{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 231,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004329004329004329,
      "grad_norm": 6.30618953704834,
      "learning_rate": 0.0,
      "loss": 0.4668,
      "step": 1
    },
    {
      "epoch": 0.008658008658008658,
      "grad_norm": 3.96525502204895,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.3887,
      "step": 2
    },
    {
      "epoch": 0.012987012987012988,
      "grad_norm": 4.628964424133301,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.4199,
      "step": 3
    },
    {
      "epoch": 0.017316017316017316,
      "grad_norm": 4.340306282043457,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.3379,
      "step": 4
    },
    {
      "epoch": 0.021645021645021644,
      "grad_norm": 10.413896560668945,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.4785,
      "step": 5
    },
    {
      "epoch": 0.025974025974025976,
      "grad_norm": 3.5524611473083496,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.3965,
      "step": 6
    },
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 3.9681403636932373,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.334,
      "step": 7
    },
    {
      "epoch": 0.03463203463203463,
      "grad_norm": 6.837435722351074,
      "learning_rate": 0.0001,
      "loss": 0.4609,
      "step": 8
    },
    {
      "epoch": 0.03896103896103896,
      "grad_norm": 6.86609411239624,
      "learning_rate": 9.999508258797877e-05,
      "loss": 0.4297,
      "step": 9
    },
    {
      "epoch": 0.04329004329004329,
      "grad_norm": 5.7420973777771,
      "learning_rate": 9.998033131915266e-05,
      "loss": 0.3691,
      "step": 10
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 5.456467151641846,
      "learning_rate": 9.995574909504435e-05,
      "loss": 0.4023,
      "step": 11
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 6.083012104034424,
      "learning_rate": 9.992134075089084e-05,
      "loss": 0.498,
      "step": 12
    },
    {
      "epoch": 0.05627705627705628,
      "grad_norm": 24.94365692138672,
      "learning_rate": 9.987711305469231e-05,
      "loss": 0.4551,
      "step": 13
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 9.905355453491211,
      "learning_rate": 9.982307470588098e-05,
      "loss": 0.4453,
      "step": 14
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 6.808415412902832,
      "learning_rate": 9.975923633360985e-05,
      "loss": 0.4434,
      "step": 15
    },
    {
      "epoch": 0.06926406926406926,
      "grad_norm": 8.532634735107422,
      "learning_rate": 9.968561049466214e-05,
      "loss": 0.4043,
      "step": 16
    },
    {
      "epoch": 0.0735930735930736,
      "grad_norm": 7.474434852600098,
      "learning_rate": 9.960221167098124e-05,
      "loss": 0.4473,
      "step": 17
    },
    {
      "epoch": 0.07792207792207792,
      "grad_norm": 31.124326705932617,
      "learning_rate": 9.950905626682228e-05,
      "loss": 0.3809,
      "step": 18
    },
    {
      "epoch": 0.08225108225108226,
      "grad_norm": 7.993480205535889,
      "learning_rate": 9.940616260552544e-05,
      "loss": 0.4863,
      "step": 19
    },
    {
      "epoch": 0.08658008658008658,
      "grad_norm": 6.659813404083252,
      "learning_rate": 9.92935509259118e-05,
      "loss": 0.3477,
      "step": 20
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 8.14820671081543,
      "learning_rate": 9.917124337830243e-05,
      "loss": 0.3809,
      "step": 21
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 6.370754718780518,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.4727,
      "step": 22
    },
    {
      "epoch": 0.09956709956709957,
      "grad_norm": 10.538797378540039,
      "learning_rate": 9.889763881136439e-05,
      "loss": 0.5156,
      "step": 23
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 8.41638469696045,
      "learning_rate": 9.874639560909117e-05,
      "loss": 0.4727,
      "step": 24
    },
    {
      "epoch": 0.10822510822510822,
      "grad_norm": 11.677297592163086,
      "learning_rate": 9.858556416234755e-05,
      "loss": 0.457,
      "step": 25
    },
    {
      "epoch": 0.11255411255411256,
      "grad_norm": 14.953546524047852,
      "learning_rate": 9.841517610611309e-05,
      "loss": 0.4336,
      "step": 26
    },
    {
      "epoch": 0.11688311688311688,
      "grad_norm": 7.516232013702393,
      "learning_rate": 9.82352649551188e-05,
      "loss": 0.3516,
      "step": 27
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 7.574983596801758,
      "learning_rate": 9.804586609725499e-05,
      "loss": 0.3926,
      "step": 28
    },
    {
      "epoch": 0.12554112554112554,
      "grad_norm": 17.778465270996094,
      "learning_rate": 9.784701678661045e-05,
      "loss": 0.4082,
      "step": 29
