },
+    {
+      "epoch": 4.95616,
+      "grad_norm": 0.47915881872177124,
+      "learning_rate": 0.00015934849912505047,
+      "loss": 3.6691,
+      "step": 3875
+    },
+    {
+      "epoch": 4.95744,
+      "grad_norm": 0.4974246323108673,
+      "learning_rate": 0.0001593081168394131,
+      "loss": 3.6795,
+      "step": 3876
+    },
+    {
+      "epoch": 4.95872,
+      "grad_norm": 0.47061559557914734,
+      "learning_rate": 0.0001592677345537757,
+      "loss": 3.6472,
+      "step": 3877
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.4836273193359375,
+      "learning_rate": 0.00015922735226813836,
+      "loss": 3.6728,
+      "step": 3878
+    },
+    {
+      "epoch": 4.96128,
+      "grad_norm": 0.5134618878364563,
+      "learning_rate": 0.000159186969982501,
+      "loss": 3.6932,
+      "step": 3879
+    },
+    {
+      "epoch": 4.96256,
+      "grad_norm": 0.49473533034324646,
+      "learning_rate": 0.00015914658769686362,
+      "loss": 3.6023,
+      "step": 3880
+    },
+    {
+      "epoch": 4.96384,
+      "grad_norm": 0.5070651173591614,
+      "learning_rate": 0.00015910620541122625,
+      "loss": 3.6855,
+      "step": 3881
+    },
+    {
+      "epoch": 4.96512,
+      "grad_norm": 0.4944706857204437,
+      "learning_rate": 0.0001590658231255889,
+      "loss": 3.6878,
+      "step": 3882
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.4829122722148895,
+      "learning_rate": 0.00015902544083995154,
+      "loss": 3.6283,
+      "step": 3883
+    },
+    {
+      "epoch": 4.96768,
+      "grad_norm": 0.47128134965896606,
+      "learning_rate": 0.00015898505855431417,
+      "loss": 3.7395,
+      "step": 3884
+    },
+    {
+      "epoch": 4.96896,
+      "grad_norm": 0.5024461150169373,
+      "learning_rate": 0.00015894467626867677,
+      "loss": 3.7182,
+      "step": 3885
+    },
+    {
+      "epoch": 4.97024,
+      "grad_norm": 0.4943237602710724,
+      "learning_rate": 0.0001589042939830394,
+      "loss": 3.705,
+      "step": 3886
+    },
+    {
+      "epoch": 4.97152,
+      "grad_norm": 0.49064311385154724,
+      "learning_rate": 0.00015886391169740206,
+      "loss": 3.6011,
+      "step": 3887
+    },
+    {
+      "epoch": 4.9728,
+      "grad_norm": 0.4886166751384735,
+      "learning_rate": 0.0001588235294117647,
+      "loss": 3.6503,
+      "step": 3888
+    },
+    {
+      "epoch": 4.97408,
+      "grad_norm": 0.4722157120704651,
+      "learning_rate": 0.00015878314712612732,
+      "loss": 3.5743,
+      "step": 3889
+    },
+    {
+      "epoch": 4.97536,
+      "grad_norm": 0.49199846386909485,
+      "learning_rate": 0.00015874276484048995,
+      "loss": 3.6846,
+      "step": 3890
+    },
+    {
+      "epoch": 4.97664,
+      "grad_norm": 0.504956066608429,
+      "learning_rate": 0.0001587023825548526,
+      "loss": 3.6009,
+      "step": 3891
+    },
+    {
+      "epoch": 4.97792,
+      "grad_norm": 0.49585065245628357,
+      "learning_rate": 0.00015866200026921524,
+      "loss": 3.6558,
+      "step": 3892
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.5018090605735779,
+      "learning_rate": 0.00015862161798357784,
+      "loss": 3.6642,
+      "step": 3893
+    },
+    {
+      "epoch": 4.98048,
+      "grad_norm": 0.49588721990585327,
+      "learning_rate": 0.00015858123569794047,
+      "loss": 3.6914,
+      "step": 3894
+    },
+    {
+      "epoch": 4.9817599999999995,
+      "grad_norm": 0.4850863516330719,
+      "learning_rate": 0.00015854085341230313,
+      "loss": 3.6784,
+      "step": 3895
+    },
+    {
+      "epoch": 4.98304,
+      "grad_norm": 0.4720815122127533,
+      "learning_rate": 0.00015850047112666576,
+      "loss": 3.6412,
+      "step": 3896
+    },
+    {
+      "epoch": 4.98432,
+      "grad_norm": 0.49307361245155334,
+      "learning_rate": 0.0001584600888410284,
+      "loss": 3.6723,
+      "step": 3897
+    },
+    {
+      "epoch": 4.9856,
+      "grad_norm": 0.48994728922843933,
+      "learning_rate": 0.00015841970655539102,
+      "loss": 3.6367,
+      "step": 3898
+    },
+    {
+      "epoch": 4.98688,
+      "grad_norm": 0.47764644026756287,
+      "learning_rate": 0.00015837932426975365,
+      "loss": 3.6513,
+      "step": 3899
+    },
+    {
+      "epoch": 4.98816,
+      "grad_norm": 0.4945572316646576,
+      "learning_rate": 0.0001583389419841163,
+      "loss": 3.6575,
+      "step": 3900
+    },
+    {
+      "epoch": 4.98944,
+      "grad_norm": 0.4753789007663727,
+      "learning_rate": 0.0001582985596984789,
+      "loss": 3.63,
+      "step": 3901
+    },
+    {
+      "epoch": 4.99072,
+      "grad_norm": 0.49885818362236023,
+      "learning_rate": 0.00015825817741284154,
+      "loss": 3.7375,
+      "step": 3902
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.4667688310146332,
+      "learning_rate": 0.00015821779512720417,
+      "loss": 3.636,
+      "step": 3903
+    },
+    {
+      "epoch": 4.99328,
+      "grad_norm": 0.46380728483200073,
+      "learning_rate": 0.00015817741284156683,
+      "loss": 3.5965,
+      "step": 3904
+    },
+    {
+      "epoch": 4.99456,
+      "grad_norm": 0.5012388229370117,
+      "learning_rate": 0.00015813703055592946,
+      "loss": 3.7338,
+      "step": 3905
+    },
+    {
+      "epoch": 4.99584,
+      "grad_norm": 0.5032443404197693,
+      "learning_rate": 0.0001580966482702921,
+      "loss": 3.6502,
+      "step": 3906
+    },
+    {
+      "epoch": 4.99712,
+      "grad_norm": 0.46951258182525635,
+      "learning_rate": 0.00015805626598465472,
+      "loss": 3.6379,
+      "step": 3907
+    },
+    {
+      "epoch": 4.9984,
+      "grad_norm": 0.502934992313385,
+      "learning_rate": 0.00015801588369901737,
+      "loss": 3.6745,
+      "step": 3908
+    },
+    {
+      "epoch": 4.99968,
+      "grad_norm": 0.46453621983528137,
+      "learning_rate": 0.00015797550141337998,
+      "loss": 3.6497,
+      "step": 3909
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.9125704169273376,
+      "learning_rate": 0.0001579351191277426,
+      "loss": 3.7091,
+      "step": 3910
+    },
+    {
+      "epoch": 5.00128,
+      "grad_norm": 0.5202215909957886,
+      "learning_rate": 0.00015789473684210524,
+      "loss": 3.5637,
+      "step": 3911
+    },
+    {
+      "epoch": 5.00256,
+      "grad_norm": 0.5078494548797607,
+      "learning_rate": 0.00015785435455646787,
+      "loss": 3.5149,
+      "step": 3912
+    },
+    {
+      "epoch": 5.00384,
+      "grad_norm": 0.463044673204422,
+      "learning_rate": 0.00015781397227083053,
+      "loss": 3.5784,
+      "step": 3913
+    },
+    {
+      "epoch": 5.00512,
+      "grad_norm": 0.5001516342163086,
+      "learning_rate": 0.00015777358998519316,
+      "loss": 3.5504,
+      "step": 3914
+    },
+    {
+      "epoch": 5.0064,
+      "grad_norm": 0.4898684620857239,
+      "learning_rate": 0.00015773320769955579,
+      "loss": 3.5146,
+      "step": 3915
+    },
+    {
+      "epoch": 5.00768,
+      "grad_norm": 0.5046645998954773,
+      "learning_rate": 0.0001576928254139184,
+      "loss": 3.5652,
+      "step": 3916
+    },
+    {
+      "epoch": 5.00896,
+      "grad_norm": 0.5051593780517578,
+      "learning_rate": 0.00015765244312828105,
+      "loss": 3.4985,
+      "step": 3917
+    },
+    {
+      "epoch": 5.01024,
+      "grad_norm": 0.49468353390693665,
+      "learning_rate": 0.00015761206084264368,
+      "loss": 3.5268,
+      "step": 3918
+    },
+    {
+      "epoch": 5.01152,
+      "grad_norm": 0.48806026577949524,
+      "learning_rate": 0.0001575716785570063,
+      "loss": 3.5752,
+      "step": 3919
+    },
+    {
+      "epoch": 5.0128,
+      "grad_norm": 0.48567843437194824,
+      "learning_rate": 0.00015753129627136894,
+      "loss": 3.5006,
+      "step": 3920
+    },
+    {
+      "epoch": 5.01408,
+      "grad_norm": 0.4762195348739624,
+      "learning_rate": 0.0001574909139857316,
+      "loss": 3.523,
+      "step": 3921
+    },
+    {
+      "epoch": 5.01536,
+      "grad_norm": 0.4921858012676239,
+      "learning_rate": 0.00015745053170009422,
+      "loss": 3.501,
+      "step": 3922
+    },
+    {
+      "epoch": 5.01664,
+      "grad_norm": 0.49300310015678406,
+      "learning_rate": 0.00015741014941445685,
+      "loss": 3.4951,
+      "step": 3923
+    },
+    {
+      "epoch": 5.01792,
+      "grad_norm": 0.5041300058364868,
+      "learning_rate": 0.00015736976712881946,
+      "loss": 3.5221,
+      "step": 3924
+    },
+    {
+      "epoch": 5.0192,
+      "grad_norm": 0.4857078790664673,
+      "learning_rate": 0.00015732938484318214,
+      "loss": 3.5616,
+      "step": 3925
+    },
+    {
+      "epoch": 5.02048,
+      "grad_norm": 0.49047553539276123,
+      "learning_rate": 0.00015728900255754474,
+      "loss": 3.4968,
+      "step": 3926
+    },
+    {
+      "epoch": 5.0217600000000004,
+      "grad_norm": 0.510486900806427,
+      "learning_rate": 0.00015724862027190737,
+      "loss": 3.5057,
+      "step": 3927
+    },
+    {
+      "epoch": 5.02304,
+      "grad_norm": 0.5110107064247131,
+      "learning_rate": 0.00015720823798627,
+      "loss": 3.5816,
+      "step": 3928
+    },
+    {
+      "epoch": 5.02432,
+      "grad_norm": 0.4852614104747772,
+      "learning_rate": 0.00015716785570063263,
+      "loss": 3.5548,
+      "step": 3929
+    },
+    {
+      "epoch": 5.0256,
+      "grad_norm": 0.4855445921421051,
+      "learning_rate": 0.0001571274734149953,
+      "loss": 3.4835,
+      "step": 3930
+    },
+    {
+      "epoch": 5.02688,
+      "grad_norm": 0.5051552057266235,
+      "learning_rate": 0.00015708709112935792,
+      "loss": 3.5332,
+      "step": 3931
+    },
+    {
+      "epoch": 5.02816,
+      "grad_norm": 0.4770444631576538,
+      "learning_rate": 0.00015704670884372052,
+      "loss": 3.5793,
+      "step": 3932
+    },
+    {
+      "epoch": 5.02944,
+      "grad_norm": 0.49611613154411316,
+      "learning_rate": 0.00015700632655808315,
+      "loss": 3.5763,
+      "step": 3933
+    },
+    {
+      "epoch": 5.03072,
+      "grad_norm": 0.5175138711929321,
+      "learning_rate": 0.0001569659442724458,
+      "loss": 3.5942,
+      "step": 3934
+    },
+    {
+      "epoch": 5.032,
+      "grad_norm": 0.5066589117050171,
+      "learning_rate": 0.00015692556198680844,
+      "loss": 3.5047,
+      "step": 3935
+    },
+    {
+      "epoch": 5.03328,
+      "grad_norm": 0.5006115436553955,
+      "learning_rate": 0.00015688517970117107,
+      "loss": 3.44,
+      "step": 3936
+    },
+    {
+      "epoch": 5.03456,
+      "grad_norm": 0.493075430393219,
+      "learning_rate": 0.0001568447974155337,
+      "loss": 3.591,
+      "step": 3937
+    },
+    {
+      "epoch": 5.03584,
+      "grad_norm": 0.5144281387329102,
+      "learning_rate": 0.00015680441512989636,
+      "loss": 3.6073,
+      "step": 3938
+    },
+    {
+      "epoch": 5.03712,
+      "grad_norm": 0.5123122930526733,
+      "learning_rate": 0.000156764032844259,
+      "loss": 3.5315,
+      "step": 3939
+    },
+    {
+      "epoch": 5.0384,
+      "grad_norm": 0.5100513696670532,
+      "learning_rate": 0.00015672365055862162,
+      "loss": 3.4975,
+      "step": 3940
+    },
+    {
+      "epoch": 5.03968,
+      "grad_norm": 0.5032030940055847,
+      "learning_rate": 0.00015668326827298422,
+      "loss": 3.58,
+      "step": 3941
+    },
+    {
+      "epoch": 5.04096,
+      "grad_norm": 0.4905944764614105,
+      "learning_rate": 0.00015664288598734685,
+      "loss": 3.6283,
+      "step": 3942
+    },
+    {
+      "epoch": 5.04224,
+      "grad_norm": 0.5140021443367004,
+      "learning_rate": 0.0001566025037017095,
+      "loss": 3.5188,
+      "step": 3943
+    },
+    {
+      "epoch": 5.04352,
+      "grad_norm": 0.4894171953201294,
+      "learning_rate": 0.00015656212141607214,
+      "loss": 3.4722,
+      "step": 3944
+    },
+    {
+      "epoch": 5.0448,
+      "grad_norm": 0.4924975335597992,
+      "learning_rate": 0.00015652173913043477,
+      "loss": 3.582,
+      "step": 3945
+    },
+    {
+      "epoch": 5.04608,
+      "grad_norm": 0.5047486424446106,
+      "learning_rate": 0.0001564813568447974,
+      "loss": 3.5844,
+      "step": 3946
+    },
+    {
+      "epoch": 5.04736,
+      "grad_norm": 0.5076228380203247,
+      "learning_rate": 0.00015644097455916006,
+      "loss": 3.4754,
+      "step": 3947
+    },
+    {
+      "epoch": 5.04864,
+      "grad_norm": 0.48156487941741943,
+      "learning_rate": 0.00015640059227352269,
+      "loss": 3.5801,
+      "step": 3948
+    },
+    {
+      "epoch": 5.04992,
+      "grad_norm": 0.52396559715271,
+      "learning_rate": 0.0001563602099878853,
+      "loss": 3.5727,
+      "step": 3949
+    },
+    {
+      "epoch": 5.0512,
+      "grad_norm": 0.5130919814109802,
+      "learning_rate": 0.00015631982770224792,
+      "loss": 3.5455,
+      "step": 3950
+    },
+    {
+      "epoch": 5.05248,
+      "grad_norm": 0.5077847242355347,
+      "learning_rate": 0.00015627944541661058,
+      "loss": 3.574,
+      "step": 3951
+    },
+    {
+      "epoch": 5.05376,
+      "grad_norm": 0.5045159459114075,
+      "learning_rate": 0.0001562390631309732,
+      "loss": 3.5511,
+      "step": 3952
+    },
+    {
+      "epoch": 5.05504,
+      "grad_norm": 0.5126574635505676,
+      "learning_rate": 0.00015619868084533584,
+      "loss": 3.6289,
+      "step": 3953
+    },
+    {
+      "epoch": 5.05632,
+      "grad_norm": 0.4960666000843048,
+      "learning_rate": 0.00015615829855969847,
+      "loss": 3.5108,
+      "step": 3954
+    },
+    {
+      "epoch": 5.0576,
+      "grad_norm": 0.49016058444976807,
+      "learning_rate": 0.0001561179162740611,
+      "loss": 3.5506,
+      "step": 3955
+    },
+    {
+      "epoch": 5.05888,
+      "grad_norm": 0.5145988464355469,
+      "learning_rate": 0.00015607753398842375,
+      "loss": 3.5351,
+      "step": 3956
+    },
+    {
+      "epoch": 5.06016,
+      "grad_norm": 0.495882123708725,
+      "learning_rate": 0.00015603715170278636,
+      "loss": 3.5469,
+      "step": 3957
+    },
+    {
+      "epoch": 5.06144,
+      "grad_norm": 0.514789879322052,
+      "learning_rate": 0.000155996769417149,
+      "loss": 3.5172,
+      "step": 3958
+    },
+    {
+      "epoch": 5.06272,
+      "grad_norm": 0.4931298494338989,
+      "learning_rate": 0.00015595638713151162,
+      "loss": 3.5787,
+      "step": 3959
+    },
+    {
+      "epoch": 5.064,
+      "grad_norm": 0.47730547189712524,
+      "learning_rate": 0.00015591600484587427,
+      "loss": 3.5478,
+      "step": 3960
+    },
+    {
+      "epoch": 5.06528,
+      "grad_norm": 0.534153938293457,
+      "learning_rate": 0.0001558756225602369,
+      "loss": 3.5449,
+      "step": 3961
+    },
+    {
+      "epoch": 5.06656,
+      "grad_norm": 0.4863246977329254,
+      "learning_rate": 0.00015583524027459953,
+      "loss": 3.4829,
+      "step": 3962
+    },
+    {
+      "epoch": 5.06784,
+      "grad_norm": 0.5165325999259949,
+      "learning_rate": 0.00015579485798896216,
+      "loss": 3.5866,
+      "step": 3963
+    },
+    {
+      "epoch": 5.06912,
+      "grad_norm": 0.5204668641090393,
+      "learning_rate": 0.00015575447570332482,
+      "loss": 3.5346,
+      "step": 3964
+    },
+    {
+      "epoch": 5.0704,
+      "grad_norm": 0.5140581130981445,
+      "learning_rate": 0.00015571409341768742,
+      "loss": 3.5652,
+      "step": 3965
+    },
+    {
+      "epoch": 5.07168,
+      "grad_norm": 0.5002838373184204,
+      "learning_rate": 0.00015567371113205005,
+      "loss": 3.5275,
+      "step": 3966
+    },
+    {
+      "epoch": 5.07296,
+      "grad_norm": 0.48074761033058167,
+      "learning_rate": 0.00015563332884641268,
+      "loss": 3.5435,
+      "step": 3967
+    },
+    {
+      "epoch": 5.07424,
+      "grad_norm": 0.49693623185157776,
+      "learning_rate": 0.00015559294656077531,
+      "loss": 3.5735,
+      "step": 3968
+    },
+    {
+      "epoch": 5.07552,
+      "grad_norm": 0.5019185543060303,
+      "learning_rate": 0.00015555256427513797,
+      "loss": 3.5511,
+      "step": 3969
+    },
+    {
+      "epoch": 5.0768,
+      "grad_norm": 0.5159568190574646,
+      "learning_rate": 0.0001555121819895006,
+      "loss": 3.5598,
+      "step": 3970
+    },
+    {
+      "epoch": 5.07808,
+      "grad_norm": 0.49248459935188293,
+      "learning_rate": 0.00015547179970386323,
+      "loss": 3.594,
+      "step": 3971
+    },
+    {
+      "epoch": 5.07936,
+      "grad_norm": 0.5076256990432739,
+      "learning_rate": 0.00015543141741822584,
+      "loss": 3.586,
+      "step": 3972
+    },
+    {
+      "epoch": 5.08064,
+      "grad_norm": 0.5257125496864319,
+      "learning_rate": 0.0001553910351325885,
+      "loss": 3.5359,
+      "step": 3973
+    },
+    {
+      "epoch": 5.08192,
+      "grad_norm": 0.48996734619140625,
+      "learning_rate": 0.00015535065284695112,
+      "loss": 3.5319,
+      "step": 3974
+    },
+    {
+      "epoch": 5.0832,
+      "grad_norm": 0.510186493396759,
+      "learning_rate": 0.00015531027056131375,
+      "loss": 3.5718,
+      "step": 3975
+    },
+    {
+      "epoch": 5.08448,
+      "grad_norm": 0.49666544795036316,
+      "learning_rate": 0.00015526988827567638,
+      "loss": 3.5401,
+      "step": 3976
+    },
+    {
+      "epoch": 5.08576,
+      "grad_norm": 0.5227068066596985,
+      "learning_rate": 0.00015522950599003904,
+      "loss": 3.6226,
+      "step": 3977
+    },
+    {
+      "epoch": 5.08704,
+      "grad_norm": 0.5159752368927002,
+      "learning_rate": 0.00015518912370440167,
+      "loss": 3.5353,
+      "step": 3978
+    },
+    {
+      "epoch": 5.08832,
+      "grad_norm": 0.49110716581344604,
+      "learning_rate": 0.0001551487414187643,
+      "loss": 3.4429,
+      "step": 3979
+    },
+    {
+      "epoch": 5.0896,
+      "grad_norm": 0.5033590793609619,
+      "learning_rate": 0.0001551083591331269,
+      "loss": 3.5877,
+      "step": 3980
+    },
+    {
+      "epoch": 5.09088,
+      "grad_norm": 0.5176244974136353,
+      "learning_rate": 0.00015506797684748953,
+      "loss": 3.5388,
+      "step": 3981
+    },
+    {
+      "epoch": 5.09216,
+      "grad_norm": 0.5007118582725525,
+      "learning_rate": 0.0001550275945618522,
+      "loss": 3.5527,
+      "step": 3982
+    },
+    {
+      "epoch": 5.09344,
+      "grad_norm": 0.4941405653953552,
+      "learning_rate": 0.00015498721227621482,
+      "loss": 3.5253,
+      "step": 3983
+    },
+    {
+      "epoch": 5.09472,
+      "grad_norm": 0.49819836020469666,
+      "learning_rate": 0.00015494682999057745,
+      "loss": 3.5868,
+      "step": 3984
+    },
+    {
+      "epoch": 5.096,
+      "grad_norm": 0.4690377712249756,
+      "learning_rate": 0.00015490644770494008,
+      "loss": 3.521,
+      "step": 3985
+    },
+    {
+      "epoch": 5.09728,
+      "grad_norm": 0.5075995326042175,
+      "learning_rate": 0.00015486606541930274,
+      "loss": 3.4984,
+      "step": 3986
+    },
+    {
+      "epoch": 5.09856,
+      "grad_norm": 0.4840170741081238,
+      "learning_rate": 0.00015482568313366537,
+      "loss": 3.5351,
+      "step": 3987
+    },
+    {
+      "epoch": 5.09984,
+      "grad_norm": 0.5013239979743958,
+      "learning_rate": 0.00015478530084802797,
+      "loss": 3.5958,
+      "step": 3988
+    },
+    {
+      "epoch": 5.10112,
+      "grad_norm": 0.500015377998352,
+      "learning_rate": 0.0001547449185623906,
+      "loss": 3.4832,
+      "step": 3989
+    },
+    {
+      "epoch": 5.1024,
+      "grad_norm": 0.5128130316734314,
+      "learning_rate": 0.00015470453627675326,
+      "loss": 3.5816,
+      "step": 3990
+    },
+    {
+      "epoch": 5.10368,
+      "grad_norm": 0.5172329545021057,
+      "learning_rate": 0.0001546641539911159,
+      "loss": 3.474,
+      "step": 3991
+    },
+    {
+      "epoch": 5.10496,
+      "grad_norm": 0.4870198369026184,
+      "learning_rate": 0.00015462377170547852,
+      "loss": 3.5239,
+      "step": 3992
+    },
+    {
+      "epoch": 5.10624,
+      "grad_norm": 0.5066513419151306,
+      "learning_rate": 0.00015458338941984115,
+      "loss": 3.6001,
+      "step": 3993
+    },
+    {
+      "epoch": 5.10752,
+      "grad_norm": 0.5186413526535034,
+      "learning_rate": 0.00015454300713420378,
+      "loss": 3.5248,
+      "step": 3994
+    },
+    {
+      "epoch": 5.1088,
+      "grad_norm": 0.4880644679069519,
+      "learning_rate": 0.00015450262484856644,
+      "loss": 3.5703,
+      "step": 3995
+    },
+    {
+      "epoch": 5.11008,
+      "grad_norm": 0.5023115277290344,
+      "learning_rate": 0.00015446224256292904,
+      "loss": 3.6021,
+      "step": 3996
+    },
+    {
+      "epoch": 5.11136,
+      "grad_norm": 0.490438312292099,
+      "learning_rate": 0.00015442186027729167,
+      "loss": 3.4991,
+      "step": 3997
+    },
+    {
+      "epoch": 5.11264,
+      "grad_norm": 0.48022663593292236,
+      "learning_rate": 0.0001543814779916543,
+      "loss": 3.411,
+      "step": 3998
+    },
+    {
+      "epoch": 5.11392,
+      "grad_norm": 0.4824524521827698,
+      "learning_rate": 0.00015434109570601696,
+      "loss": 3.5548,
+      "step": 3999
+    },
+    {
+      "epoch": 5.1152,
+      "grad_norm": 0.48679637908935547,
+      "learning_rate": 0.00015430071342037959,
+      "loss": 3.4643,
+      "step": 4000
+    },
+    {
+      "epoch": 5.11648,
+      "grad_norm": 0.5010484457015991,
+      "learning_rate": 0.00015426033113474222,
+      "loss": 3.5182,
+      "step": 4001
+    },
+    {
+      "epoch": 5.11776,
+      "grad_norm": 0.49517738819122314,
+      "learning_rate": 0.00015421994884910485,
+      "loss": 3.5063,
+      "step": 4002
+    },
+    {
+      "epoch": 5.11904,
+      "grad_norm": 0.5106388926506042,
+      "learning_rate": 0.0001541795665634675,
+      "loss": 3.6582,
+      "step": 4003
+    },
+    {
+      "epoch": 5.12032,
+      "grad_norm": 0.4887462556362152,
+      "learning_rate": 0.0001541391842778301,
+      "loss": 3.6125,
+      "step": 4004
+    },
+    {
+      "epoch": 5.1216,
+      "grad_norm": 0.4951592981815338,
+      "learning_rate": 0.00015409880199219274,
+      "loss": 3.5286,
+      "step": 4005
+    },
+    {
+      "epoch": 5.12288,
+      "grad_norm": 0.5019987225532532,
+      "learning_rate": 0.00015405841970655537,
+      "loss": 3.5732,
+      "step": 4006
+    },
+    {
+      "epoch": 5.12416,
+      "grad_norm": 0.48524710536003113,
+      "learning_rate": 0.000154018037420918,
+      "loss": 3.5385,
+      "step": 4007
+    },
+    {
+      "epoch": 5.12544,
+      "grad_norm": 0.5041416883468628,
+      "learning_rate": 0.00015397765513528065,
+      "loss": 3.6019,
+      "step": 4008
+    },
+    {
+      "epoch": 5.12672,
+      "grad_norm": 0.48453205823898315,
+      "learning_rate": 0.00015393727284964328,
+      "loss": 3.5418,
+      "step": 4009
+    },
+    {
+      "epoch": 5.128,
+      "grad_norm": 0.5173502564430237,
+      "learning_rate": 0.0001538968905640059,
+      "loss": 3.597,
+      "step": 4010
+    },
+    {
+      "epoch": 5.12928,
+      "grad_norm": 0.5132037997245789,
+      "learning_rate": 0.00015385650827836852,
+      "loss": 3.4569,
+      "step": 4011
+    },
+    {
+      "epoch": 5.13056,
+      "grad_norm": 0.5146015286445618,
+      "learning_rate": 0.0001538161259927312,
+      "loss": 3.5441,
+      "step": 4012
+    },
+    {
+      "epoch": 5.13184,
+      "grad_norm": 0.5107573866844177,
+      "learning_rate": 0.0001537757437070938,
+      "loss": 3.569,
+      "step": 4013
+    },
+    {
+      "epoch": 5.13312,
+      "grad_norm": 0.5163334012031555,
+      "learning_rate": 0.00015373536142145643,
+      "loss": 3.6092,
+      "step": 4014
+    },
+    {
+      "epoch": 5.1344,
+      "grad_norm": 0.512444019317627,
+      "learning_rate": 0.00015369497913581906,
+      "loss": 3.6463,
+      "step": 4015
+    },
+    {
+      "epoch": 5.13568,
+      "grad_norm": 0.5129110813140869,
+      "learning_rate": 0.00015365459685018172,
+      "loss": 3.5649,
+      "step": 4016
+    },
+    {
+      "epoch": 5.13696,
+      "grad_norm": 0.52217698097229,
+      "learning_rate": 0.00015361421456454435,
+      "loss": 3.5541,
+      "step": 4017
+    },
+    {
+      "epoch": 5.13824,
+      "grad_norm": 0.5022246241569519,
+      "learning_rate": 0.00015357383227890698,
+      "loss": 3.4864,
+      "step": 4018
+    },
+    {
+      "epoch": 5.13952,
+      "grad_norm": 0.5264055132865906,
+      "learning_rate": 0.00015353344999326958,
+      "loss": 3.5228,
+      "step": 4019
+    },
+    {
+      "epoch": 5.1408,
+      "grad_norm": 0.5241397023200989,
+      "learning_rate": 0.00015349306770763221,
+      "loss": 3.5233,
+      "step": 4020
+    },
+    {
+      "epoch": 5.14208,
+      "grad_norm": 0.6818204522132874,
+      "learning_rate": 0.00015345268542199487,
+      "loss": 3.639,
+      "step": 4021
+    },
+    {
+      "epoch": 5.14336,
+      "grad_norm": 0.4898463785648346,
+      "learning_rate": 0.0001534123031363575,
+      "loss": 3.4437,
+      "step": 4022
+    },
+    {
+      "epoch": 5.14464,
+      "grad_norm": 0.521457314491272,
+      "learning_rate": 0.00015337192085072013,
+      "loss": 3.5246,
+      "step": 4023
+    },
+    {
+      "epoch": 5.14592,
+      "grad_norm": 0.5136117935180664,
+      "learning_rate": 0.00015333153856508276,
+      "loss": 3.5557,
+      "step": 4024
+    },
+    {
+      "epoch": 5.1472,
+      "grad_norm": 0.49397048354148865,
+      "learning_rate": 0.00015329115627944542,
+      "loss": 3.4808,
+      "step": 4025
+    },
+    {
+      "epoch": 5.14848,
+      "grad_norm": 0.5146060585975647,
+      "learning_rate": 0.00015325077399380805,
+      "loss": 3.518,
+      "step": 4026
+    },
+    {
+      "epoch": 5.14976,
+      "grad_norm": 0.5058059692382812,
+      "learning_rate": 0.00015321039170817068,
+      "loss": 3.5399,
+      "step": 4027
+    },
+    {
+      "epoch": 5.15104,
+      "grad_norm": 0.4919331967830658,
+      "learning_rate": 0.00015317000942253328,
+      "loss": 3.5817,
+      "step": 4028
+    },
+    {
+      "epoch": 5.15232,
+      "grad_norm": 0.4798501133918762,
+      "learning_rate": 0.00015312962713689594,
+      "loss": 3.5386,
+      "step": 4029
+    },
+    {
+      "epoch": 5.1536,
+      "grad_norm": 0.49712446331977844,
+      "learning_rate": 0.00015308924485125857,
+      "loss": 3.532,
+      "step": 4030
+    },
+    {
+      "epoch": 5.15488,
+      "grad_norm": 0.5078490376472473,
+      "learning_rate": 0.0001530488625656212,
+      "loss": 3.5602,
+      "step": 4031
+    },
+    {
+      "epoch": 5.15616,
+      "grad_norm": 0.49252021312713623,
+      "learning_rate": 0.00015300848027998383,
+      "loss": 3.5433,
+      "step": 4032
+    },
+    {
+      "epoch": 5.15744,
+      "grad_norm": 0.5331763029098511,
+      "learning_rate": 0.00015296809799434646,
+      "loss": 3.5577,
+      "step": 4033
+    },
+    {
+      "epoch": 5.15872,
+      "grad_norm": 0.4990587830543518,
+      "learning_rate": 0.00015292771570870912,
+      "loss": 3.559,
+      "step": 4034
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.5130864977836609,
+      "learning_rate": 0.00015288733342307175,
+      "loss": 3.5403,
+      "step": 4035
+    },
+    {
+      "epoch": 5.16128,
+      "grad_norm": 0.4923115372657776,
+      "learning_rate": 0.00015284695113743435,
+      "loss": 3.4671,
+      "step": 4036
+    },
+    {
+      "epoch": 5.16256,
+      "grad_norm": 0.5185816287994385,
+      "learning_rate": 0.00015280656885179698,
+      "loss": 3.6086,
+      "step": 4037
+    },
+    {
+      "epoch": 5.16384,
+      "grad_norm": 0.5011401176452637,
+      "learning_rate": 0.00015276618656615964,
+      "loss": 3.5315,
+      "step": 4038
+    },
+    {
+      "epoch": 5.16512,
+      "grad_norm": 0.5076486468315125,
+      "learning_rate": 0.00015272580428052227,
+      "loss": 3.5743,
+      "step": 4039
+    },
+    {
+      "epoch": 5.1664,
+      "grad_norm": 0.49623793363571167,
+      "learning_rate": 0.0001526854219948849,
+      "loss": 3.6119,
+      "step": 4040
+    },
+    {
+      "epoch": 5.16768,
+      "grad_norm": 0.5004082918167114,
+      "learning_rate": 0.00015264503970924753,
+      "loss": 3.4876,
+      "step": 4041
+    },
+    {
+      "epoch": 5.16896,
+      "grad_norm": 0.4831116199493408,
+      "learning_rate": 0.00015260465742361018,
+      "loss": 3.5508,
+      "step": 4042
+    },
+    {
+      "epoch": 5.17024,
+      "grad_norm": 0.5067176818847656,
+      "learning_rate": 0.00015256427513797281,
+      "loss": 3.56,
+      "step": 4043
+    },
+    {
+      "epoch": 5.17152,
+      "grad_norm": 0.5060960650444031,
+      "learning_rate": 0.00015252389285233542,
+      "loss": 3.5146,
+      "step": 4044
+    },
+    {
+      "epoch": 5.1728,
+      "grad_norm": 0.5012076497077942,
+      "learning_rate": 0.00015248351056669805,
+      "loss": 3.5358,
+      "step": 4045
+    },
+    {
+      "epoch": 5.17408,
+      "grad_norm": 0.48700445890426636,
+      "learning_rate": 0.0001524431282810607,
+      "loss": 3.6407,
+      "step": 4046
+    },
+    {
+      "epoch": 5.17536,
+      "grad_norm": 0.49862363934516907,
+      "learning_rate": 0.00015240274599542333,
+      "loss": 3.5476,
+      "step": 4047
+    },
+    {
+      "epoch": 5.17664,
+      "grad_norm": 0.5184906721115112,
+      "learning_rate": 0.00015236236370978596,
+      "loss": 3.6116,
+      "step": 4048
+    },
+    {
+      "epoch": 5.17792,
+      "grad_norm": 0.4845800995826721,
+      "learning_rate": 0.0001523219814241486,
+      "loss": 3.4903,
+      "step": 4049
+    },
+    {
+      "epoch": 5.1792,
+      "grad_norm": 0.5097942352294922,
+      "learning_rate": 0.00015228159913851122,
+      "loss": 3.5238,
+      "step": 4050
+    },
+    {
+      "epoch": 5.18048,
+      "grad_norm": 0.5073378086090088,
+      "learning_rate": 0.00015224121685287388,
+      "loss": 3.6045,
+      "step": 4051
+    },
+    {
+      "epoch": 5.18176,
+      "grad_norm": 0.4886874258518219,
+      "learning_rate": 0.00015220083456723649,
+      "loss": 3.5069,
+      "step": 4052
+    },
+    {
+      "epoch": 5.18304,
+      "grad_norm": 0.49647676944732666,
+      "learning_rate": 0.00015216045228159912,
+      "loss": 3.5864,
+      "step": 4053
+    },
+    {
+      "epoch": 5.18432,
+      "grad_norm": 0.4973941445350647,
+      "learning_rate": 0.00015212006999596175,
+      "loss": 3.5571,
+      "step": 4054
+    },
+    {
+      "epoch": 5.1856,
+      "grad_norm": 0.5050147175788879,
+      "learning_rate": 0.0001520796877103244,
+      "loss": 3.5203,
+      "step": 4055
+    },
+    {
+      "epoch": 5.18688,
+      "grad_norm": 0.48217907547950745,
+      "learning_rate": 0.00015203930542468703,
+      "loss": 3.5327,
+      "step": 4056
+    },
+    {
+      "epoch": 5.18816,
+      "grad_norm": 0.5200319290161133,
+      "learning_rate": 0.00015199892313904966,
+      "loss": 3.6177,
+      "step": 4057
+    },
+    {
+      "epoch": 5.18944,
+      "grad_norm": 0.4934719204902649,
+      "learning_rate": 0.0001519585408534123,
+      "loss": 3.5692,
+      "step": 4058
+    },
+    {
+      "epoch": 5.19072,
+      "grad_norm": 0.5016127824783325,
+      "learning_rate": 0.00015191815856777495,
+      "loss": 3.5915,
+      "step": 4059
+    },
+    {
+      "epoch": 5.192,
+      "grad_norm": 0.502185046672821,
+      "learning_rate": 0.00015187777628213755,
+      "loss": 3.5411,
+      "step": 4060
+    },
+    {
+      "epoch": 5.19328,
+      "grad_norm": 0.489409863948822,
+      "learning_rate": 0.00015183739399650018,
+      "loss": 3.5464,
+      "step": 4061
+    },
+    {
+      "epoch": 5.19456,
+      "grad_norm": 0.5016448497772217,
+      "learning_rate": 0.0001517970117108628,
+      "loss": 3.5695,
+      "step": 4062
+    },
+    {
+      "epoch": 5.19584,
+      "grad_norm": 0.49411389231681824,
+      "learning_rate": 0.00015175662942522544,
+      "loss": 3.5293,
+      "step": 4063
+    },
+    {
+      "epoch": 5.19712,
+      "grad_norm": 0.5000140070915222,
+      "learning_rate": 0.0001517162471395881,
+      "loss": 3.5094,
+      "step": 4064
+    },
+    {
+      "epoch": 5.1984,
+      "grad_norm": 0.49250856041908264,
+      "learning_rate": 0.00015167586485395073,
+      "loss": 3.5954,
+      "step": 4065
+    },
+    {
+      "epoch": 5.19968,
+      "grad_norm": 0.48768413066864014,
+      "learning_rate": 0.00015163548256831336,
+      "loss": 3.4345,
+      "step": 4066
+    },
+    {
+      "epoch": 5.20096,
+      "grad_norm": 0.506270170211792,
+      "learning_rate": 0.00015159510028267596,
+      "loss": 3.5516,
+      "step": 4067
+    },
+    {
+      "epoch": 5.20224,
+      "grad_norm": 0.5164130926132202,
+      "learning_rate": 0.00015155471799703862,
+      "loss": 3.558,
+      "step": 4068
+    },
+    {
+      "epoch": 5.20352,
+      "grad_norm": 0.5122447609901428,
+      "learning_rate": 0.00015151433571140125,
+      "loss": 3.596,
+      "step": 4069
+    },
+    {
+      "epoch": 5.2048,
+      "grad_norm": 0.5188588500022888,
+      "learning_rate": 0.00015147395342576388,
+      "loss": 3.5264,
+      "step": 4070
+    },
+    {
+      "epoch": 5.20608,
+      "grad_norm": 0.49674516916275024,
+      "learning_rate": 0.0001514335711401265,
+      "loss": 3.5941,
+      "step": 4071
+    },
+    {
+      "epoch": 5.2073599999999995,
+      "grad_norm": 0.5053087472915649,
+      "learning_rate": 0.00015139318885448917,
+      "loss": 3.5922,
+      "step": 4072
+    },
+    {
+      "epoch": 5.20864,
+      "grad_norm": 0.4910738468170166,
+      "learning_rate": 0.0001513528065688518,
+      "loss": 3.516,
+      "step": 4073
+    },
+    {
+      "epoch": 5.20992,
+      "grad_norm": 0.48027095198631287,
+      "learning_rate": 0.00015131242428321443,
+      "loss": 3.4961,
+      "step": 4074
+    },
+    {
+      "epoch": 5.2112,
+      "grad_norm": 0.5024387240409851,
+      "learning_rate": 0.00015127204199757703,
+      "loss": 3.4995,
+      "step": 4075
+    },
+    {
+      "epoch": 5.21248,
+      "grad_norm": 0.5177842378616333,
+      "learning_rate": 0.00015123165971193966,
+      "loss": 3.5783,
+      "step": 4076
+    },
+    {
+      "epoch": 5.21376,
+      "grad_norm": 0.4885920584201813,
+      "learning_rate": 0.00015119127742630232,
+      "loss": 3.6155,
+      "step": 4077
+    },
+    {
+      "epoch": 5.21504,
+      "grad_norm": 0.5077448487281799,
+      "learning_rate": 0.00015115089514066495,
+      "loss": 3.581,
+      "step": 4078
+    },
+    {
+      "epoch": 5.21632,
+      "grad_norm": 0.48989179730415344,
+      "learning_rate": 0.00015111051285502758,
+      "loss": 3.5222,
+      "step": 4079
+    },
+    {
+      "epoch": 5.2176,
+      "grad_norm": 0.4760512113571167,
+      "learning_rate": 0.0001510701305693902,
+      "loss": 3.5481,
+      "step": 4080
+    },
+    {
+      "epoch": 5.21888,
+      "grad_norm": 0.49710988998413086,
+      "learning_rate": 0.00015102974828375287,
+      "loss": 3.5808,
+      "step": 4081
+    },
+    {
+      "epoch": 5.22016,
+      "grad_norm": 0.49187228083610535,
+      "learning_rate": 0.0001509893659981155,
+      "loss": 3.5191,
+      "step": 4082
+    },
+    {
+      "epoch": 5.22144,
+      "grad_norm": 0.5187274813652039,
+      "learning_rate": 0.0001509489837124781,
+      "loss": 3.6189,
+      "step": 4083
+    },
+    {
+      "epoch": 5.22272,
+      "grad_norm": 0.523126482963562,
+      "learning_rate": 0.00015090860142684073,
+      "loss": 3.6136,
+      "step": 4084
+    },
+    {
+      "epoch": 5.224,
+      "grad_norm": 0.49129125475883484,
+      "learning_rate": 0.00015086821914120339,
+      "loss": 3.5228,
+      "step": 4085
+    },
+    {
+      "epoch": 5.22528,
+      "grad_norm": 0.5184364318847656,
+      "learning_rate": 0.00015082783685556602,
+      "loss": 3.5247,
+      "step": 4086
+    },
+    {
+      "epoch": 5.22656,
+      "grad_norm": 0.5139320492744446,
+      "learning_rate": 0.00015078745456992865,
+      "loss": 3.5408,
+      "step": 4087
+    },
+    {
+      "epoch": 5.22784,
+      "grad_norm": 0.5081346035003662,
+      "learning_rate": 0.00015074707228429128,
+      "loss": 3.6262,
+      "step": 4088
+    },
+    {
+      "epoch": 5.22912,
+      "grad_norm": 0.5397776365280151,
+      "learning_rate": 0.0001507066899986539,
+      "loss": 3.5904,
+      "step": 4089
+    },
+    {
+      "epoch": 5.2304,
+      "grad_norm": 0.5284587740898132,
+      "learning_rate": 0.00015066630771301656,
+      "loss": 3.5313,
+      "step": 4090
+    },
+    {
+      "epoch": 5.23168,
+      "grad_norm": 0.4830555319786072,
+      "learning_rate": 0.00015062592542737917,
+      "loss": 3.5611,
+      "step": 4091
+    },
+    {
+      "epoch": 5.23296,
+      "grad_norm": 0.5219169855117798,
+      "learning_rate": 0.0001505855431417418,
+      "loss": 3.6486,
+      "step": 4092
+    },
+    {
+      "epoch": 5.23424,
+      "grad_norm": 0.5159496068954468,
+      "learning_rate": 0.00015054516085610443,
+      "loss": 3.4704,
+      "step": 4093
+    },
+    {
+      "epoch": 5.23552,
+      "grad_norm": 0.530392587184906,
+      "learning_rate": 0.00015050477857046708,
+      "loss": 3.517,
+      "step": 4094
+    },
+    {
+      "epoch": 5.2368,
+      "grad_norm": 0.5039182901382446,
+      "learning_rate": 0.00015046439628482971,
+      "loss": 3.5843,
+      "step": 4095
+    },
+    {
+      "epoch": 5.23808,
+      "grad_norm": 0.5051393508911133,
+      "learning_rate": 0.00015042401399919234,
+      "loss": 3.5637,
+      "step": 4096
+    },
+    {
+      "epoch": 5.23936,
+      "grad_norm": 0.5062666535377502,
+      "learning_rate": 0.00015038363171355497,
+      "loss": 3.5598,
+      "step": 4097
+    },
+    {
+      "epoch": 5.24064,
+      "grad_norm": 0.5078728199005127,
+      "learning_rate": 0.00015034324942791763,
+      "loss": 3.5031,
+      "step": 4098
+    },
+    {
+      "epoch": 5.24192,
+      "grad_norm": 0.5176907181739807,
+      "learning_rate": 0.00015030286714228023,
+      "loss": 3.5606,
+      "step": 4099
+    },
+    {
+      "epoch": 5.2432,
+      "grad_norm": 0.5005961060523987,
+      "learning_rate": 0.00015026248485664286,
+      "loss": 3.658,
+      "step": 4100
+    },
+    {
+      "epoch": 5.24448,
+      "grad_norm": 0.5151212811470032,
+      "learning_rate": 0.0001502221025710055,
+      "loss": 3.6295,
+      "step": 4101
+    },
+    {
+      "epoch": 5.24576,
+      "grad_norm": 0.5224319100379944,
+      "learning_rate": 0.00015018172028536812,
+      "loss": 3.5873,
+      "step": 4102
+    },
+    {
+      "epoch": 5.24704,
+      "grad_norm": 0.5269538164138794,
+      "learning_rate": 0.00015014133799973078,
+      "loss": 3.5697,
+      "step": 4103
+    },
+    {
+      "epoch": 5.24832,
+      "grad_norm": 0.4943110942840576,
+      "learning_rate": 0.0001501009557140934,
+      "loss": 3.4694,
+      "step": 4104
+    },
+    {
+      "epoch": 5.2496,
+      "grad_norm": 0.5208888649940491,
+      "learning_rate": 0.00015006057342845604,
+      "loss": 3.5701,
+      "step": 4105
+    },
+    {
+      "epoch": 5.25088,
+      "grad_norm": 0.5066584944725037,
+      "learning_rate": 0.00015002019114281864,
+      "loss": 3.6059,
+      "step": 4106
+    },
+    {
+      "epoch": 5.25216,
+      "grad_norm": 0.5034709572792053,
+      "learning_rate": 0.0001499798088571813,
+      "loss": 3.5807,
+      "step": 4107
+    },
+    {
+      "epoch": 5.25344,
+      "grad_norm": 0.49684327840805054,
+      "learning_rate": 0.00014993942657154393,
+      "loss": 3.5518,
+      "step": 4108
+    },
+    {
+      "epoch": 5.25472,
+      "grad_norm": 0.5177910327911377,
+      "learning_rate": 0.00014989904428590656,
+      "loss": 3.5615,
+      "step": 4109
+    },
+    {
+      "epoch": 5.256,
+      "grad_norm": 0.5041977167129517,
+      "learning_rate": 0.00014985866200026922,
+      "loss": 3.5278,
+      "step": 4110
+    },
+    {
+      "epoch": 5.25728,
+      "grad_norm": 0.5060576796531677,
+      "learning_rate": 0.00014981827971463182,
+      "loss": 3.5217,
+      "step": 4111
+    },
+    {
+      "epoch": 5.25856,
+      "grad_norm": 0.5150559544563293,
+      "learning_rate": 0.00014977789742899448,
+      "loss": 3.5016,
+      "step": 4112
+    },
+    {
+      "epoch": 5.25984,
+      "grad_norm": 0.48947054147720337,
+      "learning_rate": 0.0001497375151433571,
+      "loss": 3.5498,
+      "step": 4113
+    },
+    {
+      "epoch": 5.26112,
+      "grad_norm": 0.5176462531089783,
+      "learning_rate": 0.00014969713285771974,
+      "loss": 3.5463,
+      "step": 4114
+    },
+    {
+      "epoch": 5.2624,
+      "grad_norm": 0.501726508140564,
+      "learning_rate": 0.00014965675057208237,
+      "loss": 3.5083,
+      "step": 4115
+    },
+    {
+      "epoch": 5.26368,
+      "grad_norm": 0.5025134086608887,
+      "learning_rate": 0.000149616368286445,
+      "loss": 3.5474,
+      "step": 4116
+    },
+    {
+      "epoch": 5.26496,
+      "grad_norm": 0.4934270679950714,
+      "learning_rate": 0.00014957598600080763,
+      "loss": 3.5807,
+      "step": 4117
+    },
+    {
+      "epoch": 5.26624,
+      "grad_norm": 0.49150562286376953,
+      "learning_rate": 0.00014953560371517026,
+      "loss": 3.5932,
+      "step": 4118
+    },
+    {
+      "epoch": 5.26752,
+      "grad_norm": 0.4923061728477478,
+      "learning_rate": 0.0001494952214295329,
+      "loss": 3.5125,
+      "step": 4119
+    },
+    {
+      "epoch": 5.2688,
+      "grad_norm": 0.506479024887085,
+      "learning_rate": 0.00014945483914389552,
+      "loss": 3.6409,
+      "step": 4120
+    },
+    {
+      "epoch": 5.27008,
+      "grad_norm": 0.47946614027023315,
+      "learning_rate": 0.00014941445685825818,
+      "loss": 3.5563,
+      "step": 4121
+    },
+    {
+      "epoch": 5.27136,
+      "grad_norm": 0.5117473006248474,
+      "learning_rate": 0.0001493740745726208,
+      "loss": 3.5387,
+      "step": 4122
+    },
+    {
+      "epoch": 5.27264,
+      "grad_norm": 0.49644434452056885,
+      "learning_rate": 0.00014933369228698344,
+      "loss": 3.6186,
+      "step": 4123
+    },
+    {
+      "epoch": 5.27392,
+      "grad_norm": 0.5136969685554504,
+      "learning_rate": 0.00014929331000134607,
+      "loss": 3.6033,
+      "step": 4124
+    },
+    {
+      "epoch": 5.2752,
+      "grad_norm": 0.5244076251983643,
+      "learning_rate": 0.0001492529277157087,
+      "loss": 3.6165,
+      "step": 4125
+    },
+    {
+      "epoch": 5.27648,
+      "grad_norm": 0.49152183532714844,
+      "learning_rate": 0.00014921254543007133,
+      "loss": 3.4846,
+      "step": 4126
+    },
+    {
+      "epoch": 5.27776,
+      "grad_norm": 0.49941569566726685,
+      "learning_rate": 0.00014917216314443396,
+      "loss": 3.6035,
+      "step": 4127
+    },
+    {
+      "epoch": 5.27904,
+      "grad_norm": 0.502313494682312,
+      "learning_rate": 0.0001491317808587966,
+      "loss": 3.603,
+      "step": 4128
+    },
+    {
+      "epoch": 5.28032,
+      "grad_norm": 0.5102902054786682,
+      "learning_rate": 0.00014909139857315922,
+      "loss": 3.6138,
+      "step": 4129
+    },
+    {
+      "epoch": 5.2816,
+      "grad_norm": 0.5135358572006226,
+      "learning_rate": 0.00014905101628752187,
+      "loss": 3.5688,
+      "step": 4130
+    },
+    {
+      "epoch": 5.2828800000000005,
+      "grad_norm": 0.4999293386936188,
+      "learning_rate": 0.00014901063400188448,
+      "loss": 3.536,
+      "step": 4131
+    },
+    {
+      "epoch": 5.28416,
+      "grad_norm": 0.49614301323890686,
+      "learning_rate": 0.00014897025171624714,
+      "loss": 3.5474,
+      "step": 4132
+    },
+    {
+      "epoch": 5.28544,
+      "grad_norm": 0.5085309147834778,
+      "learning_rate": 0.00014892986943060977,
+      "loss": 3.6116,
+      "step": 4133
+    },
+    {
+      "epoch": 5.28672,
+      "grad_norm": 0.479676753282547,
+      "learning_rate": 0.0001488894871449724,
+      "loss": 3.5612,
+      "step": 4134
+    },
+    {
+      "epoch": 5.288,
+      "grad_norm": 0.4859643280506134,
+      "learning_rate": 0.00014884910485933503,
+      "loss": 3.5404,
+      "step": 4135
+    },
+    {
+      "epoch": 5.28928,
+      "grad_norm": 0.49102458357810974,
+      "learning_rate": 0.00014880872257369766,
+      "loss": 3.567,
+      "step": 4136
+    },
+    {
+      "epoch": 5.29056,
+      "grad_norm": 0.4906257688999176,
+      "learning_rate": 0.00014876834028806029,
+      "loss": 3.57,
+      "step": 4137
+    },
+    {
+      "epoch": 5.29184,
+      "grad_norm": 0.48179754614830017,
+      "learning_rate": 0.00014872795800242294,
+      "loss": 3.5837,
+      "step": 4138
+    },
+    {
+      "epoch": 5.29312,
+      "grad_norm": 0.48550358414649963,
+      "learning_rate": 0.00014868757571678555,
+      "loss": 3.4787,
+      "step": 4139
+    },
+    {
+      "epoch": 5.2943999999999996,
+      "grad_norm": 0.49071556329727173,
+      "learning_rate": 0.0001486471934311482,
+      "loss": 3.5332,
+      "step": 4140
+    },
+    {
+      "epoch": 5.29568,
+      "grad_norm": 0.4878070652484894,
+      "learning_rate": 0.00014860681114551083,
+      "loss": 3.5856,
+      "step": 4141
+    },
+    {
+      "epoch": 5.29696,
+      "grad_norm": 0.4919761121273041,
+      "learning_rate": 0.00014856642885987344,
+      "loss": 3.5238,
+      "step": 4142
+    },
+    {
+      "epoch": 5.29824,
+      "grad_norm": 0.502760648727417,
+      "learning_rate": 0.0001485260465742361,
+      "loss": 3.5677,
+      "step": 4143
+    },
+    {
+      "epoch": 5.29952,
+      "grad_norm": 0.5055625438690186,
+      "learning_rate": 0.00014848566428859872,
+      "loss": 3.6354,
+      "step": 4144
+    },
+    {
+      "epoch": 5.3008,
+      "grad_norm": 0.4847923517227173,
+      "learning_rate": 0.00014844528200296135,
+      "loss": 3.5031,
+      "step": 4145
+    },
+    {
+      "epoch": 5.30208,
+      "grad_norm": 0.4974401593208313,
+      "learning_rate": 0.00014840489971732398,
+      "loss": 3.5579,
+      "step": 4146
+    },
+    {
+      "epoch": 5.30336,
+      "grad_norm": 0.4971027076244354,
+      "learning_rate": 0.0001483645174316866,
+      "loss": 3.5142,
+      "step": 4147
+    },
+    {
+      "epoch": 5.30464,
+      "grad_norm": 0.5040155649185181,
+      "learning_rate": 0.00014832413514604924,
+      "loss": 3.629,
+      "step": 4148
+    },
+    {
+      "epoch": 5.30592,
+      "grad_norm": 0.5108280181884766,
+      "learning_rate": 0.0001482837528604119,
+      "loss": 3.5502,
+      "step": 4149
+    },
+    {
+      "epoch": 5.3072,
+      "grad_norm": 0.5081576704978943,
+      "learning_rate": 0.0001482433705747745,
+      "loss": 3.5522,
+      "step": 4150
+    },
+    {
+      "epoch": 5.30848,
+      "grad_norm": 0.49711495637893677,
+      "learning_rate": 0.00014820298828913716,
+      "loss": 3.5817,
+      "step": 4151
+    },
+    {
+      "epoch": 5.30976,
+      "grad_norm": 0.4980352520942688,
+      "learning_rate": 0.0001481626060034998,
+      "loss": 3.5515,
+      "step": 4152
+    },
+    {
+      "epoch": 5.31104,
+      "grad_norm": 0.5169786214828491,
+      "learning_rate": 0.00014812222371786242,
+      "loss": 3.5684,
+      "step": 4153
+    },
+    {
+      "epoch": 5.31232,
+      "grad_norm": 0.500923752784729,
+      "learning_rate": 0.00014808184143222505,
+      "loss": 3.5626,
+      "step": 4154
+    },
+    {
+      "epoch": 5.3136,
+      "grad_norm": 0.4993918836116791,
+      "learning_rate": 0.00014804145914658768,
+      "loss": 3.5241,
+      "step": 4155
+    },
+    {
+      "epoch": 5.31488,
+      "grad_norm": 0.5014583468437195,
+      "learning_rate": 0.0001480010768609503,
+      "loss": 3.5245,
+      "step": 4156
+    },
+    {
+      "epoch": 5.31616,
+      "grad_norm": 0.5105462670326233,
+      "learning_rate": 0.00014796069457531294,
+      "loss": 3.5115,
+      "step": 4157
+    },
+    {
+      "epoch": 5.31744,
+      "grad_norm": 0.506993293762207,
+      "learning_rate": 0.0001479203122896756,
+      "loss": 3.5967,
+      "step": 4158
+    },
+    {
+      "epoch": 5.31872,
+      "grad_norm": 0.4856204688549042,
+      "learning_rate": 0.0001478799300040382,
+      "loss": 3.5997,
+      "step": 4159
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.5217317938804626,
+      "learning_rate": 0.00014783954771840086,
+      "loss": 3.5199,
+      "step": 4160
+    },
+    {
+      "epoch": 5.32128,
+      "grad_norm": 0.4968002140522003,
+      "learning_rate": 0.0001477991654327635,
+      "loss": 3.5319,
+      "step": 4161
+    },
+    {
+      "epoch": 5.32256,
+      "grad_norm": 0.49874651432037354,
+      "learning_rate": 0.00014775878314712612,
+      "loss": 3.5956,
+      "step": 4162
+    },
+    {
+      "epoch": 5.32384,
+      "grad_norm": 0.5208538770675659,
+      "learning_rate": 0.00014771840086148875,
+      "loss": 3.5383,
+      "step": 4163
+    },
+    {
+      "epoch": 5.32512,
+      "grad_norm": 0.48870334029197693,
+      "learning_rate": 0.00014767801857585138,
+      "loss": 3.526,
+      "step": 4164
+    },
+    {
+      "epoch": 5.3264,
+      "grad_norm": 0.47953054308891296,
+      "learning_rate": 0.000147637636290214,
+      "loss": 3.5108,
+      "step": 4165
+    },
+    {
+      "epoch": 5.32768,
+      "grad_norm": 0.5066089034080505,
+      "learning_rate": 0.00014759725400457667,
+      "loss": 3.5234,
+      "step": 4166
+    },
+    {
+      "epoch": 5.32896,
+      "grad_norm": 0.5130050778388977,
+      "learning_rate": 0.00014755687171893927,
+      "loss": 3.6001,
+      "step": 4167
+    },
+    {
+      "epoch": 5.33024,
+      "grad_norm": 0.5042834281921387,
+      "learning_rate": 0.0001475164894333019,
+      "loss": 3.6354,
+      "step": 4168
+    },
+    {
+      "epoch": 5.33152,
+      "grad_norm": 0.5093918442726135,
+      "learning_rate": 0.00014747610714766456,
+      "loss": 3.5279,
+      "step": 4169
+    },
+    {
+      "epoch": 5.3328,
+      "grad_norm": 0.5074060559272766,
+      "learning_rate": 0.00014743572486202716,
+      "loss": 3.5477,
+      "step": 4170
+    },
+    {
+      "epoch": 5.33408,
+      "grad_norm": 0.5181780457496643,
+      "learning_rate": 0.00014739534257638982,
+      "loss": 3.5895,
+      "step": 4171
+    },
+    {
+      "epoch": 5.33536,
+      "grad_norm": 0.4882206618785858,
+      "learning_rate": 0.00014735496029075245,
+      "loss": 3.5639,
+      "step": 4172
+    },
+    {
+      "epoch": 5.33664,
+      "grad_norm": 0.4917638599872589,
+      "learning_rate": 0.00014731457800511508,
+      "loss": 3.5782,
+      "step": 4173
+    },
+    {
+      "epoch": 5.33792,
+      "grad_norm": 0.5091182589530945,
+      "learning_rate": 0.0001472741957194777,
+      "loss": 3.565,
+      "step": 4174
+    },
+    {
+      "epoch": 5.3392,
+      "grad_norm": 0.5003569722175598,
+      "learning_rate": 0.00014723381343384034,
+      "loss": 3.5255,
+      "step": 4175
+    },
+    {
+      "epoch": 5.34048,
+      "grad_norm": 0.5167132019996643,
+      "learning_rate": 0.00014719343114820297,
+      "loss": 3.5591,
+      "step": 4176
+    },
+    {
+      "epoch": 5.34176,
+      "grad_norm": 0.521111011505127,
+      "learning_rate": 0.00014715304886256562,
+      "loss": 3.6051,
+      "step": 4177
+    },
+    {
+      "epoch": 5.34304,
+      "grad_norm": 0.5027885437011719,
+      "learning_rate": 0.00014711266657692823,
+      "loss": 3.5387,
+      "step": 4178
+    },
+    {
+      "epoch": 5.34432,
+      "grad_norm": 0.5094881057739258,
+      "learning_rate": 0.00014707228429129088,
+      "loss": 3.4888,
+      "step": 4179
+    },
+    {
+      "epoch": 5.3456,
+      "grad_norm": 0.5104243755340576,
+      "learning_rate": 0.00014703190200565351,
+      "loss": 3.6305,
+      "step": 4180
+    },
+    {
+      "epoch": 5.34688,
+      "grad_norm": 0.5000970363616943,
+      "learning_rate": 0.00014699151972001614,
+      "loss": 3.5628,
+      "step": 4181
+    },
+    {
+      "epoch": 5.34816,
+      "grad_norm": 0.5333858132362366,
+      "learning_rate": 0.00014695113743437877,
+      "loss": 3.553,
+      "step": 4182
+    },
+    {
+      "epoch": 5.3494399999999995,
+      "grad_norm": 0.4926629364490509,
+      "learning_rate": 0.0001469107551487414,
+      "loss": 3.5841,
+      "step": 4183
+    },
+    {
+      "epoch": 5.35072,
+      "grad_norm": 0.5149261951446533,
+      "learning_rate": 0.00014687037286310403,
+      "loss": 3.5844,
+      "step": 4184
+    },
+    {
+      "epoch": 5.352,
+      "grad_norm": 0.48902156949043274,
+      "learning_rate": 0.00014682999057746666,
+      "loss": 3.5553,
+      "step": 4185
+    },
+    {
+      "epoch": 5.35328,
+      "grad_norm": 0.5008950233459473,
+      "learning_rate": 0.0001467896082918293,
+      "loss": 3.5949,
+      "step": 4186
+    },
+    {
+      "epoch": 5.35456,
+      "grad_norm": 0.5129325985908508,
+      "learning_rate": 0.00014674922600619192,
+      "loss": 3.5755,
+      "step": 4187
+    },
+    {
+      "epoch": 5.35584,
+      "grad_norm": 0.5095540881156921,
+      "learning_rate": 0.00014670884372055458,
+      "loss": 3.6056,
+      "step": 4188
+    },
+    {
+      "epoch": 5.35712,
+      "grad_norm": 0.497380793094635,
+      "learning_rate": 0.0001466684614349172,
+      "loss": 3.5998,
+      "step": 4189
+    },
+    {
+      "epoch": 5.3584,
+      "grad_norm": 0.5010343194007874,
+      "learning_rate": 0.00014662807914927984,
+      "loss": 3.5809,
+      "step": 4190
+    },
+    {
+      "epoch": 5.35968,
+      "grad_norm": 0.498839795589447,
+      "learning_rate": 0.00014658769686364247,
+      "loss": 3.6314,
+      "step": 4191
+    },
+    {
+      "epoch": 5.36096,
+      "grad_norm": 0.5233437418937683,
+      "learning_rate": 0.0001465473145780051,
+      "loss": 3.5928,
+      "step": 4192
+    },
+    {
+      "epoch": 5.36224,
+      "grad_norm": 0.4920739233493805,
+      "learning_rate": 0.00014650693229236773,
+      "loss": 3.5803,
+      "step": 4193
+    },
+    {
+      "epoch": 5.36352,
+      "grad_norm": 0.5078881978988647,
+      "learning_rate": 0.00014646655000673036,
+      "loss": 3.6682,
+      "step": 4194
+    },
+    {
+      "epoch": 5.3648,
+      "grad_norm": 0.48847028613090515,
+      "learning_rate": 0.000146426167721093,
+      "loss": 3.5167,
+      "step": 4195
+    },
+    {
+      "epoch": 5.36608,
+      "grad_norm": 0.49251747131347656,
+      "learning_rate": 0.00014638578543545562,
+      "loss": 3.6127,
+      "step": 4196
+    },
+    {
+      "epoch": 5.36736,
+      "grad_norm": 0.491445928812027,
+      "learning_rate": 0.00014634540314981828,
+      "loss": 3.6277,
+      "step": 4197
+    },
+    {
+      "epoch": 5.36864,
+      "grad_norm": 0.5016657710075378,
+      "learning_rate": 0.00014630502086418088,
+      "loss": 3.5437,
+      "step": 4198
+    },
+    {
+      "epoch": 5.3699200000000005,
+      "grad_norm": 0.4860299825668335,
+      "learning_rate": 0.00014626463857854354,
+      "loss": 3.5576,
+      "step": 4199
+    },
+    {
+      "epoch": 5.3712,
+      "grad_norm": 0.5150024890899658,
+      "learning_rate": 0.00014622425629290617,
+      "loss": 3.5746,
+      "step": 4200
+    },
+    {
+      "epoch": 5.37248,
+      "grad_norm": 0.49160733819007874,
+      "learning_rate": 0.0001461838740072688,
+      "loss": 3.5343,
+      "step": 4201
+    },
+    {
+      "epoch": 5.37376,
+      "grad_norm": 0.5088993906974792,
+      "learning_rate": 0.00014614349172163143,
+      "loss": 3.6182,
+      "step": 4202
+    },
+    {
+      "epoch": 5.37504,
+      "grad_norm": 0.4988862872123718,
+      "learning_rate": 0.00014610310943599406,
+      "loss": 3.5938,
+      "step": 4203
+    },
+    {
+      "epoch": 5.37632,
+      "grad_norm": 0.49565553665161133,
+      "learning_rate": 0.0001460627271503567,
+      "loss": 3.5703,
+      "step": 4204
+    },
+    {
+      "epoch": 5.3776,
+      "grad_norm": 0.5155361294746399,
+      "learning_rate": 0.00014602234486471935,
+      "loss": 3.5962,
+      "step": 4205
+    },
+    {
+      "epoch": 5.37888,
+      "grad_norm": 0.49401092529296875,
+      "learning_rate": 0.00014598196257908195,
+      "loss": 3.571,
+      "step": 4206
+    },
+    {
+      "epoch": 5.38016,
+      "grad_norm": 0.4905524253845215,
+      "learning_rate": 0.0001459415802934446,
+      "loss": 3.5594,
+      "step": 4207
+    },
+    {
+      "epoch": 5.38144,
+      "grad_norm": 0.5025436878204346,
+      "learning_rate": 0.00014590119800780724,
+      "loss": 3.5889,
+      "step": 4208
+    },
+    {
+      "epoch": 5.38272,
+      "grad_norm": 0.5062000155448914,
+      "learning_rate": 0.00014586081572216987,
+      "loss": 3.5641,
+      "step": 4209
+    },
+    {
+      "epoch": 5.384,
+      "grad_norm": 0.4925670027732849,
+      "learning_rate": 0.0001458204334365325,
+      "loss": 3.5346,
+      "step": 4210
+    },
+    {
+      "epoch": 5.38528,
+      "grad_norm": 0.5170202255249023,
+      "learning_rate": 0.00014578005115089513,
+      "loss": 3.5173,
+      "step": 4211
+    },
+    {
+      "epoch": 5.38656,
+      "grad_norm": 0.4998747706413269,
+      "learning_rate": 0.00014573966886525776,
+      "loss": 3.4946,
+      "step": 4212
+    },
+    {
+      "epoch": 5.38784,
+      "grad_norm": 0.5495707392692566,
+      "learning_rate": 0.0001456992865796204,
+      "loss": 3.6223,
+      "step": 4213
+    },
+    {
+      "epoch": 5.38912,
+      "grad_norm": 0.5236058831214905,
+      "learning_rate": 0.00014565890429398302,
+      "loss": 3.5723,
+      "step": 4214
+    },
+    {
+      "epoch": 5.3904,
+      "grad_norm": 0.49637115001678467,
+      "learning_rate": 0.00014561852200834565,
+      "loss": 3.5521,
+      "step": 4215
+    },
+    {
+      "epoch": 5.39168,
+      "grad_norm": 0.5067302584648132,
+      "learning_rate": 0.0001455781397227083,
+      "loss": 3.6365,
+      "step": 4216
+    },
+    {
+      "epoch": 5.39296,
+      "grad_norm": 0.5126550197601318,
+      "learning_rate": 0.00014553775743707094,
+      "loss": 3.5732,
+      "step": 4217
+    },
+    {
+      "epoch": 5.39424,
+      "grad_norm": 0.4854673743247986,
+      "learning_rate": 0.00014549737515143357,
+      "loss": 3.5872,
+      "step": 4218
+    },
+    {
+      "epoch": 5.39552,
+      "grad_norm": 0.5130731463432312,
+      "learning_rate": 0.0001454569928657962,
+      "loss": 3.5871,
+      "step": 4219
+    },
+    {
+      "epoch": 5.3968,
+      "grad_norm": 0.5053207278251648,
+      "learning_rate": 0.00014541661058015883,
+      "loss": 3.5029,
+      "step": 4220
+    },
+    {
+      "epoch": 5.39808,
+      "grad_norm": 0.5023342370986938,
+      "learning_rate": 0.00014537622829452146,
+      "loss": 3.633,
+      "step": 4221
+    },
+    {
+      "epoch": 5.39936,
+      "grad_norm": 0.4882518947124481,
+      "learning_rate": 0.00014533584600888409,
+      "loss": 3.5118,
+      "step": 4222
+    },
+    {
+      "epoch": 5.40064,
+      "grad_norm": 0.4898047149181366,
+      "learning_rate": 0.00014529546372324672,
+      "loss": 3.5909,
+      "step": 4223
+    },
+    {
+      "epoch": 5.40192,
+      "grad_norm": 0.4916389286518097,
+      "learning_rate": 0.00014525508143760935,
+      "loss": 3.5209,
+      "step": 4224
+    },
+    {
+      "epoch": 5.4032,
+      "grad_norm": 0.4937368631362915,
+      "learning_rate": 0.000145214699151972,
+      "loss": 3.5068,
+      "step": 4225
+    },
+    {
+      "epoch": 5.40448,
+      "grad_norm": 0.49129781126976013,
+      "learning_rate": 0.0001451743168663346,
+      "loss": 3.6266,
+      "step": 4226
+    },
+    {
+      "epoch": 5.40576,
+      "grad_norm": 0.5058757066726685,
+      "learning_rate": 0.00014513393458069726,
+      "loss": 3.5565,
+      "step": 4227
+    },
+    {
+      "epoch": 5.40704,
+      "grad_norm": 0.5102645754814148,
+      "learning_rate": 0.0001450935522950599,
+      "loss": 3.6193,
+      "step": 4228
+    },
+    {
+      "epoch": 5.40832,
+      "grad_norm": 0.5095773935317993,
+      "learning_rate": 0.00014505317000942252,
+      "loss": 3.5692,
+      "step": 4229
+    },
+    {
+      "epoch": 5.4096,
+      "grad_norm": 0.5054805278778076,
+      "learning_rate": 0.00014501278772378515,
+      "loss": 3.5361,
+      "step": 4230
+    },
+    {
+      "epoch": 5.41088,
+      "grad_norm": 0.4920346438884735,
+      "learning_rate": 0.00014497240543814778,
+      "loss": 3.5224,
+      "step": 4231
+    },
+    {
+      "epoch": 5.41216,
+      "grad_norm": 0.4970761239528656,
+      "learning_rate": 0.00014493202315251041,
+      "loss": 3.5793,
+      "step": 4232
+    },
+    {
+      "epoch": 5.41344,
+      "grad_norm": 0.502128005027771,
+      "learning_rate": 0.00014489164086687307,
+      "loss": 3.5549,
+      "step": 4233
+    },
+    {
+      "epoch": 5.41472,
+      "grad_norm": 0.4969402551651001,
+      "learning_rate": 0.00014485125858123567,
+      "loss": 3.5693,
+      "step": 4234
+    },
+    {
+      "epoch": 5.416,
+      "grad_norm": 0.4911673665046692,
+      "learning_rate": 0.0001448108762955983,
+      "loss": 3.5352,
+      "step": 4235
+    },
+    {
+      "epoch": 5.41728,
+      "grad_norm": 0.5134766101837158,
+      "learning_rate": 0.00014477049400996096,
+      "loss": 3.5124,
+      "step": 4236
+    },
+    {
+      "epoch": 5.41856,
+      "grad_norm": 0.5096125602722168,
+      "learning_rate": 0.00014473011172432356,
+      "loss": 3.5739,
+      "step": 4237
+    },
+    {
+      "epoch": 5.41984,
+      "grad_norm": 0.5185458064079285,
+      "learning_rate": 0.00014468972943868622,
+      "loss": 3.5937,
+      "step": 4238
+    },
+    {
+      "epoch": 5.42112,
+      "grad_norm": 0.5176596641540527,
+      "learning_rate": 0.00014464934715304885,
+      "loss": 3.5228,
+      "step": 4239
+    },
+    {
+      "epoch": 5.4224,
+      "grad_norm": 0.5246290564537048,
+      "learning_rate": 0.00014460896486741148,
+      "loss": 3.5686,
+      "step": 4240
+    },
+    {
+      "epoch": 5.42368,
+      "grad_norm": 0.5069420337677002,
+      "learning_rate": 0.0001445685825817741,
+      "loss": 3.5579,
+      "step": 4241
+    },
+    {
+      "epoch": 5.4249600000000004,
+      "grad_norm": 0.5116468667984009,
+      "learning_rate": 0.00014452820029613674,
+      "loss": 3.5462,
+      "step": 4242
+    },
+    {
+      "epoch": 5.42624,
+      "grad_norm": 0.5116104483604431,
+      "learning_rate": 0.00014448781801049937,
+      "loss": 3.587,
+      "step": 4243
+    },
+    {
+      "epoch": 5.42752,
+      "grad_norm": 0.5079525113105774,
+      "learning_rate": 0.00014444743572486203,
+      "loss": 3.5666,
+      "step": 4244
+    },
+    {
+      "epoch": 5.4288,
+      "grad_norm": 0.49015963077545166,
+      "learning_rate": 0.00014440705343922466,
+      "loss": 3.5803,
+      "step": 4245
+    },
+    {
+      "epoch": 5.43008,
+      "grad_norm": 0.5075419545173645,
+      "learning_rate": 0.0001443666711535873,
+      "loss": 3.5527,
+      "step": 4246
+    },
+    {
+      "epoch": 5.43136,
+      "grad_norm": 0.4922216534614563,
+      "learning_rate": 0.00014432628886794992,
+      "loss": 3.5723,
+      "step": 4247
+    },
+    {
+      "epoch": 5.43264,
+      "grad_norm": 0.5076229572296143,
+      "learning_rate": 0.00014428590658231255,
+      "loss": 3.5729,
+      "step": 4248
+    },
+    {
+      "epoch": 5.43392,
+      "grad_norm": 0.4939456284046173,
+      "learning_rate": 0.00014424552429667518,
+      "loss": 3.4065,
+      "step": 4249
+    },
+    {
+      "epoch": 5.4352,
+      "grad_norm": 0.5079296827316284,
+      "learning_rate": 0.0001442051420110378,
+      "loss": 3.6077,
+      "step": 4250
+    },
+    {
+      "epoch": 5.4364799999999995,
+      "grad_norm": 0.4930500090122223,
+      "learning_rate": 0.00014416475972540044,
+      "loss": 3.554,
+      "step": 4251
+    },
+    {
+      "epoch": 5.43776,
+      "grad_norm": 0.5013241767883301,
+      "learning_rate": 0.00014412437743976307,
+      "loss": 3.5551,
+      "step": 4252
+    },
+    {
+      "epoch": 5.43904,
+      "grad_norm": 0.5022532939910889,
+      "learning_rate": 0.00014408399515412573,
+      "loss": 3.5839,
+      "step": 4253
+    },
+    {
+      "epoch": 5.44032,
+      "grad_norm": 0.5082839727401733,
+      "learning_rate": 0.00014404361286848833,
+      "loss": 3.6063,
+      "step": 4254
+    },
+    {
+      "epoch": 5.4416,
+      "grad_norm": 0.5208956599235535,
+      "learning_rate": 0.000144003230582851,
+      "loss": 3.4991,
+      "step": 4255
+    },
+    {
+      "epoch": 5.44288,
+      "grad_norm": 0.5048259496688843,
+      "learning_rate": 0.00014396284829721362,
+      "loss": 3.5349,
+      "step": 4256
+    },
+    {
+      "epoch": 5.44416,
+      "grad_norm": 0.49307942390441895,
+      "learning_rate": 0.00014392246601157625,
+      "loss": 3.5114,
+      "step": 4257
+    },
+    {
+      "epoch": 5.44544,
+      "grad_norm": 0.5296189188957214,
+      "learning_rate": 0.00014388208372593888,
+      "loss": 3.5802,
+      "step": 4258
+    },
+    {
+      "epoch": 5.44672,
+      "grad_norm": 0.5138427019119263,
+      "learning_rate": 0.0001438417014403015,
+      "loss": 3.5841,
+      "step": 4259
+    },
+    {
+      "epoch": 5.448,
+      "grad_norm": 0.4858483672142029,
+      "learning_rate": 0.00014380131915466414,
+      "loss": 3.5527,
+      "step": 4260
+    },
+    {
+      "epoch": 5.44928,
+      "grad_norm": 0.5035324692726135,
+      "learning_rate": 0.00014376093686902677,
+      "loss": 3.4928,
+      "step": 4261
+    },
+    {
+      "epoch": 5.45056,
+      "grad_norm": 0.49650654196739197,
+      "learning_rate": 0.0001437205545833894,
+      "loss": 3.631,
+      "step": 4262
+    },
+    {
+      "epoch": 5.45184,
+      "grad_norm": 0.5007782578468323,
+      "learning_rate": 0.00014368017229775203,
+      "loss": 3.5251,
+      "step": 4263
+    },
+    {
+      "epoch": 5.45312,
+      "grad_norm": 0.4857237637042999,
+      "learning_rate": 0.00014363979001211468,
+      "loss": 3.5328,
+      "step": 4264
+    },
+    {
+      "epoch": 5.4544,
+      "grad_norm": 0.5186595320701599,
+      "learning_rate": 0.0001435994077264773,
+      "loss": 3.5839,
+      "step": 4265
+    },
+    {
+      "epoch": 5.45568,
+      "grad_norm": 0.48634448647499084,
+      "learning_rate": 0.00014355902544083994,
+      "loss": 3.5082,
+      "step": 4266
+    },
+    {
+      "epoch": 5.45696,
+      "grad_norm": 0.5036529302597046,
+      "learning_rate": 0.00014351864315520257,
+      "loss": 3.6215,
+      "step": 4267
+    },
+    {
+      "epoch": 5.45824,
+      "grad_norm": 0.49025043845176697,
+      "learning_rate": 0.0001434782608695652,
+      "loss": 3.5371,
+      "step": 4268
+    },
+    {
+      "epoch": 5.45952,
+      "grad_norm": 0.5002211928367615,
+      "learning_rate": 0.00014343787858392783,
+      "loss": 3.5434,
+      "step": 4269
+    },
+    {
+      "epoch": 5.4608,
+      "grad_norm": 0.48219621181488037,
+      "learning_rate": 0.00014339749629829046,
+      "loss": 3.567,
+      "step": 4270
+    },
+    {
+      "epoch": 5.46208,
+      "grad_norm": 0.4965188801288605,
+      "learning_rate": 0.0001433571140126531,
+      "loss": 3.5367,
+      "step": 4271
+    },
+    {
+      "epoch": 5.46336,
+      "grad_norm": 0.4985227584838867,
+      "learning_rate": 0.00014331673172701575,
+      "loss": 3.4608,
+      "step": 4272
+    },
+    {
+      "epoch": 5.46464,
+      "grad_norm": 0.4829920828342438,
+      "learning_rate": 0.00014327634944137836,
+      "loss": 3.5967,
+      "step": 4273
+    },
+    {
+      "epoch": 5.46592,
+      "grad_norm": 0.49400630593299866,
+      "learning_rate": 0.000143235967155741,
+      "loss": 3.5266,
+      "step": 4274
+    },
+    {
+      "epoch": 5.4672,
+      "grad_norm": 0.5178740620613098,
+      "learning_rate": 0.00014319558487010364,
+      "loss": 3.5871,
+      "step": 4275
+    },
+    {
+      "epoch": 5.46848,
+      "grad_norm": 0.49614524841308594,
+      "learning_rate": 0.00014315520258446627,
+      "loss": 3.6358,
+      "step": 4276
+    },
+    {
+      "epoch": 5.46976,
+      "grad_norm": 0.48628175258636475,
+      "learning_rate": 0.0001431148202988289,
+      "loss": 3.4888,
+      "step": 4277
+    },
+    {
+      "epoch": 5.47104,
+      "grad_norm": 0.5142548680305481,
+      "learning_rate": 0.00014307443801319153,
+      "loss": 3.5692,
+      "step": 4278
+    },
+    {
+      "epoch": 5.47232,
+      "grad_norm": 0.5002409219741821,
+      "learning_rate": 0.00014303405572755416,
+      "loss": 3.6088,
+      "step": 4279
+    },
+    {
+      "epoch": 5.4736,
+      "grad_norm": 0.5171060562133789,
+      "learning_rate": 0.0001429936734419168,
+      "loss": 3.5909,
+      "step": 4280
+    },
+    {
+      "epoch": 5.47488,
+      "grad_norm": 0.5077352523803711,
+      "learning_rate": 0.00014295329115627945,
+      "loss": 3.655,
+      "step": 4281
+    },
+    {
+      "epoch": 5.47616,
+      "grad_norm": 0.4823010563850403,
+      "learning_rate": 0.00014291290887064205,
+      "loss": 3.5184,
+      "step": 4282
+    },
+    {
+      "epoch": 5.47744,
+      "grad_norm": 0.5208946466445923,
+      "learning_rate": 0.0001428725265850047,
+      "loss": 3.6297,
+      "step": 4283
+    },
+    {
+      "epoch": 5.47872,
+      "grad_norm": 0.5040287375450134,
+      "learning_rate": 0.00014283214429936734,
+      "loss": 3.5483,
+      "step": 4284
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.5379658341407776,
+      "learning_rate": 0.00014279176201372997,
+      "loss": 3.5732,
+      "step": 4285
+    },
+    {
+      "epoch": 5.48128,
+      "grad_norm": 0.4982692003250122,
+      "learning_rate": 0.0001427513797280926,
+      "loss": 3.5691,
+      "step": 4286
+    },
+    {
+      "epoch": 5.48256,
+      "grad_norm": 0.5001688003540039,
+      "learning_rate": 0.00014271099744245523,
+      "loss": 3.6635,
+      "step": 4287
+    },
+    {
+      "epoch": 5.48384,
+      "grad_norm": 0.4979035258293152,
+      "learning_rate": 0.00014267061515681786,
+      "loss": 3.582,
+      "step": 4288
+    },
+    {
+      "epoch": 5.48512,
+      "grad_norm": 0.49051418900489807,
+      "learning_rate": 0.0001426302328711805,
+      "loss": 3.5999,
+      "step": 4289
+    },
+    {
+      "epoch": 5.4864,
+      "grad_norm": 0.4869391918182373,
+      "learning_rate": 0.00014258985058554312,
+      "loss": 3.5503,
+      "step": 4290
+    },
+    {
+      "epoch": 5.48768,
+      "grad_norm": 0.5105782747268677,
+      "learning_rate": 0.00014254946829990575,
+      "loss": 3.6143,
+      "step": 4291
+    },
+    {
+      "epoch": 5.48896,
+      "grad_norm": 0.4911419451236725,
+      "learning_rate": 0.0001425090860142684,
+      "loss": 3.5467,
+      "step": 4292
+    },
+    {
+      "epoch": 5.49024,
+      "grad_norm": 0.4994243085384369,
+      "learning_rate": 0.000142468703728631,
+      "loss": 3.6287,
+      "step": 4293
+    },
+    {
+      "epoch": 5.49152,
+      "grad_norm": 0.4881093502044678,
+      "learning_rate": 0.00014242832144299367,
+      "loss": 3.5324,
+      "step": 4294
+    },
+    {
+      "epoch": 5.4928,
+      "grad_norm": 0.4933335781097412,
+      "learning_rate": 0.0001423879391573563,
+      "loss": 3.5793,
+      "step": 4295
+    },
+    {
+      "epoch": 5.49408,
+      "grad_norm": 0.4933607876300812,
+      "learning_rate": 0.00014234755687171893,
+      "loss": 3.5533,
+      "step": 4296
+    },
+    {
+      "epoch": 5.49536,
+      "grad_norm": 0.4941942095756531,
+      "learning_rate": 0.00014230717458608156,
+      "loss": 3.5883,
+      "step": 4297
+    },
+    {
+      "epoch": 5.49664,
+      "grad_norm": 0.5198619961738586,
+      "learning_rate": 0.0001422667923004442,
+      "loss": 3.667,
+      "step": 4298
+    },
+    {
+      "epoch": 5.49792,
+      "grad_norm": 0.4889330267906189,
+      "learning_rate": 0.00014222641001480682,
+      "loss": 3.5658,
+      "step": 4299
+    },
+    {
+      "epoch": 5.4992,
+      "grad_norm": 0.48626118898391724,
+      "learning_rate": 0.00014218602772916948,
+      "loss": 3.5525,
+      "step": 4300
+    },
+    {
+      "epoch": 5.50048,
+      "grad_norm": 0.4914554953575134,
+      "learning_rate": 0.00014214564544353208,
+      "loss": 3.5968,
+      "step": 4301
+    },
+    {
+      "epoch": 5.50176,
+      "grad_norm": 0.49524810910224915,
+      "learning_rate": 0.0001421052631578947,
+      "loss": 3.5555,
+      "step": 4302
+    },
+    {
+      "epoch": 5.50304,
+      "grad_norm": 0.5117614269256592,
+      "learning_rate": 0.00014206488087225737,
+      "loss": 3.6015,
+      "step": 4303
+    },
+    {
+      "epoch": 5.50432,
+      "grad_norm": 0.4925834834575653,
+      "learning_rate": 0.00014202449858662,
+      "loss": 3.5378,
+      "step": 4304
+    },
+    {
+      "epoch": 5.5056,
+      "grad_norm": 0.494954913854599,
+      "learning_rate": 0.00014198411630098263,
+      "loss": 3.5463,
+      "step": 4305
+    },
+    {
+      "epoch": 5.50688,
+      "grad_norm": 0.5056374073028564,
+      "learning_rate": 0.00014194373401534526,
+      "loss": 3.6499,
+      "step": 4306
+    },
+    {
+      "epoch": 5.50816,
+      "grad_norm": 0.495236873626709,
+      "learning_rate": 0.00014190335172970789,
+      "loss": 3.5903,
+      "step": 4307
+    },
+    {
+      "epoch": 5.50944,
+      "grad_norm": 0.48822644352912903,
+      "learning_rate": 0.00014186296944407052,
+      "loss": 3.5119,
+      "step": 4308
+    },
+    {
+      "epoch": 5.51072,
+      "grad_norm": 0.5125917792320251,
+      "learning_rate": 0.00014182258715843315,
+      "loss": 3.5481,
+      "step": 4309
+    },
+    {
+      "epoch": 5.5120000000000005,
+      "grad_norm": 0.5041484832763672,
+      "learning_rate": 0.00014178220487279578,
+      "loss": 3.542,
+      "step": 4310
+    },
+    {
+      "epoch": 5.51328,
+      "grad_norm": 0.48397096991539,
+      "learning_rate": 0.00014174182258715843,
+      "loss": 3.5553,
+      "step": 4311
+    },
+    {
+      "epoch": 5.51456,
+      "grad_norm": 0.49939045310020447,
+      "learning_rate": 0.00014170144030152106,
+      "loss": 3.6071,
+      "step": 4312
+    },
+    {
+      "epoch": 5.51584,
+      "grad_norm": 0.4981886148452759,
+      "learning_rate": 0.0001416610580158837,
+      "loss": 3.6154,
+      "step": 4313
+    },
+    {
+      "epoch": 5.51712,
+      "grad_norm": 0.4908769130706787,
+      "learning_rate": 0.00014162067573024632,
+      "loss": 3.5811,
+      "step": 4314
+    },
+    {
+      "epoch": 5.5184,
+      "grad_norm": 0.4873631000518799,
+      "learning_rate": 0.00014158029344460895,
+      "loss": 3.5451,
+      "step": 4315
+    },
+    {
+      "epoch": 5.51968,
+      "grad_norm": 0.5031121969223022,
+      "learning_rate": 0.00014153991115897158,
+      "loss": 3.4982,
+      "step": 4316
+    },
+    {
+      "epoch": 5.52096,
+      "grad_norm": 0.5029070973396301,
+      "learning_rate": 0.00014149952887333421,
+      "loss": 3.5159,
+      "step": 4317
+    },
+    {
+      "epoch": 5.52224,
+      "grad_norm": 0.5104166865348816,
+      "learning_rate": 0.00014145914658769684,
+      "loss": 3.5713,
+      "step": 4318
+    },
+    {
+      "epoch": 5.5235199999999995,
+      "grad_norm": 0.512605607509613,
+      "learning_rate": 0.00014141876430205947,
+      "loss": 3.5637,
+      "step": 4319
+    },
+    {
+      "epoch": 5.5248,
+      "grad_norm": 0.5072982311248779,
+      "learning_rate": 0.00014137838201642213,
+      "loss": 3.5527,
+      "step": 4320
+    },
+    {
+      "epoch": 5.52608,
+      "grad_norm": 0.5111182332038879,
+      "learning_rate": 0.00014133799973078473,
+      "loss": 3.6156,
+      "step": 4321
+    },
+    {
+      "epoch": 5.52736,
+      "grad_norm": 0.5078898668289185,
+      "learning_rate": 0.0001412976174451474,
+      "loss": 3.5136,
+      "step": 4322
+    },
+    {
+      "epoch": 5.52864,
+      "grad_norm": 0.49358004331588745,
+      "learning_rate": 0.00014125723515951002,
+      "loss": 3.481,
+      "step": 4323
+    },
+    {
+      "epoch": 5.52992,
+      "grad_norm": 0.4940800070762634,
+      "learning_rate": 0.00014121685287387265,
+      "loss": 3.5878,
+      "step": 4324
+    },
+    {
+      "epoch": 5.5312,
+      "grad_norm": 0.4859899580478668,
+      "learning_rate": 0.00014117647058823528,
+      "loss": 3.5579,
+      "step": 4325
+    },
+    {
+      "epoch": 5.53248,
+      "grad_norm": 0.5090684294700623,
+      "learning_rate": 0.0001411360883025979,
+      "loss": 3.5731,
+      "step": 4326
+    },
+    {
+      "epoch": 5.53376,
+      "grad_norm": 0.5048773884773254,
+      "learning_rate": 0.00014109570601696054,
+      "loss": 3.6074,
+      "step": 4327
+    },
+    {
+      "epoch": 5.53504,
+      "grad_norm": 0.5031253099441528,
+      "learning_rate": 0.0001410553237313232,
+      "loss": 3.554,
+      "step": 4328
+    },
+    {
+      "epoch": 5.53632,
+      "grad_norm": 0.48305338621139526,
+      "learning_rate": 0.0001410149414456858,
+      "loss": 3.5785,
+      "step": 4329
+    },
+    {
+      "epoch": 5.5376,
+      "grad_norm": 0.5137234330177307,
+      "learning_rate": 0.00014097455916004843,
+      "loss": 3.5755,
+      "step": 4330
+    },
+    {
+      "epoch": 5.53888,
+      "grad_norm": 0.4992203414440155,
+      "learning_rate": 0.0001409341768744111,
+      "loss": 3.5376,
+      "step": 4331
+    },
+    {
+      "epoch": 5.54016,
+      "grad_norm": 0.5001314878463745,
+      "learning_rate": 0.00014089379458877372,
+      "loss": 3.6757,
+      "step": 4332
+    },
+    {
+      "epoch": 5.54144,
+      "grad_norm": 0.4925899803638458,
+      "learning_rate": 0.00014085341230313635,
+      "loss": 3.5308,
+      "step": 4333
+    },
+    {
+      "epoch": 5.54272,
+      "grad_norm": 0.5240253210067749,
+      "learning_rate": 0.00014081303001749898,
+      "loss": 3.539,
+      "step": 4334
+    },
+    {
+      "epoch": 5.5440000000000005,
+      "grad_norm": 0.5012069940567017,
+      "learning_rate": 0.0001407726477318616,
+      "loss": 3.5401,
+      "step": 4335
+    },
+    {
+      "epoch": 5.54528,
+      "grad_norm": 0.4972357451915741,
+      "learning_rate": 0.00014073226544622424,
+      "loss": 3.5214,
+      "step": 4336
+    },
+    {
+      "epoch": 5.54656,
+      "grad_norm": 0.5054651498794556,
+      "learning_rate": 0.00014069188316058687,
+      "loss": 3.4729,
+      "step": 4337
+    },
+    {
+      "epoch": 5.54784,
+      "grad_norm": 0.4873622953891754,
+      "learning_rate": 0.0001406515008749495,
+      "loss": 3.5886,
+      "step": 4338
+    },
+    {
+      "epoch": 5.54912,
+      "grad_norm": 0.5032194256782532,
+      "learning_rate": 0.00014061111858931216,
+      "loss": 3.4718,
+      "step": 4339
+    },
+    {
+      "epoch": 5.5504,
+      "grad_norm": 0.5260450839996338,
+      "learning_rate": 0.0001405707363036748,
+      "loss": 3.5285,
+      "step": 4340
+    },
+    {
+      "epoch": 5.55168,
+      "grad_norm": 0.4904164671897888,
+      "learning_rate": 0.00014053035401803742,
+      "loss": 3.5785,
+      "step": 4341
+    },
+    {
+      "epoch": 5.55296,
+      "grad_norm": 0.49703943729400635,
+      "learning_rate": 0.00014048997173240005,
+      "loss": 3.5338,
+      "step": 4342
+    },
+    {
+      "epoch": 5.55424,
+      "grad_norm": 0.50665283203125,
+      "learning_rate": 0.00014044958944676268,
+      "loss": 3.5355,
+      "step": 4343
+    },
+    {
+      "epoch": 5.55552,
+      "grad_norm": 0.5014834403991699,
+      "learning_rate": 0.0001404092071611253,
+      "loss": 3.5493,
+      "step": 4344
+    },
+    {
+      "epoch": 5.5568,
+      "grad_norm": 0.5009499788284302,
+      "learning_rate": 0.00014036882487548794,
+      "loss": 3.6308,
+      "step": 4345
+    },
+    {
+      "epoch": 5.55808,
+      "grad_norm": 0.4850377142429352,
+      "learning_rate": 0.00014032844258985057,
+      "loss": 3.6086,
+      "step": 4346
+    },
+    {
+      "epoch": 5.55936,
+      "grad_norm": 0.5173080563545227,
+      "learning_rate": 0.0001402880603042132,
+      "loss": 3.5599,
+      "step": 4347
+    },
+    {
+      "epoch": 5.56064,
+      "grad_norm": 0.5030738711357117,
+      "learning_rate": 0.00014024767801857585,
+      "loss": 3.618,
+      "step": 4348
+    },
+    {
+      "epoch": 5.56192,
+      "grad_norm": 0.5130670666694641,
+      "learning_rate": 0.00014020729573293846,
+      "loss": 3.5968,
+      "step": 4349
+    },
+    {
+      "epoch": 5.5632,
+      "grad_norm": 0.5004739761352539,
+      "learning_rate": 0.00014016691344730111,
+      "loss": 3.5825,
+      "step": 4350
+    },
+    {
+      "epoch": 5.56448,
+      "grad_norm": 0.5110647082328796,
+      "learning_rate": 0.00014012653116166375,
+      "loss": 3.5971,
+      "step": 4351
+    },
+    {
+      "epoch": 5.56576,
+      "grad_norm": 0.4908808767795563,
+      "learning_rate": 0.00014008614887602638,
+      "loss": 3.5662,
+      "step": 4352
+    },
+    {
+      "epoch": 5.56704,
+      "grad_norm": 0.49879395961761475,
+      "learning_rate": 0.000140045766590389,
+      "loss": 3.5705,
+      "step": 4353
+    },
+    {
+      "epoch": 5.56832,
+      "grad_norm": 0.5124648809432983,
+      "learning_rate": 0.00014000538430475164,
+      "loss": 3.618,
+      "step": 4354
+    },
+    {
+      "epoch": 5.5696,
+      "grad_norm": 0.48847079277038574,
+      "learning_rate": 0.00013996500201911427,
+      "loss": 3.5621,
+      "step": 4355
+    },
+    {
+      "epoch": 5.57088,
+      "grad_norm": 0.5030008554458618,
+      "learning_rate": 0.0001399246197334769,
+      "loss": 3.4727,
+      "step": 4356
+    },
+    {
+      "epoch": 5.57216,
+      "grad_norm": 0.5004321932792664,
+      "learning_rate": 0.00013988423744783953,
+      "loss": 3.5718,
+      "step": 4357
+    },
+    {
+      "epoch": 5.57344,
+      "grad_norm": 0.49990010261535645,
+      "learning_rate": 0.00013984385516220216,
+      "loss": 3.5607,
+      "step": 4358
+    },
+    {
+      "epoch": 5.57472,
+      "grad_norm": 0.4781895577907562,
+      "learning_rate": 0.0001398034728765648,
+      "loss": 3.5735,
+      "step": 4359
+    },
+    {
+      "epoch": 5.576,
+      "grad_norm": 0.4892998933792114,
+      "learning_rate": 0.00013976309059092742,
+      "loss": 3.6314,
+      "step": 4360
+    },
+    {
+      "epoch": 5.57728,
+      "grad_norm": 0.5173603296279907,
+      "learning_rate": 0.00013972270830529007,
+      "loss": 3.5335,
+      "step": 4361
+    },
+    {
+      "epoch": 5.5785599999999995,
+      "grad_norm": 0.4709049463272095,
+      "learning_rate": 0.0001396823260196527,
+      "loss": 3.5218,
+      "step": 4362
+    },
+    {
+      "epoch": 5.57984,
+      "grad_norm": 0.5281604528427124,
+      "learning_rate": 0.00013964194373401533,
+      "loss": 3.589,
+      "step": 4363
+    },
+    {
+      "epoch": 5.58112,
+      "grad_norm": 0.49531951546669006,
+      "learning_rate": 0.00013960156144837796,
+      "loss": 3.5535,
+      "step": 4364
+    },
+    {
+      "epoch": 5.5824,
+      "grad_norm": 0.5021586418151855,
+      "learning_rate": 0.0001395611791627406,
+      "loss": 3.461,
+      "step": 4365
+    },
+    {
+      "epoch": 5.58368,
+      "grad_norm": 0.5123895406723022,
+      "learning_rate": 0.00013952079687710322,
+      "loss": 3.4411,
+      "step": 4366
+    },
+    {
+      "epoch": 5.58496,
