{
  "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": 6.293833826007159,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.8862,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.06683060242767,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8348,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.085593029855262,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8919,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.760576316404554,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.8317,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.5955604298294395,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.8164,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.4241910432702802,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7569,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.0775478325962156,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.7122,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.9008746772989573,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.7835,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 3.6499637096171424,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.7065,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 3.897621871094942,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.7161,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.108219669394765,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6817,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 4.05540403379258,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.7038,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.568474560061959,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6703,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.9544696940564121,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.652,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.6134105187686236,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6597,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.2342699291974677,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.5939,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.7411085903434311,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.6429,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.2430514639056482,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.6238,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.310646421522415,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.6051,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.1362663542854012,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.5999,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.8889327202115898,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.5697,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.0712088916619098,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5683,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 1.0078969637180104,
      "learning_rate": 2e-05,
      "loss": 0.597,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.9422799824806541,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5507,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.9698330597252763,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.5784,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7221079971222841,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5719,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.7014980135419222,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5649,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.7633701425527275,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.55,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6417825074207438,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5935,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.708037799601626,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5589,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.7491076050282107,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5388,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.5804329603861615,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5358,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 1.0995232109558182,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.7344,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.5842012161050123,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.4806,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.704739789920105,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4871,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.736623351275328,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4807,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.6253908809510927,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.5233,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.6182900601597383,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4424,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.6681235591120583,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.4689,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.6373323329321743,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.5329,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.7047275034593882,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4655,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.6246515542250495,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4712,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.5856508173641957,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.5023,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.5962967705046875,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4177,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.7806001779998398,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.5141,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.622043734959372,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4599,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.7121168891261236,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.4406,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.6488634345123712,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.5134,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.520746079921578,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.444,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.6045948242543929,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.3933,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.5254642849482244,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.4488,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.6157649648460861,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.5049,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.5631831213957198,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.4713,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.5602537302229734,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.44,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.5914581257872572,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4619,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.549463095796092,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4267,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.5340178679294976,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4761,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.5808152206830257,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4637,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.4902800171092904,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4436,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.5390313052443518,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4885,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.5121936867897753,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4367,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.5500363533706862,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4405,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.5172272256314452,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4269,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.5483248813247145,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4215,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.59195187727915,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5136,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.6987707813692844,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.5217,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.6335148697025192,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.3458,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.5630324718459703,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.3504,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.8028572217234614,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.3902,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.5670381413787162,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.3084,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.8122168290341552,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.4264,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.4741214258962823,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.3048,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.6586253722411579,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.3685,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.561869536718334,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3297,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.6767692478813152,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.3494,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.5928821569905506,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3359,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.8164378937491322,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.3738,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.5572591269430051,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.3256,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.6300653290697507,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.3154,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.6162777886076989,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.3466,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.6058053919206401,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3428,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.560153372409703,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.3096,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.6087346002273554,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.3437,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.4955198786766178,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.3118,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.5789844176944846,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.3807,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.5309509336643291,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3003,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.6719031837507913,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3883,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.5005442690624767,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.304,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.688837206554437,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.3603,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.5394782699862007,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2964,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.5148588742942368,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.3442,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.5817362125441904,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.3446,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.5148990671834646,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.3397,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.5485534320777181,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.3069,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.5302264458421836,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.3638,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.49405999684903523,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.3712,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.5096202744829424,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3542,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.5620601845699008,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4107,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.6845312065128214,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.2537,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.5683322678824294,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.2451,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.5792932670782358,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.2225,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.9628768359244015,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.2686,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.6337138487225528,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.2613,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.6632556801107956,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.2346,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.577840594366836,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.2366,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.5967042560307922,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.2357,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.6052889150964138,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.2537,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.5526256136083812,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.2325,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.5783020114883667,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.2302,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5151150904654107,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.2379,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.5840124872421842,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.2196,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5575638981914616,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.2417,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.5676673745339996,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.2462,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5488291523193319,