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 22.582393646240234,
      "learning_rate": 9.763875613614482e-05,
      "loss": 0.4688,
      "step": 30
    },
    {
      "epoch": 0.1341991341991342,
      "grad_norm": 156.63107299804688,
      "learning_rate": 9.742112510999515e-05,
      "loss": 0.4902,
      "step": 31
    },
    {
      "epoch": 0.13852813852813853,
      "grad_norm": 22.96047592163086,
      "learning_rate": 9.719416651541839e-05,
      "loss": 0.4336,
      "step": 32
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 10.698134422302246,
      "learning_rate": 9.69579249943714e-05,
      "loss": 0.4414,
      "step": 33
    },
    {
      "epoch": 0.1471861471861472,
      "grad_norm": 7.732542514801025,
      "learning_rate": 9.671244701472999e-05,
      "loss": 0.3789,
      "step": 34
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 20.447450637817383,
      "learning_rate": 9.645778086114892e-05,
      "loss": 0.4551,
      "step": 35
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 7.715920925140381,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.4863,
      "step": 36
    },
    {
      "epoch": 0.16017316017316016,
      "grad_norm": 11.215670585632324,
      "learning_rate": 9.592108619734106e-05,
      "loss": 0.4746,
      "step": 37
    },
    {
      "epoch": 0.1645021645021645,
      "grad_norm": 14.483991622924805,
      "learning_rate": 9.563916325306594e-05,
      "loss": 0.4062,
      "step": 38
    },
    {
      "epoch": 0.16883116883116883,
      "grad_norm": 9.34101390838623,
      "learning_rate": 9.534826324599003e-05,
      "loss": 0.4941,
      "step": 39
    },
    {
      "epoch": 0.17316017316017315,
      "grad_norm": 9.180691719055176,
      "learning_rate": 9.504844339512095e-05,
      "loss": 0.4707,
      "step": 40
    },
    {
      "epoch": 0.1774891774891775,
      "grad_norm": 5.87382173538208,
      "learning_rate": 9.473976267396831e-05,
      "loss": 0.5078,
      "step": 41
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 7.854617118835449,
      "learning_rate": 9.442228179894362e-05,
      "loss": 0.4746,
      "step": 42
    },
    {
      "epoch": 0.18614718614718614,
      "grad_norm": 5.667612075805664,
      "learning_rate": 9.409606321741775e-05,
      "loss": 0.5156,
      "step": 43
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 6.341343402862549,
      "learning_rate": 9.376117109543769e-05,
      "loss": 0.498,
      "step": 44
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 6.829994201660156,
      "learning_rate": 9.341767130510528e-05,
      "loss": 0.3711,
      "step": 45
    },
    {
      "epoch": 0.19913419913419914,
      "grad_norm": 11.078237533569336,
      "learning_rate": 9.306563141162046e-05,
      "loss": 0.4434,
      "step": 46
    },
    {
      "epoch": 0.20346320346320346,
      "grad_norm": 9.45587158203125,
      "learning_rate": 9.270512065999137e-05,
      "loss": 0.5234,
      "step": 47
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 6.764471530914307,
      "learning_rate": 9.233620996141421e-05,
      "loss": 0.4219,
      "step": 48
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 23.972339630126953,
      "learning_rate": 9.195897187932512e-05,
      "loss": 0.3965,
      "step": 49
    },
    {
      "epoch": 0.21645021645021645,
      "grad_norm": 5.304466724395752,
      "learning_rate": 9.157348061512727e-05,
      "loss": 0.418,
      "step": 50
    },
    {
      "epoch": 0.22077922077922077,
      "grad_norm": 5.903518199920654,
      "learning_rate": 9.117981199359574e-05,
      "loss": 0.2969,
      "step": 51
    },
    {
      "epoch": 0.22510822510822512,
      "grad_norm": 6.200258731842041,
      "learning_rate": 9.077804344796302e-05,
      "loss": 0.4551,
      "step": 52
    },
    {
      "epoch": 0.22943722943722944,
      "grad_norm": 7.089105606079102,
      "learning_rate": 9.036825400468812e-05,
      "loss": 0.457,
      "step": 53
    },
    {
      "epoch": 0.23376623376623376,
      "grad_norm": 34.66611862182617,
      "learning_rate": 8.995052426791247e-05,
      "loss": 0.418,
      "step": 54
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 5.267077922821045,
      "learning_rate": 8.952493640360517e-05,
      "loss": 0.4082,
      "step": 55
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 6.587023735046387,
      "learning_rate": 8.90915741234015e-05,
      "loss": 0.4512,
      "step": 56
    },
    {
      "epoch": 0.24675324675324675,
      "grad_norm": 7.729217529296875,
      "learning_rate": 8.865052266813685e-05,
      "loss": 0.4258,
      "step": 57
    },
    {
      "epoch": 0.2510822510822511,
      "grad_norm": 6.812266826629639,
      "learning_rate": 8.820186879108038e-05,
      "loss": 0.4512,
      "step": 58
    },
    {
      "epoch": 0.2554112554112554,