+      "grad_norm": 0.515090823173523,
+      "learning_rate": 0.00013948041459146588,
+      "loss": 3.6073,
+      "step": 4367
+    },
+    {
+      "epoch": 5.58624,
+      "grad_norm": 0.5333694815635681,
+      "learning_rate": 0.0001394400323058285,
+      "loss": 3.6164,
+      "step": 4368
+    },
+    {
+      "epoch": 5.58752,
+      "grad_norm": 0.48728322982788086,
+      "learning_rate": 0.0001393996500201911,
+      "loss": 3.529,
+      "step": 4369
+    },
+    {
+      "epoch": 5.5888,
+      "grad_norm": 0.5210532546043396,
+      "learning_rate": 0.00013935926773455377,
+      "loss": 3.455,
+      "step": 4370
+    },
+    {
+      "epoch": 5.59008,
+      "grad_norm": 0.5110505223274231,
+      "learning_rate": 0.0001393188854489164,
+      "loss": 3.5412,
+      "step": 4371
+    },
+    {
+      "epoch": 5.59136,
+      "grad_norm": 0.4856546223163605,
+      "learning_rate": 0.00013927850316327903,
+      "loss": 3.5445,
+      "step": 4372
+    },
+    {
+      "epoch": 5.59264,
+      "grad_norm": 0.5165418386459351,
+      "learning_rate": 0.00013923812087764166,
+      "loss": 3.5215,
+      "step": 4373
+    },
+    {
+      "epoch": 5.59392,
+      "grad_norm": 0.524249255657196,
+      "learning_rate": 0.0001391977385920043,
+      "loss": 3.579,
+      "step": 4374
+    },
+    {
+      "epoch": 5.5952,
+      "grad_norm": 0.5186324715614319,
+      "learning_rate": 0.00013915735630636692,
+      "loss": 3.5448,
+      "step": 4375
+    },
+    {
+      "epoch": 5.59648,
+      "grad_norm": 0.4949183166027069,
+      "learning_rate": 0.00013911697402072958,
+      "loss": 3.5652,
+      "step": 4376
+    },
+    {
+      "epoch": 5.59776,
+      "grad_norm": 0.5009962916374207,
+      "learning_rate": 0.00013907659173509218,
+      "loss": 3.6081,
+      "step": 4377
+    },
+    {
+      "epoch": 5.5990400000000005,
+      "grad_norm": 0.4920154809951782,
+      "learning_rate": 0.00013903620944945484,
+      "loss": 3.5249,
+      "step": 4378
+    },
+    {
+      "epoch": 5.60032,
+      "grad_norm": 0.5058812499046326,
+      "learning_rate": 0.00013899582716381747,
+      "loss": 3.612,
+      "step": 4379
+    },
+    {
+      "epoch": 5.6016,
+      "grad_norm": 0.5027937889099121,
+      "learning_rate": 0.0001389554448781801,
+      "loss": 3.5013,
+      "step": 4380
+    },
+    {
+      "epoch": 5.60288,
+      "grad_norm": 0.5067436099052429,
+      "learning_rate": 0.00013891506259254273,
+      "loss": 3.5146,
+      "step": 4381
+    },
+    {
+      "epoch": 5.60416,
+      "grad_norm": 0.49127253890037537,
+      "learning_rate": 0.00013887468030690536,
+      "loss": 3.6259,
+      "step": 4382
+    },
+    {
+      "epoch": 5.60544,
+      "grad_norm": 0.5145688056945801,
+      "learning_rate": 0.000138834298021268,
+      "loss": 3.6815,
+      "step": 4383
+    },
+    {
+      "epoch": 5.60672,
+      "grad_norm": 0.5067407488822937,
+      "learning_rate": 0.00013879391573563062,
+      "loss": 3.6327,
+      "step": 4384
+    },
+    {
+      "epoch": 5.608,
+      "grad_norm": 0.5062004923820496,
+      "learning_rate": 0.00013875353344999325,
+      "loss": 3.5035,
+      "step": 4385
+    },
+    {
+      "epoch": 5.60928,
+      "grad_norm": 0.483181893825531,
+      "learning_rate": 0.00013871315116435588,
+      "loss": 3.5152,
+      "step": 4386
+    },
+    {
+      "epoch": 5.6105599999999995,
+      "grad_norm": 0.5145739912986755,
+      "learning_rate": 0.00013867276887871854,
+      "loss": 3.572,
+      "step": 4387
+    },
+    {
+      "epoch": 5.61184,
+      "grad_norm": 0.4981113374233246,
+      "learning_rate": 0.00013863238659308114,
+      "loss": 3.5428,
+      "step": 4388
+    },
+    {
+      "epoch": 5.61312,
+      "grad_norm": 0.4911549389362335,
+      "learning_rate": 0.0001385920043074438,
+      "loss": 3.5261,
+      "step": 4389
+    },
+    {
+      "epoch": 5.6144,
+      "grad_norm": 0.5160556435585022,
+      "learning_rate": 0.00013855162202180643,
+      "loss": 3.5217,
+      "step": 4390
+    },
+    {
+      "epoch": 5.61568,
+      "grad_norm": 0.4965737760066986,
+      "learning_rate": 0.00013851123973616906,
+      "loss": 3.5228,
+      "step": 4391
+    },
+    {
+      "epoch": 5.61696,
+      "grad_norm": 0.49209117889404297,
+      "learning_rate": 0.0001384708574505317,
+      "loss": 3.5664,
+      "step": 4392
+    },
+    {
+      "epoch": 5.61824,
+      "grad_norm": 0.4959764778614044,
+      "learning_rate": 0.00013843047516489432,
+      "loss": 3.6293,
+      "step": 4393
+    },
+    {
+      "epoch": 5.61952,
+      "grad_norm": 0.5043753981590271,
+      "learning_rate": 0.00013839009287925695,
+      "loss": 3.4969,
+      "step": 4394
+    },
+    {
+      "epoch": 5.6208,
+      "grad_norm": 0.5141623020172119,
+      "learning_rate": 0.0001383497105936196,
+      "loss": 3.5033,
+      "step": 4395
+    },
+    {
+      "epoch": 5.62208,
+      "grad_norm": 0.4955049753189087,
+      "learning_rate": 0.0001383093283079822,
+      "loss": 3.5877,
+      "step": 4396
+    },
+    {
+      "epoch": 5.62336,
+      "grad_norm": 0.4988468289375305,
+      "learning_rate": 0.00013826894602234484,
+      "loss": 3.5641,
+      "step": 4397
+    },
+    {
+      "epoch": 5.62464,
+      "grad_norm": 0.49312546849250793,
+      "learning_rate": 0.0001382285637367075,
+      "loss": 3.5163,
+      "step": 4398
+    },
+    {
+      "epoch": 5.62592,
+      "grad_norm": 0.49335968494415283,
+      "learning_rate": 0.00013818818145107012,
+      "loss": 3.5558,
+      "step": 4399
+    },
+    {
+      "epoch": 5.6272,
+      "grad_norm": 0.49063315987586975,
+      "learning_rate": 0.00013814779916543275,
+      "loss": 3.6151,
+      "step": 4400
+    },
+    {
+      "epoch": 5.62848,
+      "grad_norm": 0.48111316561698914,
+      "learning_rate": 0.00013810741687979538,
+      "loss": 3.5394,
+      "step": 4401
+    },
+    {
+      "epoch": 5.62976,
+      "grad_norm": 0.4941888153553009,
+      "learning_rate": 0.00013806703459415801,
+      "loss": 3.5944,
+      "step": 4402
+    },
+    {
+      "epoch": 5.6310400000000005,
+      "grad_norm": 0.5081154704093933,
+      "learning_rate": 0.00013802665230852064,
+      "loss": 3.546,
+      "step": 4403
+    },
+    {
+      "epoch": 5.63232,
+      "grad_norm": 0.49409082531929016,
+      "learning_rate": 0.0001379862700228833,
+      "loss": 3.6016,
+      "step": 4404
+    },
+    {
+      "epoch": 5.6336,
+      "grad_norm": 0.4921821355819702,
+      "learning_rate": 0.0001379458877372459,
+      "loss": 3.56,
+      "step": 4405
+    },
+    {
+      "epoch": 5.63488,
+      "grad_norm": 0.5267976522445679,
+      "learning_rate": 0.00013790550545160856,
+      "loss": 3.5694,
+      "step": 4406
+    },
+    {
+      "epoch": 5.63616,
+      "grad_norm": 0.4959132671356201,
+      "learning_rate": 0.0001378651231659712,
+      "loss": 3.5944,
+      "step": 4407
+    },
+    {
+      "epoch": 5.63744,
+      "grad_norm": 0.4983578622341156,
+      "learning_rate": 0.00013782474088033382,
+      "loss": 3.514,
+      "step": 4408
+    },
+    {
+      "epoch": 5.63872,
+      "grad_norm": 0.5071173310279846,
+      "learning_rate": 0.00013778435859469645,
+      "loss": 3.5079,
+      "step": 4409
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.4979630410671234,
+      "learning_rate": 0.00013774397630905908,
+      "loss": 3.5649,
+      "step": 4410
+    },
+    {
+      "epoch": 5.64128,
+      "grad_norm": 0.48886778950691223,
+      "learning_rate": 0.0001377035940234217,
+      "loss": 3.5758,
+      "step": 4411
+    },
+    {
+      "epoch": 5.64256,
+      "grad_norm": 0.5049862265586853,
+      "learning_rate": 0.00013766321173778434,
+      "loss": 3.5894,
+      "step": 4412
+    },
+    {
+      "epoch": 5.64384,
+      "grad_norm": 0.4826545715332031,
+      "learning_rate": 0.00013762282945214697,
+      "loss": 3.6285,
+      "step": 4413
+    },
+    {
+      "epoch": 5.64512,
+      "grad_norm": 0.5009142160415649,
+      "learning_rate": 0.0001375824471665096,
+      "loss": 3.5911,
+      "step": 4414
+    },
+    {
+      "epoch": 5.6464,
+      "grad_norm": 0.49813807010650635,
+      "learning_rate": 0.00013754206488087226,
+      "loss": 3.5388,
+      "step": 4415
+    },
+    {
+      "epoch": 5.64768,
+      "grad_norm": 0.48648300766944885,
+      "learning_rate": 0.00013750168259523486,
+      "loss": 3.5231,
+      "step": 4416
+    },
+    {
+      "epoch": 5.64896,
+      "grad_norm": 0.5059703588485718,
+      "learning_rate": 0.00013746130030959752,
+      "loss": 3.5606,
+      "step": 4417
+    },
+    {
+      "epoch": 5.65024,
+      "grad_norm": 0.49821940064430237,
+      "learning_rate": 0.00013742091802396015,
+      "loss": 3.5415,
+      "step": 4418
+    },
+    {
+      "epoch": 5.65152,
+      "grad_norm": 0.5076292157173157,
+      "learning_rate": 0.00013738053573832278,
+      "loss": 3.5893,
+      "step": 4419
+    },
+    {
+      "epoch": 5.6528,
+      "grad_norm": 0.5033696293830872,
+      "learning_rate": 0.0001373401534526854,
+      "loss": 3.5241,
+      "step": 4420
+    },
+    {
+      "epoch": 5.65408,
+      "grad_norm": 0.49491533637046814,
+      "learning_rate": 0.00013729977116704804,
+      "loss": 3.5782,
+      "step": 4421
+    },
+    {
+      "epoch": 5.65536,
+      "grad_norm": 0.5117203593254089,
+      "learning_rate": 0.00013725938888141067,
+      "loss": 3.5175,
+      "step": 4422
+    },
+    {
+      "epoch": 5.65664,
+      "grad_norm": 0.5070961117744446,
+      "learning_rate": 0.0001372190065957733,
+      "loss": 3.5301,
+      "step": 4423
+    },
+    {
+      "epoch": 5.65792,
+      "grad_norm": 0.5147043466567993,
+      "learning_rate": 0.00013717862431013593,
+      "loss": 3.5733,
+      "step": 4424
+    },
+    {
+      "epoch": 5.6592,
+      "grad_norm": 0.5103664398193359,
+      "learning_rate": 0.00013713824202449856,
+      "loss": 3.6548,
+      "step": 4425
+    },
+    {
+      "epoch": 5.66048,
+      "grad_norm": 0.4940325915813446,
+      "learning_rate": 0.00013709785973886122,
+      "loss": 3.6025,
+      "step": 4426
+    },
+    {
+      "epoch": 5.66176,
+      "grad_norm": 0.49647340178489685,
+      "learning_rate": 0.00013705747745322385,
+      "loss": 3.4952,
+      "step": 4427
+    },
+    {
+      "epoch": 5.66304,
+      "grad_norm": 0.5108480453491211,
+      "learning_rate": 0.00013701709516758648,
+      "loss": 3.6175,
+      "step": 4428
+    },
+    {
+      "epoch": 5.66432,
+      "grad_norm": 0.5075586438179016,
+      "learning_rate": 0.0001369767128819491,
+      "loss": 3.5372,
+      "step": 4429
+    },
+    {
+      "epoch": 5.6655999999999995,
+      "grad_norm": 0.49585655331611633,
+      "learning_rate": 0.00013693633059631174,
+      "loss": 3.5006,
+      "step": 4430
+    },
+    {
+      "epoch": 5.66688,
+      "grad_norm": 0.5268029570579529,
+      "learning_rate": 0.00013689594831067437,
+      "loss": 3.5752,
+      "step": 4431
+    },
+    {
+      "epoch": 5.66816,
+      "grad_norm": 0.5017437934875488,
+      "learning_rate": 0.000136855566025037,
+      "loss": 3.5115,
+      "step": 4432
+    },
+    {
+      "epoch": 5.66944,
+      "grad_norm": 0.5259951949119568,
+      "learning_rate": 0.00013681518373939963,
+      "loss": 3.5269,
+      "step": 4433
+    },
+    {
+      "epoch": 5.67072,
+      "grad_norm": 0.49649110436439514,
+      "learning_rate": 0.00013677480145376229,
+      "loss": 3.5358,
+      "step": 4434
+    },
+    {
+      "epoch": 5.672,
+      "grad_norm": 0.5151202082633972,
+      "learning_rate": 0.00013673441916812492,
+      "loss": 3.5663,
+      "step": 4435
+    },
+    {
+      "epoch": 5.67328,
+      "grad_norm": 0.5322958827018738,
+      "learning_rate": 0.00013669403688248752,
+      "loss": 3.5762,
+      "step": 4436
+    },
+    {
+      "epoch": 5.67456,
+      "grad_norm": 0.505379319190979,
+      "learning_rate": 0.00013665365459685018,
+      "loss": 3.5684,
+      "step": 4437
+    },
+    {
+      "epoch": 5.67584,
+      "grad_norm": 0.5019539594650269,
+      "learning_rate": 0.0001366132723112128,
+      "loss": 3.5522,
+      "step": 4438
+    },
+    {
+      "epoch": 5.67712,
+      "grad_norm": 0.5176437497138977,
+      "learning_rate": 0.00013657289002557544,
+      "loss": 3.5629,
+      "step": 4439
+    },
+    {
+      "epoch": 5.6784,
+      "grad_norm": 0.4964632987976074,
+      "learning_rate": 0.00013653250773993807,
+      "loss": 3.6072,
+      "step": 4440
+    },
+    {
+      "epoch": 5.67968,
+      "grad_norm": 0.49535518884658813,
+      "learning_rate": 0.0001364921254543007,
+      "loss": 3.6126,
+      "step": 4441
+    },
+    {
+      "epoch": 5.68096,
+      "grad_norm": 0.4958983063697815,
+      "learning_rate": 0.00013645174316866333,
+      "loss": 3.5765,
+      "step": 4442
+    },
+    {
+      "epoch": 5.68224,
+      "grad_norm": 0.4932209253311157,
+      "learning_rate": 0.00013641136088302598,
+      "loss": 3.5613,
+      "step": 4443
+    },
+    {
+      "epoch": 5.68352,
+      "grad_norm": 0.48881053924560547,
+      "learning_rate": 0.00013637097859738859,
+      "loss": 3.5207,
+      "step": 4444
+    },
+    {
+      "epoch": 5.6848,
+      "grad_norm": 0.4943525791168213,
+      "learning_rate": 0.00013633059631175124,
+      "loss": 3.616,
+      "step": 4445
+    },
+    {
+      "epoch": 5.6860800000000005,
+      "grad_norm": 0.504132091999054,
+      "learning_rate": 0.00013629021402611387,
+      "loss": 3.5652,
+      "step": 4446
+    },
+    {
+      "epoch": 5.68736,
+      "grad_norm": 0.5116013288497925,
+      "learning_rate": 0.0001362498317404765,
+      "loss": 3.5694,
+      "step": 4447
+    },
+    {
+      "epoch": 5.68864,
+      "grad_norm": 0.49192342162132263,
+      "learning_rate": 0.00013620944945483913,
+      "loss": 3.5206,
+      "step": 4448
+    },
+    {
+      "epoch": 5.68992,
+      "grad_norm": 0.49895188212394714,
+      "learning_rate": 0.00013616906716920176,
+      "loss": 3.5208,
+      "step": 4449
+    },
+    {
+      "epoch": 5.6912,
+      "grad_norm": 0.5205003619194031,
+      "learning_rate": 0.0001361286848835644,
+      "loss": 3.6103,
+      "step": 4450
+    },
+    {
+      "epoch": 5.69248,
+      "grad_norm": 0.5131163001060486,
+      "learning_rate": 0.00013608830259792702,
+      "loss": 3.5963,
+      "step": 4451
+    },
+    {
+      "epoch": 5.69376,
+      "grad_norm": 0.4999935030937195,
+      "learning_rate": 0.00013604792031228965,
+      "loss": 3.5195,
+      "step": 4452
+    },
+    {
+      "epoch": 5.69504,
+      "grad_norm": 0.5080338716506958,
+      "learning_rate": 0.00013600753802665228,
+      "loss": 3.5515,
+      "step": 4453
+    },
+    {
+      "epoch": 5.69632,
+      "grad_norm": 0.5050440430641174,
+      "learning_rate": 0.00013596715574101494,
+      "loss": 3.5676,
+      "step": 4454
+    },
+    {
+      "epoch": 5.6975999999999996,
+      "grad_norm": 0.4881718158721924,
+      "learning_rate": 0.00013592677345537757,
+      "loss": 3.5505,
+      "step": 4455
+    },
+    {
+      "epoch": 5.69888,
+      "grad_norm": 0.5083419680595398,
+      "learning_rate": 0.0001358863911697402,
+      "loss": 3.5002,
+      "step": 4456
+    },
+    {
+      "epoch": 5.70016,
+      "grad_norm": 0.5217411518096924,
+      "learning_rate": 0.00013584600888410283,
+      "loss": 3.5094,
+      "step": 4457
+    },
+    {
+      "epoch": 5.70144,
+      "grad_norm": 0.4996829628944397,
+      "learning_rate": 0.00013580562659846546,
+      "loss": 3.6286,
+      "step": 4458
+    },
+    {
+      "epoch": 5.70272,
+      "grad_norm": 0.5101032853126526,
+      "learning_rate": 0.0001357652443128281,
+      "loss": 3.5412,
+      "step": 4459
+    },
+    {
+      "epoch": 5.704,
+      "grad_norm": 0.5019393563270569,
+      "learning_rate": 0.00013572486202719072,
+      "loss": 3.5845,
+      "step": 4460
+    },
+    {
+      "epoch": 5.70528,
+      "grad_norm": 0.5044521689414978,
+      "learning_rate": 0.00013568447974155335,
+      "loss": 3.633,
+      "step": 4461
+    },
+    {
+      "epoch": 5.70656,
+      "grad_norm": 0.5104800462722778,
+      "learning_rate": 0.000135644097455916,
+      "loss": 3.5554,
+      "step": 4462
+    },
+    {
+      "epoch": 5.70784,
+      "grad_norm": 0.4980311989784241,
+      "learning_rate": 0.00013560371517027864,
+      "loss": 3.5151,
+      "step": 4463
+    },
+    {
+      "epoch": 5.70912,
+      "grad_norm": 0.501975417137146,
+      "learning_rate": 0.00013556333288464124,
+      "loss": 3.5999,
+      "step": 4464
+    },
+    {
+      "epoch": 5.7104,
+      "grad_norm": 0.4794447422027588,
+      "learning_rate": 0.0001355229505990039,
+      "loss": 3.5524,
+      "step": 4465
+    },
+    {
+      "epoch": 5.71168,
+      "grad_norm": 0.5068936944007874,
+      "learning_rate": 0.00013548256831336653,
+      "loss": 3.5434,
+      "step": 4466
+    },
+    {
+      "epoch": 5.71296,
+      "grad_norm": 0.5029814839363098,
+      "learning_rate": 0.00013544218602772916,
+      "loss": 3.531,
+      "step": 4467
+    },
+    {
+      "epoch": 5.71424,
+      "grad_norm": 0.48336413502693176,
+      "learning_rate": 0.0001354018037420918,
+      "loss": 3.5022,
+      "step": 4468
+    },
+    {
+      "epoch": 5.71552,
+      "grad_norm": 0.4998653829097748,
+      "learning_rate": 0.00013536142145645442,
+      "loss": 3.6426,
+      "step": 4469
+    },
+    {
+      "epoch": 5.7168,
+      "grad_norm": 0.5224689841270447,
+      "learning_rate": 0.00013532103917081705,
+      "loss": 3.6244,
+      "step": 4470
+    },
+    {
+      "epoch": 5.7180800000000005,
+      "grad_norm": 0.49238261580467224,
+      "learning_rate": 0.0001352806568851797,
+      "loss": 3.5419,
+      "step": 4471
+    },
+    {
+      "epoch": 5.71936,
+      "grad_norm": 0.5162872672080994,
+      "learning_rate": 0.0001352402745995423,
+      "loss": 3.5593,
+      "step": 4472
+    },
+    {
+      "epoch": 5.7206399999999995,
+      "grad_norm": 0.4984869658946991,
+      "learning_rate": 0.00013519989231390497,
+      "loss": 3.5436,
+      "step": 4473
+    },
+    {
+      "epoch": 5.72192,
+      "grad_norm": 0.49168428778648376,
+      "learning_rate": 0.0001351595100282676,
+      "loss": 3.5536,
+      "step": 4474
+    },
+    {
+      "epoch": 5.7232,
+      "grad_norm": 0.47858181595802307,
+      "learning_rate": 0.00013511912774263023,
+      "loss": 3.5043,
+      "step": 4475
+    },
+    {
+      "epoch": 5.72448,
+      "grad_norm": 0.5016847848892212,
+      "learning_rate": 0.00013507874545699286,
+      "loss": 3.6073,
+      "step": 4476
+    },
+    {
+      "epoch": 5.72576,
+      "grad_norm": 0.5029941201210022,
+      "learning_rate": 0.0001350383631713555,
+      "loss": 3.6127,
+      "step": 4477
+    },
+    {
+      "epoch": 5.72704,
+      "grad_norm": 0.48141196370124817,
+      "learning_rate": 0.00013499798088571812,
+      "loss": 3.574,
+      "step": 4478
+    },
+    {
+      "epoch": 5.72832,
+      "grad_norm": 0.4969176948070526,
+      "learning_rate": 0.00013495759860008075,
+      "loss": 3.5789,
+      "step": 4479
+    },
+    {
+      "epoch": 5.7296,
+      "grad_norm": 0.48834389448165894,
+      "learning_rate": 0.00013491721631444338,
+      "loss": 3.5863,
+      "step": 4480
+    },
+    {
+      "epoch": 5.73088,
+      "grad_norm": 0.49428460001945496,
+      "learning_rate": 0.000134876834028806,
+      "loss": 3.5619,
+      "step": 4481
+    },
+    {
+      "epoch": 5.73216,
+      "grad_norm": 0.49237480759620667,
+      "learning_rate": 0.00013483645174316866,
+      "loss": 3.5791,
+      "step": 4482
+    },
+    {
+      "epoch": 5.73344,
+      "grad_norm": 0.5003536343574524,
+      "learning_rate": 0.00013479606945753127,
+      "loss": 3.595,
+      "step": 4483
+    },
+    {
+      "epoch": 5.73472,
+      "grad_norm": 0.4975607395172119,
+      "learning_rate": 0.00013475568717189392,
+      "loss": 3.5885,
+      "step": 4484
+    },
+    {
+      "epoch": 5.736,
+      "grad_norm": 0.5209734439849854,
+      "learning_rate": 0.00013471530488625655,
+      "loss": 3.5795,
+      "step": 4485
+    },
+    {
+      "epoch": 5.73728,
+      "grad_norm": 0.49605241417884827,
+      "learning_rate": 0.00013467492260061918,
+      "loss": 3.5311,
+      "step": 4486
+    },
+    {
+      "epoch": 5.73856,
+      "grad_norm": 0.4768206775188446,
+      "learning_rate": 0.00013463454031498181,
+      "loss": 3.556,
+      "step": 4487
+    },
+    {
+      "epoch": 5.73984,
+      "grad_norm": 0.4915754199028015,
+      "learning_rate": 0.00013459415802934444,
+      "loss": 3.6062,
+      "step": 4488
+    },
+    {
+      "epoch": 5.7411200000000004,
+      "grad_norm": 0.4907207489013672,
+      "learning_rate": 0.00013455377574370707,
+      "loss": 3.5279,
+      "step": 4489
+    },
+    {
+      "epoch": 5.7424,
+      "grad_norm": 0.4975474178791046,
+      "learning_rate": 0.0001345133934580697,
+      "loss": 3.5248,
+      "step": 4490
+    },
+    {
+      "epoch": 5.74368,
+      "grad_norm": 0.481803834438324,
+      "learning_rate": 0.00013447301117243236,
+      "loss": 3.578,
+      "step": 4491
+    },
+    {
+      "epoch": 5.74496,
+      "grad_norm": 0.48836055397987366,
+      "learning_rate": 0.00013443262888679497,
+      "loss": 3.5618,
+      "step": 4492
+    },
+    {
+      "epoch": 5.74624,
+      "grad_norm": 0.48223766684532166,
+      "learning_rate": 0.00013439224660115762,
+      "loss": 3.5529,
+      "step": 4493
+    },
+    {
+      "epoch": 5.74752,
+      "grad_norm": 0.47637295722961426,
+      "learning_rate": 0.00013435186431552025,
+      "loss": 3.5735,
+      "step": 4494
+    },
+    {
+      "epoch": 5.7488,
+      "grad_norm": 0.48114699125289917,
+      "learning_rate": 0.00013431148202988288,
+      "loss": 3.51,
+      "step": 4495
+    },
+    {
+      "epoch": 5.75008,
+      "grad_norm": 0.49171772599220276,
+      "learning_rate": 0.0001342710997442455,
+      "loss": 3.6133,
+      "step": 4496
+    },
+    {
+      "epoch": 5.75136,
+      "grad_norm": 0.49519824981689453,
+      "learning_rate": 0.00013423071745860814,
+      "loss": 3.5715,
+      "step": 4497
+    },
+    {
+      "epoch": 5.7526399999999995,
+      "grad_norm": 0.4867844879627228,
+      "learning_rate": 0.00013419033517297077,
+      "loss": 3.5635,
+      "step": 4498
+    },
+    {
+      "epoch": 5.75392,
+      "grad_norm": 0.4920177161693573,
+      "learning_rate": 0.00013414995288733343,
+      "loss": 3.5152,
+      "step": 4499
+    },
+    {
+      "epoch": 5.7552,
+      "grad_norm": 0.4885466396808624,
+      "learning_rate": 0.00013410957060169603,
+      "loss": 3.5387,
+      "step": 4500
+    },
+    {
+      "epoch": 5.75648,
+      "grad_norm": 0.49109700322151184,
+      "learning_rate": 0.0001340691883160587,
+      "loss": 3.4059,
+      "step": 4501
+    },
+    {
+      "epoch": 5.75776,
+      "grad_norm": 0.4847142696380615,
+      "learning_rate": 0.00013402880603042132,
+      "loss": 3.4758,
+      "step": 4502
+    },
+    {
+      "epoch": 5.75904,
+      "grad_norm": 0.48357903957366943,
+      "learning_rate": 0.00013398842374478392,
+      "loss": 3.4209,
+      "step": 4503
+    },
+    {
+      "epoch": 5.76032,
+      "grad_norm": 0.4984717071056366,
+      "learning_rate": 0.00013394804145914658,
+      "loss": 3.4487,
+      "step": 4504
+    },
+    {
+      "epoch": 5.7616,
+      "grad_norm": 0.4831765592098236,
+      "learning_rate": 0.0001339076591735092,
+      "loss": 3.4414,
+      "step": 4505
+    },
+    {
+      "epoch": 5.76288,
+      "grad_norm": 0.5067592263221741,
+      "learning_rate": 0.00013386727688787184,
+      "loss": 3.4324,
+      "step": 4506
+    },
+    {
+      "epoch": 5.76416,
+      "grad_norm": 0.496747225522995,
+      "learning_rate": 0.00013382689460223447,
+      "loss": 3.5664,
+      "step": 4507
+    },
+    {
+      "epoch": 5.76544,
+      "grad_norm": 0.4953940808773041,
+      "learning_rate": 0.0001337865123165971,
+      "loss": 3.425,
+      "step": 4508
+    },
+    {
+      "epoch": 5.76672,
+      "grad_norm": 0.5052168369293213,
+      "learning_rate": 0.00013374613003095973,
+      "loss": 3.3923,
+      "step": 4509
+    },
+    {
+      "epoch": 5.768,
+      "grad_norm": 0.49418163299560547,
+      "learning_rate": 0.0001337057477453224,
+      "loss": 3.435,
+      "step": 4510
+    },
+    {
+      "epoch": 5.76928,
+      "grad_norm": 0.5005443096160889,
+      "learning_rate": 0.000133665365459685,
+      "loss": 3.4357,
+      "step": 4511
+    },
+    {
+      "epoch": 5.77056,
+      "grad_norm": 0.4933774471282959,
+      "learning_rate": 0.00013362498317404765,
+      "loss": 3.4941,
+      "step": 4512
+    },
+    {
+      "epoch": 5.77184,
+      "grad_norm": 0.49201011657714844,
+      "learning_rate": 0.00013358460088841028,
+      "loss": 3.4382,
+      "step": 4513
+    },
+    {
+      "epoch": 5.7731200000000005,
+      "grad_norm": 0.4911024570465088,
+      "learning_rate": 0.0001335442186027729,
+      "loss": 3.4343,
+      "step": 4514
+    },
+    {
+      "epoch": 5.7744,
+      "grad_norm": 0.4980873167514801,
+      "learning_rate": 0.00013350383631713554,
+      "loss": 3.4233,
+      "step": 4515
+    },
+    {
+      "epoch": 5.77568,
+      "grad_norm": 0.4919237792491913,
+      "learning_rate": 0.00013346345403149817,
+      "loss": 3.5077,
+      "step": 4516
+    },
+    {
+      "epoch": 5.77696,
+      "grad_norm": 0.5091717839241028,
+      "learning_rate": 0.0001334230717458608,
+      "loss": 3.3742,
+      "step": 4517
+    },
+    {
+      "epoch": 5.77824,
+      "grad_norm": 0.49231696128845215,
+      "learning_rate": 0.00013338268946022343,
+      "loss": 3.4706,
+      "step": 4518
+    },
+    {
+      "epoch": 5.77952,
+      "grad_norm": 0.4917304813861847,
+      "learning_rate": 0.00013334230717458606,
+      "loss": 3.4684,
+      "step": 4519
+    },
+    {
+      "epoch": 5.7808,
+      "grad_norm": 0.5030657649040222,
+      "learning_rate": 0.0001333019248889487,
+      "loss": 3.4561,
+      "step": 4520
+    },
+    {
+      "epoch": 5.78208,
+      "grad_norm": 0.5002293586730957,
+      "learning_rate": 0.00013326154260331135,
+      "loss": 3.4811,
+      "step": 4521
+    },
+    {
+      "epoch": 5.78336,
+      "grad_norm": 0.49681586027145386,
+      "learning_rate": 0.00013322116031767398,
+      "loss": 3.4188,
+      "step": 4522
+    },
+    {
+      "epoch": 5.78464,
+      "grad_norm": 0.4821189045906067,
+      "learning_rate": 0.0001331807780320366,
+      "loss": 3.3997,
+      "step": 4523
+    },
+    {
+      "epoch": 5.78592,
+      "grad_norm": 0.4861787259578705,
+      "learning_rate": 0.00013314039574639924,
+      "loss": 3.4388,
+      "step": 4524
+    },
+    {
+      "epoch": 5.7872,
+      "grad_norm": 0.48813995718955994,
+      "learning_rate": 0.00013310001346076187,
+      "loss": 3.4686,
+      "step": 4525
+    },
+    {
+      "epoch": 5.78848,
+      "grad_norm": 0.4976069927215576,
+      "learning_rate": 0.0001330596311751245,
+      "loss": 3.4327,
+      "step": 4526
+    },
+    {
+      "epoch": 5.78976,
+      "grad_norm": 0.5135701298713684,
+      "learning_rate": 0.00013301924888948715,
+      "loss": 3.4294,
+      "step": 4527
+    },
+    {
+      "epoch": 5.79104,
+      "grad_norm": 0.5241384506225586,
+      "learning_rate": 0.00013297886660384976,
+      "loss": 3.5,
+      "step": 4528
+    },
+    {
+      "epoch": 5.79232,
+      "grad_norm": 0.523514449596405,
+      "learning_rate": 0.0001329384843182124,
+      "loss": 3.4332,
+      "step": 4529
+    },
+    {
+      "epoch": 5.7936,
+      "grad_norm": 0.49370086193084717,
+      "learning_rate": 0.00013289810203257504,
+      "loss": 3.4522,
+      "step": 4530
+    },
+    {
+      "epoch": 5.79488,
+      "grad_norm": 0.5221703052520752,
+      "learning_rate": 0.00013285771974693765,
+      "loss": 3.4623,
+      "step": 4531
+    },
+    {
+      "epoch": 5.79616,
+      "grad_norm": 0.5171593427658081,
+      "learning_rate": 0.0001328173374613003,
+      "loss": 3.5817,
+      "step": 4532
+    },
+    {
+      "epoch": 5.79744,
+      "grad_norm": 0.5053053498268127,
+      "learning_rate": 0.00013277695517566293,
+      "loss": 3.4717,
+      "step": 4533
+    },
+    {
+      "epoch": 5.79872,
+      "grad_norm": 0.5156946182250977,
+      "learning_rate": 0.00013273657289002556,
+      "loss": 3.4493,
+      "step": 4534
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.5175525546073914,
+      "learning_rate": 0.0001326961906043882,
+      "loss": 3.4689,
+      "step": 4535
+    },
+    {
+      "epoch": 5.80128,
+      "grad_norm": 0.4784032106399536,
+      "learning_rate": 0.00013265580831875082,
+      "loss": 3.3858,
+      "step": 4536
+    },
+    {
+      "epoch": 5.80256,
+      "grad_norm": 0.49603456258773804,
+      "learning_rate": 0.00013261542603311345,
+      "loss": 3.4567,
+      "step": 4537
+    },
+    {
+      "epoch": 5.80384,
+      "grad_norm": 0.512304425239563,
+      "learning_rate": 0.0001325750437474761,
+      "loss": 3.4817,
+      "step": 4538
+    },
+    {
+      "epoch": 5.80512,
+      "grad_norm": 0.5141502618789673,
+      "learning_rate": 0.00013253466146183871,
+      "loss": 3.3569,
+      "step": 4539
+    },
+    {
+      "epoch": 5.8064,
+      "grad_norm": 0.5332363843917847,
+      "learning_rate": 0.00013249427917620137,
+      "loss": 3.4232,
+      "step": 4540
+    },
+    {
+      "epoch": 5.8076799999999995,
+      "grad_norm": 0.5129743814468384,
+      "learning_rate": 0.000132453896890564,
+      "loss": 3.5049,
+      "step": 4541
+    },
+    {
+      "epoch": 5.80896,
+      "grad_norm": 0.5121658444404602,
+      "learning_rate": 0.00013241351460492663,
+      "loss": 3.3905,
+      "step": 4542
+    },
+    {
+      "epoch": 5.81024,
+      "grad_norm": 0.5085654854774475,
+      "learning_rate": 0.00013237313231928926,
+      "loss": 3.4257,
+      "step": 4543
+    },
+    {
+      "epoch": 5.81152,
+      "grad_norm": 0.5045654773712158,
+      "learning_rate": 0.0001323327500336519,
+      "loss": 3.4154,
+      "step": 4544
+    },
+    {
+      "epoch": 5.8128,
+      "grad_norm": 0.49208593368530273,
+      "learning_rate": 0.00013229236774801452,
+      "loss": 3.4489,
+      "step": 4545
+    },
+    {
+      "epoch": 5.81408,
+      "grad_norm": 0.5095329880714417,
+      "learning_rate": 0.00013225198546237715,
+      "loss": 3.4906,
+      "step": 4546
+    },
+    {
+      "epoch": 5.81536,
+      "grad_norm": 0.5090464353561401,
+      "learning_rate": 0.00013221160317673978,
+      "loss": 3.4252,
+      "step": 4547
+    },
+    {
+      "epoch": 5.81664,
+      "grad_norm": 0.5093858242034912,
+      "learning_rate": 0.0001321712208911024,
+      "loss": 3.4324,
+      "step": 4548
+    },
+    {
+      "epoch": 5.81792,
+      "grad_norm": 0.5184779167175293,
+      "learning_rate": 0.00013213083860546507,
+      "loss": 3.4721,
+      "step": 4549
+    },
+    {
+      "epoch": 5.8192,
+      "grad_norm": 0.5006645321846008,
+      "learning_rate": 0.0001320904563198277,
+      "loss": 3.4371,
+      "step": 4550
+    },
+    {
+      "epoch": 5.82048,
+      "grad_norm": 0.5141642689704895,
+      "learning_rate": 0.00013205007403419033,
+      "loss": 3.3902,
+      "step": 4551
+    },
+    {
+      "epoch": 5.82176,
+      "grad_norm": 0.5118807554244995,
+      "learning_rate": 0.00013200969174855296,
+      "loss": 3.4246,
+      "step": 4552
+    },
+    {
+      "epoch": 5.82304,
+      "grad_norm": 0.5281966924667358,
+      "learning_rate": 0.0001319693094629156,
+      "loss": 3.5053,
+      "step": 4553
+    },
+    {
+      "epoch": 5.82432,
+      "grad_norm": 0.5333678126335144,
+      "learning_rate": 0.00013192892717727822,
+      "loss": 3.4423,
+      "step": 4554
+    },
+    {
+      "epoch": 5.8256,
+      "grad_norm": 0.5082951188087463,
+      "learning_rate": 0.00013188854489164085,
+      "loss": 3.4716,
+      "step": 4555
+    },
+    {
+      "epoch": 5.82688,
+      "grad_norm": 0.5443100929260254,
+      "learning_rate": 0.00013184816260600348,
+      "loss": 3.4103,
+      "step": 4556
+    },
+    {
+      "epoch": 5.8281600000000005,
+      "grad_norm": 0.5330334901809692,
+      "learning_rate": 0.0001318077803203661,
+      "loss": 3.4818,
+      "step": 4557
+    },
+    {
+      "epoch": 5.82944,
+      "grad_norm": 0.525781512260437,
+      "learning_rate": 0.00013176739803472877,
+      "loss": 3.5291,
+      "step": 4558
+    },
+    {
+      "epoch": 5.83072,
+      "grad_norm": 0.5201729536056519,
+      "learning_rate": 0.00013172701574909137,
+      "loss": 3.5319,
+      "step": 4559
+    },
+    {
+      "epoch": 5.832,
+      "grad_norm": 0.5168731212615967,
+      "learning_rate": 0.00013168663346345403,
+      "loss": 3.3785,
+      "step": 4560
+    },
+    {
+      "epoch": 5.83328,
+      "grad_norm": 0.5290531516075134,
+      "learning_rate": 0.00013164625117781666,
+      "loss": 3.4085,
+      "step": 4561
+    },
+    {
+      "epoch": 5.83456,
+      "grad_norm": 0.5436719059944153,
+      "learning_rate": 0.0001316058688921793,
+      "loss": 3.5384,
+      "step": 4562
+    },
+    {
+      "epoch": 5.83584,
+      "grad_norm": 0.5178592801094055,
+      "learning_rate": 0.00013156548660654192,
+      "loss": 3.3646,
+      "step": 4563
+    },
+    {
+      "epoch": 5.83712,
+      "grad_norm": 0.5080958604812622,
+      "learning_rate": 0.00013152510432090455,
+      "loss": 3.4061,
+      "step": 4564
+    },
+    {
+      "epoch": 5.8384,
+      "grad_norm": 0.52479487657547,
+      "learning_rate": 0.00013148472203526718,
+      "loss": 3.5095,
+      "step": 4565
+    },
+    {
+      "epoch": 5.8396799999999995,
+      "grad_norm": 0.531376302242279,
+      "learning_rate": 0.00013144433974962983,
+      "loss": 3.4566,
+      "step": 4566
+    },
+    {
+      "epoch": 5.84096,
+      "grad_norm": 0.5431889891624451,
+      "learning_rate": 0.00013140395746399244,
+      "loss": 3.508,
+      "step": 4567
+    },
+    {
+      "epoch": 5.84224,
+      "grad_norm": 0.5298098921775818,
+      "learning_rate": 0.0001313635751783551,
+      "loss": 3.4721,
+      "step": 4568
+    },
+    {
+      "epoch": 5.84352,
+      "grad_norm": 0.5101543664932251,
+      "learning_rate": 0.00013132319289271772,
+      "loss": 3.4789,
+      "step": 4569
+    },
+    {
+      "epoch": 5.8448,
+      "grad_norm": 0.5432095527648926,
+      "learning_rate": 0.00013128281060708033,
+      "loss": 3.3721,
+      "step": 4570
+    },
+    {
+      "epoch": 5.84608,
+      "grad_norm": 0.5078821778297424,
+      "learning_rate": 0.00013124242832144299,
+      "loss": 3.4997,
+      "step": 4571
+    },
+    {
+      "epoch": 5.84736,
+      "grad_norm": 0.5241081118583679,
+      "learning_rate": 0.00013120204603580562,
+      "loss": 3.3372,
+      "step": 4572
+    },
+    {
+      "epoch": 5.84864,
+      "grad_norm": 0.5384151935577393,
+      "learning_rate": 0.00013116166375016825,
+      "loss": 3.5231,
+      "step": 4573
+    },
+    {
+      "epoch": 5.84992,
+      "grad_norm": 0.5285201072692871,
+      "learning_rate": 0.00013112128146453088,
+      "loss": 3.4693,
+      "step": 4574
+    },
+    {
+      "epoch": 5.8512,
+      "grad_norm": 0.5212448239326477,
+      "learning_rate": 0.0001310808991788935,
+      "loss": 3.4739,
+      "step": 4575
+    },
+    {
+      "epoch": 5.85248,
+      "grad_norm": 0.5330734252929688,
+      "learning_rate": 0.00013104051689325614,
+      "loss": 3.4567,
+      "step": 4576
+    },
+    {
+      "epoch": 5.85376,
+      "grad_norm": 0.5170498490333557,
+      "learning_rate": 0.0001310001346076188,
+      "loss": 3.4205,
+      "step": 4577
+    },
+    {
+      "epoch": 5.85504,
+      "grad_norm": 0.5200040340423584,
+      "learning_rate": 0.00013095975232198142,
+      "loss": 3.5097,
+      "step": 4578
+    },
+    {
+      "epoch": 5.85632,
+      "grad_norm": 0.5016375184059143,
+      "learning_rate": 0.00013091937003634405,
+      "loss": 3.4698,
+      "step": 4579
+    },
+    {
+      "epoch": 5.8576,
+      "grad_norm": 0.5077860951423645,
+      "learning_rate": 0.00013087898775070668,
+      "loss": 3.4589,
+      "step": 4580
+    },
+    {
+      "epoch": 5.85888,
+      "grad_norm": 0.4943268597126007,
+      "learning_rate": 0.0001308386054650693,
+      "loss": 3.4539,
+      "step": 4581
+    },
+    {
+      "epoch": 5.8601600000000005,
+      "grad_norm": 0.5271481871604919,
+      "learning_rate": 0.00013079822317943194,
+      "loss": 3.4382,
+      "step": 4582
+    },
+    {
+      "epoch": 5.86144,
+      "grad_norm": 0.5146161913871765,
+      "learning_rate": 0.00013075784089379457,
+      "loss": 3.4512,
+      "step": 4583
+    },
+    {
+      "epoch": 5.86272,
+      "grad_norm": 0.4986138343811035,
+      "learning_rate": 0.0001307174586081572,
+      "loss": 3.4464,
+      "step": 4584
+    },
+    {
+      "epoch": 5.864,
+      "grad_norm": 0.5161074995994568,
+      "learning_rate": 0.00013067707632251983,
+      "loss": 3.4749,
+      "step": 4585
+    },
+    {
+      "epoch": 5.86528,
+      "grad_norm": 0.5155913829803467,
+      "learning_rate": 0.0001306366940368825,
+      "loss": 3.3951,
+      "step": 4586
+    },
+    {
+      "epoch": 5.86656,
+      "grad_norm": 0.511010468006134,
+      "learning_rate": 0.0001305963117512451,
+      "loss": 3.5491,
+      "step": 4587
+    },
+    {
+      "epoch": 5.86784,
+      "grad_norm": 0.5245579481124878,
+      "learning_rate": 0.00013055592946560775,
+      "loss": 3.4911,
+      "step": 4588
+    },
+    {
+      "epoch": 5.86912,
+      "grad_norm": 0.5299880504608154,
+      "learning_rate": 0.00013051554717997038,
+      "loss": 3.5118,
+      "step": 4589
+    },
+    {
+      "epoch": 5.8704,
+      "grad_norm": 0.5251659154891968,
+      "learning_rate": 0.000130475164894333,
+      "loss": 3.4269,
+      "step": 4590
+    },
+    {
+      "epoch": 5.87168,
+      "grad_norm": 0.5328729748725891,
+      "learning_rate": 0.00013043478260869564,
+      "loss": 3.4694,
+      "step": 4591
+    },
+    {
+      "epoch": 5.87296,
+      "grad_norm": 0.5197456479072571,
+      "learning_rate": 0.00013039440032305827,
+      "loss": 3.4146,
+      "step": 4592
+    },
+    {
+      "epoch": 5.87424,
+      "grad_norm": 0.5089783668518066,
+      "learning_rate": 0.0001303540180374209,
+      "loss": 3.4786,
+      "step": 4593
+    },
+    {
+      "epoch": 5.87552,
+      "grad_norm": 0.5036661624908447,
+      "learning_rate": 0.00013031363575178356,
+      "loss": 3.494,
+      "step": 4594
+    },
+    {
+      "epoch": 5.8768,
+      "grad_norm": 0.5152899622917175,
+      "learning_rate": 0.00013027325346614616,
+      "loss": 3.4804,
+      "step": 4595
+    },
+    {
+      "epoch": 5.87808,
+      "grad_norm": 0.5099830031394958,
+      "learning_rate": 0.00013023287118050882,
+      "loss": 3.5013,
+      "step": 4596
+    },
+    {
+      "epoch": 5.87936,
+      "grad_norm": 0.5049211382865906,
+      "learning_rate": 0.00013019248889487145,
+      "loss": 3.4966,
+      "step": 4597
+    },
+    {
+      "epoch": 5.88064,
+      "grad_norm": 0.5128461122512817,
+      "learning_rate": 0.00013015210660923405,
+      "loss": 3.5043,
+      "step": 4598
+    },
+    {
+      "epoch": 5.88192,
+      "grad_norm": 0.5299509763717651,
+      "learning_rate": 0.0001301117243235967,
+      "loss": 3.5566,
+      "step": 4599
+    },
+    {
+      "epoch": 5.8832,
+      "grad_norm": 0.49836984276771545,
+      "learning_rate": 0.00013007134203795934,
+      "loss": 3.4689,
+      "step": 4600
+    },
+    {
+      "epoch": 5.88448,
+      "grad_norm": 0.5294039845466614,
+      "learning_rate": 0.00013003095975232197,
+      "loss": 3.417,
+      "step": 4601
+    },
+    {
+      "epoch": 5.88576,
+      "grad_norm": 0.5025843381881714,
+      "learning_rate": 0.0001299905774666846,
+      "loss": 3.4296,
+      "step": 4602
+    },
+    {
+      "epoch": 5.88704,
+      "grad_norm": 0.5195445418357849,
+      "learning_rate": 0.00012995019518104723,
+      "loss": 3.4922,
+      "step": 4603
+    },
+    {
+      "epoch": 5.88832,
+      "grad_norm": 0.5041271448135376,
+      "learning_rate": 0.00012990981289540986,
+      "loss": 3.3655,
+      "step": 4604
+    },
+    {
+      "epoch": 5.8896,
+      "grad_norm": 0.5099116563796997,
+      "learning_rate": 0.00012986943060977252,
+      "loss": 3.3968,
+      "step": 4605
+    },
+    {
+      "epoch": 5.89088,
+      "grad_norm": 0.5230572819709778,
+      "learning_rate": 0.00012982904832413512,
+      "loss": 3.4775,
+      "step": 4606
+    },
+    {
+      "epoch": 5.89216,
+      "grad_norm": 0.4989769458770752,
+      "learning_rate": 0.00012978866603849778,
+      "loss": 3.4765,
+      "step": 4607
+    },
+    {
+      "epoch": 5.89344,
+      "grad_norm": 0.5254843831062317,
+      "learning_rate": 0.0001297482837528604,
+      "loss": 3.4976,
+      "step": 4608
+    },
+    {
+      "epoch": 5.8947199999999995,
+      "grad_norm": 0.501315176486969,
+      "learning_rate": 0.00012970790146722304,
+      "loss": 3.3925,
+      "step": 4609
+    },
+    {
+      "epoch": 5.896,
+      "grad_norm": 0.5095430016517639,
+      "learning_rate": 0.00012966751918158567,
+      "loss": 3.4135,
+      "step": 4610
+    },
+    {
+      "epoch": 5.89728,
+      "grad_norm": 0.5203284025192261,
+      "learning_rate": 0.0001296271368959483,
+      "loss": 3.5653,
+      "step": 4611
+    },
+    {
+      "epoch": 5.89856,
+      "grad_norm": 0.524929940700531,
+      "learning_rate": 0.00012958675461031093,
+      "loss": 3.4164,
+      "step": 4612
+    },
+    {
+      "epoch": 5.89984,
+      "grad_norm": 0.49622589349746704,
+      "learning_rate": 0.00012954637232467356,
+      "loss": 3.4858,
+      "step": 4613
+    },
+    {
+      "epoch": 5.90112,
+      "grad_norm": 0.5328811407089233,
+      "learning_rate": 0.00012950599003903621,
+      "loss": 3.4519,
+      "step": 4614
+    },
+    {
+      "epoch": 5.9024,
+      "grad_norm": 0.5233384370803833,
+      "learning_rate": 0.00012946560775339882,
+      "loss": 3.4802,
+      "step": 4615
+    },
+    {
+      "epoch": 5.90368,
+      "grad_norm": 0.5104706287384033,
+      "learning_rate": 0.00012942522546776147,
+      "loss": 3.503,
+      "step": 4616
+    },
+    {
+      "epoch": 5.90496,
+      "grad_norm": 0.5111770033836365,
+      "learning_rate": 0.0001293848431821241,
+      "loss": 3.3882,
+      "step": 4617
+    },
+    {
+      "epoch": 5.90624,
+      "grad_norm": 0.5340044498443604,
+      "learning_rate": 0.00012934446089648673,
+      "loss": 3.4865,
+      "step": 4618
+    },
+    {
+      "epoch": 5.90752,
+      "grad_norm": 0.5179080367088318,
+      "learning_rate": 0.00012930407861084936,
+      "loss": 3.448,
+      "step": 4619
+    },
+    {
+      "epoch": 5.9088,
+      "grad_norm": 0.5176510214805603,
+      "learning_rate": 0.000129263696325212,
+      "loss": 3.4914,
+      "step": 4620
+    },
+    {
+      "epoch": 5.91008,
+      "grad_norm": 0.5105175375938416,
+      "learning_rate": 0.00012922331403957462,
+      "loss": 3.4427,
+      "step": 4621
+    },
+    {
+      "epoch": 5.91136,
+      "grad_norm": 0.4958018362522125,
+      "learning_rate": 0.00012918293175393728,
+      "loss": 3.4691,
+      "step": 4622
+    },
+    {
+      "epoch": 5.91264,
+      "grad_norm": 0.5175390839576721,
+      "learning_rate": 0.00012914254946829988,
+      "loss": 3.4176,
+      "step": 4623
+    },
+    {
+      "epoch": 5.91392,
+      "grad_norm": 0.49366575479507446,
+      "learning_rate": 0.00012910216718266251,
+      "loss": 3.4056,
+      "step": 4624
+    },
+    {
+      "epoch": 5.9152000000000005,
+      "grad_norm": 0.49612316489219666,
+      "learning_rate": 0.00012906178489702517,
+      "loss": 3.4146,
+      "step": 4625
+    },
+    {
+      "epoch": 5.91648,
+      "grad_norm": 0.5153398513793945,
+      "learning_rate": 0.00012902140261138777,
+      "loss": 3.4182,
+      "step": 4626
+    },
+    {
+      "epoch": 5.91776,
+      "grad_norm": 0.5095019936561584,
+      "learning_rate": 0.00012898102032575043,
+      "loss": 3.4108,
+      "step": 4627
+    },
+    {
+      "epoch": 5.91904,
+      "grad_norm": 0.5022311210632324,
+      "learning_rate": 0.00012894063804011306,
+      "loss": 3.4513,
+      "step": 4628
+    },
+    {
+      "epoch": 5.92032,
+      "grad_norm": 0.5156865119934082,
+      "learning_rate": 0.0001289002557544757,
+      "loss": 3.4506,
+      "step": 4629
+    },
+    {
+      "epoch": 5.9216,
+      "grad_norm": 0.4920865595340729,
+      "learning_rate": 0.00012885987346883832,
+      "loss": 3.3831,
+      "step": 4630
+    },
+    {
+      "epoch": 5.92288,
+      "grad_norm": 0.5026581883430481,
+      "learning_rate": 0.00012881949118320095,
+      "loss": 3.4406,
+      "step": 4631
+    },
+    {
+      "epoch": 5.92416,
+      "grad_norm": 0.5237811803817749,
+      "learning_rate": 0.00012877910889756358,
+      "loss": 3.441,
+      "step": 4632
+    },
+    {
+      "epoch": 5.92544,
+      "grad_norm": 0.5131553411483765,
+      "learning_rate": 0.00012873872661192624,
+      "loss": 3.4405,
+      "step": 4633
+    },
+    {
+      "epoch": 5.9267199999999995,
+      "grad_norm": 0.5147265791893005,
+      "learning_rate": 0.00012869834432628884,
+      "loss": 3.4757,
+      "step": 4634
+    },
+    {
+      "epoch": 5.928,
+      "grad_norm": 0.5250530242919922,
+      "learning_rate": 0.0001286579620406515,
+      "loss": 3.4194,
+      "step": 4635
+    },
+    {
+      "epoch": 5.92928,
+      "grad_norm": 0.5036343336105347,
+      "learning_rate": 0.00012861757975501413,
+      "loss": 3.3476,
+      "step": 4636
+    },
+    {
+      "epoch": 5.93056,
+      "grad_norm": 0.5144938826560974,
+      "learning_rate": 0.00012857719746937676,
+      "loss": 3.4407,
+      "step": 4637
+    },
+    {
+      "epoch": 5.93184,
+      "grad_norm": 0.5358858704566956,
+      "learning_rate": 0.0001285368151837394,
+      "loss": 3.4875,
+      "step": 4638
+    },
+    {
+      "epoch": 5.93312,
+      "grad_norm": 0.503612220287323,
+      "learning_rate": 0.00012849643289810202,
+      "loss": 3.4862,
+      "step": 4639
+    },
+    {
+      "epoch": 5.9344,
+      "grad_norm": 0.5070048570632935,
+      "learning_rate": 0.00012845605061246465,
+      "loss": 3.4842,
+      "step": 4640
+    },
+    {
+      "epoch": 5.93568,
+      "grad_norm": 0.5292816758155823,
+      "learning_rate": 0.00012841566832682728,
+      "loss": 3.4764,
+      "step": 4641
+    },
+    {
+      "epoch": 5.93696,
+      "grad_norm": 0.5156516432762146,
+      "learning_rate": 0.0001283752860411899,
+      "loss": 3.4486,
+      "step": 4642
+    },
+    {
+      "epoch": 5.93824,
+      "grad_norm": 0.5131589770317078,
+      "learning_rate": 0.00012833490375555254,
+      "loss": 3.4249,
+      "step": 4643
+    },
+    {
+      "epoch": 5.93952,
+      "grad_norm": 0.5197781324386597,
+      "learning_rate": 0.0001282945214699152,
+      "loss": 3.4929,
+      "step": 4644
+    },
+    {
+      "epoch": 5.9408,
+      "grad_norm": 0.5134233236312866,
+      "learning_rate": 0.00012825413918427783,
+      "loss": 3.445,
+      "step": 4645
+    },
+    {
+      "epoch": 5.94208,
+      "grad_norm": 0.5193766355514526,
+      "learning_rate": 0.00012821375689864046,
+      "loss": 3.4006,
+      "step": 4646
+    },
+    {
+      "epoch": 5.94336,
+      "grad_norm": 0.5153205990791321,
+      "learning_rate": 0.0001281733746130031,
+      "loss": 3.4529,
+      "step": 4647
+    },
+    {
+      "epoch": 5.94464,
+      "grad_norm": 0.5201491713523865,
+      "learning_rate": 0.00012813299232736572,
+      "loss": 3.4742,
+      "step": 4648
+    },
+    {
+      "epoch": 5.94592,
+      "grad_norm": 0.5122271180152893,
+      "learning_rate": 0.00012809261004172835,
+      "loss": 3.4414,
+      "step": 4649
+    },
+    {
+      "epoch": 5.9472000000000005,
+      "grad_norm": 0.5078105926513672,
+      "learning_rate": 0.000128052227756091,
+      "loss": 3.3802,
+      "step": 4650
+    },
+    {
+      "epoch": 5.94848,
+      "grad_norm": 0.5191426873207092,
+      "learning_rate": 0.0001280118454704536,
+      "loss": 3.3924,
+      "step": 4651
+    },
+    {
+      "epoch": 5.94976,
+      "grad_norm": 0.511929452419281,
+      "learning_rate": 0.00012797146318481624,
+      "loss": 3.4573,
+      "step": 4652
+    },
+    {
+      "epoch": 5.95104,
+      "grad_norm": 0.512454628944397,
+      "learning_rate": 0.0001279310808991789,
+      "loss": 3.4888,
+      "step": 4653
+    },
+    {
+      "epoch": 5.95232,
+      "grad_norm": 0.5255744457244873,
+      "learning_rate": 0.0001278906986135415,
+      "loss": 3.5094,
+      "step": 4654
+    },
+    {
+      "epoch": 5.9536,
+      "grad_norm": 0.5149945616722107,
+      "learning_rate": 0.00012785031632790416,
+      "loss": 3.4169,
+      "step": 4655
+    },
+    {
+      "epoch": 5.95488,
+      "grad_norm": 0.507247269153595,
+      "learning_rate": 0.00012780993404226679,
+      "loss": 3.4484,
+      "step": 4656
+    },
+    {
+      "epoch": 5.95616,
+      "grad_norm": 0.5141899585723877,
+      "learning_rate": 0.00012776955175662942,
+      "loss": 3.4004,
+      "step": 4657
+    },
+    {
+      "epoch": 5.95744,
+      "grad_norm": 0.525027334690094,
+      "learning_rate": 0.00012772916947099205,
+      "loss": 3.5179,
+      "step": 4658
+    },
+    {
+      "epoch": 5.95872,
+      "grad_norm": 0.5136873126029968,
+      "learning_rate": 0.00012768878718535468,
+      "loss": 3.4939,
+      "step": 4659
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.5179911255836487,
+      "learning_rate": 0.0001276484048997173,
+      "loss": 3.4725,
+      "step": 4660
+    },
+    {
+      "epoch": 5.96128,
+      "grad_norm": 0.5283648371696472,
+      "learning_rate": 0.00012760802261407996,
+      "loss": 3.4843,
+      "step": 4661
+    },
+    {
+      "epoch": 5.96256,
+      "grad_norm": 0.5253385305404663,
+      "learning_rate": 0.00012756764032844257,
+      "loss": 3.4472,
+      "step": 4662
+    },
+    {
+      "epoch": 5.96384,
+      "grad_norm": 0.5267717242240906,
+      "learning_rate": 0.00012752725804280522,
+      "loss": 3.4637,
+      "step": 4663
+    },
+    {
+      "epoch": 5.96512,
+      "grad_norm": 0.5325003266334534,
+      "learning_rate": 0.00012748687575716785,
+      "loss": 3.4575,
+      "step": 4664
+    },
+    {
+      "epoch": 5.9664,
+      "grad_norm": 0.51334148645401,
+      "learning_rate": 0.00012744649347153048,
+      "loss": 3.4401,
+      "step": 4665
+    },
+    {
+      "epoch": 5.96768,
+      "grad_norm": 0.5157260298728943,
+      "learning_rate": 0.0001274061111858931,
+      "loss": 3.4944,
+      "step": 4666
+    },
+    {
+      "epoch": 5.96896,
+      "grad_norm": 0.5126465559005737,
+      "learning_rate": 0.00012736572890025574,
+      "loss": 3.4313,
+      "step": 4667
+    },
+    {
+      "epoch": 5.97024,
+      "grad_norm": 0.5327089428901672,
+      "learning_rate": 0.00012732534661461837,
+      "loss": 3.4903,
+      "step": 4668
+    },
+    {
+      "epoch": 5.97152,
+      "grad_norm": 0.5237742066383362,
+      "learning_rate": 0.000127284964328981,
+      "loss": 3.4409,
+      "step": 4669
+    },
+    {
+      "epoch": 5.9728,
+      "grad_norm": 0.5220187306404114,
+      "learning_rate": 0.00012724458204334363,
+      "loss": 3.4712,
+      "step": 4670
+    },
+    {
+      "epoch": 5.97408,
+      "grad_norm": 0.5200389623641968,
+      "learning_rate": 0.00012720419975770626,
+      "loss": 3.5174,
+      "step": 4671
+    },
+    {
+      "epoch": 5.97536,
+      "grad_norm": 0.5262846946716309,
+      "learning_rate": 0.00012716381747206892,
+      "loss": 3.4454,
+      "step": 4672
+    },
+    {
+      "epoch": 5.97664,
+      "grad_norm": 0.5050617456436157,
+      "learning_rate": 0.00012712343518643155,
+      "loss": 3.4899,
+      "step": 4673
+    },
+    {
+      "epoch": 5.97792,
+      "grad_norm": 0.503463089466095,
+      "learning_rate": 0.00012708305290079418,
+      "loss": 3.4283,
+      "step": 4674
+    },
+    {
+      "epoch": 5.9792,
+      "grad_norm": 0.5130547881126404,
+      "learning_rate": 0.0001270426706151568,
+      "loss": 3.4625,
+      "step": 4675
+    },
+    {
+      "epoch": 5.98048,
+      "grad_norm": 0.5259498953819275,
+      "learning_rate": 0.00012700228832951944,
+      "loss": 3.491,
+      "step": 4676
+    },
+    {
+      "epoch": 5.9817599999999995,
+      "grad_norm": 0.4981289803981781,
+      "learning_rate": 0.00012696190604388207,
+      "loss": 3.4928,
+      "step": 4677
+    },
+    {
+      "epoch": 5.98304,
+      "grad_norm": 0.520211398601532,
+      "learning_rate": 0.0001269215237582447,
+      "loss": 3.4893,
+      "step": 4678
+    },
+    {
+      "epoch": 5.98432,
+      "grad_norm": 0.5219488739967346,
+      "learning_rate": 0.00012688114147260733,
+      "loss": 3.4604,
+      "step": 4679
+    },
+    {
+      "epoch": 5.9856,
+      "grad_norm": 0.5058456063270569,
+      "learning_rate": 0.00012684075918696996,
+      "loss": 3.4903,
+      "step": 4680
+    },
+    {
+      "epoch": 5.98688,
+      "grad_norm": 0.5197397470474243,
+      "learning_rate": 0.00012680037690133262,
+      "loss": 3.493,
+      "step": 4681
+    },
+    {
+      "epoch": 5.98816,
+      "grad_norm": 0.5335855484008789,
+      "learning_rate": 0.00012675999461569522,
+      "loss": 3.5079,
+      "step": 4682
+    },
+    {
+      "epoch": 5.98944,
+      "grad_norm": 0.5145869255065918,
+      "learning_rate": 0.00012671961233005788,
+      "loss": 3.4,
+      "step": 4683
+    },
+    {
+      "epoch": 5.99072,
+      "grad_norm": 0.5230121612548828,
+      "learning_rate": 0.0001266792300444205,
+      "loss": 3.4968,
+      "step": 4684
+    },
+    {
+      "epoch": 5.992,
+      "grad_norm": 0.5136299729347229,
+      "learning_rate": 0.00012663884775878314,
+      "loss": 3.4992,
+      "step": 4685
+    },
+    {
+      "epoch": 5.99328,
+      "grad_norm": 0.5114295482635498,
+      "learning_rate": 0.00012659846547314577,
+      "loss": 3.4265,
+      "step": 4686
+    },
+    {
+      "epoch": 5.99456,
+      "grad_norm": 0.511169970035553,
+      "learning_rate": 0.0001265580831875084,
+      "loss": 3.4609,
+      "step": 4687
+    },
+    {
+      "epoch": 5.99584,
+      "grad_norm": 0.5237197279930115,
+      "learning_rate": 0.00012651770090187103,
+      "loss": 3.4253,
+      "step": 4688
+    },
+    {
+      "epoch": 5.99712,
+      "grad_norm": 0.5328317880630493,
+      "learning_rate": 0.00012647731861623369,
+      "loss": 3.3671,
+      "step": 4689
+    },
+    {
+      "epoch": 5.9984,
+      "grad_norm": 0.5123398303985596,
+      "learning_rate": 0.0001264369363305963,
+      "loss": 3.4323,
+      "step": 4690
+    },
+    {
+      "epoch": 5.99968,
+      "grad_norm": 0.5049529671669006,
+      "learning_rate": 0.00012639655404495892,
+      "loss": 3.4654,
+      "step": 4691
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.8919804692268372,
+      "learning_rate": 0.00012635617175932158,
+      "loss": 3.1534,
+      "step": 4692
+    },
+    {
+      "epoch": 6.00128,
+      "grad_norm": 0.5235251188278198,
+      "learning_rate": 0.00012631578947368418,
+      "loss": 3.476,
+      "step": 4693
+    },
+    {
+      "epoch": 6.00256,
+      "grad_norm": 0.5221917629241943,
+      "learning_rate": 0.00012627540718804684,
+      "loss": 3.4639,
+      "step": 4694
+    },
+    {
+      "epoch": 6.00384,
+      "grad_norm": 0.5179303288459778,
+      "learning_rate": 0.00012623502490240947,
+      "loss": 3.4869,
+      "step": 4695
+    },
+    {
+      "epoch": 6.00512,
+      "grad_norm": 0.509280264377594,
+      "learning_rate": 0.0001261946426167721,
+      "loss": 3.5512,
+      "step": 4696
+    },
+    {
+      "epoch": 6.0064,
+      "grad_norm": 0.5337740182876587,
+      "learning_rate": 0.00012615426033113473,
+      "loss": 3.4508,
+      "step": 4697
+    },
+    {
+      "epoch": 6.00768,
+      "grad_norm": 0.511637806892395,
+      "learning_rate": 0.00012611387804549736,
+      "loss": 3.4186,
+      "step": 4698
+    },
+    {
+      "epoch": 6.00896,
+      "grad_norm": 0.5246609449386597,
+      "learning_rate": 0.00012607349575986,
+      "loss": 3.4857,
+      "step": 4699
+    },
+    {
+      "epoch": 6.01024,
+      "grad_norm": 0.5246115922927856,
+      "learning_rate": 0.00012603311347422264,
+      "loss": 3.4306,
+      "step": 4700
+    },
+    {
+      "epoch": 6.01152,
+      "grad_norm": 0.5044547915458679,
+      "learning_rate": 0.00012599273118858527,
+      "loss": 3.4457,
+      "step": 4701
+    },
+    {
+      "epoch": 6.0128,
+      "grad_norm": 0.524234414100647,
+      "learning_rate": 0.0001259523489029479,
+      "loss": 3.4756,
+      "step": 4702
+    },
+    {
+      "epoch": 6.01408,
+      "grad_norm": 0.5325637459754944,
+      "learning_rate": 0.00012591196661731053,
+      "loss": 3.4793,
+      "step": 4703
+    },
+    {
+      "epoch": 6.01536,
+      "grad_norm": 0.5177270174026489,
+      "learning_rate": 0.00012587158433167316,
+      "loss": 3.4615,
+      "step": 4704
+    },
+    {
+      "epoch": 6.01664,
+      "grad_norm": 0.5080181360244751,
+      "learning_rate": 0.0001258312020460358,
+      "loss": 3.4701,
+      "step": 4705
+    },
+    {
+      "epoch": 6.01792,
+      "grad_norm": 0.5270411968231201,
+      "learning_rate": 0.00012579081976039842,
+      "loss": 3.4342,
+      "step": 4706
+    },
+    {
+      "epoch": 6.0192,
+      "grad_norm": 0.526719868183136,
+      "learning_rate": 0.00012575043747476105,
+      "loss": 3.4343,
+      "step": 4707
+    },
+    {
+      "epoch": 6.02048,
+      "grad_norm": 0.5216172933578491,
+      "learning_rate": 0.00012571005518912368,
+      "loss": 3.4759,
+      "step": 4708
+    },
+    {
+      "epoch": 6.0217600000000004,
+      "grad_norm": 0.5187539458274841,
+      "learning_rate": 0.00012566967290348634,
+      "loss": 3.5129,
+      "step": 4709
+    },
+    {
+      "epoch": 6.02304,
+      "grad_norm": 0.5328104496002197,
+      "learning_rate": 0.00012562929061784894,
+      "loss": 3.4285,
+      "step": 4710
+    },
+    {
+      "epoch": 6.02432,
+      "grad_norm": 0.5354236364364624,
+      "learning_rate": 0.0001255889083322116,
+      "loss": 3.4189,
+      "step": 4711
+    },
+    {
+      "epoch": 6.0256,
+      "grad_norm": 0.5221328139305115,
+      "learning_rate": 0.00012554852604657423,
+      "loss": 3.338,
+      "step": 4712
+    },
+    {
+      "epoch": 6.02688,
+      "grad_norm": 0.5087037682533264,
+      "learning_rate": 0.00012550814376093686,
+      "loss": 3.4602,
+      "step": 4713
+    },
+    {
+      "epoch": 6.02816,
+      "grad_norm": 0.5138629674911499,
+      "learning_rate": 0.0001254677614752995,
+      "loss": 3.4034,
+      "step": 4714
+    },
+    {
+      "epoch": 6.02944,
+      "grad_norm": 0.5265618562698364,
+      "learning_rate": 0.00012542737918966212,
+      "loss": 3.4126,
+      "step": 4715
+    },
+    {
+      "epoch": 6.03072,
+      "grad_norm": 0.5210886597633362,
+      "learning_rate": 0.00012538699690402475,
+      "loss": 3.4671,
+      "step": 4716
+    },
+    {
+      "epoch": 6.032,
+      "grad_norm": 0.5192465782165527,
+      "learning_rate": 0.0001253466146183874,
+      "loss": 3.3955,
+      "step": 4717
+    },
+    {
+      "epoch": 6.03328,
+      "grad_norm": 0.5116483569145203,
+      "learning_rate": 0.00012530623233275,
+      "loss": 3.4776,
+      "step": 4718
+    },
+    {
+      "epoch": 6.03456,
+      "grad_norm": 0.5206137299537659,
+      "learning_rate": 0.00012526585004711264,
+      "loss": 3.4733,
+      "step": 4719
+    },
+    {
+      "epoch": 6.03584,
+      "grad_norm": 0.523862361907959,
+      "learning_rate": 0.0001252254677614753,
+      "loss": 3.3922,
+      "step": 4720
+    },
+    {
+      "epoch": 6.03712,
+      "grad_norm": 0.5030114054679871,
+      "learning_rate": 0.0001251850854758379,
+      "loss": 3.4266,
+      "step": 4721
+    },
+    {
+      "epoch": 6.0384,
+      "grad_norm": 0.5357561111450195,
+      "learning_rate": 0.00012514470319020056,
+      "loss": 3.4608,
+      "step": 4722
+    },
+    {
+      "epoch": 6.03968,
+      "grad_norm": 0.5137104392051697,
+      "learning_rate": 0.0001251043209045632,
+      "loss": 3.4066,
+      "step": 4723
+    },
+    {
+      "epoch": 6.04096,
+      "grad_norm": 0.49622246623039246,
+      "learning_rate": 0.00012506393861892582,
+      "loss": 3.4098,
+      "step": 4724
+    },
+    {
+      "epoch": 6.04224,
+      "grad_norm": 0.5154051780700684,
+      "learning_rate": 0.00012502355633328845,
+      "loss": 3.4739,
+      "step": 4725
+    },
+    {
+      "epoch": 6.04352,
+      "grad_norm": 0.5417183041572571,
+      "learning_rate": 0.00012498317404765108,
+      "loss": 3.4566,
+      "step": 4726
+    },
+    {
+      "epoch": 6.0448,
+      "grad_norm": 0.5043236613273621,
+      "learning_rate": 0.0001249427917620137,
+      "loss": 3.5022,
+      "step": 4727
+    },
+    {
+      "epoch": 6.04608,
+      "grad_norm": 0.5146865248680115,
+      "learning_rate": 0.00012490240947637637,
+      "loss": 3.5059,
+      "step": 4728
+    },
+    {
+      "epoch": 6.04736,
+      "grad_norm": 0.5209075212478638,
+      "learning_rate": 0.00012486202719073897,
+      "loss": 3.5078,
+      "step": 4729
+    },
+    {
+      "epoch": 6.04864,
+      "grad_norm": 0.5176512002944946,
+      "learning_rate": 0.00012482164490510163,
+      "loss": 3.4633,
+      "step": 4730
+    },
+    {
+      "epoch": 6.04992,
+      "grad_norm": 0.5079405903816223,
+      "learning_rate": 0.00012478126261946426,
+      "loss": 3.4449,
+      "step": 4731
+    },
+    {
+      "epoch": 6.0512,
+      "grad_norm": 0.5069144368171692,
+      "learning_rate": 0.0001247408803338269,
+      "loss": 3.5183,
+      "step": 4732
+    },
+    {
+      "epoch": 6.05248,
+      "grad_norm": 0.5130882263183594,
+      "learning_rate": 0.00012470049804818952,
+      "loss": 3.4276,
+      "step": 4733
+    },
+    {
+      "epoch": 6.05376,
+      "grad_norm": 0.5110294222831726,
+      "learning_rate": 0.00012466011576255215,
+      "loss": 3.4404,
+      "step": 4734
+    },
+    {
+      "epoch": 6.05504,
+      "grad_norm": 0.513787031173706,
+      "learning_rate": 0.00012461973347691478,
+      "loss": 3.4809,
+      "step": 4735
+    },
+    {
+      "epoch": 6.05632,
+      "grad_norm": 0.4997158348560333,
+      "learning_rate": 0.0001245793511912774,
+      "loss": 3.4271,
+      "step": 4736
+    },
+    {
+      "epoch": 6.0576,
+      "grad_norm": 0.5131158828735352,
+      "learning_rate": 0.00012453896890564007,
+      "loss": 3.4951,
+      "step": 4737
+    },
+    {
+      "epoch": 6.05888,
+      "grad_norm": 0.511933445930481,
+      "learning_rate": 0.00012449858662000267,
+      "loss": 3.4584,
+      "step": 4738
+    },
+    {
+      "epoch": 6.06016,
+      "grad_norm": 0.5187560319900513,
+      "learning_rate": 0.00012445820433436533,
+      "loss": 3.4039,
+      "step": 4739
+    },
+    {
+      "epoch": 6.06144,
+      "grad_norm": 0.5163741111755371,
+      "learning_rate": 0.00012441782204872796,
+      "loss": 3.4948,
+      "step": 4740
+    },
+    {
+      "epoch": 6.06272,
+      "grad_norm": 0.5106804370880127,
+      "learning_rate": 0.00012437743976309059,
+      "loss": 3.4327,
+      "step": 4741
+    },
+    {
+      "epoch": 6.064,
+      "grad_norm": 0.5465947985649109,
+      "learning_rate": 0.00012433705747745322,
+      "loss": 3.5185,
+      "step": 4742
+    },
+    {
+      "epoch": 6.06528,
+      "grad_norm": 0.49931657314300537,
+      "learning_rate": 0.00012429667519181585,
+      "loss": 3.3982,
+      "step": 4743
+    },
+    {
+      "epoch": 6.06656,
+      "grad_norm": 0.510944664478302,
+      "learning_rate": 0.00012425629290617848,
+      "loss": 3.4117,
+      "step": 4744
+    },
+    {
+      "epoch": 6.06784,
+      "grad_norm": 0.5148774981498718,
+      "learning_rate": 0.0001242159106205411,
+      "loss": 3.4909,
+      "step": 4745
+    },
+    {
+      "epoch": 6.06912,
+      "grad_norm": 0.5062376260757446,
+      "learning_rate": 0.00012417552833490374,
+      "loss": 3.4504,
+      "step": 4746
+    },
+    {
+      "epoch": 6.0704,
+      "grad_norm": 0.5255651473999023,
+      "learning_rate": 0.00012413514604926637,
+      "loss": 3.5011,
+      "step": 4747
+    },
+    {
+      "epoch": 6.07168,
+      "grad_norm": 0.5072139501571655,
+      "learning_rate": 0.00012409476376362902,
+      "loss": 3.5092,
+      "step": 4748
+    },
+    {
+      "epoch": 6.07296,
+      "grad_norm": 0.5306870341300964,
+      "learning_rate": 0.00012405438147799163,
+      "loss": 3.4533,
+      "step": 4749
+    },
+    {
+      "epoch": 6.07424,
+      "grad_norm": 0.5114386677742004,
+      "learning_rate": 0.00012401399919235428,
+      "loss": 3.4855,
+      "step": 4750
+    },
+    {
+      "epoch": 6.07552,
+      "grad_norm": 0.5131011605262756,
+      "learning_rate": 0.00012397361690671691,
+      "loss": 3.4171,
+      "step": 4751
+    },
+    {
+      "epoch": 6.0768,
+      "grad_norm": 0.5139480233192444,
+      "learning_rate": 0.00012393323462107954,
+      "loss": 3.3778,
+      "step": 4752
+    },
+    {
+      "epoch": 6.07808,
+      "grad_norm": 0.5153706073760986,
+      "learning_rate": 0.00012389285233544217,
+      "loss": 3.4194,
+      "step": 4753
+    },
+    {
+      "epoch": 6.07936,
+      "grad_norm": 0.5061531066894531,
+      "learning_rate": 0.0001238524700498048,
+      "loss": 3.4817,
+      "step": 4754
+    },
+    {
+      "epoch": 6.08064,
+      "grad_norm": 0.5213887691497803,
+      "learning_rate": 0.00012381208776416743,
+      "loss": 3.4902,
+      "step": 4755
+    },
+    {
+      "epoch": 6.08192,
+      "grad_norm": 0.5053319931030273,
+      "learning_rate": 0.0001237717054785301,
+      "loss": 3.4497,
+      "step": 4756
+    },
+    {
+      "epoch": 6.0832,
+      "grad_norm": 0.5132631659507751,
+      "learning_rate": 0.0001237313231928927,
+      "loss": 3.4278,
+      "step": 4757
+    },
+    {
+      "epoch": 6.08448,
+      "grad_norm": 0.5187950730323792,
+      "learning_rate": 0.00012369094090725535,
+      "loss": 3.4636,
+      "step": 4758
+    },
+    {
+      "epoch": 6.08576,
+      "grad_norm": 0.5095174908638,
+      "learning_rate": 0.00012365055862161798,
+      "loss": 3.5092,
+      "step": 4759
+    },
+    {
+      "epoch": 6.08704,
+      "grad_norm": 0.5095679759979248,
+      "learning_rate": 0.0001236101763359806,
+      "loss": 3.4271,
+      "step": 4760
+    },
+    {
+      "epoch": 6.08832,
+      "grad_norm": 0.5283337831497192,
+      "learning_rate": 0.00012356979405034324,
+      "loss": 3.4452,
+      "step": 4761
+    },
+    {
+      "epoch": 6.0896,
+      "grad_norm": 0.5230772495269775,
+      "learning_rate": 0.00012352941176470587,
+      "loss": 3.4283,
+      "step": 4762
+    },
+    {
+      "epoch": 6.09088,
+      "grad_norm": 0.5018015503883362,
+      "learning_rate": 0.0001234890294790685,
+      "loss": 3.5196,
+      "step": 4763
+    },
+    {
+      "epoch": 6.09216,
+      "grad_norm": 0.5031048655509949,
+      "learning_rate": 0.00012344864719343113,
+      "loss": 3.4459,
+      "step": 4764
+    },
+    {
+      "epoch": 6.09344,
+      "grad_norm": 0.49931055307388306,
+      "learning_rate": 0.00012340826490779376,
+      "loss": 3.4384,
+      "step": 4765
+    },
+    {
+      "epoch": 6.09472,
+      "grad_norm": 0.5125550627708435,
+      "learning_rate": 0.0001233678826221564,
+      "loss": 3.5177,
+      "step": 4766
+    },
+    {
+      "epoch": 6.096,
+      "grad_norm": 0.5063710808753967,
+      "learning_rate": 0.00012332750033651905,
+      "loss": 3.3931,
+      "step": 4767
+    },
+    {
+      "epoch": 6.09728,
+      "grad_norm": 0.5210559368133545,
+      "learning_rate": 0.00012328711805088168,
+      "loss": 3.4212,
+      "step": 4768
+    },
+    {
+      "epoch": 6.09856,
+      "grad_norm": 0.5136427879333496,
+      "learning_rate": 0.0001232467357652443,
+      "loss": 3.513,
+      "step": 4769
+    },
+    {
+      "epoch": 6.09984,
+      "grad_norm": 0.5130379796028137,
+      "learning_rate": 0.00012320635347960694,
+      "loss": 3.4504,
+      "step": 4770
+    },
+    {
+      "epoch": 6.10112,
+      "grad_norm": 0.5283026099205017,
+      "learning_rate": 0.00012316597119396957,
+      "loss": 3.4788,
+      "step": 4771
+    },
+    {
+      "epoch": 6.1024,
+      "grad_norm": 0.5066593289375305,
+      "learning_rate": 0.0001231255889083322,
+      "loss": 3.4462,
+      "step": 4772
+    },
+    {
+      "epoch": 6.10368,
+      "grad_norm": 0.5073418021202087,
+      "learning_rate": 0.00012308520662269483,
+      "loss": 3.4944,
+      "step": 4773
+    },
+    {
+      "epoch": 6.10496,
+      "grad_norm": 0.5373370051383972,
+      "learning_rate": 0.00012304482433705746,
+      "loss": 3.4202,
+      "step": 4774
+    },
+    {
+      "epoch": 6.10624,
+      "grad_norm": 0.5177907347679138,
+      "learning_rate": 0.0001230044420514201,
+      "loss": 3.455,
+      "step": 4775
+    },
+    {
+      "epoch": 6.10752,
+      "grad_norm": 0.515060305595398,
+      "learning_rate": 0.00012296405976578275,
+      "loss": 3.3564,
+      "step": 4776
+    },
+    {
+      "epoch": 6.1088,
+      "grad_norm": 0.5271644592285156,
+      "learning_rate": 0.00012292367748014535,
+      "loss": 3.5478,
+      "step": 4777
+    },
+    {
+      "epoch": 6.11008,
+      "grad_norm": 0.5147733688354492,
+      "learning_rate": 0.000122883295194508,
+      "loss": 3.4428,
+      "step": 4778
+    },
+    {
+      "epoch": 6.11136,
+      "grad_norm": 0.5317122340202332,
+      "learning_rate": 0.00012284291290887064,
+      "loss": 3.4613,
+      "step": 4779
+    },
+    {
+      "epoch": 6.11264,
+      "grad_norm": 0.5279807448387146,
+      "learning_rate": 0.00012280253062323327,
+      "loss": 3.3885,
+      "step": 4780
+    },
+    {
+      "epoch": 6.11392,
+      "grad_norm": 0.5295637249946594,
+      "learning_rate": 0.0001227621483375959,
+      "loss": 3.4598,
+      "step": 4781
+    },
+    {
+      "epoch": 6.1152,
+      "grad_norm": 0.5437308549880981,
+      "learning_rate": 0.00012272176605195853,
+      "loss": 3.4348,
+      "step": 4782
+    },
+    {
+      "epoch": 6.11648,
+      "grad_norm": 0.5441926717758179,
+      "learning_rate": 0.00012268138376632116,
+      "loss": 3.4985,
+      "step": 4783
+    },
+    {
+      "epoch": 6.11776,
+      "grad_norm": 0.5059524774551392,
+      "learning_rate": 0.00012264100148068381,
+      "loss": 3.3963,
+      "step": 4784
+    },
+    {
+      "epoch": 6.11904,
+      "grad_norm": 0.5069935917854309,
+      "learning_rate": 0.00012260061919504642,
+      "loss": 3.4357,
+      "step": 4785
+    },
+    {
+      "epoch": 6.12032,
+      "grad_norm": 0.5232740044593811,
+      "learning_rate": 0.00012256023690940905,
+      "loss": 3.417,
+      "step": 4786
+    },
+    {
+      "epoch": 6.1216,
+      "grad_norm": 0.5216338634490967,
+      "learning_rate": 0.0001225198546237717,
+      "loss": 3.5042,
+      "step": 4787
+    },
+    {
+      "epoch": 6.12288,
+      "grad_norm": 0.5343948006629944,
+      "learning_rate": 0.00012247947233813433,
+      "loss": 3.4608,
+      "step": 4788
+    },
+    {
+      "epoch": 6.12416,
+      "grad_norm": 0.5192834138870239,
+      "learning_rate": 0.00012243909005249696,
+      "loss": 3.4506,
+      "step": 4789
+    },
+    {
+      "epoch": 6.12544,
+      "grad_norm": 0.5342220067977905,
+      "learning_rate": 0.0001223987077668596,
+      "loss": 3.5125,
+      "step": 4790
+    },
+    {
+      "epoch": 6.12672,
+      "grad_norm": 0.5249612331390381,
+      "learning_rate": 0.00012235832548122223,
+      "loss": 3.4471,
+      "step": 4791
+    },
+    {
+      "epoch": 6.128,
+      "grad_norm": 0.521793782711029,
+      "learning_rate": 0.00012231794319558486,
+      "loss": 3.4716,
+      "step": 4792
+    },
+    {
+      "epoch": 6.12928,
+      "grad_norm": 0.5028202533721924,
+      "learning_rate": 0.00012227756090994749,
+      "loss": 3.482,
+      "step": 4793
+    },
+    {
+      "epoch": 6.13056,
+      "grad_norm": 0.5394800901412964,
+      "learning_rate": 0.00012223717862431012,
+      "loss": 3.4238,
+      "step": 4794
+    },
+    {
+      "epoch": 6.13184,
+      "grad_norm": 0.5089391469955444,
+      "learning_rate": 0.00012219679633867277,
+      "loss": 3.4219,
+      "step": 4795
+    },
+    {
+      "epoch": 6.13312,
+      "grad_norm": 0.5048559904098511,
+      "learning_rate": 0.0001221564140530354,
+      "loss": 3.4201,
+      "step": 4796
+    },
+    {
+      "epoch": 6.1344,
+      "grad_norm": 0.5039951801300049,
+      "learning_rate": 0.00012211603176739803,
+      "loss": 3.492,
+      "step": 4797
+    },
+    {
+      "epoch": 6.13568,
+      "grad_norm": 0.5205702185630798,
+      "learning_rate": 0.00012207564948176066,
+      "loss": 3.4307,
+      "step": 4798
+    },
+    {
+      "epoch": 6.13696,
+      "grad_norm": 0.5324816107749939,
+      "learning_rate": 0.00012203526719612328,
+      "loss": 3.4693,
+      "step": 4799
+    },
+    {
+      "epoch": 6.13824,
+      "grad_norm": 0.4904349446296692,
+      "learning_rate": 0.00012199488491048592,
+      "loss": 3.3403,
+      "step": 4800
+    },
+    {
+      "epoch": 6.13952,
+      "grad_norm": 0.5032051801681519,
+      "learning_rate": 0.00012195450262484855,
+      "loss": 3.3592,
+      "step": 4801
+    },
+    {
+      "epoch": 6.1408,
+      "grad_norm": 0.5135837197303772,
+      "learning_rate": 0.0001219141203392112,
+      "loss": 3.4193,
+      "step": 4802
+    },
+    {
+      "epoch": 6.14208,
+      "grad_norm": 0.49779608845710754,
+      "learning_rate": 0.00012187373805357381,
+      "loss": 3.4737,
+      "step": 4803
+    },
+    {
+      "epoch": 6.14336,
+      "grad_norm": 0.5166708827018738,
+      "learning_rate": 0.00012183335576793646,
+      "loss": 3.4728,
+      "step": 4804
+    },
+    {
+      "epoch": 6.14464,
+      "grad_norm": 0.524376392364502,
+      "learning_rate": 0.00012179297348229909,
+      "loss": 3.5283,
+      "step": 4805
+    },
+    {
+      "epoch": 6.14592,
+      "grad_norm": 0.5073275566101074,
+      "learning_rate": 0.00012175259119666173,
+      "loss": 3.4763,
+      "step": 4806
+    },
+    {
+      "epoch": 6.1472,
+      "grad_norm": 0.5140321850776672,
+      "learning_rate": 0.00012171220891102435,
+      "loss": 3.5098,
+      "step": 4807
+    },
+    {
+      "epoch": 6.14848,
+      "grad_norm": 0.5006900429725647,
+      "learning_rate": 0.00012167182662538699,
+      "loss": 3.4547,
+      "step": 4808
+    },
+    {
+      "epoch": 6.14976,
+      "grad_norm": 0.5014587640762329,
+      "learning_rate": 0.00012163144433974962,
+      "loss": 3.457,
+      "step": 4809
+    },
+    {
+      "epoch": 6.15104,
+      "grad_norm": 0.5160591006278992,
+      "learning_rate": 0.00012159106205411226,
+      "loss": 3.4686,
+      "step": 4810
+    },
+    {
+      "epoch": 6.15232,
+      "grad_norm": 0.5109474658966064,
+      "learning_rate": 0.00012155067976847488,
+      "loss": 3.4054,
+      "step": 4811
+    },
+    {
+      "epoch": 6.1536,
+      "grad_norm": 0.49859675765037537,
+      "learning_rate": 0.00012151029748283751,
+      "loss": 3.4768,
+      "step": 4812
+    },
+    {
+      "epoch": 6.15488,
+      "grad_norm": 0.5176275372505188,
+      "learning_rate": 0.00012146991519720015,
+      "loss": 3.4349,
+      "step": 4813
+    },
+    {
+      "epoch": 6.15616,
+      "grad_norm": 0.5007618069648743,
+      "learning_rate": 0.00012142953291156277,
+      "loss": 3.3093,
+      "step": 4814
+    },
+    {
+      "epoch": 6.15744,
+      "grad_norm": 0.5273295044898987,
+      "learning_rate": 0.00012138915062592541,
+      "loss": 3.4368,
+      "step": 4815
+    },
+    {
+      "epoch": 6.15872,
+      "grad_norm": 0.49494895339012146,
+      "learning_rate": 0.00012134876834028804,
+      "loss": 3.4666,
+      "step": 4816
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.5105710625648499,
+      "learning_rate": 0.00012130838605465069,
+      "loss": 3.4899,
+      "step": 4817
+    },
+    {
+      "epoch": 6.16128,
+      "grad_norm": 0.5119180083274841,
+      "learning_rate": 0.0001212680037690133,
+      "loss": 3.4086,
+      "step": 4818
+    },
+    {
+      "epoch": 6.16256,
+      "grad_norm": 0.5197157263755798,
+      "learning_rate": 0.00012122762148337595,
+      "loss": 3.4382,
+      "step": 4819
+    },
+    {
+      "epoch": 6.16384,
+      "grad_norm": 0.5209784507751465,
+      "learning_rate": 0.00012118723919773858,
+      "loss": 3.4292,
+      "step": 4820
+    },
+    {
+      "epoch": 6.16512,
+      "grad_norm": 0.5159290432929993,
+      "learning_rate": 0.00012114685691210122,
+      "loss": 3.48,
+      "step": 4821
+    },
+    {
+      "epoch": 6.1664,
+      "grad_norm": 0.5117263793945312,
+      "learning_rate": 0.00012110647462646384,
+      "loss": 3.4611,
+      "step": 4822
+    },
+    {
+      "epoch": 6.16768,
+      "grad_norm": 0.5287443995475769,
+      "learning_rate": 0.00012106609234082648,
+      "loss": 3.4852,
+      "step": 4823
+    },
+    {
+      "epoch": 6.16896,
+      "grad_norm": 0.5062198042869568,
+      "learning_rate": 0.00012102571005518911,
+      "loss": 3.3852,
+      "step": 4824
+    },
+    {
+      "epoch": 6.17024,
+      "grad_norm": 0.5322122573852539,
+      "learning_rate": 0.00012098532776955176,
+      "loss": 3.5064,
+      "step": 4825
+    },
+    {
+      "epoch": 6.17152,
+      "grad_norm": 0.525723397731781,
+      "learning_rate": 0.00012094494548391439,
+      "loss": 3.486,
+      "step": 4826
+    },
+    {
+      "epoch": 6.1728,
+      "grad_norm": 0.5092231035232544,
+      "learning_rate": 0.000120904563198277,
+      "loss": 3.4674,
+      "step": 4827
+    },
+    {
+      "epoch": 6.17408,
+      "grad_norm": 0.5155490636825562,
+      "learning_rate": 0.00012086418091263965,
+      "loss": 3.4667,
+      "step": 4828
+    },
+    {
+      "epoch": 6.17536,
+      "grad_norm": 0.5157750248908997,
+      "learning_rate": 0.00012082379862700228,
+      "loss": 3.4403,
+      "step": 4829
+    },
+    {
+      "epoch": 6.17664,
+      "grad_norm": 0.49893108010292053,
+      "learning_rate": 0.00012078341634136492,
+      "loss": 3.4855,
+      "step": 4830
+    },
+    {
+      "epoch": 6.17792,
+      "grad_norm": 0.5173389911651611,
+      "learning_rate": 0.00012074303405572754,
+      "loss": 3.5103,
+      "step": 4831
+    },
+    {
+      "epoch": 6.1792,
+      "grad_norm": 0.515521228313446,
+      "learning_rate": 0.00012070265177009018,
+      "loss": 3.5259,
+      "step": 4832
+    },
+    {
+      "epoch": 6.18048,
+      "grad_norm": 0.5073827505111694,
+      "learning_rate": 0.00012066226948445281,
+      "loss": 3.46,
+      "step": 4833
+    },
+    {
+      "epoch": 6.18176,
+      "grad_norm": 0.5405731797218323,
+      "learning_rate": 0.00012062188719881545,
+      "loss": 3.4988,
+      "step": 4834
+    },
+    {
+      "epoch": 6.18304,
+      "grad_norm": 0.5233080387115479,
+      "learning_rate": 0.00012058150491317807,
+      "loss": 3.4526,
+      "step": 4835
+    },
+    {
+      "epoch": 6.18432,
+      "grad_norm": 0.5276393294334412,
+      "learning_rate": 0.00012054112262754071,
+      "loss": 3.4405,
+      "step": 4836
+    },
+    {
+      "epoch": 6.1856,
+      "grad_norm": 0.5196059346199036,
+      "learning_rate": 0.00012050074034190334,
+      "loss": 3.4805,
+      "step": 4837
+    },
+    {
+      "epoch": 6.18688,
+      "grad_norm": 0.5034788250923157,
+      "learning_rate": 0.00012046035805626599,
+      "loss": 3.5371,
+      "step": 4838
+    },
+    {
+      "epoch": 6.18816,
+      "grad_norm": 0.516704797744751,
+      "learning_rate": 0.0001204199757706286,
+      "loss": 3.4467,
+      "step": 4839
+    },
+    {
+      "epoch": 6.18944,
+      "grad_norm": 0.512168288230896,
+      "learning_rate": 0.00012037959348499123,
+      "loss": 3.4096,
+      "step": 4840
+    },
+    {
+      "epoch": 6.19072,
+      "grad_norm": 0.5281592607498169,
+      "learning_rate": 0.00012033921119935388,
+      "loss": 3.5391,
+      "step": 4841
+    },
+    {
+      "epoch": 6.192,
+      "grad_norm": 0.49945464730262756,
+      "learning_rate": 0.0001202988289137165,
+      "loss": 3.478,
+      "step": 4842
+    },
+    {
+      "epoch": 6.19328,
+      "grad_norm": 0.5051952004432678,
+      "learning_rate": 0.00012025844662807914,
+      "loss": 3.4026,
+      "step": 4843
+    },
+    {
+      "epoch": 6.19456,
+      "grad_norm": 0.5274873971939087,
+      "learning_rate": 0.00012021806434244177,
+      "loss": 3.5439,
+      "step": 4844
+    },
+    {
+      "epoch": 6.19584,
+      "grad_norm": 0.5015951991081238,
+      "learning_rate": 0.00012017768205680441,
+      "loss": 3.4883,
+      "step": 4845
+    },
+    {
+      "epoch": 6.19712,
+      "grad_norm": 0.5088681578636169,
+      "learning_rate": 0.00012013729977116703,
+      "loss": 3.4458,
+      "step": 4846
+    },
+    {
+      "epoch": 6.1984,
+      "grad_norm": 0.49841299653053284,
+      "learning_rate": 0.00012009691748552967,
+      "loss": 3.4178,
+      "step": 4847
+    },
+    {
+      "epoch": 6.19968,
+      "grad_norm": 0.5172284245491028,
+      "learning_rate": 0.0001200565351998923,
+      "loss": 3.4052,
+      "step": 4848
+    },
+    {
+      "epoch": 6.20096,
+      "grad_norm": 0.5176679491996765,
+      "learning_rate": 0.00012001615291425495,
+      "loss": 3.4595,
+      "step": 4849
+    },
+    {
+      "epoch": 6.20224,
+      "grad_norm": 0.5075709819793701,
+      "learning_rate": 0.00011997577062861756,
+      "loss": 3.439,
+      "step": 4850
+    },
+    {
+      "epoch": 6.20352,
+      "grad_norm": 0.5104464888572693,
+      "learning_rate": 0.0001199353883429802,
+      "loss": 3.4479,
+      "step": 4851
+    },
+    {
+      "epoch": 6.2048,
+      "grad_norm": 0.504559338092804,
+      "learning_rate": 0.00011989500605734284,
+      "loss": 3.4244,
+      "step": 4852
+    },
+    {
+      "epoch": 6.20608,
+      "grad_norm": 0.5160381197929382,
+      "learning_rate": 0.00011985462377170547,
+      "loss": 3.4028,
+      "step": 4853
+    },
+    {
+      "epoch": 6.2073599999999995,
+      "grad_norm": 0.532819390296936,
+      "learning_rate": 0.0001198142414860681,
+      "loss": 3.4141,
+      "step": 4854
+    },
+    {
+      "epoch": 6.20864,
+      "grad_norm": 0.5191538333892822,
+      "learning_rate": 0.00011977385920043073,
+      "loss": 3.4793,
+      "step": 4855
+    },
+    {
+      "epoch": 6.20992,
+      "grad_norm": 0.5210806727409363,
+      "learning_rate": 0.00011973347691479337,
+      "loss": 3.4534,
+      "step": 4856
+    },
+    {
+      "epoch": 6.2112,
+      "grad_norm": 0.5080645680427551,
+      "learning_rate": 0.000119693094629156,
+      "loss": 3.447,
+      "step": 4857
+    },
+    {
+      "epoch": 6.21248,
+      "grad_norm": 0.5201761722564697,
+      "learning_rate": 0.00011965271234351863,
+      "loss": 3.481,
+      "step": 4858
+    },
