{
  "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": 6.689875551000949,
      "learning_rate": 8.695652173913044e-07,
      "loss": 1.0411,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.849445220821603,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 1.06,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.688986607415651,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 1.0539,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 6.17295057672458,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 1.0234,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.824631636118133,
      "learning_rate": 4.347826086956522e-06,
      "loss": 1.005,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.967109228386207,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.9606,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.7133339652521626,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.9667,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 4.307067552348288,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.9328,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.39815824714185,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.9416,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 3.868518252302564,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.9104,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 3.7340752710348415,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.8779,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.156388934200001,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.8779,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.0918790985857574,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.8229,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.7591308291267982,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.8052,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.008219302115603,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.8185,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.4175773471523028,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.7857,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.2508900858180383,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.784,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.3351400529703412,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.7581,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.1898964303901207,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.7633,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.035618164716581,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.7508,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 1.2588535581836704,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.756,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.1617604276047977,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.7484,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.9841377574007387,
      "learning_rate": 2e-05,
      "loss": 0.7437,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.2440118406941543,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.7342,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 1.0300660603430258,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.7267,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7994354421344978,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.7148,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.9547887154481206,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.7305,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.8419282983446993,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.7112,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.8172700043560845,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.7131,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.9765213342565793,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.7068,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.8037906708063665,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.7023,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.7867836114885146,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.7089,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.9452960916004703,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.9312,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.7612295632163256,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.6469,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.8790900939266659,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.6854,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.6722911813478583,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.6264,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.8284270125532621,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.6853,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.5923231874699565,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.6191,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.677938105073477,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.6562,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.6445750513490396,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.6591,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.6591398559850276,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.6612,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.6785355561827732,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.6241,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.6640277076291504,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.6607,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.5723424852123583,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.5973,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.6129280945530802,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.6609,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.6123091193951057,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.6566,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.563780665966425,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.5956,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.6745099171218181,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.6094,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.6182295743556719,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.6534,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.6871798376763909,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.6472,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.7931097277491445,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.6351,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.7237136732899917,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.6369,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.6778553571142459,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.6105,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.833449444590497,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.6305,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.6151138014005336,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.6107,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.7431336710612528,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.6388,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.596509290604107,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.5825,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.7103672796228194,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.6686,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.6569967513611935,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.6024,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.7545111659846401,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.6166,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.7915012712287626,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.6062,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.7255250890449719,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.6111,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.644262805662098,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.6015,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.7171018109103273,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.6214,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.8148776134831176,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.7401,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.6222837355511308,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.6623,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.639255387373808,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.5927,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.5593358910252636,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.5582,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.6197086381920219,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.5404,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.5802149367130685,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.6123,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.538474276913493,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.5055,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.5199149831200351,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.5698,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.4857710491511767,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4924,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.5948737890772205,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.5793,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.6721221764728632,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.5658,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.5488349407810451,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.5228,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.6313108503170125,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.5833,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.8000920793639112,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.5486,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.7119338711789602,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.5026,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.7164819120687533,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.5361,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.7685277648765818,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.5752,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.5671292352052345,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.5008,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.6500481756936068,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.5641,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.6072746509141675,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.5247,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.5576438241361014,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.5518,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.7536130329548041,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.5802,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.6249046196175972,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.4849,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.5897098153304258,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.53,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.7724460281117684,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.577,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.5415458791892559,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5144,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.6243505252839292,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.504,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.5829233737143158,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.5807,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.6610733522406311,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.5403,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.6314309911410616,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.5283,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.5996596372228271,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.5485,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.6560318990398022,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.5511,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.47042230492960135,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.5387,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.7105360844787569,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.7318,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.6414075588950308,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.4751,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.5253463695637456,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.482,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.774967432159217,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.4803,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.5868652627563569,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.4788,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.820794326242705,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.4749,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.6077241504221031,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.4676,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.594783508239959,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.4607,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.6533478854426843,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.4734,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.5723197211508168,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.4784,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.5519310549231674,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.4598,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.6227249058550656,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.4547,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5651010498903033,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.4659,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.5822811679193834,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.4647,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5819816802343004,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.4707,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.5660083107300502,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.4654,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5185498875