      "grad_norm": 11.491703033447266,
      "learning_rate": 8.77457007408708e-05,
      "loss": 0.2832,
      "step": 59
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 5.020540714263916,
      "learning_rate": 8.728210824415827e-05,
      "loss": 0.3496,
      "step": 60
    },
    {
      "epoch": 0.26406926406926406,
      "grad_norm": 5.021598815917969,
      "learning_rate": 8.681118248795547e-05,
      "loss": 0.5,
      "step": 61
    },
    {
      "epoch": 0.2683982683982684,
      "grad_norm": 8.810498237609863,
      "learning_rate": 8.633301610170135e-05,
      "loss": 0.4551,
      "step": 62
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 7.0777411460876465,
      "learning_rate": 8.584770313904137e-05,
      "loss": 0.3926,
      "step": 63
    },
    {
      "epoch": 0.27705627705627706,
      "grad_norm": 5.617433547973633,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.3223,
      "step": 64
    },
    {
      "epoch": 0.2813852813852814,
      "grad_norm": 5.157428741455078,
      "learning_rate": 8.485602070884117e-05,
      "loss": 0.373,
      "step": 65
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 6.726741790771484,
      "learning_rate": 8.434984630174509e-05,
      "loss": 0.4219,
      "step": 66
    },
    {
      "epoch": 0.29004329004329005,
      "grad_norm": 4.791531085968018,
      "learning_rate": 8.383691540076371e-05,
      "loss": 0.4199,
      "step": 67
    },
    {
      "epoch": 0.2943722943722944,
      "grad_norm": 5.5883026123046875,
      "learning_rate": 8.33173288976002e-05,
      "loss": 0.459,
      "step": 68
    },
    {
      "epoch": 0.2987012987012987,
      "grad_norm": 7.0759663581848145,
      "learning_rate": 8.279118899309122e-05,
      "loss": 0.4629,
      "step": 69
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 7.941741466522217,
      "learning_rate": 8.225859917710439e-05,
      "loss": 0.4316,
      "step": 70
    },
    {
      "epoch": 0.30735930735930733,
      "grad_norm": 4.9386396408081055,
      "learning_rate": 8.171966420818228e-05,
      "loss": 0.3359,
      "step": 71
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 4.561506748199463,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.373,
      "step": 72
    },
    {
      "epoch": 0.31601731601731603,
      "grad_norm": 6.921361446380615,
      "learning_rate": 8.062318406519751e-05,
      "loss": 0.3809,
      "step": 73
    },
    {
      "epoch": 0.3203463203463203,
      "grad_norm": 6.0779876708984375,
      "learning_rate": 8.006585456492029e-05,
      "loss": 0.416,
      "step": 74
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 14.405391693115234,
      "learning_rate": 7.950261121685641e-05,
      "loss": 0.3711,
      "step": 75
    },
    {
      "epoch": 0.329004329004329,
      "grad_norm": 12.624444961547852,
      "learning_rate": 7.89335648089903e-05,
      "loss": 0.3691,
      "step": 76
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 8.761910438537598,
      "learning_rate": 7.835882727074779e-05,
      "loss": 0.4707,
      "step": 77
    },
    {
      "epoch": 0.33766233766233766,
      "grad_norm": 6.742737770080566,
      "learning_rate": 7.777851165098012e-05,
      "loss": 0.2949,
      "step": 78
    },
    {
      "epoch": 0.341991341991342,
      "grad_norm": 13.422913551330566,
      "learning_rate": 7.719273209572744e-05,
      "loss": 0.377,
      "step": 79
    },
    {
      "epoch": 0.3463203463203463,
      "grad_norm": 8.455282211303711,
      "learning_rate": 7.660160382576683e-05,
      "loss": 0.4707,
      "step": 80
    },
    {
      "epoch": 0.35064935064935066,
      "grad_norm": 7.551178455352783,
      "learning_rate": 7.600524311394873e-05,
      "loss": 0.4688,
      "step": 81
    },
    {
      "epoch": 0.354978354978355,
      "grad_norm": 5.217031002044678,
      "learning_rate": 7.540376726232648e-05,
      "loss": 0.4492,
      "step": 82
    },
    {
      "epoch": 0.3593073593073593,
      "grad_norm": 7.432611465454102,
      "learning_rate": 7.47972945790834e-05,
      "loss": 0.4121,
      "step": 83
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 10.472357749938965,
      "learning_rate": 7.4185944355262e-05,
      "loss": 0.4199,
      "step": 84
    },
    {
      "epoch": 0.36796536796536794,
      "grad_norm": 8.361776351928711,
      "learning_rate": 7.35698368412999e-05,
      "loss": 0.3496,
      "step": 85
    },
    {
      "epoch": 0.3722943722943723,
      "grad_norm": 8.236878395080566,
      "learning_rate": 7.294909322337689e-05,
      "loss": 0.3906,
      "step": 86
    },
    {
      "epoch": 0.37662337662337664,
      "grad_norm": 6.491207122802734,
      "learning_rate": 7.232383559957814e-05,
      "loss": 0.4277,
      "step": 87
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 9.313511848449707,