+    {
+      "epoch": 6.21376,
+      "grad_norm": 0.5427380800247192,
+      "learning_rate": 0.00011961233005788126,
+      "loss": 3.462,
+      "step": 4859
+    },
+    {
+      "epoch": 6.21504,
+      "grad_norm": 0.5049818158149719,
+      "learning_rate": 0.0001195719477722439,
+      "loss": 3.4419,
+      "step": 4860
+    },
+    {
+      "epoch": 6.21632,
+      "grad_norm": 0.5197364687919617,
+      "learning_rate": 0.00011953156548660653,
+      "loss": 3.4193,
+      "step": 4861
+    },
+    {
+      "epoch": 6.2176,
+      "grad_norm": 0.5341804623603821,
+      "learning_rate": 0.00011949118320096918,
+      "loss": 3.4411,
+      "step": 4862
+    },
+    {
+      "epoch": 6.21888,
+      "grad_norm": 0.5286295413970947,
+      "learning_rate": 0.0001194508009153318,
+      "loss": 3.4425,
+      "step": 4863
+    },
+    {
+      "epoch": 6.22016,
+      "grad_norm": 0.5018596053123474,
+      "learning_rate": 0.00011941041862969444,
+      "loss": 3.4148,
+      "step": 4864
+    },
+    {
+      "epoch": 6.22144,
+      "grad_norm": 0.5178782939910889,
+      "learning_rate": 0.00011937003634405707,
+      "loss": 3.4321,
+      "step": 4865
+    },
+    {
+      "epoch": 6.22272,
+      "grad_norm": 0.5278565287590027,
+      "learning_rate": 0.00011932965405841968,
+      "loss": 3.4431,
+      "step": 4866
+    },
+    {
+      "epoch": 6.224,
+      "grad_norm": 0.5157015919685364,
+      "learning_rate": 0.00011928927177278233,
+      "loss": 3.5341,
+      "step": 4867
+    },
+    {
+      "epoch": 6.22528,
+      "grad_norm": 0.5221292972564697,
+      "learning_rate": 0.00011924888948714496,
+      "loss": 3.4478,
+      "step": 4868
+    },
+    {
+      "epoch": 6.22656,
+      "grad_norm": 0.5204195976257324,
+      "learning_rate": 0.0001192085072015076,
+      "loss": 3.4702,
+      "step": 4869
+    },
+    {
+      "epoch": 6.22784,
+      "grad_norm": 0.5190892219543457,
+      "learning_rate": 0.00011916812491587022,
+      "loss": 3.553,
+      "step": 4870
+    },
+    {
+      "epoch": 6.22912,
+      "grad_norm": 0.52687007188797,
+      "learning_rate": 0.00011912774263023286,
+      "loss": 3.4113,
+      "step": 4871
+    },
+    {
+      "epoch": 6.2304,
+      "grad_norm": 0.5159428715705872,
+      "learning_rate": 0.00011908736034459549,
+      "loss": 3.3968,
+      "step": 4872
+    },
+    {
+      "epoch": 6.23168,
+      "grad_norm": 0.5180546641349792,
+      "learning_rate": 0.00011904697805895814,
+      "loss": 3.4233,
+      "step": 4873
+    },
+    {
+      "epoch": 6.23296,
+      "grad_norm": 0.5099559426307678,
+      "learning_rate": 0.00011900659577332075,
+      "loss": 3.419,
+      "step": 4874
+    },
+    {
+      "epoch": 6.23424,
+      "grad_norm": 0.5090861916542053,
+      "learning_rate": 0.0001189662134876834,
+      "loss": 3.4224,
+      "step": 4875
+    },
+    {
+      "epoch": 6.23552,
+      "grad_norm": 0.5132479667663574,
+      "learning_rate": 0.00011892583120204603,
+      "loss": 3.4735,
+      "step": 4876
+    },
+    {
+      "epoch": 6.2368,
+      "grad_norm": 0.5198119878768921,
+      "learning_rate": 0.00011888544891640867,
+      "loss": 3.4072,
+      "step": 4877
+    },
+    {
+      "epoch": 6.23808,
+      "grad_norm": 0.5327208042144775,
+      "learning_rate": 0.00011884506663077129,
+      "loss": 3.4841,
+      "step": 4878
+    },
+    {
+      "epoch": 6.23936,
+      "grad_norm": 0.5163946747779846,
+      "learning_rate": 0.00011880468434513392,
+      "loss": 3.483,
+      "step": 4879
+    },
+    {
+      "epoch": 6.24064,
+      "grad_norm": 0.4975844621658325,
+      "learning_rate": 0.00011876430205949656,
+      "loss": 3.5362,
+      "step": 4880
+    },
+    {
+      "epoch": 6.24192,
+      "grad_norm": 0.5258901715278625,
+      "learning_rate": 0.00011872391977385919,
+      "loss": 3.449,
+      "step": 4881
+    },
+    {
+      "epoch": 6.2432,
+      "grad_norm": 0.5171861052513123,
+      "learning_rate": 0.00011868353748822182,
+      "loss": 3.3967,
+      "step": 4882
+    },
+    {
+      "epoch": 6.24448,
+      "grad_norm": 0.5271828174591064,
+      "learning_rate": 0.00011864315520258445,
+      "loss": 3.4485,
+      "step": 4883
+    },
+    {
+      "epoch": 6.24576,
+      "grad_norm": 0.5008876323699951,
+      "learning_rate": 0.00011860277291694709,
+      "loss": 3.507,
+      "step": 4884
+    },
+    {
+      "epoch": 6.24704,
+      "grad_norm": 0.5155926942825317,
+      "learning_rate": 0.00011856239063130972,
+      "loss": 3.4627,
+      "step": 4885
+    },
+    {
+      "epoch": 6.24832,
+      "grad_norm": 0.5127007365226746,
+      "learning_rate": 0.00011852200834567235,
+      "loss": 3.4644,
+      "step": 4886
+    },
+    {
+      "epoch": 6.2496,
+      "grad_norm": 0.5121738314628601,
+      "learning_rate": 0.00011848162606003498,
+      "loss": 3.4467,
+      "step": 4887
+    },
+    {
+      "epoch": 6.25088,
+      "grad_norm": 0.5105927586555481,
+      "learning_rate": 0.00011844124377439763,
+      "loss": 3.4391,
+      "step": 4888
+    },
+    {
+      "epoch": 6.25216,
+      "grad_norm": 0.5314432382583618,
+      "learning_rate": 0.00011840086148876026,
+      "loss": 3.5517,
+      "step": 4889
+    },
+    {
+      "epoch": 6.25344,
+      "grad_norm": 0.5190565586090088,
+      "learning_rate": 0.00011836047920312289,
+      "loss": 3.41,
+      "step": 4890
+    },
+    {
+      "epoch": 6.25472,
+      "grad_norm": 0.5455322861671448,
+      "learning_rate": 0.00011832009691748552,
+      "loss": 3.4535,
+      "step": 4891
+    },
+    {
+      "epoch": 6.256,
+      "grad_norm": 0.5055553913116455,
+      "learning_rate": 0.00011827971463184816,
+      "loss": 3.5182,
+      "step": 4892
+    },
+    {
+      "epoch": 6.25728,
+      "grad_norm": 0.5350437760353088,
+      "learning_rate": 0.00011823933234621079,
+      "loss": 3.4637,
+      "step": 4893
+    },
+    {
+      "epoch": 6.25856,
+      "grad_norm": 0.5320977568626404,
+      "learning_rate": 0.00011819895006057341,
+      "loss": 3.4818,
+      "step": 4894
+    },
+    {
+      "epoch": 6.25984,
+      "grad_norm": 0.5101127624511719,
+      "learning_rate": 0.00011815856777493605,
+      "loss": 3.4325,
+      "step": 4895
+    },
+    {
+      "epoch": 6.26112,
+      "grad_norm": 0.5186399817466736,
+      "learning_rate": 0.00011811818548929868,
+      "loss": 3.4622,
+      "step": 4896
+    },
+    {
+      "epoch": 6.2624,
+      "grad_norm": 0.5287879109382629,
+      "learning_rate": 0.00011807780320366132,
+      "loss": 3.4531,
+      "step": 4897
+    },
+    {
+      "epoch": 6.26368,
+      "grad_norm": 0.5211589932441711,
+      "learning_rate": 0.00011803742091802394,
+      "loss": 3.5668,
+      "step": 4898
+    },
+    {
+      "epoch": 6.26496,
+      "grad_norm": 0.5217843651771545,
+      "learning_rate": 0.00011799703863238658,
+      "loss": 3.4539,
+      "step": 4899
+    },
+    {
+      "epoch": 6.26624,
+      "grad_norm": 0.5402528047561646,
+      "learning_rate": 0.00011795665634674921,
+      "loss": 3.4953,
+      "step": 4900
+    },
+    {
+      "epoch": 6.26752,
+      "grad_norm": 0.5263811945915222,
+      "learning_rate": 0.00011791627406111186,
+      "loss": 3.4311,
+      "step": 4901
+    },
+    {
+      "epoch": 6.2688,
+      "grad_norm": 0.5283783674240112,
+      "learning_rate": 0.00011787589177547448,
+      "loss": 3.4235,
+      "step": 4902
+    },
+    {
+      "epoch": 6.27008,
+      "grad_norm": 0.5406724214553833,
+      "learning_rate": 0.00011783550948983712,
+      "loss": 3.4439,
+      "step": 4903
+    },
+    {
+      "epoch": 6.27136,
+      "grad_norm": 0.5088178515434265,
+      "learning_rate": 0.00011779512720419975,
+      "loss": 3.4806,
+      "step": 4904
+    },
+    {
+      "epoch": 6.27264,
+      "grad_norm": 0.5212264060974121,
+      "learning_rate": 0.00011775474491856239,
+      "loss": 3.3909,
+      "step": 4905
+    },
+    {
+      "epoch": 6.27392,
+      "grad_norm": 0.5153950452804565,
+      "learning_rate": 0.00011771436263292501,
+      "loss": 3.3559,
+      "step": 4906
+    },
+    {
+      "epoch": 6.2752,
+      "grad_norm": 0.5014641880989075,
+      "learning_rate": 0.00011767398034728764,
+      "loss": 3.4949,
+      "step": 4907
+    },
+    {
+      "epoch": 6.27648,
+      "grad_norm": 0.5058307647705078,
+      "learning_rate": 0.00011763359806165028,
+      "loss": 3.4748,
+      "step": 4908
+    },
+    {
+      "epoch": 6.27776,
+      "grad_norm": 0.5182768106460571,
+      "learning_rate": 0.0001175932157760129,
+      "loss": 3.5126,
+      "step": 4909
+    },
+    {
+      "epoch": 6.27904,
+      "grad_norm": 0.50447678565979,
+      "learning_rate": 0.00011755283349037554,
+      "loss": 3.4261,
+      "step": 4910
+    },
+    {
+      "epoch": 6.28032,
+      "grad_norm": 0.5356603860855103,
+      "learning_rate": 0.00011751245120473817,
+      "loss": 3.489,
+      "step": 4911
+    },
+    {
+      "epoch": 6.2816,
+      "grad_norm": 0.5243074893951416,
+      "learning_rate": 0.00011747206891910082,
+      "loss": 3.4682,
+      "step": 4912
+    },
+    {
+      "epoch": 6.2828800000000005,
+      "grad_norm": 0.5267544984817505,
+      "learning_rate": 0.00011743168663346345,
+      "loss": 3.4675,
+      "step": 4913
+    },
+    {
+      "epoch": 6.28416,
+      "grad_norm": 0.5158196687698364,
+      "learning_rate": 0.00011739130434782608,
+      "loss": 3.4589,
+      "step": 4914
+    },
+    {
+      "epoch": 6.28544,
+      "grad_norm": 0.5127290487289429,
+      "learning_rate": 0.0001173509220621887,
+      "loss": 3.4685,
+      "step": 4915
+    },
+    {
+      "epoch": 6.28672,
+      "grad_norm": 0.5056057572364807,
+      "learning_rate": 0.00011731053977655135,
+      "loss": 3.4143,
+      "step": 4916
+    },
+    {
+      "epoch": 6.288,
+      "grad_norm": 0.5179975032806396,
+      "learning_rate": 0.00011727015749091398,
+      "loss": 3.4381,
+      "step": 4917
+    },
+    {
+      "epoch": 6.28928,
+      "grad_norm": 0.5318068265914917,
+      "learning_rate": 0.00011722977520527661,
+      "loss": 3.4787,
+      "step": 4918
+    },
+    {
+      "epoch": 6.29056,
+      "grad_norm": 0.5078819990158081,
+      "learning_rate": 0.00011718939291963924,
+      "loss": 3.4634,
+      "step": 4919
+    },
+    {
+      "epoch": 6.29184,
+      "grad_norm": 0.5218449831008911,
+      "learning_rate": 0.00011714901063400187,
+      "loss": 3.4495,
+      "step": 4920
+    },
+    {
+      "epoch": 6.29312,
+      "grad_norm": 0.5181546211242676,
+      "learning_rate": 0.00011710862834836451,
+      "loss": 3.5156,
+      "step": 4921
+    },
+    {
+      "epoch": 6.2943999999999996,
+      "grad_norm": 0.5055398344993591,
+      "learning_rate": 0.00011706824606272713,
+      "loss": 3.4951,
+      "step": 4922
+    },
+    {
+      "epoch": 6.29568,
+      "grad_norm": 0.5111787915229797,
+      "learning_rate": 0.00011702786377708977,
+      "loss": 3.4858,
+      "step": 4923
+    },
+    {
+      "epoch": 6.29696,
+      "grad_norm": 0.5231371521949768,
+      "learning_rate": 0.0001169874814914524,
+      "loss": 3.4905,
+      "step": 4924
+    },
+    {
+      "epoch": 6.29824,
+      "grad_norm": 0.5168370008468628,
+      "learning_rate": 0.00011694709920581505,
+      "loss": 3.4655,
+      "step": 4925
+    },
+    {
+      "epoch": 6.29952,
+      "grad_norm": 0.5131970047950745,
+      "learning_rate": 0.00011690671692017766,
+      "loss": 3.4392,
+      "step": 4926
+    },
+    {
+      "epoch": 6.3008,
+      "grad_norm": 0.5175589323043823,
+      "learning_rate": 0.00011686633463454031,
+      "loss": 3.4428,
+      "step": 4927
+    },
+    {
+      "epoch": 6.30208,
+      "grad_norm": 0.5290152430534363,
+      "learning_rate": 0.00011682595234890294,
+      "loss": 3.4106,
+      "step": 4928
+    },
+    {
+      "epoch": 6.30336,
+      "grad_norm": 0.5033940672874451,
+      "learning_rate": 0.00011678557006326558,
+      "loss": 3.4195,
+      "step": 4929
+    },
+    {
+      "epoch": 6.30464,
+      "grad_norm": 0.5039850473403931,
+      "learning_rate": 0.0001167451877776282,
+      "loss": 3.4448,
+      "step": 4930
+    },
+    {
+      "epoch": 6.30592,
+      "grad_norm": 0.5078590512275696,
+      "learning_rate": 0.00011670480549199084,
+      "loss": 3.4318,
+      "step": 4931
+    },
+    {
+      "epoch": 6.3072,
+      "grad_norm": 0.5021501779556274,
+      "learning_rate": 0.00011666442320635347,
+      "loss": 3.4763,
+      "step": 4932
+    },
+    {
+      "epoch": 6.30848,
+      "grad_norm": 0.5251853466033936,
+      "learning_rate": 0.00011662404092071609,
+      "loss": 3.4158,
+      "step": 4933
+    },
+    {
+      "epoch": 6.30976,
+      "grad_norm": 0.5062684416770935,
+      "learning_rate": 0.00011658365863507873,
+      "loss": 3.4314,
+      "step": 4934
+    },
+    {
+      "epoch": 6.31104,
+      "grad_norm": 0.5202977061271667,
+      "learning_rate": 0.00011654327634944136,
+      "loss": 3.4411,
+      "step": 4935
+    },
+    {
+      "epoch": 6.31232,
+      "grad_norm": 0.5242408514022827,
+      "learning_rate": 0.000116502894063804,
+      "loss": 3.5071,
+      "step": 4936
+    },
+    {
+      "epoch": 6.3136,
+      "grad_norm": 0.5003301501274109,
+      "learning_rate": 0.00011646251177816662,
+      "loss": 3.4682,
+      "step": 4937
+    },
+    {
+      "epoch": 6.31488,
+      "grad_norm": 0.5161333680152893,
+      "learning_rate": 0.00011642212949252927,
+      "loss": 3.4328,
+      "step": 4938
+    },
+    {
+      "epoch": 6.31616,
+      "grad_norm": 0.4897507429122925,
+      "learning_rate": 0.0001163817472068919,
+      "loss": 3.5237,
+      "step": 4939
+    },
+    {
+      "epoch": 6.31744,
+      "grad_norm": 0.5239113569259644,
+      "learning_rate": 0.00011634136492125454,
+      "loss": 3.4244,
+      "step": 4940
+    },
+    {
+      "epoch": 6.31872,
+      "grad_norm": 0.5347805023193359,
+      "learning_rate": 0.00011630098263561716,
+      "loss": 3.4078,
+      "step": 4941
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.49066007137298584,
+      "learning_rate": 0.0001162606003499798,
+      "loss": 3.474,
+      "step": 4942
+    },
+    {
+      "epoch": 6.32128,
+      "grad_norm": 0.5044615268707275,
+      "learning_rate": 0.00011622021806434243,
+      "loss": 3.5421,
+      "step": 4943
+    },
+    {
+      "epoch": 6.32256,
+      "grad_norm": 0.5164585113525391,
+      "learning_rate": 0.00011617983577870507,
+      "loss": 3.4537,
+      "step": 4944
+    },
+    {
+      "epoch": 6.32384,
+      "grad_norm": 0.5173794627189636,
+      "learning_rate": 0.00011613945349306769,
+      "loss": 3.4225,
+      "step": 4945
+    },
+    {
+      "epoch": 6.32512,
+      "grad_norm": 0.5148087739944458,
+      "learning_rate": 0.00011609907120743033,
+      "loss": 3.4548,
+      "step": 4946
+    },
+    {
+      "epoch": 6.3264,
+      "grad_norm": 0.533028781414032,
+      "learning_rate": 0.00011605868892179296,
+      "loss": 3.5002,
+      "step": 4947
+    },
+    {
+      "epoch": 6.32768,
+      "grad_norm": 0.5113101005554199,
+      "learning_rate": 0.0001160183066361556,
+      "loss": 3.425,
+      "step": 4948
+    },
+    {
+      "epoch": 6.32896,
+      "grad_norm": 0.5407309532165527,
+      "learning_rate": 0.00011597792435051822,
+      "loss": 3.52,
+      "step": 4949
+    },
+    {
+      "epoch": 6.33024,
+      "grad_norm": 0.5209356546401978,
+      "learning_rate": 0.00011593754206488085,
+      "loss": 3.4593,
+      "step": 4950
+    },
+    {
+      "epoch": 6.33152,
+      "grad_norm": 0.516190230846405,
+      "learning_rate": 0.0001158971597792435,
+      "loss": 3.4489,
+      "step": 4951
+    },
+    {
+      "epoch": 6.3328,
+      "grad_norm": 0.503466784954071,
+      "learning_rate": 0.00011585677749360613,
+      "loss": 3.3637,
+      "step": 4952
+    },
+    {
+      "epoch": 6.33408,
+      "grad_norm": 0.534572422504425,
+      "learning_rate": 0.00011581639520796877,
+      "loss": 3.4754,
+      "step": 4953
+    },
+    {
+      "epoch": 6.33536,
+      "grad_norm": 0.5381052494049072,
+      "learning_rate": 0.00011577601292233139,
+      "loss": 3.4082,
+      "step": 4954
+    },
+    {
+      "epoch": 6.33664,
+      "grad_norm": 0.5160523653030396,
+      "learning_rate": 0.00011573563063669403,
+      "loss": 3.4497,
+      "step": 4955
+    },
+    {
+      "epoch": 6.33792,
+      "grad_norm": 0.5168169140815735,
+      "learning_rate": 0.00011569524835105666,
+      "loss": 3.4778,
+      "step": 4956
+    },
+    {
+      "epoch": 6.3392,
+      "grad_norm": 0.5144256949424744,
+      "learning_rate": 0.0001156548660654193,
+      "loss": 3.5237,
+      "step": 4957
+    },
+    {
+      "epoch": 6.34048,
+      "grad_norm": 0.529626190662384,
+      "learning_rate": 0.00011561448377978192,
+      "loss": 3.4262,
+      "step": 4958
+    },
+    {
+      "epoch": 6.34176,
+      "grad_norm": 0.5221841335296631,
+      "learning_rate": 0.00011557410149414457,
+      "loss": 3.4335,
+      "step": 4959
+    },
+    {
+      "epoch": 6.34304,
+      "grad_norm": 0.5182304382324219,
+      "learning_rate": 0.0001155337192085072,
+      "loss": 3.3717,
+      "step": 4960
+    },
+    {
+      "epoch": 6.34432,
+      "grad_norm": 0.5277256369590759,
+      "learning_rate": 0.00011549333692286981,
+      "loss": 3.4908,
+      "step": 4961
+    },
+    {
+      "epoch": 6.3456,
+      "grad_norm": 0.5436074733734131,
+      "learning_rate": 0.00011545295463723246,
+      "loss": 3.5252,
+      "step": 4962
+    },
+    {
+      "epoch": 6.34688,
+      "grad_norm": 0.5004885196685791,
+      "learning_rate": 0.00011541257235159509,
+      "loss": 3.4502,
+      "step": 4963
+    },
+    {
+      "epoch": 6.34816,
+      "grad_norm": 0.5198299884796143,
+      "learning_rate": 0.00011537219006595773,
+      "loss": 3.4588,
+      "step": 4964
+    },
+    {
+      "epoch": 6.3494399999999995,
+      "grad_norm": 0.5136741399765015,
+      "learning_rate": 0.00011533180778032035,
+      "loss": 3.4659,
+      "step": 4965
+    },
+    {
+      "epoch": 6.35072,
+      "grad_norm": 0.5468453764915466,
+      "learning_rate": 0.00011529142549468299,
+      "loss": 3.4652,
+      "step": 4966
+    },
+    {
+      "epoch": 6.352,
+      "grad_norm": 0.4965207874774933,
+      "learning_rate": 0.00011525104320904562,
+      "loss": 3.4365,
+      "step": 4967
+    },
+    {
+      "epoch": 6.35328,
+      "grad_norm": 0.5069467425346375,
+      "learning_rate": 0.00011521066092340826,
+      "loss": 3.4168,
+      "step": 4968
+    },
+    {
+      "epoch": 6.35456,
+      "grad_norm": 0.5168161988258362,
+      "learning_rate": 0.00011517027863777088,
+      "loss": 3.519,
+      "step": 4969
+    },
+    {
+      "epoch": 6.35584,
+      "grad_norm": 0.5197111368179321,
+      "learning_rate": 0.00011512989635213352,
+      "loss": 3.4787,
+      "step": 4970
+    },
+    {
+      "epoch": 6.35712,
+      "grad_norm": 0.5127108097076416,
+      "learning_rate": 0.00011508951406649615,
+      "loss": 3.5048,
+      "step": 4971
+    },
+    {
+      "epoch": 6.3584,
+      "grad_norm": 0.49976083636283875,
+      "learning_rate": 0.0001150491317808588,
+      "loss": 3.3816,
+      "step": 4972
+    },
+    {
+      "epoch": 6.35968,
+      "grad_norm": 0.5142550468444824,
+      "learning_rate": 0.00011500874949522141,
+      "loss": 3.4407,
+      "step": 4973
+    },
+    {
+      "epoch": 6.36096,
+      "grad_norm": 0.5232443809509277,
+      "learning_rate": 0.00011496836720958404,
+      "loss": 3.4582,
+      "step": 4974
+    },
+    {
+      "epoch": 6.36224,
+      "grad_norm": 0.5081977844238281,
+      "learning_rate": 0.00011492798492394669,
+      "loss": 3.4799,
+      "step": 4975
+    },
+    {
+      "epoch": 6.36352,
+      "grad_norm": 0.49449968338012695,
+      "learning_rate": 0.00011488760263830932,
+      "loss": 3.4118,
+      "step": 4976
+    },
+    {
+      "epoch": 6.3648,
+      "grad_norm": 0.5099364519119263,
+      "learning_rate": 0.00011484722035267195,
+      "loss": 3.4475,
+      "step": 4977
+    },
+    {
+      "epoch": 6.36608,
+      "grad_norm": 0.5084264278411865,
+      "learning_rate": 0.00011480683806703458,
+      "loss": 3.5485,
+      "step": 4978
+    },
+    {
+      "epoch": 6.36736,
+      "grad_norm": 0.5079696178436279,
+      "learning_rate": 0.00011476645578139722,
+      "loss": 3.4334,
+      "step": 4979
+    },
+    {
+      "epoch": 6.36864,
+      "grad_norm": 0.5050063729286194,
+      "learning_rate": 0.00011472607349575985,
+      "loss": 3.4747,
+      "step": 4980
+    },
+    {
+      "epoch": 6.3699200000000005,
+      "grad_norm": 0.5045068264007568,
+      "learning_rate": 0.00011468569121012248,
+      "loss": 3.4687,
+      "step": 4981
+    },
+    {
+      "epoch": 6.3712,
+      "grad_norm": 0.5139387845993042,
+      "learning_rate": 0.00011464530892448511,
+      "loss": 3.4448,
+      "step": 4982
+    },
+    {
+      "epoch": 6.37248,
+      "grad_norm": 0.5034843683242798,
+      "learning_rate": 0.00011460492663884776,
+      "loss": 3.4381,
+      "step": 4983
+    },
+    {
+      "epoch": 6.37376,
+      "grad_norm": 0.5243584513664246,
+      "learning_rate": 0.00011456454435321039,
+      "loss": 3.5206,
+      "step": 4984
+    },
+    {
+      "epoch": 6.37504,
+      "grad_norm": 0.5162044167518616,
+      "learning_rate": 0.00011452416206757302,
+      "loss": 3.5582,
+      "step": 4985
+    },
+    {
+      "epoch": 6.37632,
+      "grad_norm": 0.49788084626197815,
+      "learning_rate": 0.00011448377978193565,
+      "loss": 3.4328,
+      "step": 4986
+    },
+    {
+      "epoch": 6.3776,
+      "grad_norm": 0.5192862153053284,
+      "learning_rate": 0.00011444339749629828,
+      "loss": 3.4449,
+      "step": 4987
+    },
+    {
+      "epoch": 6.37888,
+      "grad_norm": 0.5089589357376099,
+      "learning_rate": 0.00011440301521066092,
+      "loss": 3.4463,
+      "step": 4988
+    },
+    {
+      "epoch": 6.38016,
+      "grad_norm": 0.5230059623718262,
+      "learning_rate": 0.00011436263292502354,
+      "loss": 3.4992,
+      "step": 4989
+    },
+    {
+      "epoch": 6.38144,
+      "grad_norm": 0.5021970868110657,
+      "learning_rate": 0.00011432225063938618,
+      "loss": 3.461,
+      "step": 4990
+    },
+    {
+      "epoch": 6.38272,
+      "grad_norm": 0.5131734013557434,
+      "learning_rate": 0.00011428186835374881,
+      "loss": 3.4653,
+      "step": 4991
+    },
+    {
+      "epoch": 6.384,
+      "grad_norm": 0.49421629309654236,
+      "learning_rate": 0.00011424148606811145,
+      "loss": 3.4365,
+      "step": 4992
+    },
+    {
+      "epoch": 6.38528,
+      "grad_norm": 0.5171757340431213,
+      "learning_rate": 0.00011420110378247407,
+      "loss": 3.4412,
+      "step": 4993
+    },
+    {
+      "epoch": 6.38656,
+      "grad_norm": 0.530982494354248,
+      "learning_rate": 0.00011416072149683671,
+      "loss": 3.4623,
+      "step": 4994
+    },
+    {
+      "epoch": 6.38784,
+      "grad_norm": 0.5211870670318604,
+      "learning_rate": 0.00011412033921119934,
+      "loss": 3.4505,
+      "step": 4995
+    },
+    {
+      "epoch": 6.38912,
+      "grad_norm": 0.5216143131256104,
+      "learning_rate": 0.00011407995692556199,
+      "loss": 3.4959,
+      "step": 4996
+    },
+    {
+      "epoch": 6.3904,
+      "grad_norm": 0.5288113355636597,
+      "learning_rate": 0.0001140395746399246,
+      "loss": 3.4404,
+      "step": 4997
+    },
+    {
+      "epoch": 6.39168,
+      "grad_norm": 0.5123031735420227,
+      "learning_rate": 0.00011399919235428725,
+      "loss": 3.449,
+      "step": 4998
+    },
+    {
+      "epoch": 6.39296,
+      "grad_norm": 0.49802166223526,
+      "learning_rate": 0.00011395881006864988,
+      "loss": 3.3765,
+      "step": 4999
+    },
+    {
+      "epoch": 6.39424,
+      "grad_norm": 0.5179408192634583,
+      "learning_rate": 0.0001139184277830125,
+      "loss": 3.4644,
+      "step": 5000
+    },
+    {
+      "epoch": 6.39552,
+      "grad_norm": 0.5292978286743164,
+      "learning_rate": 0.00011387804549737514,
+      "loss": 3.489,
+      "step": 5001
+    },
+    {
+      "epoch": 6.3968,
+      "grad_norm": 0.5212100744247437,
+      "learning_rate": 0.00011383766321173777,
+      "loss": 3.4714,
+      "step": 5002
+    },
+    {
+      "epoch": 6.39808,
+      "grad_norm": 0.5056613087654114,
+      "learning_rate": 0.00011379728092610041,
+      "loss": 3.4111,
+      "step": 5003
+    },
+    {
+      "epoch": 6.39936,
+      "grad_norm": 0.5097994804382324,
+      "learning_rate": 0.00011375689864046304,
+      "loss": 3.4453,
+      "step": 5004
+    },
+    {
+      "epoch": 6.40064,
+      "grad_norm": 0.5117822885513306,
+      "learning_rate": 0.00011371651635482567,
+      "loss": 3.483,
+      "step": 5005
+    },
+    {
+      "epoch": 6.40192,
+      "grad_norm": 0.51289302110672,
+      "learning_rate": 0.0001136761340691883,
+      "loss": 3.5011,
+      "step": 5006
+    },
+    {
+      "epoch": 6.4032,
+      "grad_norm": 0.5105553865432739,
+      "learning_rate": 0.00011363575178355094,
+      "loss": 3.4876,
+      "step": 5007
+    },
+    {
+      "epoch": 6.40448,
+      "grad_norm": 0.5061222314834595,
+      "learning_rate": 0.00011359536949791357,
+      "loss": 3.4529,
+      "step": 5008
+    },
+    {
+      "epoch": 6.40576,
+      "grad_norm": 0.5081943869590759,
+      "learning_rate": 0.0001135549872122762,
+      "loss": 3.5208,
+      "step": 5009
+    },
+    {
+      "epoch": 6.40704,
+      "grad_norm": 0.5123940706253052,
+      "learning_rate": 0.00011351460492663883,
+      "loss": 3.4691,
+      "step": 5010
+    },
+    {
+      "epoch": 6.40832,
+      "grad_norm": 0.5150337219238281,
+      "learning_rate": 0.00011347422264100148,
+      "loss": 3.5093,
+      "step": 5011
+    },
+    {
+      "epoch": 6.4096,
+      "grad_norm": 0.5125681161880493,
+      "learning_rate": 0.00011343384035536411,
+      "loss": 3.4547,
+      "step": 5012
+    },
+    {
+      "epoch": 6.41088,
+      "grad_norm": 0.509105384349823,
+      "learning_rate": 0.00011339345806972674,
+      "loss": 3.4142,
+      "step": 5013
+    },
+    {
+      "epoch": 6.41216,
+      "grad_norm": 0.5203231573104858,
+      "learning_rate": 0.00011335307578408937,
+      "loss": 3.4432,
+      "step": 5014
+    },
+    {
+      "epoch": 6.41344,
+      "grad_norm": 0.5590437650680542,
+      "learning_rate": 0.000113312693498452,
+      "loss": 3.4524,
+      "step": 5015
+    },
+    {
+      "epoch": 6.41472,
+      "grad_norm": 0.5066128969192505,
+      "learning_rate": 0.00011327231121281464,
+      "loss": 3.5085,
+      "step": 5016
+    },
+    {
+      "epoch": 6.416,
+      "grad_norm": 0.5234223008155823,
+      "learning_rate": 0.00011323192892717726,
+      "loss": 3.5202,
+      "step": 5017
+    },
+    {
+      "epoch": 6.41728,
+      "grad_norm": 0.49386492371559143,
+      "learning_rate": 0.0001131915466415399,
+      "loss": 3.4921,
+      "step": 5018
+    },
+    {
+      "epoch": 6.41856,
+      "grad_norm": 0.5142998695373535,
+      "learning_rate": 0.00011315116435590253,
+      "loss": 3.4908,
+      "step": 5019
+    },
+    {
+      "epoch": 6.41984,
+      "grad_norm": 0.5108798742294312,
+      "learning_rate": 0.00011311078207026518,
+      "loss": 3.3707,
+      "step": 5020
+    },
+    {
+      "epoch": 6.42112,
+      "grad_norm": 0.5349586009979248,
+      "learning_rate": 0.00011307039978462779,
+      "loss": 3.5232,
+      "step": 5021
+    },
+    {
+      "epoch": 6.4224,
+      "grad_norm": 0.5002762675285339,
+      "learning_rate": 0.00011303001749899044,
+      "loss": 3.4531,
+      "step": 5022
+    },
+    {
+      "epoch": 6.42368,
+      "grad_norm": 0.5073869228363037,
+      "learning_rate": 0.00011298963521335307,
+      "loss": 3.4888,
+      "step": 5023
+    },
+    {
+      "epoch": 6.4249600000000004,
+      "grad_norm": 0.5094159245491028,
+      "learning_rate": 0.00011294925292771571,
+      "loss": 3.4729,
+      "step": 5024
+    },
+    {
+      "epoch": 6.42624,
+      "grad_norm": 0.502030074596405,
+      "learning_rate": 0.00011290887064207833,
+      "loss": 3.4268,
+      "step": 5025
+    },
+    {
+      "epoch": 6.42752,
+      "grad_norm": 0.5096508264541626,
+      "learning_rate": 0.00011286848835644097,
+      "loss": 3.4771,
+      "step": 5026
+    },
+    {
+      "epoch": 6.4288,
+      "grad_norm": 0.5121338367462158,
+      "learning_rate": 0.0001128281060708036,
+      "loss": 3.5188,
+      "step": 5027
+    },
+    {
+      "epoch": 6.43008,
+      "grad_norm": 0.5106524229049683,
+      "learning_rate": 0.00011278772378516622,
+      "loss": 3.487,
+      "step": 5028
+    },
+    {
+      "epoch": 6.43136,
+      "grad_norm": 0.5135172605514526,
+      "learning_rate": 0.00011274734149952886,
+      "loss": 3.4631,
+      "step": 5029
+    },
+    {
+      "epoch": 6.43264,
+      "grad_norm": 0.5284242033958435,
+      "learning_rate": 0.00011270695921389149,
+      "loss": 3.5371,
+      "step": 5030
+    },
+    {
+      "epoch": 6.43392,
+      "grad_norm": 0.49686378240585327,
+      "learning_rate": 0.00011266657692825413,
+      "loss": 3.4035,
+      "step": 5031
+    },
+    {
+      "epoch": 6.4352,
+      "grad_norm": 0.5379998683929443,
+      "learning_rate": 0.00011262619464261675,
+      "loss": 3.4359,
+      "step": 5032
+    },
+    {
+      "epoch": 6.4364799999999995,
+      "grad_norm": 0.5207564830780029,
+      "learning_rate": 0.0001125858123569794,
+      "loss": 3.433,
+      "step": 5033
+    },
+    {
+      "epoch": 6.43776,
+      "grad_norm": 0.49627432227134705,
+      "learning_rate": 0.00011254543007134202,
+      "loss": 3.4353,
+      "step": 5034
+    },
+    {
+      "epoch": 6.43904,
+      "grad_norm": 0.5221447944641113,
+      "learning_rate": 0.00011250504778570467,
+      "loss": 3.4024,
+      "step": 5035
+    },
+    {
+      "epoch": 6.44032,
+      "grad_norm": 0.5198942422866821,
+      "learning_rate": 0.00011246466550006728,
+      "loss": 3.4782,
+      "step": 5036
+    },
+    {
+      "epoch": 6.4416,
+      "grad_norm": 0.5337004661560059,
+      "learning_rate": 0.00011242428321442993,
+      "loss": 3.5193,
+      "step": 5037
+    },
+    {
+      "epoch": 6.44288,
+      "grad_norm": 0.5141910314559937,
+      "learning_rate": 0.00011238390092879256,
+      "loss": 3.5241,
+      "step": 5038
+    },
+    {
+      "epoch": 6.44416,
+      "grad_norm": 0.5050771832466125,
+      "learning_rate": 0.0001123435186431552,
+      "loss": 3.4287,
+      "step": 5039
+    },
+    {
+      "epoch": 6.44544,
+      "grad_norm": 0.5285760164260864,
+      "learning_rate": 0.00011230313635751783,
+      "loss": 3.4792,
+      "step": 5040
+    },
+    {
+      "epoch": 6.44672,
+      "grad_norm": 0.5192365646362305,
+      "learning_rate": 0.00011226275407188045,
+      "loss": 3.4925,
+      "step": 5041
+    },
+    {
+      "epoch": 6.448,
+      "grad_norm": 0.5274104475975037,
+      "learning_rate": 0.00011222237178624309,
+      "loss": 3.444,
+      "step": 5042
+    },
+    {
+      "epoch": 6.44928,
+      "grad_norm": 0.5355402231216431,
+      "learning_rate": 0.00011218198950060572,
+      "loss": 3.5535,
+      "step": 5043
+    },
+    {
+      "epoch": 6.45056,
+      "grad_norm": 0.5089603662490845,
+      "learning_rate": 0.00011214160721496837,
+      "loss": 3.4749,
+      "step": 5044
+    },
+    {
+      "epoch": 6.45184,
+      "grad_norm": 0.5029155611991882,
+      "learning_rate": 0.00011210122492933098,
+      "loss": 3.3917,
+      "step": 5045
+    },
+    {
+      "epoch": 6.45312,
+      "grad_norm": 0.5419932007789612,
+      "learning_rate": 0.00011206084264369363,
+      "loss": 3.4852,
+      "step": 5046
+    },
+    {
+      "epoch": 6.4544,
+      "grad_norm": 0.5222985148429871,
+      "learning_rate": 0.00011202046035805626,
+      "loss": 3.4503,
+      "step": 5047
+    },
+    {
+      "epoch": 6.45568,
+      "grad_norm": 0.5103903412818909,
+      "learning_rate": 0.0001119800780724189,
+      "loss": 3.3846,
+      "step": 5048
+    },
+    {
+      "epoch": 6.45696,
+      "grad_norm": 0.5334930419921875,
+      "learning_rate": 0.00011193969578678152,
+      "loss": 3.4451,
+      "step": 5049
+    },
+    {
+      "epoch": 6.45824,
+      "grad_norm": 0.506489098072052,
+      "learning_rate": 0.00011189931350114416,
+      "loss": 3.4588,
+      "step": 5050
+    },
+    {
+      "epoch": 6.45952,
+      "grad_norm": 0.5101386308670044,
+      "learning_rate": 0.00011185893121550679,
+      "loss": 3.5153,
+      "step": 5051
+    },
+    {
+      "epoch": 6.4608,
+      "grad_norm": 0.5294933915138245,
+      "learning_rate": 0.00011181854892986943,
+      "loss": 3.4905,
+      "step": 5052
+    },
+    {
+      "epoch": 6.46208,
+      "grad_norm": 0.5064314603805542,
+      "learning_rate": 0.00011177816664423205,
+      "loss": 3.4571,
+      "step": 5053
+    },
+    {
+      "epoch": 6.46336,
+      "grad_norm": 0.526623010635376,
+      "learning_rate": 0.00011173778435859468,
+      "loss": 3.4502,
+      "step": 5054
+    },
+    {
+      "epoch": 6.46464,
+      "grad_norm": 0.5297154784202576,
+      "learning_rate": 0.00011169740207295732,
+      "loss": 3.5096,
+      "step": 5055
+    },
+    {
+      "epoch": 6.46592,
+      "grad_norm": 0.5068486332893372,
+      "learning_rate": 0.00011165701978731994,
+      "loss": 3.4319,
+      "step": 5056
+    },
+    {
+      "epoch": 6.4672,
+      "grad_norm": 0.5235540866851807,
+      "learning_rate": 0.00011161663750168258,
+      "loss": 3.4706,
+      "step": 5057
+    },
+    {
+      "epoch": 6.46848,
+      "grad_norm": 0.5198733806610107,
+      "learning_rate": 0.00011157625521604521,
+      "loss": 3.487,
+      "step": 5058
+    },
+    {
+      "epoch": 6.46976,
+      "grad_norm": 0.5035125613212585,
+      "learning_rate": 0.00011153587293040786,
+      "loss": 3.4629,
+      "step": 5059
+    },
+    {
+      "epoch": 6.47104,
+      "grad_norm": 0.5046782493591309,
+      "learning_rate": 0.00011149549064477047,
+      "loss": 3.4192,
+      "step": 5060
+    },
+    {
+      "epoch": 6.47232,
+      "grad_norm": 0.50188809633255,
+      "learning_rate": 0.00011145510835913312,
+      "loss": 3.4571,
+      "step": 5061
+    },
+    {
+      "epoch": 6.4736,
+      "grad_norm": 0.5252716541290283,
+      "learning_rate": 0.00011141472607349575,
+      "loss": 3.4554,
+      "step": 5062
+    },
+    {
+      "epoch": 6.47488,
+      "grad_norm": 0.4985215365886688,
+      "learning_rate": 0.00011137434378785839,
+      "loss": 3.4759,
+      "step": 5063
+    },
+    {
+      "epoch": 6.47616,
+      "grad_norm": 0.5182859897613525,
+      "learning_rate": 0.00011133396150222101,
+      "loss": 3.4293,
+      "step": 5064
+    },
+    {
+      "epoch": 6.47744,
+      "grad_norm": 0.5109648108482361,
+      "learning_rate": 0.00011129357921658365,
+      "loss": 3.449,
+      "step": 5065
+    },
+    {
+      "epoch": 6.47872,
+      "grad_norm": 0.4976121485233307,
+      "learning_rate": 0.00011125319693094628,
+      "loss": 3.4391,
+      "step": 5066
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.501159131526947,
+      "learning_rate": 0.00011121281464530893,
+      "loss": 3.4683,
+      "step": 5067
+    },
+    {
+      "epoch": 6.48128,
+      "grad_norm": 0.5295117497444153,
+      "learning_rate": 0.00011117243235967154,
+      "loss": 3.4945,
+      "step": 5068
+    },
+    {
+      "epoch": 6.48256,
+      "grad_norm": 0.5260968804359436,
+      "learning_rate": 0.00011113205007403417,
+      "loss": 3.4764,
+      "step": 5069
+    },
+    {
+      "epoch": 6.48384,
+      "grad_norm": 0.5069438219070435,
+      "learning_rate": 0.00011109166778839682,
+      "loss": 3.4699,
+      "step": 5070
+    },
+    {
+      "epoch": 6.48512,
+      "grad_norm": 0.5271986126899719,
+      "learning_rate": 0.00011105128550275945,
+      "loss": 3.5268,
+      "step": 5071
+    },
+    {
+      "epoch": 6.4864,
+      "grad_norm": 0.5156261324882507,
+      "learning_rate": 0.00011101090321712208,
+      "loss": 3.4877,
+      "step": 5072
+    },
+    {
+      "epoch": 6.48768,
+      "grad_norm": 0.5246565937995911,
+      "learning_rate": 0.0001109705209314847,
+      "loss": 3.4697,
+      "step": 5073
+    },
+    {
+      "epoch": 6.48896,
+      "grad_norm": 0.5298268795013428,
+      "learning_rate": 0.00011093013864584735,
+      "loss": 3.4842,
+      "step": 5074
+    },
+    {
+      "epoch": 6.49024,
+      "grad_norm": 0.5147030353546143,
+      "learning_rate": 0.00011088975636020998,
+      "loss": 3.5063,
+      "step": 5075
+    },
+    {
+      "epoch": 6.49152,
+      "grad_norm": 0.5172019004821777,
+      "learning_rate": 0.00011084937407457262,
+      "loss": 3.4344,
+      "step": 5076
+    },
+    {
+      "epoch": 6.4928,
+      "grad_norm": 0.5194494724273682,
+      "learning_rate": 0.00011080899178893524,
+      "loss": 3.4808,
+      "step": 5077
+    },
+    {
+      "epoch": 6.49408,
+      "grad_norm": 0.5049477815628052,
+      "learning_rate": 0.00011076860950329788,
+      "loss": 3.4354,
+      "step": 5078
+    },
+    {
+      "epoch": 6.49536,
+      "grad_norm": 0.5107006430625916,
+      "learning_rate": 0.00011072822721766051,
+      "loss": 3.4447,
+      "step": 5079
+    },
+    {
+      "epoch": 6.49664,
+      "grad_norm": 0.5118290185928345,
+      "learning_rate": 0.00011068784493202316,
+      "loss": 3.4469,
+      "step": 5080
+    },
+    {
+      "epoch": 6.49792,
+      "grad_norm": 0.5051876306533813,
+      "learning_rate": 0.00011064746264638577,
+      "loss": 3.4521,
+      "step": 5081
+    },
+    {
+      "epoch": 6.4992,
+      "grad_norm": 0.4978385865688324,
+      "learning_rate": 0.0001106070803607484,
+      "loss": 3.3929,
+      "step": 5082
+    },
+    {
+      "epoch": 6.50048,
+      "grad_norm": 0.5098656415939331,
+      "learning_rate": 0.00011056669807511105,
+      "loss": 3.4236,
+      "step": 5083
+    },
+    {
+      "epoch": 6.50176,
+      "grad_norm": 0.5078462362289429,
+      "learning_rate": 0.00011052631578947366,
+      "loss": 3.4781,
+      "step": 5084
+    },
+    {
+      "epoch": 6.50304,
+      "grad_norm": 0.5070273876190186,
+      "learning_rate": 0.00011048593350383631,
+      "loss": 3.4513,
+      "step": 5085
+    },
+    {
+      "epoch": 6.50432,
+      "grad_norm": 0.4959174394607544,
+      "learning_rate": 0.00011044555121819894,
+      "loss": 3.4036,
+      "step": 5086
+    },
+    {
+      "epoch": 6.5056,
+      "grad_norm": 0.5386160016059875,
+      "learning_rate": 0.00011040516893256158,
+      "loss": 3.5321,
+      "step": 5087
+    },
+    {
+      "epoch": 6.50688,
+      "grad_norm": 0.518519937992096,
+      "learning_rate": 0.0001103647866469242,
+      "loss": 3.4958,
+      "step": 5088
+    },
+    {
+      "epoch": 6.50816,
+      "grad_norm": 0.5082813501358032,
+      "learning_rate": 0.00011032440436128684,
+      "loss": 3.4061,
+      "step": 5089
+    },
+    {
+      "epoch": 6.50944,
+      "grad_norm": 0.5264861583709717,
+      "learning_rate": 0.00011028402207564947,
+      "loss": 3.4816,
+      "step": 5090
+    },
+    {
+      "epoch": 6.51072,
+      "grad_norm": 0.5055463910102844,
+      "learning_rate": 0.00011024363979001212,
+      "loss": 3.5055,
+      "step": 5091
+    },
+    {
+      "epoch": 6.5120000000000005,
+      "grad_norm": 0.5075529217720032,
+      "learning_rate": 0.00011020325750437473,
+      "loss": 3.4639,
+      "step": 5092
+    },
+    {
+      "epoch": 6.51328,
+      "grad_norm": 0.5140883326530457,
+      "learning_rate": 0.00011016287521873738,
+      "loss": 3.4615,
+      "step": 5093
+    },
+    {
+      "epoch": 6.51456,
+      "grad_norm": 0.5120643377304077,
+      "learning_rate": 0.0001101224929331,
+      "loss": 3.4331,
+      "step": 5094
+    },
+    {
+      "epoch": 6.51584,
+      "grad_norm": 0.501859724521637,
+      "learning_rate": 0.00011008211064746264,
+      "loss": 3.4614,
+      "step": 5095
+    },
+    {
+      "epoch": 6.51712,
+      "grad_norm": 0.514984667301178,
+      "learning_rate": 0.00011004172836182527,
+      "loss": 3.4491,
+      "step": 5096
+    },
+    {
+      "epoch": 6.5184,
+      "grad_norm": 0.5052075386047363,
+      "learning_rate": 0.0001100013460761879,
+      "loss": 3.5753,
+      "step": 5097
+    },
+    {
+      "epoch": 6.51968,
+      "grad_norm": 0.5124235153198242,
+      "learning_rate": 0.00010996096379055054,
+      "loss": 3.4338,
+      "step": 5098
+    },
+    {
+      "epoch": 6.52096,
+      "grad_norm": 0.48975175619125366,
+      "learning_rate": 0.00010992058150491317,
+      "loss": 3.4508,
+      "step": 5099
+    },
+    {
+      "epoch": 6.52224,
+      "grad_norm": 0.49977898597717285,
+      "learning_rate": 0.0001098801992192758,
+      "loss": 3.3994,
+      "step": 5100
+    },
+    {
+      "epoch": 6.5235199999999995,
+      "grad_norm": 0.4916556775569916,
+      "learning_rate": 0.00010983981693363843,
+      "loss": 3.4289,
+      "step": 5101
+    },
+    {
+      "epoch": 6.5248,
+      "grad_norm": 0.5177762508392334,
+      "learning_rate": 0.00010979943464800107,
+      "loss": 3.5018,
+      "step": 5102
+    },
+    {
+      "epoch": 6.52608,
+      "grad_norm": 0.5140178799629211,
+      "learning_rate": 0.0001097590523623637,
+      "loss": 3.4727,
+      "step": 5103
+    },
+    {
+      "epoch": 6.52736,
+      "grad_norm": 0.4911567270755768,
+      "learning_rate": 0.00010971867007672633,
+      "loss": 3.5407,
+      "step": 5104
+    },
+    {
+      "epoch": 6.52864,
+      "grad_norm": 0.505172610282898,
+      "learning_rate": 0.00010967828779108896,
+      "loss": 3.3872,
+      "step": 5105
+    },
+    {
+      "epoch": 6.52992,
+      "grad_norm": 0.5191845297813416,
+      "learning_rate": 0.00010963790550545161,
+      "loss": 3.482,
+      "step": 5106
+    },
+    {
+      "epoch": 6.5312,
+      "grad_norm": 0.5074360370635986,
+      "learning_rate": 0.00010959752321981424,
+      "loss": 3.4597,
+      "step": 5107
+    },
+    {
+      "epoch": 6.53248,
+      "grad_norm": 0.5102279782295227,
+      "learning_rate": 0.00010955714093417685,
+      "loss": 3.3886,
+      "step": 5108
+    },
+    {
+      "epoch": 6.53376,
+      "grad_norm": 0.5071060061454773,
+      "learning_rate": 0.0001095167586485395,
+      "loss": 3.4508,
+      "step": 5109
+    },
+    {
+      "epoch": 6.53504,
+      "grad_norm": 0.5164380669593811,
+      "learning_rate": 0.00010947637636290213,
+      "loss": 3.446,
+      "step": 5110
+    },
+    {
+      "epoch": 6.53632,
+      "grad_norm": 0.5028985142707825,
+      "learning_rate": 0.00010943599407726477,
+      "loss": 3.5139,
+      "step": 5111
+    },
+    {
+      "epoch": 6.5376,
+      "grad_norm": 0.49001002311706543,
+      "learning_rate": 0.00010939561179162739,
+      "loss": 3.3891,
+      "step": 5112
+    },
+    {
+      "epoch": 6.53888,
+      "grad_norm": 0.5156526565551758,
+      "learning_rate": 0.00010935522950599003,
+      "loss": 3.4913,
+      "step": 5113
+    },
+    {
+      "epoch": 6.54016,
+      "grad_norm": 0.4975212514400482,
+      "learning_rate": 0.00010931484722035266,
+      "loss": 3.4462,
+      "step": 5114
+    },
+    {
+      "epoch": 6.54144,
+      "grad_norm": 0.5028743147850037,
+      "learning_rate": 0.0001092744649347153,
+      "loss": 3.396,
+      "step": 5115
+    },
+    {
+      "epoch": 6.54272,
+      "grad_norm": 0.5105941295623779,
+      "learning_rate": 0.00010923408264907792,
+      "loss": 3.5302,
+      "step": 5116
+    },
+    {
+      "epoch": 6.5440000000000005,
+      "grad_norm": 0.5078137516975403,
+      "learning_rate": 0.00010919370036344056,
+      "loss": 3.4376,
+      "step": 5117
+    },
+    {
+      "epoch": 6.54528,
+      "grad_norm": 0.4959731101989746,
+      "learning_rate": 0.0001091533180778032,
+      "loss": 3.4079,
+      "step": 5118
+    },
+    {
+      "epoch": 6.54656,
+      "grad_norm": 0.5086926817893982,
+      "learning_rate": 0.00010911293579216584,
+      "loss": 3.4873,
+      "step": 5119
+    },
+    {
+      "epoch": 6.54784,
+      "grad_norm": 0.5077411532402039,
+      "learning_rate": 0.00010907255350652845,
+      "loss": 3.4758,
+      "step": 5120
+    },
+    {
+      "epoch": 6.54912,
+      "grad_norm": 0.5054405331611633,
+      "learning_rate": 0.00010903217122089109,
+      "loss": 3.4363,
+      "step": 5121
+    },
+    {
+      "epoch": 6.5504,
+      "grad_norm": 0.5057424902915955,
+      "learning_rate": 0.00010899178893525373,
+      "loss": 3.5458,
+      "step": 5122
+    },
+    {
+      "epoch": 6.55168,
+      "grad_norm": 0.5203006267547607,
+      "learning_rate": 0.00010895140664961635,
+      "loss": 3.5374,
+      "step": 5123
+    },
+    {
+      "epoch": 6.55296,
+      "grad_norm": 0.5345029830932617,
+      "learning_rate": 0.00010891102436397899,
+      "loss": 3.4367,
+      "step": 5124
+    },
+    {
+      "epoch": 6.55424,
+      "grad_norm": 0.5119213461875916,
+      "learning_rate": 0.00010887064207834162,
+      "loss": 3.4897,
+      "step": 5125
+    },
+    {
+      "epoch": 6.55552,
+      "grad_norm": 0.5316451787948608,
+      "learning_rate": 0.00010883025979270426,
+      "loss": 3.4209,
+      "step": 5126
+    },
+    {
+      "epoch": 6.5568,
+      "grad_norm": 0.5099210739135742,
+      "learning_rate": 0.00010878987750706689,
+      "loss": 3.418,
+      "step": 5127
+    },
+    {
+      "epoch": 6.55808,
+      "grad_norm": 0.5128821730613708,
+      "learning_rate": 0.00010874949522142952,
+      "loss": 3.4507,
+      "step": 5128
+    },
+    {
+      "epoch": 6.55936,
+      "grad_norm": 0.5047348141670227,
+      "learning_rate": 0.00010870911293579215,
+      "loss": 3.4232,
+      "step": 5129
+    },
+    {
+      "epoch": 6.56064,
+      "grad_norm": 0.514415979385376,
+      "learning_rate": 0.0001086687306501548,
+      "loss": 3.5885,
+      "step": 5130
+    },
+    {
+      "epoch": 6.56192,
+      "grad_norm": 0.5269995927810669,
+      "learning_rate": 0.00010862834836451743,
+      "loss": 3.5069,
+      "step": 5131
+    },
+    {
+      "epoch": 6.5632,
+      "grad_norm": 0.5068910717964172,
+      "learning_rate": 0.00010858796607888006,
+      "loss": 3.427,
+      "step": 5132
+    },
+    {
+      "epoch": 6.56448,
+      "grad_norm": 0.514909565448761,
+      "learning_rate": 0.00010854758379324269,
+      "loss": 3.5029,
+      "step": 5133
+    },
+    {
+      "epoch": 6.56576,
+      "grad_norm": 0.5140718817710876,
+      "learning_rate": 0.00010850720150760533,
+      "loss": 3.4053,
+      "step": 5134
+    },
+    {
+      "epoch": 6.56704,
+      "grad_norm": 0.524402379989624,
+      "learning_rate": 0.00010846681922196796,
+      "loss": 3.523,
+      "step": 5135
+    },
+    {
+      "epoch": 6.56832,
+      "grad_norm": 0.519688069820404,
+      "learning_rate": 0.00010842643693633058,
+      "loss": 3.4752,
+      "step": 5136
+    },
+    {
+      "epoch": 6.5696,
+      "grad_norm": 0.5132143497467041,
+      "learning_rate": 0.00010838605465069322,
+      "loss": 3.4902,
+      "step": 5137
+    },
+    {
+      "epoch": 6.57088,
+      "grad_norm": 0.5264431834220886,
+      "learning_rate": 0.00010834567236505585,
+      "loss": 3.5565,
+      "step": 5138
+    },
+    {
+      "epoch": 6.57216,
+      "grad_norm": 0.5338211059570312,
+      "learning_rate": 0.0001083052900794185,
+      "loss": 3.4898,
+      "step": 5139
+    },
+    {
+      "epoch": 6.57344,
+      "grad_norm": 0.5249161720275879,
+      "learning_rate": 0.00010826490779378111,
+      "loss": 3.5718,
+      "step": 5140
+    },
+    {
+      "epoch": 6.57472,
+      "grad_norm": 0.5375153422355652,
+      "learning_rate": 0.00010822452550814375,
+      "loss": 3.3902,
+      "step": 5141
+    },
+    {
+      "epoch": 6.576,
+      "grad_norm": 0.5202714204788208,
+      "learning_rate": 0.00010818414322250638,
+      "loss": 3.4019,
+      "step": 5142
+    },
+    {
+      "epoch": 6.57728,
+      "grad_norm": 0.5213868618011475,
+      "learning_rate": 0.00010814376093686903,
+      "loss": 3.5119,
+      "step": 5143
+    },
+    {
+      "epoch": 6.5785599999999995,
+      "grad_norm": 0.5079566240310669,
+      "learning_rate": 0.00010810337865123164,
+      "loss": 3.4668,
+      "step": 5144
+    },
+    {
+      "epoch": 6.57984,
+      "grad_norm": 0.5234741568565369,
+      "learning_rate": 0.00010806299636559429,
+      "loss": 3.4789,
+      "step": 5145
+    },
+    {
+      "epoch": 6.58112,
+      "grad_norm": 0.5120872259140015,
+      "learning_rate": 0.00010802261407995692,
+      "loss": 3.4737,
+      "step": 5146
+    },
+    {
+      "epoch": 6.5824,
+      "grad_norm": 0.5126521587371826,
+      "learning_rate": 0.00010798223179431956,
+      "loss": 3.4625,
+      "step": 5147
+    },
+    {
+      "epoch": 6.58368,
+      "grad_norm": 0.5083807110786438,
+      "learning_rate": 0.00010794184950868218,
+      "loss": 3.5056,
+      "step": 5148
+    },
+    {
+      "epoch": 6.58496,
+      "grad_norm": 0.5182278752326965,
+      "learning_rate": 0.00010790146722304481,
+      "loss": 3.4856,
+      "step": 5149
+    },
+    {
+      "epoch": 6.58624,
+      "grad_norm": 0.5055496096611023,
+      "learning_rate": 0.00010786108493740745,
+      "loss": 3.4205,
+      "step": 5150
+    },
+    {
+      "epoch": 6.58752,
+      "grad_norm": 0.5119015574455261,
+      "learning_rate": 0.00010782070265177007,
+      "loss": 3.4647,
+      "step": 5151
+    },
+    {
+      "epoch": 6.5888,
+      "grad_norm": 0.5228475332260132,
+      "learning_rate": 0.00010778032036613271,
+      "loss": 3.4117,
+      "step": 5152
+    },
+    {
+      "epoch": 6.59008,
+      "grad_norm": 0.5190859436988831,
+      "learning_rate": 0.00010773993808049534,
+      "loss": 3.4451,
+      "step": 5153
+    },
+    {
+      "epoch": 6.59136,
+      "grad_norm": 0.509190022945404,
+      "learning_rate": 0.00010769955579485799,
+      "loss": 3.4443,
+      "step": 5154
+    },
+    {
+      "epoch": 6.59264,
+      "grad_norm": 0.4972391724586487,
+      "learning_rate": 0.0001076591735092206,
+      "loss": 3.4357,
+      "step": 5155
+    },
+    {
+      "epoch": 6.59392,
+      "grad_norm": 0.5034371614456177,
+      "learning_rate": 0.00010761879122358325,
+      "loss": 3.4543,
+      "step": 5156
+    },
+    {
+      "epoch": 6.5952,
+      "grad_norm": 0.5267638564109802,
+      "learning_rate": 0.00010757840893794588,
+      "loss": 3.4192,
+      "step": 5157
+    },
+    {
+      "epoch": 6.59648,
+      "grad_norm": 0.4983810484409332,
+      "learning_rate": 0.00010753802665230852,
+      "loss": 3.3902,
+      "step": 5158
+    },
+    {
+      "epoch": 6.59776,
+      "grad_norm": 0.49326276779174805,
+      "learning_rate": 0.00010749764436667114,
+      "loss": 3.4052,
+      "step": 5159
+    },
+    {
+      "epoch": 6.5990400000000005,
+      "grad_norm": 0.526161789894104,
+      "learning_rate": 0.00010745726208103378,
+      "loss": 3.4829,
+      "step": 5160
+    },
+    {
+      "epoch": 6.60032,
+      "grad_norm": 0.49957576394081116,
+      "learning_rate": 0.00010741687979539641,
+      "loss": 3.3863,
+      "step": 5161
+    },
+    {
+      "epoch": 6.6016,
+      "grad_norm": 0.5143072009086609,
+      "learning_rate": 0.00010737649750975904,
+      "loss": 3.502,
+      "step": 5162
+    },
+    {
+      "epoch": 6.60288,
+      "grad_norm": 0.5053294897079468,
+      "learning_rate": 0.00010733611522412168,
+      "loss": 3.4453,
+      "step": 5163
+    },
+    {
+      "epoch": 6.60416,
+      "grad_norm": 0.49979978799819946,
+      "learning_rate": 0.0001072957329384843,
+      "loss": 3.4394,
+      "step": 5164
+    },
+    {
+      "epoch": 6.60544,
+      "grad_norm": 0.5157962441444397,
+      "learning_rate": 0.00010725535065284694,
+      "loss": 3.4921,
+      "step": 5165
+    },
+    {
+      "epoch": 6.60672,
+      "grad_norm": 0.5077204704284668,
+      "learning_rate": 0.00010721496836720957,
+      "loss": 3.4838,
+      "step": 5166
+    },
+    {
+      "epoch": 6.608,
+      "grad_norm": 0.5278066992759705,
+      "learning_rate": 0.00010717458608157222,
+      "loss": 3.4713,
+      "step": 5167
+    },
+    {
+      "epoch": 6.60928,
+      "grad_norm": 0.506314218044281,
+      "learning_rate": 0.00010713420379593483,
+      "loss": 3.5477,
+      "step": 5168
+    },
+    {
+      "epoch": 6.6105599999999995,
+      "grad_norm": 0.5169037580490112,
+      "learning_rate": 0.00010709382151029748,
+      "loss": 3.5052,
+      "step": 5169
+    },
+    {
+      "epoch": 6.61184,
+      "grad_norm": 0.4995163679122925,
+      "learning_rate": 0.00010705343922466011,
+      "loss": 3.4526,
+      "step": 5170
+    },
+    {
+      "epoch": 6.61312,
+      "grad_norm": 0.516046404838562,
+      "learning_rate": 0.00010701305693902275,
+      "loss": 3.4841,
+      "step": 5171
+    },
+    {
+      "epoch": 6.6144,
+      "grad_norm": 0.5097321271896362,
+      "learning_rate": 0.00010697267465338537,
+      "loss": 3.502,
+      "step": 5172
+    },
+    {
+      "epoch": 6.61568,
+      "grad_norm": 0.5002099871635437,
+      "learning_rate": 0.00010693229236774801,
+      "loss": 3.4349,
+      "step": 5173
+    },
+    {
+      "epoch": 6.61696,
+      "grad_norm": 0.5027880668640137,
+      "learning_rate": 0.00010689191008211064,
+      "loss": 3.4844,
+      "step": 5174
+    },
+    {
+      "epoch": 6.61824,
+      "grad_norm": 0.5286634564399719,
+      "learning_rate": 0.00010685152779647326,
+      "loss": 3.4879,
+      "step": 5175
+    },
+    {
+      "epoch": 6.61952,
+      "grad_norm": 0.5041953325271606,
+      "learning_rate": 0.0001068111455108359,
+      "loss": 3.4475,
+      "step": 5176
+    },
+    {
+      "epoch": 6.6208,
+      "grad_norm": 0.5107352137565613,
+      "learning_rate": 0.00010677076322519853,
+      "loss": 3.3829,
+      "step": 5177
+    },
+    {
+      "epoch": 6.62208,
+      "grad_norm": 0.52018803358078,
+      "learning_rate": 0.00010673038093956118,
+      "loss": 3.4578,
+      "step": 5178
+    },
+    {
+      "epoch": 6.62336,
+      "grad_norm": 0.5006090998649597,
+      "learning_rate": 0.00010668999865392379,
+      "loss": 3.4867,
+      "step": 5179
+    },
+    {
+      "epoch": 6.62464,
+      "grad_norm": 0.5116918683052063,
+      "learning_rate": 0.00010664961636828644,
+      "loss": 3.3972,
+      "step": 5180
+    },
+    {
+      "epoch": 6.62592,
+      "grad_norm": 0.5370351076126099,
+      "learning_rate": 0.00010660923408264907,
+      "loss": 3.4639,
+      "step": 5181
+    },
+    {
+      "epoch": 6.6272,
+      "grad_norm": 0.5006325244903564,
+      "learning_rate": 0.00010656885179701171,
+      "loss": 3.5232,
+      "step": 5182
+    },
+    {
+      "epoch": 6.62848,
+      "grad_norm": 0.5160982608795166,
+      "learning_rate": 0.00010652846951137433,
+      "loss": 3.4108,
+      "step": 5183
+    },
+    {
+      "epoch": 6.62976,
+      "grad_norm": 0.5148961544036865,
+      "learning_rate": 0.00010648808722573697,
+      "loss": 3.5224,
+      "step": 5184
+    },
+    {
+      "epoch": 6.6310400000000005,
+      "grad_norm": 0.4979293644428253,
+      "learning_rate": 0.0001064477049400996,
+      "loss": 3.4324,
+      "step": 5185
+    },
+    {
+      "epoch": 6.63232,
+      "grad_norm": 0.5204696655273438,
+      "learning_rate": 0.00010640732265446224,
+      "loss": 3.4534,
+      "step": 5186
+    },
+    {
+      "epoch": 6.6336,
+      "grad_norm": 0.5214214324951172,
+      "learning_rate": 0.00010636694036882486,
+      "loss": 3.5314,
+      "step": 5187
+    },
+    {
+      "epoch": 6.63488,
+      "grad_norm": 0.5053163766860962,
+      "learning_rate": 0.00010632655808318749,
+      "loss": 3.3898,
+      "step": 5188
+    },
+    {
+      "epoch": 6.63616,
+      "grad_norm": 0.5069863200187683,
+      "learning_rate": 0.00010628617579755013,
+      "loss": 3.4414,
+      "step": 5189
+    },
+    {
+      "epoch": 6.63744,
+      "grad_norm": 0.5157361030578613,
+      "learning_rate": 0.00010624579351191276,
+      "loss": 3.5128,
+      "step": 5190
+    },
+    {
+      "epoch": 6.63872,
+      "grad_norm": 0.500819206237793,
+      "learning_rate": 0.0001062054112262754,
+      "loss": 3.4923,
+      "step": 5191
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.5084678530693054,
+      "learning_rate": 0.00010616502894063802,
+      "loss": 3.4921,
+      "step": 5192
+    },
+    {
+      "epoch": 6.64128,
+      "grad_norm": 0.520411491394043,
+      "learning_rate": 0.00010612464665500067,
+      "loss": 3.4166,
+      "step": 5193
+    },
+    {
+      "epoch": 6.64256,
+      "grad_norm": 0.5096445083618164,
+      "learning_rate": 0.0001060842643693633,
+      "loss": 3.4323,
+      "step": 5194
+    },
+    {
+      "epoch": 6.64384,
+      "grad_norm": 0.503450870513916,
+      "learning_rate": 0.00010604388208372593,
+      "loss": 3.3905,
+      "step": 5195
+    },
+    {
+      "epoch": 6.64512,
+      "grad_norm": 0.5244500041007996,
+      "learning_rate": 0.00010600349979808856,
+      "loss": 3.5189,
+      "step": 5196
+    },
+    {
+      "epoch": 6.6464,
+      "grad_norm": 0.5261669158935547,
+      "learning_rate": 0.0001059631175124512,
+      "loss": 3.4409,
+      "step": 5197
+    },
+    {
+      "epoch": 6.64768,
+      "grad_norm": 0.5130569338798523,
+      "learning_rate": 0.00010592273522681383,
+      "loss": 3.4309,
+      "step": 5198
+    },
+    {
+      "epoch": 6.64896,
+      "grad_norm": 0.5225698947906494,
+      "learning_rate": 0.00010588235294117647,
+      "loss": 3.4626,
+      "step": 5199
+    },
+    {
+      "epoch": 6.65024,
+      "grad_norm": 0.5169934034347534,
+      "learning_rate": 0.00010584197065553909,
+      "loss": 3.465,
+      "step": 5200
+    },
+    {
+      "epoch": 6.65152,
+      "grad_norm": 0.49897530674934387,
+      "learning_rate": 0.00010580158836990174,
+      "loss": 3.4524,
+      "step": 5201
+    },
+    {
+      "epoch": 6.6528,
+      "grad_norm": 0.5103302001953125,
+      "learning_rate": 0.00010576120608426437,
+      "loss": 3.4629,
+      "step": 5202
+    },
+    {
+      "epoch": 6.65408,
+      "grad_norm": 0.5289749503135681,
+      "learning_rate": 0.00010572082379862698,
+      "loss": 3.4733,
+      "step": 5203
+    },
+    {
+      "epoch": 6.65536,
+      "grad_norm": 0.5066429376602173,
+      "learning_rate": 0.00010568044151298963,
+      "loss": 3.4266,
+      "step": 5204
+    },
+    {
+      "epoch": 6.65664,
+      "grad_norm": 0.5125609636306763,
+      "learning_rate": 0.00010564005922735226,
+      "loss": 3.403,
+      "step": 5205
+    },
+    {
+      "epoch": 6.65792,
+      "grad_norm": 0.5179828405380249,
+      "learning_rate": 0.0001055996769417149,
+      "loss": 3.4441,
+      "step": 5206
+    },
+    {
+      "epoch": 6.6592,
+      "grad_norm": 0.5193979740142822,
+      "learning_rate": 0.00010555929465607752,
+      "loss": 3.5631,
+      "step": 5207
+    },
+    {
+      "epoch": 6.66048,
+      "grad_norm": 0.5336554050445557,
+      "learning_rate": 0.00010551891237044016,
+      "loss": 3.5912,
+      "step": 5208
+    },
+    {
+      "epoch": 6.66176,
+      "grad_norm": 0.5226610898971558,
+      "learning_rate": 0.00010547853008480279,
+      "loss": 3.461,
+      "step": 5209
+    },
+    {
+      "epoch": 6.66304,
+      "grad_norm": 0.523395836353302,
+      "learning_rate": 0.00010543814779916543,
+      "loss": 3.4231,
+      "step": 5210
+    },
+    {
+      "epoch": 6.66432,
+      "grad_norm": 0.5272761583328247,
+      "learning_rate": 0.00010539776551352805,
+      "loss": 3.4297,
+      "step": 5211
+    },
+    {
+      "epoch": 6.6655999999999995,
+      "grad_norm": 0.5125918984413147,
+      "learning_rate": 0.00010535738322789069,
+      "loss": 3.4043,
+      "step": 5212
+    },
+    {
+      "epoch": 6.66688,
+      "grad_norm": 0.4985739290714264,
+      "learning_rate": 0.00010531700094225332,
+      "loss": 3.5057,
+      "step": 5213
+    },
+    {
+      "epoch": 6.66816,
+      "grad_norm": 0.5222551226615906,
+      "learning_rate": 0.00010527661865661597,
+      "loss": 3.4881,
+      "step": 5214
+    },
+    {
+      "epoch": 6.66944,
+      "grad_norm": 0.5044218301773071,
+      "learning_rate": 0.00010523623637097858,
+      "loss": 3.4112,
+      "step": 5215
+    },
+    {
+      "epoch": 6.67072,
+      "grad_norm": 0.5089839696884155,
+      "learning_rate": 0.00010519585408534121,
+      "loss": 3.4329,
+      "step": 5216
+    },
+    {
+      "epoch": 6.672,
+      "grad_norm": 0.5202680826187134,
+      "learning_rate": 0.00010515547179970386,
+      "loss": 3.5092,
+      "step": 5217
+    },
+    {
+      "epoch": 6.67328,
+      "grad_norm": 0.5112069845199585,
+      "learning_rate": 0.00010511508951406649,
+      "loss": 3.4684,
+      "step": 5218
+    },
+    {
+      "epoch": 6.67456,
+      "grad_norm": 0.5102248787879944,
+      "learning_rate": 0.00010507470722842912,
+      "loss": 3.5571,
+      "step": 5219
+    },
+    {
+      "epoch": 6.67584,
+      "grad_norm": 0.5237480998039246,
+      "learning_rate": 0.00010503432494279175,
+      "loss": 3.5169,
+      "step": 5220
+    },
+    {
+      "epoch": 6.67712,
+      "grad_norm": 0.5136720538139343,
+      "learning_rate": 0.00010499394265715439,
+      "loss": 3.4614,
+      "step": 5221
+    },
+    {
+      "epoch": 6.6784,
+      "grad_norm": 0.5107023119926453,
+      "learning_rate": 0.00010495356037151702,
+      "loss": 3.4888,
+      "step": 5222
+    },
+    {
+      "epoch": 6.67968,
+      "grad_norm": 0.5016779899597168,
+      "learning_rate": 0.00010491317808587965,
+      "loss": 3.4266,
+      "step": 5223
+    },
+    {
+      "epoch": 6.68096,
+      "grad_norm": 0.5184606909751892,
+      "learning_rate": 0.00010487279580024228,
+      "loss": 3.4299,
+      "step": 5224
+    },
+    {
+      "epoch": 6.68224,
+      "grad_norm": 0.5030345320701599,
+      "learning_rate": 0.00010483241351460492,
+      "loss": 3.4783,
+      "step": 5225
+    },
+    {
+      "epoch": 6.68352,
+      "grad_norm": 0.5136353373527527,
+      "learning_rate": 0.00010479203122896755,
+      "loss": 3.5147,
+      "step": 5226
+    },
+    {
+      "epoch": 6.6848,
+      "grad_norm": 0.5140992999076843,
+      "learning_rate": 0.00010475164894333018,
+      "loss": 3.4818,
+      "step": 5227
+    },
+    {
+      "epoch": 6.6860800000000005,
+      "grad_norm": 0.5170648694038391,
+      "learning_rate": 0.00010471126665769281,
+      "loss": 3.4642,
+      "step": 5228
+    },
+    {
+      "epoch": 6.68736,
+      "grad_norm": 0.5128257274627686,
+      "learning_rate": 0.00010467088437205544,
+      "loss": 3.5038,
+      "step": 5229
+    },
+    {
+      "epoch": 6.68864,
+      "grad_norm": 0.5143338441848755,
+      "learning_rate": 0.00010463050208641809,
+      "loss": 3.435,
+      "step": 5230
+    },
+    {
+      "epoch": 6.68992,
+      "grad_norm": 0.5160678029060364,
+      "learning_rate": 0.0001045901198007807,
+      "loss": 3.4402,
+      "step": 5231
+    },
+    {
+      "epoch": 6.6912,
+      "grad_norm": 0.4955798387527466,
+      "learning_rate": 0.00010454973751514335,
+      "loss": 3.4803,
+      "step": 5232
+    },
+    {
+      "epoch": 6.69248,
+      "grad_norm": 0.5133627653121948,
+      "learning_rate": 0.00010450935522950598,
+      "loss": 3.5029,
+      "step": 5233
+    },
+    {
+      "epoch": 6.69376,
+      "grad_norm": 0.5118188261985779,
+      "learning_rate": 0.00010446897294386862,
+      "loss": 3.4282,
+      "step": 5234
+    },
+    {
+      "epoch": 6.69504,
+      "grad_norm": 0.5236147046089172,
+      "learning_rate": 0.00010442859065823124,
+      "loss": 3.5077,
+      "step": 5235
+    },
+    {
+      "epoch": 6.69632,
+      "grad_norm": 0.510515034198761,
+      "learning_rate": 0.00010438820837259388,
+      "loss": 3.6193,
+      "step": 5236
+    },
+    {
+      "epoch": 6.6975999999999996,
+      "grad_norm": 0.51131272315979,
+      "learning_rate": 0.00010434782608695651,
+      "loss": 3.5257,
+      "step": 5237
+    },
+    {
+      "epoch": 6.69888,
+      "grad_norm": 0.5147497653961182,
+      "learning_rate": 0.00010430744380131916,
+      "loss": 3.4566,
+      "step": 5238
+    },
+    {
+      "epoch": 6.70016,
+      "grad_norm": 0.4983549416065216,
+      "learning_rate": 0.00010426706151568177,
+      "loss": 3.5215,
+      "step": 5239
+    },
+    {
+      "epoch": 6.70144,
+      "grad_norm": 0.49420973658561707,
+      "learning_rate": 0.00010422667923004442,
+      "loss": 3.4522,
+      "step": 5240
+    },
+    {
+      "epoch": 6.70272,
+      "grad_norm": 0.5193718075752258,
+      "learning_rate": 0.00010418629694440705,
+      "loss": 3.5033,
+      "step": 5241
+    },
+    {
+      "epoch": 6.704,
+      "grad_norm": 0.5157377123832703,
+      "learning_rate": 0.00010414591465876966,
+      "loss": 3.4045,
+      "step": 5242
+    },
+    {
+      "epoch": 6.70528,
+      "grad_norm": 0.4960803985595703,
+      "learning_rate": 0.0001041055323731323,
+      "loss": 3.4672,
+      "step": 5243
+    },
+    {
+      "epoch": 6.70656,
+      "grad_norm": 0.5239757299423218,
+      "learning_rate": 0.00010406515008749494,
+      "loss": 3.57,
+      "step": 5244
+    },
+    {
+      "epoch": 6.70784,
+      "grad_norm": 0.5048173069953918,
+      "learning_rate": 0.00010402476780185758,
+      "loss": 3.4302,
+      "step": 5245
+    },
+    {
+      "epoch": 6.70912,
+      "grad_norm": 0.5351990461349487,
+      "learning_rate": 0.0001039843855162202,
+      "loss": 3.5531,
+      "step": 5246
+    },
+    {
+      "epoch": 6.7104,
+      "grad_norm": 0.49860915541648865,
+      "learning_rate": 0.00010394400323058284,
+      "loss": 3.3879,
+      "step": 5247
+    },
+    {
+      "epoch": 6.71168,
+      "grad_norm": 0.5203588604927063,
+      "learning_rate": 0.00010390362094494547,
+      "loss": 3.4665,
+      "step": 5248
+    },
+    {
+      "epoch": 6.71296,
+      "grad_norm": 0.5277810096740723,
+      "learning_rate": 0.00010386323865930811,
+      "loss": 3.4588,
+      "step": 5249
+    },
+    {
+      "epoch": 6.71424,
+      "grad_norm": 0.5041863918304443,
+      "learning_rate": 0.00010382285637367074,
+      "loss": 3.4085,
+      "step": 5250
+    },
+    {
+      "epoch": 6.71552,
+      "grad_norm": 0.5177097916603088,
+      "learning_rate": 0.00010378247408803337,
+      "loss": 3.4501,
+      "step": 5251
+    },
+    {
+      "epoch": 6.7168,
+      "grad_norm": 0.5193495750427246,
+      "learning_rate": 0.000103742091802396,
+      "loss": 3.4086,
+      "step": 5252
+    },
+    {
+      "epoch": 6.7180800000000005,
+      "grad_norm": 0.5192511081695557,
+      "learning_rate": 0.00010370170951675865,
+      "loss": 3.4711,
+      "step": 5253
+    },
+    {
+      "epoch": 6.71936,
+      "grad_norm": 0.5242570638656616,
+      "learning_rate": 0.00010366132723112128,
+      "loss": 3.4035,
+      "step": 5254
+    },
+    {
+      "epoch": 6.7206399999999995,
+      "grad_norm": 0.5168814063072205,
+      "learning_rate": 0.00010362094494548391,
+      "loss": 3.4503,
+      "step": 5255
+    },
+    {
+      "epoch": 6.72192,
+      "grad_norm": 0.506004810333252,
+      "learning_rate": 0.00010358056265984654,
+      "loss": 3.456,
+      "step": 5256
+    },
+    {
+      "epoch": 6.7232,
+      "grad_norm": 0.5204073786735535,
+      "learning_rate": 0.00010354018037420917,
+      "loss": 3.4142,
+      "step": 5257
+    },
+    {
+      "epoch": 6.72448,
+      "grad_norm": 0.4996756315231323,
+      "learning_rate": 0.00010349979808857181,
+      "loss": 3.4942,
+      "step": 5258
+    },
+    {
+      "epoch": 6.72576,
+      "grad_norm": 0.5085786581039429,
+      "learning_rate": 0.00010345941580293443,
+      "loss": 3.4686,
+      "step": 5259
+    },
+    {
+      "epoch": 6.72704,
+      "grad_norm": 0.5232467651367188,
+      "learning_rate": 0.00010341903351729707,
+      "loss": 3.5521,
+      "step": 5260
+    },
+    {
+      "epoch": 6.72832,
+      "grad_norm": 0.49586203694343567,
+      "learning_rate": 0.0001033786512316597,
+      "loss": 3.4765,
+      "step": 5261
+    },
+    {
+      "epoch": 6.7296,
+      "grad_norm": 0.5204891562461853,
+      "learning_rate": 0.00010333826894602235,
+      "loss": 3.4617,
+      "step": 5262
+    },
+    {
+      "epoch": 6.73088,
+      "grad_norm": 0.5140677094459534,
+      "learning_rate": 0.00010329788666038496,
+      "loss": 3.4447,
+      "step": 5263
+    },
+    {
+      "epoch": 6.73216,
+      "grad_norm": 0.501307487487793,
+      "learning_rate": 0.0001032575043747476,
+      "loss": 3.4758,
+      "step": 5264
+    },
+    {
+      "epoch": 6.73344,
+      "grad_norm": 0.5081751346588135,
+      "learning_rate": 0.00010321712208911024,
+      "loss": 3.4248,
+      "step": 5265
+    },
+    {
+      "epoch": 6.73472,
+      "grad_norm": 0.5316794514656067,
+      "learning_rate": 0.00010317673980347288,
+      "loss": 3.4928,
+      "step": 5266
+    },
+    {
+      "epoch": 6.736,
+      "grad_norm": 0.5068027377128601,
+      "learning_rate": 0.0001031363575178355,
+      "loss": 3.4352,
+      "step": 5267
+    },
+    {
+      "epoch": 6.73728,
+      "grad_norm": 0.5198256373405457,
+      "learning_rate": 0.00010309597523219814,
+      "loss": 3.5322,
+      "step": 5268
+    },
+    {
+      "epoch": 6.73856,
+      "grad_norm": 0.5113339424133301,
+      "learning_rate": 0.00010305559294656077,
+      "loss": 3.4244,
+      "step": 5269
+    },
+    {
+      "epoch": 6.73984,
+      "grad_norm": 0.5019405484199524,
+      "learning_rate": 0.00010301521066092339,
+      "loss": 3.5323,
+      "step": 5270
+    },
+    {
+      "epoch": 6.7411200000000004,
+      "grad_norm": 0.5108396410942078,
+      "learning_rate": 0.00010297482837528603,
+      "loss": 3.5005,
+      "step": 5271
+    },
+    {
+      "epoch": 6.7424,
+      "grad_norm": 0.5199779868125916,
+      "learning_rate": 0.00010293444608964866,
+      "loss": 3.519,
+      "step": 5272
+    },
+    {
+      "epoch": 6.74368,
+      "grad_norm": 0.5146390795707703,
+      "learning_rate": 0.0001028940638040113,
+      "loss": 3.4188,
+      "step": 5273
+    },
+    {
+      "epoch": 6.74496,
+      "grad_norm": 0.5107973217964172,
+      "learning_rate": 0.00010285368151837392,
+      "loss": 3.4814,
+      "step": 5274
+    },
+    {
+      "epoch": 6.74624,
+      "grad_norm": 0.5244498252868652,
+      "learning_rate": 0.00010281329923273656,
+      "loss": 3.4791,
+      "step": 5275
+    },
+    {
+      "epoch": 6.74752,
+      "grad_norm": 0.5274310111999512,
+      "learning_rate": 0.0001027729169470992,
+      "loss": 3.5072,
+      "step": 5276
+    },
+    {
+      "epoch": 6.7488,
+      "grad_norm": 0.528631329536438,
+      "learning_rate": 0.00010273253466146184,
+      "loss": 3.5248,
+      "step": 5277
+    },
+    {
+      "epoch": 6.75008,
+      "grad_norm": 0.5297229886054993,
+      "learning_rate": 0.00010269215237582445,
+      "loss": 3.4491,
+      "step": 5278
+    },
+    {
+      "epoch": 6.75136,
+      "grad_norm": 0.507632315158844,
+      "learning_rate": 0.0001026517700901871,
+      "loss": 3.4365,
+      "step": 5279
+    },
+    {
+      "epoch": 6.7526399999999995,
+      "grad_norm": 0.5038201808929443,
+      "learning_rate": 0.00010261138780454973,
+      "loss": 3.5131,
+      "step": 5280
+    },
+    {
+      "epoch": 6.75392,
+      "grad_norm": 0.5153136253356934,
+      "learning_rate": 0.00010257100551891237,
+      "loss": 3.4924,
+      "step": 5281
+    },
+    {
+      "epoch": 6.7552,
+      "grad_norm": 0.4991972744464874,
+      "learning_rate": 0.00010253062323327499,
+      "loss": 3.4242,
+      "step": 5282
+    },
+    {
+      "epoch": 6.75648,
+      "grad_norm": 0.4989522099494934,
+      "learning_rate": 0.00010249024094763762,
+      "loss": 3.5012,
+      "step": 5283
+    },
+    {
+      "epoch": 6.75776,
+      "grad_norm": 0.49682870507240295,
+      "learning_rate": 0.00010244985866200026,
+      "loss": 3.4454,
+      "step": 5284
+    },
+    {
+      "epoch": 6.75904,
+      "grad_norm": 0.49864479899406433,
+      "learning_rate": 0.00010240947637636289,
+      "loss": 3.443,
+      "step": 5285
+    },
+    {
+      "epoch": 6.76032,
+      "grad_norm": 0.5117205381393433,
+      "learning_rate": 0.00010236909409072552,
+      "loss": 3.4396,
+      "step": 5286
+    },
+    {
+      "epoch": 6.7616,
+      "grad_norm": 0.5127941966056824,
+      "learning_rate": 0.00010232871180508815,
+      "loss": 3.496,
+      "step": 5287
+    },
+    {
+      "epoch": 6.76288,
+      "grad_norm": 0.5159007906913757,
+      "learning_rate": 0.0001022883295194508,
+      "loss": 3.4693,
+      "step": 5288
+    },
+    {
+      "epoch": 6.76416,
+      "grad_norm": 0.5161288976669312,
+      "learning_rate": 0.00010224794723381343,
+      "loss": 3.4029,
+      "step": 5289
+    },
+    {
+      "epoch": 6.76544,
+      "grad_norm": 0.4983842670917511,
+      "learning_rate": 0.00010220756494817607,
+      "loss": 3.4671,
+      "step": 5290
+    },
+    {
+      "epoch": 6.76672,
+      "grad_norm": 0.5110688209533691,
+      "learning_rate": 0.00010216718266253869,
+      "loss": 3.435,
+      "step": 5291
+    },
+    {
+      "epoch": 6.768,
+      "grad_norm": 0.524663507938385,
+      "learning_rate": 0.00010212680037690133,
+      "loss": 3.5548,
+      "step": 5292
+    },
+    {
+      "epoch": 6.76928,
+      "grad_norm": 0.49911680817604065,
+      "learning_rate": 0.00010208641809126396,
+      "loss": 3.4678,
+      "step": 5293
+    },
+    {
+      "epoch": 6.77056,
+      "grad_norm": 0.5108251571655273,
+      "learning_rate": 0.0001020460358056266,
+      "loss": 3.4816,
+      "step": 5294
+    },
+    {
+      "epoch": 6.77184,
+      "grad_norm": 0.5047867298126221,
+      "learning_rate": 0.00010200565351998922,
+      "loss": 3.4348,
+      "step": 5295
+    },
+    {
+      "epoch": 6.7731200000000005,
+      "grad_norm": 0.5126866698265076,
+      "learning_rate": 0.00010196527123435185,
+      "loss": 3.4728,
+      "step": 5296
+    },
+    {
+      "epoch": 6.7744,
+      "grad_norm": 0.5074558258056641,
+      "learning_rate": 0.0001019248889487145,
+      "loss": 3.4696,
+      "step": 5297
+    },
+    {
+      "epoch": 6.77568,
+      "grad_norm": 0.5098884105682373,
+      "learning_rate": 0.00010188450666307711,
+      "loss": 3.4211,
+      "step": 5298
+    },
+    {
+      "epoch": 6.77696,
+      "grad_norm": 0.5483927726745605,
+      "learning_rate": 0.00010184412437743975,
+      "loss": 3.427,
+      "step": 5299
+    },
+    {
+      "epoch": 6.77824,
+      "grad_norm": 0.49813124537467957,
+      "learning_rate": 0.00010180374209180238,
+      "loss": 3.4238,
+      "step": 5300
+    },
+    {
+      "epoch": 6.77952,
+      "grad_norm": 0.5243797898292542,
+      "learning_rate": 0.00010176335980616503,
+      "loss": 3.5178,
+      "step": 5301
+    },
+    {
+      "epoch": 6.7808,
+      "grad_norm": 0.5047301054000854,
+      "learning_rate": 0.00010172297752052764,
+      "loss": 3.3934,
+      "step": 5302
+    },
+    {
+      "epoch": 6.78208,
+      "grad_norm": 0.4998399019241333,
+      "learning_rate": 0.00010168259523489029,
+      "loss": 3.4589,
+      "step": 5303
+    },
+    {
+      "epoch": 6.78336,
+      "grad_norm": 0.5172979831695557,
+      "learning_rate": 0.00010164221294925292,
+      "loss": 3.4315,
+      "step": 5304
+    },
+    {
+      "epoch": 6.78464,
+      "grad_norm": 0.5025821328163147,
+      "learning_rate": 0.00010160183066361556,
+      "loss": 3.4013,
+      "step": 5305
+    },
+    {
+      "epoch": 6.78592,
+      "grad_norm": 0.49646785855293274,
+      "learning_rate": 0.00010156144837797818,
+      "loss": 3.4618,
+      "step": 5306
+    },
+    {
+      "epoch": 6.7872,
+      "grad_norm": 0.49145373702049255,
+      "learning_rate": 0.00010152106609234082,
+      "loss": 3.4746,
+      "step": 5307
+    },
+    {
+      "epoch": 6.78848,
+      "grad_norm": 0.5159743428230286,
+      "learning_rate": 0.00010148068380670345,
+      "loss": 3.4769,
+      "step": 5308
+    },
+    {
+      "epoch": 6.78976,
+      "grad_norm": 0.4938885271549225,
+      "learning_rate": 0.00010144030152106608,
+      "loss": 3.3877,
+      "step": 5309
+    },
+    {
+      "epoch": 6.79104,
+      "grad_norm": 0.5240892767906189,
+      "learning_rate": 0.00010139991923542871,
+      "loss": 3.5444,
+      "step": 5310
+    },
+    {
+      "epoch": 6.79232,
+      "grad_norm": 0.5187338590621948,
+      "learning_rate": 0.00010135953694979134,
+      "loss": 3.5475,
+      "step": 5311
+    },
+    {
+      "epoch": 6.7936,
+      "grad_norm": 0.5051224231719971,
+      "learning_rate": 0.00010131915466415399,
+      "loss": 3.4844,
+      "step": 5312
+    },
+    {
+      "epoch": 6.79488,
+      "grad_norm": 0.5185756683349609,
+      "learning_rate": 0.00010127877237851662,
+      "loss": 3.5207,
+      "step": 5313
+    },
+    {
+      "epoch": 6.79616,
+      "grad_norm": 0.5253816843032837,
+      "learning_rate": 0.00010123839009287925,
+      "loss": 3.4069,
+      "step": 5314
+    },
+    {
+      "epoch": 6.79744,
+      "grad_norm": 0.511680543422699,
+      "learning_rate": 0.00010119800780724188,
+      "loss": 3.4456,
+      "step": 5315
+    },
+    {
+      "epoch": 6.79872,
+      "grad_norm": 0.5037193298339844,
+      "learning_rate": 0.00010115762552160452,
+      "loss": 3.5139,
+      "step": 5316
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.5148517489433289,
+      "learning_rate": 0.00010111724323596715,
+      "loss": 3.4717,
+      "step": 5317
+    },
+    {
+      "epoch": 6.80128,
+      "grad_norm": 0.4965799152851105,
+      "learning_rate": 0.00010107686095032978,
+      "loss": 3.4902,
+      "step": 5318
+    },
+    {
+      "epoch": 6.80256,
+      "grad_norm": 0.5093517899513245,
+      "learning_rate": 0.00010103647866469241,
+      "loss": 3.5074,
+      "step": 5319
+    },
+    {
+      "epoch": 6.80384,
+      "grad_norm": 0.5288740396499634,
+      "learning_rate": 0.00010099609637905505,
+      "loss": 3.4673,
+      "step": 5320
+    },
+    {
+      "epoch": 6.80512,
+      "grad_norm": 0.5054797530174255,
+      "learning_rate": 0.00010095571409341768,
+      "loss": 3.5055,
+      "step": 5321
+    },
+    {
+      "epoch": 6.8064,
+      "grad_norm": 0.5090057253837585,
+      "learning_rate": 0.00010091533180778031,
+      "loss": 3.5057,
+      "step": 5322
+    },
+    {
+      "epoch": 6.8076799999999995,
+      "grad_norm": 0.5045148730278015,
+      "learning_rate": 0.00010087494952214294,
+      "loss": 3.4295,
+      "step": 5323
+    },
+    {
+      "epoch": 6.80896,
+      "grad_norm": 0.5039887428283691,
+      "learning_rate": 0.00010083456723650557,
+      "loss": 3.4735,
+      "step": 5324
+    },
+    {
+      "epoch": 6.81024,
+      "grad_norm": 0.4997764229774475,
+      "learning_rate": 0.00010079418495086822,
+      "loss": 3.4807,
+      "step": 5325
+    },
+    {
+      "epoch": 6.81152,
+      "grad_norm": 0.5063294768333435,
+      "learning_rate": 0.00010075380266523083,
+      "loss": 3.446,
+      "step": 5326
+    },
+    {
+      "epoch": 6.8128,
+      "grad_norm": 0.5159018039703369,
+      "learning_rate": 0.00010071342037959348,
+      "loss": 3.4766,
+      "step": 5327
+    },
+    {
+      "epoch": 6.81408,
+      "grad_norm": 0.5194793343544006,
+      "learning_rate": 0.00010067303809395611,
+      "loss": 3.5094,
+      "step": 5328
+    },
+    {
+      "epoch": 6.81536,
+      "grad_norm": 0.5233056545257568,
+      "learning_rate": 0.00010063265580831875,
+      "loss": 3.5093,
+      "step": 5329
+    },
+    {
+      "epoch": 6.81664,
+      "grad_norm": 0.5220425128936768,
+      "learning_rate": 0.00010059227352268137,
+      "loss": 3.5481,
+      "step": 5330
+    },
+    {
+      "epoch": 6.81792,
+      "grad_norm": 0.4991253912448883,
+      "learning_rate": 0.00010055189123704401,
+      "loss": 3.4397,
+      "step": 5331
+    },
+    {
+      "epoch": 6.8192,
+      "grad_norm": 0.5196431279182434,
+      "learning_rate": 0.00010051150895140664,
+      "loss": 3.429,
+      "step": 5332
+    },
+    {
+      "epoch": 6.82048,
+      "grad_norm": 0.510365903377533,
+      "learning_rate": 0.00010047112666576928,
+      "loss": 3.4965,
+      "step": 5333
+    },
+    {
+      "epoch": 6.82176,
+      "grad_norm": 0.5005152821540833,
+      "learning_rate": 0.0001004307443801319,
+      "loss": 3.4985,
+      "step": 5334
+    },
+    {
+      "epoch": 6.82304,
+      "grad_norm": 0.508727490901947,
+      "learning_rate": 0.00010039036209449454,
+      "loss": 3.3829,
+      "step": 5335
+    },
+    {
+      "epoch": 6.82432,
+      "grad_norm": 0.5061600208282471,
+      "learning_rate": 0.00010034997980885717,
+      "loss": 3.4547,
+      "step": 5336
+    },
+    {
+      "epoch": 6.8256,
+      "grad_norm": 0.5172441601753235,
+      "learning_rate": 0.00010030959752321979,
+      "loss": 3.3778,
+      "step": 5337
+    },
+    {
+      "epoch": 6.82688,
+      "grad_norm": 0.4944312572479248,
+      "learning_rate": 0.00010026921523758243,
+      "loss": 3.4106,
+      "step": 5338
+    },
+    {
+      "epoch": 6.8281600000000005,
+      "grad_norm": 0.5171020030975342,
+      "learning_rate": 0.00010022883295194506,
+      "loss": 3.4994,
+      "step": 5339
+    },
+    {
+      "epoch": 6.82944,
+      "grad_norm": 0.5048860907554626,
+      "learning_rate": 0.00010018845066630771,
+      "loss": 3.4575,
+      "step": 5340
+    },
+    {
+      "epoch": 6.83072,
+      "grad_norm": 0.5153966546058655,
+      "learning_rate": 0.00010014806838067034,
+      "loss": 3.4745,
+      "step": 5341
+    },
+    {
+      "epoch": 6.832,
+      "grad_norm": 0.5013549327850342,
+      "learning_rate": 0.00010010768609503297,
+      "loss": 3.4541,
+      "step": 5342
+    },
+    {
+      "epoch": 6.83328,
+      "grad_norm": 0.520453929901123,
+      "learning_rate": 0.0001000673038093956,
+      "loss": 3.4551,
+      "step": 5343
+    },
+    {
+      "epoch": 6.83456,
+      "grad_norm": 0.5147715210914612,
+      "learning_rate": 0.00010002692152375824,
+      "loss": 3.4985,
+      "step": 5344
+    },
+    {
+      "epoch": 6.83584,
+      "grad_norm": 0.5121247172355652,
+      "learning_rate": 9.998653923812087e-05,
+      "loss": 3.4887,
+      "step": 5345
+    },
+    {
+      "epoch": 6.83712,
+      "grad_norm": 0.5188484787940979,
+      "learning_rate": 9.99461569524835e-05,
+      "loss": 3.4733,
+      "step": 5346
+    },
+    {
+      "epoch": 6.8384,
+      "grad_norm": 0.5076323747634888,
+      "learning_rate": 9.990577466684613e-05,
+      "loss": 3.4318,
+      "step": 5347
+    },
+    {
+      "epoch": 6.8396799999999995,
+      "grad_norm": 0.5205701589584351,
+      "learning_rate": 9.986539238120878e-05,
+      "loss": 3.4467,
+      "step": 5348
+    },
+    {
+      "epoch": 6.84096,
+      "grad_norm": 0.5038503408432007,
+      "learning_rate": 9.98250100955714e-05,
+      "loss": 3.3956,
+      "step": 5349
+    },
+    {
+      "epoch": 6.84224,
+      "grad_norm": 0.4975230395793915,
