{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 122,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00819672131147541,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 1.3561,
      "step": 1
    },
    {
      "epoch": 0.01639344262295082,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 1.1834,
      "step": 2
    },
    {
      "epoch": 0.02459016393442623,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 1.4623,
      "step": 3
    },
    {
      "epoch": 0.03278688524590164,
      "grad_norm": 15.98621940612793,
      "learning_rate": 0.0,
      "loss": 1.2439,
      "step": 4
    },
    {
      "epoch": 0.040983606557377046,
      "grad_norm": Infinity,
      "learning_rate": 1e-05,
      "loss": 1.1846,
      "step": 5
    },
    {
      "epoch": 0.04918032786885246,
      "grad_norm": 11.108942031860352,
      "learning_rate": 1e-05,
      "loss": 1.3067,
      "step": 6
    },
    {
      "epoch": 0.05737704918032787,
      "grad_norm": 43.746212005615234,
      "learning_rate": 2e-05,
      "loss": 1.1056,
      "step": 7
    },
    {
      "epoch": 0.06557377049180328,
      "grad_norm": 23.63981819152832,
      "learning_rate": 3e-05,
      "loss": 1.0988,
      "step": 8
    },
    {
      "epoch": 0.07377049180327869,
      "grad_norm": 13.135039329528809,
      "learning_rate": 4e-05,
      "loss": 0.9535,
      "step": 9
    },
    {
      "epoch": 0.08196721311475409,
      "grad_norm": 26.240875244140625,
      "learning_rate": 5e-05,
      "loss": 0.8525,
      "step": 10
    },
    {
      "epoch": 0.09016393442622951,
      "grad_norm": 29.70737075805664,
      "learning_rate": 6e-05,
      "loss": 0.7922,
      "step": 11
    },
    {
      "epoch": 0.09836065573770492,
      "grad_norm": 30.103702545166016,
      "learning_rate": 7e-05,
      "loss": 0.835,
      "step": 12
    },
    {
      "epoch": 0.10655737704918032,
      "grad_norm": 8.32728385925293,
      "learning_rate": 8e-05,
      "loss": 0.71,
      "step": 13
    },
    {
      "epoch": 0.11475409836065574,
      "grad_norm": 42.42969512939453,
      "learning_rate": 9e-05,
      "loss": 0.6767,
      "step": 14
    },
    {
      "epoch": 0.12295081967213115,
      "grad_norm": 48.355735778808594,
      "learning_rate": 0.0001,
      "loss": 0.9952,
      "step": 15
    },
    {
      "epoch": 0.13114754098360656,
      "grad_norm": 12.929658889770508,
      "learning_rate": 9.998033131915266e-05,
      "loss": 0.6263,
      "step": 16
    },
    {
      "epoch": 0.13934426229508196,
      "grad_norm": 33.13911437988281,
      "learning_rate": 9.992134075089084e-05,
      "loss": 0.6271,
      "step": 17
    },
    {
      "epoch": 0.14754098360655737,
      "grad_norm": 18.275238037109375,
      "learning_rate": 9.982307470588098e-05,
      "loss": 0.6788,
      "step": 18
    },
    {
      "epoch": 0.1557377049180328,
      "grad_norm": 21.644994735717773,
      "learning_rate": 9.968561049466214e-05,
      "loss": 0.7019,
      "step": 19
    },
    {
      "epoch": 0.16393442622950818,
      "grad_norm": 13.200664520263672,
      "learning_rate": 9.950905626682228e-05,
      "loss": 0.4314,
      "step": 20
    },
    {
      "epoch": 0.1721311475409836,
      "grad_norm": 30.9163875579834,
      "learning_rate": 9.92935509259118e-05,
      "loss": 0.6739,
      "step": 21
    },
    {
      "epoch": 0.18032786885245902,
      "grad_norm": 13.788652420043945,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.4574,
      "step": 22
    },
    {
      "epoch": 0.1885245901639344,
      "grad_norm": 9.37610149383545,
      "learning_rate": 9.874639560909117e-05,
      "loss": 0.504,
      "step": 23
    },
    {
      "epoch": 0.19672131147540983,
      "grad_norm": 4.596352577209473,
      "learning_rate": 9.841517610611309e-05,
      "loss": 0.381,
      "step": 24
    },
    {
      "epoch": 0.20491803278688525,
      "grad_norm": 34.62461853027344,
      "learning_rate": 9.804586609725499e-05,
      "loss": 0.5659,
      "step": 25
    },
    {
      "epoch": 0.21311475409836064,
      "grad_norm": 41.479732513427734,