      "learning_rate": 7.169418695587791e-05,
      "loss": 0.3867,
      "step": 88
    },
    {
      "epoch": 0.3852813852813853,
      "grad_norm": 13.22208309173584,
      "learning_rate": 7.106027114194855e-05,
      "loss": 0.4824,
      "step": 89
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 7.9978718757629395,
      "learning_rate": 7.042221284679982e-05,
      "loss": 0.4238,
      "step": 90
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 14.92418384552002,
      "learning_rate": 6.978013757425295e-05,
      "loss": 0.4941,
      "step": 91
    },
    {
      "epoch": 0.39826839826839827,
      "grad_norm": 8.810619354248047,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.3691,
      "step": 92
    },
    {
      "epoch": 0.4025974025974026,
      "grad_norm": 7.104898452758789,
      "learning_rate": 6.848444203803476e-05,
      "loss": 0.4609,
      "step": 93
    },
    {
      "epoch": 0.4069264069264069,
      "grad_norm": 30.174938201904297,
      "learning_rate": 6.783107663311565e-05,
      "loss": 0.375,
      "step": 94
    },
    {
      "epoch": 0.41125541125541126,
      "grad_norm": 6.395003318786621,
      "learning_rate": 6.717420391817306e-05,
      "loss": 0.4043,
      "step": 95
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 17.24394416809082,
      "learning_rate": 6.651395309775837e-05,
      "loss": 0.4434,
      "step": 96
    },
    {
      "epoch": 0.4199134199134199,
      "grad_norm": 6.928684711456299,
      "learning_rate": 6.585045404088441e-05,
      "loss": 0.4141,
      "step": 97
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 14.082657814025879,
      "learning_rate": 6.518383725548074e-05,
      "loss": 0.4414,
      "step": 98
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 21.20639991760254,
      "learning_rate": 6.451423386272312e-05,
      "loss": 0.4121,
      "step": 99
    },
    {
      "epoch": 0.4329004329004329,
      "grad_norm": 12.427319526672363,
      "learning_rate": 6.384177557124247e-05,
      "loss": 0.4258,
      "step": 100
    },
    {
      "epoch": 0.43722943722943725,
      "grad_norm": 10.986278533935547,
      "learning_rate": 6.316659465121824e-05,
      "loss": 0.3965,
      "step": 101
    },
    {
      "epoch": 0.44155844155844154,
      "grad_norm": 12.830971717834473,
      "learning_rate": 6.248882390836135e-05,
      "loss": 0.4375,
      "step": 102
    },
    {
      "epoch": 0.4458874458874459,
      "grad_norm": 14.790221214294434,
      "learning_rate": 6.180859665779172e-05,
      "loss": 0.4492,
      "step": 103
    },
    {
      "epoch": 0.45021645021645024,
      "grad_norm": 44.620018005371094,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.5703,
      "step": 104
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 13.396041870117188,
      "learning_rate": 6.04413082836085e-05,
      "loss": 0.4805,
      "step": 105
    },
    {
      "epoch": 0.4588744588744589,
      "grad_norm": 13.93975830078125,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 0.3809,
      "step": 106
    },
    {
      "epoch": 0.46320346320346323,
      "grad_norm": 18.1014404296875,
      "learning_rate": 5.9065805239014923e-05,
      "loss": 0.4707,
      "step": 107
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 7.800663948059082,
      "learning_rate": 5.837531116523682e-05,
      "loss": 0.4121,
      "step": 108
    },
    {
      "epoch": 0.47186147186147187,
      "grad_norm": 9.331329345703125,
      "learning_rate": 5.76831696972265e-05,
      "loss": 0.3926,
      "step": 109
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 8.944900512695312,
      "learning_rate": 5.698951697677498e-05,
      "loss": 0.3262,
      "step": 110
    },
    {
      "epoch": 0.4805194805194805,
      "grad_norm": 20.515888214111328,
      "learning_rate": 5.629448944293127e-05,
      "loss": 0.4648,
      "step": 111
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 9.60425853729248,
      "learning_rate": 5.559822380516539e-05,
      "loss": 0.4023,
      "step": 112
    },
    {
      "epoch": 0.48917748917748916,
      "grad_norm": 9.976903915405273,
      "learning_rate": 5.490085701647805e-05,
      "loss": 0.3438,
      "step": 113
    },
    {
      "epoch": 0.4935064935064935,
      "grad_norm": 19.132675170898438,
      "learning_rate": 5.420252624646238e-05,
      "loss": 0.3965,
      "step": 114
    },
    {
      "epoch": 0.49783549783549785,
      "grad_norm": 12.469053268432617,
      "learning_rate": 5.3503368854323366e-05,
      "loss": 0.3848,
      "step": 115
    },
    {
      "epoch": 0.5021645021645021,
      "grad_norm": 14.417160034179688,
      "learning_rate": 5.2803522361859594e-05,
      "loss": 0.459,
      "step": 116
    },
    {
      "epoch": 0.5064935064935064,
      "grad_norm": 8.167932510375977,