+      "learning_rate": 9.978462780993402e-05,
+      "loss": 3.4585,
+      "step": 5350
+    },
+    {
+      "epoch": 6.84352,
+      "grad_norm": 0.5168045163154602,
+      "learning_rate": 9.974424552429667e-05,
+      "loss": 3.4813,
+      "step": 5351
+    },
+    {
+      "epoch": 6.8448,
+      "grad_norm": 0.5278074145317078,
+      "learning_rate": 9.97038632386593e-05,
+      "loss": 3.5233,
+      "step": 5352
+    },
+    {
+      "epoch": 6.84608,
+      "grad_norm": 0.5026320219039917,
+      "learning_rate": 9.966348095302194e-05,
+      "loss": 3.3954,
+      "step": 5353
+    },
+    {
+      "epoch": 6.84736,
+      "grad_norm": 0.5222524404525757,
+      "learning_rate": 9.962309866738456e-05,
+      "loss": 3.5426,
+      "step": 5354
+    },
+    {
+      "epoch": 6.84864,
+      "grad_norm": 0.49351999163627625,
+      "learning_rate": 9.95827163817472e-05,
+      "loss": 3.5092,
+      "step": 5355
+    },
+    {
+      "epoch": 6.84992,
+      "grad_norm": 0.5170454978942871,
+      "learning_rate": 9.954233409610983e-05,
+      "loss": 3.5309,
+      "step": 5356
+    },
+    {
+      "epoch": 6.8512,
+      "grad_norm": 0.49120914936065674,
+      "learning_rate": 9.950195181047247e-05,
+      "loss": 3.4582,
+      "step": 5357
+    },
+    {
+      "epoch": 6.85248,
+      "grad_norm": 0.5041413903236389,
+      "learning_rate": 9.946156952483509e-05,
+      "loss": 3.5076,
+      "step": 5358
+    },
+    {
+      "epoch": 6.85376,
+      "grad_norm": 0.5105189085006714,
+      "learning_rate": 9.942118723919773e-05,
+      "loss": 3.5075,
+      "step": 5359
+    },
+    {
+      "epoch": 6.85504,
+      "grad_norm": 0.4975397288799286,
+      "learning_rate": 9.938080495356036e-05,
+      "loss": 3.4318,
+      "step": 5360
+    },
+    {
+      "epoch": 6.85632,
+      "grad_norm": 0.49615657329559326,
+      "learning_rate": 9.934042266792301e-05,
+      "loss": 3.4473,
+      "step": 5361
+    },
+    {
+      "epoch": 6.8576,
+      "grad_norm": 0.4980040490627289,
+      "learning_rate": 9.930004038228562e-05,
+      "loss": 3.4313,
+      "step": 5362
+    },
+    {
+      "epoch": 6.85888,
+      "grad_norm": 0.4903441071510315,
+      "learning_rate": 9.925965809664825e-05,
+      "loss": 3.4639,
+      "step": 5363
+    },
+    {
+      "epoch": 6.8601600000000005,
+      "grad_norm": 0.5194531679153442,
+      "learning_rate": 9.92192758110109e-05,
+      "loss": 3.5111,
+      "step": 5364
+    },
+    {
+      "epoch": 6.86144,
+      "grad_norm": 0.5149429440498352,
+      "learning_rate": 9.917889352537351e-05,
+      "loss": 3.4815,
+      "step": 5365
+    },
+    {
+      "epoch": 6.86272,
+      "grad_norm": 0.49559643864631653,
+      "learning_rate": 9.913851123973616e-05,
+      "loss": 3.4756,
+      "step": 5366
+    },
+    {
+      "epoch": 6.864,
+      "grad_norm": 0.5016883015632629,
+      "learning_rate": 9.909812895409879e-05,
+      "loss": 3.4241,
+      "step": 5367
+    },
+    {
+      "epoch": 6.86528,
+      "grad_norm": 0.5163260102272034,
+      "learning_rate": 9.905774666846143e-05,
+      "loss": 3.4777,
+      "step": 5368
+    },
+    {
+      "epoch": 6.86656,
+      "grad_norm": 0.5174842476844788,
+      "learning_rate": 9.901736438282405e-05,
+      "loss": 3.534,
+      "step": 5369
+    },
+    {
+      "epoch": 6.86784,
+      "grad_norm": 0.49278444051742554,
+      "learning_rate": 9.897698209718669e-05,
+      "loss": 3.453,
+      "step": 5370
+    },
+    {
+      "epoch": 6.86912,
+      "grad_norm": 0.5156992673873901,
+      "learning_rate": 9.893659981154932e-05,
+      "loss": 3.4667,
+      "step": 5371
+    },
+    {
+      "epoch": 6.8704,
+      "grad_norm": 0.5196961164474487,
+      "learning_rate": 9.889621752591197e-05,
+      "loss": 3.5111,
+      "step": 5372
+    },
+    {
+      "epoch": 6.87168,
+      "grad_norm": 0.5395616888999939,
+      "learning_rate": 9.885583524027458e-05,
+      "loss": 3.4736,
+      "step": 5373
+    },
+    {
+      "epoch": 6.87296,
+      "grad_norm": 0.49754607677459717,
+      "learning_rate": 9.881545295463723e-05,
+      "loss": 3.4057,
+      "step": 5374
+    },
+    {
+      "epoch": 6.87424,
+      "grad_norm": 0.5215137004852295,
+      "learning_rate": 9.877507066899986e-05,
+      "loss": 3.4588,
+      "step": 5375
+    },
+    {
+      "epoch": 6.87552,
+      "grad_norm": 0.5358836650848389,
+      "learning_rate": 9.87346883833625e-05,
+      "loss": 3.542,
+      "step": 5376
+    },
+    {
+      "epoch": 6.8768,
+      "grad_norm": 0.5070124268531799,
+      "learning_rate": 9.869430609772513e-05,
+      "loss": 3.4484,
+      "step": 5377
+    },
+    {
+      "epoch": 6.87808,
+      "grad_norm": 0.5137901902198792,
+      "learning_rate": 9.865392381208775e-05,
+      "loss": 3.447,
+      "step": 5378
+    },
+    {
+      "epoch": 6.87936,
+      "grad_norm": 0.5140761137008667,
+      "learning_rate": 9.861354152645039e-05,
+      "loss": 3.4393,
+      "step": 5379
+    },
+    {
+      "epoch": 6.88064,
+      "grad_norm": 0.5206302404403687,
+      "learning_rate": 9.857315924081302e-05,
+      "loss": 3.4926,
+      "step": 5380
+    },
+    {
+      "epoch": 6.88192,
+      "grad_norm": 0.5339552760124207,
+      "learning_rate": 9.853277695517566e-05,
+      "loss": 3.5058,
+      "step": 5381
+    },
+    {
+      "epoch": 6.8832,
+      "grad_norm": 0.5175068974494934,
+      "learning_rate": 9.849239466953828e-05,
+      "loss": 3.4939,
+      "step": 5382
+    },
+    {
+      "epoch": 6.88448,
+      "grad_norm": 0.4954850673675537,
+      "learning_rate": 9.845201238390092e-05,
+      "loss": 3.4226,
+      "step": 5383
+    },
+    {
+      "epoch": 6.88576,
+      "grad_norm": 0.5178061127662659,
+      "learning_rate": 9.841163009826355e-05,
+      "loss": 3.4637,
+      "step": 5384
+    },
+    {
+      "epoch": 6.88704,
+      "grad_norm": 0.5334754586219788,
+      "learning_rate": 9.83712478126262e-05,
+      "loss": 3.4926,
+      "step": 5385
+    },
+    {
+      "epoch": 6.88832,
+      "grad_norm": 0.5084394216537476,
+      "learning_rate": 9.833086552698881e-05,
+      "loss": 3.413,
+      "step": 5386
+    },
+    {
+      "epoch": 6.8896,
+      "grad_norm": 0.5201271772384644,
+      "learning_rate": 9.829048324135146e-05,
+      "loss": 3.4399,
+      "step": 5387
+    },
+    {
+      "epoch": 6.89088,
+      "grad_norm": 0.5018450021743774,
+      "learning_rate": 9.825010095571409e-05,
+      "loss": 3.4989,
+      "step": 5388
+    },
+    {
+      "epoch": 6.89216,
+      "grad_norm": 0.524716854095459,
+      "learning_rate": 9.820971867007673e-05,
+      "loss": 3.4713,
+      "step": 5389
+    },
+    {
+      "epoch": 6.89344,
+      "grad_norm": 0.5103277564048767,
+      "learning_rate": 9.816933638443935e-05,
+      "loss": 3.4293,
+      "step": 5390
+    },
+    {
+      "epoch": 6.8947199999999995,
+      "grad_norm": 0.5264323353767395,
+      "learning_rate": 9.812895409880198e-05,
+      "loss": 3.5141,
+      "step": 5391
+    },
+    {
+      "epoch": 6.896,
+      "grad_norm": 0.5120908617973328,
+      "learning_rate": 9.808857181316462e-05,
+      "loss": 3.5326,
+      "step": 5392
+    },
+    {
+      "epoch": 6.89728,
+      "grad_norm": 0.510351300239563,
+      "learning_rate": 9.804818952752724e-05,
+      "loss": 3.3871,
+      "step": 5393
+    },
+    {
+      "epoch": 6.89856,
+      "grad_norm": 0.5060296058654785,
+      "learning_rate": 9.800780724188988e-05,
+      "loss": 3.4455,
+      "step": 5394
+    },
+    {
+      "epoch": 6.89984,
+      "grad_norm": 0.5129922032356262,
+      "learning_rate": 9.796742495625251e-05,
+      "loss": 3.5359,
+      "step": 5395
+    },
+    {
+      "epoch": 6.90112,
+      "grad_norm": 0.5016512274742126,
+      "learning_rate": 9.792704267061516e-05,
+      "loss": 3.457,
+      "step": 5396
+    },
+    {
+      "epoch": 6.9024,
+      "grad_norm": 0.50436931848526,
+      "learning_rate": 9.788666038497777e-05,
+      "loss": 3.5133,
+      "step": 5397
+    },
+    {
+      "epoch": 6.90368,
+      "grad_norm": 0.5089812874794006,
+      "learning_rate": 9.784627809934042e-05,
+      "loss": 3.4324,
+      "step": 5398
+    },
+    {
+      "epoch": 6.90496,
+      "grad_norm": 0.503563642501831,
+      "learning_rate": 9.780589581370305e-05,
+      "loss": 3.4798,
+      "step": 5399
+    },
+    {
+      "epoch": 6.90624,
+      "grad_norm": 0.49585676193237305,
+      "learning_rate": 9.776551352806569e-05,
+      "loss": 3.4961,
+      "step": 5400
+    },
+    {
+      "epoch": 6.90752,
+      "grad_norm": 0.5292691588401794,
+      "learning_rate": 9.77251312424283e-05,
+      "loss": 3.4779,
+      "step": 5401
+    },
+    {
+      "epoch": 6.9088,
+      "grad_norm": 0.4982292056083679,
+      "learning_rate": 9.768474895679095e-05,
+      "loss": 3.5183,
+      "step": 5402
+    },
+    {
+      "epoch": 6.91008,
+      "grad_norm": 0.5063697099685669,
+      "learning_rate": 9.764436667115358e-05,
+      "loss": 3.5731,
+      "step": 5403
+    },
+    {
+      "epoch": 6.91136,
+      "grad_norm": 0.5067952871322632,
+      "learning_rate": 9.760398438551621e-05,
+      "loss": 3.4067,
+      "step": 5404
+    },
+    {
+      "epoch": 6.91264,
+      "grad_norm": 0.5001175403594971,
+      "learning_rate": 9.756360209987884e-05,
+      "loss": 3.4509,
+      "step": 5405
+    },
+    {
+      "epoch": 6.91392,
+      "grad_norm": 0.5022578239440918,
+      "learning_rate": 9.752321981424147e-05,
+      "loss": 3.5178,
+      "step": 5406
+    },
+    {
+      "epoch": 6.9152000000000005,
+      "grad_norm": 0.49592211842536926,
+      "learning_rate": 9.748283752860411e-05,
+      "loss": 3.4671,
+      "step": 5407
+    },
+    {
+      "epoch": 6.91648,
+      "grad_norm": 0.5008174777030945,
+      "learning_rate": 9.744245524296674e-05,
+      "loss": 3.463,
+      "step": 5408
+    },
+    {
+      "epoch": 6.91776,
+      "grad_norm": 0.49405258893966675,
+      "learning_rate": 9.740207295732937e-05,
+      "loss": 3.4687,
+      "step": 5409
+    },
+    {
+      "epoch": 6.91904,
+      "grad_norm": 0.5040400624275208,
+      "learning_rate": 9.7361690671692e-05,
+      "loss": 3.4612,
+      "step": 5410
+    },
+    {
+      "epoch": 6.92032,
+      "grad_norm": 0.5239134430885315,
+      "learning_rate": 9.732130838605465e-05,
+      "loss": 3.4698,
+      "step": 5411
+    },
+    {
+      "epoch": 6.9216,
+      "grad_norm": 0.5057811737060547,
+      "learning_rate": 9.728092610041728e-05,
+      "loss": 3.4569,
+      "step": 5412
+    },
+    {
+      "epoch": 6.92288,
+      "grad_norm": 0.49874454736709595,
+      "learning_rate": 9.724054381477992e-05,
+      "loss": 3.4118,
+      "step": 5413
+    },
+    {
+      "epoch": 6.92416,
+      "grad_norm": 0.5236087441444397,
+      "learning_rate": 9.720016152914254e-05,
+      "loss": 3.5346,
+      "step": 5414
+    },
+    {
+      "epoch": 6.92544,
+      "grad_norm": 0.5117159485816956,
+      "learning_rate": 9.715977924350518e-05,
+      "loss": 3.5194,
+      "step": 5415
+    },
+    {
+      "epoch": 6.9267199999999995,
+      "grad_norm": 0.5199319124221802,
+      "learning_rate": 9.711939695786781e-05,
+      "loss": 3.4622,
+      "step": 5416
+    },
+    {
+      "epoch": 6.928,
+      "grad_norm": 0.5007228255271912,
+      "learning_rate": 9.707901467223043e-05,
+      "loss": 3.4657,
+      "step": 5417
+    },
+    {
+      "epoch": 6.92928,
+      "grad_norm": 0.4935159683227539,
+      "learning_rate": 9.703863238659307e-05,
+      "loss": 3.473,
+      "step": 5418
+    },
+    {
+      "epoch": 6.93056,
+      "grad_norm": 0.5131574273109436,
+      "learning_rate": 9.69982501009557e-05,
+      "loss": 3.4111,
+      "step": 5419
+    },
+    {
+      "epoch": 6.93184,
+      "grad_norm": 0.4788220226764679,
+      "learning_rate": 9.695786781531834e-05,
+      "loss": 3.4845,
+      "step": 5420
+    },
+    {
+      "epoch": 6.93312,
+      "grad_norm": 0.5057500600814819,
+      "learning_rate": 9.691748552968096e-05,
+      "loss": 3.505,
+      "step": 5421
+    },
+    {
+      "epoch": 6.9344,
+      "grad_norm": 0.5060935020446777,
+      "learning_rate": 9.68771032440436e-05,
+      "loss": 3.4094,
+      "step": 5422
+    },
+    {
+      "epoch": 6.93568,
+      "grad_norm": 0.4928586184978485,
+      "learning_rate": 9.683672095840624e-05,
+      "loss": 3.4754,
+      "step": 5423
+    },
+    {
+      "epoch": 6.93696,
+      "grad_norm": 0.5091453790664673,
+      "learning_rate": 9.679633867276888e-05,
+      "loss": 3.4907,
+      "step": 5424
+    },
+    {
+      "epoch": 6.93824,
+      "grad_norm": 0.5088662505149841,
+      "learning_rate": 9.67559563871315e-05,
+      "loss": 3.4644,
+      "step": 5425
+    },
+    {
+      "epoch": 6.93952,
+      "grad_norm": 0.50926274061203,
+      "learning_rate": 9.671557410149414e-05,
+      "loss": 3.4988,
+      "step": 5426
+    },
+    {
+      "epoch": 6.9408,
+      "grad_norm": 0.5116270780563354,
+      "learning_rate": 9.667519181585677e-05,
+      "loss": 3.4536,
+      "step": 5427
+    },
+    {
+      "epoch": 6.94208,
+      "grad_norm": 0.4971657395362854,
+      "learning_rate": 9.663480953021941e-05,
+      "loss": 3.472,
+      "step": 5428
+    },
+    {
+      "epoch": 6.94336,
+      "grad_norm": 0.5055294632911682,
+      "learning_rate": 9.659442724458203e-05,
+      "loss": 3.4944,
+      "step": 5429
+    },
+    {
+      "epoch": 6.94464,
+      "grad_norm": 0.5095595121383667,
+      "learning_rate": 9.655404495894466e-05,
+      "loss": 3.4677,
+      "step": 5430
+    },
+    {
+      "epoch": 6.94592,
+      "grad_norm": 0.4956967830657959,
+      "learning_rate": 9.65136626733073e-05,
+      "loss": 3.4282,
+      "step": 5431
+    },
+    {
+      "epoch": 6.9472000000000005,
+      "grad_norm": 0.5110241174697876,
+      "learning_rate": 9.647328038766993e-05,
+      "loss": 3.4498,
+      "step": 5432
+    },
+    {
+      "epoch": 6.94848,
+      "grad_norm": 0.4950363039970398,
+      "learning_rate": 9.643289810203256e-05,
+      "loss": 3.4326,
+      "step": 5433
+    },
+    {
+      "epoch": 6.94976,
+      "grad_norm": 0.5211616158485413,
+      "learning_rate": 9.639251581639519e-05,
+      "loss": 3.512,
+      "step": 5434
+    },
+    {
+      "epoch": 6.95104,
+      "grad_norm": 0.5052586793899536,
+      "learning_rate": 9.635213353075784e-05,
+      "loss": 3.4677,
+      "step": 5435
+    },
+    {
+      "epoch": 6.95232,
+      "grad_norm": 0.509345293045044,
+      "learning_rate": 9.631175124512047e-05,
+      "loss": 3.4526,
+      "step": 5436
+    },
+    {
+      "epoch": 6.9536,
+      "grad_norm": 0.5046288967132568,
+      "learning_rate": 9.62713689594831e-05,
+      "loss": 3.4675,
+      "step": 5437
+    },
+    {
+      "epoch": 6.95488,
+      "grad_norm": 0.511081337928772,
+      "learning_rate": 9.623098667384573e-05,
+      "loss": 3.4853,
+      "step": 5438
+    },
+    {
+      "epoch": 6.95616,
+      "grad_norm": 0.512008011341095,
+      "learning_rate": 9.619060438820837e-05,
+      "loss": 3.4673,
+      "step": 5439
+    },
+    {
+      "epoch": 6.95744,
+      "grad_norm": 0.5058484077453613,
+      "learning_rate": 9.6150222102571e-05,
+      "loss": 3.4265,
+      "step": 5440
+    },
+    {
+      "epoch": 6.95872,
+      "grad_norm": 0.5225197076797485,
+      "learning_rate": 9.610983981693363e-05,
+      "loss": 3.495,
+      "step": 5441
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.5050714015960693,
+      "learning_rate": 9.606945753129626e-05,
+      "loss": 3.4189,
+      "step": 5442
+    },
+    {
+      "epoch": 6.96128,
+      "grad_norm": 0.5137665271759033,
+      "learning_rate": 9.60290752456589e-05,
+      "loss": 3.4752,
+      "step": 5443
+    },
+    {
+      "epoch": 6.96256,
+      "grad_norm": 0.5180103778839111,
+      "learning_rate": 9.598869296002153e-05,
+      "loss": 3.4813,
+      "step": 5444
+    },
+    {
+      "epoch": 6.96384,
+      "grad_norm": 0.5160430073738098,
+      "learning_rate": 9.594831067438415e-05,
+      "loss": 3.4638,
+      "step": 5445
+    },
+    {
+      "epoch": 6.96512,
+      "grad_norm": 0.511764407157898,
+      "learning_rate": 9.59079283887468e-05,
+      "loss": 3.4263,
+      "step": 5446
+    },
+    {
+      "epoch": 6.9664,
+      "grad_norm": 0.5118876099586487,
+      "learning_rate": 9.586754610310942e-05,
+      "loss": 3.4935,
+      "step": 5447
+    },
+    {
+      "epoch": 6.96768,
+      "grad_norm": 0.5115414261817932,
+      "learning_rate": 9.582716381747207e-05,
+      "loss": 3.4721,
+      "step": 5448
+    },
+    {
+      "epoch": 6.96896,
+      "grad_norm": 0.5058209896087646,
+      "learning_rate": 9.578678153183468e-05,
+      "loss": 3.464,
+      "step": 5449
+    },
+    {
+      "epoch": 6.97024,
+      "grad_norm": 0.4981795847415924,
+      "learning_rate": 9.574639924619733e-05,
+      "loss": 3.473,
+      "step": 5450
+    },
+    {
+      "epoch": 6.97152,
+      "grad_norm": 0.5040668249130249,
+      "learning_rate": 9.570601696055996e-05,
+      "loss": 3.3973,
+      "step": 5451
+    },
+    {
+      "epoch": 6.9728,
+      "grad_norm": 0.49496889114379883,
+      "learning_rate": 9.56656346749226e-05,
+      "loss": 3.372,
+      "step": 5452
+    },
+    {
+      "epoch": 6.97408,
+      "grad_norm": 0.5042381882667542,
+      "learning_rate": 9.562525238928522e-05,
+      "loss": 3.4545,
+      "step": 5453
+    },
+    {
+      "epoch": 6.97536,
+      "grad_norm": 0.5015059113502502,
+      "learning_rate": 9.558487010364786e-05,
+      "loss": 3.4237,
+      "step": 5454
+    },
+    {
+      "epoch": 6.97664,
+      "grad_norm": 0.5201092958450317,
+      "learning_rate": 9.554448781801049e-05,
+      "loss": 3.4955,
+      "step": 5455
+    },
+    {
+      "epoch": 6.97792,
+      "grad_norm": 0.5009555220603943,
+      "learning_rate": 9.550410553237314e-05,
+      "loss": 3.4304,
+      "step": 5456
+    },
+    {
+      "epoch": 6.9792,
+      "grad_norm": 0.49957409501075745,
+      "learning_rate": 9.546372324673575e-05,
+      "loss": 3.435,
+      "step": 5457
+    },
+    {
+      "epoch": 6.98048,
+      "grad_norm": 0.512995719909668,
+      "learning_rate": 9.542334096109838e-05,
+      "loss": 3.5062,
+      "step": 5458
+    },
+    {
+      "epoch": 6.9817599999999995,
+      "grad_norm": 0.5111671090126038,
+      "learning_rate": 9.538295867546103e-05,
+      "loss": 3.4766,
+      "step": 5459
+    },
+    {
+      "epoch": 6.98304,
+      "grad_norm": 0.5001948475837708,
+      "learning_rate": 9.534257638982364e-05,
+      "loss": 3.4496,
+      "step": 5460
+    },
+    {
+      "epoch": 6.98432,
+      "grad_norm": 0.5309041738510132,
+      "learning_rate": 9.530219410418629e-05,
+      "loss": 3.5057,
+      "step": 5461
+    },
+    {
+      "epoch": 6.9856,
+      "grad_norm": 0.5163221955299377,
+      "learning_rate": 9.526181181854892e-05,
+      "loss": 3.4416,
+      "step": 5462
+    },
+    {
+      "epoch": 6.98688,
+      "grad_norm": 0.4984453022480011,
+      "learning_rate": 9.522142953291156e-05,
+      "loss": 3.46,
+      "step": 5463
+    },
+    {
+      "epoch": 6.98816,
+      "grad_norm": 0.5184572339057922,
+      "learning_rate": 9.518104724727419e-05,
+      "loss": 3.4491,
+      "step": 5464
+    },
+    {
+      "epoch": 6.98944,
+      "grad_norm": 0.5225724577903748,
+      "learning_rate": 9.514066496163682e-05,
+      "loss": 3.4354,
+      "step": 5465
+    },
+    {
+      "epoch": 6.99072,
+      "grad_norm": 0.507361114025116,
+      "learning_rate": 9.510028267599945e-05,
+      "loss": 3.4749,
+      "step": 5466
+    },
+    {
+      "epoch": 6.992,
+      "grad_norm": 0.5018404126167297,
+      "learning_rate": 9.50599003903621e-05,
+      "loss": 3.4538,
+      "step": 5467
+    },
+    {
+      "epoch": 6.99328,
+      "grad_norm": 0.5017131567001343,
+      "learning_rate": 9.501951810472472e-05,
+      "loss": 3.4954,
+      "step": 5468
+    },
+    {
+      "epoch": 6.99456,
+      "grad_norm": 0.5111985802650452,
+      "learning_rate": 9.497913581908735e-05,
+      "loss": 3.4424,
+      "step": 5469
+    },
+    {
+      "epoch": 6.99584,
+      "grad_norm": 0.5290077328681946,
+      "learning_rate": 9.493875353344998e-05,
+      "loss": 3.4288,
+      "step": 5470
+    },
+    {
+      "epoch": 6.99712,
+      "grad_norm": 0.5016758441925049,
+      "learning_rate": 9.489837124781261e-05,
+      "loss": 3.4364,
+      "step": 5471
+    },
+    {
+      "epoch": 6.9984,
+      "grad_norm": 0.491237074136734,
+      "learning_rate": 9.485798896217526e-05,
+      "loss": 3.4449,
+      "step": 5472
+    },
+    {
+      "epoch": 6.99968,
+      "grad_norm": 0.5003752112388611,
+      "learning_rate": 9.481760667653787e-05,
+      "loss": 3.4199,
+      "step": 5473
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.9872975945472717,
+      "learning_rate": 9.477722439090052e-05,
+      "loss": 3.5161,
+      "step": 5474
+    },
+    {
+      "epoch": 7.00128,
+      "grad_norm": 0.5078200101852417,
+      "learning_rate": 9.473684210526315e-05,
+      "loss": 3.33,
+      "step": 5475
+    },
+    {
+      "epoch": 7.00256,
+      "grad_norm": 0.5141358971595764,
+      "learning_rate": 9.469645981962579e-05,
+      "loss": 3.333,
+      "step": 5476
+    },
+    {
+      "epoch": 7.00384,
+      "grad_norm": 0.5374777913093567,
+      "learning_rate": 9.465607753398841e-05,
+      "loss": 3.3227,
+      "step": 5477
+    },
+    {
+      "epoch": 7.00512,
+      "grad_norm": 0.5191708207130432,
+      "learning_rate": 9.461569524835105e-05,
+      "loss": 3.4003,
+      "step": 5478
+    },
+    {
+      "epoch": 7.0064,
+      "grad_norm": 0.5081069469451904,
+      "learning_rate": 9.457531296271368e-05,
+      "loss": 3.3985,
+      "step": 5479
+    },
+    {
+      "epoch": 7.00768,
+      "grad_norm": 0.5270673036575317,
+      "learning_rate": 9.453493067707633e-05,
+      "loss": 3.3482,
+      "step": 5480
+    },
+    {
+      "epoch": 7.00896,
+      "grad_norm": 0.5204823017120361,
+      "learning_rate": 9.449454839143894e-05,
+      "loss": 3.3075,
+      "step": 5481
+    },
+    {
+      "epoch": 7.01024,
+      "grad_norm": 0.5076407194137573,
+      "learning_rate": 9.445416610580159e-05,
+      "loss": 3.2981,
+      "step": 5482
+    },
+    {
+      "epoch": 7.01152,
+      "grad_norm": 0.5210766196250916,
+      "learning_rate": 9.441378382016422e-05,
+      "loss": 3.4319,
+      "step": 5483
+    },
+    {
+      "epoch": 7.0128,
+      "grad_norm": 0.5091572403907776,
+      "learning_rate": 9.437340153452683e-05,
+      "loss": 3.3706,
+      "step": 5484
+    },
+    {
+      "epoch": 7.01408,
+      "grad_norm": 0.5025621056556702,
+      "learning_rate": 9.433301924888948e-05,
+      "loss": 3.274,
+      "step": 5485
+    },
+    {
+      "epoch": 7.01536,
+      "grad_norm": 0.5144043564796448,
+      "learning_rate": 9.42926369632521e-05,
+      "loss": 3.2991,
+      "step": 5486
+    },
+    {
+      "epoch": 7.01664,
+      "grad_norm": 0.5310031175613403,
+      "learning_rate": 9.425225467761475e-05,
+      "loss": 3.3532,
+      "step": 5487
+    },
+    {
+      "epoch": 7.01792,
+      "grad_norm": 0.506899893283844,
+      "learning_rate": 9.421187239197737e-05,
+      "loss": 3.3682,
+      "step": 5488
+    },
+    {
+      "epoch": 7.0192,
+      "grad_norm": 0.5226014256477356,
+      "learning_rate": 9.417149010634001e-05,
+      "loss": 3.3653,
+      "step": 5489
+    },
+    {
+      "epoch": 7.02048,
+      "grad_norm": 0.5002744197845459,
+      "learning_rate": 9.413110782070264e-05,
+      "loss": 3.3786,
+      "step": 5490
+    },
+    {
+      "epoch": 7.0217600000000004,
+      "grad_norm": 0.5125488638877869,
+      "learning_rate": 9.409072553506528e-05,
+      "loss": 3.3929,
+      "step": 5491
+    },
+    {
+      "epoch": 7.02304,
+      "grad_norm": 0.5321469902992249,
+      "learning_rate": 9.40503432494279e-05,
+      "loss": 3.3667,
+      "step": 5492
+    },
+    {
+      "epoch": 7.02432,
+      "grad_norm": 0.5275914072990417,
+      "learning_rate": 9.400996096379054e-05,
+      "loss": 3.3928,
+      "step": 5493
+    },
+    {
+      "epoch": 7.0256,
+      "grad_norm": 0.5226854681968689,
+      "learning_rate": 9.396957867815317e-05,
+      "loss": 3.3962,
+      "step": 5494
+    },
+    {
+      "epoch": 7.02688,
+      "grad_norm": 0.5123041272163391,
+      "learning_rate": 9.392919639251582e-05,
+      "loss": 3.2311,
+      "step": 5495
+    },
+    {
+      "epoch": 7.02816,
+      "grad_norm": 0.49747586250305176,
+      "learning_rate": 9.388881410687843e-05,
+      "loss": 3.3694,
+      "step": 5496
+    },
+    {
+      "epoch": 7.02944,
+      "grad_norm": 0.5055866241455078,
+      "learning_rate": 9.384843182124106e-05,
+      "loss": 3.3472,
+      "step": 5497
+    },
+    {
+      "epoch": 7.03072,
+      "grad_norm": 0.5089617371559143,
+      "learning_rate": 9.380804953560371e-05,
+      "loss": 3.3635,
+      "step": 5498
+    },
+    {
+      "epoch": 7.032,
+      "grad_norm": 0.5178612470626831,
+      "learning_rate": 9.376766724996634e-05,
+      "loss": 3.3485,
+      "step": 5499
+    },
+    {
+      "epoch": 7.03328,
+      "grad_norm": 0.5137771964073181,
+      "learning_rate": 9.372728496432898e-05,
+      "loss": 3.3414,
+      "step": 5500
+    },
+    {
+      "epoch": 7.03456,
+      "grad_norm": 0.5222017765045166,
+      "learning_rate": 9.36869026786916e-05,
+      "loss": 3.3504,
+      "step": 5501
+    },
+    {
+      "epoch": 7.03584,
+      "grad_norm": 0.506715714931488,
+      "learning_rate": 9.364652039305424e-05,
+      "loss": 3.342,
+      "step": 5502
+    },
+    {
+      "epoch": 7.03712,
+      "grad_norm": 0.5213739275932312,
+      "learning_rate": 9.360613810741687e-05,
+      "loss": 3.3492,
+      "step": 5503
+    },
+    {
+      "epoch": 7.0384,
+      "grad_norm": 0.5213996767997742,
+      "learning_rate": 9.356575582177952e-05,
+      "loss": 3.3364,
+      "step": 5504
+    },
+    {
+      "epoch": 7.03968,
+      "grad_norm": 0.5308609008789062,
+      "learning_rate": 9.352537353614213e-05,
+      "loss": 3.2428,
+      "step": 5505
+    },
+    {
+      "epoch": 7.04096,
+      "grad_norm": 0.520922064781189,
+      "learning_rate": 9.348499125050478e-05,
+      "loss": 3.3741,
+      "step": 5506
+    },
+    {
+      "epoch": 7.04224,
+      "grad_norm": 0.49800974130630493,
+      "learning_rate": 9.34446089648674e-05,
+      "loss": 3.35,
+      "step": 5507
+    },
+    {
+      "epoch": 7.04352,
+      "grad_norm": 0.53228360414505,
+      "learning_rate": 9.340422667923005e-05,
+      "loss": 3.3215,
+      "step": 5508
+    },
+    {
+      "epoch": 7.0448,
+      "grad_norm": 0.5500988960266113,
+      "learning_rate": 9.336384439359267e-05,
+      "loss": 3.4593,
+      "step": 5509
+    },
+    {
+      "epoch": 7.04608,
+      "grad_norm": 0.5122029185295105,
+      "learning_rate": 9.332346210795531e-05,
+      "loss": 3.3557,
+      "step": 5510
+    },
+    {
+      "epoch": 7.04736,
+      "grad_norm": 0.4989195764064789,
+      "learning_rate": 9.328307982231794e-05,
+      "loss": 3.3231,
+      "step": 5511
+    },
+    {
+      "epoch": 7.04864,
+      "grad_norm": 0.521932065486908,
+      "learning_rate": 9.324269753668056e-05,
+      "loss": 3.3911,
+      "step": 5512
+    },
+    {
+      "epoch": 7.04992,
+      "grad_norm": 0.5335235595703125,
+      "learning_rate": 9.32023152510432e-05,
+      "loss": 3.3212,
+      "step": 5513
+    },
+    {
+      "epoch": 7.0512,
+      "grad_norm": 0.5094990730285645,
+      "learning_rate": 9.316193296540583e-05,
+      "loss": 3.2888,
+      "step": 5514
+    },
+    {
+      "epoch": 7.05248,
+      "grad_norm": 0.5203416347503662,
+      "learning_rate": 9.312155067976847e-05,
+      "loss": 3.4157,
+      "step": 5515
+    },
+    {
+      "epoch": 7.05376,
+      "grad_norm": 0.5284749865531921,
+      "learning_rate": 9.308116839413109e-05,
+      "loss": 3.3612,
+      "step": 5516
+    },
+    {
+      "epoch": 7.05504,
+      "grad_norm": 0.5061756372451782,
+      "learning_rate": 9.304078610849373e-05,
+      "loss": 3.4084,
+      "step": 5517
+    },
+    {
+      "epoch": 7.05632,
+      "grad_norm": 0.5212084054946899,
+      "learning_rate": 9.300040382285636e-05,
+      "loss": 3.2901,
+      "step": 5518
+    },
+    {
+      "epoch": 7.0576,
+      "grad_norm": 0.528948187828064,
+      "learning_rate": 9.296002153721901e-05,
+      "loss": 3.3865,
+      "step": 5519
+    },
+    {
+      "epoch": 7.05888,
+      "grad_norm": 0.507111132144928,
+      "learning_rate": 9.291963925158162e-05,
+      "loss": 3.4119,
+      "step": 5520
+    },
+    {
+      "epoch": 7.06016,
+      "grad_norm": 0.5240997672080994,
+      "learning_rate": 9.287925696594427e-05,
+      "loss": 3.4243,
+      "step": 5521
+    },
+    {
+      "epoch": 7.06144,
+      "grad_norm": 0.5230147242546082,
+      "learning_rate": 9.28388746803069e-05,
+      "loss": 3.376,
+      "step": 5522
+    },
+    {
+      "epoch": 7.06272,
+      "grad_norm": 0.526611864566803,
+      "learning_rate": 9.279849239466954e-05,
+      "loss": 3.3711,
+      "step": 5523
+    },
+    {
+      "epoch": 7.064,
+      "grad_norm": 0.5132043957710266,
+      "learning_rate": 9.275811010903216e-05,
+      "loss": 3.2992,
+      "step": 5524
+    },
+    {
+      "epoch": 7.06528,
+      "grad_norm": 0.4980516731739044,
+      "learning_rate": 9.271772782339479e-05,
+      "loss": 3.2729,
+      "step": 5525
+    },
+    {
+      "epoch": 7.06656,
+      "grad_norm": 0.5084584355354309,
+      "learning_rate": 9.267734553775743e-05,
+      "loss": 3.2491,
+      "step": 5526
+    },
+    {
+      "epoch": 7.06784,
+      "grad_norm": 0.515275776386261,
+      "learning_rate": 9.263696325212006e-05,
+      "loss": 3.3712,
+      "step": 5527
+    },
+    {
+      "epoch": 7.06912,
+      "grad_norm": 0.5161235928535461,
+      "learning_rate": 9.259658096648269e-05,
+      "loss": 3.4019,
+      "step": 5528
+    },
+    {
+      "epoch": 7.0704,
+      "grad_norm": 0.5367205142974854,
+      "learning_rate": 9.255619868084532e-05,
+      "loss": 3.4287,
+      "step": 5529
+    },
+    {
+      "epoch": 7.07168,
+      "grad_norm": 0.5217522978782654,
+      "learning_rate": 9.251581639520796e-05,
+      "loss": 3.4843,
+      "step": 5530
+    },
+    {
+      "epoch": 7.07296,
+      "grad_norm": 0.5207579135894775,
+      "learning_rate": 9.24754341095706e-05,
+      "loss": 3.3275,
+      "step": 5531
+    },
+    {
+      "epoch": 7.07424,
+      "grad_norm": 0.5166409611701965,
+      "learning_rate": 9.243505182393323e-05,
+      "loss": 3.293,
+      "step": 5532
+    },
+    {
+      "epoch": 7.07552,
+      "grad_norm": 0.5201752185821533,
+      "learning_rate": 9.239466953829586e-05,
+      "loss": 3.3349,
+      "step": 5533
+    },
+    {
+      "epoch": 7.0768,
+      "grad_norm": 0.5157431364059448,
+      "learning_rate": 9.23542872526585e-05,
+      "loss": 3.3274,
+      "step": 5534
+    },
+    {
+      "epoch": 7.07808,
+      "grad_norm": 0.5315886735916138,
+      "learning_rate": 9.231390496702113e-05,
+      "loss": 3.3802,
+      "step": 5535
+    },
+    {
+      "epoch": 7.07936,
+      "grad_norm": 0.5255934000015259,
+      "learning_rate": 9.227352268138377e-05,
+      "loss": 3.3859,
+      "step": 5536
+    },
+    {
+      "epoch": 7.08064,
+      "grad_norm": 0.5185539126396179,
+      "learning_rate": 9.223314039574639e-05,
+      "loss": 3.3037,
+      "step": 5537
+    },
+    {
+      "epoch": 7.08192,
+      "grad_norm": 0.5157800912857056,
+      "learning_rate": 9.219275811010902e-05,
+      "loss": 3.4169,
+      "step": 5538
+    },
+    {
+      "epoch": 7.0832,
+      "grad_norm": 0.5315351486206055,
+      "learning_rate": 9.215237582447166e-05,
+      "loss": 3.3251,
+      "step": 5539
+    },
+    {
+      "epoch": 7.08448,
+      "grad_norm": 0.5046893358230591,
+      "learning_rate": 9.211199353883428e-05,
+      "loss": 3.3554,
+      "step": 5540
+    },
+    {
+      "epoch": 7.08576,
+      "grad_norm": 0.5140607357025146,
+      "learning_rate": 9.207161125319692e-05,
+      "loss": 3.3262,
+      "step": 5541
+    },
+    {
+      "epoch": 7.08704,
+      "grad_norm": 0.508526086807251,
+      "learning_rate": 9.203122896755955e-05,
+      "loss": 3.2896,
+      "step": 5542
+    },
+    {
+      "epoch": 7.08832,
+      "grad_norm": 0.5273757576942444,
+      "learning_rate": 9.19908466819222e-05,
+      "loss": 3.4774,
+      "step": 5543
+    },
+    {
+      "epoch": 7.0896,
+      "grad_norm": 0.5205854773521423,
+      "learning_rate": 9.195046439628481e-05,
+      "loss": 3.3347,
+      "step": 5544
+    },
+    {
+      "epoch": 7.09088,
+      "grad_norm": 0.5332198739051819,
+      "learning_rate": 9.191008211064746e-05,
+      "loss": 3.3342,
+      "step": 5545
+    },
+    {
+      "epoch": 7.09216,
+      "grad_norm": 0.5330145955085754,
+      "learning_rate": 9.186969982501009e-05,
+      "loss": 3.3769,
+      "step": 5546
+    },
+    {
+      "epoch": 7.09344,
+      "grad_norm": 0.5021401047706604,
+      "learning_rate": 9.182931753937273e-05,
+      "loss": 3.3476,
+      "step": 5547
+    },
+    {
+      "epoch": 7.09472,
+      "grad_norm": 0.5273832678794861,
+      "learning_rate": 9.178893525373535e-05,
+      "loss": 3.333,
+      "step": 5548
+    },
+    {
+      "epoch": 7.096,
+      "grad_norm": 0.5313226580619812,
+      "learning_rate": 9.174855296809799e-05,
+      "loss": 3.4208,
+      "step": 5549
+    },
+    {
+      "epoch": 7.09728,
+      "grad_norm": 0.5218507051467896,
+      "learning_rate": 9.170817068246062e-05,
+      "loss": 3.3408,
+      "step": 5550
+    },
+    {
+      "epoch": 7.09856,
+      "grad_norm": 0.5151420831680298,
+      "learning_rate": 9.166778839682325e-05,
+      "loss": 3.3944,
+      "step": 5551
+    },
+    {
+      "epoch": 7.09984,
+      "grad_norm": 0.5464621186256409,
+      "learning_rate": 9.162740611118588e-05,
+      "loss": 3.3923,
+      "step": 5552
+    },
+    {
+      "epoch": 7.10112,
+      "grad_norm": 0.5159791707992554,
+      "learning_rate": 9.158702382554851e-05,
+      "loss": 3.379,
+      "step": 5553
+    },
+    {
+      "epoch": 7.1024,
+      "grad_norm": 0.526742160320282,
+      "learning_rate": 9.154664153991115e-05,
+      "loss": 3.3621,
+      "step": 5554
+    },
+    {
+      "epoch": 7.10368,
+      "grad_norm": 0.5101780891418457,
+      "learning_rate": 9.150625925427378e-05,
+      "loss": 3.3674,
+      "step": 5555
+    },
+    {
+      "epoch": 7.10496,
+      "grad_norm": 0.5260143876075745,
+      "learning_rate": 9.146587696863641e-05,
+      "loss": 3.3164,
+      "step": 5556
+    },
+    {
+      "epoch": 7.10624,
+      "grad_norm": 0.5367382168769836,
+      "learning_rate": 9.142549468299904e-05,
+      "loss": 3.3863,
+      "step": 5557
+    },
+    {
+      "epoch": 7.10752,
+      "grad_norm": 0.5072177648544312,
+      "learning_rate": 9.138511239736169e-05,
+      "loss": 3.335,
+      "step": 5558
+    },
+    {
+      "epoch": 7.1088,
+      "grad_norm": 0.5360726118087769,
+      "learning_rate": 9.134473011172432e-05,
+      "loss": 3.3996,
+      "step": 5559
+    },
+    {
+      "epoch": 7.11008,
+      "grad_norm": 0.5281416177749634,
+      "learning_rate": 9.130434782608695e-05,
+      "loss": 3.3302,
+      "step": 5560
+    },
+    {
+      "epoch": 7.11136,
+      "grad_norm": 0.526565432548523,
+      "learning_rate": 9.126396554044958e-05,
+      "loss": 3.3779,
+      "step": 5561
+    },
+    {
+      "epoch": 7.11264,
+      "grad_norm": 0.5221660733222961,
+      "learning_rate": 9.122358325481222e-05,
+      "loss": 3.3879,
+      "step": 5562
+    },
+    {
+      "epoch": 7.11392,
+      "grad_norm": 0.5545408129692078,
+      "learning_rate": 9.118320096917485e-05,
+      "loss": 3.3599,
+      "step": 5563
+    },
+    {
+      "epoch": 7.1152,
+      "grad_norm": 0.5166964530944824,
+      "learning_rate": 9.114281868353748e-05,
+      "loss": 3.3786,
+      "step": 5564
+    },
+    {
+      "epoch": 7.11648,
+      "grad_norm": 0.5115810632705688,
+      "learning_rate": 9.110243639790011e-05,
+      "loss": 3.3983,
+      "step": 5565
+    },
+    {
+      "epoch": 7.11776,
+      "grad_norm": 0.527052640914917,
+      "learning_rate": 9.106205411226274e-05,
+      "loss": 3.3886,
+      "step": 5566
+    },
+    {
+      "epoch": 7.11904,
+      "grad_norm": 0.5167108774185181,
+      "learning_rate": 9.102167182662539e-05,
+      "loss": 3.3837,
+      "step": 5567
+    },
+    {
+      "epoch": 7.12032,
+      "grad_norm": 0.5196815729141235,
+      "learning_rate": 9.0981289540988e-05,
+      "loss": 3.4007,
+      "step": 5568
+    },
+    {
+      "epoch": 7.1216,
+      "grad_norm": 0.5209911465644836,
+      "learning_rate": 9.094090725535065e-05,
+      "loss": 3.3465,
+      "step": 5569
+    },
+    {
+      "epoch": 7.12288,
+      "grad_norm": 0.5412890315055847,
+      "learning_rate": 9.090052496971328e-05,
+      "loss": 3.3554,
+      "step": 5570
+    },
+    {
+      "epoch": 7.12416,
+      "grad_norm": 0.5221083164215088,
+      "learning_rate": 9.086014268407592e-05,
+      "loss": 3.3713,
+      "step": 5571
+    },
+    {
+      "epoch": 7.12544,
+      "grad_norm": 0.5288161039352417,
+      "learning_rate": 9.081976039843854e-05,
+      "loss": 3.3445,
+      "step": 5572
+    },
+    {
+      "epoch": 7.12672,
+      "grad_norm": 0.5313680171966553,
+      "learning_rate": 9.077937811280118e-05,
+      "loss": 3.3708,
+      "step": 5573
+    },
+    {
+      "epoch": 7.128,
+      "grad_norm": 0.5292986035346985,
+      "learning_rate": 9.073899582716381e-05,
+      "loss": 3.3137,
+      "step": 5574
+    },
+    {
+      "epoch": 7.12928,
+      "grad_norm": 0.5138509273529053,
+      "learning_rate": 9.069861354152645e-05,
+      "loss": 3.3887,
+      "step": 5575
+    },
+    {
+      "epoch": 7.13056,
+      "grad_norm": 0.540255606174469,
+      "learning_rate": 9.065823125588907e-05,
+      "loss": 3.3745,
+      "step": 5576
+    },
+    {
+      "epoch": 7.13184,
+      "grad_norm": 0.5347320437431335,
+      "learning_rate": 9.061784897025171e-05,
+      "loss": 3.3579,
+      "step": 5577
+    },
+    {
+      "epoch": 7.13312,
+      "grad_norm": 0.49910813570022583,
+      "learning_rate": 9.057746668461434e-05,
+      "loss": 3.3225,
+      "step": 5578
+    },
+    {
+      "epoch": 7.1344,
+      "grad_norm": 0.5152397751808167,
+      "learning_rate": 9.053708439897696e-05,
+      "loss": 3.3496,
+      "step": 5579
+    },
+    {
+      "epoch": 7.13568,
+      "grad_norm": 0.5248477458953857,
+      "learning_rate": 9.04967021133396e-05,
+      "loss": 3.3574,
+      "step": 5580
+    },
+    {
+      "epoch": 7.13696,
+      "grad_norm": 0.5127560496330261,
+      "learning_rate": 9.045631982770223e-05,
+      "loss": 3.3692,
+      "step": 5581
+    },
+    {
+      "epoch": 7.13824,
+      "grad_norm": 0.5316688418388367,
+      "learning_rate": 9.041593754206488e-05,
+      "loss": 3.3383,
+      "step": 5582
+    },
+    {
+      "epoch": 7.13952,
+      "grad_norm": 0.5397014021873474,
+      "learning_rate": 9.03755552564275e-05,
+      "loss": 3.3939,
+      "step": 5583
+    },
+    {
+      "epoch": 7.1408,
+      "grad_norm": 0.5297314524650574,
+      "learning_rate": 9.033517297079014e-05,
+      "loss": 3.3205,
+      "step": 5584
+    },
+    {
+      "epoch": 7.14208,
+      "grad_norm": 0.5164687633514404,
+      "learning_rate": 9.029479068515277e-05,
+      "loss": 3.311,
+      "step": 5585
+    },
+    {
+      "epoch": 7.14336,
+      "grad_norm": 0.5183284878730774,
+      "learning_rate": 9.025440839951541e-05,
+      "loss": 3.3063,
+      "step": 5586
+    },
+    {
+      "epoch": 7.14464,
+      "grad_norm": 0.508436918258667,
+      "learning_rate": 9.021402611387804e-05,
+      "loss": 3.3496,
+      "step": 5587
+    },
+    {
+      "epoch": 7.14592,
+      "grad_norm": 0.5277997851371765,
+      "learning_rate": 9.017364382824067e-05,
+      "loss": 3.3782,
+      "step": 5588
+    },
+    {
+      "epoch": 7.1472,
+      "grad_norm": 0.5223129987716675,
+      "learning_rate": 9.01332615426033e-05,
+      "loss": 3.3175,
+      "step": 5589
+    },
+    {
+      "epoch": 7.14848,
+      "grad_norm": 0.4987069070339203,
+      "learning_rate": 9.009287925696595e-05,
+      "loss": 3.3346,
+      "step": 5590
+    },
+    {
+      "epoch": 7.14976,
+      "grad_norm": 0.5348743796348572,
+      "learning_rate": 9.005249697132858e-05,
+      "loss": 3.3543,
+      "step": 5591
+    },
+    {
+      "epoch": 7.15104,
+      "grad_norm": 0.5277005434036255,
+      "learning_rate": 9.001211468569119e-05,
+      "loss": 3.4288,
+      "step": 5592
+    },
+    {
+      "epoch": 7.15232,
+      "grad_norm": 0.5248242020606995,
+      "learning_rate": 8.997173240005384e-05,
+      "loss": 3.3001,
+      "step": 5593
+    },
+    {
+      "epoch": 7.1536,
+      "grad_norm": 0.5100671052932739,
+      "learning_rate": 8.993135011441647e-05,
+      "loss": 3.3571,
+      "step": 5594
+    },
+    {
+      "epoch": 7.15488,
+      "grad_norm": 0.5024705529212952,
+      "learning_rate": 8.989096782877911e-05,
+      "loss": 3.35,
+      "step": 5595
+    },
+    {
+      "epoch": 7.15616,
+      "grad_norm": 0.5296070575714111,
+      "learning_rate": 8.985058554314173e-05,
+      "loss": 3.3975,
+      "step": 5596
+    },
+    {
+      "epoch": 7.15744,
+      "grad_norm": 0.535521388053894,
+      "learning_rate": 8.981020325750437e-05,
+      "loss": 3.4045,
+      "step": 5597
+    },
+    {
+      "epoch": 7.15872,
+      "grad_norm": 0.5225574374198914,
+      "learning_rate": 8.9769820971867e-05,
+      "loss": 3.3518,
+      "step": 5598
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.509151816368103,
+      "learning_rate": 8.972943868622964e-05,
+      "loss": 3.438,
+      "step": 5599
+    },
+    {
+      "epoch": 7.16128,
+      "grad_norm": 0.5296933054924011,
+      "learning_rate": 8.968905640059226e-05,
+      "loss": 3.4388,
+      "step": 5600
+    },
+    {
+      "epoch": 7.16256,
+      "grad_norm": 0.527597963809967,
+      "learning_rate": 8.96486741149549e-05,
+      "loss": 3.36,
+      "step": 5601
+    },
+    {
+      "epoch": 7.16384,
+      "grad_norm": 0.5163085460662842,
+      "learning_rate": 8.960829182931753e-05,
+      "loss": 3.3173,
+      "step": 5602
+    },
+    {
+      "epoch": 7.16512,
+      "grad_norm": 0.5081968307495117,
+      "learning_rate": 8.956790954368018e-05,
+      "loss": 3.3379,
+      "step": 5603
+    },
+    {
+      "epoch": 7.1664,
+      "grad_norm": 0.5179334878921509,
+      "learning_rate": 8.95275272580428e-05,
+      "loss": 3.3674,
+      "step": 5604
+    },
+    {
+      "epoch": 7.16768,
+      "grad_norm": 0.5077155828475952,
+      "learning_rate": 8.948714497240542e-05,
+      "loss": 3.3791,
+      "step": 5605
+    },
+    {
+      "epoch": 7.16896,
+      "grad_norm": 0.5360821485519409,
+      "learning_rate": 8.944676268676807e-05,
+      "loss": 3.3437,
+      "step": 5606
+    },
+    {
+      "epoch": 7.17024,
+      "grad_norm": 0.5339843034744263,
+      "learning_rate": 8.940638040113068e-05,
+      "loss": 3.4416,
+      "step": 5607
+    },
+    {
+      "epoch": 7.17152,
+      "grad_norm": 0.5199359059333801,
+      "learning_rate": 8.936599811549333e-05,
+      "loss": 3.4527,
+      "step": 5608
+    },
+    {
+      "epoch": 7.1728,
+      "grad_norm": 0.5207464098930359,
+      "learning_rate": 8.932561582985596e-05,
+      "loss": 3.3807,
+      "step": 5609
+    },
+    {
+      "epoch": 7.17408,
+      "grad_norm": 0.5150240659713745,
+      "learning_rate": 8.92852335442186e-05,
+      "loss": 3.3255,
+      "step": 5610
+    },
+    {
+      "epoch": 7.17536,
+      "grad_norm": 0.5204200744628906,
+      "learning_rate": 8.924485125858122e-05,
+      "loss": 3.3301,
+      "step": 5611
+    },
+    {
+      "epoch": 7.17664,
+      "grad_norm": 0.5184539556503296,
+      "learning_rate": 8.920446897294386e-05,
+      "loss": 3.3256,
+      "step": 5612
+    },
+    {
+      "epoch": 7.17792,
+      "grad_norm": 0.5265978574752808,
+      "learning_rate": 8.916408668730649e-05,
+      "loss": 3.3752,
+      "step": 5613
+    },
+    {
+      "epoch": 7.1792,
+      "grad_norm": 0.5245677828788757,
+      "learning_rate": 8.912370440166914e-05,
+      "loss": 3.3356,
+      "step": 5614
+    },
+    {
+      "epoch": 7.18048,
+      "grad_norm": 0.5343059301376343,
+      "learning_rate": 8.908332211603175e-05,
+      "loss": 3.3374,
+      "step": 5615
+    },
+    {
+      "epoch": 7.18176,
+      "grad_norm": 0.5347888469696045,
+      "learning_rate": 8.90429398303944e-05,
+      "loss": 3.3665,
+      "step": 5616
+    },
+    {
+      "epoch": 7.18304,
+      "grad_norm": 0.5191536545753479,
+      "learning_rate": 8.900255754475703e-05,
+      "loss": 3.3387,
+      "step": 5617
+    },
+    {
+      "epoch": 7.18432,
+      "grad_norm": 0.5158880949020386,
+      "learning_rate": 8.896217525911966e-05,
+      "loss": 3.312,
+      "step": 5618
+    },
+    {
+      "epoch": 7.1856,
+      "grad_norm": 0.5173276662826538,
+      "learning_rate": 8.892179297348229e-05,
+      "loss": 3.352,
+      "step": 5619
+    },
+    {
+      "epoch": 7.18688,
+      "grad_norm": 0.515876829624176,
+      "learning_rate": 8.888141068784492e-05,
+      "loss": 3.3369,
+      "step": 5620
+    },
+    {
+      "epoch": 7.18816,
+      "grad_norm": 0.5200967788696289,
+      "learning_rate": 8.884102840220756e-05,
+      "loss": 3.3315,
+      "step": 5621
+    },
+    {
+      "epoch": 7.18944,
+      "grad_norm": 0.5234977602958679,
+      "learning_rate": 8.880064611657019e-05,
+      "loss": 3.3648,
+      "step": 5622
+    },
+    {
+      "epoch": 7.19072,
+      "grad_norm": 0.5436128377914429,
+      "learning_rate": 8.876026383093282e-05,
+      "loss": 3.3157,
+      "step": 5623
+    },
+    {
+      "epoch": 7.192,
+      "grad_norm": 0.5315131545066833,
+      "learning_rate": 8.871988154529545e-05,
+      "loss": 3.3696,
+      "step": 5624
+    },
+    {
+      "epoch": 7.19328,
+      "grad_norm": 0.5198819637298584,
+      "learning_rate": 8.867949925965809e-05,
+      "loss": 3.3841,
+      "step": 5625
+    },
+    {
+      "epoch": 7.19456,
+      "grad_norm": 0.5186936259269714,
+      "learning_rate": 8.863911697402072e-05,
+      "loss": 3.3401,
+      "step": 5626
+    },
+    {
+      "epoch": 7.19584,
+      "grad_norm": 0.5158634781837463,
+      "learning_rate": 8.859873468838337e-05,
+      "loss": 3.3416,
+      "step": 5627
+    },
+    {
+      "epoch": 7.19712,
+      "grad_norm": 0.5173478722572327,
+      "learning_rate": 8.855835240274598e-05,
+      "loss": 3.3616,
+      "step": 5628
+    },
+    {
+      "epoch": 7.1984,
+      "grad_norm": 0.5127228498458862,
+      "learning_rate": 8.851797011710863e-05,
+      "loss": 3.4288,
+      "step": 5629
+    },
+    {
+      "epoch": 7.19968,
+      "grad_norm": 0.5157126188278198,
+      "learning_rate": 8.847758783147126e-05,
+      "loss": 3.3122,
+      "step": 5630
+    },
+    {
+      "epoch": 7.20096,
+      "grad_norm": 0.5168270468711853,
+      "learning_rate": 8.84372055458339e-05,
+      "loss": 3.3206,
+      "step": 5631
+    },
+    {
+      "epoch": 7.20224,
+      "grad_norm": 0.5436537861824036,
+      "learning_rate": 8.839682326019652e-05,
+      "loss": 3.415,
+      "step": 5632
+    },
+    {
+      "epoch": 7.20352,
+      "grad_norm": 0.5260339975357056,
+      "learning_rate": 8.835644097455915e-05,
+      "loss": 3.3191,
+      "step": 5633
+    },
+    {
+      "epoch": 7.2048,
+      "grad_norm": 0.5202707648277283,
+      "learning_rate": 8.831605868892179e-05,
+      "loss": 3.3391,
+      "step": 5634
+    },
+    {
+      "epoch": 7.20608,
+      "grad_norm": 0.5285276174545288,
+      "learning_rate": 8.827567640328441e-05,
+      "loss": 3.3521,
+      "step": 5635
+    },
+    {
+      "epoch": 7.2073599999999995,
+      "grad_norm": 0.5395596027374268,
+      "learning_rate": 8.823529411764705e-05,
+      "loss": 3.3539,
+      "step": 5636
+    },
+    {
+      "epoch": 7.20864,
+      "grad_norm": 0.543609619140625,
+      "learning_rate": 8.819491183200968e-05,
+      "loss": 3.3652,
+      "step": 5637
+    },
+    {
+      "epoch": 7.20992,
+      "grad_norm": 0.5348373651504517,
+      "learning_rate": 8.815452954637232e-05,
+      "loss": 3.3671,
+      "step": 5638
+    },
+    {
+      "epoch": 7.2112,
+      "grad_norm": 0.5275009274482727,
+      "learning_rate": 8.811414726073494e-05,
+      "loss": 3.3737,
+      "step": 5639
+    },
+    {
+      "epoch": 7.21248,
+      "grad_norm": 0.5249333381652832,
+      "learning_rate": 8.807376497509758e-05,
+      "loss": 3.4308,
+      "step": 5640
+    },
+    {
+      "epoch": 7.21376,
+      "grad_norm": 0.5271218419075012,
+      "learning_rate": 8.803338268946022e-05,
+      "loss": 3.3363,
+      "step": 5641
+    },
+    {
+      "epoch": 7.21504,
+      "grad_norm": 0.5315967798233032,
+      "learning_rate": 8.799300040382286e-05,
+      "loss": 3.4132,
+      "step": 5642
+    },
+    {
+      "epoch": 7.21632,
+      "grad_norm": 0.5222828984260559,
+      "learning_rate": 8.795261811818548e-05,
+      "loss": 3.3832,
+      "step": 5643
+    },
+    {
+      "epoch": 7.2176,
+      "grad_norm": 0.5247751474380493,
+      "learning_rate": 8.791223583254812e-05,
+      "loss": 3.255,
+      "step": 5644
+    },
+    {
+      "epoch": 7.21888,
+      "grad_norm": 0.5310077667236328,
+      "learning_rate": 8.787185354691075e-05,
+      "loss": 3.3478,
+      "step": 5645
+    },
+    {
+      "epoch": 7.22016,
+      "grad_norm": 0.5162127614021301,
+      "learning_rate": 8.783147126127338e-05,
+      "loss": 3.3882,
+      "step": 5646
+    },
+    {
+      "epoch": 7.22144,
+      "grad_norm": 0.5231219530105591,
+      "learning_rate": 8.779108897563601e-05,
+      "loss": 3.3114,
+      "step": 5647
+    },
+    {
+      "epoch": 7.22272,
+      "grad_norm": 0.5083349943161011,
+      "learning_rate": 8.775070668999864e-05,
+      "loss": 3.2972,
+      "step": 5648
+    },
+    {
+      "epoch": 7.224,
+      "grad_norm": 0.5252796411514282,
+      "learning_rate": 8.771032440436128e-05,
+      "loss": 3.2982,
+      "step": 5649
+    },
+    {
+      "epoch": 7.22528,
+      "grad_norm": 0.5133306980133057,
+      "learning_rate": 8.766994211872391e-05,
+      "loss": 3.3952,
+      "step": 5650
+    },
+    {
+      "epoch": 7.22656,
+      "grad_norm": 0.5313632488250732,
+      "learning_rate": 8.762955983308654e-05,
+      "loss": 3.4222,
+      "step": 5651
+    },
+    {
+      "epoch": 7.22784,
+      "grad_norm": 0.5258992910385132,
+      "learning_rate": 8.758917754744917e-05,
+      "loss": 3.3971,
+      "step": 5652
+    },
+    {
+      "epoch": 7.22912,
+      "grad_norm": 0.5275927186012268,
+      "learning_rate": 8.754879526181182e-05,
+      "loss": 3.3187,
+      "step": 5653
+    },
+    {
+      "epoch": 7.2304,
+      "grad_norm": 0.5042412877082825,
+      "learning_rate": 8.750841297617445e-05,
+      "loss": 3.3668,
+      "step": 5654
+    },
+    {
+      "epoch": 7.23168,
+      "grad_norm": 0.528590202331543,
+      "learning_rate": 8.746803069053708e-05,
+      "loss": 3.4068,
+      "step": 5655
+    },
+    {
+      "epoch": 7.23296,
+      "grad_norm": 0.5280982255935669,
+      "learning_rate": 8.742764840489971e-05,
+      "loss": 3.4039,
+      "step": 5656
+    },
+    {
+      "epoch": 7.23424,
+      "grad_norm": 0.5211199522018433,
+      "learning_rate": 8.738726611926235e-05,
+      "loss": 3.3706,
+      "step": 5657
+    },
+    {
+      "epoch": 7.23552,
+      "grad_norm": 0.5214046239852905,
+      "learning_rate": 8.734688383362498e-05,
+      "loss": 3.3703,
+      "step": 5658
+    },
+    {
+      "epoch": 7.2368,
+      "grad_norm": 0.5262265205383301,
+      "learning_rate": 8.73065015479876e-05,
+      "loss": 3.2918,
+      "step": 5659
+    },
+    {
+      "epoch": 7.23808,
+      "grad_norm": 0.534403920173645,
+      "learning_rate": 8.726611926235024e-05,
+      "loss": 3.3363,
+      "step": 5660
+    },
+    {
+      "epoch": 7.23936,
+      "grad_norm": 0.5397913455963135,
+      "learning_rate": 8.722573697671287e-05,
+      "loss": 3.3867,
+      "step": 5661
+    },
+    {
+      "epoch": 7.24064,
+      "grad_norm": 0.5387073755264282,
+      "learning_rate": 8.718535469107551e-05,
+      "loss": 3.4315,
+      "step": 5662
+    },
+    {
+      "epoch": 7.24192,
+      "grad_norm": 0.5351658463478088,
+      "learning_rate": 8.714497240543813e-05,
+      "loss": 3.4319,
+      "step": 5663
+    },
+    {
+      "epoch": 7.2432,
+      "grad_norm": 0.5469005107879639,
+      "learning_rate": 8.710459011980077e-05,
+      "loss": 3.4048,
+      "step": 5664
+    },
+    {
+      "epoch": 7.24448,
+      "grad_norm": 0.539505660533905,
+      "learning_rate": 8.70642078341634e-05,
+      "loss": 3.4154,
+      "step": 5665
+    },
+    {
+      "epoch": 7.24576,
+      "grad_norm": 0.543161928653717,
+      "learning_rate": 8.702382554852605e-05,
+      "loss": 3.4247,
+      "step": 5666
+    },
+    {
+      "epoch": 7.24704,
+      "grad_norm": 0.5213647484779358,
+      "learning_rate": 8.698344326288866e-05,
+      "loss": 3.3403,
+      "step": 5667
+    },
+    {
+      "epoch": 7.24832,
+      "grad_norm": 0.5247822403907776,
+      "learning_rate": 8.694306097725131e-05,
+      "loss": 3.3623,
+      "step": 5668
+    },
+    {
+      "epoch": 7.2496,
+      "grad_norm": 0.5359296202659607,
+      "learning_rate": 8.690267869161394e-05,
+      "loss": 3.3437,
+      "step": 5669
+    },
+    {
+      "epoch": 7.25088,
+      "grad_norm": 0.5125774145126343,
+      "learning_rate": 8.686229640597658e-05,
+      "loss": 3.4178,
+      "step": 5670
+    },
+    {
+      "epoch": 7.25216,
+      "grad_norm": 0.5092083215713501,
+      "learning_rate": 8.68219141203392e-05,
+      "loss": 3.3376,
+      "step": 5671
+    },
+    {
+      "epoch": 7.25344,
+      "grad_norm": 0.5361836552619934,
+      "learning_rate": 8.678153183470183e-05,
+      "loss": 3.2985,
+      "step": 5672
+    },
+    {
+      "epoch": 7.25472,
+      "grad_norm": 0.525852382183075,
+      "learning_rate": 8.674114954906447e-05,
+      "loss": 3.3765,
+      "step": 5673
+    },
+    {
+      "epoch": 7.256,
+      "grad_norm": 0.5261217355728149,
+      "learning_rate": 8.670076726342709e-05,
+      "loss": 3.3436,
+      "step": 5674
+    },
+    {
+      "epoch": 7.25728,
+      "grad_norm": 0.5285188555717468,
+      "learning_rate": 8.666038497778973e-05,
+      "loss": 3.4639,
+      "step": 5675
+    },
+    {
+      "epoch": 7.25856,
+      "grad_norm": 0.5249500274658203,
+      "learning_rate": 8.662000269215236e-05,
+      "loss": 3.3356,
+      "step": 5676
+    },
+    {
+      "epoch": 7.25984,
+      "grad_norm": 0.5341816544532776,
+      "learning_rate": 8.6579620406515e-05,
+      "loss": 3.3849,
+      "step": 5677
+    },
+    {
+      "epoch": 7.26112,
+      "grad_norm": 0.538826048374176,
+      "learning_rate": 8.653923812087764e-05,
+      "loss": 3.4035,
+      "step": 5678
+    },
+    {
+      "epoch": 7.2624,
+      "grad_norm": 0.5222164392471313,
+      "learning_rate": 8.649885583524027e-05,
+      "loss": 3.3679,
+      "step": 5679
+    },
+    {
+      "epoch": 7.26368,
+      "grad_norm": 0.5226588249206543,
+      "learning_rate": 8.64584735496029e-05,
+      "loss": 3.4012,
+      "step": 5680
+    },
+    {
+      "epoch": 7.26496,
+      "grad_norm": 0.5181993246078491,
+      "learning_rate": 8.641809126396554e-05,
+      "loss": 3.3866,
+      "step": 5681
+    },
+    {
+      "epoch": 7.26624,
+      "grad_norm": 0.5294241905212402,
+      "learning_rate": 8.637770897832817e-05,
+      "loss": 3.4343,
+      "step": 5682
+    },
+    {
+      "epoch": 7.26752,
+      "grad_norm": 0.5253875851631165,
+      "learning_rate": 8.63373266926908e-05,
+      "loss": 3.3867,
+      "step": 5683
+    },
+    {
+      "epoch": 7.2688,
+      "grad_norm": 0.5317622423171997,
+      "learning_rate": 8.629694440705343e-05,
+      "loss": 3.3478,
+      "step": 5684
+    },
+    {
+      "epoch": 7.27008,
+      "grad_norm": 0.5254741907119751,
+      "learning_rate": 8.625656212141606e-05,
+      "loss": 3.4261,
+      "step": 5685
+    },
+    {
+      "epoch": 7.27136,
+      "grad_norm": 0.5245876312255859,
+      "learning_rate": 8.62161798357787e-05,
+      "loss": 3.3631,
+      "step": 5686
+    },
+    {
+      "epoch": 7.27264,
+      "grad_norm": 0.5217685699462891,
+      "learning_rate": 8.617579755014132e-05,
+      "loss": 3.3236,
+      "step": 5687
+    },
+    {
+      "epoch": 7.27392,
+      "grad_norm": 0.5240602493286133,
+      "learning_rate": 8.613541526450396e-05,
+      "loss": 3.3932,
+      "step": 5688
+    },
+    {
+      "epoch": 7.2752,
+      "grad_norm": 0.5256776809692383,
+      "learning_rate": 8.60950329788666e-05,
+      "loss": 3.4072,
+      "step": 5689
+    },
+    {
+      "epoch": 7.27648,
+      "grad_norm": 0.5401302576065063,
+      "learning_rate": 8.605465069322924e-05,
+      "loss": 3.3593,
+      "step": 5690
+    },
+    {
+      "epoch": 7.27776,
+      "grad_norm": 0.5390993356704712,
+      "learning_rate": 8.601426840759185e-05,
+      "loss": 3.3646,
+      "step": 5691
+    },
+    {
+      "epoch": 7.27904,
+      "grad_norm": 0.5312697887420654,
+      "learning_rate": 8.59738861219545e-05,
+      "loss": 3.4073,
+      "step": 5692
+    },
+    {
+      "epoch": 7.28032,
+      "grad_norm": 0.5275899171829224,
+      "learning_rate": 8.593350383631713e-05,
+      "loss": 3.3039,
+      "step": 5693
+    },
+    {
+      "epoch": 7.2816,
+      "grad_norm": 0.5374751687049866,
+      "learning_rate": 8.589312155067977e-05,
+      "loss": 3.3517,
+      "step": 5694
+    },
+    {
+      "epoch": 7.2828800000000005,
+      "grad_norm": 0.5126158595085144,
+      "learning_rate": 8.585273926504239e-05,
+      "loss": 3.3677,
+      "step": 5695
+    },
+    {
+      "epoch": 7.28416,
+      "grad_norm": 0.5134846568107605,
+      "learning_rate": 8.581235697940503e-05,
+      "loss": 3.2869,
+      "step": 5696
+    },
+    {
+      "epoch": 7.28544,
+      "grad_norm": 0.5251624584197998,
+      "learning_rate": 8.577197469376766e-05,
+      "loss": 3.3785,
+      "step": 5697
+    },
+    {
+      "epoch": 7.28672,
+      "grad_norm": 0.5131708979606628,
+      "learning_rate": 8.57315924081303e-05,
+      "loss": 3.3422,
+      "step": 5698
+    },
+    {
+      "epoch": 7.288,
+      "grad_norm": 0.5143713355064392,
+      "learning_rate": 8.569121012249292e-05,
+      "loss": 3.3171,
+      "step": 5699
+    },
+    {
+      "epoch": 7.28928,
+      "grad_norm": 0.5281633734703064,
+      "learning_rate": 8.565082783685555e-05,
+      "loss": 3.324,
+      "step": 5700
+    },
+    {
+      "epoch": 7.29056,
+      "grad_norm": 0.5322510600090027,
+      "learning_rate": 8.56104455512182e-05,
+      "loss": 3.3639,
+      "step": 5701
+    },
+    {
+      "epoch": 7.29184,
+      "grad_norm": 0.5279068946838379,
+      "learning_rate": 8.557006326558081e-05,
+      "loss": 3.3392,
+      "step": 5702
+    },
+    {
+      "epoch": 7.29312,
+      "grad_norm": 0.5267346501350403,
+      "learning_rate": 8.552968097994346e-05,
+      "loss": 3.4094,
+      "step": 5703
+    },
+    {
+      "epoch": 7.2943999999999996,
+      "grad_norm": 0.512809693813324,
+      "learning_rate": 8.548929869430609e-05,
+      "loss": 3.3388,
+      "step": 5704
+    },
+    {
+      "epoch": 7.29568,
+      "grad_norm": 0.5133838057518005,
+      "learning_rate": 8.544891640866873e-05,
+      "loss": 3.3506,
+      "step": 5705
+    },
+    {
+      "epoch": 7.29696,
+      "grad_norm": 0.5310163497924805,
+      "learning_rate": 8.540853412303135e-05,
+      "loss": 3.3633,
+      "step": 5706
+    },
+    {
+      "epoch": 7.29824,
+      "grad_norm": 0.5210456848144531,
+      "learning_rate": 8.536815183739399e-05,
+      "loss": 3.3905,
+      "step": 5707
+    },
+    {
+      "epoch": 7.29952,
+      "grad_norm": 0.5178525447845459,
+      "learning_rate": 8.532776955175662e-05,
+      "loss": 3.3811,
+      "step": 5708
+    },
+    {
+      "epoch": 7.3008,
+      "grad_norm": 0.5075005292892456,
+      "learning_rate": 8.528738726611926e-05,
+      "loss": 3.4205,
+      "step": 5709
+    },
+    {
+      "epoch": 7.30208,
+      "grad_norm": 0.5155579447746277,
+      "learning_rate": 8.524700498048188e-05,
+      "loss": 3.4302,
+      "step": 5710
+    },
+    {
+      "epoch": 7.30336,
+      "grad_norm": 0.5333580374717712,
+      "learning_rate": 8.520662269484452e-05,
+      "loss": 3.4141,
+      "step": 5711
+    },
+    {
+      "epoch": 7.30464,
+      "grad_norm": 0.520826518535614,
+      "learning_rate": 8.516624040920715e-05,
+      "loss": 3.3472,
+      "step": 5712
+    },
+    {
+      "epoch": 7.30592,
+      "grad_norm": 0.5248406529426575,
+      "learning_rate": 8.512585812356978e-05,
+      "loss": 3.3902,
+      "step": 5713
+    },
+    {
+      "epoch": 7.3072,
+      "grad_norm": 0.5231212973594666,
+      "learning_rate": 8.508547583793243e-05,
+      "loss": 3.376,
+      "step": 5714
+    },
+    {
+      "epoch": 7.30848,
+      "grad_norm": 0.5204564929008484,
+      "learning_rate": 8.504509355229504e-05,
+      "loss": 3.3655,
+      "step": 5715
+    },
+    {
+      "epoch": 7.30976,
+      "grad_norm": 0.5213709473609924,
+      "learning_rate": 8.500471126665769e-05,
+      "loss": 3.3372,
+      "step": 5716
+    },
+    {
+      "epoch": 7.31104,
+      "grad_norm": 0.5297428369522095,
+      "learning_rate": 8.496432898102032e-05,
+      "loss": 3.3941,
+      "step": 5717
+    },
+    {
+      "epoch": 7.31232,
+      "grad_norm": 0.5178290605545044,
+      "learning_rate": 8.492394669538296e-05,
+      "loss": 3.3335,
+      "step": 5718
+    },
+    {
+      "epoch": 7.3136,
+      "grad_norm": 0.537998616695404,
+      "learning_rate": 8.488356440974558e-05,
+      "loss": 3.4031,
+      "step": 5719
+    },
+    {
+      "epoch": 7.31488,
+      "grad_norm": 0.5190197825431824,
+      "learning_rate": 8.484318212410822e-05,
+      "loss": 3.435,
+      "step": 5720
+    },
+    {
+      "epoch": 7.31616,
+      "grad_norm": 0.5143560767173767,
+      "learning_rate": 8.480279983847085e-05,
+      "loss": 3.3524,
+      "step": 5721
+    },
+    {
+      "epoch": 7.31744,
+      "grad_norm": 0.5168914198875427,
+      "learning_rate": 8.47624175528335e-05,
+      "loss": 3.365,
+      "step": 5722
+    },
+    {
+      "epoch": 7.31872,
+      "grad_norm": 0.5233278870582581,
+      "learning_rate": 8.472203526719611e-05,
+      "loss": 3.3914,
+      "step": 5723
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.5222857594490051,
+      "learning_rate": 8.468165298155876e-05,
+      "loss": 3.445,
+      "step": 5724
+    },
+    {
+      "epoch": 7.32128,
+      "grad_norm": 0.5074923634529114,
+      "learning_rate": 8.464127069592139e-05,
+      "loss": 3.3527,
+      "step": 5725
+    },
+    {
+      "epoch": 7.32256,
+      "grad_norm": 0.5240098834037781,
+      "learning_rate": 8.4600888410284e-05,
+      "loss": 3.33,
+      "step": 5726
+    },
+    {
+      "epoch": 7.32384,
+      "grad_norm": 0.513421893119812,
+      "learning_rate": 8.456050612464665e-05,
+      "loss": 3.3835,
+      "step": 5727
+    },
+    {
+      "epoch": 7.32512,
+      "grad_norm": 0.524813175201416,
+      "learning_rate": 8.452012383900928e-05,
+      "loss": 3.3284,
+      "step": 5728
+    },
+    {
+      "epoch": 7.3264,
+      "grad_norm": 0.5159763693809509,
+      "learning_rate": 8.447974155337192e-05,
+      "loss": 3.3274,
+      "step": 5729
+    },
+    {
+      "epoch": 7.32768,
+      "grad_norm": 0.5181670784950256,
+      "learning_rate": 8.443935926773454e-05,
+      "loss": 3.4098,
+      "step": 5730
+    },
+    {
+      "epoch": 7.32896,
+      "grad_norm": 0.5126925706863403,
+      "learning_rate": 8.439897698209718e-05,
+      "loss": 3.2956,
+      "step": 5731
+    },
+    {
+      "epoch": 7.33024,
+      "grad_norm": 0.531437873840332,
+      "learning_rate": 8.435859469645981e-05,
+      "loss": 3.395,
+      "step": 5732
+    },
+    {
+      "epoch": 7.33152,
+      "grad_norm": 0.5225799083709717,
+      "learning_rate": 8.431821241082245e-05,
+      "loss": 3.3421,
+      "step": 5733
+    },
+    {
+      "epoch": 7.3328,
+      "grad_norm": 0.5173592567443848,
+      "learning_rate": 8.427783012518507e-05,
+      "loss": 3.3345,
+      "step": 5734
+    },
+    {
+      "epoch": 7.33408,
+      "grad_norm": 0.525583028793335,
+      "learning_rate": 8.423744783954771e-05,
+      "loss": 3.4155,
+      "step": 5735
+    },
+    {
+      "epoch": 7.33536,
+      "grad_norm": 0.5271779894828796,
+      "learning_rate": 8.419706555391034e-05,
+      "loss": 3.4121,
+      "step": 5736
+    },
+    {
+      "epoch": 7.33664,
+      "grad_norm": 0.5142373442649841,
+      "learning_rate": 8.415668326827299e-05,
+      "loss": 3.38,
+      "step": 5737
+    },
+    {
+      "epoch": 7.33792,
+      "grad_norm": 0.5095165967941284,
+      "learning_rate": 8.41163009826356e-05,
+      "loss": 3.3846,
+      "step": 5738
+    },
+    {
+      "epoch": 7.3392,
+      "grad_norm": 0.5181048512458801,
+      "learning_rate": 8.407591869699823e-05,
+      "loss": 3.4532,
+      "step": 5739
+    },
+    {
+      "epoch": 7.34048,
+      "grad_norm": 0.5084232091903687,
+      "learning_rate": 8.403553641136088e-05,
+      "loss": 3.3129,
+      "step": 5740
+    },
+    {
+      "epoch": 7.34176,
+      "grad_norm": 0.5222226977348328,
+      "learning_rate": 8.399515412572351e-05,
+      "loss": 3.3317,
+      "step": 5741
+    },
+    {
+      "epoch": 7.34304,
+      "grad_norm": 0.5268555283546448,
+      "learning_rate": 8.395477184008614e-05,
+      "loss": 3.3111,
+      "step": 5742
+    },
+    {
+      "epoch": 7.34432,
+      "grad_norm": 0.5401803851127625,
+      "learning_rate": 8.391438955444877e-05,
+      "loss": 3.3389,
+      "step": 5743
+    },
+    {
+      "epoch": 7.3456,
+      "grad_norm": 0.49834394454956055,
+      "learning_rate": 8.387400726881141e-05,
+      "loss": 3.3859,
+      "step": 5744
+    },
+    {
+      "epoch": 7.34688,
+      "grad_norm": 0.5210273265838623,
+      "learning_rate": 8.383362498317404e-05,
+      "loss": 3.371,
+      "step": 5745
+    },
+    {
+      "epoch": 7.34816,
+      "grad_norm": 0.5339854955673218,
+      "learning_rate": 8.379324269753667e-05,
+      "loss": 3.3539,
+      "step": 5746
+    },
+    {
+      "epoch": 7.3494399999999995,
+      "grad_norm": 0.5159910917282104,
+      "learning_rate": 8.37528604118993e-05,
+      "loss": 3.3968,
+      "step": 5747
+    },
+    {
+      "epoch": 7.35072,
+      "grad_norm": 0.514874279499054,
+      "learning_rate": 8.371247812626194e-05,
+      "loss": 3.4268,
+      "step": 5748
+    },
+    {
+      "epoch": 7.352,
+      "grad_norm": 0.5367988348007202,
+      "learning_rate": 8.367209584062457e-05,
+      "loss": 3.4254,
+      "step": 5749
+    },
+    {
+      "epoch": 7.35328,
+      "grad_norm": 0.5135504603385925,
+      "learning_rate": 8.363171355498722e-05,
+      "loss": 3.4124,
+      "step": 5750
+    },
+    {
+      "epoch": 7.35456,
+      "grad_norm": 0.5267705321311951,
+      "learning_rate": 8.359133126934984e-05,
+      "loss": 3.4418,
+      "step": 5751
+    },
+    {
+      "epoch": 7.35584,
+      "grad_norm": 0.5178395509719849,
+      "learning_rate": 8.355094898371248e-05,
+      "loss": 3.3394,
+      "step": 5752
+    },
+    {
+      "epoch": 7.35712,
+      "grad_norm": 0.5211403369903564,
+      "learning_rate": 8.351056669807511e-05,
+      "loss": 3.3646,
+      "step": 5753
+    },
+    {
+      "epoch": 7.3584,
+      "grad_norm": 0.5357264876365662,
+      "learning_rate": 8.347018441243773e-05,
+      "loss": 3.4218,
+      "step": 5754
+    },
+    {
+      "epoch": 7.35968,
+      "grad_norm": 0.5282320976257324,
+      "learning_rate": 8.342980212680037e-05,
+      "loss": 3.3026,
+      "step": 5755
+    },
+    {
+      "epoch": 7.36096,
+      "grad_norm": 0.5219427347183228,
+      "learning_rate": 8.3389419841163e-05,
+      "loss": 3.3327,
+      "step": 5756
+    },
+    {
+      "epoch": 7.36224,
+      "grad_norm": 0.5319931507110596,
+      "learning_rate": 8.334903755552564e-05,
+      "loss": 3.4432,
+      "step": 5757
+    },
+    {
+      "epoch": 7.36352,
+      "grad_norm": 0.5045892000198364,
+      "learning_rate": 8.330865526988826e-05,
+      "loss": 3.2485,
+      "step": 5758
+    },
+    {
+      "epoch": 7.3648,
+      "grad_norm": 0.5378910303115845,
+      "learning_rate": 8.32682729842509e-05,
+      "loss": 3.3374,
+      "step": 5759
+    },
+    {
+      "epoch": 7.36608,
+      "grad_norm": 0.5356464982032776,
+      "learning_rate": 8.322789069861353e-05,
+      "loss": 3.3517,
+      "step": 5760
+    },
+    {
+      "epoch": 7.36736,
+      "grad_norm": 0.5344239473342896,
+      "learning_rate": 8.318750841297618e-05,
+      "loss": 3.4034,
+      "step": 5761
+    },
+    {
+      "epoch": 7.36864,
+      "grad_norm": 0.530697226524353,
+      "learning_rate": 8.314712612733879e-05,
+      "loss": 3.3704,
+      "step": 5762
+    },
+    {
+      "epoch": 7.3699200000000005,
+      "grad_norm": 0.5439151525497437,
+      "learning_rate": 8.310674384170144e-05,
+      "loss": 3.3844,
+      "step": 5763
+    },
+    {
+      "epoch": 7.3712,
+      "grad_norm": 0.5485351085662842,
+      "learning_rate": 8.306636155606407e-05,
+      "loss": 3.3354,
+      "step": 5764
+    },
+    {
+      "epoch": 7.37248,
+      "grad_norm": 0.5290150046348572,
+      "learning_rate": 8.302597927042671e-05,
+      "loss": 3.3475,
+      "step": 5765
+    },
+    {
+      "epoch": 7.37376,
+      "grad_norm": 0.5363977551460266,
+      "learning_rate": 8.298559698478933e-05,
+      "loss": 3.3585,
+      "step": 5766
+    },
+    {
+      "epoch": 7.37504,
+      "grad_norm": 0.5304407477378845,
+      "learning_rate": 8.294521469915196e-05,
+      "loss": 3.3573,
+      "step": 5767
+    },
+    {
+      "epoch": 7.37632,
+      "grad_norm": 0.5171595215797424,
+      "learning_rate": 8.29048324135146e-05,
+      "loss": 3.3317,
+      "step": 5768
+    },
+    {
+      "epoch": 7.3776,
+      "grad_norm": 0.5493373870849609,
+      "learning_rate": 8.286445012787723e-05,
+      "loss": 3.4143,
+      "step": 5769
+    },
+    {
+      "epoch": 7.37888,
+      "grad_norm": 0.5111392736434937,
+      "learning_rate": 8.282406784223986e-05,
+      "loss": 3.3583,
+      "step": 5770
+    },
+    {
+      "epoch": 7.38016,
+      "grad_norm": 0.5278791189193726,
+      "learning_rate": 8.278368555660249e-05,
+      "loss": 3.3444,
+      "step": 5771
+    },
+    {
+      "epoch": 7.38144,
+      "grad_norm": 0.5462014079093933,
+      "learning_rate": 8.274330327096513e-05,
+      "loss": 3.3613,
+      "step": 5772
+    },
+    {
+      "epoch": 7.38272,
+      "grad_norm": 0.5223116874694824,
+      "learning_rate": 8.270292098532776e-05,
+      "loss": 3.3495,
+      "step": 5773
+    },
+    {
+      "epoch": 7.384,
+      "grad_norm": 0.5300905108451843,
+      "learning_rate": 8.26625386996904e-05,
+      "loss": 3.351,
+      "step": 5774
+    },
+    {
+      "epoch": 7.38528,
+      "grad_norm": 0.5270613431930542,
+      "learning_rate": 8.262215641405302e-05,
+      "loss": 3.3506,
+      "step": 5775
+    },
+    {
+      "epoch": 7.38656,
+      "grad_norm": 0.5194157958030701,
+      "learning_rate": 8.258177412841567e-05,
+      "loss": 3.3194,
+      "step": 5776
+    },
+    {
+      "epoch": 7.38784,
+      "grad_norm": 0.5169289708137512,
+      "learning_rate": 8.25413918427783e-05,
+      "loss": 3.3789,
+      "step": 5777
+    },
+    {
+      "epoch": 7.38912,
+      "grad_norm": 0.5390181541442871,
+      "learning_rate": 8.250100955714093e-05,
+      "loss": 3.3735,
+      "step": 5778
+    },
+    {
+      "epoch": 7.3904,
+      "grad_norm": 0.5157946348190308,
+      "learning_rate": 8.246062727150356e-05,
+      "loss": 3.3887,
+      "step": 5779
+    },
+    {
+      "epoch": 7.39168,
+      "grad_norm": 0.5124242901802063,
+      "learning_rate": 8.242024498586619e-05,
+      "loss": 3.3066,
+      "step": 5780
+    },
+    {
+      "epoch": 7.39296,
+      "grad_norm": 0.523602306842804,
+      "learning_rate": 8.237986270022883e-05,
+      "loss": 3.4837,
+      "step": 5781
+    },
+    {
+      "epoch": 7.39424,
+      "grad_norm": 0.525345504283905,
+      "learning_rate": 8.233948041459145e-05,
+      "loss": 3.343,
+      "step": 5782
+    },
+    {
+      "epoch": 7.39552,
+      "grad_norm": 0.5302904844284058,
+      "learning_rate": 8.229909812895409e-05,
+      "loss": 3.358,
+      "step": 5783
+    },
+    {
+      "epoch": 7.3968,
+      "grad_norm": 0.5098133087158203,
+      "learning_rate": 8.225871584331672e-05,
+      "loss": 3.3791,
+      "step": 5784
+    },
+    {
+      "epoch": 7.39808,
+      "grad_norm": 0.5192782282829285,
+      "learning_rate": 8.221833355767937e-05,
+      "loss": 3.4295,
+      "step": 5785
+    },
+    {
+      "epoch": 7.39936,
+      "grad_norm": 0.5070282816886902,
+      "learning_rate": 8.217795127204198e-05,
+      "loss": 3.3926,
+      "step": 5786
+    },
+    {
+      "epoch": 7.40064,
+      "grad_norm": 0.5000091791152954,
+      "learning_rate": 8.213756898640463e-05,
+      "loss": 3.3154,
+      "step": 5787
+    },
+    {
+      "epoch": 7.40192,
+      "grad_norm": 0.5293865203857422,
+      "learning_rate": 8.209718670076726e-05,
+      "loss": 3.3813,
+      "step": 5788
+    },
+    {
+      "epoch": 7.4032,
+      "grad_norm": 0.5276476144790649,
+      "learning_rate": 8.20568044151299e-05,
+      "loss": 3.3801,
+      "step": 5789
+    },
+    {
+      "epoch": 7.40448,
+      "grad_norm": 0.5471776127815247,
+      "learning_rate": 8.201642212949252e-05,
+      "loss": 3.4044,
+      "step": 5790
+    },
+    {
+      "epoch": 7.40576,
+      "grad_norm": 0.5328643918037415,
+      "learning_rate": 8.197603984385516e-05,
+      "loss": 3.3514,
+      "step": 5791
+    },
+    {
+      "epoch": 7.40704,
+      "grad_norm": 0.5258887410163879,
+      "learning_rate": 8.193565755821779e-05,
+      "loss": 3.3467,
+      "step": 5792
+    },
+    {
+      "epoch": 7.40832,
+      "grad_norm": 0.5234854221343994,
+      "learning_rate": 8.18952752725804e-05,
+      "loss": 3.4287,
+      "step": 5793
+    },
+    {
+      "epoch": 7.4096,
+      "grad_norm": 0.516001284122467,
+      "learning_rate": 8.185489298694305e-05,
+      "loss": 3.3589,
+      "step": 5794
+    },
+    {
+      "epoch": 7.41088,
+      "grad_norm": 0.5244291424751282,
+      "learning_rate": 8.181451070130568e-05,
+      "loss": 3.4387,
+      "step": 5795
+    },
+    {
+      "epoch": 7.41216,
+      "grad_norm": 0.5443776249885559,
+      "learning_rate": 8.177412841566832e-05,
+      "loss": 3.3916,
+      "step": 5796
+    },
+    {
+      "epoch": 7.41344,
+      "grad_norm": 0.5309216380119324,
+      "learning_rate": 8.173374613003094e-05,
+      "loss": 3.3954,
+      "step": 5797
+    },
+    {
+      "epoch": 7.41472,
+      "grad_norm": 0.5038948059082031,
+      "learning_rate": 8.169336384439358e-05,
+      "loss": 3.343,
+      "step": 5798
+    },
+    {
+      "epoch": 7.416,
+      "grad_norm": 0.52728271484375,
+      "learning_rate": 8.165298155875621e-05,
+      "loss": 3.3784,
+      "step": 5799
+    },
+    {
+      "epoch": 7.41728,
+      "grad_norm": 0.5411390066146851,
+      "learning_rate": 8.161259927311886e-05,
+      "loss": 3.4319,
+      "step": 5800
+    },
+    {
+      "epoch": 7.41856,
+      "grad_norm": 0.5185636878013611,
+      "learning_rate": 8.157221698748149e-05,
+      "loss": 3.4297,
+      "step": 5801
+    },
+    {
+      "epoch": 7.41984,
+      "grad_norm": 0.5342210531234741,
+      "learning_rate": 8.153183470184412e-05,
+      "loss": 3.3683,
+      "step": 5802
+    },
+    {
+      "epoch": 7.42112,
+      "grad_norm": 0.5324079394340515,
+      "learning_rate": 8.149145241620675e-05,
+      "loss": 3.3805,
+      "step": 5803
+    },
+    {
+      "epoch": 7.4224,
+      "grad_norm": 0.520026445388794,
+      "learning_rate": 8.145107013056939e-05,
+      "loss": 3.3222,
+      "step": 5804
+    },
+    {
+      "epoch": 7.42368,
+      "grad_norm": 0.5147387981414795,
+      "learning_rate": 8.141068784493202e-05,
+      "loss": 3.3722,
+      "step": 5805
+    },
+    {
+      "epoch": 7.4249600000000004,
+      "grad_norm": 0.5206732749938965,
+      "learning_rate": 8.137030555929464e-05,
+      "loss": 3.4011,
+      "step": 5806
+    },
+    {
+      "epoch": 7.42624,
+      "grad_norm": 0.5213762521743774,
+      "learning_rate": 8.132992327365728e-05,
+      "loss": 3.3676,
+      "step": 5807
+    },
+    {
+      "epoch": 7.42752,
+      "grad_norm": 0.5201925039291382,
+      "learning_rate": 8.128954098801991e-05,
+      "loss": 3.3553,
+      "step": 5808
+    },
+    {
+      "epoch": 7.4288,
+      "grad_norm": 0.5245599150657654,
+      "learning_rate": 8.124915870238256e-05,
+      "loss": 3.4287,
+      "step": 5809
+    },
+    {
+      "epoch": 7.43008,
+      "grad_norm": 0.5151320695877075,
+      "learning_rate": 8.120877641674517e-05,
+      "loss": 3.3672,
+      "step": 5810
+    },
+    {
+      "epoch": 7.43136,
+      "grad_norm": 0.5449185967445374,
+      "learning_rate": 8.116839413110782e-05,
+      "loss": 3.4269,
+      "step": 5811
+    },
+    {
+      "epoch": 7.43264,
+      "grad_norm": 0.5150814056396484,
+      "learning_rate": 8.112801184547045e-05,
+      "loss": 3.4574,
+      "step": 5812
+    },
+    {
+      "epoch": 7.43392,
+      "grad_norm": 0.5201151371002197,
+      "learning_rate": 8.108762955983309e-05,
+      "loss": 3.3769,
+      "step": 5813
+    },
+    {
+      "epoch": 7.4352,
+      "grad_norm": 0.5087258219718933,
+      "learning_rate": 8.10472472741957e-05,
+      "loss": 3.4079,
+      "step": 5814
+    },
+    {
+      "epoch": 7.4364799999999995,
+      "grad_norm": 0.5286344289779663,
+      "learning_rate": 8.100686498855835e-05,
+      "loss": 3.4387,
+      "step": 5815
+    },
+    {
+      "epoch": 7.43776,
+      "grad_norm": 0.5217651724815369,
+      "learning_rate": 8.096648270292098e-05,
+      "loss": 3.3826,
+      "step": 5816
+    },
+    {
+      "epoch": 7.43904,
+      "grad_norm": 0.5195396542549133,
+      "learning_rate": 8.092610041728362e-05,
+      "loss": 3.3419,
+      "step": 5817
+    },
+    {
+      "epoch": 7.44032,
+      "grad_norm": 0.5105847716331482,
+      "learning_rate": 8.088571813164624e-05,
+      "loss": 3.3915,
+      "step": 5818
+    },
+    {
+      "epoch": 7.4416,
+      "grad_norm": 0.5170384645462036,
+      "learning_rate": 8.084533584600888e-05,
+      "loss": 3.3614,
+      "step": 5819
+    },
+    {
+      "epoch": 7.44288,
+      "grad_norm": 0.5139781832695007,
+      "learning_rate": 8.080495356037151e-05,
+      "loss": 3.427,
+      "step": 5820
+    },
+    {
+      "epoch": 7.44416,
+      "grad_norm": 0.5179570317268372,
+      "learning_rate": 8.076457127473413e-05,
+      "loss": 3.4167,
+      "step": 5821
+    },
+    {
+      "epoch": 7.44544,
+      "grad_norm": 0.5289480090141296,
+      "learning_rate": 8.072418898909677e-05,
+      "loss": 3.3394,
+      "step": 5822
+    },
+    {
+      "epoch": 7.44672,
+      "grad_norm": 0.5202321410179138,
+      "learning_rate": 8.06838067034594e-05,
+      "loss": 3.3681,
+      "step": 5823
+    },
+    {
+      "epoch": 7.448,
+      "grad_norm": 0.5124974846839905,
+      "learning_rate": 8.064342441782205e-05,
+      "loss": 3.3676,
+      "step": 5824
+    },
+    {
+      "epoch": 7.44928,
+      "grad_norm": 0.5194815993309021,
+      "learning_rate": 8.060304213218466e-05,
+      "loss": 3.3824,
+      "step": 5825
+    },
+    {
+      "epoch": 7.45056,
+      "grad_norm": 0.5214881896972656,
+      "learning_rate": 8.056265984654731e-05,
+      "loss": 3.3963,
+      "step": 5826
+    },
+    {
+      "epoch": 7.45184,
+      "grad_norm": 0.5211467742919922,
+      "learning_rate": 8.052227756090994e-05,
+      "loss": 3.3754,
+      "step": 5827
+    },
+    {
+      "epoch": 7.45312,
+      "grad_norm": 0.5255600214004517,
+      "learning_rate": 8.048189527527258e-05,
+      "loss": 3.4049,
+      "step": 5828
+    },
+    {
+      "epoch": 7.4544,
+      "grad_norm": 0.5339398980140686,
+      "learning_rate": 8.04415129896352e-05,
+      "loss": 3.3593,
+      "step": 5829
+    },
+    {
+      "epoch": 7.45568,
+      "grad_norm": 0.5176902413368225,
+      "learning_rate": 8.040113070399784e-05,
+      "loss": 3.3843,
+      "step": 5830
+    },
+    {
+      "epoch": 7.45696,
+      "grad_norm": 0.5297396779060364,
+      "learning_rate": 8.036074841836047e-05,
+      "loss": 3.3186,
+      "step": 5831
+    },
+    {
+      "epoch": 7.45824,
+      "grad_norm": 0.531353235244751,
+      "learning_rate": 8.032036613272312e-05,
+      "loss": 3.3528,
+      "step": 5832
+    },
+    {
+      "epoch": 7.45952,
+      "grad_norm": 0.5194066166877747,
+      "learning_rate": 8.027998384708573e-05,
+      "loss": 3.4387,
+      "step": 5833
+    },
+    {
+      "epoch": 7.4608,
+      "grad_norm": 0.5318957567214966,
+      "learning_rate": 8.023960156144836e-05,
+      "loss": 3.3612,
+      "step": 5834
+    },
+    {
+      "epoch": 7.46208,
+      "grad_norm": 0.5308012366294861,
+      "learning_rate": 8.0199219275811e-05,
+      "loss": 3.4369,
+      "step": 5835
+    },
+    {
+      "epoch": 7.46336,
+      "grad_norm": 0.5104942321777344,
+      "learning_rate": 8.015883699017364e-05,
+      "loss": 3.4146,
+      "step": 5836
+    },
+    {
+      "epoch": 7.46464,
+      "grad_norm": 0.5228284597396851,
+      "learning_rate": 8.011845470453628e-05,
+      "loss": 3.3655,
+      "step": 5837
+    },
+    {
+      "epoch": 7.46592,
+      "grad_norm": 0.5318959951400757,
+      "learning_rate": 8.00780724188989e-05,
+      "loss": 3.3857,
+      "step": 5838
+    },
+    {
+      "epoch": 7.4672,
+      "grad_norm": 0.5241490006446838,
+      "learning_rate": 8.003769013326154e-05,
+      "loss": 3.4119,
+      "step": 5839
+    },
+    {
+      "epoch": 7.46848,
+      "grad_norm": 0.5517506003379822,
+      "learning_rate": 7.999730784762417e-05,
+      "loss": 3.395,
+      "step": 5840
+    },
+    {
+      "epoch": 7.46976,
+      "grad_norm": 0.529964804649353,
+      "learning_rate": 7.995692556198681e-05,
+      "loss": 3.3613,
+      "step": 5841
+    },
+    {
+      "epoch": 7.47104,
+      "grad_norm": 0.5344815850257874,
+      "learning_rate": 7.991654327634943e-05,
+      "loss": 3.3307,
+      "step": 5842
+    },
+    {
+      "epoch": 7.47232,
+      "grad_norm": 0.5124824643135071,
+      "learning_rate": 7.987616099071207e-05,
+      "loss": 3.3268,
+      "step": 5843
+    },
+    {
+      "epoch": 7.4736,
+      "grad_norm": 0.5115407705307007,
+      "learning_rate": 7.98357787050747e-05,
+      "loss": 3.3561,
+      "step": 5844
+    },
+    {
+      "epoch": 7.47488,
+      "grad_norm": 0.5252261757850647,