      "learning_rate": 9.763875613614482e-05,
      "loss": 0.8486,
      "step": 26
    },
    {
      "epoch": 0.22131147540983606,
      "grad_norm": 42.0236930847168,
      "learning_rate": 9.719416651541839e-05,
      "loss": 1.0237,
      "step": 27
    },
    {
      "epoch": 0.22950819672131148,
      "grad_norm": 13.166512489318848,
      "learning_rate": 9.671244701472999e-05,
      "loss": 0.4578,
      "step": 28
    },
    {
      "epoch": 0.23770491803278687,
      "grad_norm": 9.905980110168457,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.2581,
      "step": 29
    },
    {
      "epoch": 0.2459016393442623,
      "grad_norm": 15.204133033752441,
      "learning_rate": 9.563916325306594e-05,
      "loss": 0.3053,
      "step": 30
    },
    {
      "epoch": 0.2540983606557377,
      "grad_norm": 35.76464080810547,
      "learning_rate": 9.504844339512095e-05,
      "loss": 0.6698,
      "step": 31
    },
    {
      "epoch": 0.26229508196721313,
      "grad_norm": 38.28012466430664,
      "learning_rate": 9.442228179894362e-05,
      "loss": 0.6832,
      "step": 32
    },
    {
      "epoch": 0.27049180327868855,
      "grad_norm": 53.34618377685547,
      "learning_rate": 9.376117109543769e-05,
      "loss": 0.6321,
      "step": 33
    },
    {
      "epoch": 0.2786885245901639,
      "grad_norm": 37.23402786254883,
      "learning_rate": 9.306563141162046e-05,
      "loss": 0.7292,
      "step": 34
    },
    {
      "epoch": 0.28688524590163933,
      "grad_norm": 29.553579330444336,
      "learning_rate": 9.233620996141421e-05,
      "loss": 0.4497,
      "step": 35
    },
    {
      "epoch": 0.29508196721311475,
      "grad_norm": 9.313833236694336,
      "learning_rate": 9.157348061512727e-05,
      "loss": 0.4458,
      "step": 36
    },
    {
      "epoch": 0.30327868852459017,
      "grad_norm": 9.01462459564209,
      "learning_rate": 9.077804344796302e-05,
      "loss": 0.3081,
      "step": 37
    },
    {
      "epoch": 0.3114754098360656,
      "grad_norm": 13.8030424118042,
      "learning_rate": 8.995052426791247e-05,
      "loss": 0.3955,
      "step": 38
    },
    {
      "epoch": 0.319672131147541,
      "grad_norm": 30.403034210205078,
      "learning_rate": 8.90915741234015e-05,
      "loss": 0.7952,
      "step": 39
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 14.309679985046387,
      "learning_rate": 8.820186879108038e-05,
      "loss": 0.2402,
      "step": 40
    },
    {
      "epoch": 0.3360655737704918,
      "grad_norm": 33.29140853881836,
      "learning_rate": 8.728210824415827e-05,
      "loss": 0.6867,
      "step": 41
    },
    {
      "epoch": 0.3442622950819672,
      "grad_norm": 17.823509216308594,
      "learning_rate": 8.633301610170135e-05,
      "loss": 0.3311,
      "step": 42
    },
    {
      "epoch": 0.3524590163934426,
      "grad_norm": 10.712453842163086,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.3251,
      "step": 43
    },
    {
      "epoch": 0.36065573770491804,
      "grad_norm": 14.378042221069336,
      "learning_rate": 8.434984630174509e-05,
      "loss": 0.3781,
      "step": 44
    },
    {
      "epoch": 0.36885245901639346,
      "grad_norm": 11.261249542236328,
      "learning_rate": 8.33173288976002e-05,
      "loss": 0.2272,
      "step": 45
    },
    {
      "epoch": 0.3770491803278688,
      "grad_norm": 10.504888534545898,
      "learning_rate": 8.225859917710439e-05,
      "loss": 0.4378,
      "step": 46
    },
    {
      "epoch": 0.38524590163934425,
      "grad_norm": 12.617971420288086,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.3306,
      "step": 47
    },
    {
      "epoch": 0.39344262295081966,
      "grad_norm": 16.494400024414062,
      "learning_rate": 8.006585456492029e-05,
      "loss": 0.4195,
      "step": 48
    },
    {
      "epoch": 0.4016393442622951,
      "grad_norm": 4.538206100463867,
      "learning_rate": 7.89335648089903e-05,
      "loss": 0.2288,
      "step": 49
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 3.4017157554626465,
      "learning_rate": 7.777851165098012e-05,
      "loss": 0.1776,
