{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.850632911392405,
  "eval_steps": 500,
  "global_step": 224,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 6.5768450610316895,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.8696,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.55054113066675,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.9033,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.341822209665703,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8785,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.713585615723055,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.844,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.659285650305568,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.8325,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.3806476800726712,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7737,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.0741387989136433,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.788,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.9063682046207284,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.767,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.7297997187789536,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.8345,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.306569629794866,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.7822,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.383092446805944,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.7473,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.8795687768437936,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.724,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.52703742902751,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.7015,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.9959537422134273,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6991,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.5397811207280405,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6689,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.0850697899823905,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.684,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.3535608161713177,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.6344,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.2115315688531463,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.6443,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.3365063938781134,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.6553,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.9865864331608625,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.6427,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.9162459970071231,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.6256,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.034881742703691,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.6466,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.6978475083178429,
      "learning_rate": 2e-05,
      "loss": 0.6032,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.765387720808055,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.6143,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.9828574281611024,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.6059,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7990278317322586,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.6203,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.8655823249176855,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5882,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.6368231063449986,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.6039,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6731643533718409,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5639,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.6864136847043703,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5723,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.6410409875678449,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5972,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.6378425550510003,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5856,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.918979700001091,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.7859,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.7385030815975311,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.5193,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.7241620877274644,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4961,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.718650355904933,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.5712,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.8057994159955295,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.603,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.6199272117949802,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.5071,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.7000951700344361,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.5139,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.5945674216305776,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.5082,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.5965756702458194,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.5405,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.6335727318579498,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4991,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.5430643258039587,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.522,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.5015469793106819,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.48,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.597052470072264,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4904,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.5394782566605623,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.5073,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.4991702766342904,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.5109,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.5004254869480759,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4869,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.5310694460341979,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.5716,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.546943995965,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.4577,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.5805289291444264,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.5258,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.5257182732367155,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4704,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.5731413663437619,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.5519,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.5205414355108964,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.4872,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.44590613642320476,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4517,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.5413846797873679,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.5135,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.6786972311421863,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.5621,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.6014881831790463,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4929,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5644165758094571,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.5113,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.4953309187357807,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.5,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.5463217834232075,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.5161,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.5430675251106184,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.5384,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.4575981918704574,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4783,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.5652419113739153,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.5251,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.534119596417235,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5219,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.65108690707511,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.513,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.581157557571331,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.3791,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.48996681311543777,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.3978,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.715032997941402,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.3729,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.6996986060211344,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4517,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.6323798016756584,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.3596,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.6112651641721382,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.4165,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.8315313899746641,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4503,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.589256821432724,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3974,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.6462464324503571,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.3931,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.5881410290625357,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3783,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.621933470792803,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.3934,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.7114033456651335,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.4508,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.6404778451856386,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.3583,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.7507212739299634,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4212,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.5337563384530819,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3873,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.6685480938822501,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.3893,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.5616611113240841,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4185,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.5203557630130808,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.4002,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.5432784205828007,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4014,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.5549773011189598,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3957,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.4983933560896198,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3788,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.550992778245209,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.4021,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.5139335452324278,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.4165,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.46129229354063267,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3886,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.5276354748042955,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.4329,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.5371005396003636,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.3914,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.4517820528120139,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.3691,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.5493227871100687,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.41,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.4943128273595533,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.4133,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.49273159732976557,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.3876,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.47036037836033895,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3815,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.6152301119035102,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.5325,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.7293064112167361,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3202,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.588473566492545,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3362,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.6163105506568968,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.3265,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.79518313126124,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.3164,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.634260970877391,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.32,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.5701210471258467,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3043,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.5278263579358417,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.2968,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.571859767470252,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3145,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.5369179461028326,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.3158,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.47608894265411983,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.297,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.5772127762795359,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.309,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5117147266904122,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.286,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.4858001016277918,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.3028,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5010214442655788,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.3275,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.48443976872227934,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.3018,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