+      "learning_rate": 7.979539641943735e-05,
+      "loss": 3.3923,
+      "step": 5845
+    },
+    {
+      "epoch": 7.47616,
+      "grad_norm": 0.521026611328125,
+      "learning_rate": 7.975501413379996e-05,
+      "loss": 3.3531,
+      "step": 5846
+    },
+    {
+      "epoch": 7.47744,
+      "grad_norm": 0.5293687582015991,
+      "learning_rate": 7.97146318481626e-05,
+      "loss": 3.4022,
+      "step": 5847
+    },
+    {
+      "epoch": 7.47872,
+      "grad_norm": 0.5290361642837524,
+      "learning_rate": 7.967424956252524e-05,
+      "loss": 3.3675,
+      "step": 5848
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.5217385292053223,
+      "learning_rate": 7.963386727688785e-05,
+      "loss": 3.3527,
+      "step": 5849
+    },
+    {
+      "epoch": 7.48128,
+      "grad_norm": 0.5071624517440796,
+      "learning_rate": 7.95934849912505e-05,
+      "loss": 3.3443,
+      "step": 5850
+    },
+    {
+      "epoch": 7.48256,
+      "grad_norm": 0.5297966599464417,
+      "learning_rate": 7.955310270561313e-05,
+      "loss": 3.3398,
+      "step": 5851
+    },
+    {
+      "epoch": 7.48384,
+      "grad_norm": 0.5170907378196716,
+      "learning_rate": 7.951272041997577e-05,
+      "loss": 3.3573,
+      "step": 5852
+    },
+    {
+      "epoch": 7.48512,
+      "grad_norm": 0.5123015642166138,
+      "learning_rate": 7.947233813433839e-05,
+      "loss": 3.3852,
+      "step": 5853
+    },
+    {
+      "epoch": 7.4864,
+      "grad_norm": 0.5429021120071411,
+      "learning_rate": 7.943195584870103e-05,
+      "loss": 3.435,
+      "step": 5854
+    },
+    {
+      "epoch": 7.48768,
+      "grad_norm": 0.5190651416778564,
+      "learning_rate": 7.939157356306366e-05,
+      "loss": 3.4174,
+      "step": 5855
+    },
+    {
+      "epoch": 7.48896,
+      "grad_norm": 0.5276857614517212,
+      "learning_rate": 7.93511912774263e-05,
+      "loss": 3.368,
+      "step": 5856
+    },
+    {
+      "epoch": 7.49024,
+      "grad_norm": 0.5189977288246155,
+      "learning_rate": 7.931080899178892e-05,
+      "loss": 3.3205,
+      "step": 5857
+    },
+    {
+      "epoch": 7.49152,
+      "grad_norm": 0.5297253131866455,
+      "learning_rate": 7.927042670615156e-05,
+      "loss": 3.3809,
+      "step": 5858
+    },
+    {
+      "epoch": 7.4928,
+      "grad_norm": 0.513972818851471,
+      "learning_rate": 7.92300444205142e-05,
+      "loss": 3.3473,
+      "step": 5859
+    },
+    {
+      "epoch": 7.49408,
+      "grad_norm": 0.539738118648529,
+      "learning_rate": 7.918966213487682e-05,
+      "loss": 3.3157,
+      "step": 5860
+    },
+    {
+      "epoch": 7.49536,
+      "grad_norm": 0.5355290174484253,
+      "learning_rate": 7.914927984923946e-05,
+      "loss": 3.4172,
+      "step": 5861
+    },
+    {
+      "epoch": 7.49664,
+      "grad_norm": 0.5001344084739685,
+      "learning_rate": 7.910889756360209e-05,
+      "loss": 3.2782,
+      "step": 5862
+    },
+    {
+      "epoch": 7.49792,
+      "grad_norm": 0.5256953835487366,
+      "learning_rate": 7.906851527796473e-05,
+      "loss": 3.3973,
+      "step": 5863
+    },
+    {
+      "epoch": 7.4992,
+      "grad_norm": 0.526117205619812,
+      "learning_rate": 7.902813299232736e-05,
+      "loss": 3.3484,
+      "step": 5864
+    },
+    {
+      "epoch": 7.50048,
+      "grad_norm": 0.5159139037132263,
+      "learning_rate": 7.898775070668999e-05,
+      "loss": 3.2791,
+      "step": 5865
+    },
+    {
+      "epoch": 7.50176,
+      "grad_norm": 0.5434691309928894,
+      "learning_rate": 7.894736842105262e-05,
+      "loss": 3.3503,
+      "step": 5866
+    },
+    {
+      "epoch": 7.50304,
+      "grad_norm": 0.5286847949028015,
+      "learning_rate": 7.890698613541526e-05,
+      "loss": 3.3695,
+      "step": 5867
+    },
+    {
+      "epoch": 7.50432,
+      "grad_norm": 0.5285123586654663,
+      "learning_rate": 7.886660384977789e-05,
+      "loss": 3.4208,
+      "step": 5868
+    },
+    {
+      "epoch": 7.5056,
+      "grad_norm": 0.5370778441429138,
+      "learning_rate": 7.882622156414052e-05,
+      "loss": 3.4278,
+      "step": 5869
+    },
+    {
+      "epoch": 7.50688,
+      "grad_norm": 0.5134227275848389,
+      "learning_rate": 7.878583927850315e-05,
+      "loss": 3.3727,
+      "step": 5870
+    },
+    {
+      "epoch": 7.50816,
+      "grad_norm": 0.5174378156661987,
+      "learning_rate": 7.87454569928658e-05,
+      "loss": 3.3265,
+      "step": 5871
+    },
+    {
+      "epoch": 7.50944,
+      "grad_norm": 0.5034676194190979,
+      "learning_rate": 7.870507470722843e-05,
+      "loss": 3.3688,
+      "step": 5872
+    },
+    {
+      "epoch": 7.51072,
+      "grad_norm": 0.5375760197639465,
+      "learning_rate": 7.866469242159107e-05,
+      "loss": 3.3616,
+      "step": 5873
+    },
+    {
+      "epoch": 7.5120000000000005,
+      "grad_norm": 0.5142561197280884,
+      "learning_rate": 7.862431013595369e-05,
+      "loss": 3.3702,
+      "step": 5874
+    },
+    {
+      "epoch": 7.51328,
+      "grad_norm": 0.5261555314064026,
+      "learning_rate": 7.858392785031632e-05,
+      "loss": 3.372,
+      "step": 5875
+    },
+    {
+      "epoch": 7.51456,
+      "grad_norm": 0.5355465412139893,
+      "learning_rate": 7.854354556467896e-05,
+      "loss": 3.3866,
+      "step": 5876
+    },
+    {
+      "epoch": 7.51584,
+      "grad_norm": 0.5262174010276794,
+      "learning_rate": 7.850316327904158e-05,
+      "loss": 3.4209,
+      "step": 5877
+    },
+    {
+      "epoch": 7.51712,
+      "grad_norm": 0.5196813344955444,
+      "learning_rate": 7.846278099340422e-05,
+      "loss": 3.4005,
+      "step": 5878
+    },
+    {
+      "epoch": 7.5184,
+      "grad_norm": 0.5177279114723206,
+      "learning_rate": 7.842239870776685e-05,
+      "loss": 3.364,
+      "step": 5879
+    },
+    {
+      "epoch": 7.51968,
+      "grad_norm": 0.5248557329177856,
+      "learning_rate": 7.83820164221295e-05,
+      "loss": 3.3458,
+      "step": 5880
+    },
+    {
+      "epoch": 7.52096,
+      "grad_norm": 0.5306951999664307,
+      "learning_rate": 7.834163413649211e-05,
+      "loss": 3.3503,
+      "step": 5881
+    },
+    {
+      "epoch": 7.52224,
+      "grad_norm": 0.5404332876205444,
+      "learning_rate": 7.830125185085475e-05,
+      "loss": 3.2522,
+      "step": 5882
+    },
+    {
+      "epoch": 7.5235199999999995,
+      "grad_norm": 0.5501363277435303,
+      "learning_rate": 7.826086956521738e-05,
+      "loss": 3.3744,
+      "step": 5883
+    },
+    {
+      "epoch": 7.5248,
+      "grad_norm": 0.5222617387771606,
+      "learning_rate": 7.822048727958003e-05,
+      "loss": 3.3439,
+      "step": 5884
+    },
+    {
+      "epoch": 7.52608,
+      "grad_norm": 0.5240084528923035,
+      "learning_rate": 7.818010499394264e-05,
+      "loss": 3.3858,
+      "step": 5885
+    },
+    {
+      "epoch": 7.52736,
+      "grad_norm": 0.5329712629318237,
+      "learning_rate": 7.813972270830529e-05,
+      "loss": 3.3657,
+      "step": 5886
+    },
+    {
+      "epoch": 7.52864,
+      "grad_norm": 0.5383173823356628,
+      "learning_rate": 7.809934042266792e-05,
+      "loss": 3.4141,
+      "step": 5887
+    },
+    {
+      "epoch": 7.52992,
+      "grad_norm": 0.5195038318634033,
+      "learning_rate": 7.805895813703055e-05,
+      "loss": 3.43,
+      "step": 5888
+    },
+    {
+      "epoch": 7.5312,
+      "grad_norm": 0.5309056043624878,
+      "learning_rate": 7.801857585139318e-05,
+      "loss": 3.3459,
+      "step": 5889
+    },
+    {
+      "epoch": 7.53248,
+      "grad_norm": 0.5201624631881714,
+      "learning_rate": 7.797819356575581e-05,
+      "loss": 3.3979,
+      "step": 5890
+    },
+    {
+      "epoch": 7.53376,
+      "grad_norm": 0.520959198474884,
+      "learning_rate": 7.793781128011845e-05,
+      "loss": 3.3557,
+      "step": 5891
+    },
+    {
+      "epoch": 7.53504,
+      "grad_norm": 0.5142325758934021,
+      "learning_rate": 7.789742899448108e-05,
+      "loss": 3.4258,
+      "step": 5892
+    },
+    {
+      "epoch": 7.53632,
+      "grad_norm": 0.5339785814285278,
+      "learning_rate": 7.785704670884371e-05,
+      "loss": 3.3424,
+      "step": 5893
+    },
+    {
+      "epoch": 7.5376,
+      "grad_norm": 0.5324097275733948,
+      "learning_rate": 7.781666442320634e-05,
+      "loss": 3.3253,
+      "step": 5894
+    },
+    {
+      "epoch": 7.53888,
+      "grad_norm": 0.5203735828399658,
+      "learning_rate": 7.777628213756899e-05,
+      "loss": 3.422,
+      "step": 5895
+    },
+    {
+      "epoch": 7.54016,
+      "grad_norm": 0.5267855525016785,
+      "learning_rate": 7.773589985193162e-05,
+      "loss": 3.3743,
+      "step": 5896
+    },
+    {
+      "epoch": 7.54144,
+      "grad_norm": 0.51844722032547,
+      "learning_rate": 7.769551756629425e-05,
+      "loss": 3.2686,
+      "step": 5897
+    },
+    {
+      "epoch": 7.54272,
+      "grad_norm": 0.5209645628929138,
+      "learning_rate": 7.765513528065688e-05,
+      "loss": 3.3467,
+      "step": 5898
+    },
+    {
+      "epoch": 7.5440000000000005,
+      "grad_norm": 0.530583918094635,
+      "learning_rate": 7.761475299501952e-05,
+      "loss": 3.437,
+      "step": 5899
+    },
+    {
+      "epoch": 7.54528,
+      "grad_norm": 0.5081465840339661,
+      "learning_rate": 7.757437070938215e-05,
+      "loss": 3.3694,
+      "step": 5900
+    },
+    {
+      "epoch": 7.54656,
+      "grad_norm": 0.5179983973503113,
+      "learning_rate": 7.753398842374477e-05,
+      "loss": 3.3602,
+      "step": 5901
+    },
+    {
+      "epoch": 7.54784,
+      "grad_norm": 0.5412976145744324,
+      "learning_rate": 7.749360613810741e-05,
+      "loss": 3.3402,
+      "step": 5902
+    },
+    {
+      "epoch": 7.54912,
+      "grad_norm": 0.5343058705329895,
+      "learning_rate": 7.745322385247004e-05,
+      "loss": 3.4374,
+      "step": 5903
+    },
+    {
+      "epoch": 7.5504,
+      "grad_norm": 0.5409019589424133,
+      "learning_rate": 7.741284156683268e-05,
+      "loss": 3.323,
+      "step": 5904
+    },
+    {
+      "epoch": 7.55168,
+      "grad_norm": 0.5322821736335754,
+      "learning_rate": 7.73724592811953e-05,
+      "loss": 3.4257,
+      "step": 5905
+    },
+    {
+      "epoch": 7.55296,
+      "grad_norm": 0.5157960653305054,
+      "learning_rate": 7.733207699555794e-05,
+      "loss": 3.3649,
+      "step": 5906
+    },
+    {
+      "epoch": 7.55424,
+      "grad_norm": 0.5299696922302246,
+      "learning_rate": 7.729169470992057e-05,
+      "loss": 3.3594,
+      "step": 5907
+    },
+    {
+      "epoch": 7.55552,
+      "grad_norm": 0.5254634618759155,
+      "learning_rate": 7.725131242428322e-05,
+      "loss": 3.4228,
+      "step": 5908
+    },
+    {
+      "epoch": 7.5568,
+      "grad_norm": 0.5235505104064941,
+      "learning_rate": 7.721093013864583e-05,
+      "loss": 3.357,
+      "step": 5909
+    },
+    {
+      "epoch": 7.55808,
+      "grad_norm": 0.5225314497947693,
+      "learning_rate": 7.717054785300848e-05,
+      "loss": 3.3528,
+      "step": 5910
+    },
+    {
+      "epoch": 7.55936,
+      "grad_norm": 0.51315838098526,
+      "learning_rate": 7.713016556737111e-05,
+      "loss": 3.4245,
+      "step": 5911
+    },
+    {
+      "epoch": 7.56064,
+      "grad_norm": 0.536595344543457,
+      "learning_rate": 7.708978328173375e-05,
+      "loss": 3.3816,
+      "step": 5912
+    },
+    {
+      "epoch": 7.56192,
+      "grad_norm": 0.5261996984481812,
+      "learning_rate": 7.704940099609637e-05,
+      "loss": 3.3648,
+      "step": 5913
+    },
+    {
+      "epoch": 7.5632,
+      "grad_norm": 0.5061861276626587,
+      "learning_rate": 7.7009018710459e-05,
+      "loss": 3.4095,
+      "step": 5914
+    },
+    {
+      "epoch": 7.56448,
+      "grad_norm": 0.52558434009552,
+      "learning_rate": 7.696863642482164e-05,
+      "loss": 3.4055,
+      "step": 5915
+    },
+    {
+      "epoch": 7.56576,
+      "grad_norm": 0.5246904492378235,
+      "learning_rate": 7.692825413918426e-05,
+      "loss": 3.341,
+      "step": 5916
+    },
+    {
+      "epoch": 7.56704,
+      "grad_norm": 0.5209987759590149,
+      "learning_rate": 7.68878718535469e-05,
+      "loss": 3.3632,
+      "step": 5917
+    },
+    {
+      "epoch": 7.56832,
+      "grad_norm": 0.5232482552528381,
+      "learning_rate": 7.684748956790953e-05,
+      "loss": 3.3235,
+      "step": 5918
+    },
+    {
+      "epoch": 7.5696,
+      "grad_norm": 0.5276979207992554,
+      "learning_rate": 7.680710728227218e-05,
+      "loss": 3.4077,
+      "step": 5919
+    },
+    {
+      "epoch": 7.57088,
+      "grad_norm": 0.5282018184661865,
+      "learning_rate": 7.676672499663479e-05,
+      "loss": 3.2798,
+      "step": 5920
+    },
+    {
+      "epoch": 7.57216,
+      "grad_norm": 0.5135851502418518,
+      "learning_rate": 7.672634271099744e-05,
+      "loss": 3.4252,
+      "step": 5921
+    },
+    {
+      "epoch": 7.57344,
+      "grad_norm": 0.5193025469779968,
+      "learning_rate": 7.668596042536007e-05,
+      "loss": 3.2833,
+      "step": 5922
+    },
+    {
+      "epoch": 7.57472,
+      "grad_norm": 0.5133464336395264,
+      "learning_rate": 7.664557813972271e-05,
+      "loss": 3.3827,
+      "step": 5923
+    },
+    {
+      "epoch": 7.576,
+      "grad_norm": 0.5185580253601074,
+      "learning_rate": 7.660519585408534e-05,
+      "loss": 3.3816,
+      "step": 5924
+    },
+    {
+      "epoch": 7.57728,
+      "grad_norm": 0.5164629220962524,
+      "learning_rate": 7.656481356844797e-05,
+      "loss": 3.4205,
+      "step": 5925
+    },
+    {
+      "epoch": 7.5785599999999995,
+      "grad_norm": 0.5288913249969482,
+      "learning_rate": 7.65244312828106e-05,
+      "loss": 3.3824,
+      "step": 5926
+    },
+    {
+      "epoch": 7.57984,
+      "grad_norm": 0.518733024597168,
+      "learning_rate": 7.648404899717323e-05,
+      "loss": 3.3578,
+      "step": 5927
+    },
+    {
+      "epoch": 7.58112,
+      "grad_norm": 0.5246930718421936,
+      "learning_rate": 7.644366671153587e-05,
+      "loss": 3.3632,
+      "step": 5928
+    },
+    {
+      "epoch": 7.5824,
+      "grad_norm": 0.5315341949462891,
+      "learning_rate": 7.640328442589849e-05,
+      "loss": 3.3564,
+      "step": 5929
+    },
+    {
+      "epoch": 7.58368,
+      "grad_norm": 0.5214976668357849,
+      "learning_rate": 7.636290214026113e-05,
+      "loss": 3.3396,
+      "step": 5930
+    },
+    {
+      "epoch": 7.58496,
+      "grad_norm": 0.5225616097450256,
+      "learning_rate": 7.632251985462376e-05,
+      "loss": 3.4089,
+      "step": 5931
+    },
+    {
+      "epoch": 7.58624,
+      "grad_norm": 0.5472208857536316,
+      "learning_rate": 7.628213756898641e-05,
+      "loss": 3.397,
+      "step": 5932
+    },
+    {
+      "epoch": 7.58752,
+      "grad_norm": 0.5345306396484375,
+      "learning_rate": 7.624175528334902e-05,
+      "loss": 3.395,
+      "step": 5933
+    },
+    {
+      "epoch": 7.5888,
+      "grad_norm": 0.5214905142784119,
+      "learning_rate": 7.620137299771167e-05,
+      "loss": 3.3558,
+      "step": 5934
+    },
+    {
+      "epoch": 7.59008,
+      "grad_norm": 0.53025221824646,
+      "learning_rate": 7.61609907120743e-05,
+      "loss": 3.3575,
+      "step": 5935
+    },
+    {
+      "epoch": 7.59136,
+      "grad_norm": 0.5355333685874939,
+      "learning_rate": 7.612060842643694e-05,
+      "loss": 3.3473,
+      "step": 5936
+    },
+    {
+      "epoch": 7.59264,
+      "grad_norm": 0.5342869758605957,
+      "learning_rate": 7.608022614079956e-05,
+      "loss": 3.3687,
+      "step": 5937
+    },
+    {
+      "epoch": 7.59392,
+      "grad_norm": 0.5135994553565979,
+      "learning_rate": 7.60398438551622e-05,
+      "loss": 3.3285,
+      "step": 5938
+    },
+    {
+      "epoch": 7.5952,
+      "grad_norm": 0.523440957069397,
+      "learning_rate": 7.599946156952483e-05,
+      "loss": 3.3989,
+      "step": 5939
+    },
+    {
+      "epoch": 7.59648,
+      "grad_norm": 0.5180022716522217,
+      "learning_rate": 7.595907928388747e-05,
+      "loss": 3.3353,
+      "step": 5940
+    },
+    {
+      "epoch": 7.59776,
+      "grad_norm": 0.5203397870063782,
+      "learning_rate": 7.591869699825009e-05,
+      "loss": 3.3903,
+      "step": 5941
+    },
+    {
+      "epoch": 7.5990400000000005,
+      "grad_norm": 0.5081101059913635,
+      "learning_rate": 7.587831471261272e-05,
+      "loss": 3.3339,
+      "step": 5942
+    },
+    {
+      "epoch": 7.60032,
+      "grad_norm": 0.5131347179412842,
+      "learning_rate": 7.583793242697537e-05,
+      "loss": 3.4133,
+      "step": 5943
+    },
+    {
+      "epoch": 7.6016,
+      "grad_norm": 0.5185050964355469,
+      "learning_rate": 7.579755014133798e-05,
+      "loss": 3.2982,
+      "step": 5944
+    },
+    {
+      "epoch": 7.60288,
+      "grad_norm": 0.5262473821640015,
+      "learning_rate": 7.575716785570063e-05,
+      "loss": 3.4203,
+      "step": 5945
+    },
+    {
+      "epoch": 7.60416,
+      "grad_norm": 0.5225714445114136,
+      "learning_rate": 7.571678557006326e-05,
+      "loss": 3.4279,
+      "step": 5946
+    },
+    {
+      "epoch": 7.60544,
+      "grad_norm": 0.5234572887420654,
+      "learning_rate": 7.56764032844259e-05,
+      "loss": 3.4388,
+      "step": 5947
+    },
+    {
+      "epoch": 7.60672,
+      "grad_norm": 0.51933753490448,
+      "learning_rate": 7.563602099878852e-05,
+      "loss": 3.4044,
+      "step": 5948
+    },
+    {
+      "epoch": 7.608,
+      "grad_norm": 0.5347757339477539,
+      "learning_rate": 7.559563871315116e-05,
+      "loss": 3.4157,
+      "step": 5949
+    },
+    {
+      "epoch": 7.60928,
+      "grad_norm": 0.5258873105049133,
+      "learning_rate": 7.555525642751379e-05,
+      "loss": 3.4091,
+      "step": 5950
+    },
+    {
+      "epoch": 7.6105599999999995,
+      "grad_norm": 0.5191522836685181,
+      "learning_rate": 7.551487414187643e-05,
+      "loss": 3.2995,
+      "step": 5951
+    },
+    {
+      "epoch": 7.61184,
+      "grad_norm": 0.5296809673309326,
+      "learning_rate": 7.547449185623905e-05,
+      "loss": 3.4695,
+      "step": 5952
+    },
+    {
+      "epoch": 7.61312,
+      "grad_norm": 0.5205982327461243,
+      "learning_rate": 7.543410957060169e-05,
+      "loss": 3.3205,
+      "step": 5953
+    },
+    {
+      "epoch": 7.6144,
+      "grad_norm": 0.5297776460647583,
+      "learning_rate": 7.539372728496432e-05,
+      "loss": 3.4047,
+      "step": 5954
+    },
+    {
+      "epoch": 7.61568,
+      "grad_norm": 0.5211127996444702,
+      "learning_rate": 7.535334499932695e-05,
+      "loss": 3.2961,
+      "step": 5955
+    },
+    {
+      "epoch": 7.61696,
+      "grad_norm": 0.5078245401382446,
+      "learning_rate": 7.531296271368958e-05,
+      "loss": 3.3458,
+      "step": 5956
+    },
+    {
+      "epoch": 7.61824,
+      "grad_norm": 0.5190492272377014,
+      "learning_rate": 7.527258042805221e-05,
+      "loss": 3.3822,
+      "step": 5957
+    },
+    {
+      "epoch": 7.61952,
+      "grad_norm": 0.506162703037262,
+      "learning_rate": 7.523219814241486e-05,
+      "loss": 3.3511,
+      "step": 5958
+    },
+    {
+      "epoch": 7.6208,
+      "grad_norm": 0.5184814929962158,
+      "learning_rate": 7.519181585677749e-05,
+      "loss": 3.3438,
+      "step": 5959
+    },
+    {
+      "epoch": 7.62208,
+      "grad_norm": 0.516089916229248,
+      "learning_rate": 7.515143357114012e-05,
+      "loss": 3.4503,
+      "step": 5960
+    },
+    {
+      "epoch": 7.62336,
+      "grad_norm": 0.5319845676422119,
+      "learning_rate": 7.511105128550275e-05,
+      "loss": 3.3838,
+      "step": 5961
+    },
+    {
+      "epoch": 7.62464,
+      "grad_norm": 0.5250706076622009,
+      "learning_rate": 7.507066899986539e-05,
+      "loss": 3.4235,
+      "step": 5962
+    },
+    {
+      "epoch": 7.62592,
+      "grad_norm": 0.512394368648529,
+      "learning_rate": 7.503028671422802e-05,
+      "loss": 3.3163,
+      "step": 5963
+    },
+    {
+      "epoch": 7.6272,
+      "grad_norm": 0.5351422429084778,
+      "learning_rate": 7.498990442859065e-05,
+      "loss": 3.4371,
+      "step": 5964
+    },
+    {
+      "epoch": 7.62848,
+      "grad_norm": 0.5451759696006775,
+      "learning_rate": 7.494952214295328e-05,
+      "loss": 3.3743,
+      "step": 5965
+    },
+    {
+      "epoch": 7.62976,
+      "grad_norm": 0.5345867872238159,
+      "learning_rate": 7.490913985731591e-05,
+      "loss": 3.3628,
+      "step": 5966
+    },
+    {
+      "epoch": 7.6310400000000005,
+      "grad_norm": 0.5288296341896057,
+      "learning_rate": 7.486875757167855e-05,
+      "loss": 3.3488,
+      "step": 5967
+    },
+    {
+      "epoch": 7.63232,
+      "grad_norm": 0.5270797610282898,
+      "learning_rate": 7.482837528604118e-05,
+      "loss": 3.3402,
+      "step": 5968
+    },
+    {
+      "epoch": 7.6336,
+      "grad_norm": 0.5452633500099182,
+      "learning_rate": 7.478799300040381e-05,
+      "loss": 3.3927,
+      "step": 5969
+    },
+    {
+      "epoch": 7.63488,
+      "grad_norm": 0.5138344764709473,
+      "learning_rate": 7.474761071476644e-05,
+      "loss": 3.3503,
+      "step": 5970
+    },
+    {
+      "epoch": 7.63616,
+      "grad_norm": 0.5245453715324402,
+      "learning_rate": 7.470722842912909e-05,
+      "loss": 3.3953,
+      "step": 5971
+    },
+    {
+      "epoch": 7.63744,
+      "grad_norm": 0.5355793237686157,
+      "learning_rate": 7.466684614349172e-05,
+      "loss": 3.3593,
+      "step": 5972
+    },
+    {
+      "epoch": 7.63872,
+      "grad_norm": 0.5104573369026184,
+      "learning_rate": 7.462646385785435e-05,
+      "loss": 3.347,
+      "step": 5973
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.5135825276374817,
+      "learning_rate": 7.458608157221698e-05,
+      "loss": 3.3998,
+      "step": 5974
+    },
+    {
+      "epoch": 7.64128,
+      "grad_norm": 0.5211784243583679,
+      "learning_rate": 7.454569928657961e-05,
+      "loss": 3.3404,
+      "step": 5975
+    },
+    {
+      "epoch": 7.64256,
+      "grad_norm": 0.5213080644607544,
+      "learning_rate": 7.450531700094224e-05,
+      "loss": 3.41,
+      "step": 5976
+    },
+    {
+      "epoch": 7.64384,
+      "grad_norm": 0.518886923789978,
+      "learning_rate": 7.446493471530488e-05,
+      "loss": 3.41,
+      "step": 5977
+    },
+    {
+      "epoch": 7.64512,
+      "grad_norm": 0.5232623815536499,
+      "learning_rate": 7.442455242966751e-05,
+      "loss": 3.324,
+      "step": 5978
+    },
+    {
+      "epoch": 7.6464,
+      "grad_norm": 0.5307015776634216,
+      "learning_rate": 7.438417014403014e-05,
+      "loss": 3.3834,
+      "step": 5979
+    },
+    {
+      "epoch": 7.64768,
+      "grad_norm": 0.5195972323417664,
+      "learning_rate": 7.434378785839277e-05,
+      "loss": 3.4509,
+      "step": 5980
+    },
+    {
+      "epoch": 7.64896,
+      "grad_norm": 0.5420621037483215,
+      "learning_rate": 7.430340557275542e-05,
+      "loss": 3.3845,
+      "step": 5981
+    },
+    {
+      "epoch": 7.65024,
+      "grad_norm": 0.5237659215927124,
+      "learning_rate": 7.426302328711805e-05,
+      "loss": 3.3393,
+      "step": 5982
+    },
+    {
+      "epoch": 7.65152,
+      "grad_norm": 0.5215779542922974,
+      "learning_rate": 7.422264100148068e-05,
+      "loss": 3.3277,
+      "step": 5983
+    },
+    {
+      "epoch": 7.6528,
+      "grad_norm": 0.5218585729598999,
+      "learning_rate": 7.41822587158433e-05,
+      "loss": 3.3724,
+      "step": 5984
+    },
+    {
+      "epoch": 7.65408,
+      "grad_norm": 0.5390816926956177,
+      "learning_rate": 7.414187643020595e-05,
+      "loss": 3.3104,
+      "step": 5985
+    },
+    {
+      "epoch": 7.65536,
+      "grad_norm": 0.5215725302696228,
+      "learning_rate": 7.410149414456858e-05,
+      "loss": 3.3578,
+      "step": 5986
+    },
+    {
+      "epoch": 7.65664,
+      "grad_norm": 0.52614825963974,
+      "learning_rate": 7.406111185893121e-05,
+      "loss": 3.3742,
+      "step": 5987
+    },
+    {
+      "epoch": 7.65792,
+      "grad_norm": 0.5331546664237976,
+      "learning_rate": 7.402072957329384e-05,
+      "loss": 3.4524,
+      "step": 5988
+    },
+    {
+      "epoch": 7.6592,
+      "grad_norm": 0.5238944888114929,
+      "learning_rate": 7.398034728765647e-05,
+      "loss": 3.3707,
+      "step": 5989
+    },
+    {
+      "epoch": 7.66048,
+      "grad_norm": 0.5260823965072632,
+      "learning_rate": 7.39399650020191e-05,
+      "loss": 3.4557,
+      "step": 5990
+    },
+    {
+      "epoch": 7.66176,
+      "grad_norm": 0.5286345481872559,
+      "learning_rate": 7.389958271638174e-05,
+      "loss": 3.3191,
+      "step": 5991
+    },
+    {
+      "epoch": 7.66304,
+      "grad_norm": 0.5183789134025574,
+      "learning_rate": 7.385920043074437e-05,
+      "loss": 3.3766,
+      "step": 5992
+    },
+    {
+      "epoch": 7.66432,
+      "grad_norm": 0.5158443450927734,
+      "learning_rate": 7.3818818145107e-05,
+      "loss": 3.3405,
+      "step": 5993
+    },
+    {
+      "epoch": 7.6655999999999995,
+      "grad_norm": 0.5058552026748657,
+      "learning_rate": 7.377843585946963e-05,
+      "loss": 3.3266,
+      "step": 5994
+    },
+    {
+      "epoch": 7.66688,
+      "grad_norm": 0.5316519737243652,
+      "learning_rate": 7.373805357383228e-05,
+      "loss": 3.4087,
+      "step": 5995
+    },
+    {
+      "epoch": 7.66816,
+      "grad_norm": 0.5409033298492432,
+      "learning_rate": 7.369767128819491e-05,
+      "loss": 3.3138,
+      "step": 5996
+    },
+    {
+      "epoch": 7.66944,
+      "grad_norm": 0.5183295607566833,
+      "learning_rate": 7.365728900255754e-05,
+      "loss": 3.3506,
+      "step": 5997
+    },
+    {
+      "epoch": 7.67072,
+      "grad_norm": 0.5203860998153687,
+      "learning_rate": 7.361690671692017e-05,
+      "loss": 3.3259,
+      "step": 5998
+    },
+    {
+      "epoch": 7.672,
+      "grad_norm": 0.5194439888000488,
+      "learning_rate": 7.357652443128281e-05,
+      "loss": 3.3583,
+      "step": 5999
+    },
+    {
+      "epoch": 7.67328,
+      "grad_norm": 0.5380210280418396,
+      "learning_rate": 7.353614214564544e-05,
+      "loss": 3.3976,
+      "step": 6000
+    },
+    {
+      "epoch": 7.67456,
+      "grad_norm": 0.5133769512176514,
+      "learning_rate": 7.349575986000807e-05,
+      "loss": 3.3776,
+      "step": 6001
+    },
+    {
+      "epoch": 7.67584,
+      "grad_norm": 0.5389159321784973,
+      "learning_rate": 7.34553775743707e-05,
+      "loss": 3.3787,
+      "step": 6002
+    },
+    {
+      "epoch": 7.67712,
+      "grad_norm": 0.5545948147773743,
+      "learning_rate": 7.341499528873333e-05,
+      "loss": 3.3445,
+      "step": 6003
+    },
+    {
+      "epoch": 7.6784,
+      "grad_norm": 0.5201895236968994,
+      "learning_rate": 7.337461300309596e-05,
+      "loss": 3.2763,
+      "step": 6004
+    },
+    {
+      "epoch": 7.67968,
+      "grad_norm": 0.538422167301178,
+      "learning_rate": 7.33342307174586e-05,
+      "loss": 3.3572,
+      "step": 6005
+    },
+    {
+      "epoch": 7.68096,
+      "grad_norm": 0.5424717664718628,
+      "learning_rate": 7.329384843182124e-05,
+      "loss": 3.3556,
+      "step": 6006
+    },
+    {
+      "epoch": 7.68224,
+      "grad_norm": 0.5273032784461975,
+      "learning_rate": 7.325346614618387e-05,
+      "loss": 3.3478,
+      "step": 6007
+    },
+    {
+      "epoch": 7.68352,
+      "grad_norm": 0.5107418298721313,
+      "learning_rate": 7.32130838605465e-05,
+      "loss": 3.4369,
+      "step": 6008
+    },
+    {
+      "epoch": 7.6848,
+      "grad_norm": 0.5244372487068176,
+      "learning_rate": 7.317270157490914e-05,
+      "loss": 3.3214,
+      "step": 6009
+    },
+    {
+      "epoch": 7.6860800000000005,
+      "grad_norm": 0.5354893207550049,
+      "learning_rate": 7.313231928927177e-05,
+      "loss": 3.3396,
+      "step": 6010
+    },
+    {
+      "epoch": 7.68736,
+      "grad_norm": 0.5245274901390076,
+      "learning_rate": 7.30919370036344e-05,
+      "loss": 3.423,
+      "step": 6011
+    },
+    {
+      "epoch": 7.68864,
+      "grad_norm": 0.528121829032898,
+      "learning_rate": 7.305155471799703e-05,
+      "loss": 3.4042,
+      "step": 6012
+    },
+    {
+      "epoch": 7.68992,
+      "grad_norm": 0.5323206782341003,
+      "learning_rate": 7.301117243235967e-05,
+      "loss": 3.4324,
+      "step": 6013
+    },
+    {
+      "epoch": 7.6912,
+      "grad_norm": 0.5250928401947021,
+      "learning_rate": 7.29707901467223e-05,
+      "loss": 3.3946,
+      "step": 6014
+    },
+    {
+      "epoch": 7.69248,
+      "grad_norm": 0.5359845757484436,
+      "learning_rate": 7.293040786108493e-05,
+      "loss": 3.3705,
+      "step": 6015
+    },
+    {
+      "epoch": 7.69376,
+      "grad_norm": 0.5194463729858398,
+      "learning_rate": 7.289002557544756e-05,
+      "loss": 3.3554,
+      "step": 6016
+    },
+    {
+      "epoch": 7.69504,
+      "grad_norm": 0.5283463597297668,
+      "learning_rate": 7.28496432898102e-05,
+      "loss": 3.3502,
+      "step": 6017
+    },
+    {
+      "epoch": 7.69632,
+      "grad_norm": 0.5234790444374084,
+      "learning_rate": 7.280926100417282e-05,
+      "loss": 3.3667,
+      "step": 6018
+    },
+    {
+      "epoch": 7.6975999999999996,
+      "grad_norm": 0.5353789329528809,
+      "learning_rate": 7.276887871853547e-05,
+      "loss": 3.4113,
+      "step": 6019
+    },
+    {
+      "epoch": 7.69888,
+      "grad_norm": 0.5060636401176453,
+      "learning_rate": 7.27284964328981e-05,
+      "loss": 3.4047,
+      "step": 6020
+    },
+    {
+      "epoch": 7.70016,
+      "grad_norm": 0.5186939239501953,
+      "learning_rate": 7.268811414726073e-05,
+      "loss": 3.3966,
+      "step": 6021
+    },
+    {
+      "epoch": 7.70144,
+      "grad_norm": 0.5250869989395142,
+      "learning_rate": 7.264773186162336e-05,
+      "loss": 3.4134,
+      "step": 6022
+    },
+    {
+      "epoch": 7.70272,
+      "grad_norm": 0.5262473821640015,
+      "learning_rate": 7.2607349575986e-05,
+      "loss": 3.3685,
+      "step": 6023
+    },
+    {
+      "epoch": 7.704,
+      "grad_norm": 0.5311341881752014,
+      "learning_rate": 7.256696729034863e-05,
+      "loss": 3.3809,
+      "step": 6024
+    },
+    {
+      "epoch": 7.70528,
+      "grad_norm": 0.522276759147644,
+      "learning_rate": 7.252658500471126e-05,
+      "loss": 3.3186,
+      "step": 6025
+    },
+    {
+      "epoch": 7.70656,
+      "grad_norm": 0.5295814871788025,
+      "learning_rate": 7.248620271907389e-05,
+      "loss": 3.3868,
+      "step": 6026
+    },
+    {
+      "epoch": 7.70784,
+      "grad_norm": 0.5124343037605286,
+      "learning_rate": 7.244582043343654e-05,
+      "loss": 3.4317,
+      "step": 6027
+    },
+    {
+      "epoch": 7.70912,
+      "grad_norm": 0.5283524990081787,
+      "learning_rate": 7.240543814779915e-05,
+      "loss": 3.3263,
+      "step": 6028
+    },
+    {
+      "epoch": 7.7104,
+      "grad_norm": 0.5237511992454529,
+      "learning_rate": 7.236505586216178e-05,
+      "loss": 3.3889,
+      "step": 6029
+    },
+    {
+      "epoch": 7.71168,
+      "grad_norm": 0.5321406126022339,
+      "learning_rate": 7.232467357652443e-05,
+      "loss": 3.3616,
+      "step": 6030
+    },
+    {
+      "epoch": 7.71296,
+      "grad_norm": 0.5302574634552002,
+      "learning_rate": 7.228429129088706e-05,
+      "loss": 3.4753,
+      "step": 6031
+    },
+    {
+      "epoch": 7.71424,
+      "grad_norm": 0.5124385356903076,
+      "learning_rate": 7.224390900524969e-05,
+      "loss": 3.3905,
+      "step": 6032
+    },
+    {
+      "epoch": 7.71552,
+      "grad_norm": 0.5328637361526489,
+      "learning_rate": 7.220352671961233e-05,
+      "loss": 3.4005,
+      "step": 6033
+    },
+    {
+      "epoch": 7.7168,
+      "grad_norm": 0.5470924377441406,
+      "learning_rate": 7.216314443397496e-05,
+      "loss": 3.3782,
+      "step": 6034
+    },
+    {
+      "epoch": 7.7180800000000005,
+      "grad_norm": 0.553785502910614,
+      "learning_rate": 7.212276214833759e-05,
+      "loss": 3.3522,
+      "step": 6035
+    },
+    {
+      "epoch": 7.71936,
+      "grad_norm": 0.533015251159668,
+      "learning_rate": 7.208237986270022e-05,
+      "loss": 3.3962,
+      "step": 6036
+    },
+    {
+      "epoch": 7.7206399999999995,
+      "grad_norm": 0.5438454747200012,
+      "learning_rate": 7.204199757706286e-05,
+      "loss": 3.3756,
+      "step": 6037
+    },
+    {
+      "epoch": 7.72192,
+      "grad_norm": 0.5408405661582947,
+      "learning_rate": 7.20016152914255e-05,
+      "loss": 3.3191,
+      "step": 6038
+    },
+    {
+      "epoch": 7.7232,
+      "grad_norm": 0.543483316898346,
+      "learning_rate": 7.196123300578812e-05,
+      "loss": 3.4062,
+      "step": 6039
+    },
+    {
+      "epoch": 7.72448,
+      "grad_norm": 0.49515143036842346,
+      "learning_rate": 7.192085072015075e-05,
+      "loss": 3.3466,
+      "step": 6040
+    },
+    {
+      "epoch": 7.72576,
+      "grad_norm": 0.5400490164756775,
+      "learning_rate": 7.188046843451338e-05,
+      "loss": 3.3837,
+      "step": 6041
+    },
+    {
+      "epoch": 7.72704,
+      "grad_norm": 0.5192726850509644,
+      "learning_rate": 7.184008614887601e-05,
+      "loss": 3.3819,
+      "step": 6042
+    },
+    {
+      "epoch": 7.72832,
+      "grad_norm": 0.5206571221351624,
+      "learning_rate": 7.179970386323864e-05,
+      "loss": 3.4132,
+      "step": 6043
+    },
+    {
+      "epoch": 7.7296,
+      "grad_norm": 0.5262637138366699,
+      "learning_rate": 7.175932157760129e-05,
+      "loss": 3.4121,
+      "step": 6044
+    },
+    {
+      "epoch": 7.73088,
+      "grad_norm": 0.5163931846618652,
+      "learning_rate": 7.171893929196392e-05,
+      "loss": 3.3883,
+      "step": 6045
+    },
+    {
+      "epoch": 7.73216,
+      "grad_norm": 0.5316890478134155,
+      "learning_rate": 7.167855700632655e-05,
+      "loss": 3.3518,
+      "step": 6046
+    },
+    {
+      "epoch": 7.73344,
+      "grad_norm": 0.5298048853874207,
+      "learning_rate": 7.163817472068918e-05,
+      "loss": 3.4238,
+      "step": 6047
+    },
+    {
+      "epoch": 7.73472,
+      "grad_norm": 0.5294202566146851,
+      "learning_rate": 7.159779243505182e-05,
+      "loss": 3.3889,
+      "step": 6048
+    },
+    {
+      "epoch": 7.736,
+      "grad_norm": 0.5132120847702026,
+      "learning_rate": 7.155741014941445e-05,
+      "loss": 3.365,
+      "step": 6049
+    },
+    {
+      "epoch": 7.73728,
+      "grad_norm": 0.5124995708465576,
+      "learning_rate": 7.151702786377708e-05,
+      "loss": 3.3348,
+      "step": 6050
+    },
+    {
+      "epoch": 7.73856,
+      "grad_norm": 0.5307555794715881,
+      "learning_rate": 7.147664557813973e-05,
+      "loss": 3.3825,
+      "step": 6051
+    },
+    {
+      "epoch": 7.73984,
+      "grad_norm": 0.5224402546882629,
+      "learning_rate": 7.143626329250236e-05,
+      "loss": 3.3489,
+      "step": 6052
+    },
+    {
+      "epoch": 7.7411200000000004,
+      "grad_norm": 0.5100345015525818,
+      "learning_rate": 7.139588100686499e-05,
+      "loss": 3.4441,
+      "step": 6053
+    },
+    {
+      "epoch": 7.7424,
+      "grad_norm": 0.5374388694763184,
+      "learning_rate": 7.135549872122762e-05,
+      "loss": 3.3539,
+      "step": 6054
+    },
+    {
+      "epoch": 7.74368,
+      "grad_norm": 0.5344933867454529,
+      "learning_rate": 7.131511643559025e-05,
+      "loss": 3.3818,
+      "step": 6055
+    },
+    {
+      "epoch": 7.74496,
+      "grad_norm": 0.5270034670829773,
+      "learning_rate": 7.127473414995288e-05,
+      "loss": 3.3496,
+      "step": 6056
+    },
+    {
+      "epoch": 7.74624,
+      "grad_norm": 0.5249508619308472,
+      "learning_rate": 7.12343518643155e-05,
+      "loss": 3.4065,
+      "step": 6057
+    },
+    {
+      "epoch": 7.74752,
+      "grad_norm": 0.5132743716239929,
+      "learning_rate": 7.119396957867815e-05,
+      "loss": 3.3701,
+      "step": 6058
+    },
+    {
+      "epoch": 7.7488,
+      "grad_norm": 0.5167114734649658,
+      "learning_rate": 7.115358729304078e-05,
+      "loss": 3.385,
+      "step": 6059
+    },
+    {
+      "epoch": 7.75008,
+      "grad_norm": 0.5301070809364319,
+      "learning_rate": 7.111320500740341e-05,
+      "loss": 3.3957,
+      "step": 6060
+    },
+    {
+      "epoch": 7.75136,
+      "grad_norm": 0.5281143188476562,
+      "learning_rate": 7.107282272176604e-05,
+      "loss": 3.4052,
+      "step": 6061
+    },
+    {
+      "epoch": 7.7526399999999995,
+      "grad_norm": 0.5261697173118591,
+      "learning_rate": 7.103244043612868e-05,
+      "loss": 3.4694,
+      "step": 6062
+    },
+    {
+      "epoch": 7.75392,
+      "grad_norm": 0.5167369246482849,
+      "learning_rate": 7.099205815049131e-05,
+      "loss": 3.378,
+      "step": 6063
+    },
+    {
+      "epoch": 7.7552,
+      "grad_norm": 0.5419825911521912,
+      "learning_rate": 7.095167586485394e-05,
+      "loss": 3.4047,
+      "step": 6064
+    },
+    {
+      "epoch": 7.75648,
+      "grad_norm": 0.5268843173980713,
+      "learning_rate": 7.091129357921657e-05,
+      "loss": 3.3606,
+      "step": 6065
+    },
+    {
+      "epoch": 7.75776,
+      "grad_norm": 0.5122654438018799,
+      "learning_rate": 7.087091129357922e-05,
+      "loss": 3.3676,
+      "step": 6066
+    },
+    {
+      "epoch": 7.75904,
+      "grad_norm": 0.5293072462081909,
+      "learning_rate": 7.083052900794185e-05,
+      "loss": 3.3321,
+      "step": 6067
+    },
+    {
+      "epoch": 7.76032,
+      "grad_norm": 0.5248163342475891,
+      "learning_rate": 7.079014672230448e-05,
+      "loss": 3.348,
+      "step": 6068
+    },
+    {
+      "epoch": 7.7616,
+      "grad_norm": 0.5331951975822449,
+      "learning_rate": 7.074976443666711e-05,
+      "loss": 3.3557,
+      "step": 6069
+    },
+    {
+      "epoch": 7.76288,
+      "grad_norm": 0.5159996151924133,
+      "learning_rate": 7.070938215102974e-05,
+      "loss": 3.3483,
+      "step": 6070
+    },
+    {
+      "epoch": 7.76416,
+      "grad_norm": 0.519347071647644,
+      "learning_rate": 7.066899986539237e-05,
+      "loss": 3.3843,
+      "step": 6071
+    },
+    {
+      "epoch": 7.76544,
+      "grad_norm": 0.5147475004196167,
+      "learning_rate": 7.062861757975501e-05,
+      "loss": 3.3476,
+      "step": 6072
+    },
+    {
+      "epoch": 7.76672,
+      "grad_norm": 0.5350569486618042,
+      "learning_rate": 7.058823529411764e-05,
+      "loss": 3.4024,
+      "step": 6073
+    },
+    {
+      "epoch": 7.768,
+      "grad_norm": 0.5119604468345642,
+      "learning_rate": 7.054785300848027e-05,
+      "loss": 3.3335,
+      "step": 6074
+    },
+    {
+      "epoch": 7.76928,
+      "grad_norm": 0.5323851108551025,
+      "learning_rate": 7.05074707228429e-05,
+      "loss": 3.4564,
+      "step": 6075
+    },
+    {
+      "epoch": 7.77056,
+      "grad_norm": 0.5254625678062439,
+      "learning_rate": 7.046708843720554e-05,
+      "loss": 3.4561,
+      "step": 6076
+    },
+    {
+      "epoch": 7.77184,
+      "grad_norm": 0.5181812644004822,
+      "learning_rate": 7.042670615156817e-05,
+      "loss": 3.3622,
+      "step": 6077
+    },
+    {
+      "epoch": 7.7731200000000005,
+      "grad_norm": 0.5225897431373596,
+      "learning_rate": 7.03863238659308e-05,
+      "loss": 3.4193,
+      "step": 6078
+    },
+    {
+      "epoch": 7.7744,
+      "grad_norm": 0.5444203615188599,
+      "learning_rate": 7.034594158029343e-05,
+      "loss": 3.4026,
+      "step": 6079
+    },
+    {
+      "epoch": 7.77568,
+      "grad_norm": 0.5264379978179932,
+      "learning_rate": 7.030555929465608e-05,
+      "loss": 3.448,
+      "step": 6080
+    },
+    {
+      "epoch": 7.77696,
+      "grad_norm": 0.5221927762031555,
+      "learning_rate": 7.026517700901871e-05,
+      "loss": 3.3665,
+      "step": 6081
+    },
+    {
+      "epoch": 7.77824,
+      "grad_norm": 0.5278477072715759,
+      "learning_rate": 7.022479472338134e-05,
+      "loss": 3.3952,
+      "step": 6082
+    },
+    {
+      "epoch": 7.77952,
+      "grad_norm": 0.5419744849205017,
+      "learning_rate": 7.018441243774397e-05,
+      "loss": 3.4158,
+      "step": 6083
+    },
+    {
+      "epoch": 7.7808,
+      "grad_norm": 0.5510439276695251,
+      "learning_rate": 7.01440301521066e-05,
+      "loss": 3.3966,
+      "step": 6084
+    },
+    {
+      "epoch": 7.78208,
+      "grad_norm": 0.5095901489257812,
+      "learning_rate": 7.010364786646923e-05,
+      "loss": 3.3745,
+      "step": 6085
+    },
+    {
+      "epoch": 7.78336,
+      "grad_norm": 0.5376297831535339,
+      "learning_rate": 7.006326558083187e-05,
+      "loss": 3.3438,
+      "step": 6086
+    },
+    {
+      "epoch": 7.78464,
+      "grad_norm": 0.5214537382125854,
+      "learning_rate": 7.00228832951945e-05,
+      "loss": 3.4108,
+      "step": 6087
+    },
+    {
+      "epoch": 7.78592,
+      "grad_norm": 0.5194911956787109,
+      "learning_rate": 6.998250100955713e-05,
+      "loss": 3.4137,
+      "step": 6088
+    },
+    {
+      "epoch": 7.7872,
+      "grad_norm": 0.528439462184906,
+      "learning_rate": 6.994211872391976e-05,
+      "loss": 3.3615,
+      "step": 6089
+    },
+    {
+      "epoch": 7.78848,
+      "grad_norm": 0.5447664856910706,
+      "learning_rate": 6.99017364382824e-05,
+      "loss": 3.3933,
+      "step": 6090
+    },
+    {
+      "epoch": 7.78976,
+      "grad_norm": 0.5379071235656738,
+      "learning_rate": 6.986135415264504e-05,
+      "loss": 3.4011,
+      "step": 6091
+    },
+    {
+      "epoch": 7.79104,
+      "grad_norm": 0.5375171303749084,
+      "learning_rate": 6.982097186700767e-05,
+      "loss": 3.4155,
+      "step": 6092
+    },
+    {
+      "epoch": 7.79232,
+      "grad_norm": 0.5257745385169983,
+      "learning_rate": 6.97805895813703e-05,
+      "loss": 3.4416,
+      "step": 6093
+    },
+    {
+      "epoch": 7.7936,
+      "grad_norm": 0.5261110663414001,
+      "learning_rate": 6.974020729573294e-05,
+      "loss": 3.4095,
+      "step": 6094
+    },
+    {
+      "epoch": 7.79488,
+      "grad_norm": 0.5397837162017822,
+      "learning_rate": 6.969982501009556e-05,
+      "loss": 3.2971,
+      "step": 6095
+    },
+    {
+      "epoch": 7.79616,
+      "grad_norm": 0.5269252061843872,
+      "learning_rate": 6.96594427244582e-05,
+      "loss": 3.3842,
+      "step": 6096
+    },
+    {
+      "epoch": 7.79744,
+      "grad_norm": 0.514562726020813,
+      "learning_rate": 6.961906043882083e-05,
+      "loss": 3.4729,
+      "step": 6097
+    },
+    {
+      "epoch": 7.79872,
+      "grad_norm": 0.5287720561027527,
+      "learning_rate": 6.957867815318346e-05,
+      "loss": 3.3533,
+      "step": 6098
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.5171098709106445,
+      "learning_rate": 6.953829586754609e-05,
+      "loss": 3.356,
+      "step": 6099
+    },
+    {
+      "epoch": 7.80128,
+      "grad_norm": 0.5358878374099731,
+      "learning_rate": 6.949791358190873e-05,
+      "loss": 3.3566,
+      "step": 6100
+    },
+    {
+      "epoch": 7.80256,
+      "grad_norm": 0.5136488080024719,
+      "learning_rate": 6.945753129627136e-05,
+      "loss": 3.4024,
+      "step": 6101
+    },
+    {
+      "epoch": 7.80384,
+      "grad_norm": 0.5352822542190552,
+      "learning_rate": 6.9417149010634e-05,
+      "loss": 3.4127,
+      "step": 6102
+    },
+    {
+      "epoch": 7.80512,
+      "grad_norm": 0.5107256174087524,
+      "learning_rate": 6.937676672499662e-05,
+      "loss": 3.3742,
+      "step": 6103
+    },
+    {
+      "epoch": 7.8064,
+      "grad_norm": 0.527970552444458,
+      "learning_rate": 6.933638443935927e-05,
+      "loss": 3.4109,
+      "step": 6104
+    },
+    {
+      "epoch": 7.8076799999999995,
+      "grad_norm": 0.5363277196884155,
+      "learning_rate": 6.92960021537219e-05,
+      "loss": 3.3898,
+      "step": 6105
+    },
+    {
+      "epoch": 7.80896,
+      "grad_norm": 0.5282407999038696,
+      "learning_rate": 6.925561986808453e-05,
+      "loss": 3.3657,
+      "step": 6106
+    },
+    {
+      "epoch": 7.81024,
+      "grad_norm": 0.5155186057090759,
+      "learning_rate": 6.921523758244716e-05,
+      "loss": 3.3866,
+      "step": 6107
+    },
+    {
+      "epoch": 7.81152,
+      "grad_norm": 0.5109987258911133,
+      "learning_rate": 6.91748552968098e-05,
+      "loss": 3.3924,
+      "step": 6108
+    },
+    {
+      "epoch": 7.8128,
+      "grad_norm": 0.5373751521110535,
+      "learning_rate": 6.913447301117242e-05,
+      "loss": 3.3803,
+      "step": 6109
+    },
+    {
+      "epoch": 7.81408,
+      "grad_norm": 0.5129071474075317,
+      "learning_rate": 6.909409072553506e-05,
+      "loss": 3.3581,
+      "step": 6110
+    },
+    {
+      "epoch": 7.81536,
+      "grad_norm": 0.5419071316719055,
+      "learning_rate": 6.905370843989769e-05,
+      "loss": 3.367,
+      "step": 6111
+    },
+    {
+      "epoch": 7.81664,
+      "grad_norm": 0.5342729687690735,
+      "learning_rate": 6.901332615426032e-05,
+      "loss": 3.5005,
+      "step": 6112
+    },
+    {
+      "epoch": 7.81792,
+      "grad_norm": 0.5251415371894836,
+      "learning_rate": 6.897294386862295e-05,
+      "loss": 3.3677,
+      "step": 6113
+    },
+    {
+      "epoch": 7.8192,
+      "grad_norm": 0.5298883318901062,
+      "learning_rate": 6.89325615829856e-05,
+      "loss": 3.4143,
+      "step": 6114
+    },
+    {
+      "epoch": 7.82048,
+      "grad_norm": 0.5260635018348694,
+      "learning_rate": 6.889217929734823e-05,
+      "loss": 3.4216,
+      "step": 6115
+    },
+    {
+      "epoch": 7.82176,
+      "grad_norm": 0.5399680733680725,
+      "learning_rate": 6.885179701171086e-05,
+      "loss": 3.3927,
+      "step": 6116
+    },
+    {
+      "epoch": 7.82304,
+      "grad_norm": 0.5211296081542969,
+      "learning_rate": 6.881141472607349e-05,
+      "loss": 3.4147,
+      "step": 6117
+    },
+    {
+      "epoch": 7.82432,
+      "grad_norm": 0.5149162411689758,
+      "learning_rate": 6.877103244043613e-05,
+      "loss": 3.4004,
+      "step": 6118
+    },
+    {
+      "epoch": 7.8256,
+      "grad_norm": 0.5208600759506226,
+      "learning_rate": 6.873065015479876e-05,
+      "loss": 3.305,
+      "step": 6119
+    },
+    {
+      "epoch": 7.82688,
+      "grad_norm": 0.5258411765098572,
+      "learning_rate": 6.869026786916139e-05,
+      "loss": 3.3602,
+      "step": 6120
+    },
+    {
+      "epoch": 7.8281600000000005,
+      "grad_norm": 0.5164110660552979,
+      "learning_rate": 6.864988558352402e-05,
+      "loss": 3.295,
+      "step": 6121
+    },
+    {
+      "epoch": 7.82944,
+      "grad_norm": 0.539932131767273,
+      "learning_rate": 6.860950329788665e-05,
+      "loss": 3.3754,
+      "step": 6122
+    },
+    {
+      "epoch": 7.83072,
+      "grad_norm": 0.5369081497192383,
+      "learning_rate": 6.856912101224928e-05,
+      "loss": 3.3797,
+      "step": 6123
+    },
+    {
+      "epoch": 7.832,
+      "grad_norm": 0.5187609791755676,
+      "learning_rate": 6.852873872661192e-05,
+      "loss": 3.3654,
+      "step": 6124
+    },
+    {
+      "epoch": 7.83328,
+      "grad_norm": 0.5491527915000916,
+      "learning_rate": 6.848835644097455e-05,
+      "loss": 3.4431,
+      "step": 6125
+    },
+    {
+      "epoch": 7.83456,
+      "grad_norm": 0.5081817507743835,
+      "learning_rate": 6.844797415533718e-05,
+      "loss": 3.4171,
+      "step": 6126
+    },
+    {
+      "epoch": 7.83584,
+      "grad_norm": 0.5206778645515442,
+      "learning_rate": 6.840759186969981e-05,
+      "loss": 3.38,
+      "step": 6127
+    },
+    {
+      "epoch": 7.83712,
+      "grad_norm": 0.5148397088050842,
+      "learning_rate": 6.836720958406246e-05,
+      "loss": 3.3939,
+      "step": 6128
+    },
+    {
+      "epoch": 7.8384,
+      "grad_norm": 0.5235511660575867,
+      "learning_rate": 6.832682729842509e-05,
+      "loss": 3.3396,
+      "step": 6129
+    },
+    {
+      "epoch": 7.8396799999999995,
+      "grad_norm": 0.5200296640396118,
+      "learning_rate": 6.828644501278772e-05,
+      "loss": 3.3397,
+      "step": 6130
+    },
+    {
+      "epoch": 7.84096,
+      "grad_norm": 0.5297372341156006,
+      "learning_rate": 6.824606272715035e-05,
+      "loss": 3.4623,
+      "step": 6131
+    },
+    {
+      "epoch": 7.84224,
+      "grad_norm": 0.5361821055412292,
+      "learning_rate": 6.820568044151299e-05,
+      "loss": 3.381,
+      "step": 6132
+    },
+    {
+      "epoch": 7.84352,
+      "grad_norm": 0.5185701847076416,
+      "learning_rate": 6.816529815587562e-05,
+      "loss": 3.3566,
+      "step": 6133
+    },
+    {
+      "epoch": 7.8448,
+      "grad_norm": 0.5242226123809814,
+      "learning_rate": 6.812491587023825e-05,
+      "loss": 3.3903,
+      "step": 6134
+    },
+    {
+      "epoch": 7.84608,
+      "grad_norm": 0.532424807548523,
+      "learning_rate": 6.808453358460088e-05,
+      "loss": 3.4556,
+      "step": 6135
+    },
+    {
+      "epoch": 7.84736,
+      "grad_norm": 0.5150805115699768,
+      "learning_rate": 6.804415129896351e-05,
+      "loss": 3.3373,
+      "step": 6136
+    },
+    {
+      "epoch": 7.84864,
+      "grad_norm": 0.5233043432235718,
+      "learning_rate": 6.800376901332614e-05,
+      "loss": 3.4448,
+      "step": 6137
+    },
+    {
+      "epoch": 7.84992,
+      "grad_norm": 0.517623245716095,
+      "learning_rate": 6.796338672768879e-05,
+      "loss": 3.3695,
+      "step": 6138
+    },
+    {
+      "epoch": 7.8512,
+      "grad_norm": 0.51863032579422,
+      "learning_rate": 6.792300444205142e-05,
+      "loss": 3.3963,
+      "step": 6139
+    },
+    {
+      "epoch": 7.85248,
+      "grad_norm": 0.52508544921875,
+      "learning_rate": 6.788262215641405e-05,
+      "loss": 3.3959,
+      "step": 6140
+    },
+    {
+      "epoch": 7.85376,
+      "grad_norm": 0.525643527507782,
+      "learning_rate": 6.784223987077668e-05,
+      "loss": 3.3873,
+      "step": 6141
+    },
+    {
+      "epoch": 7.85504,
+      "grad_norm": 0.5046698451042175,
+      "learning_rate": 6.780185758513932e-05,
+      "loss": 3.4145,
+      "step": 6142
+    },
+    {
+      "epoch": 7.85632,
+      "grad_norm": 0.5362650156021118,
+      "learning_rate": 6.776147529950195e-05,
+      "loss": 3.4785,
+      "step": 6143
+    },
+    {
+      "epoch": 7.8576,
+      "grad_norm": 0.5106462240219116,
+      "learning_rate": 6.772109301386458e-05,
+      "loss": 3.4154,
+      "step": 6144
+    },
+    {
+      "epoch": 7.85888,
+      "grad_norm": 0.5172496438026428,
+      "learning_rate": 6.768071072822721e-05,
+      "loss": 3.4247,
+      "step": 6145
+    },
+    {
+      "epoch": 7.8601600000000005,
+      "grad_norm": 0.5204678773880005,
+      "learning_rate": 6.764032844258985e-05,
+      "loss": 3.4397,
+      "step": 6146
+    },
+    {
+      "epoch": 7.86144,
+      "grad_norm": 0.5216001868247986,
+      "learning_rate": 6.759994615695248e-05,
+      "loss": 3.4111,
+      "step": 6147
+    },
+    {
+      "epoch": 7.86272,
+      "grad_norm": 0.5186067819595337,
+      "learning_rate": 6.755956387131511e-05,
+      "loss": 3.3803,
+      "step": 6148
+    },
+    {
+      "epoch": 7.864,
+      "grad_norm": 0.5313200950622559,
+      "learning_rate": 6.751918158567774e-05,
+      "loss": 3.3722,
+      "step": 6149
+    },
+    {
+      "epoch": 7.86528,
+      "grad_norm": 0.530690610408783,
+      "learning_rate": 6.747879930004037e-05,
+      "loss": 3.3698,
+      "step": 6150
+    },
+    {
+      "epoch": 7.86656,
+      "grad_norm": 0.5053249001502991,
+      "learning_rate": 6.7438417014403e-05,
+      "loss": 3.2738,
+      "step": 6151
+    },
+    {
+      "epoch": 7.86784,
+      "grad_norm": 0.5100921988487244,
+      "learning_rate": 6.739803472876563e-05,
+      "loss": 3.3805,
+      "step": 6152
+    },
+    {
+      "epoch": 7.86912,
+      "grad_norm": 0.5200990438461304,
+      "learning_rate": 6.735765244312828e-05,
+      "loss": 3.3805,
+      "step": 6153
+    },
+    {
+      "epoch": 7.8704,
+      "grad_norm": 0.5229262113571167,
+      "learning_rate": 6.731727015749091e-05,
+      "loss": 3.3675,
+      "step": 6154
+    },
+    {
+      "epoch": 7.87168,
+      "grad_norm": 0.536566436290741,
+      "learning_rate": 6.727688787185354e-05,
+      "loss": 3.4692,
+      "step": 6155
+    },
+    {
+      "epoch": 7.87296,
+      "grad_norm": 0.5252888798713684,
+      "learning_rate": 6.723650558621618e-05,
+      "loss": 3.375,
+      "step": 6156
+    },
+    {
+      "epoch": 7.87424,
+      "grad_norm": 0.5295806527137756,
+      "learning_rate": 6.719612330057881e-05,
+      "loss": 3.3758,
+      "step": 6157
+    },
+    {
+      "epoch": 7.87552,
+      "grad_norm": 0.5360879302024841,
+      "learning_rate": 6.715574101494144e-05,
+      "loss": 3.394,
+      "step": 6158
+    },
+    {
+      "epoch": 7.8768,
+      "grad_norm": 0.5231373310089111,
+      "learning_rate": 6.711535872930407e-05,
+      "loss": 3.3922,
+      "step": 6159
+    },
+    {
+      "epoch": 7.87808,
+      "grad_norm": 0.5426596999168396,
+      "learning_rate": 6.707497644366671e-05,
+      "loss": 3.4305,
+      "step": 6160
+    },
+    {
+      "epoch": 7.87936,
+      "grad_norm": 0.5190793871879578,
+      "learning_rate": 6.703459415802935e-05,
+      "loss": 3.3645,
+      "step": 6161
+    },
+    {
+      "epoch": 7.88064,
+      "grad_norm": 0.5220685005187988,
+      "learning_rate": 6.699421187239196e-05,
+      "loss": 3.3472,
+      "step": 6162
+    },
+    {
+      "epoch": 7.88192,
+      "grad_norm": 0.5117445588111877,
+      "learning_rate": 6.69538295867546e-05,
+      "loss": 3.3592,
+      "step": 6163
+    },
+    {
+      "epoch": 7.8832,
+      "grad_norm": 0.520155668258667,
+      "learning_rate": 6.691344730111724e-05,
+      "loss": 3.4109,
+      "step": 6164
+    },
+    {
+      "epoch": 7.88448,
+      "grad_norm": 0.5211771726608276,
+      "learning_rate": 6.687306501547987e-05,
+      "loss": 3.359,
+      "step": 6165
+    },
+    {
+      "epoch": 7.88576,
+      "grad_norm": 0.5182546973228455,
+      "learning_rate": 6.68326827298425e-05,
+      "loss": 3.3939,
+      "step": 6166
+    },
+    {
+      "epoch": 7.88704,
+      "grad_norm": 0.5254797339439392,
+      "learning_rate": 6.679230044420514e-05,
+      "loss": 3.359,
+      "step": 6167
+    },
+    {
+      "epoch": 7.88832,
+      "grad_norm": 0.5227944254875183,
+      "learning_rate": 6.675191815856777e-05,
+      "loss": 3.4111,
+      "step": 6168
+    },
+    {
+      "epoch": 7.8896,
+      "grad_norm": 0.5160208940505981,
+      "learning_rate": 6.67115358729304e-05,
+      "loss": 3.3495,
+      "step": 6169
+    },
+    {
+      "epoch": 7.89088,
+      "grad_norm": 0.5098345279693604,
+      "learning_rate": 6.667115358729303e-05,
+      "loss": 3.329,
+      "step": 6170
+    },
+    {
+      "epoch": 7.89216,
+      "grad_norm": 0.5212696194648743,
+      "learning_rate": 6.663077130165567e-05,
+      "loss": 3.4255,
+      "step": 6171
+    },
+    {
+      "epoch": 7.89344,
+      "grad_norm": 0.5201159715652466,
+      "learning_rate": 6.65903890160183e-05,
+      "loss": 3.3864,
+      "step": 6172
+    },
+    {
+      "epoch": 7.8947199999999995,
+      "grad_norm": 0.5263365507125854,
+      "learning_rate": 6.655000673038093e-05,
+      "loss": 3.3921,
+      "step": 6173
+    },
+    {
+      "epoch": 7.896,
+      "grad_norm": 0.5230242013931274,
+      "learning_rate": 6.650962444474358e-05,
+      "loss": 3.3675,
+      "step": 6174
+    },
+    {
+      "epoch": 7.89728,
+      "grad_norm": 0.5143063068389893,
+      "learning_rate": 6.64692421591062e-05,
+      "loss": 3.3708,
+      "step": 6175
+    },
+    {
+      "epoch": 7.89856,
+      "grad_norm": 0.5212923288345337,
+      "learning_rate": 6.642885987346882e-05,
+      "loss": 3.4216,
+      "step": 6176
+    },
+    {
+      "epoch": 7.89984,
+      "grad_norm": 0.5219258069992065,
+      "learning_rate": 6.638847758783147e-05,
+      "loss": 3.3633,
+      "step": 6177
+    },
+    {
+      "epoch": 7.90112,
+      "grad_norm": 0.5209653377532959,
+      "learning_rate": 6.63480953021941e-05,
+      "loss": 3.3671,
+      "step": 6178
+    },
+    {
+      "epoch": 7.9024,
+      "grad_norm": 0.5172528624534607,
+      "learning_rate": 6.630771301655673e-05,
+      "loss": 3.398,
+      "step": 6179
+    },
+    {
+      "epoch": 7.90368,
+      "grad_norm": 0.5204746127128601,
+      "learning_rate": 6.626733073091936e-05,
+      "loss": 3.3294,
+      "step": 6180
+    },
+    {
+      "epoch": 7.90496,
+      "grad_norm": 0.5181044936180115,
+      "learning_rate": 6.6226948445282e-05,
+      "loss": 3.2986,
+      "step": 6181
+    },
+    {
+      "epoch": 7.90624,
+      "grad_norm": 0.503993034362793,
+      "learning_rate": 6.618656615964463e-05,
+      "loss": 3.2642,
+      "step": 6182
+    },
+    {
+      "epoch": 7.90752,
+      "grad_norm": 0.5091139674186707,
+      "learning_rate": 6.614618387400726e-05,
+      "loss": 3.3122,
+      "step": 6183
+    },
+    {
+      "epoch": 7.9088,
+      "grad_norm": 0.5170398950576782,
+      "learning_rate": 6.610580158836989e-05,
+      "loss": 3.3814,
+      "step": 6184
+    },
+    {
+      "epoch": 7.91008,
+      "grad_norm": 0.5216673016548157,
+      "learning_rate": 6.606541930273253e-05,
+      "loss": 3.3738,
+      "step": 6185
+    },
+    {
+      "epoch": 7.91136,
+      "grad_norm": 0.5129463076591492,
+      "learning_rate": 6.602503701709516e-05,
+      "loss": 3.285,
+      "step": 6186
+    },
+    {
+      "epoch": 7.91264,
+      "grad_norm": 0.5073363184928894,
+      "learning_rate": 6.59846547314578e-05,
+      "loss": 3.3271,
+      "step": 6187
+    },
+    {
+      "epoch": 7.91392,
+      "grad_norm": 0.5254284143447876,
+      "learning_rate": 6.594427244582042e-05,
+      "loss": 3.3854,
+      "step": 6188
+    },
+    {
+      "epoch": 7.9152000000000005,
+      "grad_norm": 0.5197746157646179,
+      "learning_rate": 6.590389016018305e-05,
+      "loss": 3.4062,
+      "step": 6189
+    },
+    {
+      "epoch": 7.91648,
+      "grad_norm": 0.5055711269378662,
+      "learning_rate": 6.586350787454568e-05,
+      "loss": 3.3037,
+      "step": 6190
+    },
+    {
+      "epoch": 7.91776,
+      "grad_norm": 0.523854672908783,
+      "learning_rate": 6.582312558890833e-05,
+      "loss": 3.3311,
+      "step": 6191
+    },
+    {
+      "epoch": 7.91904,
+      "grad_norm": 0.5296027064323425,
+      "learning_rate": 6.578274330327096e-05,
+      "loss": 3.4084,
+      "step": 6192
+    },
+    {
+      "epoch": 7.92032,
+      "grad_norm": 0.5208480954170227,
+      "learning_rate": 6.574236101763359e-05,
+      "loss": 3.4572,
+      "step": 6193
+    },
+    {
+      "epoch": 7.9216,
+      "grad_norm": 0.5234736800193787,
+      "learning_rate": 6.570197873199622e-05,
+      "loss": 3.4112,
+      "step": 6194
+    },
+    {
+      "epoch": 7.92288,
+      "grad_norm": 0.5259544849395752,
+      "learning_rate": 6.566159644635886e-05,
+      "loss": 3.4369,
+      "step": 6195
+    },
+    {
+      "epoch": 7.92416,
+      "grad_norm": 0.5302821397781372,
+      "learning_rate": 6.562121416072149e-05,
+      "loss": 3.3975,
+      "step": 6196
+    },
+    {
+      "epoch": 7.92544,
+      "grad_norm": 0.5262648463249207,
+      "learning_rate": 6.558083187508412e-05,
+      "loss": 3.4066,
+      "step": 6197
+    },
+    {
+      "epoch": 7.9267199999999995,
+      "grad_norm": 0.5254402756690979,
+      "learning_rate": 6.554044958944675e-05,
+      "loss": 3.3711,
+      "step": 6198
+    },
+    {
+      "epoch": 7.928,
+      "grad_norm": 0.5265082120895386,
+      "learning_rate": 6.55000673038094e-05,
+      "loss": 3.367,
+      "step": 6199
+    },
+    {
+      "epoch": 7.92928,
+      "grad_norm": 0.5290309190750122,
+      "learning_rate": 6.545968501817203e-05,
+      "loss": 3.3912,
+      "step": 6200
+    },
+    {
+      "epoch": 7.93056,
+      "grad_norm": 0.5364521145820618,
+      "learning_rate": 6.541930273253466e-05,
+      "loss": 3.466,
+      "step": 6201
+    },
+    {
+      "epoch": 7.93184,
+      "grad_norm": 0.5207248330116272,
+      "learning_rate": 6.537892044689729e-05,
+      "loss": 3.3728,
+      "step": 6202
+    },
+    {
+      "epoch": 7.93312,
+      "grad_norm": 0.533179759979248,
+      "learning_rate": 6.533853816125992e-05,
+      "loss": 3.354,
+      "step": 6203
+    },
+    {
+      "epoch": 7.9344,
+      "grad_norm": 0.5137979388237,
+      "learning_rate": 6.529815587562255e-05,
+      "loss": 3.3751,
+      "step": 6204
+    },
+    {
+      "epoch": 7.93568,
+      "grad_norm": 0.5287781953811646,
+      "learning_rate": 6.525777358998519e-05,
+      "loss": 3.3851,
+      "step": 6205
+    },
+    {
+      "epoch": 7.93696,
+      "grad_norm": 0.5349233746528625,
+      "learning_rate": 6.521739130434782e-05,
+      "loss": 3.4174,
+      "step": 6206
+    },
+    {
+      "epoch": 7.93824,
+      "grad_norm": 0.5217055082321167,
+      "learning_rate": 6.517700901871045e-05,
+      "loss": 3.3981,
+      "step": 6207
+    },
+    {
+      "epoch": 7.93952,
+      "grad_norm": 0.5294625759124756,
+      "learning_rate": 6.513662673307308e-05,
+      "loss": 3.4748,
+      "step": 6208
+    },
+    {
+      "epoch": 7.9408,
+      "grad_norm": 0.5194503664970398,
+      "learning_rate": 6.509624444743572e-05,
+      "loss": 3.3751,
+      "step": 6209
+    },
+    {
+      "epoch": 7.94208,
+      "grad_norm": 0.528069257736206,
+      "learning_rate": 6.505586216179835e-05,
+      "loss": 3.3722,
+      "step": 6210
+    },
+    {
+      "epoch": 7.94336,
+      "grad_norm": 0.5228124856948853,
+      "learning_rate": 6.501547987616098e-05,
+      "loss": 3.3791,
+      "step": 6211
+    },
+    {
+      "epoch": 7.94464,
+      "grad_norm": 0.5172755718231201,
+      "learning_rate": 6.497509759052361e-05,
+      "loss": 3.4098,
+      "step": 6212
+    },
+    {
+      "epoch": 7.94592,
+      "grad_norm": 0.5152941346168518,
+      "learning_rate": 6.493471530488626e-05,
+      "loss": 3.4003,
+      "step": 6213
+    },
+    {
+      "epoch": 7.9472000000000005,
+      "grad_norm": 0.53346186876297,
+      "learning_rate": 6.489433301924889e-05,
+      "loss": 3.3025,
+      "step": 6214
+    },
+    {
+      "epoch": 7.94848,
+      "grad_norm": 0.5327422022819519,
+      "learning_rate": 6.485395073361152e-05,
+      "loss": 3.422,
+      "step": 6215
+    },
+    {
+      "epoch": 7.94976,
+      "grad_norm": 0.5168991088867188,
+      "learning_rate": 6.481356844797415e-05,
+      "loss": 3.3234,
+      "step": 6216
+    },
+    {
+      "epoch": 7.95104,
+      "grad_norm": 0.5401486754417419,
+      "learning_rate": 6.477318616233678e-05,
+      "loss": 3.4159,
+      "step": 6217
+    },
+    {
+      "epoch": 7.95232,
+      "grad_norm": 0.5209543704986572,
+      "learning_rate": 6.473280387669941e-05,
+      "loss": 3.3988,
+      "step": 6218
+    },
+    {
+      "epoch": 7.9536,
+      "grad_norm": 0.5165715217590332,
+      "learning_rate": 6.469242159106205e-05,
+      "loss": 3.3658,
+      "step": 6219
+    },
+    {
+      "epoch": 7.95488,
+      "grad_norm": 0.5266816020011902,
+      "learning_rate": 6.465203930542468e-05,
+      "loss": 3.4037,
+      "step": 6220
+    },
+    {
+      "epoch": 7.95616,
+      "grad_norm": 0.5167005658149719,
+      "learning_rate": 6.461165701978731e-05,
+      "loss": 3.343,
+      "step": 6221
+    },
+    {
+      "epoch": 7.95744,
+      "grad_norm": 0.5324759483337402,
+      "learning_rate": 6.457127473414994e-05,
+      "loss": 3.3873,
+      "step": 6222
+    },
+    {
+      "epoch": 7.95872,
+      "grad_norm": 0.5235298871994019,
+      "learning_rate": 6.453089244851259e-05,
+      "loss": 3.4334,
+      "step": 6223
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.5192801356315613,
+      "learning_rate": 6.449051016287522e-05,
+      "loss": 3.3825,
+      "step": 6224
+    },
+    {
+      "epoch": 7.96128,
+      "grad_norm": 0.5173263549804688,
+      "learning_rate": 6.445012787723785e-05,
+      "loss": 3.4143,
+      "step": 6225
+    },
+    {
+      "epoch": 7.96256,
+      "grad_norm": 0.5249988436698914,
+      "learning_rate": 6.440974559160048e-05,
+      "loss": 3.3573,
+      "step": 6226
+    },
+    {
+      "epoch": 7.96384,
+      "grad_norm": 0.5259899497032166,
+      "learning_rate": 6.436936330596312e-05,
+      "loss": 3.3955,
+      "step": 6227
+    },
+    {
+      "epoch": 7.96512,
+      "grad_norm": 0.5165438055992126,
+      "learning_rate": 6.432898102032575e-05,
+      "loss": 3.4257,
+      "step": 6228
+    },
+    {
+      "epoch": 7.9664,
+      "grad_norm": 0.5043834447860718,
+      "learning_rate": 6.428859873468838e-05,
+      "loss": 3.3238,
+      "step": 6229
+    },
+    {
+      "epoch": 7.96768,
+      "grad_norm": 0.5150227546691895,
+      "learning_rate": 6.424821644905101e-05,
+      "loss": 3.3857,
+      "step": 6230
+    },
+    {
+      "epoch": 7.96896,
+      "grad_norm": 0.5162477493286133,
+      "learning_rate": 6.420783416341364e-05,
+      "loss": 3.351,
+      "step": 6231
+    },
+    {
+      "epoch": 7.97024,
+      "grad_norm": 0.5149223804473877,
+      "learning_rate": 6.416745187777627e-05,
+      "loss": 3.3808,
+      "step": 6232
+    },
+    {
+      "epoch": 7.97152,
+      "grad_norm": 0.5162880420684814,
+      "learning_rate": 6.412706959213891e-05,
+      "loss": 3.3048,
+      "step": 6233
+    },
+    {
+      "epoch": 7.9728,
+      "grad_norm": 0.5108062624931335,
+      "learning_rate": 6.408668730650154e-05,
+      "loss": 3.442,
+      "step": 6234
+    },
+    {
+      "epoch": 7.97408,
+      "grad_norm": 0.5145354270935059,
+      "learning_rate": 6.404630502086417e-05,
+      "loss": 3.3521,
+      "step": 6235
+    },
+    {
+      "epoch": 7.97536,
+      "grad_norm": 0.5244855284690857,
+      "learning_rate": 6.40059227352268e-05,
+      "loss": 3.376,
+      "step": 6236
+    },
+    {
+      "epoch": 7.97664,
+      "grad_norm": 0.5116865038871765,
+      "learning_rate": 6.396554044958945e-05,
+      "loss": 3.3633,
+      "step": 6237
+    },
+    {
+      "epoch": 7.97792,
+      "grad_norm": 0.5205397009849548,
+      "learning_rate": 6.392515816395208e-05,
+      "loss": 3.4108,
+      "step": 6238
+    },
+    {
+      "epoch": 7.9792,
+      "grad_norm": 0.5130834579467773,
+      "learning_rate": 6.388477587831471e-05,
+      "loss": 3.3463,
+      "step": 6239
+    },
+    {
+      "epoch": 7.98048,
+      "grad_norm": 0.5261620283126831,
+      "learning_rate": 6.384439359267734e-05,
+      "loss": 3.4818,
+      "step": 6240
+    },
+    {
+      "epoch": 7.9817599999999995,
+      "grad_norm": 0.5290659070014954,
+      "learning_rate": 6.380401130703998e-05,
+      "loss": 3.4639,
+      "step": 6241
+    },
+    {
+      "epoch": 7.98304,
+      "grad_norm": 0.5278947353363037,
+      "learning_rate": 6.376362902140261e-05,
+      "loss": 3.4368,
+      "step": 6242
+    },
+    {
+      "epoch": 7.98432,
+      "grad_norm": 0.5171101093292236,
+      "learning_rate": 6.372324673576524e-05,
+      "loss": 3.3449,
+      "step": 6243
+    },
+    {
+      "epoch": 7.9856,
+      "grad_norm": 0.5271726250648499,
+      "learning_rate": 6.368286445012787e-05,
+      "loss": 3.3427,
+      "step": 6244
+    },
+    {
+      "epoch": 7.98688,
+      "grad_norm": 0.5079402923583984,
+      "learning_rate": 6.36424821644905e-05,
+      "loss": 3.3444,
+      "step": 6245
+    },
+    {
+      "epoch": 7.98816,
+      "grad_norm": 0.5159564018249512,
+      "learning_rate": 6.360209987885313e-05,
+      "loss": 3.3274,
+      "step": 6246
+    },
+    {
+      "epoch": 7.98944,
+      "grad_norm": 0.5179126858711243,
+      "learning_rate": 6.356171759321578e-05,
+      "loss": 3.2998,
+      "step": 6247
+    },
+    {
+      "epoch": 7.99072,
+      "grad_norm": 0.5274621248245239,
+      "learning_rate": 6.35213353075784e-05,
+      "loss": 3.2867,
+      "step": 6248
+    },
+    {
+      "epoch": 7.992,
+      "grad_norm": 0.5217450857162476,
+      "learning_rate": 6.348095302194104e-05,
+      "loss": 3.4232,
+      "step": 6249
+    },
+    {
+      "epoch": 7.99328,
+      "grad_norm": 0.5240947604179382,
+      "learning_rate": 6.344057073630367e-05,
+      "loss": 3.4344,
+      "step": 6250
+    },
+    {
+      "epoch": 7.99456,
+      "grad_norm": 0.5134342312812805,
+      "learning_rate": 6.340018845066631e-05,
+      "loss": 3.365,
+      "step": 6251
+    },
+    {
+      "epoch": 7.99584,
+      "grad_norm": 0.5243435502052307,
+      "learning_rate": 6.335980616502894e-05,
+      "loss": 3.3951,
+      "step": 6252
+    },
+    {
+      "epoch": 7.99712,
+      "grad_norm": 0.5112805366516113,
+      "learning_rate": 6.331942387939157e-05,
+      "loss": 3.3713,
+      "step": 6253
+    },
+    {
+      "epoch": 7.9984,
+      "grad_norm": 0.532479465007782,
+      "learning_rate": 6.32790415937542e-05,
+      "loss": 3.4131,
+      "step": 6254
+    },
+    {
+      "epoch": 7.99968,
+      "grad_norm": 0.5240690112113953,
+      "learning_rate": 6.323865930811684e-05,
+      "loss": 3.3779,
+      "step": 6255
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.0870201587677002,
+      "learning_rate": 6.319827702247946e-05,
+      "loss": 3.4098,
+      "step": 6256
+    },
+    {
+      "epoch": 8.00128,
+      "grad_norm": 0.5369760394096375,
+      "learning_rate": 6.315789473684209e-05,
+      "loss": 3.3303,
+      "step": 6257
+    },
+    {
+      "epoch": 8.00256,
+      "grad_norm": 0.5155583620071411,
+      "learning_rate": 6.311751245120473e-05,
+      "loss": 3.3235,
+      "step": 6258
+    },
+    {
+      "epoch": 8.00384,
+      "grad_norm": 0.50521320104599,
+      "learning_rate": 6.307713016556736e-05,
+      "loss": 3.3097,
+      "step": 6259
+    },
+    {
+      "epoch": 8.00512,
+      "grad_norm": 0.5237175226211548,
+      "learning_rate": 6.303674787993e-05,
+      "loss": 3.3051,
+      "step": 6260
+    },
+    {
+      "epoch": 8.0064,
+      "grad_norm": 0.521220862865448,
+      "learning_rate": 6.299636559429264e-05,
+      "loss": 3.2319,
+      "step": 6261
+    },
+    {
+      "epoch": 8.00768,
+      "grad_norm": 0.5294133424758911,
+      "learning_rate": 6.295598330865527e-05,
+      "loss": 3.2934,
+      "step": 6262
+    },
+    {
+      "epoch": 8.00896,
+      "grad_norm": 0.5125667452812195,
+      "learning_rate": 6.29156010230179e-05,
+      "loss": 3.1766,
+      "step": 6263
+    },
+    {
+      "epoch": 8.01024,
+      "grad_norm": 0.5298846364021301,
+      "learning_rate": 6.287521873738053e-05,
+      "loss": 3.2827,
+      "step": 6264
+    },
+    {
+      "epoch": 8.01152,
+      "grad_norm": 0.5239753127098083,
+      "learning_rate": 6.283483645174317e-05,
+      "loss": 3.256,
+      "step": 6265
+    },
+    {
+      "epoch": 8.0128,
+      "grad_norm": 0.5199098587036133,
+      "learning_rate": 6.27944541661058e-05,
+      "loss": 3.2353,
+      "step": 6266
+    },
+    {
+      "epoch": 8.01408,
+      "grad_norm": 0.5292956233024597,
+      "learning_rate": 6.275407188046843e-05,
+      "loss": 3.2311,
+      "step": 6267
+    },
+    {
+      "epoch": 8.01536,
+      "grad_norm": 0.5216575860977173,
+      "learning_rate": 6.271368959483106e-05,
+      "loss": 3.3264,
+      "step": 6268
+    },
+    {
+      "epoch": 8.01664,
+      "grad_norm": 0.514209508895874,
+      "learning_rate": 6.26733073091937e-05,
+      "loss": 3.2135,
+      "step": 6269
+    },
+    {
+      "epoch": 8.01792,
+      "grad_norm": 0.5441368818283081,
+      "learning_rate": 6.263292502355632e-05,
+      "loss": 3.3422,
+      "step": 6270
+    },
+    {
+      "epoch": 8.0192,
+      "grad_norm": 0.5475019216537476,
+      "learning_rate": 6.259254273791895e-05,
+      "loss": 3.2504,
+      "step": 6271
+    },
+    {
+      "epoch": 8.02048,
+      "grad_norm": 0.5108786821365356,
+      "learning_rate": 6.25521604522816e-05,
+      "loss": 3.291,
+      "step": 6272
+    },
+    {
+      "epoch": 8.02176,
+      "grad_norm": 0.532529890537262,
+      "learning_rate": 6.251177816664423e-05,
+      "loss": 3.3117,
+      "step": 6273
+    },
+    {
+      "epoch": 8.02304,
+      "grad_norm": 0.5336194634437561,
+      "learning_rate": 6.247139588100686e-05,
+      "loss": 3.2797,
+      "step": 6274
+    },
+    {
+      "epoch": 8.02432,
+      "grad_norm": 0.5180751085281372,
+      "learning_rate": 6.243101359536949e-05,
+      "loss": 3.2857,
+      "step": 6275
+    },
+    {
+      "epoch": 8.0256,
+      "grad_norm": 0.5190191864967346,
+      "learning_rate": 6.239063130973213e-05,
+      "loss": 3.2449,
+      "step": 6276
+    },
+    {
+      "epoch": 8.02688,
+      "grad_norm": 0.533854603767395,
+      "learning_rate": 6.235024902409476e-05,
+      "loss": 3.3588,
+      "step": 6277
+    },
+    {
+      "epoch": 8.02816,
+      "grad_norm": 0.5298550724983215,
+      "learning_rate": 6.230986673845739e-05,
+      "loss": 3.2968,
+      "step": 6278
+    },
+    {
+      "epoch": 8.02944,
+      "grad_norm": 0.5271071791648865,
+      "learning_rate": 6.226948445282003e-05,
+      "loss": 3.18,
+      "step": 6279
+    },
+    {
+      "epoch": 8.03072,
+      "grad_norm": 0.5124052166938782,
+      "learning_rate": 6.222910216718266e-05,
+      "loss": 3.3069,
+      "step": 6280
+    },
+    {
+      "epoch": 8.032,
+      "grad_norm": 0.5358180403709412,
+      "learning_rate": 6.218871988154529e-05,
+      "loss": 3.2968,
+      "step": 6281
+    },
+    {
+      "epoch": 8.03328,
+      "grad_norm": 0.5310022234916687,
+      "learning_rate": 6.214833759590792e-05,
+      "loss": 3.3983,
+      "step": 6282
+    },
+    {
+      "epoch": 8.03456,
+      "grad_norm": 0.5346135497093201,
+      "learning_rate": 6.210795531027055e-05,
+      "loss": 3.2622,
+      "step": 6283
+    },
+    {
+      "epoch": 8.03584,
+      "grad_norm": 0.5122181177139282,
+      "learning_rate": 6.206757302463318e-05,
+      "loss": 3.3501,
+      "step": 6284
+    },
+    {
+      "epoch": 8.03712,
+      "grad_norm": 0.5202990770339966,
+      "learning_rate": 6.202719073899581e-05,
+      "loss": 3.2425,
+      "step": 6285
+    },
+    {
+      "epoch": 8.0384,
+      "grad_norm": 0.5240968465805054,
+      "learning_rate": 6.198680845335846e-05,
+      "loss": 3.2669,
+      "step": 6286
+    },
+    {
+      "epoch": 8.03968,
+      "grad_norm": 0.5298383831977844,
+      "learning_rate": 6.194642616772109e-05,
+      "loss": 3.3057,
+      "step": 6287
+    },
+    {
+      "epoch": 8.04096,
+      "grad_norm": 0.5407096743583679,
+      "learning_rate": 6.190604388208372e-05,
+      "loss": 3.2781,
+      "step": 6288
+    },
+    {
+      "epoch": 8.04224,
+      "grad_norm": 0.5193719267845154,
+      "learning_rate": 6.186566159644635e-05,
+      "loss": 3.3438,
+      "step": 6289
+    },
+    {
+      "epoch": 8.043520000000001,
+      "grad_norm": 0.5216055512428284,
+      "learning_rate": 6.182527931080899e-05,
+      "loss": 3.2728,
+      "step": 6290
+    },
+    {
+      "epoch": 8.0448,
+      "grad_norm": 0.5255246758460999,
+      "learning_rate": 6.178489702517162e-05,
+      "loss": 3.2996,
+      "step": 6291
+    },
+    {
+      "epoch": 8.04608,
+      "grad_norm": 0.5254226326942444,
+      "learning_rate": 6.174451473953425e-05,
+      "loss": 3.293,
+      "step": 6292
+    },
+    {
+      "epoch": 8.04736,
+      "grad_norm": 0.5267927050590515,
+      "learning_rate": 6.170413245389688e-05,
+      "loss": 3.2104,
+      "step": 6293
+    },
+    {
+      "epoch": 8.04864,
+      "grad_norm": 0.5204477906227112,
+      "learning_rate": 6.166375016825952e-05,
+      "loss": 3.297,
+      "step": 6294
+    },
+    {
+      "epoch": 8.04992,
+      "grad_norm": 0.5418780446052551,
+      "learning_rate": 6.162336788262215e-05,
+      "loss": 3.3136,
+      "step": 6295
+    },
+    {
+      "epoch": 8.0512,
+      "grad_norm": 0.5511976480484009,
+      "learning_rate": 6.158298559698478e-05,
+      "loss": 3.332,
+      "step": 6296
+    },
+    {
+      "epoch": 8.05248,
+      "grad_norm": 0.5536786317825317,
+      "learning_rate": 6.154260331134741e-05,
+      "loss": 3.3544,
+      "step": 6297
+    },
+    {
+      "epoch": 8.05376,
+      "grad_norm": 0.5378971099853516,
+      "learning_rate": 6.150222102571004e-05,
+      "loss": 3.2811,
+      "step": 6298
+    },
+    {
+      "epoch": 8.05504,
+      "grad_norm": 0.5422542095184326,
+      "learning_rate": 6.146183874007267e-05,
+      "loss": 3.2559,
+      "step": 6299
+    },
+    {
+      "epoch": 8.05632,
+      "grad_norm": 0.5333569049835205,
+      "learning_rate": 6.142145645443532e-05,
+      "loss": 3.3139,
+      "step": 6300
+    },
+    {
+      "epoch": 8.0576,
+      "grad_norm": 0.5348880290985107,
+      "learning_rate": 6.138107416879795e-05,
+      "loss": 3.3467,
+      "step": 6301
+    },
+    {
+      "epoch": 8.05888,
+      "grad_norm": 0.5167289972305298,
+      "learning_rate": 6.134069188316058e-05,
+      "loss": 3.2289,
+      "step": 6302
+    },
+    {
+      "epoch": 8.06016,
+      "grad_norm": 0.5306516289710999,
+      "learning_rate": 6.130030959752321e-05,
+      "loss": 3.3242,
+      "step": 6303
+    },
+    {
+      "epoch": 8.06144,
+      "grad_norm": 0.5348290801048279,
+      "learning_rate": 6.125992731188585e-05,
+      "loss": 3.36,
+      "step": 6304
+    },
+    {
+      "epoch": 8.06272,
+      "grad_norm": 0.5243771076202393,
+      "learning_rate": 6.121954502624848e-05,
+      "loss": 3.2549,
+      "step": 6305
+    },
+    {
+      "epoch": 8.064,
+      "grad_norm": 0.5279974341392517,
+      "learning_rate": 6.117916274061111e-05,
+      "loss": 3.3017,
+      "step": 6306
+    },
+    {
+      "epoch": 8.06528,
+      "grad_norm": 0.539170503616333,
+      "learning_rate": 6.113878045497374e-05,
+      "loss": 3.2932,
+      "step": 6307
+    },
+    {
+      "epoch": 8.06656,
+      "grad_norm": 0.5232791304588318,
+      "learning_rate": 6.109839816933639e-05,
+      "loss": 3.2744,
+      "step": 6308
+    },
+    {
+      "epoch": 8.06784,
+      "grad_norm": 0.5283025503158569,
+      "learning_rate": 6.105801588369902e-05,
+      "loss": 3.3529,
+      "step": 6309
+    },
+    {
+      "epoch": 8.06912,
+      "grad_norm": 0.5246297121047974,
+      "learning_rate": 6.101763359806164e-05,
+      "loss": 3.2392,
+      "step": 6310
+    },
+    {
+      "epoch": 8.0704,
+      "grad_norm": 0.5458753108978271,
+      "learning_rate": 6.0977251312424276e-05,
+      "loss": 3.2406,
+      "step": 6311
+    },
+    {
+      "epoch": 8.07168,
+      "grad_norm": 0.5204522609710693,
+      "learning_rate": 6.0936869026786907e-05,
+      "loss": 3.3418,
+      "step": 6312
+    },
+    {
+      "epoch": 8.07296,
+      "grad_norm": 0.5313632488250732,
+      "learning_rate": 6.089648674114954e-05,
+      "loss": 3.3164,
+      "step": 6313
+    },
+    {
+      "epoch": 8.07424,
+      "grad_norm": 0.5227934122085571,
+      "learning_rate": 6.0856104455512173e-05,
+      "loss": 3.2728,
+      "step": 6314
+    },
+    {
+      "epoch": 8.07552,
+      "grad_norm": 0.5291623473167419,
+      "learning_rate": 6.081572216987481e-05,
+      "loss": 3.307,
+      "step": 6315
+    },
+    {
+      "epoch": 8.0768,
+      "grad_norm": 0.5221740007400513,
+      "learning_rate": 6.077533988423744e-05,
+      "loss": 3.301,
+      "step": 6316
+    },
+    {
+      "epoch": 8.07808,
+      "grad_norm": 0.528735339641571,
+      "learning_rate": 6.073495759860008e-05,
+      "loss": 3.2217,
+      "step": 6317
+    },
+    {
+      "epoch": 8.07936,
+      "grad_norm": 0.5461089611053467,
+      "learning_rate": 6.069457531296271e-05,
+      "loss": 3.2215,
+      "step": 6318
+    },
+    {
+      "epoch": 8.08064,
+      "grad_norm": 0.5363235473632812,
+      "learning_rate": 6.0654193027325344e-05,
+      "loss": 3.2331,
+      "step": 6319
+    },
+    {
+      "epoch": 8.08192,
+      "grad_norm": 0.5322678089141846,
+      "learning_rate": 6.0613810741687974e-05,
+      "loss": 3.2927,
+      "step": 6320
+    },
+    {
+      "epoch": 8.0832,
+      "grad_norm": 0.5260446071624756,
+      "learning_rate": 6.057342845605061e-05,
+      "loss": 3.2021,
+      "step": 6321
+    },
+    {
+      "epoch": 8.08448,
+      "grad_norm": 0.5323824286460876,
+      "learning_rate": 6.053304617041324e-05,
+      "loss": 3.3024,
+      "step": 6322
+    },
+    {
+      "epoch": 8.08576,
+      "grad_norm": 0.5511333346366882,
+      "learning_rate": 6.049266388477588e-05,
+      "loss": 3.2838,
+      "step": 6323
+    },
+    {
+      "epoch": 8.08704,
+      "grad_norm": 0.5331417322158813,
+      "learning_rate": 6.04522815991385e-05,
+      "loss": 3.3127,
+      "step": 6324
+    },
+    {
+      "epoch": 8.08832,
+      "grad_norm": 0.5371091961860657,
+      "learning_rate": 6.041189931350114e-05,
+      "loss": 3.3004,
+      "step": 6325
+    },
+    {
+      "epoch": 8.0896,
+      "grad_norm": 0.5571906566619873,
+      "learning_rate": 6.037151702786377e-05,
+      "loss": 3.2509,
+      "step": 6326
+    },
+    {
+      "epoch": 8.09088,
+      "grad_norm": 0.5286134481430054,
+      "learning_rate": 6.0331134742226405e-05,
+      "loss": 3.3488,
+      "step": 6327
+    },
+    {
+      "epoch": 8.09216,
+      "grad_norm": 0.5509176254272461,
+      "learning_rate": 6.0290752456589035e-05,
+      "loss": 3.3017,
+      "step": 6328
+    },
+    {
+      "epoch": 8.09344,
+      "grad_norm": 0.5480915307998657,
+      "learning_rate": 6.025037017095167e-05,
+      "loss": 3.2503,
+      "step": 6329
+    },
+    {
+      "epoch": 8.09472,
+      "grad_norm": 0.5310816764831543,
+      "learning_rate": 6.02099878853143e-05,
+      "loss": 3.309,
+      "step": 6330
+    },
+    {
+      "epoch": 8.096,
+      "grad_norm": 0.5437869429588318,
+      "learning_rate": 6.016960559967694e-05,
+      "loss": 3.2537,
+      "step": 6331
+    },
+    {
+      "epoch": 8.09728,
+      "grad_norm": 0.5316812992095947,
+      "learning_rate": 6.012922331403957e-05,
+      "loss": 3.2957,
+      "step": 6332
+    },
+    {
+      "epoch": 8.09856,
+      "grad_norm": 0.5401766896247864,
+      "learning_rate": 6.0088841028402206e-05,
+      "loss": 3.2455,
+      "step": 6333
+    },
+    {
+      "epoch": 8.09984,
+      "grad_norm": 0.518835723400116,
+      "learning_rate": 6.0048458742764836e-05,
+      "loss": 3.298,
+      "step": 6334
+    },
+    {
+      "epoch": 8.10112,
+      "grad_norm": 0.531012237071991,
+      "learning_rate": 6.000807645712747e-05,
+      "loss": 3.2884,
+      "step": 6335
+    },
+    {
+      "epoch": 8.1024,
+      "grad_norm": 0.5368117094039917,
+      "learning_rate": 5.99676941714901e-05,
+      "loss": 3.2962,
+      "step": 6336
+    },
+    {
+      "epoch": 8.10368,
+      "grad_norm": 0.5301125049591064,
+      "learning_rate": 5.992731188585273e-05,
+      "loss": 3.366,
+      "step": 6337
+    },
+    {
+      "epoch": 8.10496,
+      "grad_norm": 0.5062065720558167,
+      "learning_rate": 5.988692960021536e-05,
+      "loss": 3.2497,
+      "step": 6338
+    },
+    {
+      "epoch": 8.10624,
+      "grad_norm": 0.5318953990936279,
+      "learning_rate": 5.9846547314578e-05,
+      "loss": 3.3522,