      "step": 50
    },
    {
      "epoch": 0.4180327868852459,
      "grad_norm": 6.894968509674072,
      "learning_rate": 7.660160382576683e-05,
      "loss": 0.2644,
      "step": 51
    },
    {
      "epoch": 0.4262295081967213,
      "grad_norm": 7.190054893493652,
      "learning_rate": 7.540376726232648e-05,
      "loss": 0.2066,
      "step": 52
    },
    {
      "epoch": 0.4344262295081967,
      "grad_norm": 33.77849578857422,
      "learning_rate": 7.4185944355262e-05,
      "loss": 0.4325,
      "step": 53
    },
    {
      "epoch": 0.4426229508196721,
      "grad_norm": 16.263246536254883,
      "learning_rate": 7.294909322337689e-05,
      "loss": 0.2738,
      "step": 54
    },
    {
      "epoch": 0.45081967213114754,
      "grad_norm": 17.391746520996094,
      "learning_rate": 7.169418695587791e-05,
      "loss": 0.3253,
      "step": 55
    },
    {
      "epoch": 0.45901639344262296,
      "grad_norm": 20.385494232177734,
      "learning_rate": 7.042221284679982e-05,
      "loss": 0.3355,
      "step": 56
    },
    {
      "epoch": 0.4672131147540984,
      "grad_norm": 25.4639892578125,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.5549,
      "step": 57
    },
    {
      "epoch": 0.47540983606557374,
      "grad_norm": 18.658313751220703,
      "learning_rate": 6.783107663311565e-05,
      "loss": 0.2025,
      "step": 58
    },
    {
      "epoch": 0.48360655737704916,
      "grad_norm": 20.296470642089844,
      "learning_rate": 6.651395309775837e-05,
      "loss": 0.5972,
      "step": 59
    },
    {
      "epoch": 0.4918032786885246,
      "grad_norm": 21.115795135498047,
      "learning_rate": 6.518383725548074e-05,
      "loss": 0.4646,
      "step": 60
    },
    {
      "epoch": 0.5,
      "grad_norm": 52.6274299621582,
      "learning_rate": 6.384177557124247e-05,
      "loss": 0.8483,
      "step": 61
    },
    {
      "epoch": 0.5081967213114754,
      "grad_norm": 36.52357864379883,
      "learning_rate": 6.248882390836135e-05,
      "loss": 0.9086,
      "step": 62
    },
    {
      "epoch": 0.5163934426229508,
      "grad_norm": 15.678121566772461,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.5953,
      "step": 63
    },
    {
      "epoch": 0.5245901639344263,
      "grad_norm": 18.18658447265625,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 0.4511,
      "step": 64
    },
    {
      "epoch": 0.5327868852459017,
      "grad_norm": 5.188011646270752,
      "learning_rate": 5.837531116523682e-05,
      "loss": 0.3454,
      "step": 65
    },
    {
      "epoch": 0.5409836065573771,
      "grad_norm": 17.617431640625,
      "learning_rate": 5.698951697677498e-05,
      "loss": 0.6944,
      "step": 66
    },
    {
      "epoch": 0.5491803278688525,
      "grad_norm": 18.674060821533203,
      "learning_rate": 5.559822380516539e-05,
      "loss": 0.6018,
      "step": 67
    },
    {
      "epoch": 0.5573770491803278,
      "grad_norm": 9.069574356079102,
      "learning_rate": 5.420252624646238e-05,
      "loss": 0.3661,
      "step": 68
    },
    {
      "epoch": 0.5655737704918032,
      "grad_norm": 32.6479606628418,
      "learning_rate": 5.2803522361859594e-05,
      "loss": 0.4797,
      "step": 69
    },
    {
      "epoch": 0.5737704918032787,
      "grad_norm": 5.975889205932617,
      "learning_rate": 5.140231281379345e-05,
      "loss": 0.1784,
      "step": 70
    },
    {
      "epoch": 0.5819672131147541,
      "grad_norm": 5.846829414367676,
      "learning_rate": 5e-05,
      "loss": 0.2431,
      "step": 71
    },
    {
      "epoch": 0.5901639344262295,
      "grad_norm": 11.340662956237793,
      "learning_rate": 4.859768718620656e-05,
      "loss": 0.3237,
      "step": 72
    },
    {
      "epoch": 0.5983606557377049,
      "grad_norm": 9.923660278320312,
      "learning_rate": 4.7196477638140404e-05,
      "loss": 0.3903,
      "step": 73
    },
    {
      "epoch": 0.6065573770491803,
      "grad_norm": 3.1883232593536377,
      "learning_rate": 4.579747375353763e-05,
      "loss": 0.248,
      "step": 74
    },
    {
      "epoch": 0.6147540983606558,