411866,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.4663,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.5116411329423655,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.4712,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.5209997436098156,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.4743,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.4871230760841342,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.4617,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.457102675321537,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.4745,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.5127974866098665,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.4527,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.42245001835461105,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.4708,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.49188080220386793,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.4584,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.44167618457063434,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.4712,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.5039688673027475,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.4615,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.4205361183926625,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.4625,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.43233021008075545,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.4751,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.487910224415194,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.476,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.4030581513496535,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.4764,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.4393249352179175,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.4693,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.44104563278521997,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.4707,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.4002212987011352,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.4651,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.648026859841576,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.6193,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.561221316671484,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.4262,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.45269085949811677,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.3715,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.6779101453871519,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.4212,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.5497678034724917,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.398,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.5003797530147505,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.3865,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.5580549407868024,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.407,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.5172188041194234,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.38,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.48915210633040296,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.4108,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.558908230314928,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.4117,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.49602986574082164,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.3829,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.4609800715774028,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.4057,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.46096456188788704,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.3917,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.4585718750789489,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.4182,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.41614902867794823,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.3778,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.4297531850892571,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.372,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.4404023195294183,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.408,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.45167223263998,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.3948,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.41539136420911377,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.3861,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.42074220510910704,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.4038,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.4141695669267265,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.3865,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.41079371804642845,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.4039,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.3884188627873057,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.3861,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.37953254241519824,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.4395,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.39987415094795503,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.4035,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.3877759370729739,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.3661,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.389059575644023,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4107,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.345400783073805,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.3779,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.39166180661464217,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.4036,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.38343303718359834,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.4067,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.3880946514795974,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.4307,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.34848503794352365,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.3579,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.45449586177166873,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.5092,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.5062638334113145,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.3992,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.5312794593990976,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.3741,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.3549986667359863,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.3439,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.42625586846437513,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.3518,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.6145040463567123,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.3535,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.5190472940259413,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.3497,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.4037654231868124,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.3744,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.39010492521577306,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.3417,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.43621092475439815,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.3453,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.45622542720212206,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.3755,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.3808402384317074,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.3329,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.37053760936822366,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.3386,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.4073764552671374,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.3327,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.4290701109825632,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.3602,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.40941863501824033,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.3285,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.4139293691174814,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.3701,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.3777113984274894,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.3447,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.36629086304895164,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.3332,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.35730174201686227,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.3255,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.33898170898210067,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.3236,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.35556052455015685,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.3343,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.3628091740313485,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.3671,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.3528887633239016,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.3627,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.35540816712948214,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.3452,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.35604246573404463,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.3251,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.3322343669563559,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.3543,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.34493178170408967,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.3578,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.3593545359875567,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.336,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.359523898968028,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.3468,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.35103113454869783,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.335,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.35416910936066975,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.3708,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.34424571317134517,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.3713,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.45934808464555765,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.4364,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.44962762325821687,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.3312,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.4478929520440433,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.3133,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.38955085059958916,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.3235,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.34255835552872055,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.3187,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.3181615279992102,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.3285,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.3563527291119814,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.3165,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.3401868281682868,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.3267,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.35960005423937524,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.3341,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.3763544106961004,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.3216,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.40579998510222787,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.3374,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.3686629640169421,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.3172,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.3799925645700284,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.3332,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.33580814479557997,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.3177,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.3583375161475769,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.3235,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.33920282172817046,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.3299,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.33097595841372207,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.3187,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.32482371977032953,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.3283,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.3166975241806231,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.3212,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.30366787288195024,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.327,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.33313199684156974,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.3241,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.3186026476638579,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.3321,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.3453794528323902,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.3053,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.3205422906400857,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.325,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.3132049065793132,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.3388,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.3185655590073017,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.329,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.3054909895467588,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.3282,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.3155921368834683,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.32,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.33481857262300907,
      "learning_rate": 0.0,
      "loss": 0.3219,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 5.742441502835999e+17,
      "train_loss": 0.5152867310015219,
      "train_runtime": 24342.4741,
      "train_samples_per_second": 0.909,
      "train_steps_per_second": 0.009
    }
  ],
  "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": 5.742441502835999e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}