+      "step": 6339
+    },
+    {
+      "epoch": 8.10752,
+      "grad_norm": 0.5129379034042358,
+      "learning_rate": 5.980616502894063e-05,
+      "loss": 3.2826,
+      "step": 6340
+    },
+    {
+      "epoch": 8.1088,
+      "grad_norm": 0.5241764187812805,
+      "learning_rate": 5.976578274330327e-05,
+      "loss": 3.295,
+      "step": 6341
+    },
+    {
+      "epoch": 8.11008,
+      "grad_norm": 0.5228458046913147,
+      "learning_rate": 5.97254004576659e-05,
+      "loss": 3.2485,
+      "step": 6342
+    },
+    {
+      "epoch": 8.11136,
+      "grad_norm": 0.523200273513794,
+      "learning_rate": 5.9685018172028534e-05,
+      "loss": 3.306,
+      "step": 6343
+    },
+    {
+      "epoch": 8.11264,
+      "grad_norm": 0.5235135555267334,
+      "learning_rate": 5.9644635886391164e-05,
+      "loss": 3.2701,
+      "step": 6344
+    },
+    {
+      "epoch": 8.11392,
+      "grad_norm": 0.5147677063941956,
+      "learning_rate": 5.96042536007538e-05,
+      "loss": 3.2849,
+      "step": 6345
+    },
+    {
+      "epoch": 8.1152,
+      "grad_norm": 0.5381238460540771,
+      "learning_rate": 5.956387131511643e-05,
+      "loss": 3.3421,
+      "step": 6346
+    },
+    {
+      "epoch": 8.11648,
+      "grad_norm": 0.5454177260398865,
+      "learning_rate": 5.952348902947907e-05,
+      "loss": 3.3036,
+      "step": 6347
+    },
+    {
+      "epoch": 8.11776,
+      "grad_norm": 0.5195689797401428,
+      "learning_rate": 5.94831067438417e-05,
+      "loss": 3.2856,
+      "step": 6348
+    },
+    {
+      "epoch": 8.11904,
+      "grad_norm": 0.5343729853630066,
+      "learning_rate": 5.9442724458204335e-05,
+      "loss": 3.2902,
+      "step": 6349
+    },
+    {
+      "epoch": 8.12032,
+      "grad_norm": 0.5302938222885132,
+      "learning_rate": 5.940234217256696e-05,
+      "loss": 3.2755,
+      "step": 6350
+    },
+    {
+      "epoch": 8.1216,
+      "grad_norm": 0.5336959958076477,
+      "learning_rate": 5.9361959886929595e-05,
+      "loss": 3.2973,
+      "step": 6351
+    },
+    {
+      "epoch": 8.12288,
+      "grad_norm": 0.525851845741272,
+      "learning_rate": 5.9321577601292225e-05,
+      "loss": 3.2702,
+      "step": 6352
+    },
+    {
+      "epoch": 8.12416,
+      "grad_norm": 0.5346911549568176,
+      "learning_rate": 5.928119531565486e-05,
+      "loss": 3.3683,
+      "step": 6353
+    },
+    {
+      "epoch": 8.12544,
+      "grad_norm": 0.5326193571090698,
+      "learning_rate": 5.924081303001749e-05,
+      "loss": 3.3173,
+      "step": 6354
+    },
+    {
+      "epoch": 8.12672,
+      "grad_norm": 0.5400025844573975,
+      "learning_rate": 5.920043074438013e-05,
+      "loss": 3.279,
+      "step": 6355
+    },
+    {
+      "epoch": 8.128,
+      "grad_norm": 0.5323216915130615,
+      "learning_rate": 5.916004845874276e-05,
+      "loss": 3.2818,
+      "step": 6356
+    },
+    {
+      "epoch": 8.12928,
+      "grad_norm": 0.5249941945075989,
+      "learning_rate": 5.9119666173105395e-05,
+      "loss": 3.2571,
+      "step": 6357
+    },
+    {
+      "epoch": 8.13056,
+      "grad_norm": 0.5245072245597839,
+      "learning_rate": 5.9079283887468026e-05,
+      "loss": 3.2323,
+      "step": 6358
+    },
+    {
+      "epoch": 8.13184,
+      "grad_norm": 0.5230678915977478,
+      "learning_rate": 5.903890160183066e-05,
+      "loss": 3.2932,
+      "step": 6359
+    },
+    {
+      "epoch": 8.13312,
+      "grad_norm": 0.5385542511940002,
+      "learning_rate": 5.899851931619329e-05,
+      "loss": 3.3357,
+      "step": 6360
+    },
+    {
+      "epoch": 8.1344,
+      "grad_norm": 0.5333831310272217,
+      "learning_rate": 5.895813703055593e-05,
+      "loss": 3.301,
+      "step": 6361
+    },
+    {
+      "epoch": 8.13568,
+      "grad_norm": 0.5369766354560852,
+      "learning_rate": 5.891775474491856e-05,
+      "loss": 3.2642,
+      "step": 6362
+    },
+    {
+      "epoch": 8.13696,
+      "grad_norm": 0.5322566032409668,
+      "learning_rate": 5.8877372459281196e-05,
+      "loss": 3.2542,
+      "step": 6363
+    },
+    {
+      "epoch": 8.13824,
+      "grad_norm": 0.5302499532699585,
+      "learning_rate": 5.883699017364382e-05,
+      "loss": 3.2494,
+      "step": 6364
+    },
+    {
+      "epoch": 8.13952,
+      "grad_norm": 0.5144988298416138,
+      "learning_rate": 5.879660788800645e-05,
+      "loss": 3.1926,
+      "step": 6365
+    },
+    {
+      "epoch": 8.1408,
+      "grad_norm": 0.5225915908813477,
+      "learning_rate": 5.8756225602369086e-05,
+      "loss": 3.3514,
+      "step": 6366
+    },
+    {
+      "epoch": 8.14208,
+      "grad_norm": 0.5414659976959229,
+      "learning_rate": 5.871584331673172e-05,
+      "loss": 3.3684,
+      "step": 6367
+    },
+    {
+      "epoch": 8.14336,
+      "grad_norm": 0.5391842722892761,
+      "learning_rate": 5.867546103109435e-05,
+      "loss": 3.3141,
+      "step": 6368
+    },
+    {
+      "epoch": 8.14464,
+      "grad_norm": 0.5198903679847717,
+      "learning_rate": 5.863507874545699e-05,
+      "loss": 3.284,
+      "step": 6369
+    },
+    {
+      "epoch": 8.14592,
+      "grad_norm": 0.5301628112792969,
+      "learning_rate": 5.859469645981962e-05,
+      "loss": 3.2225,
+      "step": 6370
+    },
+    {
+      "epoch": 8.1472,
+      "grad_norm": 0.5428996682167053,
+      "learning_rate": 5.855431417418226e-05,
+      "loss": 3.3104,
+      "step": 6371
+    },
+    {
+      "epoch": 8.14848,
+      "grad_norm": 0.5339773297309875,
+      "learning_rate": 5.851393188854489e-05,
+      "loss": 3.3018,
+      "step": 6372
+    },
+    {
+      "epoch": 8.14976,
+      "grad_norm": 0.52174973487854,
+      "learning_rate": 5.8473549602907524e-05,
+      "loss": 3.2703,
+      "step": 6373
+    },
+    {
+      "epoch": 8.15104,
+      "grad_norm": 0.5308415293693542,
+      "learning_rate": 5.8433167317270154e-05,
+      "loss": 3.2691,
+      "step": 6374
+    },
+    {
+      "epoch": 8.15232,
+      "grad_norm": 0.5573276877403259,
+      "learning_rate": 5.839278503163279e-05,
+      "loss": 3.34,
+      "step": 6375
+    },
+    {
+      "epoch": 8.1536,
+      "grad_norm": 0.5412933230400085,
+      "learning_rate": 5.835240274599542e-05,
+      "loss": 3.286,
+      "step": 6376
+    },
+    {
+      "epoch": 8.15488,
+      "grad_norm": 0.5277814269065857,
+      "learning_rate": 5.8312020460358044e-05,
+      "loss": 3.3651,
+      "step": 6377
+    },
+    {
+      "epoch": 8.15616,
+      "grad_norm": 0.5270769596099854,
+      "learning_rate": 5.827163817472068e-05,
+      "loss": 3.3258,
+      "step": 6378
+    },
+    {
+      "epoch": 8.15744,
+      "grad_norm": 0.5277435779571533,
+      "learning_rate": 5.823125588908331e-05,
+      "loss": 3.1801,
+      "step": 6379
+    },
+    {
+      "epoch": 8.15872,
+      "grad_norm": 0.5209450125694275,
+      "learning_rate": 5.819087360344595e-05,
+      "loss": 3.2954,
+      "step": 6380
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.5338136553764343,
+      "learning_rate": 5.815049131780858e-05,
+      "loss": 3.2964,
+      "step": 6381
+    },
+    {
+      "epoch": 8.16128,
+      "grad_norm": 0.547455370426178,
+      "learning_rate": 5.8110109032171215e-05,
+      "loss": 3.3137,
+      "step": 6382
+    },
+    {
+      "epoch": 8.16256,
+      "grad_norm": 0.5422819256782532,
+      "learning_rate": 5.8069726746533845e-05,
+      "loss": 3.2683,
+      "step": 6383
+    },
+    {
+      "epoch": 8.16384,
+      "grad_norm": 0.536233127117157,
+      "learning_rate": 5.802934446089648e-05,
+      "loss": 3.3036,
+      "step": 6384
+    },
+    {
+      "epoch": 8.16512,
+      "grad_norm": 0.5339389443397522,
+      "learning_rate": 5.798896217525911e-05,
+      "loss": 3.3953,
+      "step": 6385
+    },
+    {
+      "epoch": 8.1664,
+      "grad_norm": 0.5347949266433716,
+      "learning_rate": 5.794857988962175e-05,
+      "loss": 3.3728,
+      "step": 6386
+    },
+    {
+      "epoch": 8.16768,
+      "grad_norm": 0.5407658219337463,
+      "learning_rate": 5.7908197603984386e-05,
+      "loss": 3.3447,
+      "step": 6387
+    },
+    {
+      "epoch": 8.16896,
+      "grad_norm": 0.5457347631454468,
+      "learning_rate": 5.7867815318347016e-05,
+      "loss": 3.3002,
+      "step": 6388
+    },
+    {
+      "epoch": 8.17024,
+      "grad_norm": 0.5405694842338562,
+      "learning_rate": 5.782743303270965e-05,
+      "loss": 3.2992,
+      "step": 6389
+    },
+    {
+      "epoch": 8.17152,
+      "grad_norm": 0.5365909934043884,
+      "learning_rate": 5.778705074707228e-05,
+      "loss": 3.3,
+      "step": 6390
+    },
+    {
+      "epoch": 8.1728,
+      "grad_norm": 0.5313602089881897,
+      "learning_rate": 5.7746668461434906e-05,
+      "loss": 3.2851,
+      "step": 6391
+    },
+    {
+      "epoch": 8.17408,
+      "grad_norm": 0.5397671461105347,
+      "learning_rate": 5.770628617579754e-05,
+      "loss": 3.3395,
+      "step": 6392
+    },
+    {
+      "epoch": 8.17536,
+      "grad_norm": 0.5422872304916382,
+      "learning_rate": 5.766590389016017e-05,
+      "loss": 3.2891,
+      "step": 6393
+    },
+    {
+      "epoch": 8.17664,
+      "grad_norm": 0.5259747505187988,
+      "learning_rate": 5.762552160452281e-05,
+      "loss": 3.2912,
+      "step": 6394
+    },
+    {
+      "epoch": 8.17792,
+      "grad_norm": 0.5186034440994263,
+      "learning_rate": 5.758513931888544e-05,
+      "loss": 3.2599,
+      "step": 6395
+    },
+    {
+      "epoch": 8.1792,
+      "grad_norm": 0.5253992080688477,
+      "learning_rate": 5.754475703324808e-05,
+      "loss": 3.2672,
+      "step": 6396
+    },
+    {
+      "epoch": 8.18048,
+      "grad_norm": 0.5303462743759155,
+      "learning_rate": 5.750437474761071e-05,
+      "loss": 3.3237,
+      "step": 6397
+    },
+    {
+      "epoch": 8.18176,
+      "grad_norm": 0.5276507139205933,
+      "learning_rate": 5.7463992461973344e-05,
+      "loss": 3.2787,
+      "step": 6398
+    },
+    {
+      "epoch": 8.18304,
+      "grad_norm": 0.5117890238761902,
+      "learning_rate": 5.7423610176335974e-05,
+      "loss": 3.2535,
+      "step": 6399
+    },
+    {
+      "epoch": 8.18432,
+      "grad_norm": 0.5390210151672363,
+      "learning_rate": 5.738322789069861e-05,
+      "loss": 3.2985,
+      "step": 6400
+    },
+    {
+      "epoch": 8.1856,
+      "grad_norm": 0.5502768754959106,
+      "learning_rate": 5.734284560506124e-05,
+      "loss": 3.2703,
+      "step": 6401
+    },
+    {
+      "epoch": 8.18688,
+      "grad_norm": 0.533101499080658,
+      "learning_rate": 5.730246331942388e-05,
+      "loss": 3.3369,
+      "step": 6402
+    },
+    {
+      "epoch": 8.18816,
+      "grad_norm": 0.5256611108779907,
+      "learning_rate": 5.726208103378651e-05,
+      "loss": 3.286,
+      "step": 6403
+    },
+    {
+      "epoch": 8.18944,
+      "grad_norm": 0.5344136357307434,
+      "learning_rate": 5.722169874814914e-05,
+      "loss": 3.2811,
+      "step": 6404
+    },
+    {
+      "epoch": 8.19072,
+      "grad_norm": 0.53043532371521,
+      "learning_rate": 5.718131646251177e-05,
+      "loss": 3.1944,
+      "step": 6405
+    },
+    {
+      "epoch": 8.192,
+      "grad_norm": 0.5280171036720276,
+      "learning_rate": 5.7140934176874405e-05,
+      "loss": 3.3055,
+      "step": 6406
+    },
+    {
+      "epoch": 8.19328,
+      "grad_norm": 0.5500734448432922,
+      "learning_rate": 5.7100551891237035e-05,
+      "loss": 3.3447,
+      "step": 6407
+    },
+    {
+      "epoch": 8.19456,
+      "grad_norm": 0.5249525308609009,
+      "learning_rate": 5.706016960559967e-05,
+      "loss": 3.2479,
+      "step": 6408
+    },
+    {
+      "epoch": 8.19584,
+      "grad_norm": 0.5362176895141602,
+      "learning_rate": 5.70197873199623e-05,
+      "loss": 3.3179,
+      "step": 6409
+    },
+    {
+      "epoch": 8.19712,
+      "grad_norm": 0.53925621509552,
+      "learning_rate": 5.697940503432494e-05,
+      "loss": 3.3497,
+      "step": 6410
+    },
+    {
+      "epoch": 8.1984,
+      "grad_norm": 0.5505580306053162,
+      "learning_rate": 5.693902274868757e-05,
+      "loss": 3.3724,
+      "step": 6411
+    },
+    {
+      "epoch": 8.19968,
+      "grad_norm": 0.5253210067749023,
+      "learning_rate": 5.6898640463050205e-05,
+      "loss": 3.2298,
+      "step": 6412
+    },
+    {
+      "epoch": 8.20096,
+      "grad_norm": 0.5256701707839966,
+      "learning_rate": 5.6858258177412836e-05,
+      "loss": 3.3066,
+      "step": 6413
+    },
+    {
+      "epoch": 8.20224,
+      "grad_norm": 0.5349451899528503,
+      "learning_rate": 5.681787589177547e-05,
+      "loss": 3.3191,
+      "step": 6414
+    },
+    {
+      "epoch": 8.20352,
+      "grad_norm": 0.5368797779083252,
+      "learning_rate": 5.67774936061381e-05,
+      "loss": 3.2995,
+      "step": 6415
+    },
+    {
+      "epoch": 8.2048,
+      "grad_norm": 0.5200009942054749,
+      "learning_rate": 5.673711132050074e-05,
+      "loss": 3.3066,
+      "step": 6416
+    },
+    {
+      "epoch": 8.20608,
+      "grad_norm": 0.5384882092475891,
+      "learning_rate": 5.669672903486337e-05,
+      "loss": 3.2356,
+      "step": 6417
+    },
+    {
+      "epoch": 8.20736,
+      "grad_norm": 0.5600325465202332,
+      "learning_rate": 5.6656346749226e-05,
+      "loss": 3.341,
+      "step": 6418
+    },
+    {
+      "epoch": 8.20864,
+      "grad_norm": 0.5298986434936523,
+      "learning_rate": 5.661596446358863e-05,
+      "loss": 3.278,
+      "step": 6419
+    },
+    {
+      "epoch": 8.20992,
+      "grad_norm": 0.5479633808135986,
+      "learning_rate": 5.6575582177951266e-05,
+      "loss": 3.313,
+      "step": 6420
+    },
+    {
+      "epoch": 8.2112,
+      "grad_norm": 0.5331804156303406,
+      "learning_rate": 5.6535199892313896e-05,
+      "loss": 3.2668,
+      "step": 6421
+    },
+    {
+      "epoch": 8.21248,
+      "grad_norm": 0.5298319458961487,
+      "learning_rate": 5.649481760667653e-05,
+      "loss": 3.2764,
+      "step": 6422
+    },
+    {
+      "epoch": 8.21376,
+      "grad_norm": 0.5214959383010864,
+      "learning_rate": 5.645443532103916e-05,
+      "loss": 3.2612,
+      "step": 6423
+    },
+    {
+      "epoch": 8.21504,
+      "grad_norm": 0.5243988633155823,
+      "learning_rate": 5.64140530354018e-05,
+      "loss": 3.2165,
+      "step": 6424
+    },
+    {
+      "epoch": 8.21632,
+      "grad_norm": 0.5409027934074402,
+      "learning_rate": 5.637367074976443e-05,
+      "loss": 3.3805,
+      "step": 6425
+    },
+    {
+      "epoch": 8.2176,
+      "grad_norm": 0.5333059430122375,
+      "learning_rate": 5.633328846412707e-05,
+      "loss": 3.3277,
+      "step": 6426
+    },
+    {
+      "epoch": 8.21888,
+      "grad_norm": 0.5418739914894104,
+      "learning_rate": 5.62929061784897e-05,
+      "loss": 3.3094,
+      "step": 6427
+    },
+    {
+      "epoch": 8.22016,
+      "grad_norm": 0.544068455696106,
+      "learning_rate": 5.6252523892852334e-05,
+      "loss": 3.3722,
+      "step": 6428
+    },
+    {
+      "epoch": 8.22144,
+      "grad_norm": 0.5321391820907593,
+      "learning_rate": 5.6212141607214964e-05,
+      "loss": 3.2523,
+      "step": 6429
+    },
+    {
+      "epoch": 8.22272,
+      "grad_norm": 0.538314163684845,
+      "learning_rate": 5.61717593215776e-05,
+      "loss": 3.3437,
+      "step": 6430
+    },
+    {
+      "epoch": 8.224,
+      "grad_norm": 0.519304096698761,
+      "learning_rate": 5.6131377035940224e-05,
+      "loss": 3.3433,
+      "step": 6431
+    },
+    {
+      "epoch": 8.22528,
+      "grad_norm": 0.5347556471824646,
+      "learning_rate": 5.609099475030286e-05,
+      "loss": 3.2909,
+      "step": 6432
+    },
+    {
+      "epoch": 8.22656,
+      "grad_norm": 0.5274844169616699,
+      "learning_rate": 5.605061246466549e-05,
+      "loss": 3.31,
+      "step": 6433
+    },
+    {
+      "epoch": 8.22784,
+      "grad_norm": 0.516024112701416,
+      "learning_rate": 5.601023017902813e-05,
+      "loss": 3.2947,
+      "step": 6434
+    },
+    {
+      "epoch": 8.22912,
+      "grad_norm": 0.5266866087913513,
+      "learning_rate": 5.596984789339076e-05,
+      "loss": 3.2854,
+      "step": 6435
+    },
+    {
+      "epoch": 8.2304,
+      "grad_norm": 0.537972629070282,
+      "learning_rate": 5.5929465607753395e-05,
+      "loss": 3.3274,
+      "step": 6436
+    },
+    {
+      "epoch": 8.23168,
+      "grad_norm": 0.5341406464576721,
+      "learning_rate": 5.5889083322116025e-05,
+      "loss": 3.2714,
+      "step": 6437
+    },
+    {
+      "epoch": 8.23296,
+      "grad_norm": 0.5305014252662659,
+      "learning_rate": 5.584870103647866e-05,
+      "loss": 3.2615,
+      "step": 6438
+    },
+    {
+      "epoch": 8.23424,
+      "grad_norm": 0.5244127511978149,
+      "learning_rate": 5.580831875084129e-05,
+      "loss": 3.328,
+      "step": 6439
+    },
+    {
+      "epoch": 8.23552,
+      "grad_norm": 0.54160475730896,
+      "learning_rate": 5.576793646520393e-05,
+      "loss": 3.258,
+      "step": 6440
+    },
+    {
+      "epoch": 8.2368,
+      "grad_norm": 0.5337183475494385,
+      "learning_rate": 5.572755417956656e-05,
+      "loss": 3.279,
+      "step": 6441
+    },
+    {
+      "epoch": 8.23808,
+      "grad_norm": 0.5248258709907532,
+      "learning_rate": 5.5687171893929196e-05,
+      "loss": 3.2938,
+      "step": 6442
+    },
+    {
+      "epoch": 8.23936,
+      "grad_norm": 0.5294501185417175,
+      "learning_rate": 5.5646789608291826e-05,
+      "loss": 3.2288,
+      "step": 6443
+    },
+    {
+      "epoch": 8.24064,
+      "grad_norm": 0.5414791703224182,
+      "learning_rate": 5.560640732265446e-05,
+      "loss": 3.2559,
+      "step": 6444
+    },
+    {
+      "epoch": 8.24192,
+      "grad_norm": 0.5217268466949463,
+      "learning_rate": 5.5566025037017086e-05,
+      "loss": 3.341,
+      "step": 6445
+    },
+    {
+      "epoch": 8.2432,
+      "grad_norm": 0.5463381409645081,
+      "learning_rate": 5.552564275137972e-05,
+      "loss": 3.3459,
+      "step": 6446
+    },
+    {
+      "epoch": 8.24448,
+      "grad_norm": 0.5450925827026367,
+      "learning_rate": 5.548526046574235e-05,
+      "loss": 3.287,
+      "step": 6447
+    },
+    {
+      "epoch": 8.24576,
+      "grad_norm": 0.5325472950935364,
+      "learning_rate": 5.544487818010499e-05,
+      "loss": 3.285,
+      "step": 6448
+    },
+    {
+      "epoch": 8.24704,
+      "grad_norm": 0.5504860877990723,
+      "learning_rate": 5.540449589446762e-05,
+      "loss": 3.3032,
+      "step": 6449
+    },
+    {
+      "epoch": 8.24832,
+      "grad_norm": 0.5341525077819824,
+      "learning_rate": 5.536411360883026e-05,
+      "loss": 3.2986,
+      "step": 6450
+    },
+    {
+      "epoch": 8.2496,
+      "grad_norm": 0.5327278971672058,
+      "learning_rate": 5.532373132319289e-05,
+      "loss": 3.2495,
+      "step": 6451
+    },
+    {
+      "epoch": 8.25088,
+      "grad_norm": 0.5406553745269775,
+      "learning_rate": 5.5283349037555524e-05,
+      "loss": 3.3444,
+      "step": 6452
+    },
+    {
+      "epoch": 8.25216,
+      "grad_norm": 0.5242081880569458,
+      "learning_rate": 5.5242966751918154e-05,
+      "loss": 3.264,
+      "step": 6453
+    },
+    {
+      "epoch": 8.25344,
+      "grad_norm": 0.534663200378418,
+      "learning_rate": 5.520258446628079e-05,
+      "loss": 3.3117,
+      "step": 6454
+    },
+    {
+      "epoch": 8.25472,
+      "grad_norm": 0.5145909190177917,
+      "learning_rate": 5.516220218064342e-05,
+      "loss": 3.255,
+      "step": 6455
+    },
+    {
+      "epoch": 8.256,
+      "grad_norm": 0.5177584886550903,
+      "learning_rate": 5.512181989500606e-05,
+      "loss": 3.2581,
+      "step": 6456
+    },
+    {
+      "epoch": 8.25728,
+      "grad_norm": 0.5454705357551575,
+      "learning_rate": 5.508143760936869e-05,
+      "loss": 3.2777,
+      "step": 6457
+    },
+    {
+      "epoch": 8.25856,
+      "grad_norm": 0.5337566137313843,
+      "learning_rate": 5.504105532373132e-05,
+      "loss": 3.2425,
+      "step": 6458
+    },
+    {
+      "epoch": 8.25984,
+      "grad_norm": 0.5299128293991089,
+      "learning_rate": 5.500067303809395e-05,
+      "loss": 3.3318,
+      "step": 6459
+    },
+    {
+      "epoch": 8.26112,
+      "grad_norm": 0.5477068424224854,
+      "learning_rate": 5.4960290752456585e-05,
+      "loss": 3.3498,
+      "step": 6460
+    },
+    {
+      "epoch": 8.2624,
+      "grad_norm": 0.5416424870491028,
+      "learning_rate": 5.4919908466819215e-05,
+      "loss": 3.3292,
+      "step": 6461
+    },
+    {
+      "epoch": 8.26368,
+      "grad_norm": 0.5232448577880859,
+      "learning_rate": 5.487952618118185e-05,
+      "loss": 3.282,
+      "step": 6462
+    },
+    {
+      "epoch": 8.26496,
+      "grad_norm": 0.5199217796325684,
+      "learning_rate": 5.483914389554448e-05,
+      "loss": 3.3211,
+      "step": 6463
+    },
+    {
+      "epoch": 8.26624,
+      "grad_norm": 0.5238966941833496,
+      "learning_rate": 5.479876160990712e-05,
+      "loss": 3.2954,
+      "step": 6464
+    },
+    {
+      "epoch": 8.26752,
+      "grad_norm": 0.5222386121749878,
+      "learning_rate": 5.475837932426975e-05,
+      "loss": 3.3141,
+      "step": 6465
+    },
+    {
+      "epoch": 8.2688,
+      "grad_norm": 0.5436253547668457,
+      "learning_rate": 5.4717997038632385e-05,
+      "loss": 3.3027,
+      "step": 6466
+    },
+    {
+      "epoch": 8.27008,
+      "grad_norm": 0.5344957709312439,
+      "learning_rate": 5.4677614752995015e-05,
+      "loss": 3.3603,
+      "step": 6467
+    },
+    {
+      "epoch": 8.27136,
+      "grad_norm": 0.5461180806159973,
+      "learning_rate": 5.463723246735765e-05,
+      "loss": 3.3247,
+      "step": 6468
+    },
+    {
+      "epoch": 8.272639999999999,
+      "grad_norm": 0.5331898927688599,
+      "learning_rate": 5.459685018172028e-05,
+      "loss": 3.2901,
+      "step": 6469
+    },
+    {
+      "epoch": 8.27392,
+      "grad_norm": 0.5247555375099182,
+      "learning_rate": 5.455646789608292e-05,
+      "loss": 3.2163,
+      "step": 6470
+    },
+    {
+      "epoch": 8.2752,
+      "grad_norm": 0.5188766121864319,
+      "learning_rate": 5.451608561044554e-05,
+      "loss": 3.2779,
+      "step": 6471
+    },
+    {
+      "epoch": 8.27648,
+      "grad_norm": 0.5405585169792175,
+      "learning_rate": 5.447570332480817e-05,
+      "loss": 3.2365,
+      "step": 6472
+    },
+    {
+      "epoch": 8.27776,
+      "grad_norm": 0.529931366443634,
+      "learning_rate": 5.443532103917081e-05,
+      "loss": 3.3514,
+      "step": 6473
+    },
+    {
+      "epoch": 8.27904,
+      "grad_norm": 0.5279687643051147,
+      "learning_rate": 5.4394938753533446e-05,
+      "loss": 3.2361,
+      "step": 6474
+    },
+    {
+      "epoch": 8.28032,
+      "grad_norm": 0.537344753742218,
+      "learning_rate": 5.4354556467896076e-05,
+      "loss": 3.2811,
+      "step": 6475
+    },
+    {
+      "epoch": 8.2816,
+      "grad_norm": 0.5340175628662109,
+      "learning_rate": 5.431417418225871e-05,
+      "loss": 3.2844,
+      "step": 6476
+    },
+    {
+      "epoch": 8.28288,
+      "grad_norm": 0.5298900604248047,
+      "learning_rate": 5.427379189662134e-05,
+      "loss": 3.3131,
+      "step": 6477
+    },
+    {
+      "epoch": 8.28416,
+      "grad_norm": 0.5220656394958496,
+      "learning_rate": 5.423340961098398e-05,
+      "loss": 3.2696,
+      "step": 6478
+    },
+    {
+      "epoch": 8.28544,
+      "grad_norm": 0.5230215191841125,
+      "learning_rate": 5.419302732534661e-05,
+      "loss": 3.2052,
+      "step": 6479
+    },
+    {
+      "epoch": 8.28672,
+      "grad_norm": 0.5259697437286377,
+      "learning_rate": 5.415264503970925e-05,
+      "loss": 3.2847,
+      "step": 6480
+    },
+    {
+      "epoch": 8.288,
+      "grad_norm": 0.5399904251098633,
+      "learning_rate": 5.411226275407188e-05,
+      "loss": 3.3553,
+      "step": 6481
+    },
+    {
+      "epoch": 8.28928,
+      "grad_norm": 0.5356093645095825,
+      "learning_rate": 5.4071880468434514e-05,
+      "loss": 3.2998,
+      "step": 6482
+    },
+    {
+      "epoch": 8.29056,
+      "grad_norm": 0.5491902232170105,
+      "learning_rate": 5.4031498182797144e-05,
+      "loss": 3.3246,
+      "step": 6483
+    },
+    {
+      "epoch": 8.29184,
+      "grad_norm": 0.5284157991409302,
+      "learning_rate": 5.399111589715978e-05,
+      "loss": 3.3414,
+      "step": 6484
+    },
+    {
+      "epoch": 8.29312,
+      "grad_norm": 0.5413424372673035,
+      "learning_rate": 5.3950733611522404e-05,
+      "loss": 3.3315,
+      "step": 6485
+    },
+    {
+      "epoch": 8.2944,
+      "grad_norm": 0.5393568873405457,
+      "learning_rate": 5.3910351325885034e-05,
+      "loss": 3.3165,
+      "step": 6486
+    },
+    {
+      "epoch": 8.29568,
+      "grad_norm": 0.5338280200958252,
+      "learning_rate": 5.386996904024767e-05,
+      "loss": 3.2115,
+      "step": 6487
+    },
+    {
+      "epoch": 8.29696,
+      "grad_norm": 0.5379251837730408,
+      "learning_rate": 5.38295867546103e-05,
+      "loss": 3.2415,
+      "step": 6488
+    },
+    {
+      "epoch": 8.29824,
+      "grad_norm": 0.5346460938453674,
+      "learning_rate": 5.378920446897294e-05,
+      "loss": 3.2432,
+      "step": 6489
+    },
+    {
+      "epoch": 8.29952,
+      "grad_norm": 0.5359485149383545,
+      "learning_rate": 5.374882218333557e-05,
+      "loss": 3.285,
+      "step": 6490
+    },
+    {
+      "epoch": 8.3008,
+      "grad_norm": 0.5321866869926453,
+      "learning_rate": 5.3708439897698205e-05,
+      "loss": 3.2818,
+      "step": 6491
+    },
+    {
+      "epoch": 8.30208,
+      "grad_norm": 0.5524401068687439,
+      "learning_rate": 5.366805761206084e-05,
+      "loss": 3.317,
+      "step": 6492
+    },
+    {
+      "epoch": 8.30336,
+      "grad_norm": 0.5479095578193665,
+      "learning_rate": 5.362767532642347e-05,
+      "loss": 3.3955,
+      "step": 6493
+    },
+    {
+      "epoch": 8.30464,
+      "grad_norm": 0.5439134836196899,
+      "learning_rate": 5.358729304078611e-05,
+      "loss": 3.3354,
+      "step": 6494
+    },
+    {
+      "epoch": 8.30592,
+      "grad_norm": 0.5380696058273315,
+      "learning_rate": 5.354691075514874e-05,
+      "loss": 3.2792,
+      "step": 6495
+    },
+    {
+      "epoch": 8.3072,
+      "grad_norm": 0.537322998046875,
+      "learning_rate": 5.3506528469511376e-05,
+      "loss": 3.3121,
+      "step": 6496
+    },
+    {
+      "epoch": 8.30848,
+      "grad_norm": 0.5562627911567688,
+      "learning_rate": 5.3466146183874006e-05,
+      "loss": 3.3261,
+      "step": 6497
+    },
+    {
+      "epoch": 8.30976,
+      "grad_norm": 0.5425490140914917,
+      "learning_rate": 5.342576389823663e-05,
+      "loss": 3.3206,
+      "step": 6498
+    },
+    {
+      "epoch": 8.31104,
+      "grad_norm": 0.5292763113975525,
+      "learning_rate": 5.3385381612599266e-05,
+      "loss": 3.283,
+      "step": 6499
+    },
+    {
+      "epoch": 8.31232,
+      "grad_norm": 0.5407996773719788,
+      "learning_rate": 5.3344999326961896e-05,
+      "loss": 3.2909,
+      "step": 6500
+    },
+    {
+      "epoch": 8.3136,
+      "grad_norm": 0.547214925289154,
+      "learning_rate": 5.330461704132453e-05,
+      "loss": 3.3512,
+      "step": 6501
+    },
+    {
+      "epoch": 8.31488,
+      "grad_norm": 0.5352652668952942,
+      "learning_rate": 5.326423475568716e-05,
+      "loss": 3.337,
+      "step": 6502
+    },
+    {
+      "epoch": 8.31616,
+      "grad_norm": 0.5354547500610352,
+      "learning_rate": 5.32238524700498e-05,
+      "loss": 3.2523,
+      "step": 6503
+    },
+    {
+      "epoch": 8.31744,
+      "grad_norm": 0.5254250168800354,
+      "learning_rate": 5.318347018441243e-05,
+      "loss": 3.3103,
+      "step": 6504
+    },
+    {
+      "epoch": 8.31872,
+      "grad_norm": 0.5346370935440063,
+      "learning_rate": 5.314308789877507e-05,
+      "loss": 3.2874,
+      "step": 6505
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.5400400161743164,
+      "learning_rate": 5.31027056131377e-05,
+      "loss": 3.3643,
+      "step": 6506
+    },
+    {
+      "epoch": 8.32128,
+      "grad_norm": 0.5396955609321594,
+      "learning_rate": 5.3062323327500334e-05,
+      "loss": 3.2664,
+      "step": 6507
+    },
+    {
+      "epoch": 8.32256,
+      "grad_norm": 0.5290980339050293,
+      "learning_rate": 5.3021941041862964e-05,
+      "loss": 3.2937,
+      "step": 6508
+    },
+    {
+      "epoch": 8.32384,
+      "grad_norm": 0.5446150898933411,
+      "learning_rate": 5.29815587562256e-05,
+      "loss": 3.3385,
+      "step": 6509
+    },
+    {
+      "epoch": 8.32512,
+      "grad_norm": 0.5492910146713257,
+      "learning_rate": 5.294117647058824e-05,
+      "loss": 3.2376,
+      "step": 6510
+    },
+    {
+      "epoch": 8.3264,
+      "grad_norm": 0.518017053604126,
+      "learning_rate": 5.290079418495087e-05,
+      "loss": 3.2401,
+      "step": 6511
+    },
+    {
+      "epoch": 8.32768,
+      "grad_norm": 0.527169406414032,
+      "learning_rate": 5.286041189931349e-05,
+      "loss": 3.2828,
+      "step": 6512
+    },
+    {
+      "epoch": 8.32896,
+      "grad_norm": 0.5413787364959717,
+      "learning_rate": 5.282002961367613e-05,
+      "loss": 3.338,
+      "step": 6513
+    },
+    {
+      "epoch": 8.33024,
+      "grad_norm": 0.5491598844528198,
+      "learning_rate": 5.277964732803876e-05,
+      "loss": 3.3348,
+      "step": 6514
+    },
+    {
+      "epoch": 8.33152,
+      "grad_norm": 0.529373049736023,
+      "learning_rate": 5.2739265042401395e-05,
+      "loss": 3.3055,
+      "step": 6515
+    },
+    {
+      "epoch": 8.3328,
+      "grad_norm": 0.5360579490661621,
+      "learning_rate": 5.2698882756764025e-05,
+      "loss": 3.2354,
+      "step": 6516
+    },
+    {
+      "epoch": 8.33408,
+      "grad_norm": 0.5181642770767212,
+      "learning_rate": 5.265850047112666e-05,
+      "loss": 3.2713,
+      "step": 6517
+    },
+    {
+      "epoch": 8.33536,
+      "grad_norm": 0.5401854515075684,
+      "learning_rate": 5.261811818548929e-05,
+      "loss": 3.314,
+      "step": 6518
+    },
+    {
+      "epoch": 8.33664,
+      "grad_norm": 0.5461652874946594,
+      "learning_rate": 5.257773589985193e-05,
+      "loss": 3.2553,
+      "step": 6519
+    },
+    {
+      "epoch": 8.33792,
+      "grad_norm": 0.5387153625488281,
+      "learning_rate": 5.253735361421456e-05,
+      "loss": 3.3327,
+      "step": 6520
+    },
+    {
+      "epoch": 8.3392,
+      "grad_norm": 0.5197851657867432,
+      "learning_rate": 5.2496971328577195e-05,
+      "loss": 3.3562,
+      "step": 6521
+    },
+    {
+      "epoch": 8.34048,
+      "grad_norm": 0.5335418581962585,
+      "learning_rate": 5.2456589042939825e-05,
+      "loss": 3.3083,
+      "step": 6522
+    },
+    {
+      "epoch": 8.34176,
+      "grad_norm": 0.5208085775375366,
+      "learning_rate": 5.241620675730246e-05,
+      "loss": 3.2347,
+      "step": 6523
+    },
+    {
+      "epoch": 8.34304,
+      "grad_norm": 0.5384852290153503,
+      "learning_rate": 5.237582447166509e-05,
+      "loss": 3.285,
+      "step": 6524
+    },
+    {
+      "epoch": 8.34432,
+      "grad_norm": 0.5363668203353882,
+      "learning_rate": 5.233544218602772e-05,
+      "loss": 3.3247,
+      "step": 6525
+    },
+    {
+      "epoch": 8.3456,
+      "grad_norm": 0.5379889011383057,
+      "learning_rate": 5.229505990039035e-05,
+      "loss": 3.3292,
+      "step": 6526
+    },
+    {
+      "epoch": 8.34688,
+      "grad_norm": 0.5510496497154236,
+      "learning_rate": 5.225467761475299e-05,
+      "loss": 3.3561,
+      "step": 6527
+    },
+    {
+      "epoch": 8.34816,
+      "grad_norm": 0.5368717908859253,
+      "learning_rate": 5.221429532911562e-05,
+      "loss": 3.2615,
+      "step": 6528
+    },
+    {
+      "epoch": 8.34944,
+      "grad_norm": 0.5422154068946838,
+      "learning_rate": 5.2173913043478256e-05,
+      "loss": 3.2918,
+      "step": 6529
+    },
+    {
+      "epoch": 8.35072,
+      "grad_norm": 0.5283968448638916,
+      "learning_rate": 5.2133530757840886e-05,
+      "loss": 3.2817,
+      "step": 6530
+    },
+    {
+      "epoch": 8.352,
+      "grad_norm": 0.5295470952987671,
+      "learning_rate": 5.209314847220352e-05,
+      "loss": 3.2432,
+      "step": 6531
+    },
+    {
+      "epoch": 8.35328,
+      "grad_norm": 0.5349634885787964,
+      "learning_rate": 5.205276618656615e-05,
+      "loss": 3.2913,
+      "step": 6532
+    },
+    {
+      "epoch": 8.35456,
+      "grad_norm": 0.5370112657546997,
+      "learning_rate": 5.201238390092879e-05,
+      "loss": 3.2895,
+      "step": 6533
+    },
+    {
+      "epoch": 8.35584,
+      "grad_norm": 0.5439594984054565,
+      "learning_rate": 5.197200161529142e-05,
+      "loss": 3.3035,
+      "step": 6534
+    },
+    {
+      "epoch": 8.35712,
+      "grad_norm": 0.5383702516555786,
+      "learning_rate": 5.193161932965406e-05,
+      "loss": 3.3121,
+      "step": 6535
+    },
+    {
+      "epoch": 8.3584,
+      "grad_norm": 0.5429950952529907,
+      "learning_rate": 5.189123704401669e-05,
+      "loss": 3.3598,
+      "step": 6536
+    },
+    {
+      "epoch": 8.35968,
+      "grad_norm": 0.5377165675163269,
+      "learning_rate": 5.1850854758379324e-05,
+      "loss": 3.2763,
+      "step": 6537
+    },
+    {
+      "epoch": 8.36096,
+      "grad_norm": 0.5333889126777649,
+      "learning_rate": 5.1810472472741954e-05,
+      "loss": 3.2872,
+      "step": 6538
+    },
+    {
+      "epoch": 8.36224,
+      "grad_norm": 0.5367612242698669,
+      "learning_rate": 5.1770090187104584e-05,
+      "loss": 3.2694,
+      "step": 6539
+    },
+    {
+      "epoch": 8.36352,
+      "grad_norm": 0.5359439253807068,
+      "learning_rate": 5.1729707901467214e-05,
+      "loss": 3.2822,
+      "step": 6540
+    },
+    {
+      "epoch": 8.3648,
+      "grad_norm": 0.531587541103363,
+      "learning_rate": 5.168932561582985e-05,
+      "loss": 3.3343,
+      "step": 6541
+    },
+    {
+      "epoch": 8.36608,
+      "grad_norm": 0.5479581952095032,
+      "learning_rate": 5.164894333019248e-05,
+      "loss": 3.2928,
+      "step": 6542
+    },
+    {
+      "epoch": 8.36736,
+      "grad_norm": 0.5197001099586487,
+      "learning_rate": 5.160856104455512e-05,
+      "loss": 3.3011,
+      "step": 6543
+    },
+    {
+      "epoch": 8.36864,
+      "grad_norm": 0.5361149311065674,
+      "learning_rate": 5.156817875891775e-05,
+      "loss": 3.2673,
+      "step": 6544
+    },
+    {
+      "epoch": 8.36992,
+      "grad_norm": 0.5395957827568054,
+      "learning_rate": 5.1527796473280385e-05,
+      "loss": 3.3285,
+      "step": 6545
+    },
+    {
+      "epoch": 8.3712,
+      "grad_norm": 0.538285493850708,
+      "learning_rate": 5.1487414187643015e-05,
+      "loss": 3.3397,
+      "step": 6546
+    },
+    {
+      "epoch": 8.37248,
+      "grad_norm": 0.5274959206581116,
+      "learning_rate": 5.144703190200565e-05,
+      "loss": 3.2938,
+      "step": 6547
+    },
+    {
+      "epoch": 8.37376,
+      "grad_norm": 0.5292668342590332,
+      "learning_rate": 5.140664961636828e-05,
+      "loss": 3.2236,
+      "step": 6548
+    },
+    {
+      "epoch": 8.37504,
+      "grad_norm": 0.5355281233787537,
+      "learning_rate": 5.136626733073092e-05,
+      "loss": 3.3204,
+      "step": 6549
+    },
+    {
+      "epoch": 8.37632,
+      "grad_norm": 0.5475792288780212,
+      "learning_rate": 5.132588504509355e-05,
+      "loss": 3.2975,
+      "step": 6550
+    },
+    {
+      "epoch": 8.3776,
+      "grad_norm": 0.5429788827896118,
+      "learning_rate": 5.1285502759456186e-05,
+      "loss": 3.2851,
+      "step": 6551
+    },
+    {
+      "epoch": 8.37888,
+      "grad_norm": 0.5370329022407532,
+      "learning_rate": 5.124512047381881e-05,
+      "loss": 3.32,
+      "step": 6552
+    },
+    {
+      "epoch": 8.38016,
+      "grad_norm": 0.5378831028938293,
+      "learning_rate": 5.1204738188181446e-05,
+      "loss": 3.2567,
+      "step": 6553
+    },
+    {
+      "epoch": 8.38144,
+      "grad_norm": 0.5425311326980591,
+      "learning_rate": 5.1164355902544076e-05,
+      "loss": 3.2856,
+      "step": 6554
+    },
+    {
+      "epoch": 8.38272,
+      "grad_norm": 0.5385202765464783,
+      "learning_rate": 5.112397361690671e-05,
+      "loss": 3.3203,
+      "step": 6555
+    },
+    {
+      "epoch": 8.384,
+      "grad_norm": 0.5249854922294617,
+      "learning_rate": 5.108359133126934e-05,
+      "loss": 3.2775,
+      "step": 6556
+    },
+    {
+      "epoch": 8.38528,
+      "grad_norm": 0.5497921705245972,
+      "learning_rate": 5.104320904563198e-05,
+      "loss": 3.375,
+      "step": 6557
+    },
+    {
+      "epoch": 8.38656,
+      "grad_norm": 0.5290823578834534,
+      "learning_rate": 5.100282675999461e-05,
+      "loss": 3.4186,
+      "step": 6558
+    },
+    {
+      "epoch": 8.38784,
+      "grad_norm": 0.5338512659072876,
+      "learning_rate": 5.096244447435725e-05,
+      "loss": 3.3006,
+      "step": 6559
+    },
+    {
+      "epoch": 8.38912,
+      "grad_norm": 0.5281364321708679,
+      "learning_rate": 5.092206218871988e-05,
+      "loss": 3.3595,
+      "step": 6560
+    },
+    {
+      "epoch": 8.3904,
+      "grad_norm": 0.5322358012199402,
+      "learning_rate": 5.0881679903082514e-05,
+      "loss": 3.2731,
+      "step": 6561
+    },
+    {
+      "epoch": 8.39168,
+      "grad_norm": 0.5335200428962708,
+      "learning_rate": 5.0841297617445144e-05,
+      "loss": 3.2838,
+      "step": 6562
+    },
+    {
+      "epoch": 8.39296,
+      "grad_norm": 0.5297821760177612,
+      "learning_rate": 5.080091533180778e-05,
+      "loss": 3.3354,
+      "step": 6563
+    },
+    {
+      "epoch": 8.39424,
+      "grad_norm": 0.5368463397026062,
+      "learning_rate": 5.076053304617041e-05,
+      "loss": 3.3192,
+      "step": 6564
+    },
+    {
+      "epoch": 8.39552,
+      "grad_norm": 0.5247287154197693,
+      "learning_rate": 5.072015076053304e-05,
+      "loss": 3.3062,
+      "step": 6565
+    },
+    {
+      "epoch": 8.3968,
+      "grad_norm": 0.540341317653656,
+      "learning_rate": 5.067976847489567e-05,
+      "loss": 3.3337,
+      "step": 6566
+    },
+    {
+      "epoch": 8.39808,
+      "grad_norm": 0.5336794853210449,
+      "learning_rate": 5.063938618925831e-05,
+      "loss": 3.2744,
+      "step": 6567
+    },
+    {
+      "epoch": 8.39936,
+      "grad_norm": 0.5233598351478577,
+      "learning_rate": 5.059900390362094e-05,
+      "loss": 3.2749,
+      "step": 6568
+    },
+    {
+      "epoch": 8.40064,
+      "grad_norm": 0.5248898267745972,
+      "learning_rate": 5.0558621617983575e-05,
+      "loss": 3.2279,
+      "step": 6569
+    },
+    {
+      "epoch": 8.40192,
+      "grad_norm": 0.5336484313011169,
+      "learning_rate": 5.0518239332346205e-05,
+      "loss": 3.2535,
+      "step": 6570
+    },
+    {
+      "epoch": 8.4032,
+      "grad_norm": 0.550398051738739,
+      "learning_rate": 5.047785704670884e-05,
+      "loss": 3.3059,
+      "step": 6571
+    },
+    {
+      "epoch": 8.40448,
+      "grad_norm": 0.5399869084358215,
+      "learning_rate": 5.043747476107147e-05,
+      "loss": 3.3524,
+      "step": 6572
+    },
+    {
+      "epoch": 8.40576,
+      "grad_norm": 0.5220796465873718,
+      "learning_rate": 5.039709247543411e-05,
+      "loss": 3.2817,
+      "step": 6573
+    },
+    {
+      "epoch": 8.40704,
+      "grad_norm": 0.5332768559455872,
+      "learning_rate": 5.035671018979674e-05,
+      "loss": 3.2817,
+      "step": 6574
+    },
+    {
+      "epoch": 8.40832,
+      "grad_norm": 0.5253673195838928,
+      "learning_rate": 5.0316327904159375e-05,
+      "loss": 3.2993,
+      "step": 6575
+    },
+    {
+      "epoch": 8.4096,
+      "grad_norm": 0.5454734563827515,
+      "learning_rate": 5.0275945618522005e-05,
+      "loss": 3.2808,
+      "step": 6576
+    },
+    {
+      "epoch": 8.41088,
+      "grad_norm": 0.5498039126396179,
+      "learning_rate": 5.023556333288464e-05,
+      "loss": 3.3277,
+      "step": 6577
+    },
+    {
+      "epoch": 8.41216,
+      "grad_norm": 0.52730792760849,
+      "learning_rate": 5.019518104724727e-05,
+      "loss": 3.2521,
+      "step": 6578
+    },
+    {
+      "epoch": 8.41344,
+      "grad_norm": 0.5146779417991638,
+      "learning_rate": 5.0154798761609896e-05,
+      "loss": 3.3197,
+      "step": 6579
+    },
+    {
+      "epoch": 8.414719999999999,
+      "grad_norm": 0.5284165740013123,
+      "learning_rate": 5.011441647597253e-05,
+      "loss": 3.3388,
+      "step": 6580
+    },
+    {
+      "epoch": 8.416,
+      "grad_norm": 0.5253511667251587,
+      "learning_rate": 5.007403419033517e-05,
+      "loss": 3.324,
+      "step": 6581
+    },
+    {
+      "epoch": 8.41728,
+      "grad_norm": 0.53730309009552,
+      "learning_rate": 5.00336519046978e-05,
+      "loss": 3.3254,
+      "step": 6582
+    },
+    {
+      "epoch": 8.41856,
+      "grad_norm": 0.5318365693092346,
+      "learning_rate": 4.9993269619060436e-05,
+      "loss": 3.3386,
+      "step": 6583
+    },
+    {
+      "epoch": 8.41984,
+      "grad_norm": 0.522120475769043,
+      "learning_rate": 4.9952887333423066e-05,
+      "loss": 3.2756,
+      "step": 6584
+    },
+    {
+      "epoch": 8.42112,
+      "grad_norm": 0.5341548919677734,
+      "learning_rate": 4.99125050477857e-05,
+      "loss": 3.2815,
+      "step": 6585
+    },
+    {
+      "epoch": 8.4224,
+      "grad_norm": 0.5254530310630798,
+      "learning_rate": 4.987212276214833e-05,
+      "loss": 3.3148,
+      "step": 6586
+    },
+    {
+      "epoch": 8.42368,
+      "grad_norm": 0.516562283039093,
+      "learning_rate": 4.983174047651097e-05,
+      "loss": 3.3143,
+      "step": 6587
+    },
+    {
+      "epoch": 8.42496,
+      "grad_norm": 0.5405133366584778,
+      "learning_rate": 4.97913581908736e-05,
+      "loss": 3.2954,
+      "step": 6588
+    },
+    {
+      "epoch": 8.42624,
+      "grad_norm": 0.5392136573791504,
+      "learning_rate": 4.975097590523624e-05,
+      "loss": 3.2129,
+      "step": 6589
+    },
+    {
+      "epoch": 8.42752,
+      "grad_norm": 0.5335177779197693,
+      "learning_rate": 4.971059361959887e-05,
+      "loss": 3.2511,
+      "step": 6590
+    },
+    {
+      "epoch": 8.4288,
+      "grad_norm": 0.5130221843719482,
+      "learning_rate": 4.9670211333961504e-05,
+      "loss": 3.2803,
+      "step": 6591
+    },
+    {
+      "epoch": 8.43008,
+      "grad_norm": 0.5377788543701172,
+      "learning_rate": 4.962982904832413e-05,
+      "loss": 3.3044,
+      "step": 6592
+    },
+    {
+      "epoch": 8.43136,
+      "grad_norm": 0.535243034362793,
+      "learning_rate": 4.958944676268676e-05,
+      "loss": 3.3071,
+      "step": 6593
+    },
+    {
+      "epoch": 8.43264,
+      "grad_norm": 0.5345697999000549,
+      "learning_rate": 4.9549064477049394e-05,
+      "loss": 3.3025,
+      "step": 6594
+    },
+    {
+      "epoch": 8.43392,
+      "grad_norm": 0.5292974710464478,
+      "learning_rate": 4.9508682191412024e-05,
+      "loss": 3.3006,
+      "step": 6595
+    },
+    {
+      "epoch": 8.4352,
+      "grad_norm": 0.5351077318191528,
+      "learning_rate": 4.946829990577466e-05,
+      "loss": 3.3633,
+      "step": 6596
+    },
+    {
+      "epoch": 8.43648,
+      "grad_norm": 0.546981155872345,
+      "learning_rate": 4.942791762013729e-05,
+      "loss": 3.3898,
+      "step": 6597
+    },
+    {
+      "epoch": 8.43776,
+      "grad_norm": 0.5353581309318542,
+      "learning_rate": 4.938753533449993e-05,
+      "loss": 3.2787,
+      "step": 6598
+    },
+    {
+      "epoch": 8.43904,
+      "grad_norm": 0.5259009003639221,
+      "learning_rate": 4.9347153048862565e-05,
+      "loss": 3.204,
+      "step": 6599
+    },
+    {
+      "epoch": 8.44032,
+      "grad_norm": 0.5294379591941833,
+      "learning_rate": 4.9306770763225195e-05,
+      "loss": 3.3074,
+      "step": 6600
+    },
+    {
+      "epoch": 8.4416,
+      "grad_norm": 0.5439087748527527,
+      "learning_rate": 4.926638847758783e-05,
+      "loss": 3.2858,
+      "step": 6601
+    },
+    {
+      "epoch": 8.44288,
+      "grad_norm": 0.5535233020782471,
+      "learning_rate": 4.922600619195046e-05,
+      "loss": 3.2854,
+      "step": 6602
+    },
+    {
+      "epoch": 8.44416,
+      "grad_norm": 0.5227919220924377,
+      "learning_rate": 4.91856239063131e-05,
+      "loss": 3.2956,
+      "step": 6603
+    },
+    {
+      "epoch": 8.44544,
+      "grad_norm": 0.525883674621582,
+      "learning_rate": 4.914524162067573e-05,
+      "loss": 3.3417,
+      "step": 6604
+    },
+    {
+      "epoch": 8.44672,
+      "grad_norm": 0.5234944224357605,
+      "learning_rate": 4.9104859335038366e-05,
+      "loss": 3.2156,
+      "step": 6605
+    },
+    {
+      "epoch": 8.448,
+      "grad_norm": 0.5246995091438293,
+      "learning_rate": 4.906447704940099e-05,
+      "loss": 3.2844,
+      "step": 6606
+    },
+    {
+      "epoch": 8.44928,
+      "grad_norm": 0.5292709469795227,
+      "learning_rate": 4.902409476376362e-05,
+      "loss": 3.3131,
+      "step": 6607
+    },
+    {
+      "epoch": 8.45056,
+      "grad_norm": 0.5411702394485474,
+      "learning_rate": 4.8983712478126256e-05,
+      "loss": 3.2209,
+      "step": 6608
+    },
+    {
+      "epoch": 8.45184,
+      "grad_norm": 0.5438616871833801,
+      "learning_rate": 4.8943330192488886e-05,
+      "loss": 3.3588,
+      "step": 6609
+    },
+    {
+      "epoch": 8.45312,
+      "grad_norm": 0.5124935507774353,
+      "learning_rate": 4.890294790685152e-05,
+      "loss": 3.2722,
+      "step": 6610
+    },
+    {
+      "epoch": 8.4544,
+      "grad_norm": 0.5373281240463257,
+      "learning_rate": 4.886256562121415e-05,
+      "loss": 3.2317,
+      "step": 6611
+    },
+    {
+      "epoch": 8.45568,
+      "grad_norm": 0.5373220443725586,
+      "learning_rate": 4.882218333557679e-05,
+      "loss": 3.2904,
+      "step": 6612
+    },
+    {
+      "epoch": 8.45696,
+      "grad_norm": 0.526901125907898,
+      "learning_rate": 4.878180104993942e-05,
+      "loss": 3.3668,
+      "step": 6613
+    },
+    {
+      "epoch": 8.45824,
+      "grad_norm": 0.5479453802108765,
+      "learning_rate": 4.874141876430206e-05,
+      "loss": 3.3688,
+      "step": 6614
+    },
+    {
+      "epoch": 8.45952,
+      "grad_norm": 0.5343961715698242,
+      "learning_rate": 4.870103647866469e-05,
+      "loss": 3.3287,
+      "step": 6615
+    },
+    {
+      "epoch": 8.4608,
+      "grad_norm": 0.5295287370681763,
+      "learning_rate": 4.8660654193027324e-05,
+      "loss": 3.3605,
+      "step": 6616
+    },
+    {
+      "epoch": 8.46208,
+      "grad_norm": 0.5407337546348572,
+      "learning_rate": 4.862027190738996e-05,
+      "loss": 3.2565,
+      "step": 6617
+    },
+    {
+      "epoch": 8.46336,
+      "grad_norm": 0.5319713950157166,
+      "learning_rate": 4.857988962175259e-05,
+      "loss": 3.2649,
+      "step": 6618
+    },
+    {
+      "epoch": 8.46464,
+      "grad_norm": 0.5339657664299011,
+      "learning_rate": 4.8539507336115214e-05,
+      "loss": 3.2722,
+      "step": 6619
+    },
+    {
+      "epoch": 8.46592,
+      "grad_norm": 0.5170288681983948,
+      "learning_rate": 4.849912505047785e-05,
+      "loss": 3.2732,
+      "step": 6620
+    },
+    {
+      "epoch": 8.4672,
+      "grad_norm": 0.5304208397865295,
+      "learning_rate": 4.845874276484048e-05,
+      "loss": 3.2746,
+      "step": 6621
+    },
+    {
+      "epoch": 8.46848,
+      "grad_norm": 0.5211367011070251,
+      "learning_rate": 4.841836047920312e-05,
+      "loss": 3.2367,
+      "step": 6622
+    },
+    {
+      "epoch": 8.46976,
+      "grad_norm": 0.5377680659294128,
+      "learning_rate": 4.837797819356575e-05,
+      "loss": 3.3002,
+      "step": 6623
+    },
+    {
+      "epoch": 8.47104,
+      "grad_norm": 0.532772421836853,
+      "learning_rate": 4.8337595907928385e-05,
+      "loss": 3.3249,
+      "step": 6624
+    },
+    {
+      "epoch": 8.47232,
+      "grad_norm": 0.5326812863349915,
+      "learning_rate": 4.8297213622291015e-05,
+      "loss": 3.277,
+      "step": 6625
+    },
+    {
+      "epoch": 8.4736,
+      "grad_norm": 0.532048761844635,
+      "learning_rate": 4.825683133665365e-05,
+      "loss": 3.3765,
+      "step": 6626
+    },
+    {
+      "epoch": 8.47488,
+      "grad_norm": 0.5287336707115173,
+      "learning_rate": 4.821644905101628e-05,
+      "loss": 3.3689,
+      "step": 6627
+    },
+    {
+      "epoch": 8.47616,
+      "grad_norm": 0.5333081483840942,
+      "learning_rate": 4.817606676537892e-05,
+      "loss": 3.3058,
+      "step": 6628
+    },
+    {
+      "epoch": 8.47744,
+      "grad_norm": 0.5384047031402588,
+      "learning_rate": 4.813568447974155e-05,
+      "loss": 3.2469,
+      "step": 6629
+    },
+    {
+      "epoch": 8.47872,
+      "grad_norm": 0.5469292998313904,
+      "learning_rate": 4.8095302194104185e-05,
+      "loss": 3.2969,
+      "step": 6630
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.5489159226417542,
+      "learning_rate": 4.8054919908466815e-05,
+      "loss": 3.3041,
+      "step": 6631
+    },
+    {
+      "epoch": 8.48128,
+      "grad_norm": 0.5341904759407043,
+      "learning_rate": 4.801453762282945e-05,
+      "loss": 3.2527,
+      "step": 6632
+    },
+    {
+      "epoch": 8.48256,
+      "grad_norm": 0.5368700623512268,
+      "learning_rate": 4.7974155337192076e-05,
+      "loss": 3.2925,
+      "step": 6633
+    },
+    {
+      "epoch": 8.48384,
+      "grad_norm": 0.5526073575019836,
+      "learning_rate": 4.793377305155471e-05,
+      "loss": 3.2979,
+      "step": 6634
+    },
+    {
+      "epoch": 8.48512,
+      "grad_norm": 0.5394701361656189,
+      "learning_rate": 4.789339076591734e-05,
+      "loss": 3.3142,
+      "step": 6635
+    },
+    {
+      "epoch": 8.4864,
+      "grad_norm": 0.5406348705291748,
+      "learning_rate": 4.785300848027998e-05,
+      "loss": 3.277,
+      "step": 6636
+    },
+    {
+      "epoch": 8.48768,
+      "grad_norm": 0.5281654000282288,
+      "learning_rate": 4.781262619464261e-05,
+      "loss": 3.2689,
+      "step": 6637
+    },
+    {
+      "epoch": 8.48896,
+      "grad_norm": 0.519383430480957,
+      "learning_rate": 4.7772243909005246e-05,
+      "loss": 3.2823,
+      "step": 6638
+    },
+    {
+      "epoch": 8.49024,
+      "grad_norm": 0.5329931974411011,
+      "learning_rate": 4.7731861623367876e-05,
+      "loss": 3.3037,
+      "step": 6639
+    },
+    {
+      "epoch": 8.49152,
+      "grad_norm": 0.5432103872299194,
+      "learning_rate": 4.769147933773051e-05,
+      "loss": 3.3509,
+      "step": 6640
+    },
+    {
+      "epoch": 8.4928,
+      "grad_norm": 0.523330807685852,
+      "learning_rate": 4.765109705209314e-05,
+      "loss": 3.3401,
+      "step": 6641
+    },
+    {
+      "epoch": 8.49408,
+      "grad_norm": 0.5295400619506836,
+      "learning_rate": 4.761071476645578e-05,
+      "loss": 3.2433,
+      "step": 6642
+    },
+    {
+      "epoch": 8.49536,
+      "grad_norm": 0.5371250510215759,
+      "learning_rate": 4.757033248081841e-05,
+      "loss": 3.2771,
+      "step": 6643
+    },
+    {
+      "epoch": 8.49664,
+      "grad_norm": 0.5461635589599609,
+      "learning_rate": 4.752995019518105e-05,
+      "loss": 3.3759,
+      "step": 6644
+    },
+    {
+      "epoch": 8.49792,
+      "grad_norm": 0.5435198545455933,
+      "learning_rate": 4.748956790954368e-05,
+      "loss": 3.343,
+      "step": 6645
+    },
+    {
+      "epoch": 8.4992,
+      "grad_norm": 0.5389981269836426,
+      "learning_rate": 4.744918562390631e-05,
+      "loss": 3.2387,
+      "step": 6646
+    },
+    {
+      "epoch": 8.50048,
+      "grad_norm": 0.5361154675483704,
+      "learning_rate": 4.740880333826894e-05,
+      "loss": 3.312,
+      "step": 6647
+    },
+    {
+      "epoch": 8.50176,
+      "grad_norm": 0.5247445702552795,
+      "learning_rate": 4.7368421052631574e-05,
+      "loss": 3.2876,
+      "step": 6648
+    },
+    {
+      "epoch": 8.50304,
+      "grad_norm": 0.5444439053535461,
+      "learning_rate": 4.7328038766994204e-05,
+      "loss": 3.2557,
+      "step": 6649
+    },
+    {
+      "epoch": 8.50432,
+      "grad_norm": 0.549291729927063,
+      "learning_rate": 4.728765648135684e-05,
+      "loss": 3.291,
+      "step": 6650
+    },
+    {
+      "epoch": 8.5056,
+      "grad_norm": 0.5521546602249146,
+      "learning_rate": 4.724727419571947e-05,
+      "loss": 3.3253,
+      "step": 6651
+    },
+    {
+      "epoch": 8.50688,
+      "grad_norm": 0.5204415917396545,
+      "learning_rate": 4.720689191008211e-05,
+      "loss": 3.2887,
+      "step": 6652
+    },
+    {
+      "epoch": 8.50816,
+      "grad_norm": 0.5449312329292297,
+      "learning_rate": 4.716650962444474e-05,
+      "loss": 3.2632,
+      "step": 6653
+    },
+    {
+      "epoch": 8.50944,
+      "grad_norm": 0.558323323726654,
+      "learning_rate": 4.7126127338807375e-05,
+      "loss": 3.356,
+      "step": 6654
+    },
+    {
+      "epoch": 8.51072,
+      "grad_norm": 0.5528385639190674,
+      "learning_rate": 4.7085745053170005e-05,
+      "loss": 3.3238,
+      "step": 6655
+    },
+    {
+      "epoch": 8.512,
+      "grad_norm": 0.5442092418670654,
+      "learning_rate": 4.704536276753264e-05,
+      "loss": 3.1951,
+      "step": 6656
+    },
+    {
+      "epoch": 8.51328,
+      "grad_norm": 0.5273060202598572,
+      "learning_rate": 4.700498048189527e-05,
+      "loss": 3.2714,
+      "step": 6657
+    },
+    {
+      "epoch": 8.51456,
+      "grad_norm": 0.5554311871528625,
+      "learning_rate": 4.696459819625791e-05,
+      "loss": 3.3199,
+      "step": 6658
+    },
+    {
+      "epoch": 8.51584,
+      "grad_norm": 0.540524423122406,
+      "learning_rate": 4.692421591062053e-05,
+      "loss": 3.3171,
+      "step": 6659
+    },
+    {
+      "epoch": 8.51712,
+      "grad_norm": 0.543006420135498,
+      "learning_rate": 4.688383362498317e-05,
+      "loss": 3.3168,
+      "step": 6660
+    },
+    {
+      "epoch": 8.5184,
+      "grad_norm": 0.552527666091919,
+      "learning_rate": 4.68434513393458e-05,
+      "loss": 3.3671,
+      "step": 6661
+    },
+    {
+      "epoch": 8.51968,
+      "grad_norm": 0.5287879109382629,
+      "learning_rate": 4.6803069053708436e-05,
+      "loss": 3.2603,
+      "step": 6662
+    },
+    {
+      "epoch": 8.52096,
+      "grad_norm": 0.5465594530105591,
+      "learning_rate": 4.6762686768071066e-05,
+      "loss": 3.287,
+      "step": 6663
+    },
+    {
+      "epoch": 8.52224,
+      "grad_norm": 0.5492598414421082,
+      "learning_rate": 4.67223044824337e-05,
+      "loss": 3.299,
+      "step": 6664
+    },
+    {
+      "epoch": 8.52352,
+      "grad_norm": 0.534814178943634,
+      "learning_rate": 4.668192219679633e-05,
+      "loss": 3.2982,
+      "step": 6665
+    },
+    {
+      "epoch": 8.5248,
+      "grad_norm": 0.5302172303199768,
+      "learning_rate": 4.664153991115897e-05,
+      "loss": 3.2304,
+      "step": 6666
+    },
+    {
+      "epoch": 8.52608,
+      "grad_norm": 0.5542375445365906,
+      "learning_rate": 4.66011576255216e-05,
+      "loss": 3.3307,
+      "step": 6667
+    },
+    {
+      "epoch": 8.52736,
+      "grad_norm": 0.5232375860214233,
+      "learning_rate": 4.6560775339884237e-05,
+      "loss": 3.2925,
+      "step": 6668
+    },
+    {
+      "epoch": 8.52864,
+      "grad_norm": 0.5362772941589355,
+      "learning_rate": 4.652039305424687e-05,
+      "loss": 3.3613,
+      "step": 6669
+    },
+    {
+      "epoch": 8.52992,
+      "grad_norm": 0.5412434935569763,
+      "learning_rate": 4.6480010768609504e-05,
+      "loss": 3.3056,
+      "step": 6670
+    },
+    {
+      "epoch": 8.5312,
+      "grad_norm": 0.5264190435409546,
+      "learning_rate": 4.6439628482972134e-05,
+      "loss": 3.2676,
+      "step": 6671
+    },
+    {
+      "epoch": 8.53248,
+      "grad_norm": 0.5494565367698669,
+      "learning_rate": 4.639924619733477e-05,
+      "loss": 3.3138,
+      "step": 6672
+    },
+    {
+      "epoch": 8.533760000000001,
+      "grad_norm": 0.5212265253067017,
+      "learning_rate": 4.6358863911697394e-05,
+      "loss": 3.3393,
+      "step": 6673
+    },
+    {
+      "epoch": 8.53504,
+      "grad_norm": 0.5425857901573181,
+      "learning_rate": 4.631848162606003e-05,
+      "loss": 3.3338,
+      "step": 6674
+    },
+    {
+      "epoch": 8.53632,
+      "grad_norm": 0.5471029877662659,
+      "learning_rate": 4.627809934042266e-05,
+      "loss": 3.2664,
+      "step": 6675
+    },
+    {
+      "epoch": 8.5376,
+      "grad_norm": 0.5544648170471191,
+      "learning_rate": 4.62377170547853e-05,
+      "loss": 3.2952,
+      "step": 6676
+    },
+    {
+      "epoch": 8.53888,
+      "grad_norm": 0.5421332716941833,
+      "learning_rate": 4.619733476914793e-05,
+      "loss": 3.3276,
+      "step": 6677
+    },
+    {
+      "epoch": 8.54016,
+      "grad_norm": 0.5286380052566528,
+      "learning_rate": 4.6156952483510564e-05,
+      "loss": 3.2644,
+      "step": 6678
+    },
+    {
+      "epoch": 8.54144,
+      "grad_norm": 0.5330486297607422,
+      "learning_rate": 4.6116570197873195e-05,
+      "loss": 3.3331,
+      "step": 6679
+    },
+    {
+      "epoch": 8.54272,
+      "grad_norm": 0.5347878932952881,
+      "learning_rate": 4.607618791223583e-05,
+      "loss": 3.2441,
+      "step": 6680
+    },
+    {
+      "epoch": 8.544,
+      "grad_norm": 0.5383884906768799,
+      "learning_rate": 4.603580562659846e-05,
+      "loss": 3.3088,
+      "step": 6681
+    },
+    {
+      "epoch": 8.54528,
+      "grad_norm": 0.5470399856567383,
+      "learning_rate": 4.59954233409611e-05,
+      "loss": 3.2866,
+      "step": 6682
+    },
+    {
+      "epoch": 8.54656,
+      "grad_norm": 0.5385159254074097,
+      "learning_rate": 4.595504105532373e-05,
+      "loss": 3.248,
+      "step": 6683
+    },
+    {
+      "epoch": 8.54784,
+      "grad_norm": 0.5333342552185059,
+      "learning_rate": 4.5914658769686365e-05,
+      "loss": 3.3123,
+      "step": 6684
+    },
+    {
+      "epoch": 8.54912,
+      "grad_norm": 0.5450199842453003,
+      "learning_rate": 4.5874276484048995e-05,
+      "loss": 3.2946,
+      "step": 6685
+    },
+    {
+      "epoch": 8.5504,
+      "grad_norm": 0.552905797958374,
+      "learning_rate": 4.5833894198411625e-05,
+      "loss": 3.3255,
+      "step": 6686
+    },
+    {
+      "epoch": 8.55168,
+      "grad_norm": 0.5411385297775269,
+      "learning_rate": 4.5793511912774255e-05,
+      "loss": 3.3116,
+      "step": 6687
+    },
+    {
+      "epoch": 8.55296,
+      "grad_norm": 0.5458307266235352,
+      "learning_rate": 4.575312962713689e-05,
+      "loss": 3.3065,
+      "step": 6688
+    },
+    {
+      "epoch": 8.55424,
+      "grad_norm": 0.526465892791748,
+      "learning_rate": 4.571274734149952e-05,
+      "loss": 3.3649,
+      "step": 6689
+    },
+    {
+      "epoch": 8.55552,
+      "grad_norm": 0.5384424924850464,
+      "learning_rate": 4.567236505586216e-05,
+      "loss": 3.3591,
+      "step": 6690
+    },
+    {
+      "epoch": 8.556799999999999,
+      "grad_norm": 0.5338189601898193,
+      "learning_rate": 4.563198277022479e-05,
+      "loss": 3.2857,
+      "step": 6691
+    },
+    {
+      "epoch": 8.55808,
+      "grad_norm": 0.5296444296836853,
+      "learning_rate": 4.5591600484587426e-05,
+      "loss": 3.2774,
+      "step": 6692
+    },
+    {
+      "epoch": 8.55936,
+      "grad_norm": 0.5314897894859314,
+      "learning_rate": 4.5551218198950056e-05,
+      "loss": 3.3251,
+      "step": 6693
+    },
+    {
+      "epoch": 8.56064,
+      "grad_norm": 0.5258769989013672,
+      "learning_rate": 4.551083591331269e-05,
+      "loss": 3.285,
+      "step": 6694
+    },
+    {
+      "epoch": 8.56192,
+      "grad_norm": 0.5228273868560791,
+      "learning_rate": 4.547045362767532e-05,
+      "loss": 3.327,
+      "step": 6695
+    },
+    {
+      "epoch": 8.5632,
+      "grad_norm": 0.5333476662635803,
+      "learning_rate": 4.543007134203796e-05,
+      "loss": 3.2938,
+      "step": 6696
+    },
+    {
+      "epoch": 8.56448,
+      "grad_norm": 0.5421110987663269,
+      "learning_rate": 4.538968905640059e-05,
+      "loss": 3.3763,
+      "step": 6697
+    },
+    {
+      "epoch": 8.565760000000001,
+      "grad_norm": 0.526771605014801,
+      "learning_rate": 4.534930677076323e-05,
+      "loss": 3.2587,
+      "step": 6698
+    },
+    {
+      "epoch": 8.56704,
+      "grad_norm": 0.5357040762901306,
+      "learning_rate": 4.530892448512586e-05,
+      "loss": 3.3598,
+      "step": 6699
+    },
+    {
+      "epoch": 8.56832,
+      "grad_norm": 0.5350444316864014,
+      "learning_rate": 4.526854219948848e-05,
+      "loss": 3.2902,
+      "step": 6700
+    },
+    {
+      "epoch": 8.5696,
+      "grad_norm": 0.524205207824707,
+      "learning_rate": 4.522815991385112e-05,
+      "loss": 3.3145,
+      "step": 6701
+    },
+    {
+      "epoch": 8.57088,
+      "grad_norm": 0.5205404758453369,
+      "learning_rate": 4.518777762821375e-05,
+      "loss": 3.2893,
+      "step": 6702
+    },
+    {
+      "epoch": 8.57216,
+      "grad_norm": 0.5430803894996643,
+      "learning_rate": 4.5147395342576384e-05,
+      "loss": 3.3511,
+      "step": 6703
+    },
+    {
+      "epoch": 8.57344,
+      "grad_norm": 0.5332580804824829,
+      "learning_rate": 4.510701305693902e-05,
+      "loss": 3.3326,
+      "step": 6704
+    },
+    {
+      "epoch": 8.57472,
+      "grad_norm": 0.5290676951408386,
+      "learning_rate": 4.506663077130165e-05,
+      "loss": 3.2984,
+      "step": 6705
+    },
+    {
+      "epoch": 8.576,
+      "grad_norm": 0.5440507531166077,
+      "learning_rate": 4.502624848566429e-05,
+      "loss": 3.3492,
+      "step": 6706
+    },
+    {
+      "epoch": 8.57728,
+      "grad_norm": 0.5364778637886047,
+      "learning_rate": 4.498586620002692e-05,
+      "loss": 3.2265,
+      "step": 6707
+    },
+    {
+      "epoch": 8.57856,
+      "grad_norm": 0.5364824533462524,
+      "learning_rate": 4.4945483914389555e-05,
+      "loss": 3.291,
+      "step": 6708
+    },
+    {
+      "epoch": 8.57984,
+      "grad_norm": 0.5298408269882202,
+      "learning_rate": 4.4905101628752185e-05,
+      "loss": 3.3376,
+      "step": 6709
+    },
+    {
+      "epoch": 8.58112,
+      "grad_norm": 0.5235030651092529,
+      "learning_rate": 4.486471934311482e-05,
+      "loss": 3.3014,
+      "step": 6710
+    },
+    {
+      "epoch": 8.5824,
+      "grad_norm": 0.5457448959350586,
+      "learning_rate": 4.482433705747745e-05,
+      "loss": 3.3132,
+      "step": 6711
+    },
+    {
+      "epoch": 8.58368,
+      "grad_norm": 0.5244759321212769,
+      "learning_rate": 4.478395477184009e-05,
+      "loss": 3.3322,
+      "step": 6712
+    },
+    {
+      "epoch": 8.58496,
+      "grad_norm": 0.5247620344161987,
+      "learning_rate": 4.474357248620271e-05,
+      "loss": 3.3435,
+      "step": 6713
+    },
+    {
+      "epoch": 8.58624,
+      "grad_norm": 0.5359330177307129,
+      "learning_rate": 4.470319020056534e-05,
+      "loss": 3.2807,
+      "step": 6714
+    },
+    {
+      "epoch": 8.58752,
+      "grad_norm": 0.5311152338981628,
+      "learning_rate": 4.466280791492798e-05,
+      "loss": 3.2728,
+      "step": 6715
+    },
+    {
+      "epoch": 8.588799999999999,
+      "grad_norm": 0.5377166867256165,
+      "learning_rate": 4.462242562929061e-05,
+      "loss": 3.2665,
+      "step": 6716
+    },
+    {
+      "epoch": 8.59008,
+      "grad_norm": 0.5189564824104309,
+      "learning_rate": 4.4582043343653246e-05,
+      "loss": 3.2582,
+      "step": 6717
+    },
+    {
+      "epoch": 8.59136,
+      "grad_norm": 0.5309081673622131,
+      "learning_rate": 4.4541661058015876e-05,
+      "loss": 3.3129,
+      "step": 6718
+    },
+    {
+      "epoch": 8.59264,
+      "grad_norm": 0.536917507648468,
+      "learning_rate": 4.450127877237851e-05,
+      "loss": 3.3365,
+      "step": 6719
+    },
+    {
+      "epoch": 8.59392,
+      "grad_norm": 0.5495189428329468,
+      "learning_rate": 4.446089648674114e-05,
+      "loss": 3.2733,
+      "step": 6720
+    },
+    {
+      "epoch": 8.5952,
+      "grad_norm": 0.5371974110603333,
+      "learning_rate": 4.442051420110378e-05,
+      "loss": 3.2837,
+      "step": 6721
+    },
+    {
+      "epoch": 8.59648,
+      "grad_norm": 0.5370466113090515,
+      "learning_rate": 4.438013191546641e-05,
+      "loss": 3.3062,
+      "step": 6722
+    },
+    {
+      "epoch": 8.59776,
+      "grad_norm": 0.5342411398887634,
+      "learning_rate": 4.4339749629829047e-05,
+      "loss": 3.3036,
+      "step": 6723
+    },
+    {
+      "epoch": 8.59904,
+      "grad_norm": 0.5245928764343262,
+      "learning_rate": 4.4299367344191683e-05,
+      "loss": 3.2969,
+      "step": 6724
+    },
+    {
+      "epoch": 8.60032,
+      "grad_norm": 0.5341728329658508,
+      "learning_rate": 4.4258985058554314e-05,
+      "loss": 3.3438,
+      "step": 6725
+    },
+    {
+      "epoch": 8.6016,
+      "grad_norm": 0.5377516150474548,
+      "learning_rate": 4.421860277291695e-05,
+      "loss": 3.2733,
+      "step": 6726
+    },
+    {
+      "epoch": 8.60288,
+      "grad_norm": 0.5328512191772461,
+      "learning_rate": 4.4178220487279574e-05,
+      "loss": 3.3902,
+      "step": 6727
+    },
+    {
+      "epoch": 8.60416,
+      "grad_norm": 0.5404960513114929,
+      "learning_rate": 4.4137838201642204e-05,
+      "loss": 3.3356,
+      "step": 6728
+    },
+    {
+      "epoch": 8.60544,
+      "grad_norm": 0.5246291160583496,
+      "learning_rate": 4.409745591600484e-05,
+      "loss": 3.2186,
+      "step": 6729
+    },
+    {
+      "epoch": 8.60672,
+      "grad_norm": 0.5350708961486816,
+      "learning_rate": 4.405707363036747e-05,
+      "loss": 3.3441,
+      "step": 6730
+    },
+    {
+      "epoch": 8.608,
+      "grad_norm": 0.525844395160675,
+      "learning_rate": 4.401669134473011e-05,
+      "loss": 3.3156,
+      "step": 6731
+    },
+    {
+      "epoch": 8.60928,
+      "grad_norm": 0.5407573580741882,
+      "learning_rate": 4.397630905909274e-05,
+      "loss": 3.2902,
+      "step": 6732
+    },
+    {
+      "epoch": 8.61056,
+      "grad_norm": 0.5487748384475708,
+      "learning_rate": 4.3935926773455374e-05,
+      "loss": 3.2765,
+      "step": 6733
+    },
+    {
+      "epoch": 8.61184,
+      "grad_norm": 0.56423020362854,
+      "learning_rate": 4.3895544487818005e-05,
+      "loss": 3.2918,
+      "step": 6734
+    },
+    {
+      "epoch": 8.61312,
+      "grad_norm": 0.5294039249420166,
+      "learning_rate": 4.385516220218064e-05,
+      "loss": 3.3598,
+      "step": 6735
+    },
+    {
+      "epoch": 8.6144,
+      "grad_norm": 0.5347636342048645,
+      "learning_rate": 4.381477991654327e-05,
+      "loss": 3.2215,
+      "step": 6736
+    },
+    {
+      "epoch": 8.61568,
+      "grad_norm": 0.5291707515716553,
+      "learning_rate": 4.377439763090591e-05,
+      "loss": 3.3315,
+      "step": 6737
+    },
+    {
+      "epoch": 8.61696,
+      "grad_norm": 0.5520625710487366,
+      "learning_rate": 4.373401534526854e-05,
+      "loss": 3.3297,
+      "step": 6738
+    },
+    {
+      "epoch": 8.61824,
+      "grad_norm": 0.5564367771148682,
+      "learning_rate": 4.3693633059631175e-05,
+      "loss": 3.3697,
+      "step": 6739
+    },
+    {
+      "epoch": 8.61952,
+      "grad_norm": 0.5352851748466492,
+      "learning_rate": 4.36532507739938e-05,
+      "loss": 3.322,
+      "step": 6740
+    },
+    {
+      "epoch": 8.6208,
+      "grad_norm": 0.5492033362388611,
+      "learning_rate": 4.3612868488356435e-05,
+      "loss": 3.3269,
+      "step": 6741
+    },
+    {
+      "epoch": 8.62208,
+      "grad_norm": 0.5320399403572083,
+      "learning_rate": 4.3572486202719065e-05,
+      "loss": 3.3004,
+      "step": 6742
+    },
+    {
+      "epoch": 8.62336,
+      "grad_norm": 0.5452033877372742,
+      "learning_rate": 4.35321039170817e-05,
+      "loss": 3.3551,
+      "step": 6743
+    },
+    {
+      "epoch": 8.62464,
+      "grad_norm": 0.5418961644172668,
+      "learning_rate": 4.349172163144433e-05,
+      "loss": 3.3384,
+      "step": 6744
+    },
+    {
+      "epoch": 8.62592,
+      "grad_norm": 0.5382152199745178,
+      "learning_rate": 4.345133934580697e-05,
+      "loss": 3.3212,
+      "step": 6745
+    },
+    {
+      "epoch": 8.6272,
+      "grad_norm": 0.5336072444915771,
+      "learning_rate": 4.34109570601696e-05,
+      "loss": 3.2777,
+      "step": 6746
+    },
+    {
+      "epoch": 8.62848,
+      "grad_norm": 0.5362292528152466,
+      "learning_rate": 4.3370574774532236e-05,
+      "loss": 3.3193,
+      "step": 6747
+    },
+    {
+      "epoch": 8.62976,
+      "grad_norm": 0.5334807634353638,
+      "learning_rate": 4.3330192488894866e-05,
+      "loss": 3.2778,
+      "step": 6748
+    },
+    {
+      "epoch": 8.63104,
+      "grad_norm": 0.5414783954620361,
+      "learning_rate": 4.32898102032575e-05,
+      "loss": 3.3341,
+      "step": 6749
+    },
+    {
+      "epoch": 8.63232,
+      "grad_norm": 0.5455512404441833,
+      "learning_rate": 4.324942791762013e-05,
+      "loss": 3.2904,
+      "step": 6750
+    },
+    {
+      "epoch": 8.6336,
+      "grad_norm": 0.5280025601387024,
+      "learning_rate": 4.320904563198277e-05,
+      "loss": 3.2243,
+      "step": 6751
+    },
+    {
+      "epoch": 8.63488,
+      "grad_norm": 0.5390767455101013,
+      "learning_rate": 4.31686633463454e-05,
+      "loss": 3.3436,
+      "step": 6752
+    },
+    {
+      "epoch": 8.63616,
+      "grad_norm": 0.5390846133232117,
+      "learning_rate": 4.312828106070803e-05,
+      "loss": 3.261,
+      "step": 6753
+    },
+    {
+      "epoch": 8.63744,
+      "grad_norm": 0.5264686942100525,
+      "learning_rate": 4.308789877507066e-05,
+      "loss": 3.3319,
+      "step": 6754
+    },
+    {
+      "epoch": 8.63872,
+      "grad_norm": 0.5262269377708435,
+      "learning_rate": 4.30475164894333e-05,
+      "loss": 3.2829,
+      "step": 6755
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.5335661768913269,
+      "learning_rate": 4.300713420379593e-05,
+      "loss": 3.2751,
+      "step": 6756
+    },
+    {
+      "epoch": 8.64128,
+      "grad_norm": 0.5311907529830933,
+      "learning_rate": 4.2966751918158564e-05,
+      "loss": 3.3261,
+      "step": 6757
+    },
+    {
+      "epoch": 8.64256,
+      "grad_norm": 0.5352111458778381,
+      "learning_rate": 4.2926369632521194e-05,
+      "loss": 3.2754,
+      "step": 6758
+    },
+    {
+      "epoch": 8.64384,
+      "grad_norm": 0.5307966470718384,
+      "learning_rate": 4.288598734688383e-05,
+      "loss": 3.3116,
+      "step": 6759
+    },
+    {
+      "epoch": 8.64512,
+      "grad_norm": 0.536453127861023,
+      "learning_rate": 4.284560506124646e-05,
+      "loss": 3.3126,
+      "step": 6760
+    },
+    {
+      "epoch": 8.6464,
+      "grad_norm": 0.5201568007469177,
+      "learning_rate": 4.28052227756091e-05,
+      "loss": 3.2846,
+      "step": 6761
+    },
+    {
+      "epoch": 8.64768,
+      "grad_norm": 0.5415171980857849,
+      "learning_rate": 4.276484048997173e-05,
+      "loss": 3.3514,
+      "step": 6762
+    },
+    {
+      "epoch": 8.64896,
+      "grad_norm": 0.5305605530738831,
+      "learning_rate": 4.2724458204334365e-05,
+      "loss": 3.2887,
+      "step": 6763
+    },
+    {
+      "epoch": 8.65024,
+      "grad_norm": 0.5250943303108215,
+      "learning_rate": 4.2684075918696995e-05,
+      "loss": 3.3334,
+      "step": 6764
+    },
+    {
+      "epoch": 8.65152,
+      "grad_norm": 0.5324791073799133,
+      "learning_rate": 4.264369363305963e-05,
+      "loss": 3.246,
+      "step": 6765
+    },
+    {
+      "epoch": 8.6528,
+      "grad_norm": 0.5431379675865173,
+      "learning_rate": 4.260331134742226e-05,
+      "loss": 3.2801,
+      "step": 6766
+    },
+    {
+      "epoch": 8.65408,
+      "grad_norm": 0.5285992622375488,
+      "learning_rate": 4.256292906178489e-05,
+      "loss": 3.2725,
+      "step": 6767
+    },
+    {
+      "epoch": 8.65536,
+      "grad_norm": 0.5373889803886414,
+      "learning_rate": 4.252254677614752e-05,
+      "loss": 3.3643,
+      "step": 6768
+    },
+    {
+      "epoch": 8.65664,
+      "grad_norm": 0.5431947708129883,
+      "learning_rate": 4.248216449051016e-05,
+      "loss": 3.3618,
+      "step": 6769
+    },
+    {
+      "epoch": 8.65792,
+      "grad_norm": 0.5260449647903442,
+      "learning_rate": 4.244178220487279e-05,
+      "loss": 3.265,
+      "step": 6770
+    },
+    {
+      "epoch": 8.6592,
+      "grad_norm": 0.5484868288040161,
+      "learning_rate": 4.2401399919235426e-05,
+      "loss": 3.3274,
+      "step": 6771
+    },
+    {
+      "epoch": 8.66048,
+      "grad_norm": 0.5380119681358337,
+      "learning_rate": 4.2361017633598056e-05,
+      "loss": 3.2953,
+      "step": 6772
+    },
+    {
+      "epoch": 8.66176,
+      "grad_norm": 0.539348840713501,
+      "learning_rate": 4.232063534796069e-05,
+      "loss": 3.2966,
+      "step": 6773
+    },
+    {
+      "epoch": 8.66304,
+      "grad_norm": 0.528852641582489,
+      "learning_rate": 4.228025306232332e-05,
+      "loss": 3.3094,
+      "step": 6774
+    },
+    {
+      "epoch": 8.66432,
+      "grad_norm": 0.5262014269828796,
+      "learning_rate": 4.223987077668596e-05,
+      "loss": 3.237,
+      "step": 6775
+    },
+    {
+      "epoch": 8.6656,
+      "grad_norm": 0.5200825333595276,
+      "learning_rate": 4.219948849104859e-05,
+      "loss": 3.2816,
+      "step": 6776
+    },
+    {
+      "epoch": 8.66688,
+      "grad_norm": 0.5474440455436707,
+      "learning_rate": 4.2159106205411227e-05,
+      "loss": 3.3406,
+      "step": 6777
+    },
+    {
+      "epoch": 8.66816,
+      "grad_norm": 0.5352291464805603,
+      "learning_rate": 4.2118723919773857e-05,
+      "loss": 3.279,
+      "step": 6778
+    },
+    {
+      "epoch": 8.66944,
+      "grad_norm": 0.5264368653297424,
+      "learning_rate": 4.2078341634136493e-05,
+      "loss": 3.2559,
+      "step": 6779
+    },
+    {
+      "epoch": 8.67072,
+      "grad_norm": 0.5307362675666809,
+      "learning_rate": 4.203795934849912e-05,
+      "loss": 3.2877,
+      "step": 6780
+    },
+    {
+      "epoch": 8.672,
+      "grad_norm": 0.5339285135269165,
+      "learning_rate": 4.1997577062861754e-05,
+      "loss": 3.3039,
+      "step": 6781
+    },
+    {
+      "epoch": 8.67328,
+      "grad_norm": 0.5298980474472046,
+      "learning_rate": 4.1957194777224384e-05,
+      "loss": 3.2824,
+      "step": 6782
+    },
+    {
+      "epoch": 8.67456,
+      "grad_norm": 0.5294161438941956,
+      "learning_rate": 4.191681249158702e-05,
+      "loss": 3.3326,
+      "step": 6783
+    },
+    {
+      "epoch": 8.67584,
+      "grad_norm": 0.5197745561599731,
+      "learning_rate": 4.187643020594965e-05,
+      "loss": 3.2917,
+      "step": 6784
+    },
+    {
+      "epoch": 8.67712,
+      "grad_norm": 0.5530399680137634,
+      "learning_rate": 4.183604792031229e-05,
+      "loss": 3.2526,
+      "step": 6785
+    },
+    {
+      "epoch": 8.6784,
+      "grad_norm": 0.534653902053833,
+      "learning_rate": 4.179566563467492e-05,
+      "loss": 3.3122,
+      "step": 6786
+    },
+    {
+      "epoch": 8.67968,
+      "grad_norm": 0.5447145700454712,
+      "learning_rate": 4.1755283349037554e-05,
+      "loss": 3.3016,
+      "step": 6787
+    },
+    {
+      "epoch": 8.68096,
+      "grad_norm": 0.5226910710334778,
+      "learning_rate": 4.1714901063400184e-05,
+      "loss": 3.2551,
+      "step": 6788
+    },
+    {
+      "epoch": 8.68224,
+      "grad_norm": 0.5207378268241882,
+      "learning_rate": 4.167451877776282e-05,
+      "loss": 3.247,
+      "step": 6789
+    },
+    {
+      "epoch": 8.68352,
+      "grad_norm": 0.5464200973510742,
+      "learning_rate": 4.163413649212545e-05,
+      "loss": 3.2968,
+      "step": 6790
+    },
+    {
+      "epoch": 8.6848,
+      "grad_norm": 0.5588473081588745,
+      "learning_rate": 4.159375420648809e-05,
+      "loss": 3.301,
+      "step": 6791
+    },
+    {
+      "epoch": 8.68608,
+      "grad_norm": 0.5364888906478882,
+      "learning_rate": 4.155337192085072e-05,
+      "loss": 3.2832,
+      "step": 6792
+    },
+    {
+      "epoch": 8.68736,
+      "grad_norm": 0.5507677793502808,
+      "learning_rate": 4.1512989635213355e-05,
+      "loss": 3.3312,
+      "step": 6793
+    },
+    {
+      "epoch": 8.68864,
+      "grad_norm": 0.5295298099517822,
+      "learning_rate": 4.147260734957598e-05,
+      "loss": 3.3066,
+      "step": 6794
+    },
+    {
+      "epoch": 8.68992,
+      "grad_norm": 0.5314168334007263,
+      "learning_rate": 4.1432225063938615e-05,
+      "loss": 3.2516,
+      "step": 6795
+    },
+    {
+      "epoch": 8.6912,
+      "grad_norm": 0.5450182557106018,
+      "learning_rate": 4.1391842778301245e-05,
+      "loss": 3.309,
+      "step": 6796
+    },
+    {
+      "epoch": 8.69248,
+      "grad_norm": 0.5385010242462158,
+      "learning_rate": 4.135146049266388e-05,
+      "loss": 3.1977,
+      "step": 6797
+    },
+    {
+      "epoch": 8.69376,
+      "grad_norm": 0.5317445397377014,
+      "learning_rate": 4.131107820702651e-05,
+      "loss": 3.2691,
+      "step": 6798
+    },
+    {
+      "epoch": 8.69504,
+      "grad_norm": 0.536864697933197,
+      "learning_rate": 4.127069592138915e-05,
+      "loss": 3.2888,
+      "step": 6799
+    },
+    {
+      "epoch": 8.69632,
+      "grad_norm": 0.5189259648323059,
+      "learning_rate": 4.123031363575178e-05,
+      "loss": 3.2824,
+      "step": 6800
+    },
+    {
+      "epoch": 8.6976,
+      "grad_norm": 0.5167062282562256,
+      "learning_rate": 4.1189931350114416e-05,
+      "loss": 3.2385,
+      "step": 6801
+    },
+    {
+      "epoch": 8.698879999999999,
+      "grad_norm": 0.5420131683349609,
+      "learning_rate": 4.1149549064477046e-05,
+      "loss": 3.4164,
+      "step": 6802
+    },
+    {
+      "epoch": 8.70016,
+      "grad_norm": 0.533211886882782,
+      "learning_rate": 4.110916677883968e-05,
+      "loss": 3.2774,
+      "step": 6803
+    },
+    {
+      "epoch": 8.70144,
+      "grad_norm": 0.5351250767707825,
+      "learning_rate": 4.106878449320231e-05,
+      "loss": 3.3219,
+      "step": 6804
+    },
+    {
+      "epoch": 8.70272,
+      "grad_norm": 0.5335774421691895,
+      "learning_rate": 4.102840220756495e-05,
+      "loss": 3.272,
+      "step": 6805
+    },
+    {
+      "epoch": 8.704,
+      "grad_norm": 0.5470417737960815,
+      "learning_rate": 4.098801992192758e-05,
+      "loss": 3.265,
+      "step": 6806
+    },
+    {
+      "epoch": 8.70528,
+      "grad_norm": 0.5419219732284546,
+      "learning_rate": 4.09476376362902e-05,
+      "loss": 3.3382,
+      "step": 6807
+    },
+    {
+      "epoch": 8.70656,
+      "grad_norm": 0.5278633832931519,
+      "learning_rate": 4.090725535065284e-05,
+      "loss": 3.341,
+      "step": 6808
+    },
+    {
+      "epoch": 8.707840000000001,
+      "grad_norm": 0.5203990936279297,
+      "learning_rate": 4.086687306501547e-05,
+      "loss": 3.3018,
+      "step": 6809
+    },
+    {
+      "epoch": 8.70912,
+      "grad_norm": 0.530105710029602,
+      "learning_rate": 4.082649077937811e-05,
+      "loss": 3.3272,
+      "step": 6810
+    },
+    {
+      "epoch": 8.7104,
+      "grad_norm": 0.5211985111236572,
+      "learning_rate": 4.0786108493740744e-05,
+      "loss": 3.2695,
+      "step": 6811
+    },
+    {
+      "epoch": 8.71168,
+      "grad_norm": 0.5373910665512085,
+      "learning_rate": 4.0745726208103374e-05,
+      "loss": 3.3031,
+      "step": 6812
+    },
+    {
+      "epoch": 8.71296,
+      "grad_norm": 0.5470989942550659,
+      "learning_rate": 4.070534392246601e-05,
+      "loss": 3.3386,
+      "step": 6813
+    },
+    {
+      "epoch": 8.71424,
+      "grad_norm": 0.5321961045265198,
+      "learning_rate": 4.066496163682864e-05,
+      "loss": 3.2882,
+      "step": 6814
+    },
+    {
+      "epoch": 8.71552,
+      "grad_norm": 0.5434589385986328,
+      "learning_rate": 4.062457935119128e-05,
+      "loss": 3.3023,
+      "step": 6815
+    },
+    {
+      "epoch": 8.7168,
+      "grad_norm": 0.5319420099258423,
+      "learning_rate": 4.058419706555391e-05,
+      "loss": 3.3207,
+      "step": 6816
+    },
+    {
+      "epoch": 8.71808,
+      "grad_norm": 0.5350143313407898,
+      "learning_rate": 4.0543814779916545e-05,
+      "loss": 3.3401,
+      "step": 6817
+    },
+    {
+      "epoch": 8.71936,
+      "grad_norm": 0.5368631482124329,
+      "learning_rate": 4.0503432494279175e-05,
+      "loss": 3.2471,
+      "step": 6818
+    },
+    {
+      "epoch": 8.72064,
+      "grad_norm": 0.541717529296875,
+      "learning_rate": 4.046305020864181e-05,
+      "loss": 3.3548,
+      "step": 6819
+    },
+    {
+      "epoch": 8.72192,
+      "grad_norm": 0.5320764183998108,
+      "learning_rate": 4.042266792300444e-05,
+      "loss": 3.2743,
+      "step": 6820
+    },
+    {
+      "epoch": 8.7232,
+      "grad_norm": 0.536125659942627,
+      "learning_rate": 4.0382285637367065e-05,
+      "loss": 3.3071,
+      "step": 6821
+    },
+    {
+      "epoch": 8.72448,
+      "grad_norm": 0.5400855541229248,
+      "learning_rate": 4.03419033517297e-05,
+      "loss": 3.3063,
+      "step": 6822
+    },
+    {
+      "epoch": 8.72576,
+      "grad_norm": 0.5443190336227417,
+      "learning_rate": 4.030152106609233e-05,
+      "loss": 3.3044,
+      "step": 6823
+    },
+    {
+      "epoch": 8.72704,
+      "grad_norm": 0.5323216319084167,
+      "learning_rate": 4.026113878045497e-05,
+      "loss": 3.2328,
+      "step": 6824
+    },
+    {
+      "epoch": 8.72832,
+      "grad_norm": 0.5174821615219116,
+      "learning_rate": 4.02207564948176e-05,
+      "loss": 3.3146,
+      "step": 6825
+    },
+    {
+      "epoch": 8.7296,
+      "grad_norm": 0.5285962224006653,
+      "learning_rate": 4.0180374209180236e-05,
+      "loss": 3.3673,
+      "step": 6826
+    },
+    {
+      "epoch": 8.730879999999999,
+      "grad_norm": 0.525948703289032,
+      "learning_rate": 4.0139991923542866e-05,
+      "loss": 3.3524,
+      "step": 6827
+    },
+    {
+      "epoch": 8.73216,
+      "grad_norm": 0.526118814945221,
+      "learning_rate": 4.00996096379055e-05,
+      "loss": 3.3361,
+      "step": 6828
+    },
+    {
+      "epoch": 8.73344,
+      "grad_norm": 0.5256155729293823,
+      "learning_rate": 4.005922735226814e-05,
+      "loss": 3.35,
+      "step": 6829
+    },
+    {
+      "epoch": 8.73472,
+      "grad_norm": 0.5380839705467224,
+      "learning_rate": 4.001884506663077e-05,
+      "loss": 3.3115,
+      "step": 6830
+    },
+    {
+      "epoch": 8.736,
+      "grad_norm": 0.5296849012374878,
+      "learning_rate": 3.9978462780993406e-05,
+      "loss": 3.2831,
+      "step": 6831
+    },
+    {
+      "epoch": 8.73728,
+      "grad_norm": 0.5297383069992065,
+      "learning_rate": 3.9938080495356037e-05,
+      "loss": 3.2457,
+      "step": 6832
+    },
+    {
+      "epoch": 8.73856,
+      "grad_norm": 0.528474748134613,
+      "learning_rate": 3.989769820971867e-05,
+      "loss": 3.2796,
+      "step": 6833
+    },
+    {
+      "epoch": 8.739840000000001,