      "learning_rate": 5.2103124426413264e-05,
      "loss": 0.3711,
      "step": 117
    },
    {
      "epoch": 0.5108225108225108,
      "grad_norm": 9.628777503967285,
      "learning_rate": 5.140231281379345e-05,
      "loss": 0.2695,
      "step": 118
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 13.074538230895996,
      "learning_rate": 5.070122537117812e-05,
      "loss": 0.3633,
      "step": 119
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 8.415633201599121,
      "learning_rate": 5e-05,
      "loss": 0.4395,
      "step": 120
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 16.27584457397461,
      "learning_rate": 4.929877462882189e-05,
      "loss": 0.416,
      "step": 121
    },
    {
      "epoch": 0.5281385281385281,
      "grad_norm": 5.963536262512207,
      "learning_rate": 4.859768718620656e-05,
      "loss": 0.3633,
      "step": 122
    },
    {
      "epoch": 0.5324675324675324,
      "grad_norm": 17.088741302490234,
      "learning_rate": 4.7896875573586755e-05,
      "loss": 0.3672,
      "step": 123
    },
    {
      "epoch": 0.5367965367965368,
      "grad_norm": 16.568267822265625,
      "learning_rate": 4.7196477638140404e-05,
      "loss": 0.4121,
      "step": 124
    },
    {
      "epoch": 0.5411255411255411,
      "grad_norm": 15.208264350891113,
      "learning_rate": 4.649663114567663e-05,
      "loss": 0.4082,
      "step": 125
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 11.11156177520752,
      "learning_rate": 4.579747375353763e-05,
      "loss": 0.375,
      "step": 126
    },
    {
      "epoch": 0.5497835497835498,
      "grad_norm": 15.04819107055664,
      "learning_rate": 4.509914298352197e-05,
      "loss": 0.3594,
      "step": 127
    },
    {
      "epoch": 0.5541125541125541,
      "grad_norm": 6.2651286125183105,
      "learning_rate": 4.4401776194834613e-05,
      "loss": 0.2793,
      "step": 128
    },
    {
      "epoch": 0.5584415584415584,
      "grad_norm": 9.082189559936523,
      "learning_rate": 4.370551055706874e-05,
      "loss": 0.3906,
      "step": 129
    },
    {
      "epoch": 0.5627705627705628,
      "grad_norm": 7.258723258972168,
      "learning_rate": 4.3010483023225045e-05,
      "loss": 0.3086,
      "step": 130
    },
    {
      "epoch": 0.5670995670995671,
      "grad_norm": 9.980514526367188,
      "learning_rate": 4.231683030277349e-05,
      "loss": 0.4043,
      "step": 131
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 13.367783546447754,
      "learning_rate": 4.162468883476319e-05,
      "loss": 0.3223,
      "step": 132
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 5.13516902923584,
      "learning_rate": 4.093419476098509e-05,
      "loss": 0.3027,
      "step": 133
    },
    {
      "epoch": 0.5800865800865801,
      "grad_norm": 6.814506530761719,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 0.4395,
      "step": 134
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 9.574424743652344,
      "learning_rate": 3.955869171639152e-05,
      "loss": 0.3027,
      "step": 135
    },
    {
      "epoch": 0.5887445887445888,
      "grad_norm": 6.566686630249023,
      "learning_rate": 3.887395330218429e-05,
      "loss": 0.3262,
      "step": 136
    },
    {
      "epoch": 0.5930735930735931,
      "grad_norm": 14.991905212402344,
      "learning_rate": 3.81914033422083e-05,
      "loss": 0.3926,
      "step": 137
    },
    {
      "epoch": 0.5974025974025974,
      "grad_norm": 6.989403247833252,
      "learning_rate": 3.7511176091638653e-05,
      "loss": 0.416,
      "step": 138
    },
    {
      "epoch": 0.6017316017316018,
      "grad_norm": 99.71111297607422,
      "learning_rate": 3.683340534878176e-05,
      "loss": 0.3555,
      "step": 139
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 7.185469150543213,
      "learning_rate": 3.6158224428757535e-05,
      "loss": 0.3555,
      "step": 140
    },
    {
      "epoch": 0.6103896103896104,
      "grad_norm": 48.649051666259766,
      "learning_rate": 3.5485766137276894e-05,
      "loss": 0.3906,
      "step": 141
    },
    {
      "epoch": 0.6147186147186147,
      "grad_norm": 8.177955627441406,
      "learning_rate": 3.4816162744519263e-05,
      "loss": 0.4043,
      "step": 142
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 7.434495449066162,
      "learning_rate": 3.4149545959115605e-05,
      "loss": 0.3867,
      "step": 143
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 10.178008079528809,
      "learning_rate": 3.3486046902241664e-05,
      "loss": 0.4141,
      "step": 144
    },
    {
      "epoch": 0.6277056277056277,
      "grad_norm": 6.520373821258545,
      "learning_rate": 3.282579608182694e-05,
      "loss": 0.3301,
      "step": 145
    },
    {
      "epoch": 0.6320346320346321,
      "grad_norm": 15.099567413330078,