.4898261795469956,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.2789,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.43691759974845557,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.2867,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.4759881237119461,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.312,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.45297986644062854,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.2927,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.5016232402474848,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.2968,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.45090411536064756,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.2991,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.42221877497278953,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.267,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.5252573885337772,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.2877,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.5025432182891634,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3239,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.4411908372944242,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.2837,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.5188116619829167,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.2764,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.5052626872547239,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.2952,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.4642271148692705,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.3002,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.4958762765624688,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.3085,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.47172160376478783,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.3094,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.4644247293069529,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.3031,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.42203002589352623,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.3014,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.6184016649984608,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3787,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.5868672020554305,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2274,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.48349819018937273,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.2553,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5819147547323128,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.2345,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.653555071120802,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.2291,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.4987384252383243,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.2003,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.5192071655910175,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.2337,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.4877663607790642,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.2252,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.4708202117958089,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.211,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.48342260754402555,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.2102,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.4290822317771518,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2057,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.48097352887405265,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.2645,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.4269640138979185,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.2078,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.4849597627109131,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.2418,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.4366815623230591,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.1972,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.4824469266819184,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.2152,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.41209168122138534,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.2099,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.4200533132794104,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.2158,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.42459254312524186,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.2187,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.4285408493594532,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.2249,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.4036270452789849,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.192,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.47276982640574583,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.2369,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.40696925082974317,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.2143,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.39963882297896436,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.2166,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.38554788751462793,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.2073,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.37793339811870846,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.2044,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.3801636361324258,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.228,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.36310328395218033,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.1918,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.41730482962316867,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.2321,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.3775878229415627,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.2042,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.38392332876210566,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.2255,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.3984646208836156,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.1998,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.4691583743976479,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2548,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.5161995559667772,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.2206,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.527150501963232,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1781,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.4606166502705677,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.1693,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.3899195488596434,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.19,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.4811235639786534,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.1486,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.622109994821902,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.1778,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.5340682334816561,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1597,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.42249987569616093,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.1552,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.4326287136288023,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1751,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.36480021181382283,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1431,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.47927027386671583,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.1823,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.35559621801537933,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.1491,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.3545064614847448,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.1527,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.4000814536936177,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.1701,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.41396706156868784,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1423,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.3931014138628955,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.1691,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.3911896599681565,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.1624,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.37077963107598727,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.146,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.3675346401968028,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1642,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.32791237781614085,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.1514,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.38590288444420046,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1773,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.3733101628698152,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.169,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.35381071180000695,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1557,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.3756856019173942,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.1813,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3374705155560338,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1461,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.3408453474473912,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.16,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.3727260748990488,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.1812,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.39021624680966305,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.1746,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.3314609203978931,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.1495,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.3397438613008862,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1644,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.33263487388737834,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1432,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3852724991074481,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1942,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.4705236010236445,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.2015,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.3998159107579952,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.141,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.3772889404474953,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1396,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.3750130456703218,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.1366,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.34408946364448384,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.1319,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.3204329372348201,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.1346,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.3246545140356931,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1293,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.3255511779082857,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.1496,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.31939613604740436,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1444,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.3252451221013161,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.1463,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.3718184473325323,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.1536,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.33826284008927243,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.1412,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.38139890797647724,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1521,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.3040129430195829,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.139,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.31535902524561954,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.1427,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.3154171353677096,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.1363,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.3790261251497049,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.1555,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.3343557777574481,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1447,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.32568052273128223,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.1308,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.31552829892523987,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.1329,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.31491429623723605,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.1445,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.317502168089278,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1427,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.31714425876076857,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.1399,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.31179217318936164,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.1428,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.3357842216657128,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.145,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.3437985123770126,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.1334,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.3017928174610383,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1388,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.3161483121715175,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.1466,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.3153522565599406,
      "learning_rate": 0.0,
      "loss": 0.1367,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.795397184103383e+17,
      "train_loss": 0.35499650366338237,
      "train_runtime": 17226.3441,
      "train_samples_per_second": 1.284,
      "train_steps_per_second": 0.013
    }
  ],
  "logging_steps": 1,
  "max_steps": 224,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.795397184103383e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}