+      "grad_norm": 0.5146127939224243,
+      "learning_rate": 3.98573159240813e-05,
+      "loss": 3.2998,
+      "step": 6834
+    },
+    {
+      "epoch": 8.74112,
+      "grad_norm": 0.5278341174125671,
+      "learning_rate": 3.981693363844393e-05,
+      "loss": 3.2445,
+      "step": 6835
+    },
+    {
+      "epoch": 8.7424,
+      "grad_norm": 0.5160564184188843,
+      "learning_rate": 3.9776551352806564e-05,
+      "loss": 3.307,
+      "step": 6836
+    },
+    {
+      "epoch": 8.74368,
+      "grad_norm": 0.5266233682632446,
+      "learning_rate": 3.9736169067169194e-05,
+      "loss": 3.2835,
+      "step": 6837
+    },
+    {
+      "epoch": 8.74496,
+      "grad_norm": 0.5305745601654053,
+      "learning_rate": 3.969578678153183e-05,
+      "loss": 3.3538,
+      "step": 6838
+    },
+    {
+      "epoch": 8.74624,
+      "grad_norm": 0.5437188744544983,
+      "learning_rate": 3.965540449589446e-05,
+      "loss": 3.3189,
+      "step": 6839
+    },
+    {
+      "epoch": 8.74752,
+      "grad_norm": 0.5442191958427429,
+      "learning_rate": 3.96150222102571e-05,
+      "loss": 3.3207,
+      "step": 6840
+    },
+    {
+      "epoch": 8.7488,
+      "grad_norm": 0.5521953701972961,
+      "learning_rate": 3.957463992461973e-05,
+      "loss": 3.3707,
+      "step": 6841
+    },
+    {
+      "epoch": 8.75008,
+      "grad_norm": 0.5367425084114075,
+      "learning_rate": 3.9534257638982364e-05,
+      "loss": 3.3494,
+      "step": 6842
+    },
+    {
+      "epoch": 8.75136,
+      "grad_norm": 0.5323737263679504,
+      "learning_rate": 3.9493875353344994e-05,
+      "loss": 3.3304,
+      "step": 6843
+    },
+    {
+      "epoch": 8.75264,
+      "grad_norm": 0.5559576749801636,
+      "learning_rate": 3.945349306770763e-05,
+      "loss": 3.2795,
+      "step": 6844
+    },
+    {
+      "epoch": 8.75392,
+      "grad_norm": 0.5402018427848816,
+      "learning_rate": 3.941311078207026e-05,
+      "loss": 3.2885,
+      "step": 6845
+    },
+    {
+      "epoch": 8.7552,
+      "grad_norm": 0.5311586856842041,
+      "learning_rate": 3.93727284964329e-05,
+      "loss": 3.2718,
+      "step": 6846
+    },
+    {
+      "epoch": 8.75648,
+      "grad_norm": 0.5346354246139526,
+      "learning_rate": 3.9332346210795535e-05,
+      "loss": 3.2709,
+      "step": 6847
+    },
+    {
+      "epoch": 8.75776,
+      "grad_norm": 0.5428527593612671,
+      "learning_rate": 3.929196392515816e-05,
+      "loss": 3.296,
+      "step": 6848
+    },
+    {
+      "epoch": 8.75904,
+      "grad_norm": 0.5390904545783997,
+      "learning_rate": 3.925158163952079e-05,
+      "loss": 3.2799,
+      "step": 6849
+    },
+    {
+      "epoch": 8.76032,
+      "grad_norm": 0.5223475694656372,
+      "learning_rate": 3.9211199353883425e-05,
+      "loss": 3.3201,
+      "step": 6850
+    },
+    {
+      "epoch": 8.7616,
+      "grad_norm": 0.5381412506103516,
+      "learning_rate": 3.9170817068246055e-05,
+      "loss": 3.3655,
+      "step": 6851
+    },
+    {
+      "epoch": 8.76288,
+      "grad_norm": 0.5343440175056458,
+      "learning_rate": 3.913043478260869e-05,
+      "loss": 3.264,
+      "step": 6852
+    },
+    {
+      "epoch": 8.76416,
+      "grad_norm": 0.5274061560630798,
+      "learning_rate": 3.909005249697132e-05,
+      "loss": 3.35,
+      "step": 6853
+    },
+    {
+      "epoch": 8.76544,
+      "grad_norm": 0.5369771718978882,
+      "learning_rate": 3.904967021133396e-05,
+      "loss": 3.3448,
+      "step": 6854
+    },
+    {
+      "epoch": 8.76672,
+      "grad_norm": 0.5344805717468262,
+      "learning_rate": 3.900928792569659e-05,
+      "loss": 3.3441,
+      "step": 6855
+    },
+    {
+      "epoch": 8.768,
+      "grad_norm": 0.5286930203437805,
+      "learning_rate": 3.8968905640059226e-05,
+      "loss": 3.3262,
+      "step": 6856
+    },
+    {
+      "epoch": 8.76928,
+      "grad_norm": 0.5351747870445251,
+      "learning_rate": 3.8928523354421856e-05,
+      "loss": 3.3277,
+      "step": 6857
+    },
+    {
+      "epoch": 8.77056,
+      "grad_norm": 0.5334113836288452,
+      "learning_rate": 3.888814106878449e-05,
+      "loss": 3.3462,
+      "step": 6858
+    },
+    {
+      "epoch": 8.77184,
+      "grad_norm": 0.5328750014305115,
+      "learning_rate": 3.884775878314712e-05,
+      "loss": 3.257,
+      "step": 6859
+    },
+    {
+      "epoch": 8.77312,
+      "grad_norm": 0.5237380266189575,
+      "learning_rate": 3.880737649750976e-05,
+      "loss": 3.2508,
+      "step": 6860
+    },
+    {
+      "epoch": 8.7744,
+      "grad_norm": 0.542219877243042,
+      "learning_rate": 3.876699421187238e-05,
+      "loss": 3.3518,
+      "step": 6861
+    },
+    {
+      "epoch": 8.77568,
+      "grad_norm": 0.5408708453178406,
+      "learning_rate": 3.872661192623502e-05,
+      "loss": 3.4064,
+      "step": 6862
+    },
+    {
+      "epoch": 8.77696,
+      "grad_norm": 0.543810248374939,
+      "learning_rate": 3.868622964059765e-05,
+      "loss": 3.3481,
+      "step": 6863
+    },
+    {
+      "epoch": 8.77824,
+      "grad_norm": 0.5442402958869934,
+      "learning_rate": 3.864584735496029e-05,
+      "loss": 3.3938,
+      "step": 6864
+    },
+    {
+      "epoch": 8.77952,
+      "grad_norm": 0.5541428923606873,
+      "learning_rate": 3.860546506932292e-05,
+      "loss": 3.3481,
+      "step": 6865
+    },
+    {
+      "epoch": 8.7808,
+      "grad_norm": 0.5401941537857056,
+      "learning_rate": 3.8565082783685554e-05,
+      "loss": 3.3197,
+      "step": 6866
+    },
+    {
+      "epoch": 8.78208,
+      "grad_norm": 0.5313037633895874,
+      "learning_rate": 3.8524700498048184e-05,
+      "loss": 3.2591,
+      "step": 6867
+    },
+    {
+      "epoch": 8.78336,
+      "grad_norm": 0.5377905368804932,
+      "learning_rate": 3.848431821241082e-05,
+      "loss": 3.3435,
+      "step": 6868
+    },
+    {
+      "epoch": 8.78464,
+      "grad_norm": 0.5325095057487488,
+      "learning_rate": 3.844393592677345e-05,
+      "loss": 3.3114,
+      "step": 6869
+    },
+    {
+      "epoch": 8.78592,
+      "grad_norm": 0.540044367313385,
+      "learning_rate": 3.840355364113609e-05,
+      "loss": 3.3397,
+      "step": 6870
+    },
+    {
+      "epoch": 8.7872,
+      "grad_norm": 0.531450092792511,
+      "learning_rate": 3.836317135549872e-05,
+      "loss": 3.2683,
+      "step": 6871
+    },
+    {
+      "epoch": 8.78848,
+      "grad_norm": 0.5396316647529602,
+      "learning_rate": 3.8322789069861355e-05,
+      "loss": 3.2794,
+      "step": 6872
+    },
+    {
+      "epoch": 8.78976,
+      "grad_norm": 0.5314230918884277,
+      "learning_rate": 3.8282406784223985e-05,
+      "loss": 3.305,
+      "step": 6873
+    },
+    {
+      "epoch": 8.79104,
+      "grad_norm": 0.5359961986541748,
+      "learning_rate": 3.8242024498586615e-05,
+      "loss": 3.3323,
+      "step": 6874
+    },
+    {
+      "epoch": 8.79232,
+      "grad_norm": 0.5214905738830566,
+      "learning_rate": 3.8201642212949245e-05,
+      "loss": 3.2747,
+      "step": 6875
+    },
+    {
+      "epoch": 8.7936,
+      "grad_norm": 0.5209900140762329,
+      "learning_rate": 3.816125992731188e-05,
+      "loss": 3.2886,
+      "step": 6876
+    },
+    {
+      "epoch": 8.79488,
+      "grad_norm": 0.5249080061912537,
+      "learning_rate": 3.812087764167451e-05,
+      "loss": 3.3115,
+      "step": 6877
+    },
+    {
+      "epoch": 8.79616,
+      "grad_norm": 0.561204195022583,
+      "learning_rate": 3.808049535603715e-05,
+      "loss": 3.3694,
+      "step": 6878
+    },
+    {
+      "epoch": 8.79744,
+      "grad_norm": 0.5340297222137451,
+      "learning_rate": 3.804011307039978e-05,
+      "loss": 3.2916,
+      "step": 6879
+    },
+    {
+      "epoch": 8.79872,
+      "grad_norm": 0.5248726010322571,
+      "learning_rate": 3.7999730784762416e-05,
+      "loss": 3.3112,
+      "step": 6880
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.5359779596328735,
+      "learning_rate": 3.7959348499125046e-05,
+      "loss": 3.3402,
+      "step": 6881
+    },
+    {
+      "epoch": 8.80128,
+      "grad_norm": 0.5329644083976746,
+      "learning_rate": 3.791896621348768e-05,
+      "loss": 3.3107,
+      "step": 6882
+    },
+    {
+      "epoch": 8.80256,
+      "grad_norm": 0.5371171236038208,
+      "learning_rate": 3.787858392785031e-05,
+      "loss": 3.3409,
+      "step": 6883
+    },
+    {
+      "epoch": 8.80384,
+      "grad_norm": 0.5442470908164978,
+      "learning_rate": 3.783820164221295e-05,
+      "loss": 3.3277,
+      "step": 6884
+    },
+    {
+      "epoch": 8.80512,
+      "grad_norm": 0.5545779466629028,
+      "learning_rate": 3.779781935657558e-05,
+      "loss": 3.3126,
+      "step": 6885
+    },
+    {
+      "epoch": 8.8064,
+      "grad_norm": 0.5447943210601807,
+      "learning_rate": 3.7757437070938216e-05,
+      "loss": 3.2873,
+      "step": 6886
+    },
+    {
+      "epoch": 8.80768,
+      "grad_norm": 0.5211206674575806,
+      "learning_rate": 3.7717054785300847e-05,
+      "loss": 3.3024,
+      "step": 6887
+    },
+    {
+      "epoch": 8.80896,
+      "grad_norm": 0.5305944681167603,
+      "learning_rate": 3.7676672499663477e-05,
+      "loss": 3.3325,
+      "step": 6888
+    },
+    {
+      "epoch": 8.81024,
+      "grad_norm": 0.5540784001350403,
+      "learning_rate": 3.763629021402611e-05,
+      "loss": 3.3269,
+      "step": 6889
+    },
+    {
+      "epoch": 8.81152,
+      "grad_norm": 0.5730512738227844,
+      "learning_rate": 3.7595907928388744e-05,
+      "loss": 3.3965,
+      "step": 6890
+    },
+    {
+      "epoch": 8.8128,
+      "grad_norm": 0.5502961874008179,
+      "learning_rate": 3.7555525642751374e-05,
+      "loss": 3.2786,
+      "step": 6891
+    },
+    {
+      "epoch": 8.81408,
+      "grad_norm": 0.5448852777481079,
+      "learning_rate": 3.751514335711401e-05,
+      "loss": 3.3031,
+      "step": 6892
+    },
+    {
+      "epoch": 8.81536,
+      "grad_norm": 0.5359817743301392,
+      "learning_rate": 3.747476107147664e-05,
+      "loss": 3.318,
+      "step": 6893
+    },
+    {
+      "epoch": 8.81664,
+      "grad_norm": 0.5279159545898438,
+      "learning_rate": 3.743437878583928e-05,
+      "loss": 3.3592,
+      "step": 6894
+    },
+    {
+      "epoch": 8.81792,
+      "grad_norm": 0.5456517338752747,
+      "learning_rate": 3.739399650020191e-05,
+      "loss": 3.3106,
+      "step": 6895
+    },
+    {
+      "epoch": 8.8192,
+      "grad_norm": 0.5353114604949951,
+      "learning_rate": 3.7353614214564544e-05,
+      "loss": 3.3135,
+      "step": 6896
+    },
+    {
+      "epoch": 8.82048,
+      "grad_norm": 0.5308430194854736,
+      "learning_rate": 3.7313231928927174e-05,
+      "loss": 3.2659,
+      "step": 6897
+    },
+    {
+      "epoch": 8.82176,
+      "grad_norm": 0.5371167659759521,
+      "learning_rate": 3.7272849643289804e-05,
+      "loss": 3.244,
+      "step": 6898
+    },
+    {
+      "epoch": 8.82304,
+      "grad_norm": 0.5514832735061646,
+      "learning_rate": 3.723246735765244e-05,
+      "loss": 3.3401,
+      "step": 6899
+    },
+    {
+      "epoch": 8.82432,
+      "grad_norm": 0.536336362361908,
+      "learning_rate": 3.719208507201507e-05,
+      "loss": 3.3118,
+      "step": 6900
+    },
+    {
+      "epoch": 8.8256,
+      "grad_norm": 0.5184128284454346,
+      "learning_rate": 3.715170278637771e-05,
+      "loss": 3.1998,
+      "step": 6901
+    },
+    {
+      "epoch": 8.82688,
+      "grad_norm": 0.5253881812095642,
+      "learning_rate": 3.711132050074034e-05,
+      "loss": 3.1437,
+      "step": 6902
+    },
+    {
+      "epoch": 8.82816,
+      "grad_norm": 0.5361356735229492,
+      "learning_rate": 3.7070938215102975e-05,
+      "loss": 3.1978,
+      "step": 6903
+    },
+    {
+      "epoch": 8.82944,
+      "grad_norm": 0.533018946647644,
+      "learning_rate": 3.7030555929465605e-05,
+      "loss": 3.2561,
+      "step": 6904
+    },
+    {
+      "epoch": 8.83072,
+      "grad_norm": 0.5180771350860596,
+      "learning_rate": 3.6990173643828235e-05,
+      "loss": 3.2524,
+      "step": 6905
+    },
+    {
+      "epoch": 8.832,
+      "grad_norm": 0.5097092986106873,
+      "learning_rate": 3.694979135819087e-05,
+      "loss": 3.1094,
+      "step": 6906
+    },
+    {
+      "epoch": 8.83328,
+      "grad_norm": 0.5372468829154968,
+      "learning_rate": 3.69094090725535e-05,
+      "loss": 3.1413,
+      "step": 6907
+    },
+    {
+      "epoch": 8.83456,
+      "grad_norm": 0.5473780632019043,
+      "learning_rate": 3.686902678691614e-05,
+      "loss": 3.2555,
+      "step": 6908
+    },
+    {
+      "epoch": 8.83584,
+      "grad_norm": 0.537969172000885,
+      "learning_rate": 3.682864450127877e-05,
+      "loss": 3.0862,
+      "step": 6909
+    },
+    {
+      "epoch": 8.83712,
+      "grad_norm": 0.5439502596855164,
+      "learning_rate": 3.6788262215641406e-05,
+      "loss": 3.1283,
+      "step": 6910
+    },
+    {
+      "epoch": 8.8384,
+      "grad_norm": 0.5461182594299316,
+      "learning_rate": 3.6747879930004036e-05,
+      "loss": 3.2333,
+      "step": 6911
+    },
+    {
+      "epoch": 8.83968,
+      "grad_norm": 0.5343428254127502,
+      "learning_rate": 3.6707497644366666e-05,
+      "loss": 3.1861,
+      "step": 6912
+    },
+    {
+      "epoch": 8.84096,
+      "grad_norm": 0.5404158234596252,
+      "learning_rate": 3.66671153587293e-05,
+      "loss": 3.2274,
+      "step": 6913
+    },
+    {
+      "epoch": 8.84224,
+      "grad_norm": 0.5451922416687012,
+      "learning_rate": 3.662673307309193e-05,
+      "loss": 3.1915,
+      "step": 6914
+    },
+    {
+      "epoch": 8.84352,
+      "grad_norm": 0.5269055366516113,
+      "learning_rate": 3.658635078745457e-05,
+      "loss": 3.2036,
+      "step": 6915
+    },
+    {
+      "epoch": 8.8448,
+      "grad_norm": 0.5376463532447815,
+      "learning_rate": 3.65459685018172e-05,
+      "loss": 3.0966,
+      "step": 6916
+    },
+    {
+      "epoch": 8.84608,
+      "grad_norm": 0.5396373271942139,
+      "learning_rate": 3.650558621617984e-05,
+      "loss": 3.2148,
+      "step": 6917
+    },
+    {
+      "epoch": 8.84736,
+      "grad_norm": 0.5474383234977722,
+      "learning_rate": 3.646520393054247e-05,
+      "loss": 3.0508,
+      "step": 6918
+    },
+    {
+      "epoch": 8.84864,
+      "grad_norm": 0.5413127541542053,
+      "learning_rate": 3.64248216449051e-05,
+      "loss": 3.2371,
+      "step": 6919
+    },
+    {
+      "epoch": 8.849920000000001,
+      "grad_norm": 0.5485610365867615,
+      "learning_rate": 3.6384439359267734e-05,
+      "loss": 3.183,
+      "step": 6920
+    },
+    {
+      "epoch": 8.8512,
+      "grad_norm": 0.551889955997467,
+      "learning_rate": 3.6344057073630364e-05,
+      "loss": 3.2028,
+      "step": 6921
+    },
+    {
+      "epoch": 8.85248,
+      "grad_norm": 0.5297598242759705,
+      "learning_rate": 3.6303674787993e-05,
+      "loss": 3.1678,
+      "step": 6922
+    },
+    {
+      "epoch": 8.85376,
+      "grad_norm": 0.5409109592437744,
+      "learning_rate": 3.626329250235563e-05,
+      "loss": 3.1554,
+      "step": 6923
+    },
+    {
+      "epoch": 8.85504,
+      "grad_norm": 0.5404353737831116,
+      "learning_rate": 3.622291021671827e-05,
+      "loss": 3.2286,
+      "step": 6924
+    },
+    {
+      "epoch": 8.85632,
+      "grad_norm": 0.5331123471260071,
+      "learning_rate": 3.618252793108089e-05,
+      "loss": 3.1951,
+      "step": 6925
+    },
+    {
+      "epoch": 8.8576,
+      "grad_norm": 0.53252112865448,
+      "learning_rate": 3.614214564544353e-05,
+      "loss": 3.1818,
+      "step": 6926
+    },
+    {
+      "epoch": 8.85888,
+      "grad_norm": 0.5230981707572937,
+      "learning_rate": 3.6101763359806165e-05,
+      "loss": 3.1821,
+      "step": 6927
+    },
+    {
+      "epoch": 8.86016,
+      "grad_norm": 0.5399007797241211,
+      "learning_rate": 3.6061381074168795e-05,
+      "loss": 3.1559,
+      "step": 6928
+    },
+    {
+      "epoch": 8.86144,
+      "grad_norm": 0.5414989590644836,
+      "learning_rate": 3.602099878853143e-05,
+      "loss": 3.1675,
+      "step": 6929
+    },
+    {
+      "epoch": 8.86272,
+      "grad_norm": 0.5386769771575928,
+      "learning_rate": 3.598061650289406e-05,
+      "loss": 3.1707,
+      "step": 6930
+    },
+    {
+      "epoch": 8.864,
+      "grad_norm": 0.5298625230789185,
+      "learning_rate": 3.594023421725669e-05,
+      "loss": 3.2093,
+      "step": 6931
+    },
+    {
+      "epoch": 8.86528,
+      "grad_norm": 0.5370173454284668,
+      "learning_rate": 3.589985193161932e-05,
+      "loss": 3.1237,
+      "step": 6932
+    },
+    {
+      "epoch": 8.86656,
+      "grad_norm": 0.5402517914772034,
+      "learning_rate": 3.585946964598196e-05,
+      "loss": 3.2736,
+      "step": 6933
+    },
+    {
+      "epoch": 8.86784,
+      "grad_norm": 0.5565125346183777,
+      "learning_rate": 3.581908736034459e-05,
+      "loss": 3.2115,
+      "step": 6934
+    },
+    {
+      "epoch": 8.86912,
+      "grad_norm": 0.5417731404304504,
+      "learning_rate": 3.5778705074707226e-05,
+      "loss": 3.2296,
+      "step": 6935
+    },
+    {
+      "epoch": 8.8704,
+      "grad_norm": 0.5352853536605835,
+      "learning_rate": 3.573832278906986e-05,
+      "loss": 3.1484,
+      "step": 6936
+    },
+    {
+      "epoch": 8.87168,
+      "grad_norm": 0.5440704226493835,
+      "learning_rate": 3.569794050343249e-05,
+      "loss": 3.1939,
+      "step": 6937
+    },
+    {
+      "epoch": 8.872959999999999,
+      "grad_norm": 0.553403913974762,
+      "learning_rate": 3.565755821779512e-05,
+      "loss": 3.1381,
+      "step": 6938
+    },
+    {
+      "epoch": 8.87424,
+      "grad_norm": 0.5392211079597473,
+      "learning_rate": 3.561717593215775e-05,
+      "loss": 3.2004,
+      "step": 6939
+    },
+    {
+      "epoch": 8.87552,
+      "grad_norm": 0.5346492528915405,
+      "learning_rate": 3.557679364652039e-05,
+      "loss": 3.2119,
+      "step": 6940
+    },
+    {
+      "epoch": 8.8768,
+      "grad_norm": 0.5390061140060425,
+      "learning_rate": 3.553641136088302e-05,
+      "loss": 3.2046,
+      "step": 6941
+    },
+    {
+      "epoch": 8.87808,
+      "grad_norm": 0.5476921796798706,
+      "learning_rate": 3.5496029075245657e-05,
+      "loss": 3.2266,
+      "step": 6942
+    },
+    {
+      "epoch": 8.87936,
+      "grad_norm": 0.5377036333084106,
+      "learning_rate": 3.5455646789608287e-05,
+      "loss": 3.2189,
+      "step": 6943
+    },
+    {
+      "epoch": 8.88064,
+      "grad_norm": 0.5338594317436218,
+      "learning_rate": 3.5415264503970923e-05,
+      "loss": 3.2326,
+      "step": 6944
+    },
+    {
+      "epoch": 8.881920000000001,
+      "grad_norm": 0.5427373647689819,
+      "learning_rate": 3.5374882218333554e-05,
+      "loss": 3.2784,
+      "step": 6945
+    },
+    {
+      "epoch": 8.8832,
+      "grad_norm": 0.5387793779373169,
+      "learning_rate": 3.5334499932696184e-05,
+      "loss": 3.1931,
+      "step": 6946
+    },
+    {
+      "epoch": 8.88448,
+      "grad_norm": 0.5435247421264648,
+      "learning_rate": 3.529411764705882e-05,
+      "loss": 3.1342,
+      "step": 6947
+    },
+    {
+      "epoch": 8.88576,
+      "grad_norm": 0.5299807786941528,
+      "learning_rate": 3.525373536142145e-05,
+      "loss": 3.1364,
+      "step": 6948
+    },
+    {
+      "epoch": 8.88704,
+      "grad_norm": 0.5515544414520264,
+      "learning_rate": 3.521335307578409e-05,
+      "loss": 3.2109,
+      "step": 6949
+    },
+    {
+      "epoch": 8.88832,
+      "grad_norm": 0.5315101146697998,
+      "learning_rate": 3.517297079014672e-05,
+      "loss": 3.0987,
+      "step": 6950
+    },
+    {
+      "epoch": 8.8896,
+      "grad_norm": 0.5465390682220459,
+      "learning_rate": 3.5132588504509354e-05,
+      "loss": 3.1256,
+      "step": 6951
+    },
+    {
+      "epoch": 8.89088,
+      "grad_norm": 0.5575345158576965,
+      "learning_rate": 3.5092206218871984e-05,
+      "loss": 3.194,
+      "step": 6952
+    },
+    {
+      "epoch": 8.89216,
+      "grad_norm": 0.5439948439598083,
+      "learning_rate": 3.5051823933234614e-05,
+      "loss": 3.2022,
+      "step": 6953
+    },
+    {
+      "epoch": 8.89344,
+      "grad_norm": 0.5493448972702026,
+      "learning_rate": 3.501144164759725e-05,
+      "loss": 3.2155,
+      "step": 6954
+    },
+    {
+      "epoch": 8.89472,
+      "grad_norm": 0.5295661091804504,
+      "learning_rate": 3.497105936195988e-05,
+      "loss": 3.1321,
+      "step": 6955
+    },
+    {
+      "epoch": 8.896,
+      "grad_norm": 0.5430306792259216,
+      "learning_rate": 3.493067707632252e-05,
+      "loss": 3.1363,
+      "step": 6956
+    },
+    {
+      "epoch": 8.89728,
+      "grad_norm": 0.5532288551330566,
+      "learning_rate": 3.489029479068515e-05,
+      "loss": 3.2848,
+      "step": 6957
+    },
+    {
+      "epoch": 8.89856,
+      "grad_norm": 0.5648680329322815,
+      "learning_rate": 3.484991250504778e-05,
+      "loss": 3.1427,
+      "step": 6958
+    },
+    {
+      "epoch": 8.89984,
+      "grad_norm": 0.5307405591011047,
+      "learning_rate": 3.4809530219410415e-05,
+      "loss": 3.2077,
+      "step": 6959
+    },
+    {
+      "epoch": 8.90112,
+      "grad_norm": 0.5643923282623291,
+      "learning_rate": 3.4769147933773045e-05,
+      "loss": 3.1602,
+      "step": 6960
+    },
+    {
+      "epoch": 8.9024,
+      "grad_norm": 0.5337897539138794,
+      "learning_rate": 3.472876564813568e-05,
+      "loss": 3.186,
+      "step": 6961
+    },
+    {
+      "epoch": 8.90368,
+      "grad_norm": 0.5449624061584473,
+      "learning_rate": 3.468838336249831e-05,
+      "loss": 3.223,
+      "step": 6962
+    },
+    {
+      "epoch": 8.904959999999999,
+      "grad_norm": 0.5331872701644897,
+      "learning_rate": 3.464800107686095e-05,
+      "loss": 3.1097,
+      "step": 6963
+    },
+    {
+      "epoch": 8.90624,
+      "grad_norm": 0.5646026134490967,
+      "learning_rate": 3.460761879122358e-05,
+      "loss": 3.203,
+      "step": 6964
+    },
+    {
+      "epoch": 8.90752,
+      "grad_norm": 0.554621696472168,
+      "learning_rate": 3.456723650558621e-05,
+      "loss": 3.1597,
+      "step": 6965
+    },
+    {
+      "epoch": 8.9088,
+      "grad_norm": 0.5355739593505859,
+      "learning_rate": 3.4526854219948846e-05,
+      "loss": 3.2187,
+      "step": 6966
+    },
+    {
+      "epoch": 8.91008,
+      "grad_norm": 0.5411005020141602,
+      "learning_rate": 3.4486471934311476e-05,
+      "loss": 3.1622,
+      "step": 6967
+    },
+    {
+      "epoch": 8.91136,
+      "grad_norm": 0.5598929524421692,
+      "learning_rate": 3.444608964867411e-05,
+      "loss": 3.1857,
+      "step": 6968
+    },
+    {
+      "epoch": 8.91264,
+      "grad_norm": 0.5395918488502502,
+      "learning_rate": 3.440570736303674e-05,
+      "loss": 3.1413,
+      "step": 6969
+    },
+    {
+      "epoch": 8.91392,
+      "grad_norm": 0.5290459394454956,
+      "learning_rate": 3.436532507739938e-05,
+      "loss": 3.1358,
+      "step": 6970
+    },
+    {
+      "epoch": 8.9152,
+      "grad_norm": 0.5369711518287659,
+      "learning_rate": 3.432494279176201e-05,
+      "loss": 3.1303,
+      "step": 6971
+    },
+    {
+      "epoch": 8.91648,
+      "grad_norm": 0.5442224740982056,
+      "learning_rate": 3.428456050612464e-05,
+      "loss": 3.1369,
+      "step": 6972
+    },
+    {
+      "epoch": 8.91776,
+      "grad_norm": 0.5366105437278748,
+      "learning_rate": 3.424417822048728e-05,
+      "loss": 3.1323,
+      "step": 6973
+    },
+    {
+      "epoch": 8.91904,
+      "grad_norm": 0.5568270087242126,
+      "learning_rate": 3.420379593484991e-05,
+      "loss": 3.1774,
+      "step": 6974
+    },
+    {
+      "epoch": 8.92032,
+      "grad_norm": 0.5447999835014343,
+      "learning_rate": 3.4163413649212544e-05,
+      "loss": 3.1728,
+      "step": 6975
+    },
+    {
+      "epoch": 8.9216,
+      "grad_norm": 0.5445603728294373,
+      "learning_rate": 3.4123031363575174e-05,
+      "loss": 3.1144,
+      "step": 6976
+    },
+    {
+      "epoch": 8.92288,
+      "grad_norm": 0.5516496896743774,
+      "learning_rate": 3.408264907793781e-05,
+      "loss": 3.1628,
+      "step": 6977
+    },
+    {
+      "epoch": 8.92416,
+      "grad_norm": 0.5592992901802063,
+      "learning_rate": 3.404226679230044e-05,
+      "loss": 3.1663,
+      "step": 6978
+    },
+    {
+      "epoch": 8.92544,
+      "grad_norm": 0.5563246607780457,
+      "learning_rate": 3.400188450666307e-05,
+      "loss": 3.1543,
+      "step": 6979
+    },
+    {
+      "epoch": 8.92672,
+      "grad_norm": 0.5511194467544556,
+      "learning_rate": 3.396150222102571e-05,
+      "loss": 3.2003,
+      "step": 6980
+    },
+    {
+      "epoch": 8.928,
+      "grad_norm": 0.5466040372848511,
+      "learning_rate": 3.392111993538834e-05,
+      "loss": 3.1386,
+      "step": 6981
+    },
+    {
+      "epoch": 8.92928,
+      "grad_norm": 0.5524888038635254,
+      "learning_rate": 3.3880737649750975e-05,
+      "loss": 3.0766,
+      "step": 6982
+    },
+    {
+      "epoch": 8.93056,
+      "grad_norm": 0.5536906123161316,
+      "learning_rate": 3.3840355364113605e-05,
+      "loss": 3.1638,
+      "step": 6983
+    },
+    {
+      "epoch": 8.93184,
+      "grad_norm": 0.5587788224220276,
+      "learning_rate": 3.379997307847624e-05,
+      "loss": 3.2024,
+      "step": 6984
+    },
+    {
+      "epoch": 8.93312,
+      "grad_norm": 0.5503929257392883,
+      "learning_rate": 3.375959079283887e-05,
+      "loss": 3.2101,
+      "step": 6985
+    },
+    {
+      "epoch": 8.9344,
+      "grad_norm": 0.5357226729393005,
+      "learning_rate": 3.37192085072015e-05,
+      "loss": 3.2078,
+      "step": 6986
+    },
+    {
+      "epoch": 8.93568,
+      "grad_norm": 0.5584990382194519,
+      "learning_rate": 3.367882622156414e-05,
+      "loss": 3.186,
+      "step": 6987
+    },
+    {
+      "epoch": 8.93696,
+      "grad_norm": 0.5508469939231873,
+      "learning_rate": 3.363844393592677e-05,
+      "loss": 3.1558,
+      "step": 6988
+    },
+    {
+      "epoch": 8.93824,
+      "grad_norm": 0.5476589202880859,
+      "learning_rate": 3.3598061650289406e-05,
+      "loss": 3.1501,
+      "step": 6989
+    },
+    {
+      "epoch": 8.93952,
+      "grad_norm": 0.5468451380729675,
+      "learning_rate": 3.3557679364652036e-05,
+      "loss": 3.211,
+      "step": 6990
+    },
+    {
+      "epoch": 8.9408,
+      "grad_norm": 0.5536555647850037,
+      "learning_rate": 3.351729707901467e-05,
+      "loss": 3.1646,
+      "step": 6991
+    },
+    {
+      "epoch": 8.94208,
+      "grad_norm": 0.5505756735801697,
+      "learning_rate": 3.34769147933773e-05,
+      "loss": 3.1244,
+      "step": 6992
+    },
+    {
+      "epoch": 8.94336,
+      "grad_norm": 0.560417890548706,
+      "learning_rate": 3.343653250773993e-05,
+      "loss": 3.1672,
+      "step": 6993
+    },
+    {
+      "epoch": 8.94464,
+      "grad_norm": 0.5642468929290771,
+      "learning_rate": 3.339615022210257e-05,
+      "loss": 3.1955,
+      "step": 6994
+    },
+    {
+      "epoch": 8.94592,
+      "grad_norm": 0.5576351881027222,
+      "learning_rate": 3.33557679364652e-05,
+      "loss": 3.1591,
+      "step": 6995
+    },
+    {
+      "epoch": 8.9472,
+      "grad_norm": 0.543550968170166,
+      "learning_rate": 3.3315385650827836e-05,
+      "loss": 3.1024,
+      "step": 6996
+    },
+    {
+      "epoch": 8.94848,
+      "grad_norm": 0.5292938947677612,
+      "learning_rate": 3.3275003365190467e-05,
+      "loss": 3.1107,
+      "step": 6997
+    },
+    {
+      "epoch": 8.94976,
+      "grad_norm": 0.5450161695480347,
+      "learning_rate": 3.32346210795531e-05,
+      "loss": 3.1862,
+      "step": 6998
+    },
+    {
+      "epoch": 8.95104,
+      "grad_norm": 0.5542306303977966,
+      "learning_rate": 3.3194238793915733e-05,
+      "loss": 3.209,
+      "step": 6999
+    },
+    {
+      "epoch": 8.95232,
+      "grad_norm": 0.552590548992157,
+      "learning_rate": 3.3153856508278364e-05,
+      "loss": 3.2279,
+      "step": 7000
+    },
+    {
+      "epoch": 8.9536,
+      "grad_norm": 0.548201858997345,
+      "learning_rate": 3.3113474222641e-05,
+      "loss": 3.1338,
+      "step": 7001
+    },
+    {
+      "epoch": 8.95488,
+      "grad_norm": 0.5407779216766357,
+      "learning_rate": 3.307309193700363e-05,
+      "loss": 3.1704,
+      "step": 7002
+    },
+    {
+      "epoch": 8.95616,
+      "grad_norm": 0.5337063074111938,
+      "learning_rate": 3.303270965136627e-05,
+      "loss": 3.1294,
+      "step": 7003
+    },
+    {
+      "epoch": 8.95744,
+      "grad_norm": 0.5435624718666077,
+      "learning_rate": 3.29923273657289e-05,
+      "loss": 3.2299,
+      "step": 7004
+    },
+    {
+      "epoch": 8.95872,
+      "grad_norm": 0.5566171407699585,
+      "learning_rate": 3.295194508009153e-05,
+      "loss": 3.2128,
+      "step": 7005
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.5542343854904175,
+      "learning_rate": 3.2911562794454164e-05,
+      "loss": 3.1888,
+      "step": 7006
+    },
+    {
+      "epoch": 8.96128,
+      "grad_norm": 0.5571079254150391,
+      "learning_rate": 3.2871180508816794e-05,
+      "loss": 3.1994,
+      "step": 7007
+    },
+    {
+      "epoch": 8.96256,
+      "grad_norm": 0.5469478964805603,
+      "learning_rate": 3.283079822317943e-05,
+      "loss": 3.1812,
+      "step": 7008
+    },
+    {
+      "epoch": 8.96384,
+      "grad_norm": 0.5388615727424622,
+      "learning_rate": 3.279041593754206e-05,
+      "loss": 3.1808,
+      "step": 7009
+    },
+    {
+      "epoch": 8.96512,
+      "grad_norm": 0.5441920161247253,
+      "learning_rate": 3.27500336519047e-05,
+      "loss": 3.1783,
+      "step": 7010
+    },
+    {
+      "epoch": 8.9664,
+      "grad_norm": 0.5513103604316711,
+      "learning_rate": 3.270965136626733e-05,
+      "loss": 3.1525,
+      "step": 7011
+    },
+    {
+      "epoch": 8.96768,
+      "grad_norm": 0.5584927201271057,
+      "learning_rate": 3.266926908062996e-05,
+      "loss": 3.207,
+      "step": 7012
+    },
+    {
+      "epoch": 8.96896,
+      "grad_norm": 0.5390376448631287,
+      "learning_rate": 3.2628886794992595e-05,
+      "loss": 3.1688,
+      "step": 7013
+    },
+    {
+      "epoch": 8.97024,
+      "grad_norm": 0.5638677477836609,
+      "learning_rate": 3.2588504509355225e-05,
+      "loss": 3.2089,
+      "step": 7014
+    },
+    {
+      "epoch": 8.97152,
+      "grad_norm": 0.5538638830184937,
+      "learning_rate": 3.254812222371786e-05,
+      "loss": 3.1597,
+      "step": 7015
+    },
+    {
+      "epoch": 8.9728,
+      "grad_norm": 0.5503858923912048,
+      "learning_rate": 3.250773993808049e-05,
+      "loss": 3.1927,
+      "step": 7016
+    },
+    {
+      "epoch": 8.97408,
+      "grad_norm": 0.5431429147720337,
+      "learning_rate": 3.246735765244313e-05,
+      "loss": 3.2362,
+      "step": 7017
+    },
+    {
+      "epoch": 8.97536,
+      "grad_norm": 0.5614336133003235,
+      "learning_rate": 3.242697536680576e-05,
+      "loss": 3.1585,
+      "step": 7018
+    },
+    {
+      "epoch": 8.97664,
+      "grad_norm": 0.540317714214325,
+      "learning_rate": 3.238659308116839e-05,
+      "loss": 3.2157,
+      "step": 7019
+    },
+    {
+      "epoch": 8.97792,
+      "grad_norm": 0.5396054983139038,
+      "learning_rate": 3.2346210795531026e-05,
+      "loss": 3.1543,
+      "step": 7020
+    },
+    {
+      "epoch": 8.9792,
+      "grad_norm": 0.541448712348938,
+      "learning_rate": 3.2305828509893656e-05,
+      "loss": 3.1709,
+      "step": 7021
+    },
+    {
+      "epoch": 8.98048,
+      "grad_norm": 0.553905725479126,
+      "learning_rate": 3.226544622425629e-05,
+      "loss": 3.1969,
+      "step": 7022
+    },
+    {
+      "epoch": 8.98176,
+      "grad_norm": 0.5426627397537231,
+      "learning_rate": 3.222506393861892e-05,
+      "loss": 3.2246,
+      "step": 7023
+    },
+    {
+      "epoch": 8.98304,
+      "grad_norm": 0.5612876415252686,
+      "learning_rate": 3.218468165298156e-05,
+      "loss": 3.2059,
+      "step": 7024
+    },
+    {
+      "epoch": 8.98432,
+      "grad_norm": 0.5525398850440979,
+      "learning_rate": 3.214429936734419e-05,
+      "loss": 3.1842,
+      "step": 7025
+    },
+    {
+      "epoch": 8.9856,
+      "grad_norm": 0.5441053509712219,
+      "learning_rate": 3.210391708170682e-05,
+      "loss": 3.2081,
+      "step": 7026
+    },
+    {
+      "epoch": 8.98688,
+      "grad_norm": 0.5565687417984009,
+      "learning_rate": 3.206353479606946e-05,
+      "loss": 3.2089,
+      "step": 7027
+    },
+    {
+      "epoch": 8.98816,
+      "grad_norm": 0.553962230682373,
+      "learning_rate": 3.202315251043209e-05,
+      "loss": 3.2249,
+      "step": 7028
+    },
+    {
+      "epoch": 8.98944,
+      "grad_norm": 0.5383572578430176,
+      "learning_rate": 3.1982770224794724e-05,
+      "loss": 3.1249,
+      "step": 7029
+    },
+    {
+      "epoch": 8.99072,
+      "grad_norm": 0.5608500242233276,
+      "learning_rate": 3.1942387939157354e-05,
+      "loss": 3.2204,
+      "step": 7030
+    },
+    {
+      "epoch": 8.992,
+      "grad_norm": 0.5564176440238953,
+      "learning_rate": 3.190200565351999e-05,
+      "loss": 3.222,
+      "step": 7031
+    },
+    {
+      "epoch": 8.99328,
+      "grad_norm": 0.5383712649345398,
+      "learning_rate": 3.186162336788262e-05,
+      "loss": 3.1404,
+      "step": 7032
+    },
+    {
+      "epoch": 8.99456,
+      "grad_norm": 0.5401707291603088,
+      "learning_rate": 3.182124108224525e-05,
+      "loss": 3.1736,
+      "step": 7033
+    },
+    {
+      "epoch": 8.99584,
+      "grad_norm": 0.5486315488815308,
+      "learning_rate": 3.178085879660789e-05,
+      "loss": 3.1421,
+      "step": 7034
+    },
+    {
+      "epoch": 8.99712,
+      "grad_norm": 0.5637720227241516,
+      "learning_rate": 3.174047651097052e-05,
+      "loss": 3.0799,
+      "step": 7035
+    },
+    {
+      "epoch": 8.9984,
+      "grad_norm": 0.5368375182151794,
+      "learning_rate": 3.1700094225333155e-05,
+      "loss": 3.1562,
+      "step": 7036
+    },
+    {
+      "epoch": 8.99968,
+      "grad_norm": 0.5359020233154297,
+      "learning_rate": 3.1659711939695785e-05,
+      "loss": 3.1867,
+      "step": 7037
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.9711993932723999,
+      "learning_rate": 3.161932965405842e-05,
+      "loss": 2.7479,
+      "step": 7038
+    },
+    {
+      "epoch": 9.00128,
+      "grad_norm": 0.5583388209342957,
+      "learning_rate": 3.1578947368421045e-05,
+      "loss": 3.2842,
+      "step": 7039
+    },
+    {
+      "epoch": 9.00256,
+      "grad_norm": 0.5551260113716125,
+      "learning_rate": 3.153856508278368e-05,
+      "loss": 3.2511,
+      "step": 7040
+    },
+    {
+      "epoch": 9.00384,
+      "grad_norm": 0.5703940987586975,
+      "learning_rate": 3.149818279714632e-05,
+      "loss": 3.2821,
+      "step": 7041
+    },
+    {
+      "epoch": 9.00512,
+      "grad_norm": 0.5474501848220825,
+      "learning_rate": 3.145780051150895e-05,
+      "loss": 3.2544,
+      "step": 7042
+    },
+    {
+      "epoch": 9.0064,
+      "grad_norm": 0.5608713030815125,
+      "learning_rate": 3.1417418225871586e-05,
+      "loss": 3.3234,
+      "step": 7043
+    },
+    {
+      "epoch": 9.00768,
+      "grad_norm": 0.5595000386238098,
+      "learning_rate": 3.1377035940234216e-05,
+      "loss": 3.2516,
+      "step": 7044
+    },
+    {
+      "epoch": 9.00896,
+      "grad_norm": 0.5540786385536194,
+      "learning_rate": 3.133665365459685e-05,
+      "loss": 3.3641,
+      "step": 7045
+    },
+    {
+      "epoch": 9.01024,
+      "grad_norm": 0.5481628775596619,
+      "learning_rate": 3.1296271368959476e-05,
+      "loss": 3.1462,
+      "step": 7046
+    },
+    {
+      "epoch": 9.01152,
+      "grad_norm": 0.545194685459137,
+      "learning_rate": 3.125588908332211e-05,
+      "loss": 3.2292,
+      "step": 7047
+    },
+    {
+      "epoch": 9.0128,
+      "grad_norm": 0.5309725403785706,
+      "learning_rate": 3.121550679768474e-05,
+      "loss": 3.244,
+      "step": 7048
+    },
+    {
+      "epoch": 9.01408,
+      "grad_norm": 0.5396099090576172,
+      "learning_rate": 3.117512451204738e-05,
+      "loss": 3.3037,
+      "step": 7049
+    },
+    {
+      "epoch": 9.01536,
+      "grad_norm": 0.5477572083473206,
+      "learning_rate": 3.1134742226410016e-05,
+      "loss": 3.1592,
+      "step": 7050
+    },
+    {
+      "epoch": 9.01664,
+      "grad_norm": 0.5614184737205505,
+      "learning_rate": 3.1094359940772646e-05,
+      "loss": 3.1785,
+      "step": 7051
+    },
+    {
+      "epoch": 9.01792,
+      "grad_norm": 0.5425687432289124,
+      "learning_rate": 3.1053977655135277e-05,
+      "loss": 3.2016,
+      "step": 7052
+    },
+    {
+      "epoch": 9.0192,
+      "grad_norm": 0.5536109209060669,
+      "learning_rate": 3.1013595369497907e-05,
+      "loss": 3.247,
+      "step": 7053
+    },
+    {
+      "epoch": 9.02048,
+      "grad_norm": 0.5437664985656738,
+      "learning_rate": 3.0973213083860543e-05,
+      "loss": 3.2428,
+      "step": 7054
+    },
+    {
+      "epoch": 9.02176,
+      "grad_norm": 0.5411404371261597,
+      "learning_rate": 3.0932830798223174e-05,
+      "loss": 3.2555,
+      "step": 7055
+    },
+    {
+      "epoch": 9.02304,
+      "grad_norm": 0.5475705862045288,
+      "learning_rate": 3.089244851258581e-05,
+      "loss": 3.2672,
+      "step": 7056
+    },
+    {
+      "epoch": 9.02432,
+      "grad_norm": 0.5438557863235474,
+      "learning_rate": 3.085206622694844e-05,
+      "loss": 3.1994,
+      "step": 7057
+    },
+    {
+      "epoch": 9.0256,
+      "grad_norm": 0.54364413022995,
+      "learning_rate": 3.081168394131108e-05,
+      "loss": 3.249,
+      "step": 7058
+    },
+    {
+      "epoch": 9.02688,
+      "grad_norm": 0.5349510908126831,
+      "learning_rate": 3.077130165567371e-05,
+      "loss": 3.2396,
+      "step": 7059
+    },
+    {
+      "epoch": 9.02816,
+      "grad_norm": 0.5500809550285339,
+      "learning_rate": 3.073091937003634e-05,
+      "loss": 3.2324,
+      "step": 7060
+    },
+    {
+      "epoch": 9.02944,
+      "grad_norm": 0.5365533232688904,
+      "learning_rate": 3.0690537084398974e-05,
+      "loss": 3.1999,
+      "step": 7061
+    },
+    {
+      "epoch": 9.03072,
+      "grad_norm": 0.5576481819152832,
+      "learning_rate": 3.0650154798761604e-05,
+      "loss": 3.3374,
+      "step": 7062
+    },
+    {
+      "epoch": 9.032,
+      "grad_norm": 0.5433608293533325,
+      "learning_rate": 3.060977251312424e-05,
+      "loss": 3.2957,
+      "step": 7063
+    },
+    {
+      "epoch": 9.03328,
+      "grad_norm": 0.5531395673751831,
+      "learning_rate": 3.056939022748687e-05,
+      "loss": 3.2484,
+      "step": 7064
+    },
+    {
+      "epoch": 9.03456,
+      "grad_norm": 0.5414525866508484,
+      "learning_rate": 3.052900794184951e-05,
+      "loss": 3.218,
+      "step": 7065
+    },
+    {
+      "epoch": 9.03584,
+      "grad_norm": 0.5335577726364136,
+      "learning_rate": 3.0488625656212138e-05,
+      "loss": 3.1524,
+      "step": 7066
+    },
+    {
+      "epoch": 9.03712,
+      "grad_norm": 0.5346528887748718,
+      "learning_rate": 3.044824337057477e-05,
+      "loss": 3.172,
+      "step": 7067
+    },
+    {
+      "epoch": 9.0384,
+      "grad_norm": 0.5395179390907288,
+      "learning_rate": 3.0407861084937405e-05,
+      "loss": 3.1621,
+      "step": 7068
+    },
+    {
+      "epoch": 9.03968,
+      "grad_norm": 0.544449508190155,
+      "learning_rate": 3.036747879930004e-05,
+      "loss": 3.2452,
+      "step": 7069
+    },
+    {
+      "epoch": 9.04096,
+      "grad_norm": 0.5622183680534363,
+      "learning_rate": 3.0327096513662672e-05,
+      "loss": 3.2826,
+      "step": 7070
+    },
+    {
+      "epoch": 9.04224,
+      "grad_norm": 0.543907880783081,
+      "learning_rate": 3.0286714228025306e-05,
+      "loss": 3.2684,
+      "step": 7071
+    },
+    {
+      "epoch": 9.043520000000001,
+      "grad_norm": 0.5473181009292603,
+      "learning_rate": 3.024633194238794e-05,
+      "loss": 3.2676,
+      "step": 7072
+    },
+    {
+      "epoch": 9.0448,
+      "grad_norm": 0.5329403877258301,
+      "learning_rate": 3.020594965675057e-05,
+      "loss": 3.2042,
+      "step": 7073
+    },
+    {
+      "epoch": 9.04608,
+      "grad_norm": 0.519966185092926,
+      "learning_rate": 3.0165567371113203e-05,
+      "loss": 3.1961,
+      "step": 7074
+    },
+    {
+      "epoch": 9.04736,
+      "grad_norm": 0.5505262017250061,
+      "learning_rate": 3.0125185085475836e-05,
+      "loss": 3.2669,
+      "step": 7075
+    },
+    {
+      "epoch": 9.04864,
+      "grad_norm": 0.5391356945037842,
+      "learning_rate": 3.008480279983847e-05,
+      "loss": 3.2735,
+      "step": 7076
+    },
+    {
+      "epoch": 9.04992,
+      "grad_norm": 0.5384522676467896,
+      "learning_rate": 3.0044420514201103e-05,
+      "loss": 3.3018,
+      "step": 7077
+    },
+    {
+      "epoch": 9.0512,
+      "grad_norm": 0.5468092560768127,
+      "learning_rate": 3.0004038228563736e-05,
+      "loss": 3.2427,
+      "step": 7078
+    },
+    {
+      "epoch": 9.05248,
+      "grad_norm": 0.5462827682495117,
+      "learning_rate": 2.9963655942926366e-05,
+      "loss": 3.2236,
+      "step": 7079
+    },
+    {
+      "epoch": 9.05376,
+      "grad_norm": 0.5457635521888733,
+      "learning_rate": 2.9923273657289e-05,
+      "loss": 3.1881,
+      "step": 7080
+    },
+    {
+      "epoch": 9.05504,
+      "grad_norm": 0.5283154249191284,
+      "learning_rate": 2.9882891371651633e-05,
+      "loss": 3.2305,
+      "step": 7081
+    },
+    {
+      "epoch": 9.05632,
+      "grad_norm": 0.533507227897644,
+      "learning_rate": 2.9842509086014267e-05,
+      "loss": 3.2302,
+      "step": 7082
+    },
+    {
+      "epoch": 9.0576,
+      "grad_norm": 0.5426484942436218,
+      "learning_rate": 2.98021268003769e-05,
+      "loss": 3.2233,
+      "step": 7083
+    },
+    {
+      "epoch": 9.05888,
+      "grad_norm": 0.5597418546676636,
+      "learning_rate": 2.9761744514739534e-05,
+      "loss": 3.2381,
+      "step": 7084
+    },
+    {
+      "epoch": 9.06016,
+      "grad_norm": 0.5480257868766785,
+      "learning_rate": 2.9721362229102167e-05,
+      "loss": 3.2169,
+      "step": 7085
+    },
+    {
+      "epoch": 9.06144,
+      "grad_norm": 0.5565598011016846,
+      "learning_rate": 2.9680979943464797e-05,
+      "loss": 3.2586,
+      "step": 7086
+    },
+    {
+      "epoch": 9.06272,
+      "grad_norm": 0.5538652539253235,
+      "learning_rate": 2.964059765782743e-05,
+      "loss": 3.211,
+      "step": 7087
+    },
+    {
+      "epoch": 9.064,
+      "grad_norm": 0.5413698554039001,
+      "learning_rate": 2.9600215372190064e-05,
+      "loss": 3.2349,
+      "step": 7088
+    },
+    {
+      "epoch": 9.06528,
+      "grad_norm": 0.5424516797065735,
+      "learning_rate": 2.9559833086552698e-05,
+      "loss": 3.2359,
+      "step": 7089
+    },
+    {
+      "epoch": 9.06656,
+      "grad_norm": 0.5624558925628662,
+      "learning_rate": 2.951945080091533e-05,
+      "loss": 3.2008,
+      "step": 7090
+    },
+    {
+      "epoch": 9.06784,
+      "grad_norm": 0.5545729398727417,
+      "learning_rate": 2.9479068515277965e-05,
+      "loss": 3.2058,
+      "step": 7091
+    },
+    {
+      "epoch": 9.06912,
+      "grad_norm": 0.5462035536766052,
+      "learning_rate": 2.9438686229640598e-05,
+      "loss": 3.2632,
+      "step": 7092
+    },
+    {
+      "epoch": 9.0704,
+      "grad_norm": 0.5403680801391602,
+      "learning_rate": 2.9398303944003225e-05,
+      "loss": 3.2468,
+      "step": 7093
+    },
+    {
+      "epoch": 9.07168,
+      "grad_norm": 0.5381686091423035,
+      "learning_rate": 2.935792165836586e-05,
+      "loss": 3.2539,
+      "step": 7094
+    },
+    {
+      "epoch": 9.07296,
+      "grad_norm": 0.5552157759666443,
+      "learning_rate": 2.9317539372728495e-05,
+      "loss": 3.2792,
+      "step": 7095
+    },
+    {
+      "epoch": 9.07424,
+      "grad_norm": 0.5554720759391785,
+      "learning_rate": 2.927715708709113e-05,
+      "loss": 3.1714,
+      "step": 7096
+    },
+    {
+      "epoch": 9.07552,
+      "grad_norm": 0.5406906604766846,
+      "learning_rate": 2.9236774801453762e-05,
+      "loss": 3.1821,
+      "step": 7097
+    },
+    {
+      "epoch": 9.0768,
+      "grad_norm": 0.5409752130508423,
+      "learning_rate": 2.9196392515816396e-05,
+      "loss": 3.2008,
+      "step": 7098
+    },
+    {
+      "epoch": 9.07808,
+      "grad_norm": 0.5609580874443054,
+      "learning_rate": 2.9156010230179022e-05,
+      "loss": 3.2278,
+      "step": 7099
+    },
+    {
+      "epoch": 9.07936,
+      "grad_norm": 0.5392387509346008,
+      "learning_rate": 2.9115627944541656e-05,
+      "loss": 3.2108,
+      "step": 7100
+    },
+    {
+      "epoch": 9.08064,
+      "grad_norm": 0.5388525128364563,
+      "learning_rate": 2.907524565890429e-05,
+      "loss": 3.2953,
+      "step": 7101
+    },
+    {
+      "epoch": 9.08192,
+      "grad_norm": 0.544260561466217,
+      "learning_rate": 2.9034863373266923e-05,
+      "loss": 3.2398,
+      "step": 7102
+    },
+    {
+      "epoch": 9.0832,
+      "grad_norm": 0.5434393882751465,
+      "learning_rate": 2.8994481087629556e-05,
+      "loss": 3.2731,
+      "step": 7103
+    },
+    {
+      "epoch": 9.08448,
+      "grad_norm": 0.5499718189239502,
+      "learning_rate": 2.8954098801992193e-05,
+      "loss": 3.1798,
+      "step": 7104
+    },
+    {
+      "epoch": 9.08576,
+      "grad_norm": 0.53836590051651,
+      "learning_rate": 2.8913716516354826e-05,
+      "loss": 3.2279,
+      "step": 7105
+    },
+    {
+      "epoch": 9.08704,
+      "grad_norm": 0.5385804772377014,
+      "learning_rate": 2.8873334230717453e-05,
+      "loss": 3.2543,
+      "step": 7106
+    },
+    {
+      "epoch": 9.08832,
+      "grad_norm": 0.5505902767181396,
+      "learning_rate": 2.8832951945080087e-05,
+      "loss": 3.1762,
+      "step": 7107
+    },
+    {
+      "epoch": 9.0896,
+      "grad_norm": 0.5531958937644958,
+      "learning_rate": 2.879256965944272e-05,
+      "loss": 3.197,
+      "step": 7108
+    },
+    {
+      "epoch": 9.09088,
+      "grad_norm": 0.5575078725814819,
+      "learning_rate": 2.8752187373805353e-05,
+      "loss": 3.2469,
+      "step": 7109
+    },
+    {
+      "epoch": 9.09216,
+      "grad_norm": 0.5386976003646851,
+      "learning_rate": 2.8711805088167987e-05,
+      "loss": 3.2092,
+      "step": 7110
+    },
+    {
+      "epoch": 9.09344,
+      "grad_norm": 0.5384331345558167,
+      "learning_rate": 2.867142280253062e-05,
+      "loss": 3.2562,
+      "step": 7111
+    },
+    {
+      "epoch": 9.09472,
+      "grad_norm": 0.549574077129364,
+      "learning_rate": 2.8631040516893254e-05,
+      "loss": 3.2311,
+      "step": 7112
+    },
+    {
+      "epoch": 9.096,
+      "grad_norm": 0.5679108500480652,
+      "learning_rate": 2.8590658231255884e-05,
+      "loss": 3.3449,
+      "step": 7113
+    },
+    {
+      "epoch": 9.09728,
+      "grad_norm": 0.5485793948173523,
+      "learning_rate": 2.8550275945618517e-05,
+      "loss": 3.2654,
+      "step": 7114
+    },
+    {
+      "epoch": 9.09856,
+      "grad_norm": 0.5479905605316162,
+      "learning_rate": 2.850989365998115e-05,
+      "loss": 3.252,
+      "step": 7115
+    },
+    {
+      "epoch": 9.09984,
+      "grad_norm": 0.5610731840133667,
+      "learning_rate": 2.8469511374343784e-05,
+      "loss": 3.2753,
+      "step": 7116
+    },
+    {
+      "epoch": 9.10112,
+      "grad_norm": 0.5504294633865356,
+      "learning_rate": 2.8429129088706418e-05,
+      "loss": 3.2685,
+      "step": 7117
+    },
+    {
+      "epoch": 9.1024,
+      "grad_norm": 0.5451478958129883,
+      "learning_rate": 2.838874680306905e-05,
+      "loss": 3.1972,
+      "step": 7118
+    },
+    {
+      "epoch": 9.10368,
+      "grad_norm": 0.5582462549209595,
+      "learning_rate": 2.8348364517431685e-05,
+      "loss": 3.2463,
+      "step": 7119
+    },
+    {
+      "epoch": 9.10496,
+      "grad_norm": 0.5593787431716919,
+      "learning_rate": 2.8307982231794315e-05,
+      "loss": 3.304,
+      "step": 7120
+    },
+    {
+      "epoch": 9.10624,
+      "grad_norm": 0.5443241596221924,
+      "learning_rate": 2.8267599946156948e-05,
+      "loss": 3.1912,
+      "step": 7121
+    },
+    {
+      "epoch": 9.10752,
+      "grad_norm": 0.52846759557724,
+      "learning_rate": 2.822721766051958e-05,
+      "loss": 3.1861,
+      "step": 7122
+    },
+    {
+      "epoch": 9.1088,
+      "grad_norm": 0.5365118384361267,
+      "learning_rate": 2.8186835374882215e-05,
+      "loss": 3.227,
+      "step": 7123
+    },
+    {
+      "epoch": 9.11008,
+      "grad_norm": 0.5490257740020752,
+      "learning_rate": 2.814645308924485e-05,
+      "loss": 3.2294,
+      "step": 7124
+    },
+    {
+      "epoch": 9.11136,
+      "grad_norm": 0.5475121140480042,
+      "learning_rate": 2.8106070803607482e-05,
+      "loss": 3.2356,
+      "step": 7125
+    },
+    {
+      "epoch": 9.11264,
+      "grad_norm": 0.5491126775741577,
+      "learning_rate": 2.8065688517970112e-05,
+      "loss": 3.2843,
+      "step": 7126
+    },
+    {
+      "epoch": 9.11392,
+      "grad_norm": 0.5404393076896667,
+      "learning_rate": 2.8025306232332746e-05,
+      "loss": 3.2315,
+      "step": 7127
+    },
+    {
+      "epoch": 9.1152,
+      "grad_norm": 0.5436875224113464,
+      "learning_rate": 2.798492394669538e-05,
+      "loss": 3.2436,
+      "step": 7128
+    },
+    {
+      "epoch": 9.11648,
+      "grad_norm": 0.5332518219947815,
+      "learning_rate": 2.7944541661058013e-05,
+      "loss": 3.2059,
+      "step": 7129
+    },
+    {
+      "epoch": 9.11776,
+      "grad_norm": 0.5293616056442261,
+      "learning_rate": 2.7904159375420646e-05,
+      "loss": 3.2773,
+      "step": 7130
+    },
+    {
+      "epoch": 9.11904,
+      "grad_norm": 0.5454232692718506,
+      "learning_rate": 2.786377708978328e-05,
+      "loss": 3.2034,
+      "step": 7131
+    },
+    {
+      "epoch": 9.12032,
+      "grad_norm": 0.5401408076286316,
+      "learning_rate": 2.7823394804145913e-05,
+      "loss": 3.1661,
+      "step": 7132
+    },
+    {
+      "epoch": 9.1216,
+      "grad_norm": 0.5544297695159912,
+      "learning_rate": 2.7783012518508543e-05,
+      "loss": 3.2872,
+      "step": 7133
+    },
+    {
+      "epoch": 9.12288,
+      "grad_norm": 0.5589653253555298,
+      "learning_rate": 2.7742630232871176e-05,
+      "loss": 3.3172,
+      "step": 7134
+    },
+    {
+      "epoch": 9.12416,
+      "grad_norm": 0.5525814294815063,
+      "learning_rate": 2.770224794723381e-05,
+      "loss": 3.2713,
+      "step": 7135
+    },
+    {
+      "epoch": 9.12544,
+      "grad_norm": 0.5349526405334473,
+      "learning_rate": 2.7661865661596443e-05,
+      "loss": 3.2127,
+      "step": 7136
+    },
+    {
+      "epoch": 9.12672,
+      "grad_norm": 0.5305134654045105,
+      "learning_rate": 2.7621483375959077e-05,
+      "loss": 3.1684,
+      "step": 7137
+    },
+    {
+      "epoch": 9.128,
+      "grad_norm": 0.5401234030723572,
+      "learning_rate": 2.758110109032171e-05,
+      "loss": 3.3105,
+      "step": 7138
+    },
+    {
+      "epoch": 9.12928,
+      "grad_norm": 0.5535856485366821,
+      "learning_rate": 2.7540718804684344e-05,
+      "loss": 3.2225,
+      "step": 7139
+    },
+    {
+      "epoch": 9.13056,
+      "grad_norm": 0.5476309061050415,
+      "learning_rate": 2.7500336519046974e-05,
+      "loss": 3.2879,
+      "step": 7140
+    },
+    {
+      "epoch": 9.13184,
+      "grad_norm": 0.5413089394569397,
+      "learning_rate": 2.7459954233409607e-05,
+      "loss": 3.1868,
+      "step": 7141
+    },
+    {
+      "epoch": 9.13312,
+      "grad_norm": 0.548028290271759,
+      "learning_rate": 2.741957194777224e-05,
+      "loss": 3.2499,
+      "step": 7142
+    },
+    {
+      "epoch": 9.1344,
+      "grad_norm": 0.5540904402732849,
+      "learning_rate": 2.7379189662134874e-05,
+      "loss": 3.2888,
+      "step": 7143
+    },
+    {
+      "epoch": 9.13568,
+      "grad_norm": 0.5426977872848511,
+      "learning_rate": 2.7338807376497508e-05,
+      "loss": 3.227,
+      "step": 7144
+    },
+    {
+      "epoch": 9.13696,
+      "grad_norm": 0.5503297448158264,
+      "learning_rate": 2.729842509086014e-05,
+      "loss": 3.1835,
+      "step": 7145
+    },
+    {
+      "epoch": 9.13824,
+      "grad_norm": 0.5338129997253418,
+      "learning_rate": 2.725804280522277e-05,
+      "loss": 3.257,
+      "step": 7146
+    },
+    {
+      "epoch": 9.13952,
+      "grad_norm": 0.5652978420257568,
+      "learning_rate": 2.7217660519585405e-05,
+      "loss": 3.2995,
+      "step": 7147
+    },
+    {
+      "epoch": 9.1408,
+      "grad_norm": 0.5556164979934692,
+      "learning_rate": 2.7177278233948038e-05,
+      "loss": 3.2905,
+      "step": 7148
+    },
+    {
+      "epoch": 9.14208,
+      "grad_norm": 0.5410715341567993,
+      "learning_rate": 2.713689594831067e-05,
+      "loss": 3.3382,
+      "step": 7149
+    },
+    {
+      "epoch": 9.14336,
+      "grad_norm": 0.546202540397644,
+      "learning_rate": 2.7096513662673305e-05,
+      "loss": 3.321,
+      "step": 7150
+    },
+    {
+      "epoch": 9.14464,
+      "grad_norm": 0.5364200472831726,
+      "learning_rate": 2.705613137703594e-05,
+      "loss": 3.2471,
+      "step": 7151
+    },
+    {
+      "epoch": 9.14592,
+      "grad_norm": 0.5380752086639404,
+      "learning_rate": 2.7015749091398572e-05,
+      "loss": 3.3008,
+      "step": 7152
+    },
+    {
+      "epoch": 9.1472,
+      "grad_norm": 0.538922131061554,
+      "learning_rate": 2.6975366805761202e-05,
+      "loss": 3.2591,
+      "step": 7153
+    },
+    {
+      "epoch": 9.14848,
+      "grad_norm": 0.5580816864967346,
+      "learning_rate": 2.6934984520123836e-05,
+      "loss": 3.2522,
+      "step": 7154
+    },
+    {
+      "epoch": 9.14976,
+      "grad_norm": 0.5336586236953735,
+      "learning_rate": 2.689460223448647e-05,
+      "loss": 3.1542,
+      "step": 7155
+    },
+    {
+      "epoch": 9.15104,
+      "grad_norm": 0.5401532649993896,
+      "learning_rate": 2.6854219948849103e-05,
+      "loss": 3.1945,
+      "step": 7156
+    },
+    {
+      "epoch": 9.15232,
+      "grad_norm": 0.5493281483650208,
+      "learning_rate": 2.6813837663211736e-05,
+      "loss": 3.3135,
+      "step": 7157
+    },
+    {
+      "epoch": 9.1536,
+      "grad_norm": 0.545318603515625,
+      "learning_rate": 2.677345537757437e-05,
+      "loss": 3.2359,
+      "step": 7158
+    },
+    {
+      "epoch": 9.15488,
+      "grad_norm": 0.5534164309501648,
+      "learning_rate": 2.6733073091937003e-05,
+      "loss": 3.2863,
+      "step": 7159
+    },
+    {
+      "epoch": 9.15616,
+      "grad_norm": 0.547370970249176,
+      "learning_rate": 2.6692690806299633e-05,
+      "loss": 3.2479,
+      "step": 7160
+    },
+    {
+      "epoch": 9.15744,
+      "grad_norm": 0.5454353094100952,
+      "learning_rate": 2.6652308520662266e-05,
+      "loss": 3.2092,
+      "step": 7161
+    },
+    {
+      "epoch": 9.15872,
+      "grad_norm": 0.5439633727073669,
+      "learning_rate": 2.66119262350249e-05,
+      "loss": 3.2338,
+      "step": 7162
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 0.5375586152076721,
+      "learning_rate": 2.6571543949387533e-05,
+      "loss": 3.2332,
+      "step": 7163
+    },
+    {
+      "epoch": 9.16128,
+      "grad_norm": 0.5413592457771301,
+      "learning_rate": 2.6531161663750167e-05,
+      "loss": 3.1831,
+      "step": 7164
+    },
+    {
+      "epoch": 9.16256,
+      "grad_norm": 0.5374849438667297,
+      "learning_rate": 2.64907793781128e-05,
+      "loss": 3.2544,
+      "step": 7165
+    },
+    {
+      "epoch": 9.16384,
+      "grad_norm": 0.5411040782928467,
+      "learning_rate": 2.6450397092475434e-05,
+      "loss": 3.2554,
+      "step": 7166
+    },
+    {
+      "epoch": 9.16512,
+      "grad_norm": 0.5422086119651794,
+      "learning_rate": 2.6410014806838064e-05,
+      "loss": 3.1815,
+      "step": 7167
+    },
+    {
+      "epoch": 9.1664,
+      "grad_norm": 0.5530745983123779,
+      "learning_rate": 2.6369632521200697e-05,
+      "loss": 3.329,
+      "step": 7168
+    },
+    {
+      "epoch": 9.16768,
+      "grad_norm": 0.535277247428894,
+      "learning_rate": 2.632925023556333e-05,
+      "loss": 3.1514,
+      "step": 7169
+    },
+    {
+      "epoch": 9.16896,
+      "grad_norm": 0.5430591106414795,
+      "learning_rate": 2.6288867949925964e-05,
+      "loss": 3.2385,
+      "step": 7170
+    },
+    {
+      "epoch": 9.17024,
+      "grad_norm": 0.5327318906784058,
+      "learning_rate": 2.6248485664288598e-05,
+      "loss": 3.2665,
+      "step": 7171
+    },
+    {
+      "epoch": 9.17152,
+      "grad_norm": 0.5375852584838867,
+      "learning_rate": 2.620810337865123e-05,
+      "loss": 3.1724,
+      "step": 7172
+    },
+    {
+      "epoch": 9.1728,
+      "grad_norm": 0.5394271612167358,
+      "learning_rate": 2.616772109301386e-05,
+      "loss": 3.1937,
+      "step": 7173
+    },
+    {
+      "epoch": 9.17408,
+      "grad_norm": 0.5466704368591309,
+      "learning_rate": 2.6127338807376495e-05,
+      "loss": 3.1974,
+      "step": 7174
+    },
+    {
+      "epoch": 9.17536,
+      "grad_norm": 0.5488426685333252,
+      "learning_rate": 2.6086956521739128e-05,
+      "loss": 3.2674,
+      "step": 7175
+    },
+    {
+      "epoch": 9.17664,
+      "grad_norm": 0.5353536009788513,
+      "learning_rate": 2.604657423610176e-05,
+      "loss": 3.2576,
+      "step": 7176
+    },
+    {
+      "epoch": 9.17792,
+      "grad_norm": 0.5377178192138672,
+      "learning_rate": 2.6006191950464395e-05,
+      "loss": 3.2777,
+      "step": 7177
+    },
+    {
+      "epoch": 9.1792,
+      "grad_norm": 0.5338087677955627,
+      "learning_rate": 2.596580966482703e-05,
+      "loss": 3.2165,
+      "step": 7178
+    },
+    {
+      "epoch": 9.18048,
+      "grad_norm": 0.5402014851570129,
+      "learning_rate": 2.5925427379189662e-05,
+      "loss": 3.2682,
+      "step": 7179
+    },
+    {
+      "epoch": 9.18176,
+      "grad_norm": 0.5575425028800964,
+      "learning_rate": 2.5885045093552292e-05,
+      "loss": 3.2025,
+      "step": 7180
+    },
+    {
+      "epoch": 9.18304,
+      "grad_norm": 0.5500099658966064,
+      "learning_rate": 2.5844662807914926e-05,
+      "loss": 3.2428,
+      "step": 7181
+    },
+    {
+      "epoch": 9.18432,
+      "grad_norm": 0.5393756628036499,
+      "learning_rate": 2.580428052227756e-05,
+      "loss": 3.3224,
+      "step": 7182
+    },
+    {
+      "epoch": 9.1856,
+      "grad_norm": 0.5401633381843567,
+      "learning_rate": 2.5763898236640192e-05,
+      "loss": 3.1761,
+      "step": 7183
+    },
+    {
+      "epoch": 9.18688,
+      "grad_norm": 0.5423944592475891,
+      "learning_rate": 2.5723515951002826e-05,
+      "loss": 3.2521,
+      "step": 7184
+    },
+    {
+      "epoch": 9.18816,
+      "grad_norm": 0.5446013808250427,
+      "learning_rate": 2.568313366536546e-05,
+      "loss": 3.2659,
+      "step": 7185
+    },
+    {
+      "epoch": 9.18944,
+      "grad_norm": 0.5417162775993347,
+      "learning_rate": 2.5642751379728093e-05,
+      "loss": 3.2457,
+      "step": 7186
+    },
+    {
+      "epoch": 9.19072,
+      "grad_norm": 0.5435442328453064,
+      "learning_rate": 2.5602369094090723e-05,
+      "loss": 3.2624,
+      "step": 7187
+    },
+    {
+      "epoch": 9.192,
+      "grad_norm": 0.5303104519844055,
+      "learning_rate": 2.5561986808453356e-05,
+      "loss": 3.2494,
+      "step": 7188
+    },
+    {
+      "epoch": 9.19328,
+      "grad_norm": 0.5464146137237549,
+      "learning_rate": 2.552160452281599e-05,
+      "loss": 3.2261,
+      "step": 7189
+    },
+    {
+      "epoch": 9.19456,
+      "grad_norm": 0.5271885395050049,
+      "learning_rate": 2.5481222237178623e-05,
+      "loss": 3.2895,
+      "step": 7190
+    },
+    {
+      "epoch": 9.19584,
+      "grad_norm": 0.5422387719154358,
+      "learning_rate": 2.5440839951541257e-05,
+      "loss": 3.2853,
+      "step": 7191
+    },
+    {
+      "epoch": 9.19712,
+      "grad_norm": 0.533074676990509,
+      "learning_rate": 2.540045766590389e-05,
+      "loss": 3.2502,
+      "step": 7192
+    },
+    {
+      "epoch": 9.1984,
+      "grad_norm": 0.5399423837661743,
+      "learning_rate": 2.536007538026652e-05,
+      "loss": 3.2562,
+      "step": 7193
+    },
+    {
+      "epoch": 9.19968,
+      "grad_norm": 0.5299935936927795,
+      "learning_rate": 2.5319693094629154e-05,
+      "loss": 3.1815,
+      "step": 7194
+    },
+    {
+      "epoch": 9.20096,
+      "grad_norm": 0.538884162902832,
+      "learning_rate": 2.5279310808991787e-05,
+      "loss": 3.1976,
+      "step": 7195
+    },
+    {
+      "epoch": 9.20224,
+      "grad_norm": 0.551018238067627,
+      "learning_rate": 2.523892852335442e-05,
+      "loss": 3.2776,
+      "step": 7196
+    },
+    {
+      "epoch": 9.20352,
+      "grad_norm": 0.5342896580696106,
+      "learning_rate": 2.5198546237717054e-05,
+      "loss": 3.1877,
+      "step": 7197
+    },
+    {
+      "epoch": 9.2048,
+      "grad_norm": 0.5489623546600342,
+      "learning_rate": 2.5158163952079688e-05,
+      "loss": 3.2537,
+      "step": 7198
+    },
+    {
+      "epoch": 9.20608,
+      "grad_norm": 0.5403774380683899,
+      "learning_rate": 2.511778166644232e-05,
+      "loss": 3.2378,
+      "step": 7199
+    },
+    {
+      "epoch": 9.20736,
+      "grad_norm": 0.5302898287773132,
+      "learning_rate": 2.5077399380804948e-05,
+      "loss": 3.1942,
+      "step": 7200
+    },
+    {
+      "epoch": 9.20864,
+      "grad_norm": 0.5362948179244995,
+      "learning_rate": 2.5037017095167585e-05,
+      "loss": 3.2845,
+      "step": 7201
+    },
+    {
+      "epoch": 9.20992,
+      "grad_norm": 0.5381386280059814,
+      "learning_rate": 2.4996634809530218e-05,
+      "loss": 3.2257,
+      "step": 7202
+    },
+    {
+      "epoch": 9.2112,
+      "grad_norm": 0.5591046214103699,
+      "learning_rate": 2.495625252389285e-05,
+      "loss": 3.2319,
+      "step": 7203
+    },
+    {
+      "epoch": 9.21248,
+      "grad_norm": 0.5458407998085022,
+      "learning_rate": 2.4915870238255485e-05,
+      "loss": 3.24,
+      "step": 7204
+    },
+    {
+      "epoch": 9.21376,
+      "grad_norm": 0.5337607860565186,
+      "learning_rate": 2.487548795261812e-05,
+      "loss": 3.3132,
+      "step": 7205
+    },
+    {
+      "epoch": 9.21504,
+      "grad_norm": 0.5361645817756653,
+      "learning_rate": 2.4835105666980752e-05,
+      "loss": 3.2828,
+      "step": 7206
+    },
+    {
+      "epoch": 9.21632,
+      "grad_norm": 0.5330992937088013,
+      "learning_rate": 2.479472338134338e-05,
+      "loss": 3.2427,
+      "step": 7207
+    },
+    {
+      "epoch": 9.2176,
+      "grad_norm": 0.5395091772079468,
+      "learning_rate": 2.4754341095706012e-05,
+      "loss": 3.2042,
+      "step": 7208
+    },
+    {
+      "epoch": 9.21888,
+      "grad_norm": 0.5380380749702454,
+      "learning_rate": 2.4713958810068646e-05,
+      "loss": 3.2095,
+      "step": 7209
+    },
+    {
+      "epoch": 9.22016,
+      "grad_norm": 0.5278270244598389,
+      "learning_rate": 2.4673576524431282e-05,
+      "loss": 3.2077,
+      "step": 7210
+    },
+    {
+      "epoch": 9.22144,
+      "grad_norm": 0.5470373034477234,
+      "learning_rate": 2.4633194238793916e-05,
+      "loss": 3.1724,
+      "step": 7211
+    },
+    {
+      "epoch": 9.22272,
+      "grad_norm": 0.5408812165260315,
+      "learning_rate": 2.459281195315655e-05,
+      "loss": 3.3655,
+      "step": 7212
+    },
+    {
+      "epoch": 9.224,
+      "grad_norm": 0.5187749266624451,
+      "learning_rate": 2.4552429667519183e-05,
+      "loss": 3.2324,
+      "step": 7213
+    },
+    {
+      "epoch": 9.22528,
+      "grad_norm": 0.5471019148826599,
+      "learning_rate": 2.451204738188181e-05,
+      "loss": 3.227,
+      "step": 7214
+    },
+    {
+      "epoch": 9.22656,
+      "grad_norm": 0.5407528281211853,
+      "learning_rate": 2.4471665096244443e-05,
+      "loss": 3.2282,
+      "step": 7215
+    },
+    {
+      "epoch": 9.22784,
+      "grad_norm": 0.5427590608596802,
+      "learning_rate": 2.4431282810607076e-05,
+      "loss": 3.2541,
+      "step": 7216
+    },
+    {
+      "epoch": 9.22912,
+      "grad_norm": 0.5527137517929077,
+      "learning_rate": 2.439090052496971e-05,
+      "loss": 3.3002,
+      "step": 7217
+    },
+    {
+      "epoch": 9.2304,
+      "grad_norm": 0.5542083978652954,
+      "learning_rate": 2.4350518239332343e-05,
+      "loss": 3.2398,
+      "step": 7218
+    },
+    {
+      "epoch": 9.23168,
+      "grad_norm": 0.5390406250953674,
+      "learning_rate": 2.431013595369498e-05,
+      "loss": 3.2449,
+      "step": 7219
+    },
+    {
+      "epoch": 9.23296,
+      "grad_norm": 0.5343725681304932,
+      "learning_rate": 2.4269753668057607e-05,
+      "loss": 3.205,
+      "step": 7220
+    },
+    {
+      "epoch": 9.23424,
+      "grad_norm": 0.53874671459198,
+      "learning_rate": 2.422937138242024e-05,
+      "loss": 3.2123,
+      "step": 7221
+    },
+    {
+      "epoch": 9.23552,
+      "grad_norm": 0.5471300482749939,
+      "learning_rate": 2.4188989096782874e-05,
+      "loss": 3.2171,
+      "step": 7222
+    },
+    {
+      "epoch": 9.2368,
+      "grad_norm": 0.5401138067245483,
+      "learning_rate": 2.4148606811145507e-05,
+      "loss": 3.2517,
+      "step": 7223
+    },
+    {
+      "epoch": 9.23808,
+      "grad_norm": 0.5442971587181091,
+      "learning_rate": 2.410822452550814e-05,
+      "loss": 3.226,
+      "step": 7224
+    },
+    {
+      "epoch": 9.23936,
+      "grad_norm": 0.5361701846122742,
+      "learning_rate": 2.4067842239870774e-05,
+      "loss": 3.2387,
+      "step": 7225
+    },
+    {
+      "epoch": 9.24064,
+      "grad_norm": 0.5273582935333252,
+      "learning_rate": 2.4027459954233408e-05,
+      "loss": 3.2343,
+      "step": 7226
+    },
+    {
+      "epoch": 9.24192,
+      "grad_norm": 0.5558703541755676,
+      "learning_rate": 2.3987077668596038e-05,
+      "loss": 3.291,
+      "step": 7227
+    },
+    {
+      "epoch": 9.2432,
+      "grad_norm": 0.5563510656356812,
+      "learning_rate": 2.394669538295867e-05,
+      "loss": 3.2446,
+      "step": 7228
+    },
+    {
+      "epoch": 9.24448,
+      "grad_norm": 0.5549482107162476,
+      "learning_rate": 2.3906313097321305e-05,
+      "loss": 3.1956,
+      "step": 7229
+    },
+    {
+      "epoch": 9.24576,
+      "grad_norm": 0.5281810760498047,
+      "learning_rate": 2.3865930811683938e-05,
+      "loss": 3.3095,
+      "step": 7230
+    },
+    {
+      "epoch": 9.24704,
+      "grad_norm": 0.5421765446662903,
+      "learning_rate": 2.382554852604657e-05,
+      "loss": 3.2243,
+      "step": 7231
+    },
+    {
+      "epoch": 9.24832,
+      "grad_norm": 0.5279957056045532,
+      "learning_rate": 2.3785166240409205e-05,
+      "loss": 3.2774,
+      "step": 7232
+    },
+    {
+      "epoch": 9.2496,
+      "grad_norm": 0.5412895679473877,
+      "learning_rate": 2.374478395477184e-05,
+      "loss": 3.195,
+      "step": 7233
+    },
+    {
+      "epoch": 9.25088,
+      "grad_norm": 0.5304553508758545,
+      "learning_rate": 2.370440166913447e-05,
+      "loss": 3.2945,
+      "step": 7234
+    },
+    {
+      "epoch": 9.25216,
+      "grad_norm": 0.536499559879303,
+      "learning_rate": 2.3664019383497102e-05,
+      "loss": 3.1877,
+      "step": 7235
+    },
+    {
+      "epoch": 9.25344,
+      "grad_norm": 0.5327797532081604,
+      "learning_rate": 2.3623637097859736e-05,
+      "loss": 3.186,
+      "step": 7236
+    },
+    {
+      "epoch": 9.25472,
+      "grad_norm": 0.5580500960350037,
+      "learning_rate": 2.358325481222237e-05,
+      "loss": 3.2472,
+      "step": 7237
+    },
+    {
+      "epoch": 9.256,
+      "grad_norm": 0.5418148040771484,
+      "learning_rate": 2.3542872526585002e-05,
+      "loss": 3.293,
+      "step": 7238
+    },
+    {
+      "epoch": 9.25728,
+      "grad_norm": 0.5535669922828674,
+      "learning_rate": 2.3502490240947636e-05,
+      "loss": 3.247,
+      "step": 7239
+    },
+    {
+      "epoch": 9.25856,
+      "grad_norm": 0.5439664721488953,
+      "learning_rate": 2.3462107955310266e-05,
+      "loss": 3.2334,
+      "step": 7240
+    },
+    {
+      "epoch": 9.25984,
+      "grad_norm": 0.5313609838485718,
+      "learning_rate": 2.34217256696729e-05,
+      "loss": 3.1706,
+      "step": 7241
+    },
+    {
+      "epoch": 9.26112,
+      "grad_norm": 0.5202425122261047,
+      "learning_rate": 2.3381343384035533e-05,
+      "loss": 3.3,
+      "step": 7242
+    },
+    {
+      "epoch": 9.2624,
+      "grad_norm": 0.5400013327598572,
+      "learning_rate": 2.3340961098398166e-05,
+      "loss": 3.2212,
+      "step": 7243
+    },
+    {
+      "epoch": 9.26368,
+      "grad_norm": 0.5324732661247253,
+      "learning_rate": 2.33005788127608e-05,
+      "loss": 3.1796,
+      "step": 7244
+    },
+    {
+      "epoch": 9.26496,
+      "grad_norm": 0.5566893815994263,
+      "learning_rate": 2.3260196527123433e-05,
+      "loss": 3.2211,
+      "step": 7245
+    },
+    {
+      "epoch": 9.26624,
+      "grad_norm": 0.5362138152122498,
+      "learning_rate": 2.3219814241486067e-05,
+      "loss": 3.2751,
+      "step": 7246
+    },
+    {
+      "epoch": 9.26752,
+      "grad_norm": 0.5280265808105469,
+      "learning_rate": 2.3179431955848697e-05,
+      "loss": 3.3033,
+      "step": 7247
+    },
+    {
+      "epoch": 9.2688,
+      "grad_norm": 0.5292081236839294,
+      "learning_rate": 2.313904967021133e-05,
+      "loss": 3.2031,
+      "step": 7248
+    },
+    {
+      "epoch": 9.27008,
+      "grad_norm": 0.529188334941864,
+      "learning_rate": 2.3098667384573964e-05,
+      "loss": 3.1846,
+      "step": 7249
+    },
+    {
+      "epoch": 9.27136,
+      "grad_norm": 0.5365098118782043,
+      "learning_rate": 2.3058285098936597e-05,
+      "loss": 3.2667,
+      "step": 7250
+    },
+    {
+      "epoch": 9.272639999999999,
+      "grad_norm": 0.5411892533302307,
+      "learning_rate": 2.301790281329923e-05,
+      "loss": 3.2621,
+      "step": 7251
+    },
+    {
+      "epoch": 9.27392,
+      "grad_norm": 0.5581626296043396,
+      "learning_rate": 2.2977520527661864e-05,
+      "loss": 3.2695,
+      "step": 7252
+    },
+    {
+      "epoch": 9.2752,
+      "grad_norm": 0.5382561087608337,
+      "learning_rate": 2.2937138242024498e-05,
+      "loss": 3.1889,
+      "step": 7253
+    },
+    {
+      "epoch": 9.27648,
+      "grad_norm": 0.5616065859794617,
+      "learning_rate": 2.2896755956387128e-05,
+      "loss": 3.3357,
+      "step": 7254
+    },
+    {
+      "epoch": 9.27776,
+      "grad_norm": 0.5404537320137024,
+      "learning_rate": 2.285637367074976e-05,
+      "loss": 3.2573,
+      "step": 7255
+    },
+    {
+      "epoch": 9.27904,
+      "grad_norm": 0.5330450534820557,
+      "learning_rate": 2.2815991385112395e-05,
+      "loss": 3.1843,
+      "step": 7256
+    },
+    {
+      "epoch": 9.28032,
+      "grad_norm": 0.5430814027786255,
+      "learning_rate": 2.2775609099475028e-05,
+      "loss": 3.1601,
+      "step": 7257
+    },
+    {
+      "epoch": 9.2816,
+      "grad_norm": 0.5439402461051941,
+      "learning_rate": 2.273522681383766e-05,
+      "loss": 3.2283,
+      "step": 7258
+    },
+    {
+      "epoch": 9.28288,
+      "grad_norm": 0.5354278087615967,
+      "learning_rate": 2.2694844528200295e-05,
+      "loss": 3.1896,
+      "step": 7259
+    },
+    {
+      "epoch": 9.28416,
+      "grad_norm": 0.5329548120498657,
+      "learning_rate": 2.265446224256293e-05,
+      "loss": 3.278,
+      "step": 7260
+    },
+    {
+      "epoch": 9.28544,
+      "grad_norm": 0.5459061861038208,
+      "learning_rate": 2.261407995692556e-05,
+      "loss": 3.2793,
+      "step": 7261
+    },
+    {
+      "epoch": 9.28672,
+      "grad_norm": 0.5458545684814453,
+      "learning_rate": 2.2573697671288192e-05,
+      "loss": 3.2462,
+      "step": 7262
+    },
+    {
+      "epoch": 9.288,
+      "grad_norm": 0.5448905229568481,
+      "learning_rate": 2.2533315385650826e-05,
+      "loss": 3.1968,
+      "step": 7263
+    },
+    {
+      "epoch": 9.28928,
+      "grad_norm": 0.5559900403022766,
+      "learning_rate": 2.249293310001346e-05,
+      "loss": 3.21,
+      "step": 7264
+    },
+    {
+      "epoch": 9.29056,
+      "grad_norm": 0.5367174744606018,
+      "learning_rate": 2.2452550814376092e-05,
+      "loss": 3.137,
+      "step": 7265
+    },
+    {
+      "epoch": 9.29184,
+      "grad_norm": 0.5573728084564209,
+      "learning_rate": 2.2412168528738726e-05,
+      "loss": 3.2597,
+      "step": 7266
+    },
+    {
+      "epoch": 9.29312,
+      "grad_norm": 0.5518680214881897,
+      "learning_rate": 2.2371786243101356e-05,
+      "loss": 3.3036,
+      "step": 7267
+    },
+    {
+      "epoch": 9.2944,
+      "grad_norm": 0.5395819544792175,
+      "learning_rate": 2.233140395746399e-05,
+      "loss": 3.2503,
+      "step": 7268
+    },
+    {
+      "epoch": 9.29568,
+      "grad_norm": 0.528838038444519,
+      "learning_rate": 2.2291021671826623e-05,
+      "loss": 3.2374,
+      "step": 7269
+    },
+    {
+      "epoch": 9.29696,
+      "grad_norm": 0.5213932394981384,
+      "learning_rate": 2.2250639386189256e-05,
+      "loss": 3.2252,
+      "step": 7270
+    },
+    {
+      "epoch": 9.29824,
+      "grad_norm": 0.5346904993057251,
+      "learning_rate": 2.221025710055189e-05,
+      "loss": 3.2775,
+      "step": 7271
+    },
+    {
+      "epoch": 9.29952,
+      "grad_norm": 0.5518351197242737,
+      "learning_rate": 2.2169874814914523e-05,
+      "loss": 3.2446,
+      "step": 7272
+    },
+    {
+      "epoch": 9.3008,
+      "grad_norm": 0.5426953434944153,
+      "learning_rate": 2.2129492529277157e-05,
+      "loss": 3.2718,
+      "step": 7273
+    },
+    {
+      "epoch": 9.30208,
+      "grad_norm": 0.5330793261528015,
+      "learning_rate": 2.2089110243639787e-05,
+      "loss": 3.2078,
+      "step": 7274
+    },
+    {
+      "epoch": 9.30336,
+      "grad_norm": 0.5355795621871948,
+      "learning_rate": 2.204872795800242e-05,
+      "loss": 3.2268,
+      "step": 7275
+    },
+    {
+      "epoch": 9.30464,
+      "grad_norm": 0.5403215885162354,
+      "learning_rate": 2.2008345672365054e-05,
+      "loss": 3.2271,
+      "step": 7276
+    },
+    {
+      "epoch": 9.30592,
+      "grad_norm": 0.5434122085571289,
+      "learning_rate": 2.1967963386727687e-05,
+      "loss": 3.2366,
+      "step": 7277
+    },
+    {
+      "epoch": 9.3072,
+      "grad_norm": 0.5365995168685913,
+      "learning_rate": 2.192758110109032e-05,
+      "loss": 3.2595,
+      "step": 7278
+    },
+    {
+      "epoch": 9.30848,
+      "grad_norm": 0.537118136882782,
+      "learning_rate": 2.1887198815452954e-05,
+      "loss": 3.2239,
+      "step": 7279
+    },
+    {
+      "epoch": 9.30976,
+      "grad_norm": 0.5411548614501953,
+      "learning_rate": 2.1846816529815588e-05,
+      "loss": 3.3102,
+      "step": 7280
+    },
+    {
+      "epoch": 9.31104,
+      "grad_norm": 0.5356780290603638,
+      "learning_rate": 2.1806434244178218e-05,
+      "loss": 3.3315,
+      "step": 7281
+    },
+    {
+      "epoch": 9.31232,
+      "grad_norm": 0.5491350889205933,
+      "learning_rate": 2.176605195854085e-05,
+      "loss": 3.3068,
+      "step": 7282
+    },
+    {
+      "epoch": 9.3136,
+      "grad_norm": 0.5464149713516235,
+      "learning_rate": 2.1725669672903485e-05,
+      "loss": 3.2792,
+      "step": 7283
+    },
+    {
+      "epoch": 9.31488,
+      "grad_norm": 0.5239377617835999,
+      "learning_rate": 2.1685287387266118e-05,
+      "loss": 3.1874,
+      "step": 7284
+    },
+    {
+      "epoch": 9.31616,
+      "grad_norm": 0.5476245284080505,
+      "learning_rate": 2.164490510162875e-05,
+      "loss": 3.2341,
+      "step": 7285
+    },
+    {
+      "epoch": 9.31744,
+      "grad_norm": 0.546082079410553,
+      "learning_rate": 2.1604522815991385e-05,
+      "loss": 3.237,
+      "step": 7286
+    },
+    {
+      "epoch": 9.31872,
+      "grad_norm": 0.5425671339035034,
+      "learning_rate": 2.1564140530354015e-05,
+      "loss": 3.2401,
+      "step": 7287
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 0.5253366827964783,
+      "learning_rate": 2.152375824471665e-05,
+      "loss": 3.2288,
+      "step": 7288
+    },
+    {
+      "epoch": 9.32128,
+      "grad_norm": 0.5394009351730347,
+      "learning_rate": 2.1483375959079282e-05,
+      "loss": 3.2209,
+      "step": 7289
+    },
+    {
+      "epoch": 9.32256,
+      "grad_norm": 0.5472865104675293,
+      "learning_rate": 2.1442993673441915e-05,
+      "loss": 3.2802,
+      "step": 7290
+    },
+    {
+      "epoch": 9.32384,
+      "grad_norm": 0.5333254933357239,
+      "learning_rate": 2.140261138780455e-05,
+      "loss": 3.2069,
+      "step": 7291
+    },
+    {
+      "epoch": 9.32512,
+      "grad_norm": 0.5474021434783936,
+      "learning_rate": 2.1362229102167182e-05,
+      "loss": 3.2606,
+      "step": 7292
+    },
+    {
+      "epoch": 9.3264,
+      "grad_norm": 0.5382992625236511,
+      "learning_rate": 2.1321846816529816e-05,
+      "loss": 3.2327,
+      "step": 7293
+    },
+    {
+      "epoch": 9.32768,
+      "grad_norm": 0.533406674861908,
+      "learning_rate": 2.1281464530892446e-05,
+      "loss": 3.2464,
+      "step": 7294
+    },
+    {
+      "epoch": 9.32896,
+      "grad_norm": 0.5368064641952515,
+      "learning_rate": 2.124108224525508e-05,
+      "loss": 3.2933,
+      "step": 7295
+    },
+    {
+      "epoch": 9.33024,
+      "grad_norm": 0.5427320599555969,
+      "learning_rate": 2.1200699959617713e-05,
+      "loss": 3.2846,
+      "step": 7296
+    },
+    {
+      "epoch": 9.33152,
+      "grad_norm": 0.5371754765510559,
+      "learning_rate": 2.1160317673980346e-05,
+      "loss": 3.3316,
+      "step": 7297
+    },
+    {
+      "epoch": 9.3328,
+      "grad_norm": 0.5448208451271057,
+      "learning_rate": 2.111993538834298e-05,
+      "loss": 3.2638,
+      "step": 7298
+    },
+    {
+      "epoch": 9.33408,
+      "grad_norm": 0.5369369983673096,
+      "learning_rate": 2.1079553102705613e-05,
+      "loss": 3.172,
+      "step": 7299
+    },
+    {
+      "epoch": 9.33536,
+      "grad_norm": 0.5618038773536682,
+      "learning_rate": 2.1039170817068247e-05,
+      "loss": 3.2398,
+      "step": 7300
+    },
+    {
+      "epoch": 9.33664,
+      "grad_norm": 0.5481551289558411,
+      "learning_rate": 2.0998788531430877e-05,
+      "loss": 3.25,
+      "step": 7301
+    },
+    {
+      "epoch": 9.33792,
+      "grad_norm": 0.5384693741798401,
+      "learning_rate": 2.095840624579351e-05,
+      "loss": 3.2027,
+      "step": 7302
+    },
+    {
+      "epoch": 9.3392,
+      "grad_norm": 0.5455001592636108,
+      "learning_rate": 2.0918023960156144e-05,
+      "loss": 3.2312,
+      "step": 7303
+    },
+    {
+      "epoch": 9.34048,
+      "grad_norm": 0.5411433577537537,
+      "learning_rate": 2.0877641674518777e-05,
+      "loss": 3.2789,
+      "step": 7304
+    },
+    {
+      "epoch": 9.34176,
+      "grad_norm": 0.5493059158325195,
+      "learning_rate": 2.083725938888141e-05,
+      "loss": 3.2225,
+      "step": 7305
+    },
+    {
+      "epoch": 9.34304,
+      "grad_norm": 0.5578423738479614,
+      "learning_rate": 2.0796877103244044e-05,
+      "loss": 3.2215,
+      "step": 7306
+    },
+    {
+      "epoch": 9.34432,
+      "grad_norm": 0.5362764596939087,
+      "learning_rate": 2.0756494817606678e-05,
+      "loss": 3.1921,
+      "step": 7307
+    },
+    {
+      "epoch": 9.3456,
+      "grad_norm": 0.5485124588012695,
+      "learning_rate": 2.0716112531969308e-05,
+      "loss": 3.2242,
+      "step": 7308
+    },
+    {
+      "epoch": 9.34688,
+      "grad_norm": 0.5391074419021606,
+      "learning_rate": 2.067573024633194e-05,
+      "loss": 3.2832,
+      "step": 7309
+    },
+    {
+      "epoch": 9.34816,
+      "grad_norm": 0.5347970128059387,
+      "learning_rate": 2.0635347960694575e-05,
+      "loss": 3.2889,
+      "step": 7310
+    },
+    {
+      "epoch": 9.34944,
+      "grad_norm": 0.5423051714897156,
+      "learning_rate": 2.0594965675057208e-05,
+      "loss": 3.2891,
+      "step": 7311
+    },
+    {
+      "epoch": 9.35072,
+      "grad_norm": 0.5293038487434387,
+      "learning_rate": 2.055458338941984e-05,
+      "loss": 3.224,
+      "step": 7312
+    },
+    {
+      "epoch": 9.352,
+      "grad_norm": 0.5548036098480225,
+      "learning_rate": 2.0514201103782475e-05,
+      "loss": 3.2699,
+      "step": 7313
+    },
+    {
+      "epoch": 9.35328,
+      "grad_norm": 0.534817636013031,
+      "learning_rate": 2.04738188181451e-05,
+      "loss": 3.1907,
+      "step": 7314
+    },
+    {
+      "epoch": 9.35456,
+      "grad_norm": 0.5452345013618469,
+      "learning_rate": 2.0433436532507735e-05,
+      "loss": 3.2226,
+      "step": 7315
+    },
+    {
+      "epoch": 9.35584,
+      "grad_norm": 0.5356956124305725,
+      "learning_rate": 2.0393054246870372e-05,
+      "loss": 3.2319,
+      "step": 7316
+    },
+    {
+      "epoch": 9.35712,
+      "grad_norm": 0.5347426533699036,
+      "learning_rate": 2.0352671961233005e-05,
+      "loss": 3.2123,
+      "step": 7317
+    },
+    {
+      "epoch": 9.3584,
+      "grad_norm": 0.533959686756134,
+      "learning_rate": 2.031228967559564e-05,
+      "loss": 3.213,
+      "step": 7318
+    },
+    {
+      "epoch": 9.35968,
+      "grad_norm": 0.549071729183197,
+      "learning_rate": 2.0271907389958272e-05,
+      "loss": 3.3334,
+      "step": 7319
+    },
+    {
+      "epoch": 9.36096,
+      "grad_norm": 0.5512160658836365,
+      "learning_rate": 2.0231525104320906e-05,
+      "loss": 3.2198,
+      "step": 7320
+    },
+    {
+      "epoch": 9.36224,
+      "grad_norm": 0.5341069102287292,
+      "learning_rate": 2.0191142818683533e-05,
+      "loss": 3.1991,
+      "step": 7321
+    },
+    {
+      "epoch": 9.36352,
+      "grad_norm": 0.536259114742279,
+      "learning_rate": 2.0150760533046166e-05,
+      "loss": 3.2656,
+      "step": 7322
+    },
+    {
+      "epoch": 9.3648,
+      "grad_norm": 0.5421985387802124,
+      "learning_rate": 2.01103782474088e-05,
+      "loss": 3.1539,
+      "step": 7323
+    },
+    {
+      "epoch": 9.36608,
+      "grad_norm": 0.5535771250724792,
+      "learning_rate": 2.0069995961771433e-05,
+      "loss": 3.2589,
+      "step": 7324
+    },
+    {
+      "epoch": 9.36736,
+      "grad_norm": 0.5412119626998901,
+      "learning_rate": 2.002961367613407e-05,
+      "loss": 3.3689,
+      "step": 7325
+    },
+    {
+      "epoch": 9.36864,
+      "grad_norm": 0.5178043842315674,
+      "learning_rate": 1.9989231390496703e-05,
+      "loss": 3.177,
+      "step": 7326
+    },
+    {
+      "epoch": 9.36992,
+      "grad_norm": 0.5386024117469788,
+      "learning_rate": 1.9948849104859337e-05,
+      "loss": 3.23,
+      "step": 7327
+    },
+    {
+      "epoch": 9.3712,