      "grad_norm": 8.905538558959961,
      "learning_rate": 4.4401776194834613e-05,
      "loss": 0.1613,
      "step": 75
    },
    {
      "epoch": 0.6229508196721312,
      "grad_norm": 19.726444244384766,
      "learning_rate": 4.3010483023225045e-05,
      "loss": 0.2736,
      "step": 76
    },
    {
      "epoch": 0.6311475409836066,
      "grad_norm": 7.428593158721924,
      "learning_rate": 4.162468883476319e-05,
      "loss": 0.1273,
      "step": 77
    },
    {
      "epoch": 0.639344262295082,
      "grad_norm": 9.217852592468262,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 0.1887,
      "step": 78
    },
    {
      "epoch": 0.6475409836065574,
      "grad_norm": 15.629423141479492,
      "learning_rate": 3.887395330218429e-05,
      "loss": 0.3924,
      "step": 79
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 9.791216850280762,
      "learning_rate": 3.7511176091638653e-05,
      "loss": 0.3166,
      "step": 80
    },
    {
      "epoch": 0.6639344262295082,
      "grad_norm": 8.492615699768066,
      "learning_rate": 3.6158224428757535e-05,
      "loss": 0.2088,
      "step": 81
    },
    {
      "epoch": 0.6721311475409836,
      "grad_norm": 18.473737716674805,
      "learning_rate": 3.4816162744519263e-05,
      "loss": 0.4732,
      "step": 82
    },
    {
      "epoch": 0.680327868852459,
      "grad_norm": 9.715383529663086,
      "learning_rate": 3.3486046902241664e-05,
      "loss": 0.5531,
      "step": 83
    },
    {
      "epoch": 0.6885245901639344,
      "grad_norm": 13.272611618041992,
      "learning_rate": 3.216892336688435e-05,
      "loss": 0.1446,
      "step": 84
    },
    {
      "epoch": 0.6967213114754098,
      "grad_norm": 6.765244960784912,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0961,
      "step": 85
    },
    {
      "epoch": 0.7049180327868853,
      "grad_norm": 13.067438125610352,
      "learning_rate": 2.9577787153200197e-05,
      "loss": 0.1131,
      "step": 86
    },
    {
      "epoch": 0.7131147540983607,
      "grad_norm": 7.727385997772217,
      "learning_rate": 2.8305813044122097e-05,
      "loss": 0.4602,
      "step": 87
    },
    {
      "epoch": 0.7213114754098361,
      "grad_norm": 10.270739555358887,
      "learning_rate": 2.705090677662311e-05,
      "loss": 0.4261,
      "step": 88
    },
    {
      "epoch": 0.7295081967213115,
      "grad_norm": 3.8513522148132324,
      "learning_rate": 2.581405564473801e-05,
      "loss": 0.356,
      "step": 89
    },
    {
      "epoch": 0.7377049180327869,
      "grad_norm": 17.689189910888672,
      "learning_rate": 2.459623273767354e-05,
      "loss": 0.5393,
      "step": 90
    },
    {
      "epoch": 0.7459016393442623,
      "grad_norm": 14.543610572814941,
      "learning_rate": 2.3398396174233178e-05,
      "loss": 0.37,
      "step": 91
    },
    {
      "epoch": 0.7540983606557377,
      "grad_norm": Infinity,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.8326,
      "step": 92
    },
    {
      "epoch": 0.7622950819672131,
      "grad_norm": 19.77420997619629,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.3247,
      "step": 93
    },
    {
      "epoch": 0.7704918032786885,
      "grad_norm": 19.445480346679688,
      "learning_rate": 2.1066435191009715e-05,
      "loss": 0.4273,
      "step": 94
    },
    {
      "epoch": 0.7786885245901639,
      "grad_norm": 49.049041748046875,
      "learning_rate": 1.9934145435079702e-05,
      "loss": 0.4686,
      "step": 95
    },
    {
      "epoch": 0.7868852459016393,
      "grad_norm": 23.511159896850586,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.3971,
      "step": 96
    },
    {
      "epoch": 0.7950819672131147,
      "grad_norm": 10.313678741455078,
      "learning_rate": 1.774140082289563e-05,
      "loss": 0.399,
      "step": 97
    },
    {
      "epoch": 0.8032786885245902,
      "grad_norm": 21.878589630126953,
      "learning_rate": 1.6682671102399805e-05,
      "loss": 0.3466,
      "step": 98
    },
    {
      "epoch": 0.8114754098360656,
      "grad_norm": 10.881017684936523,
      "learning_rate": 1.5650153698254916e-05,