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.2291,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.5704098450792927,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.2374,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.4867272584844033,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.2376,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.5529530918394492,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.242,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.5192190664599996,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.2476,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.45987096851195997,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.2364,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.4956300338006397,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.2354,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.5040046760144843,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.2284,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.48330329822327217,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.2413,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.45658137067510934,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.2467,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.4734591767237467,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.213,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.505360568959351,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.244,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.42730450106912954,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.2324,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.44477229642302935,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.2338,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.5032262389671527,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.2551,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.4353746960923058,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.2134,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.4647540232924408,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.2247,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.6339386720042828,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3163,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.6530531217301823,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2007,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.43875095031695605,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.1541,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5625986310025605,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.1569,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.758344673638472,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.1675,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.5742624637188126,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.1713,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.47306195670055207,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.1529,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.5082834459351141,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.1629,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.5149170887831235,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.1496,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.5396696154356401,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.147,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.4639127622235571,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.1713,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.4642291200280393,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.1422,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.5277818211438168,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.1547,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.5011866251927994,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.163,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.42343022420856047,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.1362,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.4787131077725175,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.1605,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.46626711693552353,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.1657,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.4198237310944781,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.1417,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.4538075285568228,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.1598,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.4496647677446296,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.1659,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.37645967832153626,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.1232,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4483913968217151,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.1407,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.48675690589639076,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.188,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.40360767268137376,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.139,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.41779812899318686,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.1642,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.3961846709703568,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.1471,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.40133842476598025,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1367,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.43745906466801676,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.1452,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.41254691993311743,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.1428,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.3992517987662612,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.1564,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.42501443450592047,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.1586,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.421487348641718,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.1401,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.48888887508046985,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.1777,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.532368006164555,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.1548,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.5126184772174526,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1256,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.42535194636650964,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.0947,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.3771851719739927,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.0933,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.45196639950573986,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.122,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.45157600306964624,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.0971,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.5169697816786871,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1095,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.5431066086388939,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.113,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.4731235647267685,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1125,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.38763624259915713,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1058,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.38932153933623975,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.0954,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.3958162438015379,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.106,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.37217291947862674,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.0931,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.3789283664630874,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.0908,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.3664134539407824,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1057,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.35771266382078304,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.0986,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.3472703082619096,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.0877,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.3515962152406197,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.099,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.34854961186427696,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.0958,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.40154731325881876,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.1123,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.36183221654182146,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.0934,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.4295934552136442,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.1252,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.32970138402708415,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.0801,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.3695821342065737,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.0998,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3672619778582091,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1002,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.41056065686731014,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1115,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.35028287274755804,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.0986,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.35138499534126416,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.108,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.3464988655002844,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.0913,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.36519734216811445,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1042,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.33907718375369833,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1083,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3385410561678952,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1117,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.4671555068109887,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.1091,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.41214886246914095,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.0816,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.39539527834164195,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.0799,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.36386501874919946,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.0739,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.31706993387029775,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.0758,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.34669274298668185,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.0949,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.31834158848784283,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.0876,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.29201858105248085,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.087,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.32292934350297237,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.0895,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.28146694255115595,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.0889,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.2900228526044847,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.0703,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.2920659823821498,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.0796,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.35055198081769356,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.0968,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.3152896231902338,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.0825,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.31340595951494493,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.0887,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.3060587528938938,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.0764,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.31527091622895953,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.0887,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.3202497921531981,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.0806,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.34033534010942323,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.0864,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.31351196332151854,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.0867,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.3245746897018996,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.0833,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.3159564900300316,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.0833,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.2692316727079578,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.073,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.3240983350991465,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.0851,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.3080977226597638,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.0793,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.2963952964292608,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.0734,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.3343021300081363,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.0939,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.3065995855334038,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.0796,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.31520626619705966,
      "learning_rate": 0.0,
      "loss": 0.0811,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 2.86943794002133e+17,
      "train_loss": 0.2993284181159522,
      "train_runtime": 12424.2407,
      "train_samples_per_second": 1.78,
      "train_steps_per_second": 0.018
    }
  ],
  "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": 2.86943794002133e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}