+      "grad_norm": 0.5399037599563599,
+      "learning_rate": 1.9908466819221963e-05,
+      "loss": 3.274,
+      "step": 7328
+    },
+    {
+      "epoch": 9.37248,
+      "grad_norm": 0.5423693060874939,
+      "learning_rate": 1.9868084533584597e-05,
+      "loss": 3.1935,
+      "step": 7329
+    },
+    {
+      "epoch": 9.37376,
+      "grad_norm": 0.5459950566291809,
+      "learning_rate": 1.982770224794723e-05,
+      "loss": 3.1973,
+      "step": 7330
+    },
+    {
+      "epoch": 9.37504,
+      "grad_norm": 0.5486062169075012,
+      "learning_rate": 1.9787319962309864e-05,
+      "loss": 3.1877,
+      "step": 7331
+    },
+    {
+      "epoch": 9.37632,
+      "grad_norm": 0.552658200263977,
+      "learning_rate": 1.9746937676672497e-05,
+      "loss": 3.2007,
+      "step": 7332
+    },
+    {
+      "epoch": 9.3776,
+      "grad_norm": 0.5424798130989075,
+      "learning_rate": 1.970655539103513e-05,
+      "loss": 3.2671,
+      "step": 7333
+    },
+    {
+      "epoch": 9.37888,
+      "grad_norm": 0.5408488512039185,
+      "learning_rate": 1.9666173105397768e-05,
+      "loss": 3.222,
+      "step": 7334
+    },
+    {
+      "epoch": 9.38016,
+      "grad_norm": 0.537208080291748,
+      "learning_rate": 1.9625790819760394e-05,
+      "loss": 3.3263,
+      "step": 7335
+    },
+    {
+      "epoch": 9.38144,
+      "grad_norm": 0.5427209734916687,
+      "learning_rate": 1.9585408534123028e-05,
+      "loss": 3.2509,
+      "step": 7336
+    },
+    {
+      "epoch": 9.38272,
+      "grad_norm": 0.5269715189933777,
+      "learning_rate": 1.954502624848566e-05,
+      "loss": 3.1778,
+      "step": 7337
+    },
+    {
+      "epoch": 9.384,
+      "grad_norm": 0.5249371528625488,
+      "learning_rate": 1.9504643962848295e-05,
+      "loss": 3.1646,
+      "step": 7338
+    },
+    {
+      "epoch": 9.38528,
+      "grad_norm": 0.5350185632705688,
+      "learning_rate": 1.9464261677210928e-05,
+      "loss": 3.1933,
+      "step": 7339
+    },
+    {
+      "epoch": 9.38656,
+      "grad_norm": 0.5533794164657593,
+      "learning_rate": 1.942387939157356e-05,
+      "loss": 3.3036,
+      "step": 7340
+    },
+    {
+      "epoch": 9.38784,
+      "grad_norm": 0.537314236164093,
+      "learning_rate": 1.938349710593619e-05,
+      "loss": 3.266,
+      "step": 7341
+    },
+    {
+      "epoch": 9.38912,
+      "grad_norm": 0.5408215522766113,
+      "learning_rate": 1.9343114820298825e-05,
+      "loss": 3.2319,
+      "step": 7342
+    },
+    {
+      "epoch": 9.3904,
+      "grad_norm": 0.5351347923278809,
+      "learning_rate": 1.930273253466146e-05,
+      "loss": 3.2069,
+      "step": 7343
+    },
+    {
+      "epoch": 9.39168,
+      "grad_norm": 0.56521075963974,
+      "learning_rate": 1.9262350249024092e-05,
+      "loss": 3.2888,
+      "step": 7344
+    },
+    {
+      "epoch": 9.39296,
+      "grad_norm": 0.5516117811203003,
+      "learning_rate": 1.9221967963386725e-05,
+      "loss": 3.2223,
+      "step": 7345
+    },
+    {
+      "epoch": 9.39424,
+      "grad_norm": 0.5425495505332947,
+      "learning_rate": 1.918158567774936e-05,
+      "loss": 3.2373,
+      "step": 7346
+    },
+    {
+      "epoch": 9.39552,
+      "grad_norm": 0.5260109901428223,
+      "learning_rate": 1.9141203392111992e-05,
+      "loss": 3.2283,
+      "step": 7347
+    },
+    {
+      "epoch": 9.3968,
+      "grad_norm": 0.530563473701477,
+      "learning_rate": 1.9100821106474622e-05,
+      "loss": 3.2014,
+      "step": 7348
+    },
+    {
+      "epoch": 9.39808,
+      "grad_norm": 0.5480346083641052,
+      "learning_rate": 1.9060438820837256e-05,
+      "loss": 3.3066,
+      "step": 7349
+    },
+    {
+      "epoch": 9.39936,
+      "grad_norm": 0.5326750874519348,
+      "learning_rate": 1.902005653519989e-05,
+      "loss": 3.2109,
+      "step": 7350
+    },
+    {
+      "epoch": 9.40064,
+      "grad_norm": 0.5301714539527893,
+      "learning_rate": 1.8979674249562523e-05,
+      "loss": 3.2464,
+      "step": 7351
+    },
+    {
+      "epoch": 9.40192,
+      "grad_norm": 0.5453739166259766,
+      "learning_rate": 1.8939291963925156e-05,
+      "loss": 3.2734,
+      "step": 7352
+    },
+    {
+      "epoch": 9.4032,
+      "grad_norm": 0.5450626611709595,
+      "learning_rate": 1.889890967828779e-05,
+      "loss": 3.2435,
+      "step": 7353
+    },
+    {
+      "epoch": 9.40448,
+      "grad_norm": 0.5502852201461792,
+      "learning_rate": 1.8858527392650423e-05,
+      "loss": 3.3081,
+      "step": 7354
+    },
+    {
+      "epoch": 9.40576,
+      "grad_norm": 0.5369656682014465,
+      "learning_rate": 1.8818145107013053e-05,
+      "loss": 3.2784,
+      "step": 7355
+    },
+    {
+      "epoch": 9.40704,
+      "grad_norm": 0.5568560361862183,
+      "learning_rate": 1.8777762821375687e-05,
+      "loss": 3.3034,
+      "step": 7356
+    },
+    {
+      "epoch": 9.40832,
+      "grad_norm": 0.541373074054718,
+      "learning_rate": 1.873738053573832e-05,
+      "loss": 3.2594,
+      "step": 7357
+    },
+    {
+      "epoch": 9.4096,
+      "grad_norm": 0.5414525866508484,
+      "learning_rate": 1.8696998250100954e-05,
+      "loss": 3.2231,
+      "step": 7358
+    },
+    {
+      "epoch": 9.41088,
+      "grad_norm": 0.535603940486908,
+      "learning_rate": 1.8656615964463587e-05,
+      "loss": 3.2663,
+      "step": 7359
+    },
+    {
+      "epoch": 9.41216,
+      "grad_norm": 0.5469928979873657,
+      "learning_rate": 1.861623367882622e-05,
+      "loss": 3.2296,
+      "step": 7360
+    },
+    {
+      "epoch": 9.41344,
+      "grad_norm": 0.5573539733886719,
+      "learning_rate": 1.8575851393188854e-05,
+      "loss": 3.2841,
+      "step": 7361
+    },
+    {
+      "epoch": 9.414719999999999,
+      "grad_norm": 0.5386171340942383,
+      "learning_rate": 1.8535469107551488e-05,
+      "loss": 3.2654,
+      "step": 7362
+    },
+    {
+      "epoch": 9.416,
+      "grad_norm": 0.5473566651344299,
+      "learning_rate": 1.8495086821914118e-05,
+      "loss": 3.206,
+      "step": 7363
+    },
+    {
+      "epoch": 9.41728,
+      "grad_norm": 0.54440838098526,
+      "learning_rate": 1.845470453627675e-05,
+      "loss": 3.2718,
+      "step": 7364
+    },
+    {
+      "epoch": 9.41856,
+      "grad_norm": 0.5463802814483643,
+      "learning_rate": 1.8414322250639385e-05,
+      "loss": 3.268,
+      "step": 7365
+    },
+    {
+      "epoch": 9.41984,
+      "grad_norm": 0.5378354787826538,
+      "learning_rate": 1.8373939965002018e-05,
+      "loss": 3.2652,
+      "step": 7366
+    },
+    {
+      "epoch": 9.42112,
+      "grad_norm": 0.5412339568138123,
+      "learning_rate": 1.833355767936465e-05,
+      "loss": 3.1854,
+      "step": 7367
+    },
+    {
+      "epoch": 9.4224,
+      "grad_norm": 0.5401472449302673,
+      "learning_rate": 1.8293175393727285e-05,
+      "loss": 3.3367,
+      "step": 7368
+    },
+    {
+      "epoch": 9.42368,
+      "grad_norm": 0.5603813529014587,
+      "learning_rate": 1.825279310808992e-05,
+      "loss": 3.2289,
+      "step": 7369
+    },
+    {
+      "epoch": 9.42496,
+      "grad_norm": 0.5412847399711609,
+      "learning_rate": 1.821241082245255e-05,
+      "loss": 3.2606,
+      "step": 7370
+    },
+    {
+      "epoch": 9.42624,
+      "grad_norm": 0.5373712778091431,
+      "learning_rate": 1.8172028536815182e-05,
+      "loss": 3.2097,
+      "step": 7371
+    },
+    {
+      "epoch": 9.42752,
+      "grad_norm": 0.5573616623878479,
+      "learning_rate": 1.8131646251177815e-05,
+      "loss": 3.2397,
+      "step": 7372
+    },
+    {
+      "epoch": 9.4288,
+      "grad_norm": 0.5405897498130798,
+      "learning_rate": 1.8091263965540446e-05,
+      "loss": 3.259,
+      "step": 7373
+    },
+    {
+      "epoch": 9.43008,
+      "grad_norm": 0.5547781586647034,
+      "learning_rate": 1.8050881679903082e-05,
+      "loss": 3.247,
+      "step": 7374
+    },
+    {
+      "epoch": 9.43136,
+      "grad_norm": 0.5367437601089478,
+      "learning_rate": 1.8010499394265716e-05,
+      "loss": 3.25,
+      "step": 7375
+    },
+    {
+      "epoch": 9.43264,
+      "grad_norm": 0.5429502725601196,
+      "learning_rate": 1.7970117108628346e-05,
+      "loss": 3.2392,
+      "step": 7376
+    },
+    {
+      "epoch": 9.43392,
+      "grad_norm": 0.5472865700721741,
+      "learning_rate": 1.792973482299098e-05,
+      "loss": 3.2329,
+      "step": 7377
+    },
+    {
+      "epoch": 9.4352,
+      "grad_norm": 0.542381227016449,
+      "learning_rate": 1.7889352537353613e-05,
+      "loss": 3.2845,
+      "step": 7378
+    },
+    {
+      "epoch": 9.43648,
+      "grad_norm": 0.5476444959640503,
+      "learning_rate": 1.7848970251716246e-05,
+      "loss": 3.2889,
+      "step": 7379
+    },
+    {
+      "epoch": 9.43776,
+      "grad_norm": 0.5548306107521057,
+      "learning_rate": 1.7808587966078876e-05,
+      "loss": 3.2567,
+      "step": 7380
+    },
+    {
+      "epoch": 9.43904,
+      "grad_norm": 0.5338513851165771,
+      "learning_rate": 1.776820568044151e-05,
+      "loss": 3.1739,
+      "step": 7381
+    },
+    {
+      "epoch": 9.44032,
+      "grad_norm": 0.5605148673057556,
+      "learning_rate": 1.7727823394804143e-05,
+      "loss": 3.2783,
+      "step": 7382
+    },
+    {
+      "epoch": 9.4416,
+      "grad_norm": 0.5437026619911194,
+      "learning_rate": 1.7687441109166777e-05,
+      "loss": 3.1916,
+      "step": 7383
+    },
+    {
+      "epoch": 9.44288,
+      "grad_norm": 0.5323542952537537,
+      "learning_rate": 1.764705882352941e-05,
+      "loss": 3.2599,
+      "step": 7384
+    },
+    {
+      "epoch": 9.44416,
+      "grad_norm": 0.5453789830207825,
+      "learning_rate": 1.7606676537892044e-05,
+      "loss": 3.2619,
+      "step": 7385
+    },
+    {
+      "epoch": 9.44544,
+      "grad_norm": 0.5387235283851624,
+      "learning_rate": 1.7566294252254677e-05,
+      "loss": 3.2936,
+      "step": 7386
+    },
+    {
+      "epoch": 9.44672,
+      "grad_norm": 0.5475628972053528,
+      "learning_rate": 1.7525911966617307e-05,
+      "loss": 3.244,
+      "step": 7387
+    },
+    {
+      "epoch": 9.448,
+      "grad_norm": 0.5442694425582886,
+      "learning_rate": 1.748552968097994e-05,
+      "loss": 3.2876,
+      "step": 7388
+    },
+    {
+      "epoch": 9.44928,
+      "grad_norm": 0.5453842282295227,
+      "learning_rate": 1.7445147395342574e-05,
+      "loss": 3.2533,
+      "step": 7389
+    },
+    {
+      "epoch": 9.45056,
+      "grad_norm": 0.5407310128211975,
+      "learning_rate": 1.7404765109705208e-05,
+      "loss": 3.2382,
+      "step": 7390
+    },
+    {
+      "epoch": 9.45184,
+      "grad_norm": 0.531166672706604,
+      "learning_rate": 1.736438282406784e-05,
+      "loss": 3.2376,
+      "step": 7391
+    },
+    {
+      "epoch": 9.45312,
+      "grad_norm": 0.5313123464584351,
+      "learning_rate": 1.7324000538430475e-05,
+      "loss": 3.2362,
+      "step": 7392
+    },
+    {
+      "epoch": 9.4544,
+      "grad_norm": 0.5363158583641052,
+      "learning_rate": 1.7283618252793105e-05,
+      "loss": 3.2088,
+      "step": 7393
+    },
+    {
+      "epoch": 9.45568,
+      "grad_norm": 0.5450273752212524,
+      "learning_rate": 1.7243235967155738e-05,
+      "loss": 3.26,
+      "step": 7394
+    },
+    {
+      "epoch": 9.45696,
+      "grad_norm": 0.57196044921875,
+      "learning_rate": 1.720285368151837e-05,
+      "loss": 3.2556,
+      "step": 7395
+    },
+    {
+      "epoch": 9.45824,
+      "grad_norm": 0.5576905012130737,
+      "learning_rate": 1.7162471395881005e-05,
+      "loss": 3.2441,
+      "step": 7396
+    },
+    {
+      "epoch": 9.45952,
+      "grad_norm": 0.5438629388809204,
+      "learning_rate": 1.712208911024364e-05,
+      "loss": 3.1963,
+      "step": 7397
+    },
+    {
+      "epoch": 9.4608,
+      "grad_norm": 0.5366012454032898,
+      "learning_rate": 1.7081706824606272e-05,
+      "loss": 3.2966,
+      "step": 7398
+    },
+    {
+      "epoch": 9.46208,
+      "grad_norm": 0.5356854796409607,
+      "learning_rate": 1.7041324538968905e-05,
+      "loss": 3.22,
+      "step": 7399
+    },
+    {
+      "epoch": 9.46336,
+      "grad_norm": 0.5522456169128418,
+      "learning_rate": 1.7000942253331535e-05,
+      "loss": 3.2367,
+      "step": 7400
+    },
+    {
+      "epoch": 9.46464,
+      "grad_norm": 0.5475519299507141,
+      "learning_rate": 1.696055996769417e-05,
+      "loss": 3.2766,
+      "step": 7401
+    },
+    {
+      "epoch": 9.46592,
+      "grad_norm": 0.5508553385734558,
+      "learning_rate": 1.6920177682056802e-05,
+      "loss": 3.2885,
+      "step": 7402
+    },
+    {
+      "epoch": 9.4672,
+      "grad_norm": 0.5404077172279358,
+      "learning_rate": 1.6879795396419436e-05,
+      "loss": 3.2627,
+      "step": 7403
+    },
+    {
+      "epoch": 9.46848,
+      "grad_norm": 0.5281044840812683,
+      "learning_rate": 1.683941311078207e-05,
+      "loss": 3.2113,
+      "step": 7404
+    },
+    {
+      "epoch": 9.46976,
+      "grad_norm": 0.5426114201545715,
+      "learning_rate": 1.6799030825144703e-05,
+      "loss": 3.2066,
+      "step": 7405
+    },
+    {
+      "epoch": 9.47104,
+      "grad_norm": 0.5397725701332092,
+      "learning_rate": 1.6758648539507336e-05,
+      "loss": 3.2879,
+      "step": 7406
+    },
+    {
+      "epoch": 9.47232,
+      "grad_norm": 0.5343126058578491,
+      "learning_rate": 1.6718266253869966e-05,
+      "loss": 3.21,
+      "step": 7407
+    },
+    {
+      "epoch": 9.4736,
+      "grad_norm": 0.527040421962738,
+      "learning_rate": 1.66778839682326e-05,
+      "loss": 3.2397,
+      "step": 7408
+    },
+    {
+      "epoch": 9.47488,
+      "grad_norm": 0.5362902879714966,
+      "learning_rate": 1.6637501682595233e-05,
+      "loss": 3.2548,
+      "step": 7409
+    },
+    {
+      "epoch": 9.47616,
+      "grad_norm": 0.5329915285110474,
+      "learning_rate": 1.6597119396957867e-05,
+      "loss": 3.2235,
+      "step": 7410
+    },
+    {
+      "epoch": 9.47744,
+      "grad_norm": 0.5368509292602539,
+      "learning_rate": 1.65567371113205e-05,
+      "loss": 3.1814,
+      "step": 7411
+    },
+    {
+      "epoch": 9.47872,
+      "grad_norm": 0.5509939193725586,
+      "learning_rate": 1.6516354825683134e-05,
+      "loss": 3.2374,
+      "step": 7412
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 0.5431427359580994,
+      "learning_rate": 1.6475972540045764e-05,
+      "loss": 3.1822,
+      "step": 7413
+    },
+    {
+      "epoch": 9.48128,
+      "grad_norm": 0.5423570275306702,
+      "learning_rate": 1.6435590254408397e-05,
+      "loss": 3.2381,
+      "step": 7414
+    },
+    {
+      "epoch": 9.48256,
+      "grad_norm": 0.5203372240066528,
+      "learning_rate": 1.639520796877103e-05,
+      "loss": 3.238,
+      "step": 7415
+    },
+    {
+      "epoch": 9.48384,
+      "grad_norm": 0.5459218621253967,
+      "learning_rate": 1.6354825683133664e-05,
+      "loss": 3.2343,
+      "step": 7416
+    },
+    {
+      "epoch": 9.48512,
+      "grad_norm": 0.5473526120185852,
+      "learning_rate": 1.6314443397496298e-05,
+      "loss": 3.263,
+      "step": 7417
+    },
+    {
+      "epoch": 9.4864,
+      "grad_norm": 0.5523941516876221,
+      "learning_rate": 1.627406111185893e-05,
+      "loss": 3.2685,
+      "step": 7418
+    },
+    {
+      "epoch": 9.48768,
+      "grad_norm": 0.541957437992096,
+      "learning_rate": 1.6233678826221565e-05,
+      "loss": 3.3066,
+      "step": 7419
+    },
+    {
+      "epoch": 9.48896,
+      "grad_norm": 0.5321203470230103,
+      "learning_rate": 1.6193296540584195e-05,
+      "loss": 3.2308,
+      "step": 7420
+    },
+    {
+      "epoch": 9.49024,
+      "grad_norm": 0.5349709391593933,
+      "learning_rate": 1.6152914254946828e-05,
+      "loss": 3.2747,
+      "step": 7421
+    },
+    {
+      "epoch": 9.49152,
+      "grad_norm": 0.5336790084838867,
+      "learning_rate": 1.611253196930946e-05,
+      "loss": 3.2135,
+      "step": 7422
+    },
+    {
+      "epoch": 9.4928,
+      "grad_norm": 0.551705539226532,
+      "learning_rate": 1.6072149683672095e-05,
+      "loss": 3.2396,
+      "step": 7423
+    },
+    {
+      "epoch": 9.49408,
+      "grad_norm": 0.547311007976532,
+      "learning_rate": 1.603176739803473e-05,
+      "loss": 3.3115,
+      "step": 7424
+    },
+    {
+      "epoch": 9.49536,
+      "grad_norm": 0.558070182800293,
+      "learning_rate": 1.5991385112397362e-05,
+      "loss": 3.2538,
+      "step": 7425
+    },
+    {
+      "epoch": 9.49664,
+      "grad_norm": 0.5429165363311768,
+      "learning_rate": 1.5951002826759995e-05,
+      "loss": 3.1978,
+      "step": 7426
+    },
+    {
+      "epoch": 9.49792,
+      "grad_norm": 0.5304774045944214,
+      "learning_rate": 1.5910620541122625e-05,
+      "loss": 3.2507,
+      "step": 7427
+    },
+    {
+      "epoch": 9.4992,
+      "grad_norm": 0.5391626358032227,
+      "learning_rate": 1.587023825548526e-05,
+      "loss": 3.2713,
+      "step": 7428
+    },
+    {
+      "epoch": 9.50048,
+      "grad_norm": 0.5373072028160095,
+      "learning_rate": 1.5829855969847892e-05,
+      "loss": 3.2066,
+      "step": 7429
+    },
+    {
+      "epoch": 9.50176,
+      "grad_norm": 0.52620929479599,
+      "learning_rate": 1.5789473684210522e-05,
+      "loss": 3.2278,
+      "step": 7430
+    },
+    {
+      "epoch": 9.50304,
+      "grad_norm": 0.5313057899475098,
+      "learning_rate": 1.574909139857316e-05,
+      "loss": 3.2863,
+      "step": 7431
+    },
+    {
+      "epoch": 9.50432,
+      "grad_norm": 0.555780291557312,
+      "learning_rate": 1.5708709112935793e-05,
+      "loss": 3.3081,
+      "step": 7432
+    },
+    {
+      "epoch": 9.5056,
+      "grad_norm": 0.5542112588882446,
+      "learning_rate": 1.5668326827298426e-05,
+      "loss": 3.2342,
+      "step": 7433
+    },
+    {
+      "epoch": 9.50688,
+      "grad_norm": 0.5538204312324524,
+      "learning_rate": 1.5627944541661056e-05,
+      "loss": 3.2699,
+      "step": 7434
+    },
+    {
+      "epoch": 9.50816,
+      "grad_norm": 0.5440041422843933,
+      "learning_rate": 1.558756225602369e-05,
+      "loss": 3.195,
+      "step": 7435
+    },
+    {
+      "epoch": 9.50944,
+      "grad_norm": 0.529220700263977,
+      "learning_rate": 1.5547179970386323e-05,
+      "loss": 3.2317,
+      "step": 7436
+    },
+    {
+      "epoch": 9.51072,
+      "grad_norm": 0.5430290102958679,
+      "learning_rate": 1.5506797684748953e-05,
+      "loss": 3.2327,
+      "step": 7437
+    },
+    {
+      "epoch": 9.512,
+      "grad_norm": 0.5630149245262146,
+      "learning_rate": 1.5466415399111587e-05,
+      "loss": 3.3014,
+      "step": 7438
+    },
+    {
+      "epoch": 9.51328,
+      "grad_norm": 0.5409126877784729,
+      "learning_rate": 1.542603311347422e-05,
+      "loss": 3.2794,
+      "step": 7439
+    },
+    {
+      "epoch": 9.51456,
+      "grad_norm": 0.5274693369865417,
+      "learning_rate": 1.5385650827836854e-05,
+      "loss": 3.2504,
+      "step": 7440
+    },
+    {
+      "epoch": 9.51584,
+      "grad_norm": 0.5387745499610901,
+      "learning_rate": 1.5345268542199487e-05,
+      "loss": 3.1959,
+      "step": 7441
+    },
+    {
+      "epoch": 9.51712,
+      "grad_norm": 0.5413409471511841,
+      "learning_rate": 1.530488625656212e-05,
+      "loss": 3.208,
+      "step": 7442
+    },
+    {
+      "epoch": 9.5184,
+      "grad_norm": 0.5244672298431396,
+      "learning_rate": 1.5264503970924754e-05,
+      "loss": 3.1847,
+      "step": 7443
+    },
+    {
+      "epoch": 9.51968,
+      "grad_norm": 0.5482396483421326,
+      "learning_rate": 1.5224121685287386e-05,
+      "loss": 3.2254,
+      "step": 7444
+    },
+    {
+      "epoch": 9.52096,
+      "grad_norm": 0.5391595363616943,
+      "learning_rate": 1.518373939965002e-05,
+      "loss": 3.174,
+      "step": 7445
+    },
+    {
+      "epoch": 9.52224,
+      "grad_norm": 0.5382583737373352,
+      "learning_rate": 1.5143357114012653e-05,
+      "loss": 3.2514,
+      "step": 7446
+    },
+    {
+      "epoch": 9.52352,
+      "grad_norm": 0.5451508164405823,
+      "learning_rate": 1.5102974828375285e-05,
+      "loss": 3.3266,
+      "step": 7447
+    },
+    {
+      "epoch": 9.5248,
+      "grad_norm": 0.5408745408058167,
+      "learning_rate": 1.5062592542737918e-05,
+      "loss": 3.2239,
+      "step": 7448
+    },
+    {
+      "epoch": 9.52608,
+      "grad_norm": 0.5346300601959229,
+      "learning_rate": 1.5022210257100551e-05,
+      "loss": 3.1678,
+      "step": 7449
+    },
+    {
+      "epoch": 9.52736,
+      "grad_norm": 0.54719078540802,
+      "learning_rate": 1.4981827971463183e-05,
+      "loss": 3.3131,
+      "step": 7450
+    },
+    {
+      "epoch": 9.52864,
+      "grad_norm": 0.5220864415168762,
+      "learning_rate": 1.4941445685825817e-05,
+      "loss": 3.2922,
+      "step": 7451
+    },
+    {
+      "epoch": 9.52992,
+      "grad_norm": 0.538643479347229,
+      "learning_rate": 1.490106340018845e-05,
+      "loss": 3.2359,
+      "step": 7452
+    },
+    {
+      "epoch": 9.5312,
+      "grad_norm": 0.5405166745185852,
+      "learning_rate": 1.4860681114551084e-05,
+      "loss": 3.1694,
+      "step": 7453
+    },
+    {
+      "epoch": 9.53248,
+      "grad_norm": 0.5491073131561279,
+      "learning_rate": 1.4820298828913715e-05,
+      "loss": 3.2585,
+      "step": 7454
+    },
+    {
+      "epoch": 9.533760000000001,
+      "grad_norm": 0.5413920879364014,
+      "learning_rate": 1.4779916543276349e-05,
+      "loss": 3.1619,
+      "step": 7455
+    },
+    {
+      "epoch": 9.53504,
+      "grad_norm": 0.5279734134674072,
+      "learning_rate": 1.4739534257638982e-05,
+      "loss": 3.1632,
+      "step": 7456
+    },
+    {
+      "epoch": 9.53632,
+      "grad_norm": 0.5433555245399475,
+      "learning_rate": 1.4699151972001612e-05,
+      "loss": 3.2692,
+      "step": 7457
+    },
+    {
+      "epoch": 9.5376,
+      "grad_norm": 0.5342807173728943,
+      "learning_rate": 1.4658769686364248e-05,
+      "loss": 3.2832,
+      "step": 7458
+    },
+    {
+      "epoch": 9.53888,
+      "grad_norm": 0.5404999256134033,
+      "learning_rate": 1.4618387400726881e-05,
+      "loss": 3.2091,
+      "step": 7459
+    },
+    {
+      "epoch": 9.54016,
+      "grad_norm": 0.5424739718437195,
+      "learning_rate": 1.4578005115089511e-05,
+      "loss": 3.192,
+      "step": 7460
+    },
+    {
+      "epoch": 9.54144,
+      "grad_norm": 0.5365021228790283,
+      "learning_rate": 1.4537622829452145e-05,
+      "loss": 3.2372,
+      "step": 7461
+    },
+    {
+      "epoch": 9.54272,
+      "grad_norm": 0.5258615016937256,
+      "learning_rate": 1.4497240543814778e-05,
+      "loss": 3.2472,
+      "step": 7462
+    },
+    {
+      "epoch": 9.544,
+      "grad_norm": 0.543051540851593,
+      "learning_rate": 1.4456858258177413e-05,
+      "loss": 3.2782,
+      "step": 7463
+    },
+    {
+      "epoch": 9.54528,
+      "grad_norm": 0.5488424897193909,
+      "learning_rate": 1.4416475972540043e-05,
+      "loss": 3.2517,
+      "step": 7464
+    },
+    {
+      "epoch": 9.54656,
+      "grad_norm": 0.5342796444892883,
+      "learning_rate": 1.4376093686902677e-05,
+      "loss": 3.2171,
+      "step": 7465
+    },
+    {
+      "epoch": 9.54784,
+      "grad_norm": 0.5366568565368652,
+      "learning_rate": 1.433571140126531e-05,
+      "loss": 3.2127,
+      "step": 7466
+    },
+    {
+      "epoch": 9.54912,
+      "grad_norm": 0.5328598022460938,
+      "learning_rate": 1.4295329115627942e-05,
+      "loss": 3.1996,
+      "step": 7467
+    },
+    {
+      "epoch": 9.5504,
+      "grad_norm": 0.5342961549758911,
+      "learning_rate": 1.4254946829990575e-05,
+      "loss": 3.2439,
+      "step": 7468
+    },
+    {
+      "epoch": 9.55168,
+      "grad_norm": 0.5400911569595337,
+      "learning_rate": 1.4214564544353209e-05,
+      "loss": 3.2113,
+      "step": 7469
+    },
+    {
+      "epoch": 9.55296,
+      "grad_norm": 0.536578893661499,
+      "learning_rate": 1.4174182258715842e-05,
+      "loss": 3.2135,
+      "step": 7470
+    },
+    {
+      "epoch": 9.55424,
+      "grad_norm": 0.5431190729141235,
+      "learning_rate": 1.4133799973078474e-05,
+      "loss": 3.2889,
+      "step": 7471
+    },
+    {
+      "epoch": 9.55552,
+      "grad_norm": 0.554277241230011,
+      "learning_rate": 1.4093417687441108e-05,
+      "loss": 3.2935,
+      "step": 7472
+    },
+    {
+      "epoch": 9.556799999999999,
+      "grad_norm": 0.5410692095756531,
+      "learning_rate": 1.4053035401803741e-05,
+      "loss": 3.2235,
+      "step": 7473
+    },
+    {
+      "epoch": 9.55808,
+      "grad_norm": 0.5578665733337402,
+      "learning_rate": 1.4012653116166373e-05,
+      "loss": 3.246,
+      "step": 7474
+    },
+    {
+      "epoch": 9.55936,
+      "grad_norm": 0.5440622568130493,
+      "learning_rate": 1.3972270830529006e-05,
+      "loss": 3.2371,
+      "step": 7475
+    },
+    {
+      "epoch": 9.56064,
+      "grad_norm": 0.5429546236991882,
+      "learning_rate": 1.393188854489164e-05,
+      "loss": 3.1835,
+      "step": 7476
+    },
+    {
+      "epoch": 9.56192,
+      "grad_norm": 0.5361295342445374,
+      "learning_rate": 1.3891506259254272e-05,
+      "loss": 3.1896,
+      "step": 7477
+    },
+    {
+      "epoch": 9.5632,
+      "grad_norm": 0.5314315557479858,
+      "learning_rate": 1.3851123973616905e-05,
+      "loss": 3.279,
+      "step": 7478
+    },
+    {
+      "epoch": 9.56448,
+      "grad_norm": 0.5387744903564453,
+      "learning_rate": 1.3810741687979538e-05,
+      "loss": 3.2884,
+      "step": 7479
+    },
+    {
+      "epoch": 9.565760000000001,
+      "grad_norm": 0.5472334027290344,
+      "learning_rate": 1.3770359402342172e-05,
+      "loss": 3.2379,
+      "step": 7480
+    },
+    {
+      "epoch": 9.56704,
+      "grad_norm": 0.5438302755355835,
+      "learning_rate": 1.3729977116704804e-05,
+      "loss": 3.2577,
+      "step": 7481
+    },
+    {
+      "epoch": 9.56832,
+      "grad_norm": 0.5295417904853821,
+      "learning_rate": 1.3689594831067437e-05,
+      "loss": 3.3273,
+      "step": 7482
+    },
+    {
+      "epoch": 9.5696,
+      "grad_norm": 0.5329756140708923,
+      "learning_rate": 1.364921254543007e-05,
+      "loss": 3.1549,
+      "step": 7483
+    },
+    {
+      "epoch": 9.57088,
+      "grad_norm": 0.5324873328208923,
+      "learning_rate": 1.3608830259792702e-05,
+      "loss": 3.245,
+      "step": 7484
+    },
+    {
+      "epoch": 9.57216,
+      "grad_norm": 0.5475909113883972,
+      "learning_rate": 1.3568447974155336e-05,
+      "loss": 3.3146,
+      "step": 7485
+    },
+    {
+      "epoch": 9.57344,
+      "grad_norm": 0.552155077457428,
+      "learning_rate": 1.352806568851797e-05,
+      "loss": 3.22,
+      "step": 7486
+    },
+    {
+      "epoch": 9.57472,
+      "grad_norm": 0.5365369915962219,
+      "learning_rate": 1.3487683402880601e-05,
+      "loss": 3.2338,
+      "step": 7487
+    },
+    {
+      "epoch": 9.576,
+      "grad_norm": 0.5464675426483154,
+      "learning_rate": 1.3447301117243235e-05,
+      "loss": 3.3097,
+      "step": 7488
+    },
+    {
+      "epoch": 9.57728,
+      "grad_norm": 0.5443135499954224,
+      "learning_rate": 1.3406918831605868e-05,
+      "loss": 3.2514,
+      "step": 7489
+    },
+    {
+      "epoch": 9.57856,
+      "grad_norm": 0.5426671504974365,
+      "learning_rate": 1.3366536545968501e-05,
+      "loss": 3.2772,
+      "step": 7490
+    },
+    {
+      "epoch": 9.57984,
+      "grad_norm": 0.5442038774490356,
+      "learning_rate": 1.3326154260331133e-05,
+      "loss": 3.2219,
+      "step": 7491
+    },
+    {
+      "epoch": 9.58112,
+      "grad_norm": 0.5287876129150391,
+      "learning_rate": 1.3285771974693767e-05,
+      "loss": 3.246,
+      "step": 7492
+    },
+    {
+      "epoch": 9.5824,
+      "grad_norm": 0.5401532649993896,
+      "learning_rate": 1.32453896890564e-05,
+      "loss": 3.2587,
+      "step": 7493
+    },
+    {
+      "epoch": 9.58368,
+      "grad_norm": 0.538408637046814,
+      "learning_rate": 1.3205007403419032e-05,
+      "loss": 3.2394,
+      "step": 7494
+    },
+    {
+      "epoch": 9.58496,
+      "grad_norm": 0.5716665387153625,
+      "learning_rate": 1.3164625117781665e-05,
+      "loss": 3.2705,
+      "step": 7495
+    },
+    {
+      "epoch": 9.58624,
+      "grad_norm": 0.5246546864509583,
+      "learning_rate": 1.3124242832144299e-05,
+      "loss": 3.2281,
+      "step": 7496
+    },
+    {
+      "epoch": 9.58752,
+      "grad_norm": 0.5313910245895386,
+      "learning_rate": 1.308386054650693e-05,
+      "loss": 3.2713,
+      "step": 7497
+    },
+    {
+      "epoch": 9.588799999999999,
+      "grad_norm": 0.5274931788444519,
+      "learning_rate": 1.3043478260869564e-05,
+      "loss": 3.2255,
+      "step": 7498
+    },
+    {
+      "epoch": 9.59008,
+      "grad_norm": 0.5328890085220337,
+      "learning_rate": 1.3003095975232198e-05,
+      "loss": 3.2064,
+      "step": 7499
+    },
+    {
+      "epoch": 9.59136,
+      "grad_norm": 0.5413006544113159,
+      "learning_rate": 1.2962713689594831e-05,
+      "loss": 3.2733,
+      "step": 7500
+    },
+    {
+      "epoch": 9.59264,
+      "grad_norm": 0.5474052429199219,
+      "learning_rate": 1.2922331403957463e-05,
+      "loss": 3.2836,
+      "step": 7501
+    },
+    {
+      "epoch": 9.59392,
+      "grad_norm": 0.5338591933250427,
+      "learning_rate": 1.2881949118320096e-05,
+      "loss": 3.2265,
+      "step": 7502
+    },
+    {
+      "epoch": 9.5952,
+      "grad_norm": 0.533933699131012,
+      "learning_rate": 1.284156683268273e-05,
+      "loss": 3.2688,
+      "step": 7503
+    },
+    {
+      "epoch": 9.59648,
+      "grad_norm": 0.5405238270759583,
+      "learning_rate": 1.2801184547045361e-05,
+      "loss": 3.3086,
+      "step": 7504
+    },
+    {
+      "epoch": 9.59776,
+      "grad_norm": 0.5274151563644409,
+      "learning_rate": 1.2760802261407995e-05,
+      "loss": 3.2516,
+      "step": 7505
+    },
+    {
+      "epoch": 9.59904,
+      "grad_norm": 0.533254861831665,
+      "learning_rate": 1.2720419975770628e-05,
+      "loss": 3.2087,
+      "step": 7506
+    },
+    {
+      "epoch": 9.60032,
+      "grad_norm": 0.5390244722366333,
+      "learning_rate": 1.268003769013326e-05,
+      "loss": 3.1609,
+      "step": 7507
+    },
+    {
+      "epoch": 9.6016,
+      "grad_norm": 0.5491555333137512,
+      "learning_rate": 1.2639655404495894e-05,
+      "loss": 3.261,
+      "step": 7508
+    },
+    {
+      "epoch": 9.60288,
+      "grad_norm": 0.5407535433769226,
+      "learning_rate": 1.2599273118858527e-05,
+      "loss": 3.2147,
+      "step": 7509
+    },
+    {
+      "epoch": 9.60416,
+      "grad_norm": 0.549811065196991,
+      "learning_rate": 1.255889083322116e-05,
+      "loss": 3.2559,
+      "step": 7510
+    },
+    {
+      "epoch": 9.60544,
+      "grad_norm": 0.5414533019065857,
+      "learning_rate": 1.2518508547583792e-05,
+      "loss": 3.2521,
+      "step": 7511
+    },
+    {
+      "epoch": 9.60672,
+      "grad_norm": 0.5452784895896912,
+      "learning_rate": 1.2478126261946426e-05,
+      "loss": 3.2758,
+      "step": 7512
+    },
+    {
+      "epoch": 9.608,
+      "grad_norm": 0.5487382411956787,
+      "learning_rate": 1.243774397630906e-05,
+      "loss": 3.2842,
+      "step": 7513
+    },
+    {
+      "epoch": 9.60928,
+      "grad_norm": 0.5465408563613892,
+      "learning_rate": 1.239736169067169e-05,
+      "loss": 3.242,
+      "step": 7514
+    },
+    {
+      "epoch": 9.61056,
+      "grad_norm": 0.5443041920661926,
+      "learning_rate": 1.2356979405034323e-05,
+      "loss": 3.2918,
+      "step": 7515
+    },
+    {
+      "epoch": 9.61184,
+      "grad_norm": 0.5370469689369202,
+      "learning_rate": 1.2316597119396958e-05,
+      "loss": 3.2174,
+      "step": 7516
+    },
+    {
+      "epoch": 9.61312,
+      "grad_norm": 0.5332541465759277,
+      "learning_rate": 1.2276214833759591e-05,
+      "loss": 3.2014,
+      "step": 7517
+    },
+    {
+      "epoch": 9.6144,
+      "grad_norm": 0.5249244570732117,
+      "learning_rate": 1.2235832548122221e-05,
+      "loss": 3.2625,
+      "step": 7518
+    },
+    {
+      "epoch": 9.61568,
+      "grad_norm": 0.5379579663276672,
+      "learning_rate": 1.2195450262484855e-05,
+      "loss": 3.2321,
+      "step": 7519
+    },
+    {
+      "epoch": 9.61696,
+      "grad_norm": 0.5287061333656311,
+      "learning_rate": 1.215506797684749e-05,
+      "loss": 3.2617,
+      "step": 7520
+    },
+    {
+      "epoch": 9.61824,
+      "grad_norm": 0.5358967185020447,
+      "learning_rate": 1.211468569121012e-05,
+      "loss": 3.2692,
+      "step": 7521
+    },
+    {
+      "epoch": 9.61952,
+      "grad_norm": 0.5491884350776672,
+      "learning_rate": 1.2074303405572754e-05,
+      "loss": 3.208,
+      "step": 7522
+    },
+    {
+      "epoch": 9.6208,
+      "grad_norm": 0.5433226227760315,
+      "learning_rate": 1.2033921119935387e-05,
+      "loss": 3.2463,
+      "step": 7523
+    },
+    {
+      "epoch": 9.62208,
+      "grad_norm": 0.548875093460083,
+      "learning_rate": 1.1993538834298019e-05,
+      "loss": 3.2841,
+      "step": 7524
+    },
+    {
+      "epoch": 9.62336,
+      "grad_norm": 0.5372313261032104,
+      "learning_rate": 1.1953156548660652e-05,
+      "loss": 3.2922,
+      "step": 7525
+    },
+    {
+      "epoch": 9.62464,
+      "grad_norm": 0.5375744700431824,
+      "learning_rate": 1.1912774263023286e-05,
+      "loss": 3.2362,
+      "step": 7526
+    },
+    {
+      "epoch": 9.62592,
+      "grad_norm": 0.5384906530380249,
+      "learning_rate": 1.187239197738592e-05,
+      "loss": 3.3354,
+      "step": 7527
+    },
+    {
+      "epoch": 9.6272,
+      "grad_norm": 0.532680869102478,
+      "learning_rate": 1.1832009691748551e-05,
+      "loss": 3.2397,
+      "step": 7528
+    },
+    {
+      "epoch": 9.62848,
+      "grad_norm": 0.5564149618148804,
+      "learning_rate": 1.1791627406111185e-05,
+      "loss": 3.2319,
+      "step": 7529
+    },
+    {
+      "epoch": 9.62976,
+      "grad_norm": 0.5441488027572632,
+      "learning_rate": 1.1751245120473818e-05,
+      "loss": 3.2088,
+      "step": 7530
+    },
+    {
+      "epoch": 9.63104,
+      "grad_norm": 0.5175425410270691,
+      "learning_rate": 1.171086283483645e-05,
+      "loss": 3.2514,
+      "step": 7531
+    },
+    {
+      "epoch": 9.63232,
+      "grad_norm": 0.5204359292984009,
+      "learning_rate": 1.1670480549199083e-05,
+      "loss": 3.2264,
+      "step": 7532
+    },
+    {
+      "epoch": 9.6336,
+      "grad_norm": 0.5370702743530273,
+      "learning_rate": 1.1630098263561717e-05,
+      "loss": 3.2164,
+      "step": 7533
+    },
+    {
+      "epoch": 9.63488,
+      "grad_norm": 0.562423050403595,
+      "learning_rate": 1.1589715977924348e-05,
+      "loss": 3.2344,
+      "step": 7534
+    },
+    {
+      "epoch": 9.63616,
+      "grad_norm": 0.5344887375831604,
+      "learning_rate": 1.1549333692286982e-05,
+      "loss": 3.1734,
+      "step": 7535
+    },
+    {
+      "epoch": 9.63744,
+      "grad_norm": 0.5403535962104797,
+      "learning_rate": 1.1508951406649615e-05,
+      "loss": 3.2056,
+      "step": 7536
+    },
+    {
+      "epoch": 9.63872,
+      "grad_norm": 0.5409168004989624,
+      "learning_rate": 1.1468569121012249e-05,
+      "loss": 3.2126,
+      "step": 7537
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 0.5526571273803711,
+      "learning_rate": 1.142818683537488e-05,
+      "loss": 3.2779,
+      "step": 7538
+    },
+    {
+      "epoch": 9.64128,
+      "grad_norm": 0.5267760753631592,
+      "learning_rate": 1.1387804549737514e-05,
+      "loss": 3.2548,
+      "step": 7539
+    },
+    {
+      "epoch": 9.64256,
+      "grad_norm": 0.5352032780647278,
+      "learning_rate": 1.1347422264100148e-05,
+      "loss": 3.2621,
+      "step": 7540
+    },
+    {
+      "epoch": 9.64384,
+      "grad_norm": 0.5344724059104919,
+      "learning_rate": 1.130703997846278e-05,
+      "loss": 3.1993,
+      "step": 7541
+    },
+    {
+      "epoch": 9.64512,
+      "grad_norm": 0.5381539463996887,
+      "learning_rate": 1.1266657692825413e-05,
+      "loss": 3.2864,
+      "step": 7542
+    },
+    {
+      "epoch": 9.6464,
+      "grad_norm": 0.5417296886444092,
+      "learning_rate": 1.1226275407188046e-05,
+      "loss": 3.2113,
+      "step": 7543
+    },
+    {
+      "epoch": 9.64768,
+      "grad_norm": 0.5437041521072388,
+      "learning_rate": 1.1185893121550678e-05,
+      "loss": 3.1934,
+      "step": 7544
+    },
+    {
+      "epoch": 9.64896,
+      "grad_norm": 0.5453625321388245,
+      "learning_rate": 1.1145510835913311e-05,
+      "loss": 3.2435,
+      "step": 7545
+    },
+    {
+      "epoch": 9.65024,
+      "grad_norm": 0.5361076593399048,
+      "learning_rate": 1.1105128550275945e-05,
+      "loss": 3.2032,
+      "step": 7546
+    },
+    {
+      "epoch": 9.65152,
+      "grad_norm": 0.5454299449920654,
+      "learning_rate": 1.1064746264638578e-05,
+      "loss": 3.3266,
+      "step": 7547
+    },
+    {
+      "epoch": 9.6528,
+      "grad_norm": 0.5517051219940186,
+      "learning_rate": 1.102436397900121e-05,
+      "loss": 3.2157,
+      "step": 7548
+    },
+    {
+      "epoch": 9.65408,
+      "grad_norm": 0.5611749291419983,
+      "learning_rate": 1.0983981693363844e-05,
+      "loss": 3.2855,
+      "step": 7549
+    },
+    {
+      "epoch": 9.65536,
+      "grad_norm": 0.5386665463447571,
+      "learning_rate": 1.0943599407726477e-05,
+      "loss": 3.2633,
+      "step": 7550
+    },
+    {
+      "epoch": 9.65664,
+      "grad_norm": 0.545172929763794,
+      "learning_rate": 1.0903217122089109e-05,
+      "loss": 3.2414,
+      "step": 7551
+    },
+    {
+      "epoch": 9.65792,
+      "grad_norm": 0.5379675030708313,
+      "learning_rate": 1.0862834836451742e-05,
+      "loss": 3.2504,
+      "step": 7552
+    },
+    {
+      "epoch": 9.6592,
+      "grad_norm": 0.545353353023529,
+      "learning_rate": 1.0822452550814376e-05,
+      "loss": 3.2483,
+      "step": 7553
+    },
+    {
+      "epoch": 9.66048,
+      "grad_norm": 0.5340268015861511,
+      "learning_rate": 1.0782070265177008e-05,
+      "loss": 3.2026,
+      "step": 7554
+    },
+    {
+      "epoch": 9.66176,
+      "grad_norm": 0.5404039621353149,
+      "learning_rate": 1.0741687979539641e-05,
+      "loss": 3.2664,
+      "step": 7555
+    },
+    {
+      "epoch": 9.66304,
+      "grad_norm": 0.5477861762046814,
+      "learning_rate": 1.0701305693902274e-05,
+      "loss": 3.2691,
+      "step": 7556
+    },
+    {
+      "epoch": 9.66432,
+      "grad_norm": 0.5429335832595825,
+      "learning_rate": 1.0660923408264908e-05,
+      "loss": 3.2404,
+      "step": 7557
+    },
+    {
+      "epoch": 9.6656,
+      "grad_norm": 0.5497188568115234,
+      "learning_rate": 1.062054112262754e-05,
+      "loss": 3.1482,
+      "step": 7558
+    },
+    {
+      "epoch": 9.66688,
+      "grad_norm": 0.5364313721656799,
+      "learning_rate": 1.0580158836990173e-05,
+      "loss": 3.3282,
+      "step": 7559
+    },
+    {
+      "epoch": 9.66816,
+      "grad_norm": 0.5295746326446533,
+      "learning_rate": 1.0539776551352807e-05,
+      "loss": 3.2642,
+      "step": 7560
+    },
+    {
+      "epoch": 9.66944,
+      "grad_norm": 0.5566303133964539,
+      "learning_rate": 1.0499394265715438e-05,
+      "loss": 3.3054,
+      "step": 7561
+    },
+    {
+      "epoch": 9.67072,
+      "grad_norm": 0.5400784611701965,
+      "learning_rate": 1.0459011980078072e-05,
+      "loss": 3.2225,
+      "step": 7562
+    },
+    {
+      "epoch": 9.672,
+      "grad_norm": 0.546852707862854,
+      "learning_rate": 1.0418629694440705e-05,
+      "loss": 3.2187,
+      "step": 7563
+    },
+    {
+      "epoch": 9.67328,
+      "grad_norm": 0.5458729267120361,
+      "learning_rate": 1.0378247408803339e-05,
+      "loss": 3.265,
+      "step": 7564
+    },
+    {
+      "epoch": 9.67456,
+      "grad_norm": 0.5483717322349548,
+      "learning_rate": 1.033786512316597e-05,
+      "loss": 3.2788,
+      "step": 7565
+    },
+    {
+      "epoch": 9.67584,
+      "grad_norm": 0.5389772057533264,
+      "learning_rate": 1.0297482837528604e-05,
+      "loss": 3.2733,
+      "step": 7566
+    },
+    {
+      "epoch": 9.67712,
+      "grad_norm": 0.5320543646812439,
+      "learning_rate": 1.0257100551891237e-05,
+      "loss": 3.2597,
+      "step": 7567
+    },
+    {
+      "epoch": 9.6784,
+      "grad_norm": 0.5460016131401062,
+      "learning_rate": 1.0216718266253868e-05,
+      "loss": 3.226,
+      "step": 7568
+    },
+    {
+      "epoch": 9.67968,
+      "grad_norm": 0.5420553684234619,
+      "learning_rate": 1.0176335980616503e-05,
+      "loss": 3.2611,
+      "step": 7569
+    },
+    {
+      "epoch": 9.68096,
+      "grad_norm": 0.5544966459274292,
+      "learning_rate": 1.0135953694979136e-05,
+      "loss": 3.2986,
+      "step": 7570
+    },
+    {
+      "epoch": 9.68224,
+      "grad_norm": 0.5373578071594238,
+      "learning_rate": 1.0095571409341766e-05,
+      "loss": 3.2232,
+      "step": 7571
+    },
+    {
+      "epoch": 9.68352,
+      "grad_norm": 0.5499645471572876,
+      "learning_rate": 1.00551891237044e-05,
+      "loss": 3.2186,
+      "step": 7572
+    },
+    {
+      "epoch": 9.6848,
+      "grad_norm": 0.5397508144378662,
+      "learning_rate": 1.0014806838067035e-05,
+      "loss": 3.284,
+      "step": 7573
+    },
+    {
+      "epoch": 9.68608,
+      "grad_norm": 0.5304678082466125,
+      "learning_rate": 9.974424552429668e-06,
+      "loss": 3.2274,
+      "step": 7574
+    },
+    {
+      "epoch": 9.68736,
+      "grad_norm": 0.5234605073928833,
+      "learning_rate": 9.934042266792298e-06,
+      "loss": 3.216,
+      "step": 7575
+    },
+    {
+      "epoch": 9.68864,
+      "grad_norm": 0.5570342540740967,
+      "learning_rate": 9.893659981154932e-06,
+      "loss": 3.2837,
+      "step": 7576
+    },
+    {
+      "epoch": 9.68992,
+      "grad_norm": 0.5394232273101807,
+      "learning_rate": 9.853277695517565e-06,
+      "loss": 3.1624,
+      "step": 7577
+    },
+    {
+      "epoch": 9.6912,
+      "grad_norm": 0.5388901829719543,
+      "learning_rate": 9.812895409880197e-06,
+      "loss": 3.2089,
+      "step": 7578
+    },
+    {
+      "epoch": 9.69248,
+      "grad_norm": 0.5200719833374023,
+      "learning_rate": 9.77251312424283e-06,
+      "loss": 3.198,
+      "step": 7579
+    },
+    {
+      "epoch": 9.69376,
+      "grad_norm": 0.550902247428894,
+      "learning_rate": 9.732130838605464e-06,
+      "loss": 3.2799,
+      "step": 7580
+    },
+    {
+      "epoch": 9.69504,
+      "grad_norm": 0.531865656375885,
+      "learning_rate": 9.691748552968096e-06,
+      "loss": 3.242,
+      "step": 7581
+    },
+    {
+      "epoch": 9.69632,
+      "grad_norm": 0.533011257648468,
+      "learning_rate": 9.65136626733073e-06,
+      "loss": 3.2183,
+      "step": 7582
+    },
+    {
+      "epoch": 9.6976,
+      "grad_norm": 0.538938581943512,
+      "learning_rate": 9.610983981693363e-06,
+      "loss": 3.2257,
+      "step": 7583
+    },
+    {
+      "epoch": 9.698879999999999,
+      "grad_norm": 0.5364640951156616,
+      "learning_rate": 9.570601696055996e-06,
+      "loss": 3.2082,
+      "step": 7584
+    },
+    {
+      "epoch": 9.70016,
+      "grad_norm": 0.5598052740097046,
+      "learning_rate": 9.530219410418628e-06,
+      "loss": 3.3729,
+      "step": 7585
+    },
+    {
+      "epoch": 9.70144,
+      "grad_norm": 0.5431958436965942,
+      "learning_rate": 9.489837124781261e-06,
+      "loss": 3.2299,
+      "step": 7586
+    },
+    {
+      "epoch": 9.70272,
+      "grad_norm": 0.5261279344558716,
+      "learning_rate": 9.449454839143895e-06,
+      "loss": 3.2334,
+      "step": 7587
+    },
+    {
+      "epoch": 9.704,
+      "grad_norm": 0.5326124429702759,
+      "learning_rate": 9.409072553506527e-06,
+      "loss": 3.2051,
+      "step": 7588
+    },
+    {
+      "epoch": 9.70528,
+      "grad_norm": 0.551518440246582,
+      "learning_rate": 9.36869026786916e-06,
+      "loss": 3.1614,
+      "step": 7589
+    },
+    {
+      "epoch": 9.70656,
+      "grad_norm": 0.5355722308158875,
+      "learning_rate": 9.328307982231794e-06,
+      "loss": 3.2427,
+      "step": 7590
+    },
+    {
+      "epoch": 9.707840000000001,
+      "grad_norm": 0.5317171812057495,
+      "learning_rate": 9.287925696594427e-06,
+      "loss": 3.1419,
+      "step": 7591
+    },
+    {
+      "epoch": 9.70912,
+      "grad_norm": 0.5412980914115906,
+      "learning_rate": 9.247543410957059e-06,
+      "loss": 3.2759,
+      "step": 7592
+    },
+    {
+      "epoch": 9.7104,
+      "grad_norm": 0.546062171459198,
+      "learning_rate": 9.207161125319692e-06,
+      "loss": 3.2773,
+      "step": 7593
+    },
+    {
+      "epoch": 9.71168,
+      "grad_norm": 0.5412468314170837,
+      "learning_rate": 9.166778839682326e-06,
+      "loss": 3.247,
+      "step": 7594
+    },
+    {
+      "epoch": 9.71296,
+      "grad_norm": 0.5379220247268677,
+      "learning_rate": 9.12639655404496e-06,
+      "loss": 3.2106,
+      "step": 7595
+    },
+    {
+      "epoch": 9.71424,
+      "grad_norm": 0.5455538034439087,
+      "learning_rate": 9.086014268407591e-06,
+      "loss": 3.2229,
+      "step": 7596
+    },
+    {
+      "epoch": 9.71552,
+      "grad_norm": 0.5402326583862305,
+      "learning_rate": 9.045631982770223e-06,
+      "loss": 3.2599,
+      "step": 7597
+    },
+    {
+      "epoch": 9.7168,
+      "grad_norm": 0.5508577823638916,
+      "learning_rate": 9.005249697132858e-06,
+      "loss": 3.272,
+      "step": 7598
+    },
+    {
+      "epoch": 9.71808,
+      "grad_norm": 0.5405802726745605,
+      "learning_rate": 8.96486741149549e-06,
+      "loss": 3.2912,
+      "step": 7599
+    },
+    {
+      "epoch": 9.71936,
+      "grad_norm": 0.5341350436210632,
+      "learning_rate": 8.924485125858123e-06,
+      "loss": 3.2955,
+      "step": 7600
+    },
+    {
+      "epoch": 9.72064,
+      "grad_norm": 0.5520155429840088,
+      "learning_rate": 8.884102840220755e-06,
+      "loss": 3.272,
+      "step": 7601
+    },
+    {
+      "epoch": 9.72192,
+      "grad_norm": 0.5376763939857483,
+      "learning_rate": 8.843720554583388e-06,
+      "loss": 3.215,
+      "step": 7602
+    },
+    {
+      "epoch": 9.7232,
+      "grad_norm": 0.5320829749107361,
+      "learning_rate": 8.803338268946022e-06,
+      "loss": 3.2852,
+      "step": 7603
+    },
+    {
+      "epoch": 9.72448,
+      "grad_norm": 0.5295441746711731,
+      "learning_rate": 8.762955983308654e-06,
+      "loss": 3.1911,
+      "step": 7604
+    },
+    {
+      "epoch": 9.72576,
+      "grad_norm": 0.5424656271934509,
+      "learning_rate": 8.722573697671287e-06,
+      "loss": 3.2145,
+      "step": 7605
+    },
+    {
+      "epoch": 9.72704,
+      "grad_norm": 0.5340449810028076,
+      "learning_rate": 8.68219141203392e-06,
+      "loss": 3.275,
+      "step": 7606
+    },
+    {
+      "epoch": 9.72832,
+      "grad_norm": 0.532345175743103,
+      "learning_rate": 8.641809126396552e-06,
+      "loss": 3.1572,
+      "step": 7607
+    },
+    {
+      "epoch": 9.7296,
+      "grad_norm": 0.5268622636795044,
+      "learning_rate": 8.601426840759186e-06,
+      "loss": 3.2741,
+      "step": 7608
+    },
+    {
+      "epoch": 9.730879999999999,
+      "grad_norm": 0.5382117629051208,
+      "learning_rate": 8.56104455512182e-06,
+      "loss": 3.2102,
+      "step": 7609
+    },
+    {
+      "epoch": 9.73216,
+      "grad_norm": 0.5525277256965637,
+      "learning_rate": 8.520662269484453e-06,
+      "loss": 3.3048,
+      "step": 7610
+    },
+    {
+      "epoch": 9.73344,
+      "grad_norm": 0.5279812216758728,
+      "learning_rate": 8.480279983847084e-06,
+      "loss": 3.256,
+      "step": 7611
+    },
+    {
+      "epoch": 9.73472,
+      "grad_norm": 0.5394153594970703,
+      "learning_rate": 8.439897698209718e-06,
+      "loss": 3.2207,
+      "step": 7612
+    },
+    {
+      "epoch": 9.736,
+      "grad_norm": 0.533865213394165,
+      "learning_rate": 8.399515412572351e-06,
+      "loss": 3.1895,
+      "step": 7613
+    },
+    {
+      "epoch": 9.73728,
+      "grad_norm": 0.5335356593132019,
+      "learning_rate": 8.359133126934983e-06,
+      "loss": 3.2414,
+      "step": 7614
+    },
+    {
+      "epoch": 9.73856,
+      "grad_norm": 0.5455939173698425,
+      "learning_rate": 8.318750841297617e-06,
+      "loss": 3.2906,
+      "step": 7615
+    },
+    {
+      "epoch": 9.739840000000001,
+      "grad_norm": 0.5451316237449646,
+      "learning_rate": 8.27836855566025e-06,
+      "loss": 3.2626,
+      "step": 7616
+    },
+    {
+      "epoch": 9.74112,
+      "grad_norm": 0.5487070679664612,
+      "learning_rate": 8.237986270022882e-06,
+      "loss": 3.2446,
+      "step": 7617
+    },
+    {
+      "epoch": 9.7424,
+      "grad_norm": 0.5351135730743408,
+      "learning_rate": 8.197603984385515e-06,
+      "loss": 3.2191,
+      "step": 7618
+    },
+    {
+      "epoch": 9.74368,
+      "grad_norm": 0.5554419755935669,
+      "learning_rate": 8.157221698748149e-06,
+      "loss": 3.272,
+      "step": 7619
+    },
+    {
+      "epoch": 9.74496,
+      "grad_norm": 0.5496061444282532,
+      "learning_rate": 8.116839413110782e-06,
+      "loss": 3.2893,
+      "step": 7620
+    },
+    {
+      "epoch": 9.74624,
+      "grad_norm": 0.5537092685699463,
+      "learning_rate": 8.076457127473414e-06,
+      "loss": 3.3493,
+      "step": 7621
+    },
+    {
+      "epoch": 9.74752,
+      "grad_norm": 0.5267013907432556,
+      "learning_rate": 8.036074841836047e-06,
+      "loss": 3.2105,
+      "step": 7622
+    },
+    {
+      "epoch": 9.7488,
+      "grad_norm": 0.537086009979248,
+      "learning_rate": 7.995692556198681e-06,
+      "loss": 3.2392,
+      "step": 7623
+    },
+    {
+      "epoch": 9.75008,
+      "grad_norm": 0.5526744723320007,
+      "learning_rate": 7.955310270561313e-06,
+      "loss": 3.2584,
+      "step": 7624
+    },
+    {
+      "epoch": 9.75136,
+      "grad_norm": 0.5359621047973633,
+      "learning_rate": 7.914927984923946e-06,
+      "loss": 3.2106,
+      "step": 7625
+    },
+    {
+      "epoch": 9.75264,
+      "grad_norm": 0.5292626619338989,
+      "learning_rate": 7.87454569928658e-06,
+      "loss": 3.2777,
+      "step": 7626
+    },
+    {
+      "epoch": 9.75392,
+      "grad_norm": 0.5385765433311462,
+      "learning_rate": 7.834163413649213e-06,
+      "loss": 3.2815,
+      "step": 7627
+    },
+    {
+      "epoch": 9.7552,
+      "grad_norm": 0.5359296798706055,
+      "learning_rate": 7.793781128011845e-06,
+      "loss": 3.2214,
+      "step": 7628
+    },
+    {
+      "epoch": 9.75648,
+      "grad_norm": 0.5338163375854492,
+      "learning_rate": 7.753398842374477e-06,
+      "loss": 3.2129,
+      "step": 7629
+    },
+    {
+      "epoch": 9.75776,
+      "grad_norm": 0.5449706315994263,
+      "learning_rate": 7.71301655673711e-06,
+      "loss": 3.2514,
+      "step": 7630
+    },
+    {
+      "epoch": 9.75904,
+      "grad_norm": 0.5432578325271606,
+      "learning_rate": 7.672634271099744e-06,
+      "loss": 3.2642,
+      "step": 7631
+    },
+    {
+      "epoch": 9.76032,
+      "grad_norm": 0.5408291816711426,
+      "learning_rate": 7.632251985462377e-06,
+      "loss": 3.2562,
+      "step": 7632
+    },
+    {
+      "epoch": 9.7616,
+      "grad_norm": 0.5344693064689636,
+      "learning_rate": 7.59186969982501e-06,
+      "loss": 3.2389,
+      "step": 7633
+    },
+    {
+      "epoch": 9.76288,
+      "grad_norm": 0.5311311483383179,
+      "learning_rate": 7.551487414187642e-06,
+      "loss": 3.2397,
+      "step": 7634
+    },
+    {
+      "epoch": 9.76416,
+      "grad_norm": 0.5357692241668701,
+      "learning_rate": 7.511105128550276e-06,
+      "loss": 3.2205,
+      "step": 7635
+    },
+    {
+      "epoch": 9.76544,
+      "grad_norm": 0.5423558950424194,
+      "learning_rate": 7.470722842912908e-06,
+      "loss": 3.2897,
+      "step": 7636
+    },
+    {
+      "epoch": 9.76672,
+      "grad_norm": 0.5260913372039795,
+      "learning_rate": 7.430340557275542e-06,
+      "loss": 3.1949,
+      "step": 7637
+    },
+    {
+      "epoch": 9.768,
+      "grad_norm": 0.5533113479614258,
+      "learning_rate": 7.3899582716381744e-06,
+      "loss": 3.2199,
+      "step": 7638
+    },
+    {
+      "epoch": 9.76928,
+      "grad_norm": 0.5424081087112427,
+      "learning_rate": 7.349575986000806e-06,
+      "loss": 3.3368,
+      "step": 7639
+    },
+    {
+      "epoch": 9.77056,
+      "grad_norm": 0.5297884941101074,
+      "learning_rate": 7.3091937003634405e-06,
+      "loss": 3.1818,
+      "step": 7640
+    },
+    {
+      "epoch": 9.77184,
+      "grad_norm": 0.5412716865539551,
+      "learning_rate": 7.268811414726072e-06,
+      "loss": 3.219,
+      "step": 7641
+    },
+    {
+      "epoch": 9.77312,
+      "grad_norm": 0.5297017693519592,
+      "learning_rate": 7.228429129088707e-06,
+      "loss": 3.2751,
+      "step": 7642
+    },
+    {
+      "epoch": 9.7744,
+      "grad_norm": 0.5230127573013306,
+      "learning_rate": 7.188046843451338e-06,
+      "loss": 3.2886,
+      "step": 7643
+    },
+    {
+      "epoch": 9.77568,
+      "grad_norm": 0.5418264865875244,
+      "learning_rate": 7.147664557813971e-06,
+      "loss": 3.2826,
+      "step": 7644
+    },
+    {
+      "epoch": 9.77696,
+      "grad_norm": 0.5246114134788513,
+      "learning_rate": 7.1072822721766044e-06,
+      "loss": 3.2256,
+      "step": 7645
+    },
+    {
+      "epoch": 9.77824,
+      "grad_norm": 0.5591127872467041,
+      "learning_rate": 7.066899986539237e-06,
+      "loss": 3.2657,
+      "step": 7646
+    },
+    {
+      "epoch": 9.77952,
+      "grad_norm": 0.5307941436767578,
+      "learning_rate": 7.0265177009018705e-06,
+      "loss": 3.2904,
+      "step": 7647
+    },
+    {
+      "epoch": 9.7808,
+      "grad_norm": 0.5432814359664917,
+      "learning_rate": 6.986135415264503e-06,
+      "loss": 3.3016,
+      "step": 7648
+    },
+    {
+      "epoch": 9.78208,
+      "grad_norm": 0.53673255443573,
+      "learning_rate": 6.945753129627136e-06,
+      "loss": 3.1894,
+      "step": 7649
+    },
+    {
+      "epoch": 9.78336,
+      "grad_norm": 0.5456019043922424,
+      "learning_rate": 6.905370843989769e-06,
+      "loss": 3.251,
+      "step": 7650
+    },
+    {
+      "epoch": 9.78464,
+      "grad_norm": 0.5382410287857056,
+      "learning_rate": 6.864988558352402e-06,
+      "loss": 3.2033,
+      "step": 7651
+    },
+    {
+      "epoch": 9.78592,
+      "grad_norm": 0.5357476472854614,
+      "learning_rate": 6.824606272715035e-06,
+      "loss": 3.2886,
+      "step": 7652
+    },
+    {
+      "epoch": 9.7872,
+      "grad_norm": 0.5439364314079285,
+      "learning_rate": 6.784223987077668e-06,
+      "loss": 3.314,
+      "step": 7653
+    },
+    {
+      "epoch": 9.78848,
+      "grad_norm": 0.5283563137054443,
+      "learning_rate": 6.7438417014403005e-06,
+      "loss": 3.2845,
+      "step": 7654
+    },
+    {
+      "epoch": 9.78976,
+      "grad_norm": 0.5235844254493713,
+      "learning_rate": 6.703459415802934e-06,
+      "loss": 3.2595,
+      "step": 7655
+    },
+    {
+      "epoch": 9.79104,
+      "grad_norm": 0.5161533951759338,
+      "learning_rate": 6.663077130165567e-06,
+      "loss": 3.2127,
+      "step": 7656
+    },
+    {
+      "epoch": 9.79232,
+      "grad_norm": 0.5272259712219238,
+      "learning_rate": 6.6226948445282e-06,
+      "loss": 3.2172,
+      "step": 7657
+    },
+    {
+      "epoch": 9.7936,
+      "grad_norm": 0.5525578260421753,
+      "learning_rate": 6.582312558890833e-06,
+      "loss": 3.2411,
+      "step": 7658
+    },
+    {
+      "epoch": 9.79488,
+      "grad_norm": 0.5354979634284973,
+      "learning_rate": 6.541930273253465e-06,
+      "loss": 3.1923,
+      "step": 7659
+    },
+    {
+      "epoch": 9.79616,
+      "grad_norm": 0.5332154035568237,
+      "learning_rate": 6.501547987616099e-06,
+      "loss": 3.204,
+      "step": 7660
+    },
+    {
+      "epoch": 9.79744,
+      "grad_norm": 0.5478113889694214,
+      "learning_rate": 6.461165701978731e-06,
+      "loss": 3.2707,
+      "step": 7661
+    },
+    {
+      "epoch": 9.79872,
+      "grad_norm": 0.5272064208984375,
+      "learning_rate": 6.420783416341365e-06,
+      "loss": 3.2359,
+      "step": 7662
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 0.5392633676528931,
+      "learning_rate": 6.3804011307039975e-06,
+      "loss": 3.1821,
+      "step": 7663
+    },
+    {
+      "epoch": 9.80128,
+      "grad_norm": 0.5203908681869507,
+      "learning_rate": 6.34001884506663e-06,
+      "loss": 3.2749,
+      "step": 7664
+    },
+    {
+      "epoch": 9.80256,
+      "grad_norm": 0.5322408080101013,
+      "learning_rate": 6.2996365594292635e-06,
+      "loss": 3.2694,
+      "step": 7665
+    },
+    {
+      "epoch": 9.80384,
+      "grad_norm": 0.5292795896530151,
+      "learning_rate": 6.259254273791896e-06,
+      "loss": 3.2264,
+      "step": 7666
+    },
+    {
+      "epoch": 9.80512,
+      "grad_norm": 0.5364894866943359,
+      "learning_rate": 6.21887198815453e-06,
+      "loss": 3.1952,
+      "step": 7667
+    },
+    {
+      "epoch": 9.8064,
+      "grad_norm": 0.5564066171646118,
+      "learning_rate": 6.178489702517161e-06,
+      "loss": 3.3316,
+      "step": 7668
+    },
+    {
+      "epoch": 9.80768,
+      "grad_norm": 0.5362501740455627,
+      "learning_rate": 6.138107416879796e-06,
+      "loss": 3.2155,
+      "step": 7669
+    },
+    {
+      "epoch": 9.80896,
+      "grad_norm": 0.5605980753898621,
+      "learning_rate": 6.0977251312424275e-06,
+      "loss": 3.2895,
+      "step": 7670
+    },
+    {
+      "epoch": 9.81024,
+      "grad_norm": 0.538100004196167,
+      "learning_rate": 6.05734284560506e-06,
+      "loss": 3.2186,
+      "step": 7671
+    },
+    {
+      "epoch": 9.81152,
+      "grad_norm": 0.5335180163383484,
+      "learning_rate": 6.0169605599676936e-06,
+      "loss": 3.2623,
+      "step": 7672
+    },
+    {
+      "epoch": 9.8128,
+      "grad_norm": 0.535961389541626,
+      "learning_rate": 5.976578274330326e-06,
+      "loss": 3.2338,
+      "step": 7673
+    },
+    {
+      "epoch": 9.81408,
+      "grad_norm": 0.5322915315628052,
+      "learning_rate": 5.93619598869296e-06,
+      "loss": 3.1992,
+      "step": 7674
+    },
+    {
+      "epoch": 9.81536,
+      "grad_norm": 0.52436763048172,
+      "learning_rate": 5.895813703055592e-06,
+      "loss": 3.257,
+      "step": 7675
+    },
+    {
+      "epoch": 9.81664,
+      "grad_norm": 0.5428990125656128,
+      "learning_rate": 5.855431417418225e-06,
+      "loss": 3.1951,
+      "step": 7676
+    },
+    {
+      "epoch": 9.81792,
+      "grad_norm": 0.5562485456466675,
+      "learning_rate": 5.815049131780858e-06,
+      "loss": 3.303,
+      "step": 7677
+    },
+    {
+      "epoch": 9.8192,
+      "grad_norm": 0.5489171743392944,
+      "learning_rate": 5.774666846143491e-06,
+      "loss": 3.2214,
+      "step": 7678
+    },
+    {
+      "epoch": 9.82048,
+      "grad_norm": 0.5357405543327332,
+      "learning_rate": 5.734284560506124e-06,
+      "loss": 3.2282,
+      "step": 7679
+    },
+    {
+      "epoch": 9.82176,
+      "grad_norm": 0.5341373682022095,
+      "learning_rate": 5.693902274868757e-06,
+      "loss": 3.2049,
+      "step": 7680
+    },
+    {
+      "epoch": 9.82304,
+      "grad_norm": 0.5386019349098206,
+      "learning_rate": 5.65351998923139e-06,
+      "loss": 3.2545,
+      "step": 7681
+    },
+    {
+      "epoch": 9.82432,
+      "grad_norm": 0.5332474708557129,
+      "learning_rate": 5.613137703594023e-06,
+      "loss": 3.207,
+      "step": 7682
+    },
+    {
+      "epoch": 9.8256,
+      "grad_norm": 0.5416889786720276,
+      "learning_rate": 5.572755417956656e-06,
+      "loss": 3.2406,
+      "step": 7683
+    },
+    {
+      "epoch": 9.82688,
+      "grad_norm": 0.5246055722236633,
+      "learning_rate": 5.532373132319289e-06,
+      "loss": 3.3026,
+      "step": 7684
+    },
+    {
+      "epoch": 9.82816,
+      "grad_norm": 0.5286515951156616,
+      "learning_rate": 5.491990846681922e-06,
+      "loss": 3.2483,
+      "step": 7685
+    },
+    {
+      "epoch": 9.82944,
+      "grad_norm": 0.5369952917098999,
+      "learning_rate": 5.451608561044554e-06,
+      "loss": 3.2452,
+      "step": 7686
+    },
+    {
+      "epoch": 9.83072,
+      "grad_norm": 0.5319638252258301,
+      "learning_rate": 5.411226275407188e-06,
+      "loss": 3.2861,
+      "step": 7687
+    },
+    {
+      "epoch": 9.832,
+      "grad_norm": 0.5259891748428345,
+      "learning_rate": 5.3708439897698205e-06,
+      "loss": 3.2605,
+      "step": 7688
+    },
+    {
+      "epoch": 9.83328,
+      "grad_norm": 0.5339581966400146,
+      "learning_rate": 5.330461704132454e-06,
+      "loss": 3.2469,
+      "step": 7689
+    },
+    {
+      "epoch": 9.83456,
+      "grad_norm": 0.5350625514984131,
+      "learning_rate": 5.290079418495087e-06,
+      "loss": 3.2347,
+      "step": 7690
+    },
+    {
+      "epoch": 9.83584,
+      "grad_norm": 0.5377787947654724,
+      "learning_rate": 5.249697132857719e-06,
+      "loss": 3.1947,
+      "step": 7691
+    },
+    {
+      "epoch": 9.83712,
+      "grad_norm": 0.5324146747589111,
+      "learning_rate": 5.209314847220353e-06,
+      "loss": 3.2842,
+      "step": 7692
+    },
+    {
+      "epoch": 9.8384,
+      "grad_norm": 0.5433526635169983,
+      "learning_rate": 5.168932561582985e-06,
+      "loss": 3.1448,
+      "step": 7693
+    },
+    {
+      "epoch": 9.83968,
+      "grad_norm": 0.5497397184371948,
+      "learning_rate": 5.128550275945619e-06,
+      "loss": 3.2761,
+      "step": 7694
+    },
+    {
+      "epoch": 9.84096,
+      "grad_norm": 0.5405837893486023,
+      "learning_rate": 5.088167990308251e-06,
+      "loss": 3.2843,
+      "step": 7695
+    },
+    {
+      "epoch": 9.84224,
+      "grad_norm": 0.5413019061088562,
+      "learning_rate": 5.047785704670883e-06,
+      "loss": 3.2202,
+      "step": 7696
+    },
+    {
+      "epoch": 9.84352,
+      "grad_norm": 0.5248864889144897,
+      "learning_rate": 5.0074034190335174e-06,
+      "loss": 3.234,
+      "step": 7697
+    },
+    {
+      "epoch": 9.8448,
+      "grad_norm": 0.5380563139915466,
+      "learning_rate": 4.967021133396149e-06,
+      "loss": 3.2331,
+      "step": 7698
+    },
+    {
+      "epoch": 9.84608,
+      "grad_norm": 0.5344839096069336,
+      "learning_rate": 4.926638847758783e-06,
+      "loss": 3.3208,
+      "step": 7699
+    },
+    {
+      "epoch": 9.84736,
+      "grad_norm": 0.544310450553894,
+      "learning_rate": 4.886256562121415e-06,
+      "loss": 3.242,
+      "step": 7700
+    },
+    {
+      "epoch": 9.84864,
+      "grad_norm": 0.5436192750930786,
+      "learning_rate": 4.845874276484048e-06,
+      "loss": 3.2463,
+      "step": 7701
+    },
+    {
+      "epoch": 9.849920000000001,
+      "grad_norm": 0.5517845749855042,
+      "learning_rate": 4.805491990846681e-06,
+      "loss": 3.2587,
+      "step": 7702
+    },
+    {
+      "epoch": 9.8512,
+      "grad_norm": 0.5308093428611755,
+      "learning_rate": 4.765109705209314e-06,
+      "loss": 3.238,
+      "step": 7703
+    },
+    {
+      "epoch": 9.85248,
+      "grad_norm": 0.5324238538742065,
+      "learning_rate": 4.7247274195719474e-06,
+      "loss": 3.1847,
+      "step": 7704
+    },
+    {
+      "epoch": 9.85376,
+      "grad_norm": 0.5277994871139526,
+      "learning_rate": 4.68434513393458e-06,
+      "loss": 3.1705,
+      "step": 7705
+    },
+    {
+      "epoch": 9.85504,
+      "grad_norm": 0.5271157026290894,
+      "learning_rate": 4.6439628482972135e-06,
+      "loss": 3.1905,
+      "step": 7706
+    },
+    {
+      "epoch": 9.85632,
+      "grad_norm": 0.531658947467804,
+      "learning_rate": 4.603580562659846e-06,
+      "loss": 3.2695,
+      "step": 7707
+    },
+    {
+      "epoch": 9.8576,
+      "grad_norm": 0.5260719060897827,
+      "learning_rate": 4.56319827702248e-06,
+      "loss": 3.2373,
+      "step": 7708
+    },
+    {
+      "epoch": 9.85888,
+      "grad_norm": 0.5422466993331909,
+      "learning_rate": 4.522815991385111e-06,
+      "loss": 3.2401,
+      "step": 7709
+    },
+    {
+      "epoch": 9.86016,
+      "grad_norm": 0.5454497337341309,
+      "learning_rate": 4.482433705747745e-06,
+      "loss": 3.2724,
+      "step": 7710
+    },
+    {
+      "epoch": 9.86144,
+      "grad_norm": 0.5371284484863281,
+      "learning_rate": 4.4420514201103775e-06,
+      "loss": 3.267,
+      "step": 7711
+    },
+    {
+      "epoch": 9.86272,
+      "grad_norm": 0.5461208820343018,
+      "learning_rate": 4.401669134473011e-06,
+      "loss": 3.2026,
+      "step": 7712
+    },
+    {
+      "epoch": 9.864,
+      "grad_norm": 0.5462895631790161,
+      "learning_rate": 4.3612868488356435e-06,
+      "loss": 3.1647,
+      "step": 7713
+    },
+    {
+      "epoch": 9.86528,
+      "grad_norm": 0.5370737314224243,
+      "learning_rate": 4.320904563198276e-06,
+      "loss": 3.2734,
+      "step": 7714
+    },
+    {
+      "epoch": 9.86656,
+      "grad_norm": 0.5331903100013733,
+      "learning_rate": 4.28052227756091e-06,
+      "loss": 3.1304,
+      "step": 7715
+    },
+    {
+      "epoch": 9.86784,
+      "grad_norm": 0.5273880958557129,
+      "learning_rate": 4.240139991923542e-06,
+      "loss": 3.2982,
+      "step": 7716
+    },
+    {
+      "epoch": 9.86912,
+      "grad_norm": 0.5413378477096558,
+      "learning_rate": 4.199757706286176e-06,
+      "loss": 3.1814,
+      "step": 7717
+    },
+    {
+      "epoch": 9.8704,
+      "grad_norm": 0.5351929664611816,
+      "learning_rate": 4.159375420648808e-06,
+      "loss": 3.2605,
+      "step": 7718
+    },
+    {
+      "epoch": 9.87168,
+      "grad_norm": 0.545540452003479,
+      "learning_rate": 4.118993135011441e-06,
+      "loss": 3.2398,
+      "step": 7719
+    },
+    {
+      "epoch": 9.872959999999999,
+      "grad_norm": 0.5313085913658142,
+      "learning_rate": 4.078610849374074e-06,
+      "loss": 3.2208,
+      "step": 7720
+    },
+    {
+      "epoch": 9.87424,
+      "grad_norm": 0.5450189709663391,
+      "learning_rate": 4.038228563736707e-06,
+      "loss": 3.2366,
+      "step": 7721
+    },
+    {
+      "epoch": 9.87552,
+      "grad_norm": 0.5320997834205627,
+      "learning_rate": 3.9978462780993405e-06,
+      "loss": 3.2865,
+      "step": 7722
+    },
+    {
+      "epoch": 9.8768,
+      "grad_norm": 0.5449676513671875,
+      "learning_rate": 3.957463992461973e-06,
+      "loss": 3.3148,
+      "step": 7723
+    },
+    {
+      "epoch": 9.87808,
+      "grad_norm": 0.525892972946167,
+      "learning_rate": 3.9170817068246066e-06,
+      "loss": 3.1522,
+      "step": 7724
+    },
+    {
+      "epoch": 9.87936,
+      "grad_norm": 0.5414352416992188,
+      "learning_rate": 3.876699421187238e-06,
+      "loss": 3.2166,
+      "step": 7725
+    },
+    {
+      "epoch": 9.88064,
+      "grad_norm": 0.5205792188644409,
+      "learning_rate": 3.836317135549872e-06,
+      "loss": 3.2689,
+      "step": 7726
+    },
+    {
+      "epoch": 9.881920000000001,
+      "grad_norm": 0.5402050614356995,
+      "learning_rate": 3.795934849912505e-06,
+      "loss": 3.2732,
+      "step": 7727
+    },
+    {
+      "epoch": 9.8832,
+      "grad_norm": 0.5389068722724915,
+      "learning_rate": 3.755552564275138e-06,
+      "loss": 3.2579,
+      "step": 7728
+    },
+    {
+      "epoch": 9.88448,
+      "grad_norm": 0.5307827591896057,
+      "learning_rate": 3.715170278637771e-06,
+      "loss": 3.2471,
+      "step": 7729
+    },
+    {
+      "epoch": 9.88576,
+      "grad_norm": 0.5479539036750793,
+      "learning_rate": 3.674787993000403e-06,
+      "loss": 3.2576,
+      "step": 7730
+    },
+    {
+      "epoch": 9.88704,
+      "grad_norm": 0.5323429703712463,
+      "learning_rate": 3.634405707363036e-06,
+      "loss": 3.2332,
+      "step": 7731
+    },
+    {
+      "epoch": 9.88832,
+      "grad_norm": 0.5382372736930847,
+      "learning_rate": 3.594023421725669e-06,
+      "loss": 3.2954,
+      "step": 7732
+    },
+    {
+      "epoch": 9.8896,
+      "grad_norm": 0.5464025735855103,
+      "learning_rate": 3.5536411360883022e-06,
+      "loss": 3.2248,
+      "step": 7733
+    },
+    {
+      "epoch": 9.89088,
+      "grad_norm": 0.5193501114845276,
+      "learning_rate": 3.5132588504509353e-06,
+      "loss": 3.2544,
+      "step": 7734
+    },
+    {
+      "epoch": 9.89216,
+      "grad_norm": 0.5276758074760437,
+      "learning_rate": 3.472876564813568e-06,
+      "loss": 3.1927,
+      "step": 7735
+    },
+    {
+      "epoch": 9.89344,
+      "grad_norm": 0.5477931499481201,
+      "learning_rate": 3.432494279176201e-06,
+      "loss": 3.2572,
+      "step": 7736
+    },
+    {
+      "epoch": 9.89472,
+      "grad_norm": 0.5434842705726624,
+      "learning_rate": 3.392111993538834e-06,
+      "loss": 3.3412,
+      "step": 7737
+    },
+    {
+      "epoch": 9.896,
+      "grad_norm": 0.5318534970283508,
+      "learning_rate": 3.351729707901467e-06,
+      "loss": 3.2676,
+      "step": 7738
+    },
+    {
+      "epoch": 9.89728,
+      "grad_norm": 0.5352578163146973,
+      "learning_rate": 3.3113474222641e-06,
+      "loss": 3.2455,
+      "step": 7739
+    },
+    {
+      "epoch": 9.89856,
+      "grad_norm": 0.5339382290840149,
+      "learning_rate": 3.2709651366267327e-06,
+      "loss": 3.2947,
+      "step": 7740
+    },
+    {
+      "epoch": 9.89984,
+      "grad_norm": 0.5460047721862793,
+      "learning_rate": 3.2305828509893657e-06,
+      "loss": 3.3013,
+      "step": 7741
+    },
+    {
+      "epoch": 9.90112,
+      "grad_norm": 0.5352959632873535,
+      "learning_rate": 3.1902005653519987e-06,
+      "loss": 3.2407,
+      "step": 7742
+    },
+    {
+      "epoch": 9.9024,
+      "grad_norm": 0.5354017019271851,
+      "learning_rate": 3.1498182797146318e-06,
+      "loss": 3.1999,
+      "step": 7743
+    },
+    {
+      "epoch": 9.90368,
+      "grad_norm": 0.5433909296989441,
+      "learning_rate": 3.109435994077265e-06,
+      "loss": 3.1953,
+      "step": 7744
+    },
+    {
+      "epoch": 9.904959999999999,
+      "grad_norm": 0.5294229388237,
+      "learning_rate": 3.069053708439898e-06,
+      "loss": 3.19,
+      "step": 7745
+    },
+    {
+      "epoch": 9.90624,
+      "grad_norm": 0.5373355150222778,
+      "learning_rate": 3.02867142280253e-06,
+      "loss": 3.2349,
+      "step": 7746
+    },
+    {
+      "epoch": 9.90752,
+      "grad_norm": 0.5573491454124451,
+      "learning_rate": 2.988289137165163e-06,
+      "loss": 3.1777,
+      "step": 7747
+    },
+    {
+      "epoch": 9.9088,
+      "grad_norm": 0.54659104347229,
+      "learning_rate": 2.947906851527796e-06,
+      "loss": 3.2335,
+      "step": 7748
+    },
+    {
+      "epoch": 9.91008,
+      "grad_norm": 0.5471231341362,
+      "learning_rate": 2.907524565890429e-06,
+      "loss": 3.2676,
+      "step": 7749
+    },
+    {
+      "epoch": 9.91136,
+      "grad_norm": 0.5398557186126709,
+      "learning_rate": 2.867142280253062e-06,
+      "loss": 3.2567,
+      "step": 7750
+    },
+    {
+      "epoch": 9.91264,
+      "grad_norm": 0.5324544310569763,
+      "learning_rate": 2.826759994615695e-06,
+      "loss": 3.2461,
+      "step": 7751
+    },
+    {
+      "epoch": 9.91392,
+      "grad_norm": 0.5316013693809509,
+      "learning_rate": 2.786377708978328e-06,
+      "loss": 3.2755,
+      "step": 7752
+    },
+    {
+      "epoch": 9.9152,
+      "grad_norm": 0.5322742462158203,
+      "learning_rate": 2.745995423340961e-06,
+      "loss": 3.2667,
+      "step": 7753
+    },
+    {
+      "epoch": 9.91648,
+      "grad_norm": 0.5327667593955994,
+      "learning_rate": 2.705613137703594e-06,
+      "loss": 3.278,
+      "step": 7754
+    },
+    {
+      "epoch": 9.91776,
+      "grad_norm": 0.5432612895965576,
+      "learning_rate": 2.665230852066227e-06,
+      "loss": 3.2554,
+      "step": 7755
+    },
+    {
+      "epoch": 9.91904,
+      "grad_norm": 0.5206037163734436,
+      "learning_rate": 2.6248485664288596e-06,
+      "loss": 3.2419,
+      "step": 7756
+    },
+    {
+      "epoch": 9.92032,
+      "grad_norm": 0.5430627465248108,
+      "learning_rate": 2.5844662807914926e-06,
+      "loss": 3.2313,
+      "step": 7757
+    },
+    {
+      "epoch": 9.9216,
+      "grad_norm": 0.527530312538147,
+      "learning_rate": 2.5440839951541257e-06,
+      "loss": 3.2726,
+      "step": 7758
+    },
+    {
+      "epoch": 9.92288,
+      "grad_norm": 0.5246833562850952,
+      "learning_rate": 2.5037017095167587e-06,
+      "loss": 3.1722,
+      "step": 7759
+    },
+    {
+      "epoch": 9.92416,
+      "grad_norm": 0.5283114314079285,
+      "learning_rate": 2.4633194238793913e-06,
+      "loss": 3.2337,
+      "step": 7760
+    },
+    {
+      "epoch": 9.92544,
+      "grad_norm": 0.5455198884010315,
+      "learning_rate": 2.422937138242024e-06,
+      "loss": 3.2814,
+      "step": 7761
+    },
+    {
+      "epoch": 9.92672,
+      "grad_norm": 0.5268814563751221,
+      "learning_rate": 2.382554852604657e-06,
+      "loss": 3.2407,
+      "step": 7762
+    },
+    {
+      "epoch": 9.928,
+      "grad_norm": 0.5490137338638306,
+      "learning_rate": 2.34217256696729e-06,
+      "loss": 3.2843,
+      "step": 7763
+    },
+    {
+      "epoch": 9.92928,
+      "grad_norm": 0.525651216506958,
+      "learning_rate": 2.301790281329923e-06,
+      "loss": 3.2615,
+      "step": 7764
+    },
+    {
+      "epoch": 9.93056,
+      "grad_norm": 0.5373636484146118,
+      "learning_rate": 2.2614079956925557e-06,
+      "loss": 3.2135,
+      "step": 7765
+    },
+    {
+      "epoch": 9.93184,
+      "grad_norm": 0.5554747581481934,
+      "learning_rate": 2.2210257100551887e-06,
+      "loss": 3.2698,
+      "step": 7766
+    },
+    {
+      "epoch": 9.93312,
+      "grad_norm": 0.5517007112503052,
+      "learning_rate": 2.1806434244178218e-06,
+      "loss": 3.3233,
+      "step": 7767
+    },
+    {
+      "epoch": 9.9344,
+      "grad_norm": 0.5402884483337402,
+      "learning_rate": 2.140261138780455e-06,
+      "loss": 3.2164,
+      "step": 7768
+    },
+    {
+      "epoch": 9.93568,
+      "grad_norm": 0.5363087058067322,
+      "learning_rate": 2.099878853143088e-06,
+      "loss": 3.3029,
+      "step": 7769
+    },
+    {
+      "epoch": 9.93696,
+      "grad_norm": 0.547885537147522,
+      "learning_rate": 2.0594965675057205e-06,
+      "loss": 3.2803,
+      "step": 7770
+    },
+    {
+      "epoch": 9.93824,
+      "grad_norm": 0.5402307510375977,
+      "learning_rate": 2.0191142818683535e-06,
+      "loss": 3.2742,
+      "step": 7771
+    },
+    {
+      "epoch": 9.93952,
+      "grad_norm": 0.5349573493003845,
+      "learning_rate": 1.9787319962309865e-06,
+      "loss": 3.2538,
+      "step": 7772
+    },
+    {
+      "epoch": 9.9408,
+      "grad_norm": 0.5216020345687866,
+      "learning_rate": 1.938349710593619e-06,
+      "loss": 3.1519,
+      "step": 7773
+    },
+    {
+      "epoch": 9.94208,
+      "grad_norm": 0.5340332388877869,
+      "learning_rate": 1.8979674249562524e-06,
+      "loss": 3.2481,
+      "step": 7774
+    },
+    {
+      "epoch": 9.94336,
+      "grad_norm": 0.5314613580703735,
+      "learning_rate": 1.8575851393188855e-06,
+      "loss": 3.2759,
+      "step": 7775
+    },
+    {
+      "epoch": 9.94464,
+      "grad_norm": 0.5448631048202515,
+      "learning_rate": 1.817202853681518e-06,
+      "loss": 3.2906,
+      "step": 7776
+    },
+    {
+      "epoch": 9.94592,
+      "grad_norm": 0.5468592047691345,
+      "learning_rate": 1.7768205680441511e-06,
+      "loss": 3.2446,
+      "step": 7777
+    },
+    {
+      "epoch": 9.9472,
+      "grad_norm": 0.5331081748008728,
+      "learning_rate": 1.736438282406784e-06,
+      "loss": 3.2681,
+      "step": 7778
+    },
+    {
+      "epoch": 9.94848,
+      "grad_norm": 0.5275120139122009,
+      "learning_rate": 1.696055996769417e-06,
+      "loss": 3.2593,
+      "step": 7779
+    },
+    {
+      "epoch": 9.94976,
+      "grad_norm": 0.5483401417732239,
+      "learning_rate": 1.65567371113205e-06,
+      "loss": 3.2909,
+      "step": 7780
+    },
+    {
+      "epoch": 9.95104,
+      "grad_norm": 0.5375556349754333,
+      "learning_rate": 1.6152914254946828e-06,
+      "loss": 3.219,
+      "step": 7781
+    },
+    {
+      "epoch": 9.95232,
+      "grad_norm": 0.532954752445221,
+      "learning_rate": 1.5749091398573159e-06,
+      "loss": 3.2613,
+      "step": 7782
+    },
+    {
+      "epoch": 9.9536,
+      "grad_norm": 0.534887969493866,
+      "learning_rate": 1.534526854219949e-06,
+      "loss": 3.2319,
+      "step": 7783
+    },
+    {
+      "epoch": 9.95488,
+      "grad_norm": 0.5328458547592163,
+      "learning_rate": 1.4941445685825815e-06,
+      "loss": 3.2237,
+      "step": 7784
+    },
+    {
+      "epoch": 9.95616,
+      "grad_norm": 0.530035674571991,
+      "learning_rate": 1.4537622829452146e-06,
+      "loss": 3.2497,
+      "step": 7785
+    },
+    {
+      "epoch": 9.95744,
+      "grad_norm": 0.5311540365219116,
+      "learning_rate": 1.4133799973078474e-06,
+      "loss": 3.1767,
+      "step": 7786
+    },
+    {
+      "epoch": 9.95872,
+      "grad_norm": 0.5378389358520508,
+      "learning_rate": 1.3729977116704805e-06,
+      "loss": 3.253,
+      "step": 7787
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 0.5423367619514465,
+      "learning_rate": 1.3326154260331135e-06,
+      "loss": 3.2811,
+      "step": 7788
+    },
+    {
+      "epoch": 9.96128,
+      "grad_norm": 0.5347367525100708,
+      "learning_rate": 1.2922331403957463e-06,
+      "loss": 3.2527,
+      "step": 7789
+    },
+    {
+      "epoch": 9.96256,
+      "grad_norm": 0.547399640083313,
+      "learning_rate": 1.2518508547583794e-06,
+      "loss": 3.2159,
+      "step": 7790
+    },
+    {
+      "epoch": 9.96384,
+      "grad_norm": 0.5291503667831421,
+      "learning_rate": 1.211468569121012e-06,
+      "loss": 3.1828,
+      "step": 7791
+    },
+    {
+      "epoch": 9.96512,
+      "grad_norm": 0.5376866459846497,
+      "learning_rate": 1.171086283483645e-06,
+      "loss": 3.2118,
+      "step": 7792
+    },
+    {
+      "epoch": 9.9664,
+      "grad_norm": 0.5406926274299622,
+      "learning_rate": 1.1307039978462778e-06,
+      "loss": 3.1903,
+      "step": 7793
+    },
+    {
+      "epoch": 9.96768,
+      "grad_norm": 0.5449708104133606,
+      "learning_rate": 1.0903217122089109e-06,
+      "loss": 3.2392,
+      "step": 7794
+    },
+    {
+      "epoch": 9.96896,
+      "grad_norm": 0.5226523876190186,
+      "learning_rate": 1.049939426571544e-06,
+      "loss": 3.2205,
+      "step": 7795
+    },
+    {
+      "epoch": 9.97024,
+      "grad_norm": 0.5401772260665894,
+      "learning_rate": 1.0095571409341768e-06,
+      "loss": 3.2496,
+      "step": 7796
+    },
+    {
+      "epoch": 9.97152,
+      "grad_norm": 0.5378510355949402,
+      "learning_rate": 9.691748552968096e-07,
+      "loss": 3.1916,
+      "step": 7797
+    },
+    {
+      "epoch": 9.9728,
+      "grad_norm": 0.5274825096130371,
+      "learning_rate": 9.287925696594427e-07,
+      "loss": 3.2405,
+      "step": 7798
+    },
+    {
+      "epoch": 9.97408,
+      "grad_norm": 0.5408015251159668,
+      "learning_rate": 8.884102840220756e-07,
+      "loss": 3.2419,
+      "step": 7799
+    },
+    {
+      "epoch": 9.97536,
+      "grad_norm": 0.5279608964920044,
+      "learning_rate": 8.480279983847085e-07,
+      "loss": 3.2378,
+      "step": 7800
+    },
+    {
+      "epoch": 9.97664,
+      "grad_norm": 0.5284964442253113,
+      "learning_rate": 8.076457127473414e-07,
+      "loss": 3.2899,
+      "step": 7801
+    },
+    {
+      "epoch": 9.97792,
+      "grad_norm": 0.537851095199585,
+      "learning_rate": 7.672634271099745e-07,
+      "loss": 3.308,
+      "step": 7802
+    },
+    {
+      "epoch": 9.9792,
+      "grad_norm": 0.5350565314292908,
+      "learning_rate": 7.268811414726073e-07,
+      "loss": 3.1784,
+      "step": 7803
+    },
+    {
+      "epoch": 9.98048,
+      "grad_norm": 0.5333359837532043,
+      "learning_rate": 6.864988558352402e-07,
+      "loss": 3.2109,
+      "step": 7804
+    },
+    {
+      "epoch": 9.98176,
+      "grad_norm": 0.5407792925834656,
+      "learning_rate": 6.461165701978732e-07,
+      "loss": 3.2474,
+      "step": 7805
+    },
+    {
+      "epoch": 9.98304,
+      "grad_norm": 0.545315146446228,
+      "learning_rate": 6.05734284560506e-07,
+      "loss": 3.2454,
+      "step": 7806
+    },
+    {
+      "epoch": 9.98432,
+      "grad_norm": 0.5328691601753235,
+      "learning_rate": 5.653519989231389e-07,
+      "loss": 3.2104,
+      "step": 7807
+    },
+    {
+      "epoch": 9.9856,
+      "grad_norm": 0.5297470092773438,
+      "learning_rate": 5.24969713285772e-07,
+      "loss": 3.2225,
+      "step": 7808
+    },
+    {
+      "epoch": 9.98688,
+      "grad_norm": 0.5331732630729675,
+      "learning_rate": 4.845874276484048e-07,
+      "loss": 3.2141,
+      "step": 7809
+    },
+    {
+      "epoch": 9.98816,
+      "grad_norm": 0.5387186408042908,
+      "learning_rate": 4.442051420110378e-07,
+      "loss": 3.2828,
+      "step": 7810
+    },
+    {
+      "epoch": 9.98944,
+      "grad_norm": 0.5281053781509399,
+      "learning_rate": 4.038228563736707e-07,
+      "loss": 3.2234,
+      "step": 7811
+    },
+    {
+      "epoch": 9.99072,
+      "grad_norm": 0.5405586957931519,
+      "learning_rate": 3.6344057073630365e-07,
+      "loss": 3.3145,
+      "step": 7812
+    },
+    {
+      "epoch": 9.992,
+      "grad_norm": 0.5372722744941711,
+      "learning_rate": 3.230582850989366e-07,
+      "loss": 3.2676,
+      "step": 7813
+    },
+    {
+      "epoch": 9.99328,
+      "grad_norm": 0.5271605253219604,
+      "learning_rate": 2.8267599946156946e-07,
+      "loss": 3.2855,
+      "step": 7814
+    },
+    {
+      "epoch": 9.99456,
+      "grad_norm": 0.5379496812820435,
+      "learning_rate": 2.422937138242024e-07,
+      "loss": 3.1735,
+      "step": 7815
+    },
+    {
+      "epoch": 9.99584,
+      "grad_norm": 0.5309686660766602,
+      "learning_rate": 2.0191142818683536e-07,
+      "loss": 3.2678,
+      "step": 7816
+    },
+    {
+      "epoch": 9.99712,
+      "grad_norm": 0.5377652645111084,
+      "learning_rate": 1.615291425494683e-07,
+      "loss": 3.3243,
+      "step": 7817
+    },
+    {
+      "epoch": 9.9984,
+      "grad_norm": 0.5389779806137085,
+      "learning_rate": 1.211468569121012e-07,
+      "loss": 3.2626,
+      "step": 7818
+    },
+    {
+      "epoch": 9.99968,
+      "grad_norm": 0.54366135597229,
+      "learning_rate": 8.076457127473414e-08,
+      "loss": 3.2713,
+      "step": 7819
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.0505841970443726,
+      "learning_rate": 4.038228563736707e-08,
+      "loss": 3.2862,
+      "step": 7820
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7820,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.5963011260928e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}