      "learning_rate": 3.216892336688435e-05,
      "loss": 0.3496,
      "step": 146
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 14.876191139221191,
      "learning_rate": 3.151555796196525e-05,
      "loss": 0.3691,
      "step": 147
    },
    {
      "epoch": 0.6406926406926406,
      "grad_norm": 10.065027236938477,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.3379,
      "step": 148
    },
    {
      "epoch": 0.645021645021645,
      "grad_norm": 7.905645370483398,
      "learning_rate": 3.021986242574707e-05,
      "loss": 0.4004,
      "step": 149
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 10.962891578674316,
      "learning_rate": 2.9577787153200197e-05,
      "loss": 0.4141,
      "step": 150
    },
    {
      "epoch": 0.6536796536796536,
      "grad_norm": 6.561282157897949,
      "learning_rate": 2.893972885805148e-05,
      "loss": 0.4844,
      "step": 151
    },
    {
      "epoch": 0.658008658008658,
      "grad_norm": 15.016473770141602,
      "learning_rate": 2.8305813044122097e-05,
      "loss": 0.3418,
      "step": 152
    },
    {
      "epoch": 0.6623376623376623,
      "grad_norm": 8.82343864440918,
      "learning_rate": 2.7676164400421862e-05,
      "loss": 0.3438,
      "step": 153
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 6.84242582321167,
      "learning_rate": 2.705090677662311e-05,
      "loss": 0.3438,
      "step": 154
    },
    {
      "epoch": 0.670995670995671,
      "grad_norm": 7.3810930252075195,
      "learning_rate": 2.6430163158700115e-05,
      "loss": 0.3906,
      "step": 155
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 8.569120407104492,
      "learning_rate": 2.581405564473801e-05,
      "loss": 0.4004,
      "step": 156
    },
    {
      "epoch": 0.6796536796536796,
      "grad_norm": 13.025605201721191,
      "learning_rate": 2.5202705420916627e-05,
      "loss": 0.373,
      "step": 157
    },
    {
      "epoch": 0.683982683982684,
      "grad_norm": 8.868095397949219,
      "learning_rate": 2.459623273767354e-05,
      "loss": 0.4199,
      "step": 158
    },
    {
      "epoch": 0.6883116883116883,
      "grad_norm": 6.385418891906738,
      "learning_rate": 2.3994756886051268e-05,
      "loss": 0.3203,
      "step": 159
    },
    {
      "epoch": 0.6926406926406926,
      "grad_norm": 8.916970252990723,
      "learning_rate": 2.3398396174233178e-05,
      "loss": 0.2949,
      "step": 160
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 9.499614715576172,
      "learning_rate": 2.280726790427258e-05,
      "loss": 0.2832,
      "step": 161
    },
    {
      "epoch": 0.7012987012987013,
      "grad_norm": 14.341413497924805,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.3125,
      "step": 162
    },
    {
      "epoch": 0.7056277056277056,
      "grad_norm": 6.781824111938477,
      "learning_rate": 2.164117272925221e-05,
      "loss": 0.332,
      "step": 163
    },
    {
      "epoch": 0.70995670995671,
      "grad_norm": 44.00325393676758,
      "learning_rate": 2.1066435191009715e-05,
      "loss": 0.4141,
      "step": 164
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 13.864439010620117,
      "learning_rate": 2.0497388783143602e-05,
      "loss": 0.4199,
      "step": 165
    },
    {
      "epoch": 0.7186147186147186,
      "grad_norm": 8.958086967468262,
      "learning_rate": 1.9934145435079702e-05,
      "loss": 0.3535,
      "step": 166
    },
    {
      "epoch": 0.7229437229437229,
      "grad_norm": 4.822065830230713,
      "learning_rate": 1.9376815934802496e-05,
      "loss": 0.3828,
      "step": 167
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 8.322967529296875,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.4258,
      "step": 168
    },
    {
      "epoch": 0.7316017316017316,
      "grad_norm": 13.969364166259766,
      "learning_rate": 1.8280335791817733e-05,
      "loss": 0.3672,
      "step": 169
    },
    {
      "epoch": 0.7359307359307359,
      "grad_norm": 6.8569722175598145,
      "learning_rate": 1.774140082289563e-05,
      "loss": 0.3789,
      "step": 170
    },
    {
      "epoch": 0.7402597402597403,
      "grad_norm": 6.253945827484131,
      "learning_rate": 1.7208811006908798e-05,
      "loss": 0.3379,
      "step": 171
    },
    {
      "epoch": 0.7445887445887446,
      "grad_norm": 5.97335958480835,
      "learning_rate": 1.6682671102399805e-05,
      "loss": 0.3672,
      "step": 172
    },
    {
      "epoch": 0.7489177489177489,
      "grad_norm": 21.618064880371094,
      "learning_rate": 1.6163084599236278e-05,
      "loss": 0.4297,
      "step": 173
    },
    {
      "epoch": 0.7532467532467533,
      "grad_norm": 6.355106353759766,
      "learning_rate": 1.5650153698254916e-05,
      "loss": 0.2969,
      "step": 174
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 11.884471893310547,