      "loss": 0.2696,
      "step": 99
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 5.622803211212158,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.719,
      "step": 100
    },
    {
      "epoch": 0.8278688524590164,
      "grad_norm": 20.312349319458008,
      "learning_rate": 1.3666983898298657e-05,
      "loss": 0.3302,
      "step": 101
    },
    {
      "epoch": 0.8360655737704918,
      "grad_norm": 1.5690298080444336,
      "learning_rate": 1.2717891755841722e-05,
      "loss": 0.1621,
      "step": 102
    },
    {
      "epoch": 0.8442622950819673,
      "grad_norm": 14.779884338378906,
      "learning_rate": 1.1798131208919627e-05,
      "loss": 0.4733,
      "step": 103
    },
    {
      "epoch": 0.8524590163934426,
      "grad_norm": 18.317291259765625,
      "learning_rate": 1.090842587659851e-05,
      "loss": 0.327,
      "step": 104
    },
    {
      "epoch": 0.860655737704918,
      "grad_norm": 5.87141227722168,
      "learning_rate": 1.004947573208756e-05,
      "loss": 0.4066,
      "step": 105
    },
    {
      "epoch": 0.8688524590163934,
      "grad_norm": 8.89691162109375,
      "learning_rate": 9.221956552036992e-06,
      "loss": 0.1725,
      "step": 106
    },
    {
      "epoch": 0.8770491803278688,
      "grad_norm": 22.22539520263672,
      "learning_rate": 8.426519384872733e-06,
      "loss": 0.4812,
      "step": 107
    },
    {
      "epoch": 0.8852459016393442,
      "grad_norm": 16.512264251708984,
      "learning_rate": 7.663790038585793e-06,
      "loss": 0.2801,
      "step": 108
    },
    {
      "epoch": 0.8934426229508197,
      "grad_norm": 11.542386054992676,
      "learning_rate": 6.934368588379553e-06,
      "loss": 0.3944,
      "step": 109
    },
    {
      "epoch": 0.9016393442622951,
      "grad_norm": 2.5426712036132812,
      "learning_rate": 6.238828904562316e-06,
      "loss": 0.2072,
      "step": 110
    },
    {
      "epoch": 0.9098360655737705,
      "grad_norm": 6.43900203704834,
      "learning_rate": 5.577718201056392e-06,
      "loss": 0.2337,
      "step": 111
    },
    {
      "epoch": 0.9180327868852459,
      "grad_norm": 2.6949198246002197,
      "learning_rate": 4.951556604879048e-06,
      "loss": 0.2164,
      "step": 112
    },
    {
      "epoch": 0.9262295081967213,
      "grad_norm": 12.75400161743164,
      "learning_rate": 4.360836746934055e-06,
      "loss": 0.2642,
      "step": 113
    },
    {
      "epoch": 0.9344262295081968,
      "grad_norm": 7.314509868621826,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.2681,
      "step": 114
    },
    {
      "epoch": 0.9426229508196722,
      "grad_norm": 7.261117458343506,
      "learning_rate": 3.2875529852700147e-06,
      "loss": 0.1757,
      "step": 115
    },
    {
      "epoch": 0.9508196721311475,
      "grad_norm": 10.839629173278809,
      "learning_rate": 2.8058334845816213e-06,
      "loss": 0.1706,
      "step": 116
    },
    {
      "epoch": 0.9590163934426229,
      "grad_norm": 6.028248310089111,
      "learning_rate": 2.361243863855184e-06,
      "loss": 0.351,
      "step": 117
    },
    {
      "epoch": 0.9672131147540983,
      "grad_norm": 5.811429023742676,
      "learning_rate": 1.9541339027450256e-06,
      "loss": 0.2582,
      "step": 118
    },
    {
      "epoch": 0.9754098360655737,
      "grad_norm": 11.192886352539062,
      "learning_rate": 1.584823893886933e-06,
      "loss": 0.392,
      "step": 119
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 11.7086820602417,
      "learning_rate": 1.2536043909088191e-06,
      "loss": 0.1863,
      "step": 120
    },
    {
      "epoch": 0.9918032786885246,
      "grad_norm": 25.053173065185547,
      "learning_rate": 9.607359798384785e-07,
      "loss": 0.4225,
      "step": 121
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.306795120239258,
      "learning_rate": 7.064490740882057e-07,
      "loss": 0.1589,
      "step": 122
    }
  ],
  "logging_steps": 1,
  "max_steps": 122,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 6324556423495680.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}