      "learning_rate": 1.5143979291158838e-05,
      "loss": 0.4531,
      "step": 175
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 8.2210054397583,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.3535,
      "step": 176
    },
    {
      "epoch": 0.7662337662337663,
      "grad_norm": 5.43289852142334,
      "learning_rate": 1.4152296860958642e-05,
      "loss": 0.3281,
      "step": 177
    },
    {
      "epoch": 0.7705627705627706,
      "grad_norm": 13.447378158569336,
      "learning_rate": 1.3666983898298657e-05,
      "loss": 0.4336,
      "step": 178
    },
    {
      "epoch": 0.7748917748917749,
      "grad_norm": 9.235896110534668,
      "learning_rate": 1.3188817512044544e-05,
      "loss": 0.3516,
      "step": 179
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 6.7336602210998535,
      "learning_rate": 1.2717891755841722e-05,
      "loss": 0.3223,
      "step": 180
    },
    {
      "epoch": 0.7835497835497836,
      "grad_norm": 6.973309516906738,
      "learning_rate": 1.225429925912921e-05,
      "loss": 0.3301,
      "step": 181
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 30.55255699157715,
      "learning_rate": 1.1798131208919627e-05,
      "loss": 0.3027,
      "step": 182
    },
    {
      "epoch": 0.7922077922077922,
      "grad_norm": 10.4804048538208,
      "learning_rate": 1.134947733186315e-05,
      "loss": 0.3848,
      "step": 183
    },
    {
      "epoch": 0.7965367965367965,
      "grad_norm": 5.046676158905029,
      "learning_rate": 1.090842587659851e-05,
      "loss": 0.4805,
      "step": 184
    },
    {
      "epoch": 0.8008658008658008,
      "grad_norm": 8.961450576782227,
      "learning_rate": 1.047506359639483e-05,
      "loss": 0.3398,
      "step": 185
    },
    {
      "epoch": 0.8051948051948052,
      "grad_norm": 10.69762134552002,
      "learning_rate": 1.004947573208756e-05,
      "loss": 0.3945,
      "step": 186
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 9.194847106933594,
      "learning_rate": 9.63174599531188e-06,
      "loss": 0.3379,
      "step": 187
    },
    {
      "epoch": 0.8138528138528138,
      "grad_norm": 7.735859394073486,
      "learning_rate": 9.221956552036992e-06,
      "loss": 0.373,
      "step": 188
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 7.634536266326904,
      "learning_rate": 8.820188006404268e-06,
      "loss": 0.3418,
      "step": 189
    },
    {
      "epoch": 0.8225108225108225,
      "grad_norm": 12.410811424255371,
      "learning_rate": 8.426519384872733e-06,
      "loss": 0.3125,
      "step": 190
    },
    {
      "epoch": 0.8268398268398268,
      "grad_norm": 44.61256408691406,
      "learning_rate": 8.041028120674893e-06,
      "loss": 0.3418,
      "step": 191
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 8.871509552001953,
      "learning_rate": 7.663790038585793e-06,
      "loss": 0.4141,
      "step": 192
    },
    {
      "epoch": 0.8354978354978355,
      "grad_norm": 7.508159160614014,
      "learning_rate": 7.2948793400086315e-06,
      "loss": 0.4004,
      "step": 193
    },
    {
      "epoch": 0.8398268398268398,
      "grad_norm": 33.30258560180664,
      "learning_rate": 6.934368588379553e-06,
      "loss": 0.3574,
      "step": 194
    },
    {
      "epoch": 0.8441558441558441,
      "grad_norm": 21.46922492980957,
      "learning_rate": 6.582328694894729e-06,
      "loss": 0.4023,
      "step": 195
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 31.504789352416992,
      "learning_rate": 6.238828904562316e-06,
      "loss": 0.377,
      "step": 196
    },
    {
      "epoch": 0.8528138528138528,
      "grad_norm": 9.979594230651855,
      "learning_rate": 5.903936782582253e-06,
      "loss": 0.4668,
      "step": 197
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 7.933041095733643,
      "learning_rate": 5.577718201056392e-06,
      "loss": 0.4102,
      "step": 198
    },
    {
      "epoch": 0.8614718614718615,
      "grad_norm": 5.786583423614502,
      "learning_rate": 5.260237326031697e-06,
      "loss": 0.4414,
      "step": 199
    },
    {
      "epoch": 0.8658008658008658,
      "grad_norm": 5.894907474517822,
      "learning_rate": 4.951556604879048e-06,
      "loss": 0.3848,
      "step": 200
    },
    {
      "epoch": 0.8701298701298701,
      "grad_norm": 32.59992599487305,
      "learning_rate": 4.651736754009972e-06,
      "loss": 0.3672,
      "step": 201
    },
    {
      "epoch": 0.8744588744588745,
      "grad_norm": 10.215508460998535,
      "learning_rate": 4.360836746934055e-06,
      "loss": 0.3496,
      "step": 202
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 10.185182571411133,
      "learning_rate": 4.078913802658946e-06,
      "loss": 0.4062,
      "step": 203
    },
    {
      "epoch": 0.8831168831168831,
      "grad_norm": 5.9914021492004395,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.3633,
      "step": 204
    },
    {
      "epoch": 0.8874458874458875,
      "grad_norm": 18.541730880737305,
      "learning_rate": 3.542219138851094e-06,
      "loss": 0.3691,
      "step": 205
    },
    {
      "epoch": 0.8917748917748918,
      "grad_norm": 5.432623863220215,
      "learning_rate": 3.2875529852700147e-06,
      "loss": 0.4297,
      "step": 206
    },
    {
      "epoch": 0.8961038961038961,
      "grad_norm": 15.46761417388916,
      "learning_rate": 3.0420750056286195e-06,
      "loss": 0.3379,
      "step": 207
    },
    {
      "epoch": 0.9004329004329005,
      "grad_norm": 6.300500392913818,
      "learning_rate": 2.8058334845816213e-06,
      "loss": 0.2754,
      "step": 208
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 8.409278869628906,
      "learning_rate": 2.5788748900048676e-06,
      "loss": 0.293,
      "step": 209
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 12.198765754699707,
      "learning_rate": 2.361243863855184e-06,
      "loss": 0.3789,
      "step": 210
    },
    {
      "epoch": 0.9134199134199135,
      "grad_norm": 6.002485275268555,
      "learning_rate": 2.152983213389559e-06,
      "loss": 0.3652,
      "step": 211
    },
    {
      "epoch": 0.9177489177489178,
      "grad_norm": 5.1724090576171875,
      "learning_rate": 1.9541339027450256e-06,
      "loss": 0.3125,
      "step": 212
    },
    {
      "epoch": 0.922077922077922,
      "grad_norm": 19.717180252075195,
      "learning_rate": 1.7647350448812106e-06,
      "loss": 0.3516,
      "step": 213
    },
    {
      "epoch": 0.9264069264069265,
      "grad_norm": 17.20496368408203,
      "learning_rate": 1.584823893886933e-06,
      "loss": 0.3789,
      "step": 214
    },
    {
      "epoch": 0.9307359307359307,
      "grad_norm": 6.270045280456543,
      "learning_rate": 1.4144358376524503e-06,
      "loss": 0.3672,
      "step": 215
    },
    {
      "epoch": 0.935064935064935,
      "grad_norm": 10.227100372314453,
      "learning_rate": 1.2536043909088191e-06,
      "loss": 0.4395,
      "step": 216
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 8.13603401184082,
      "learning_rate": 1.10236118863562e-06,
      "loss": 0.3418,
      "step": 217
    },
    {
      "epoch": 0.9437229437229437,
      "grad_norm": 7.209720611572266,
      "learning_rate": 9.607359798384785e-07,
      "loss": 0.3945,
      "step": 218
    },
    {
      "epoch": 0.948051948051948,
      "grad_norm": 8.675386428833008,
      "learning_rate": 8.287566216975795e-07,
      "loss": 0.3242,
      "step": 219
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 13.124395370483398,
      "learning_rate": 7.064490740882057e-07,
      "loss": 0.3398,
      "step": 220
    },
    {
      "epoch": 0.9567099567099567,
      "grad_norm": 5.071322917938232,
      "learning_rate": 5.938373944745612e-07,
      "loss": 0.2812,
      "step": 221
    },
    {
      "epoch": 0.961038961038961,
      "grad_norm": 6.736144542694092,
      "learning_rate": 4.909437331777179e-07,
      "loss": 0.3906,
      "step": 222
    },
    {
      "epoch": 0.9653679653679653,
      "grad_norm": 6.963472366333008,
      "learning_rate": 3.9778832901876675e-07,
      "loss": 0.3945,
      "step": 223
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 11.631999015808105,
      "learning_rate": 3.143895053378698e-07,
      "loss": 0.418,
      "step": 224
    },
    {
      "epoch": 0.974025974025974,
      "grad_norm": 6.837718963623047,
      "learning_rate": 2.407636663901591e-07,
      "loss": 0.334,
      "step": 225
    },
    {
      "epoch": 0.9783549783549783,
      "grad_norm": 71.07975006103516,
      "learning_rate": 1.7692529411904578e-07,
      "loss": 0.3926,
      "step": 226
    },
    {
      "epoch": 0.9826839826839827,
      "grad_norm": 9.980896949768066,
      "learning_rate": 1.228869453076986e-07,
      "loss": 0.3398,
      "step": 227
    },
    {
      "epoch": 0.987012987012987,
      "grad_norm": 34.9681282043457,
      "learning_rate": 7.865924910916977e-08,
      "loss": 0.4414,
      "step": 228
    },
    {
      "epoch": 0.9913419913419913,
      "grad_norm": 23.342512130737305,
      "learning_rate": 4.4250904955656095e-08,
      "loss": 0.4336,
      "step": 229
    },
    {
      "epoch": 0.9956709956709957,
      "grad_norm": 15.746844291687012,
      "learning_rate": 1.9668680847356735e-08,
      "loss": 0.375,
      "step": 230
    },
    {
      "epoch": 1.0,
      "grad_norm": 19.062803268432617,
      "learning_rate": 4.917412021249179e-09,
      "loss": 0.3711,
      "step": 231
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 116,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.922842866902368e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}